Nuclear Medicine - Radiological Society of North …archive.rsna.org/2018/NuclearMedicine.pdfNM104-ED-X Spectrum of Benign Bone and Soft Tissue Findings on Sodium-Fluoride (NaF) PET

NuclearMedicine

NM001-EB-X

The Sum of Two Powers: High Resolution Breast PET with Mammography Fusion

All Day Room: NA Hardcopy Backboard

ParticipantsCristian Perez, MD, Mexico City, Mexico (Presenter) Nothing to DiscloseGisela Estrada, MD,PhD, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseMaria Lara, MD, Mexico, Mexico (Abstract Co-Author) Nothing to DiscloseMary C. Herrera-Zarza, MD, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseJorge Martin Schalch Ponce De Leon, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseJose Criales Cortes, Mexico City, Mexico (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

To learn the variable rates of false positives and negatives and normal findings of the digital breast mammography that continue tobe the standard screening in breast cancer. To show that the breast cancer cells have higher metabolism than normal cells andexhibit increased uptake of the 18F-FDG radiotracer. Illustrate that the fusion of high resolution breast PET and mammography isbetter because it provides correlation with morphological and metabolic information to recognize breast cancer variants and non-malignant conditions and asses of response to chemotherapy. To recognize that high resolution breast PET has higher spatialresolution than whole body PET - CT, it can detect lesions measuring less than two millimeters. To remark that high resolutionBreast PET is not affected by either breast density, presence of implants or a woman's hormonal status. To know that highresolution breast PET has the same sensitivity than breast MRI, but greater specificity and has fewer false-positive results.

TABLE OF CONTENTS/OUTLINE

Introduction. Breast cancer: • Role of digital mammography. • Role of High Resolution Breast PET. • Usefulness of the fusion: HighResolution Breast PET with mammography. Clinical cases. Future implications.

NM100-ED-X

A Pictorial Review of Arthritis on FDG-PET/CT: Key Features and Uptake Patterns for Differential Diagnosis

All Day Room: NA Digital Education Exhibit

AwardsCertificate of MeritIdentified for RadioGraphics

ParticipantsMasatoshi Hotta, Shinjuku, Japan (Presenter) Nothing to DiscloseRyogo Minamimoto, MD, PhD, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseTuyoshi Tajima, MD, PhD, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseTomoyuki Noguchi, MD, PhD, Shinjuku-Ku, Japan (Abstract Co-Author) Nothing to DiscloseYoshitaka Shida, MD, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseTakashi Okafuji, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose

For information about this presentation, contact:

[email protected]

TEACHING POINTS

1) To review the role of FDG-PET/CT in the diagnosis and management of arthritis 2) To demonstrate FDG-PET/CT images ofvarious joint disorders including rheumatoid arthritis and related diseases, spondyloarthritis, and miscellaneous joint disorders 3) Todescribe the key features of FDG-PET/CT for differentiating between rheumatoid arthritis and other joint disorders


This presentation provides FDG-PET/CT images of: 1. Rheumatoid arthritis and related diseases (adult-onset Still's disease (AOSD),polymyalgia rheumatica (PMR), remitting seronegative symmetrical synovitis with pitting edema (RS3PE)) 2. Spondyloarthritis(ankylosing spondylitis (AS), psoriatic arthritis (PsA), reactive arthritis (ReA), inflammatory bowel disease arthritis, SAPHOsyndrome) 3. Miscellaneous joint disorders (sarcoidosis, amyloid deposition, hemophilia, multicentric reticulohistiocytosis (MRH),chronic recurrent multifocal osteomyelitis (CRMO))

NM101-ED-X

Incidental Hematopoietic Retroperitoneal Mass: Hands off! Proposal of a Non-Invasive Diagnostic Approachand Review of the Literature


ParticipantsMaria Angeles M. Martin, MD, Palma de Mallorca, Spain (Presenter) Nothing to DiscloseElena Usamentiaga, MD, Palma De Mallorca, Spain (Abstract Co-Author) Nothing to DiscloseAlberto Ramiro Biesa Campos, Inca, Spain (Abstract Co-Author) Nothing to DiscloseMarco Antonio Martinez Ortega, Palma de Mallorca, Spain (Abstract Co-Author) Nothing to DiscloseJaime Daumal Domenech, Palma de Mallorca, Spain (Abstract Co-Author) Nothing to DiscloseBeatriz Miriam Rodriguez Chikri, MD, PALMA DE MALLORCA, Spain (Abstract Co-Author) Nothing to DisclosePaula Roig Egea, Palma de Mallorca, Spain (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1.To depict on CT and MR the imaging features of hematopoietic masses (extaadrenal myelolipoma and extramedullaryhematopoiesis), focusing on the retroperitoneal and presacral spaces. 2.To underline the importance of the patients´clinicalbackground in order to differentiate between these hematopoietic masses. 3.To expose the usefulness of 99m-Tc-nanocolloidSPECT highlighting their hematopoietic composition. 4.To propose an algorithm based on non-invasive imaging techniques to identifyand differentiate them from malignancy .


1.Introduction 1.a.Etiology 2.Pathophysiology 3.Clinical findings 4.Imaging findings (CT, MRI, SPECT-BM) 4.a.Anatomy of thepresacral space 5.Differential diagnosis 6.Follow up 6.a.Spectrum of possible behaviors 7.Interventional management (goldstandard) 8.Algorithm proposal: SPECT-BM 9.Sample cases

NM102-ED-X

"Wild & Crazy" Musculoskeletal Scintigraphy - The Bones and Beyond: A Quiz-based Approach


AwardsCertificate of Merit

ParticipantsAmanda M. Romesberg, DO, Lexington, KY (Presenter) Nothing to DiscloseAhmed A. Faraq, MD, Lexington, KY (Abstract Co-Author) Nothing to DiscloseRiham H. El Khouli, MD, PhD, Nicholasville, KY (Abstract Co-Author) Nothing to DiscloseM. Elizabeth Oates, MD, Lexington, KY (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1. To understand the expected and unexpected biodistribution of Tc-99m MDP and other radiopharmaceuticals with respect to themusculoskeletal system. 2. To describe typical patterns and characterize diagnostic features of various conditions involving themusculoskeletal system on dynamic and static planar scintigraphy, SPECT and SPECT/CT. 3. To recognize unexpected findings thatmay be clinically relevant.


I. Overview: Mechanism of localization of Tc-99m MDP and other radiopharmaceuticals in the skeleton and beyond. II. Review:Typical musculoskeletal scintigraphy protocols. III. For each case: Clinical history, key image(s), quiz questions & answers,diagnosis, discussion, and diagnostic pearls. IV. 12 representative cases: Tc-99m MDP: 1) arterial injection, 2) stomach activity inhyperparathyroidism, 3) colovesical fistula, 4) progressive unilateral renal atrophy, 5) hydrocele, 6) osteoid osteoma of ilium, 7)polyostotic fibrous dysplasia, 8) hypertrophic osteoarthropathy; Tc-99m MDP plus In-111 WBCs: 9) osteomyelitis at MTP joint, 10)Charcot foot; Ga-67 citrate: 11) vertebral osteomyelitis with psoas abscess; Tc-99m RBCs: 12) calvarial hemangioma. V. Take-home highlights. VI. Selected references.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Riham H. El Khouli, MD, PhD - 2012 Honored Educator

NM103-ED-X

Imaging of Merkel Cell Carcinoma: What Imaging Experts Should Know


FDA Discussions may include off-label uses.


ParticipantsGensuke Akaike, MD, Seattle, WA (Presenter) Nothing to DiscloseTomoko Akaike, MD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseKristina Lachance, MS, Seattle, WA (Abstract Co-Author) Nothing to DisclosePaul Nghiem, MD, PhD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseSanaz Behnia, MD, Seattle, WA (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

The purpose of this exhibit is to understand and review1. Pathophysiology and clinical behavior of Merkel cell carcinoma (MCC).2.Its aggressive features and the important role of imaging in the management. 3. The Merkel cell polyomavirus (MCPyV) as a causeof MCC and utility of a blood test that detects antibodies to the MCPyV.4. Different clinical approaches and use of imaging inantibody producers and non-antibody producers. 5. Use of F18-FDG PET/CT in the initial staging and follow-up.6. Neuroendocrinefeatures of MCC with somatostatin receptor (SSTR) expression and SSTR seeking nuclear medicine studies. 7. Various imagingmanifestations of MCC on different modalities.


-Pathophysiology and prognosis.-Current diagnosis and treatment.-MCPyV and antibody test.-SSTR expression and its clinicalrelevance.-Review of various imaging manifestations and role of each modality. *Computed Tomography (CT) and MagneticResonance Imaging (MRI). *Sentinel node lymphoscintigraphy. *Bone scintigraphy *F18-FDG PET/CT *SSTR seeking imaging (e.g.In111 pentetreotide scintigraphy, Ga68-Dotatate-PET/CT) *Peptide Radionuclide Receptor Therapy (e.g. Lu177 Dotatate): Currentevidence review.

NM104-ED-X

Spectrum of Benign Bone and Soft Tissue Findings on Sodium-Fluoride (NaF) PET


ParticipantsNada Almenieir, MBBS, MS, Montreal, QC (Presenter) Nothing to DiscloseMarc Hickeson, MD,FRCPC, Montreal, QC (Abstract Co-Author) Nothing to DiscloseJavier A. Novales-Diaz, MD, Montreal, QC (Abstract Co-Author) Nothing to DiscloseVilma A. Derbekyan, FRCPC, MD, Montreal, QC (Abstract Co-Author) Nothing to DiscloseRobert Lisbona, MD, Montreal, QC (Abstract Co-Author) Nothing to DiscloseGad Abikhzer, MD, Haifa, Israel (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Sodium-Fluoride PET is a sensitive method to detect altered bone mineralization. Its increasing use in routine clinical practice hasidentified many benign bone lesions and soft tissue findings. A spectrum of NaF PET scan cases with benign bone lesions and softtissue findings are presented, including whole body PET bone scan and selected PET/CT, CT or MR images of the region of interest.Key differential diagnostic points will be highlighted in the discussion of each case. By reviewing these cases, the viewer will be ableto: 1- Improve the knowledge related to the clinical presentation of the aformentioned entities, some of them rare. 2- Recognizethem based on PET bone scans and CT/MR imaging patterns.


Benign bone lesions: Paget's disease Enchondroma Hemangioma Fibrous dysplasia Chondroblastoma Giant Cell Tumor Langerhans' cellhistiocytosis Desmoplastic fibroma Intra-osseous lipoma Osteochondroma Osteoid osteoma Fractures: Ilial fracture Compressionfracture Infection/Inflammation: Spondylodiscitis and epidural abscess Frostbite Soft tissue uptake: Heterotopic ossification Softtissue bowel Soft tissue lung

NM105-ED-X

Emerging Clinical Use of Technetium-99m Pyrophosphate Scintigraphy for Diagnosing Transthyretin-RelatedCardiac Amyloidosis: What the Radiologist Needs to Know


ParticipantsJasleen Saini, DO, Jersey City, NJ (Presenter) Nothing to DiscloseOsmani Deochand, MD, Morristown, NJ (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

The purpose of this exhibit is to: 1. Review the clinical presentation and pathophysiology of transthyretin (ATTR) andimmunoglobulin light chain (AL) cardiac amyloidosis subtypes. 2. Discuss the two different approaches to classifying myocardialtechnetium-99m pyrophosphate uptake for diagnosing ATTR cardiac amyloidosis. 3. Learn the appropriate criteria for classifying thepossibility of positive ATTR cardiac amyloidosis.


1. Discussion of cardiac amyloidosis subtypes (ATTR and AL) as well as the pathophysiology and clinical presentation of eachsubtype 2. Introduction of utilizing Tc-99m pyrophosphate study for evaluation of ATTR cardiac amyloidosis 3. Classification of thevisual grading of myocardial Tc-99m pyrophosphate uptake 4. Classification of the quantitation of cardiac Tc-99m pyrophosphateuptake using the heart to contralateral lung (H/CL) ratio 5. Imaging examples of each type of classification: -Grade 0, H/CL ratio<1.5 -Grade 1, H/CL ratio <1.5 -Grade 2, H/CL ratio <1.5 -Grade 3, H/CL ratio >1.5 6. Discussion of how to categorize overallfindings into: -Not suggestive of ATTR cardiac amyloidosis -Strongly suggestive of ATTR cardiac amyloidosis -Equivocal for ATTRcardiac amyloidosis 7. Discussion of the pitfalls of the study 8. Discussion of the usefulness and limitations of the study

NM106-ED-X

Congenital Heart: Role of Split Dose Ventilation-Perfusion Scan in the Biodirectional Glenn and FontanCompletion Patient


ParticipantsAmy B. Farkas, MD, Jackson, MS (Presenter) Nothing to DiscloseVani Vijayakumar, MD, Ridgeland, MS (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

At the end viewers will learn: a. The anatomy of the Glenn and Fontan procedures b. The altered hemodynamics and redistributionof pulmonary perfusion resulting from these procedures c. The role of ventilation-perfusion (VQ) scans in the evaluation of thepost-operative patient d. Rationale of split dose VQ studies in this setting to compensate for and evaluate altered physiology


The Glenn procedure and Fontan completion are surgically created cardiac shunts used in the management of tricuspid atresia andhypoplastic left heart syndrome. The Glenn procedure routes systemic venous return from the SVC to the right pulmonary artery,while the Fontan completion routes systemic venous return from the IVC to the right pulmonary artery. Hypoperfusion of the lungsoccurs as the entirety of systemic venous return circulation bypasses the right heart. Nuclear medicine studies are used toevaluate for pulmonary stenosis and have a distinct advantage over angiography. VQ scans provide valuable prognostic informationat a significant dose reduction, which of particular importance in a pediatric patient population that requires serial imaging. Theimaging protocol and dosing of the VQ scans in this patient population will be reviewed, including rationale for splitting theradiopharmaceutical dose to avoid preferential flow to either lung.

NM107-ED-X

A Comprehensive Review of the Valuable Role of Nuclear Myocardial Perfusion Imaging in Children with HeartConditions


ParticipantsDivya Shakti, MBBS,MPH, Jackson, MS (Presenter) Nothing to DiscloseVani Vijayakumar, MD, Ridgeland, MS (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

MPI in children: Role. Indications. Safety considerations with stress MPI. Special considerations, with planning and interpretation.


Nuclear myocardial perfusion imaging (MPI) is a useful tool in the evaluation and surveillance of a variety of heart conditions inchildren. The indications of MPI, and stress MPI in children include acquired and congenital cardiac conditions as well ascardiomyopathies and post-transplant evaluations. When considering stress MPI, pediatric considerations include safety; selectionof the type of stress (exercise vs. phramacologic); selection of appropriate and safe pharmacologic agent; age and disease relatedfactors. There are differences in normal MPI study in children (compared to adults) by age, body habitus and gender that should bekept in mind to prevent erroneous interpretation. Knowledge of the strengths and limitations of MPI and other modalities is helpfulto select the best, and safest test to answer the given clinical question. Table 1. Role of Nuclear myocardial perfusion imaging(MPI) in children Table 2. Indications of MPI in children Figure 1. Thallium NM viability study example in congenital heart defectTable 3. Safety considerations in children with Stress MPI Table 4. Special considerations in children by age and disease

NM108-ED-X

Multimodality Imaging of Dementia Patients: Clinical Importance and Role of Integrated Anatomic andMolecular Imaging


AwardsIdentified for RadioGraphics

ParticipantsKunal P. Patel, MD, Miami Beach, FL (Presenter) Nothing to DiscloseDavid T. Wymer, MD, Miami, FL (Abstract Co-Author) Nothing to DiscloseRanjan Duara, Miami Beach, FL (Abstract Co-Author) Nothing to DiscloseVinay K. Bhatia, MD, Miami, FL (Abstract Co-Author) Nothing to DiscloseChetan D. Rajadhyaksha, MD, Miami Beach, FL (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

-Discuss the pertinent anatomy, clinical presentation, and differing pathophysiology of neurodegenerative disorders, includingAlzheimer dementia, Lewy Body disease and Parkinson dementia, fronto-temporal dementia, and multi-infarct dementia -Discuss thedifferent nuclear modalities for diagnosing neurodegenerative disorders and their molecular pathways including 123I-ioflupane, 18F-FDG, and 18F amyloid imaging (florbetapir, florbetaben) -Discuss the salient imaging findings of each neurodegenerative disorderwith correlation with multimodality imaging, including changes on traditional MRI and their relationship with the associated nuclearstudies, including both qualitative and quantitative nuclear assessment


-Review of the anatomy and physiology of the limbic system and nigrostriatal pathway -Review of each major neurodegenerativedisorder with clinical presentation and pathophysiology, with emphasis on how this relates to use of radiopharmaceuticals formolecular imaging -Review of the details of each of the radiopharmaceuticals used in molecular imaging ofdementias/neurodegenerative diseases including the technique related to performing the studies -Review of each majorneurodegenerative disorder as it appears on molecular imaging, with comparison to multi-modality qualitative and quantitativeassessment

NM109-ED-X

Needle in the Hay: The Role of F-CHOLINE PET CT Scan in the Detection of Ectopic Parathyroid Tissue


ParticipantsJulio R. Coronil, MD, Buenos Aires, Argentina (Presenter) Nothing to DiscloseJulieta Iorio, MD, Buenos Aires, Argentina (Abstract Co-Author) Nothing to DiscloseMaria A. Brisco, Buenos Aires City, Argentina (Abstract Co-Author) Nothing to DiscloseLeandro Fassola, Buenos Aires, Argentina (Abstract Co-Author) Nothing to DiscloseConstanza Montenegro, CABA, Argentina (Abstract Co-Author) Nothing to DiscloseMaria F. Bambaci, MD, Buenos Aires, Argentina (Abstract Co-Author) Nothing to DiscloseDavid C. Polillo, MD, Buenos Aires, Argentina (Abstract Co-Author) Nothing to DiscloseLucrecia Cuneo, MMedSc, Buenos Aires, Argentina (Abstract Co-Author) Nothing to DiscloseMartin A. Eleta, MD, Capital Federal, Argentina (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

To enumerate the main indication of F CHOLINE PET-CT scan in primary hyperparathyroidismTo identify key imaging features andfrequent pitfallsTo describe F CHOLINE PET-CT protocol


F CHOLINE PET CT protocol and principal findingsSample cases

NM110-ED-X

PET/MRI Staging and Restaging Pediatric Lymphoma and Sarcoma in Less Than 30 Minutes-Initial PET/MRIExperience in a Stand-Alone Children Hospital



ParticipantsJing Qi, MD, Milwaukee, WI (Presenter) Nothing to DiscloseJoe Ackerman, Milwaukee, WI (Abstract Co-Author) Nothing to DisclosePooja D. Thakrar, MD, Portland, OR (Abstract Co-Author) Nothing to DiscloseDavid C. Gregg, MD, Whitefish Bay, WI (Abstract Co-Author) Nothing to DiscloseKevin P. Boyd, DO, Whitefish Bay, WI (Abstract Co-Author) Nothing to DiscloseMatthew Plunk, Milwaukee, WI (Abstract Co-Author) Nothing to DiscloseSachin S. Kumbhar, MBBS, MD, Milwaukee, WI (Abstract Co-Author) Nothing to DiscloseDavid C. Moe, MD, Pewaukee, WI (Abstract Co-Author) Nothing to DiscloseNghia Vo, MD, Seattle, WA (Abstract Co-Author) Nothing to Disclose


-

TEACHING POINTS

Perform PET/MRI within 30 minutes is technically feasible and clinically satisfactory to stage/restage cancer. Short PET/MRIimproves patient compliance and decrease sedation time in pediatric patients.


Previous studies have shown equivalent performance of PET/MRI to PET/CT with significant reduction in ionizing radiation. Thedisadvantage of PET/MRI is long scan time. Since PET/MRI acquires PET and MRI data simultaneously, we tailored MRI sequences tofit PET acquisition time, making scan time equivalent to PET/CT. 23 PET/MRIs were performed during our initial 2 months service,including 15 lymphoma/sarcoma and a few other indications. 11 patients (48%) need sedation. FDG dose 0.1 mCi/kg. PET isacquired 3 min/bed position, along with 3 MRI sequences: MRAC for attenuation correction, Ax 3D LAVA Flex for anatomicregistration, and Ax T2 FSE flex for pathology survey. Total scan time is 18-30 minutes (Fig 1). Our PET/MRI protocols areadequate for staging/restaging lymphoma/sarcoma. Lesions are well characterized even with tailored sequences (Fig 2-4). Only 1out of 10 sarcoma patients required diagnostic MRI at the time of PET/MRI. Dotatate PET/MRI is sensitive and specific to evaluatepheochromacytoma (Fig 5). Our continuing work is to compare the efficacy of 30-minute PET/MRI to extensive PET/MRI in pediatriccancer management.

NM111-ED-X

Nuclear Medicine Myocardial Perfusion Studies: A Glance Beyond Coronary Abnormalities


AwardsCum Laude

ParticipantsHardik U. Shah, MBBS, MD, Seattle, WA (Presenter) Nothing to DiscloseBhagya Sannananja, MD, San Antonio, TX (Abstract Co-Author) Nothing to DiscloseManuela C. Matesan, MD, PhD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseClaudia Zacharias, MD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseLaurie A. Soine, PhD, RN, Seatle, WA (Abstract Co-Author) Nothing to DiscloseJitesh Ahuja, MD, MBBS, Seattle, WA (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

At the conclusion of this activity, participants will be able to: 1. Perform basic interpretation of nuclear medicine myocardial rest-stress perfusion studies 2. Pay attention to extra-cardiac findings on nuclear medicine myocardial perfusion studies 3. Practicecaution in the interpretation of cardiac findings secondary to pathologies other than coronary abnormalities.


1. Introduction 2. Basic principles in the interpretation of myocardial perfusion studies. 3. Imaging findings outside the box ofcoronary abnormalities. a. Extra-cardiac findings: Although the purpose of a myocardial perfusion study is detection of coronarycirculation abnormalities, few findings point towads extra-cardiac abnormalities requiring further follow-up or management.Examples: Malignancies of the breast, chest wall and intrathoracic structures; hypertrophic cardiomyopathy and Takotsubocardiomyopathy. b. Artifacts: are one of the major issues and present challenges to interpretation. These can be caused by amultitude of factors; for instance medical devices, metallic objects or variable breast position. c. Iatrogenic: Medical interventionsor procedures occasionally give rise to imaging findings that mimic perfusion abnormalities. Examples include: Konno procedure andalcohol ablation of interventricular septum. 4. Conclusion.

NM112-ED-X

Evaluation of the Shielding Ability of Radiological Shielding Materials in Nuclear Medicine Including AnnihilationRadiation


ParticipantsTakeo Honda, RT, Fukuoka, Japan (Presenter) Nothing to DiscloseYuichi Yokoyama, RT, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseKeisuke Hamada, RT, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseNobuyuki Tabata, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseShinichi Orita, RT, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Molecular imaging using Radioisotopes is expected to increase in the future. Radioprotection during the exam is necessary toprotect the staff. Based on the effective dose transmittance factor using ICRP 107 date of radiological shielding materials,1cm doseequivalent was measured,and radioprotection of the staff was evaluated.


In nuclear medicine,radiation exposure including annihilation radiation must be managed using appropriate radiological shieldingmaterials. The effective dose transmittance factor of annihilation radiation and gamma rays of 99mTc is presented using the date ofICRP Publication 107,but not through data obtained in the clinical setting. In this paper,transmittance at the clinical site wascalculated and evaluated using a 1cm dose equivalent. The effective dose transmittance factor using ICRP 107 data of radiologicalshielding material was different from the transmittance at the clinical site as determined by measuring an equivalent 1cm dose. Inorder to measure and safely manage radiation,it is necessary to acquire and maintain transmittance data at the clinical site.

NM113-ED-X

Many Hands Make Light Work-99mTc-MDP Bone SPECT/CT Scintigraphy of the Hands and Wrists: Pearls andPitfalls



ParticipantsRamin Mandegaran, MBBS, London, United Kingdom (Presenter) Nothing to DiscloseHajira Ilyas, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseAmidevi Desai, MRCP, FRCR, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSanjay Vijayanathan, MBBS, Harrow, United Kingdom (Abstract Co-Author) Nothing to DiscloseCharlotte Fowler, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseAmy E. Eccles, MBBCHIR,FRCR, London, United Kingdom (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

1.Clinical assessment is limited in ability to identify specific pain generators in the hand/wrist 2.Conventional imaging modalities(including CT/MRI) identify morphological changes, which may not represent active pain generators 3.Imaging is further complicatedpost-operatively with potential added limitation of metal artefact 4.SPECT/CT can identify active pain generators & add incrementalvalue over conventional imaging modalities - particularly in the context of widespread morphological changes & complex post-operative cases 5.SPECT/CT can play a unique role & should be considered when conventional modalities fail to elucidate the causeof pain


1.Discussion of aetiologies of hand/wrist pain by anatomical location 2.Summarise current literature on role/limitations of radiologicalimaging in hand/wrist pain & potential indications of SPECT/CT 3.Image acquisition - technical considerations 4.Illustratedpresentation of SPECT/CT appearances in patients with hand/wrist pain with/without previous surgery: a.Normal & abnormal scans:degenerative disease, inflammatory arthropathy, scaphoid/other carpal fractures, scaphoid non-union, scapholunate instability,radiocarpal/intercarpal/CMC fusions, fusion non-union, carpal bone resections, abutments, carpal AVN, infection etc b.Normalvariants c.Artifacts & false positives

NM114-ED-X

Treatment Related Changes on PET/CT: Imaging Patterns


ParticipantsNemi Gandy, MBChB,FRCR, London, United Kingdom (Presenter) Nothing to DiscloseMubarik A. Arshad, MBBS, BSC, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSamuel Nguku, Nairobi, Kenya (Abstract Co-Author) Nothing to DiscloseKathryn L. Wallitt, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSameer Khan, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseTara D. Barwick, MBChB, London, United Kingdom (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Purpose of the exhibit: to be aware of treatment related effects/changes encountered on PET-CT to understand mechanisms tominimise these pitfalls


Introduction of PET/CT Role in post treatment imaging PET/CT imaging protocols with specific emphasis on reducing treatmentrelated pitfalls Examples of changes related to: Surgery/intervention Targeted therapies with RFA/cyberknife ChemotherapyRadiotherapy Immunotherapy

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Tara D. Barwick, MBChB - 2017 Honored Educator

NM115-ED-X

CT and PETCT Detection of Tumor Thrombus in Various Atypical Malignancies and Their Incidence


ParticipantsArvind Reddy, MBBS, Hyderabad , India (Presenter) Nothing to DisclosePathapati Deepthi, MD, Hyderabad, India (Abstract Co-Author) Nothing to DiscloseGati Rajitha JR, MBBS, Hyderabad, India (Abstract Co-Author) Nothing to DiscloseMeghana Kandati, MD, Hyderabad, India (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Tumor thrombus is rare complication of solid cancers, mainly in cases of renal cell carcinoma, wilms tumor, testicular carcinoma,adrenocortical carcionoma and hepatocellular carcinomas. Tumor thrombus in atypical malignancies is very rare and is reported infew cases. Involvement of major vessels is necessary to rule out venous thromboembolism and unnecessary anti coagulant therapyEarly identification of tumor thrombus with appropriate treatment can increase patient survival and quality of life.


F18-FDG PET/CT has a definite role in identifying tumor thrombus and to differentiate between tumor thrombus and venousthromboembolism, which is also a frequent complication of malignancies. PET combined with contrast enhanced CT is likely to bethe most accurate diagnostic modality as neovascularisation of tumor thrombus with intraluminal filling defect can be identified inmost of the cases in the arterial phase. It is classically called as the "thread and streak" sign. Detection of tumor thrombus inpatients with solid tumors and differentiating it from bland thrombus is important to guide correct therapy and likely to improvepatient prognosis. F18-FDG PET combined with contrast-enhanced computed tomography (CECT) is likely to be a superior modalityin detection and prognosis.

NM116-ED-X

Clinical PET Imaging in Pediatric Brain Tumors



ParticipantsYuka Yamamoto, MD, PhD, Kagawa, Japan (Presenter) Nothing to DiscloseTakashi Norikane, Kita-gun, Japan (Abstract Co-Author) Nothing to DiscloseYoshihiro Nishiyama, MD, Kagawa, Japan (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

To understand the role of clinically available PET tracers in pediatric brain tumor for detecting and monitoring therapeuticassessment.


1. Review the clinically available PET tracers in pediatric brain tumor 2. Review the role of each PET tracer for detecting pediatricbrain tumor 3. Review the role of each PET tracer for monitoring therapeutic assessment - 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG) - L-[methyl-11C]-methionine (11C-MET) - 3'-deoxy-3'-18F-fluorothymidine (18F-FLT)

NM117-ED-X

FDG PET/CT Imaging in Systemic Autoimmune Disease


ParticipantsYoshihiro Nishiyama, MD, Kagawa, Japan (Presenter) Nothing to DiscloseTakashi Norikane, Kita-gun, Japan (Abstract Co-Author) Nothing to DiscloseYuka Yamamoto, MD, PhD, Kagawa, Japan (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

FDG accumulates not only in malignant tissues but also at sites of infection and inflammation and in autoimmune diseases. Thisexhibit describes the impact of FDG PET/CT in the diagnostic work-up of patients with systemic autoimmune diseases. 1. Reviewthe usefulness of FDG PET/CT for detecting systemic involvement in autoimmune disease 2. Review the follow-up FDG PET/CT formonitoring therapeutic assessment


1. FDG PET/CT imaging in patients with systemic autoimmune disease - Rheumatoid arthritis - Polymyositis/ Dermatomyositis -Systemic sclerosis - Sjögren syndrome - IgG4-related disease - Vasculitis syndrome - Others 2. Follow-up FDG PET/CT imagingafter therapy

NM118-ED-X

The Peak of the Uptake: 68Ga-PSMA PET-CT in Prostate Cancer, Correlation Between Levels of PSA andPattern of Dissemination Through a Semi-Quantitative Scale


ParticipantsCristian Perez, MD, Mexico City, Mexico (Presenter) Nothing to DiscloseMary C. Herrera-Zarza, MD, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseJorge Martin Schalch Ponce De Leon, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseGisela Estrada, MD,PhD, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseJose Criales Cortes, Mexico City, Mexico (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

68Ga-PSMA PET-CT is effective for imaging disease in the prostate, lymph nodes, soft tissue, and bone. There is emerging evidencefor its clinical value in staging of high-risk primary prostate cancer and localization of disease in biochemical recurrence. Isredefining patterns of disease spread compared with those seen at conventional imaging, with identification of subcentimeterprostate cancer lesions. Knowledge of its physiologic distribution and other causes of uptake is essential to minimize false-positiveimaging findings. Illustrate the patterns of uptake and relation between SUVlmax and PSA serum value.


Introduction • Overview of prostate cancer and its imaging modalities • PSA as a tracer of etapification in prostate cancer • Role of68GA- PSMA PET-CT in prostate cancer; indications Illustrative cases and show relation between PSA and SUVlmax in 68GA- PSMAPET-CT

NM119-ED-X

Current Status and Future Perspective of Molecular Imaging Studies in Prostate Cancer



ParticipantsAkira Toriihara, Stanford, CA (Presenter) Nothing to DiscloseTomomi Nobashi, MD,PhD, Palo Alto, CA (Abstract Co-Author) Nothing to DiscloseLucia Baratto, Stanford, CA (Abstract Co-Author) Nothing to DiscloseSonya Y. Park, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseNegin Hatami, Stanford, CA (Abstract Co-Author) Nothing to DiscloseHeying Duan, Stanford, CA (Abstract Co-Author) Nothing to DiscloseAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric Company

TEACHING POINTS

The purposes of this exhibit are as follows: 1. To learn mechanism of representative molecular imaging studies used to evaluatepatients with prostate cancer. 2. To learn advantage or disadvantage of each radiotracer through the actual patients' images andliterature review. 3. To discuss future perspective of approach to patients with prostate cancer in terms of nuclear medicine.


Overview of prostate cancer Morphological imaging study: CT and MRI Molecular imaging study - FDG-PET/CT - Bone scan and NaF-PET/CT (evaluation of bone metastases) - Prostate cancer-specific radiotracers: 68Ga-PSMA and 68Ga-RM2 - Integrated PET/MRIFuture perspective (including 'theranostics')

NM120-ED-X

Differentiated Thyroid Carcinoma Radioiodine Refractory Patients: How Can FDG-PET/CT Help in theTheranostic Era?


ParticipantsFlavia P. Lopes, MD, Rio de Janeiro, Brazil (Presenter) Nothing to DiscloseRenata Moreira, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseJose Leite, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseMayra L. Moreira, MD,MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseRoberta Hespanhol, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseSumara A. Gouveia, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1) 5 - 15% of patients with differentiated thyroid cancers will have local recurrence or distal metastasis. From those 2/3 willbecome radioactive iodine-refractory 2) FDG-PET/CT has high sensitivity in the evaluation of aggressive disease in patients'radioiodine refractory. 3) FDG uptake on metastatic disease is associated with worse radioiodine treatment treatment, as wellworse predictor of overall survival 4) Multimodal nuclear imaging is essential to design the lesion-based multimodal treatmentstrategy for patients with multiple heterogeneous metastatic lesions


Purpose: To review the basic principles of radioactive iodine-refractory differentiated thyroid cancer / To provide a pictorial reviewof PET/TC on DTC Definition of DTC radioactive iodine-refractory: clinical, laboratorial and imaging aspects How to identify andclassify DTC radioactive iodine-refractory? Flip-Flop Phenomenon Concept of Tumor Heterogeneity and Theranostic Era Follow upevaluation: how to do it? FDG-PET/CT: basic concepts and uptake mechanism Indications: diagnosis, follow up, treatment responseand therapeutic planning PET/TC on DTC Future Perspectives

NM121-ED-X

Pearls and Pitfalls of PET/CT in Infection and Inflammation: Be Aware!



ParticipantsLucas Farias, MD , Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseIgor G. Padilha, MD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseDaniel C. Menezes, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMaria L. Soares, MD, Maceio, Brazil (Abstract Co-Author) Nothing to DiscloseRaphael M. Fernandes, MD, Maceio, Brazil (Abstract Co-Author) Nothing to DiscloseYves B. Costa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarla J. Justo Dos Santos, Maceio, Brazil (Abstract Co-Author) Nothing to DiscloseMarcia R. Lemos, MD, Maceio, Brazil (Abstract Co-Author) Nothing to DiscloseChristiana M. Miranda, MD, Maceio, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

The purposes of this exhibit are to: - Review structural and functional imaging abnormalities on positron emission tomography /computed tomography (PET/CT); - Illustrate the concepts of 18F-FDG uptake due to infection and inflammatory injuries; -Correlate important findings with the clinical, anatomy and pathogenesis of non-oncological thoracic diseases; - Highlight theircharacteristics in order to familiarize radiologists with these conditions, preventing unfavorable patient outcome.


- How does PET/CT work? - Pathophysiology: 18F-FDG uptake in infection and inflammatory injuries. - Review of imaging findingsaccording to illustrative cases of: o Neuroradiology / Head & Neck; o Chest; o Breast; o Abdominal / Urogenital; o Vascular; oMusculoskeletal. - Sample cases of pearls, pitfalls, diagnostic difficulties and mimics. - Summary and take home messages.

NM122-ED-X

Is There A Role for 18F-FDG-PET/CT in the Evaluation of Neuroendocrine Tumors in the DOTATATE Era?


ParticipantsSumara A. Gouveia, Rio de Janeiro, Brazil (Presenter) Nothing to DiscloseJose Leite, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseRoberta Hespanhol, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseRenata Moreira, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseMayra L. Moreira, MD,MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseFlavia P. Lopes, MD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

• We are experiencing an increasing incidence of desdifferentiation of neuroendocrine tumors (NET) worldwide; • Low grade tumorspresent a larger number of somatostatine receptors compared to high grade tumors which have more GLUT transporters; • G1 NET:68Ga-DOTATATE PET/CT is recommended for diagnosis and elegibility to radionuclide therapy; • G3 NET: 18F-FDG-PET/CT isrecommended for follow up and as a prognostic biomarker when positive; • 18F-FDG is an essential tool in the follow up ofprogressive disease to in order to change the failure therapy;


1 - Definition and characterization of neuroendocrine tumors (NET) 2 - Pathophysiology and classification of NET 3 - Uptakemechanisms of 18F-FDG 4 - Uptake mechanisms of 68-Ga-DOTATATE in NET 5 - PET/CT as an essential tool for diagnosis, follow-upand prognosis of aggressive NET.

NM123-ED-X

Pearls, Pitfalls, Interpretative Criteria and Utility of F18-Fluciclovine PET CT and Prostate MRI in BiochemicallyRecurrent Prostate Cancer - A Case Based Review


ParticipantsVaraha Tammisetti, MD, Houston, TX (Presenter) Nothing to DiscloseDavid Q. Wan, MD, Houston, TX (Abstract Co-Author) Nothing to DiscloseHye Ryung Yang, San Antonio, TX (Abstract Co-Author) Nothing to DiscloseSpencer C. Behr, MD, Burlingame, CA (Abstract Co-Author) Research Grant, General Electric Company Consultant, General ElectricCompany Consultant, Navidea Biopharmaceuticals, Inc Grant, Navidea Biopharmaceuticals, IncAndrew B. Rosenkrantz, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseBaris Turkbey, MD, Bethesda, MD (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1. Discuss different disease states of biochemically recurrent and metastatic prostate cancer and also role of imaging in thesesettings. 2. Understand and illustrate interpretative crieria, imaging patterns, pitfalls and limitations of Fluciclovine PET by casebased approach. Discuss role of prostate MRI with Fluciclovine PET. 3. Discuss current challenges with Fluciclovine PET and theimpact on patient management.


1. Current concepts in the understanding and challenges of biochemical recurrence of prodstate cancer including oligometastaticdisease, metastatic castration sensitiove prostate cancer, metastatic castration resistant prostate cancer, chemorefractorymCRPC,. Role of imaging in each of the scenarios. 2. Mechanism of action and imaging protocol of Fluciclovine PET and comparisonwith other PET tracers. Impact of PSA levels and PSA kinetics. Role of prostate MRI with Fluciclovine PET. 3. Case based review ofimaging patterns, diagnostic criteria, differential diagnosis and pitfalls. 4. Current challenges and limitations with Fluciclovine PETincluding lack of theranostic applications. 5. Impact of Fluciclovine PET on patient management.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Spencer C. Behr, MD - 2017 Honored Educator

NM124-ED-X

Look at the Top Before You Sign Off: Incidental Intracranial Findings on Whole Body FDG PET/CT ScansObtained for Non-CNS Malignancies


ParticipantsJoseph M. Curcio, DO, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseLisa Walker, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseNils Grosse Hokamp, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseKai Roman Laukamp, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseShiraz Rahim, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAmit Gupta, MD, Cleveland, OH (Presenter) Nothing to Disclose


[email protected]

TEACHING POINTS

To understand the pattern of physiological FDG uptake in the brain, which is helpful in differentiating physiologic uptake from truelesions. To highlight importance of using different adjustments with CT and FDG data to increase the sensitivity of lesion detection.To illustrate the spectrum of brain abnormalities that can be incidentally found on whole body scans, and can affect patientmanagement. To demonstrate the utility of including brain in the field of view for routine oncologic examinations


Normal patterns of FDG uptake in the brain. Intense upatekin the cortex and deep grey matter structures. Relative photopenia inthe white matter and the ventricles. Importance of including brain in the field of the view on routine torso PET/CT examinations.Manoeuvers to increase lesion detection. Use of brain CT window setting Importance of scaling the PET data to account forphysiologic intense cortical uptake. Realization of the fact that even a low dose CT is adequate in detecting gross abnormalitiesand may help in lesion characterization. Spectrum of incidental intracranial finings: Brain metastases(Hypermetabolic andhypometabolic) Benign cystic lesions Infarct and encephalomalacia Epileptogenic activity Post- surgical and post therapy relatedchanges. Ventricluar system abnormalities

NM125-ED-X

Pearls and Pitfalls of Quantitative Software Analysis of Dopamine Transporter Single Photon EmissionComputed Tomography for Parkinsonian Syndromes


ParticipantsRaza Mushtaq, MD, Tucson, AZ (Presenter) Nothing to DiscloseLaura Steinmeyer, Tucson, AZ (Abstract Co-Author) Nothing to DiscloseRyan J. Avery, MD, Tucson, AZ (Abstract Co-Author) Nothing to DisclosePhillip Kuo, MD,PhD, Tucson, AZ (Abstract Co-Author) Author, MD Training at Home; Research Grant, Astellas Group; Consultant,Endocyte, Inc; Consultant, General Electric Company; Education Grant, General Electric Company; Speakers Bureau, Eli Lilly andCompany; Consultant, inviCRO, LLC; Consultant, Imaging Endpoints; Consultant, Progenics Pharmaceuticals, Inc


[email protected]

TEACHING POINTS

Dopamine transporter (DaT) single photon emission computed tomography (SPECT) with I-123 ioflupane can aid in the diagnosis ofParkinsonian syndromes by demonstrating the loss of DaT in the striatum and correspondingly dopaminergic neurons in thenigrostriatal pathway. DaT SPECT is approved for visual interpretation by the FDA; however, quantitative software analysis is ahelpful adjunct to visual interpretation. Through case-based learning, we demonstrate how quantitative software analysis assistsvisual interpretation by adjusting for patient's age, correcting for head tilt/orientation and avoiding misinterpretations due to scalingartifacts from salivary gland or atypically intense caudate activity. We will also show cases where the software analysis fails suchas in vascular Parkinsonism and normal variation.


DaTQUANT - How to interpret Normal aging Artifact from head tilt/orientation Variant size/morphology of striatum Salivary glandartifact Hot caudate sign Lewy body symmetry DaTQUANT failure - Vascular Parkinsonism DaTQUANT failure - Patient variation

NM126-ED-X

Beyond a Needle and a Hot Spot: Biopsy of Osseous Lesions Identified on PET


AwardsCum Laude

ParticipantsStephen J. Nogel, MD, Rochester, MN (Presenter) Nothing to DiscloseAnn Packard, MD, Rochester, MN (Abstract Co-Author) Nothing to DiscloseAnil N. Kurup, MD, Rochester, MN (Abstract Co-Author) Research Grant, Galil Medical Ltd; Research Grant, EDDA Technology, Inc;Royalties, Wolters Kluwer nvMark A. Nathan, MD, Rochester, MN (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1. Describe the various imaging findings for bony lesions on PET/CT and PET/MRI which can lead to biopsy. 2. Discuss theprocedural techniques of CT-guided bone biopsy, with particular attention to guidance from prior PET/CT and PET/MRI. 3. Discussthe imaging pitfalls with osseous lesions on PET/CT and PET/MRI. 4. Evaluate the differences between bony lesions on PET/CT andPET/MRI.


A. Imaging findings 1. FGD uptake 2. Other radiotracers 3. CT and MRI findings B. Procedural techniques 1. Biopsy planning based onPET 2. Anatomic considerations 3. Sampling technique C. Pitfalls 1. Benign lesions 2. PET/CT and MRI challenges -Co-localization -Artifact -Background uptake D. Comparing PET/CT and PET/MRI

NM127-ED-X

Non-Prostatic Diseases on PSMA-PET Imaging: A Spectrum of Benign and Malignant Findings



ParticipantsMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseFelipe D. Barbosa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRafael F. Nunes, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseElaine C. Zaniboni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseLarissa B. Costa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJose F. Marin, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Teaching points i. Recognize the spectrum of diseases that need to be reported but are not related to prostatic disease ii. Highlightimaging key features for differential diagnosis with metastatic disease iii. Discuss new potential non-prostate indications for PSMAPET imaging


i. Non-prostatic surgery-related a. Foreign-body reaction ii. Bone-related a. Fracture, hemangioma, Fibrous osseous defect, Fibrousdysplasia, Osteomyelitis, Multiple Myeloma, Paget's disease iii. Inflammatory/Infectious a. Sarcoidosis, Tuberculosis, Renal abscess,Chronic pancreatitis iv. Neoplastic a. Benign i. Meningioma, Thyroid Follicular Adenoma b. Malignant i. Lung adenocarcinoma, Rectumadenocarcinoma, Lymphoma, Thymus carcinoma, Adenoid-Cystic carcinoma.

NM128-ED-X

Experience and Clinical Benefit of SPECT CT for Sentinel Node Biopsy in Oral Cavity SCC


ParticipantsIan McLaughlin, MBChB, Glasgow, United Kingdom (Presenter) Nothing to DiscloseAlice Nicol, PhD,MSc, Glasgow, United Kingdom (Abstract Co-Author) Nothing to DiscloseJeremy McMahon, MBBCh,BDS, Glasgow, United Kingdom (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Head & Neck Cancer is common and the incidence of cervical nodal metastasis has significant prognostic and treatmentimplications. The incidence of occult nodal metastasis is 16% in stage 1 oral cavity, and 34% in tongue cancer. Sentinel nodebiopsy (SNB) helps in the detection of occult cervical nodal metastasis and enables stratification of patients with stage 1 oralcavity SCC into watch & wait groups avoiding morbid adjuvant treatments, and into those requiring neck dissection or radiotherapy.SNB has previously been performed using planar images. We now describe the benefits of a large series of SPECT CT. This exhibitaims to: 1) Review the incidence of Head & Neck SCC and importance of identification of occult nodal metastasis for prognosis andtreatment planning. 2) Describe technical aspects of SPECT CT and familiarize the reader with a newer technique and benefits. 3)Show pictorial examples and suggest a structured report providing the surgeon with what they need to know. 4) Make the readeraware of pitfalls in technique and interpretation.


1) Introduction: Incidence of oral cavity SCC. Importance of nodal staging for treatment stratification. 2) Technique: Injection,SPECT CT, image fusion. 3) Pictorial review of cases. Example of structured report. 4) Pitfalls and difficulties. 5) Conclusion. Keylearning points.

NM129-ED-X

68Ga-DOTATATE PET Neuroimaging: A Case-Based Pictorial Review



ParticipantsFang M. Zhu, MD, PhD, St. Louis, MO (Presenter) Nothing to DiscloseMaria R. Ponisio, MD, St. Louis, MO (Abstract Co-Author) Speakers Bureau, Siemens AG


[email protected]

TEACHING POINTS

1. Provides representative cases of neuroradiological findings observed in patients undergoing 68Ga-DOTATATE whole-body PET. 2.In several cases, the 68Ga-DOTATATE PET findings modified treatment planning.


Background - Pathophysiology of SSTRs positive neoplasms Review of imaging findings • 68Ga-DOTATATE PET images •Otherconventional images Sample cases Summary Background: We present a pictorial review of neuroimaging with 68Ga-DOTATATE PET-CT and PET-MRI, in patients with known or suspected NET undergoing whole-body PET. Methods: On retrospective reviewed of 233patients who had undergone whole-body 68Ga-DOTATATE PET imaging, 19 patients had neuroradiological 68Ga positive findings.When available, the PET images were correlated with conventional imaging. The cases included meningiomas, paraganglioma, jugularglomus tumor, pituitary macroadenoma, vertebral hemangiomas, multiple hemangioblastomas, and intracranial/skull base metastasisof the neuroendocrine tumor. Syndromic associations were noted, including von Hippel-Lindau syndrome, familial paragangliomasyndrome and multiple endocrine neoplasia syndromes. Summary: 68Ga-DOTATATE PET images accurately identified theaforementioned lesions, detecting lesions not identified on MRI or FDG/PET, adding value in identifying small or missed lesions incomplex anatomical regions.

NM130-ED-X

Precision Medicine: Y 90 Therasphere and Sirsphere Tumor Targeting in Liver Hepatocellular Carcinoma


ParticipantsVani Vijayakumar, MD, Ridgeland, MS (Presenter) Nothing to DiscloseKevin A. Zand, MD, Jackson, MS (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Present Y 90 Therasphere and Sirsphere Indications and differences Discuss Pre and post Multimodality Imaging - Tc MAA Liverplanar, SPECT, Fusion with MRI , PET CT Precise tumor targeting- Mechanism of action Show case examples


Selective internal radiation therapy (SIRT) has become a widely employed brachytherapy for the treatment of primary andmetastatic hepatic malignancies. The main difference between glass spheres (Theraspheres) and resin spheres ( Sirspheres) is theactivity in each sphere, being much higher in a glass sphere (about 2,500 Bq) at production time, with respect to about 50 Bq inone resin sphere. Commercially available vials of glass spheres contain up to 20 GBq, while resin sphere ones up to 3 GBq. Due tothe increased vascularity of tumor nodules in Hepatocellular carcinoma ( HCC) and the hepatic artery being the main blood supply ofthe liver nodules, SIRT provides a treatment option for targeted therapy for primary HCC and hepatic metastasis. TargetedTherasphere and Sirsphere imaging localization and therapy shown with precise localization in caudate lobe and left lobe of liver(Fig1-5). SIRT delivers personalized precision tumor killing and hence prolongs patient survival . SIRT can be repeated if showsimprovement in tumor shrinking. SIRT success is due to precision target and delivery.

NM131-ED-X

The Hunt is On: Finding the Culprit in Patients with Hyperparathyroidism


ParticipantsMatthew H. Jenson, MD, St. Louis, MO (Presenter) Nothing to DiscloseKatherine Lambert, DO, Jacksonville, FL (Abstract Co-Author) Nothing to DiscloseAlexandra High, DO, Jacksonville, FL (Abstract Co-Author) Nothing to DiscloseFranz W. Toro Pape, BS, MD, San Juan, PR (Abstract Co-Author) Nothing to DiscloseSavas Ozdemir, MD, Jacksonville, FL (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

Radiology plays an important role in evaluating patients with hyperparathyroidism, primarily in identifying parathyroid adenomas. Avariety of imaging protocols are available to be used in this endeavor. It is incumbent on the practicing radiologist to be familiarwith these different protocols in order to better serve patients with hyperparathyroidism. At the end of this educational exhibit thereader should be able to: 1. Understand available parathyroid imaging protocols. 2. Accurately diagnose parathyroid pathology usingdifferent imaging protocols.


I. Background II. Overview of Protocols: a) 99mTc-sestamibi planar imaging b) 99mTc-sestamibi SPECT c) 99mTc-sestamibiSPECT/CT d) Dual isotope exam with 99mTc-sestamibi and 99mTc-pertechnetate e) Dual isotope exam with 99mTc-sestamibi and123I f) 4D-CT g) 99mTc-sestamibi SPECT/4D-CT h) Ultrasound III. Cases: a) Unknown cases utilizing the above listed protocols willbe presented in a quiz format. b) Cases will include parathyroid adenomas, ectopic parathyroid adenomas, and parathyroidhyperplasia.

NM132-ED-X

Forget Something? Update and Review on Dementia and Parkinson's Disease Nuclear Medicine Imaging


ParticipantsHenry Chen, MD, Cleveland, OH (Presenter) Nothing to DiscloseShivani Pahwa, MD, Clevelnad, OH (Abstract Co-Author) Nothing to DiscloseRobert S. Jones, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseJames K. O'Donnell, MD, Cleveland, OH (Abstract Co-Author) Speakers Bureau, Eli Lilly and Company

TEACHING POINTS

• Review current understanding of Dementia, etiologies and pathophysiology • Understand current radiotracers used for Dementiaimaging, protocols and pertinent findings • Understand the advances in Dementia imaging with newly approved radiopharmaceuticalincluding F-18 (florbetapir, Amyvid)


Summary of Dementia and Parkinson's Disease Epidemiology Clinical Course Physiology Prognosis Current Imaging Clinical indicationsfor Nuclear Medicine Imaging in Dementia and Parkinson's disease Radiopharmaceutical Physiology (F-18 FDG, I-123 Datscan, F-18DOPA) Protocol at our institution and Image Acquisition Differential diagnosis Pertinent differentiating imaging findings Advances inImaging - F-18 (florbetapir, Amyvid) Clinical indications Radiopharmaceutical Physiology Protocol and Image Acquisition Pertinentimaging findings Caveats with Impression

NM133-ED-X

Indium-111 WBC and Gallium-67 Citrate SPECT/CT Evaluation of Infection: A Head-To-Toe Case-BasedReview



ParticipantsBradley A. Nowack, MD, Shreveport, LA (Presenter) Nothing to DiscloseMingxia Shi, Shreveport, LA (Abstract Co-Author) Nothing to DiscloseMeghna Chadha, MD, MBBS, Shreveport, LA (Abstract Co-Author) Nothing to DiscloseChiranjiv Virk, Shreveport, LA (Abstract Co-Author) Nothing to DiscloseMuzammil Aziz, Shreveport, LA (Abstract Co-Author) Nothing to DiscloseZhiyun Yang, MD, Shreveport, LA (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

1. Discuss specific advantages of In-111 oxide WBC and Ga-67 citrate SPECT/CT for detection of infections.2. Demonstrateinterpretative pitfalls in infection imaging.3. Present a novel simultaneous SPECT/CT acquisition technique to define infection sites.


Infection remains a major diagnostic challenge. Early diagnosis and precise localization are critical for effective treatment, especiallywhen planning surgery. Structural abnormalities from infection may be detected by high-resolution anatomical imaging, but typicallymanifest later and are often non-specific. WBC and Ga planar and SPECT imaging can provide evidence of early infectious changes,but lack anatomic detail. Through a case-based review, we demonstrate the advantages of SPECT/CT, particularly in providinganatomical and functional information simultaneously. SPECT/CT can cover multiple anatomic regions of interest, secure earlierdiagnosis, pinpoint infectious sites, and monitor treatment response.I. SPECT/CT for early diagnosis of multiple sites of infection.II.SPECT/CT for identifying extent of disease and monitoring treatment response.III. SPECT/CT infection imaging pitfalls.IV.Simultaneous WBC and Tc-99m-sulfur colloid SPECT/CT protocol for bone and soft tissue infections.V. Simultaneous Ga and Tc-99m-MDP SPECT/CT protocol for bone and soft tissue infections.

NM134-ED-X

Normal Variants and Abnormal Patterns of Pancreatic Uptake on Ga-68 DOTATATE PET/CT



ParticipantsJeanna M. Harvey Barnes, MD, Providence, RI (Presenter) Nothing to DiscloseDon C. Yoo, MD, E Greenwich, RI (Abstract Co-Author) Consultant, Endocyte, Inc


[email protected]

TEACHING POINTS

Draw on examples from over 30 patients over a 10 month period since FDA approval to illustrate normal and abnormal appearancesof the pancreas on Ga-68 DOTATATE PET/CT. Review the mechanism for physiologic Ga-68 DOTATATE uptake in the pancreas.Discuss and illustrate common pathologic findings within the pancreas seen on Ga-68 DOTATATE PET/CT.


Illustrate and discuss normal physiologic uptake within the pancreas showing the appearance of multiple normal variants.Demonstrate characteristic imaging findings of abnormal and pathologic uptake within the pancreas on Ga-68 DOTATATE PET/CTusing CT and MRI correlates when possible. Postsurgical appearance of the pancreas Neuroendocrine masses and metastases withinthe pancreas Pitfalls of imaging the pancreas, including adjacent duodenal and splenic lesions

NM135-ED-X

Update on Hodgkin's Disease from a Radiology Perspective


ParticipantsIsaac Levine, Cleveland, OH (Presenter) Nothing to DiscloseKevin R. Kalisz, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAmit Gupta, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseKianoush Ansari-Gilani, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseNikhil H. Ramaiya, MD, Jamaica Plain, MA (Abstract Co-Author) Nothing to DiscloseFrancesco Alessandrino, MD, Boston, MA (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

To provide an overview of the types, stages, and prognostic classifications of Hodgkin lymphoma (HL) To describe the current roleof imaging in HL To describe the evolution in treatment of HL from early chemotherapy and radiation therapy (RT) to the currentarray of treatment options for HL, in view of the targeted therapies, stem cell transplantation and immune checkpoint inhibitors


Overview of HL Types Classical HL Nodular lymphocyte-predominant HL Staging: Ann Arbor (1974) staging criteria: Stages I-IV withmodifiers Cotswolds (1988) modifications Lugano (2014) criteria International Prognostic Score Favorable risk disease, unfavorablerisk disease Current role of imaging in HL Representative imaging examples for diagnosis, staging and defining prognosisChemotherapy From nitrogen mustard to MOPP, ABVD, and BEACOPP (1992) RT From extended field radiation therapy to involvedsite radiation therapy (2014) Current treatment of HL Current role of chemotherapy and RT Stem cell transplantation Targetedtherapies: - Example: monoclonal antibodies, including anti-CD30 brentixumab vedotin (FDA approval 2011) Immune checkpointinhibitors: - Example: pembrolizumab (FDA approval 2017) Representative imaging examples for all treatments in various phases ofdisease

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Nikhil H. Ramaiya, MD - 2017 Honored Educator

NM136-ED-X

FDG-PET Evaluation of Cardiac Sarcoidosis: Technique, Diagnosis, and Extracardiac Findings


ParticipantsSardius Chen, MD , Los Angeles, CA (Presenter) Nothing to DiscloseLouise J. Thomson, MBCHB, Los Angeles, CA (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

Cardiac sarcoidosis is a high mortality and morbidity condition that can be diagnosed with high sensitivity and specificity with F18-FDG PET. Study quality achieved through proper patient preparation is significant for accurate diagnosis and early treatment. Mostrecent American Society of Nuclear Cardiology guidelines for producing consistent and high quality studies prescribe a low-carbohydrate, high-fat and high-protein diet. F18-FDG PET can be used to follow treatment effects and for longitudinal follow-up.Extracardiac involvement can be identified on cardiac PET studies.


Indications, clinical concerns, pathophysiology of sarcoidosis. Technique and protocols. Patient preparation: suppression ofmyocardial glucose utilization, ASNC guidelines, sample imaging flowchart and protocol. Interpretation, Nuances and Pitfalls:Common patterns of FDG uptake, Role of myocardial perfusion studies, Quality control. Case series: Diagnosis, evaluation oftreatment effects, longitudinal follow-up. Alternative Diagnoses: Cardiac device infection, myocarditis. Common extracardiacfindings on PET.

NM137-ED-X

Scintigraphic Presentation of Benign Bone Lesions in Children


ParticipantsMiller Jennings, DO, Jackson, MS (Presenter) Nothing to DiscloseVani Vijayakumar, MD, Ridgeland, MS (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

At the end viewers will learn: Common Benign Bone Lesions in Children, Scintigraphic Presentation of the lesions, Several Caseexamples


Benign bone lesions in the pediatric population are commonly discovered as incidental findings during workup for other conditions orcomplaints. Some of the most common benign bone lesions in the pediatric and adolescent population include( Table 1) Fibrousdysplasia (FD) is one of the common benign bone lesions in which normal bone is replaced by fibrous tissue. Commonly presentingduring adolescence and young adulthood, may occur in single or multiple bones: most commonly occurring in the proximal femur,tibia, ribs, and skull. In widespread disease, the skull and jaw are usually always involved. FD typically presents with pain, swelling,and even pathologic fractures. Symptoms of headache can be found in patients with skull involvement. FD lesions are among themost intensely avid lesions observed on MDP bone scintigraphy (Fig 1). The degree of uptake is reliant on the degree ofosteoclastic activity as well as blood flow. Enchondroma ( Fig2),Langerhans cell Histiocytosis ( Fig 3), Sickle Cell Disease, andosteomyelitis (Fig 4) are few other benign lesions which will be presented.

NM138-ED-X

PET/CT Imaging of Tumor Response and Immune-Related Side Effects of Checkpoint Inhibitor Therapy inCutaneous Melanoma


ParticipantsJames T. Frencher, MD, San Francisco, CA (Presenter) Nothing to DiscloseAdil Daud, San Francisco, CA (Abstract Co-Author) Nothing to DiscloseSpencer C. Behr, MD, Burlingame, CA (Abstract Co-Author) Research Grant, General Electric Company Consultant, General ElectricCompany Consultant, Navidea Biopharmaceuticals, Inc Grant, Navidea Biopharmaceuticals, IncMiguel Pampaloni, MD, San Francisco, CA (Abstract Co-Author) Nothing to DiscloseRobert R. Flavell, MD, PhD, San Francisco, CA (Abstract Co-Author) Nothing to DiscloseThomas A. Hope, MD, San Francisco, CA (Abstract Co-Author) Research support, General Electric Company

TEACHING POINTS

1. Checkpoint inhibitors represent a new class of highly effective cancer therapies which block the tumor's ability to suppresstumor-specific lymphocytes, resulting in augmented antit-tumor host response. 2. Immunotherapy-induced lymphocyte activationand tumor infiltration may causes apparent increase in size and FDG avidity of tumor on PET/CT despite response to treatment(pseudo-progression). irRECIST criteria represents a modified set of rules to evaluate response to treatment with checkpointinhibitors. 3. Clinically significant organ-specific immune related adverse events are visible on PET/CT evaluation of treatedpatients.


Table of Contents/Outline: 1. Checkpoint inhibitors a. Mechanism of Action b. Checkpoint inhibitors currently FDA approved c.Immune activation cancer therapies in development 2. Determining response and progression on CT, MRI, and PET/CT followupimaging a. RECIST b. PERCIST c. irRECIST 3. Imaging of immune-related adverse events using FDG PET/CT 4. Imaging correlates ofearly response to treatment a. Immune-related adverse events b. FDG uptake in lymphoid tissues

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Spencer C. Behr, MD - 2017 Honored Educator

NM139-ED-X

Patterns of Lung Cancer Recurrence on 18F-FDG PET/CT: Loco-Regional and Distant Disease


ParticipantsCarson Sibley, MD, Dallas, TX (Presenter) Nothing to DiscloseAsha Kandathil, MD, Dallas, TX (Abstract Co-Author) Nothing to DiscloseRathan M. Subramaniam, MD,PhD, Dallas, TX (Abstract Co-Author) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue EarthDiagnostics Ltd


[email protected]

TEACHING POINTS

- Although therapeutic options for patients with lung cancer have evolved dramatically in the last decade, patients have a high riskof loco-regional and/or distant tumor recurrence, especially within the first two years.- In patients with suspected loco-regionalrecurrence on surveillance chest CT, NCCN guidelines recommend 18F-FDG PET/CT for detection of distant metastases and fordifferentiating true malignancy from post treatment changes.- F18-FDG PET/CT provides a guide to biopsy and may identify distantmetastases , impacting management.- Subsequent therapy and survival depends on extent of recurrent disease, which is bestevaluated with 18F-FDG PET/CT.- The role of 18F-FDG PET/CT in post treatment surveillance of lung cancer patients is evolving asprognostic and therapeutic data emerge.


- Review current therapeutic options for patients with lung cancer and expected post treatment changes- Illustrate patterns oflung cancer recurrence: loco-regional and distant disease - Illustrate the role of 18F-FDG PET/CT in detection of biopsy provenloco-regional and distant recurrence.- Illustrate pitfalls of 18F-FDG PET/CT in evaluation of post treatment recurrence such as falsepositive uptake in post radiation changes and pleurodesis.

NM140-ED-X

The Peritoneum: What Nuclear Radiologists Need to Know



ParticipantsMohammed S. Bermo, MD, Seattle , WA (Presenter) Nothing to DiscloseMeena Kumar, MD, Galveston, TX (Abstract Co-Author) Nothing to DiscloseBhasker Rao Koppula, MBBS, Salt Lake City, UT (Abstract Co-Author) Nothing to DiscloseAntoine Leblond, MD, Montreal, QC (Abstract Co-Author) Nothing to DiscloseManuela C. Matesan, MD, PhD, Seattle, WA (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

1- The peritoneum is a serosal surface enclosing the peritoneal cavity. The peritoneal cavity is composed of multiple communicatingspaces containing trace amount of peritoneal fluid. The most common sign of peritoneal disease is exudation of peritoneal fluid.2-Anatomical abnormalities of the peritoneal cavity include abnormal communication between the peritoneal cavity and scrotal tunicalspace through a patent processus vaginalis, communication between the peritoneal cavity and pleural cavities throughpleuroperitoneal channels, and adhesions resulting in compartmentalization of part of the peritoneal cavity. 3-Peritoneal pathologyencountered in nuclear medicine practice is usually inflammatory or neoplastic, whether primary or secondary.


1-Peritoneal embryology and anatomy: Peritoneal spacesIntraperitoneal and retroperitoneal organs Potential peritonealcommunications 2-Common peritoneal pathologies: Abnormal communication with the scrotum or pleural spaces Peritoneal adhesionsPeritoneal collectionsPeritoneal inflammation Primary peritoneal neoplasmsSecondary peritoneal neoplasms3-Diagnostic tools FDG-PET Peritoneal ScintigraphyRadiolabeled white blood cellsOther Nuclear Medicine techniques

SPPH01A PET/CT Introduction and Clinical Applications

SPPH01B PET/MR Introduction and Clinical Applications

SPPH01C Quantitative SPECT

SPPH01

AAPM/RSNA Physics Tutorial Session 1

Saturday, Nov. 24 12:00PM - 2:00PM Room: E351

BQ GI MR NM PH

AMA PRA Category 1 Credits ™: 2.00ARRT Category A+ Credits: 2.25


ParticipantsThaddeus A. Wilson, PhD, Madison, WI (Moderator) Nothing to Disclose

Sub-Events

ParticipantsOsama R. Mawlawi, PhD, Houston, TX (Presenter) Research Grant, General Electric Company; Research Grant, Siemens AG


[email protected]

LEARNING OBJECTIVES

1) Describe the latest advances in hardware and software for PET/CT imaging. 2) Explain how these advances affect PET imagequality and quantification. 3) Describe novel PET/CT clinical applications using new radiopharmaceuticals. 4) Discuss futuredevelopments and clinical applications of PET/CT imaging.

ParticipantsRobert A. Pooley, PhD, Jacksonville, FL (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Describe reasons for combining PET and MR into a single scanner. 2) Explain how PET instrumentation can affect MR, and howMR instrumentation can affect PET. 3) Identify PETMR protocol acquisition strategies. 4) Describe clinical applications of PETMR.

ParticipantsBenjamin M. Tsui, PhD, Baltimore, MD (Presenter) Researcher, Koninklijke Philips NV; License agreement, General Electric Company;


[email protected]

LEARNING OBJECTIVES

1) Define quantitation and quantitative SPECT. 2) List and describe the image degrading factors of SPECT. 3) Describe methods tocompensate for the SPECT image degrading factors. 4) Assess quality and quantitative accuracy improvements of quantitativeSPECT images. 5) Apply quantitative SPECT to clinical practices.

ABSTRACT

Recent development and application of quantitative SPECT have provided significantly improved image quality and quantitativeaccuracy that aid in clinical diagnosis and treatment of diseases. In this educational course, we will define quantitation,quantitative SPECT and its goals. The image degrading factors of SPECT will be listed and described. Methods that compensates forthe image degradation factors will be presented and explained. Examples of improvements in image quality and quantitativeaccuracy in various clinical applications will be presented.

Active Handout:Benjamin M. Tsui

http://abstract.rsna.org/uploads/2018/18000347/RSNA-QuantitatveSPECT HandoutSPPH01C.pdf

SPPH02A The Nuts and Bolts of Dosimetry in Nuclear Medicine and its Application

SPPH02B Theranostics Introduction and Applications

SPPH02

AAPM/RSNA Physics Tutorial Session 2

Saturday, Nov. 24 2:15PM - 4:15PM Room: E351

NM PH



ParticipantsThaddeus A. Wilson, PhD, Madison, WI (Moderator) Nothing to Disclose

Sub-Events

ParticipantsMichael G. Stabin, PhD, Nashville, TN (Presenter) Software, Vanderbilt University

LEARNING OBJECTIVES

1) Understand current models and methods used in radiation dosimetry for radiopharmaceuticals. 2) Understand methods for newdrug approval with the FDA. 3) Understand current experience with radiopharmaceuticals used in therapy: types of compounds,clinical experience, biological effects.

ParticipantsHossein Jadvar, MD, PhD, Pasadena, CA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Define theranostisc. 2) Review the history and current clinical applications of theranostics. 3) Describe potential outlook fortheranostics in the era of precision medicine.

ABSTRACT

Advances in the understanding of cancer biology, developments in diagnostic technologies, and expansion of therapeutic optionshave all contributed to the concept of personalized cancer care. Theranostics is the systematic integration of targeted diagnosticsand therapeutics. The theranostic platform includes an imaging component that 'sees' the lesions followed by administration of thecompanion therapy agent that 'treats' the same lesions. This strategy leads to enhanced therapy efficacy, manageable adverseevents, improved patient outcome, and lower overall costs. In this lecture, I review the the concept, history, recent developments,current challenges, and outlook for radiotheranostics (use of radionuclides in theranostics) in the management of patients withcancer (Jadvar H et al. Radiotheranostics in Cancer Diagnosis and Management. Radiology 2018; 286:388-400).

ED010-SU

Nuclear Medicine Sunday Case of the Day

Sunday, Nov. 25 7:00AM - 11:59PM Room: Case of Day, Learning Center

AMA PRA Category 1 Credit ™: .50

ParticipantsLevi Sokol, MD, New York, NY (Presenter) Nothing to DiscloseRichard K. Brown, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseMurray D. Becker, MD, PhD, East Brunswick, NJ (Abstract Co-Author) Nothing to DiscloseJeffrey S. Kempf, MD, New Brunswick, NJ (Abstract Co-Author) Nothing to DiscloseAni Peshtani, DO, New Brunswick, NJ (Abstract Co-Author) Nothing to DiscloseGregory A. Ngo, DO, New Brunswick, NJ (Abstract Co-Author) Nothing to DiscloseEric Hu, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseTarun Jindal, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseAndrew Kaiser, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseYashesh Shah, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DisclosePeter Girgis, MD, New Brunswick, NJ (Abstract Co-Author) Nothing to DiscloseMonica N. Abghari-Gerst, MD, Bloomfield Hills, MI (Abstract Co-Author) Nothing to DiscloseDrew Kempf, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

1) Immunotherapy-related pneumonitis: Recognize pneumonitis as a complication of immunotherapy and clinical implications. 2)Peritoneal carcinomatosis on bone scan. Recognize this as one cause of extra-osseous bone scan uptake. 3) Cardiac amyloid onbone scan. Amyloidosis can be a cause of extra-osseous bone scan uptake. 4) Complex regional pain syndrome/RSD: Recognize RSDon triple phase bone scan. 5) Lipomatous hypertrophy of intra-atrial septum: Recognize this benign cause of hypermetabolic uptakeat intra-atrial septum on PET/CT. Discuss the etiology and clinical implications.

SSA13-01 Template-Enhanced ZTE Attenuation Correction for Brain FDG-PET/MR Imaging

Sunday, Nov. 25 10:45AM - 10:55AM Room: S504CD

SSA13-02 Advantages of 325ps Time-of-Flight Digital Photon Counting PET/CT in Low Dose Brain PETPerformance for Assessing Breast Cancer Metastasis


SSA13

Molecular Imaging (Brain)

Sunday, Nov. 25 10:45AM - 12:15PM Room: S504CD

MR MI NR NM

AMA PRA Category 1 Credits ™: 1.50ARRT Category A+ Credit: 1.75


ParticipantsPeter Herscovitch, MD, Bethesda, MD (Moderator) Nothing to DiscloseDima A. Hammoud, MD, Bethesda, MD (Moderator) Nothing to Disclose

Sub-Events

ParticipantsTetsuro Sekine, MD, PhD, Tokyo, Japan (Presenter) Nothing to DiscloseBradley J. Kemp, PhD, London, ON (Abstract Co-Author) Nothing to DiscloseSandeep Kaushik, MS, Bangalore, India (Abstract Co-Author) Employee, General Electric CompanyFlorian Wiesinger, Garching bei Munchen, Germany (Abstract Co-Author) Employee, General Electric Company Gaspar Delso, PhD, Cambridge, United Kingdom (Abstract Co-Author) Employee, General Electric Company;


[email protected]

PURPOSE

The impact of MR-based attenuation correction on PET quantitation accuracy is an ongoing cause of concern for advanced brainresearch with PET/MR. The purpose of this study was to evaluate a new, template-enhanced zero-echo-time (ZTE) attenuationcorrection method for PET/MR scanners.

METHOD AND MATERIALS

30 subjects underwent a clinically-indicated 18F-FDG-PET/CT, followed by PET/MR on a GE SIGNA PET/MR. For each patient, a 42-second ZTE sequence was used to generate two attenuation maps: one with the standard ZTE segmentation-based method; andanother with a modification of the method, wherein pre-registered anatomical templates and CT data were used to enhance thesegmentation. CT data, was used as gold standard. Reconstructed PET images were qualified visually and quantified in 68 volumes-of-interest using a standardized brain atlas.

RESULTS

Attenuation maps were successfully generated in all cases, without manual intervention or parameter tuning. The PET bias withtemplate-enhanced ZTE attenuation correction was measured to be -0.9% ± 0.9%, compared with -1.4% ± 1.1% with regular ZTEattenuation correction. In terms of absolute bias, the new method yielded 1.1% ± 0.7%, compared with 1.6% ± 0.9% with regularZTE. Statistically significant bias reduction was obtained in the frontal region (from -2.0% to -1.0%), temporal (from -1.2% to -0.2%), parietal (from -1.9% to -1.1%), occipital (from -2.0% to -1.1%) and insula (from -1.4% to -1.1%).

CONCLUSION

These results indicate that the co-registration of pre-recorded anatomical templates to ZTE data is feasible in clinical practice andcan be effectively used to improve the performance of segmentation-based attenuation correction.

CLINICAL RELEVANCE/APPLICATION

The accuracy of PET/MR attenuation correction based on zero echo time (ZTE) data can be increased with the incorporation ofregistered anatomical priors in the segmentation procedure.

ParticipantsJun Zhang, PhD, Columbus, OH (Presenter) Nothing to DiscloseKatherine Binzel, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseMichelle I. Knopp, Columbus, OH (Abstract Co-Author) Nothing to DiscloseMaryam Lustberg, Columbus, OH (Abstract Co-Author) Nothing to DiscloseMichael V. Knopp, MD, PhD, Columbus, OH (Abstract Co-Author) Nothing to Disclose

SSA13-03 The Association Between Perihematomal Edema and Iron Deposition Using Diffusion Tensor Imagingand T2* Mapping after Experimental Intracerebral Hemorrhage


PURPOSE

To evaluate the feasibility and advantages of 325ps time-of-flight (TOF) digital photon counting (DPC) PET in low dose FDG brainPET performance in patients with breast cancer.


18 low dose (3.1±0.2mCi) brain FDG-PET scans from 0-75min of 9 patients (BMI=33±6) were performed on a 325ps DPC PET/CT(Philips Vereos) to assess neurometabolic changes for breast cancer before and after chemotherapy (interval=77±10 days). 5groups of PET reconstructions (10, 5, 2, 1 and 0.5 min) centered at 65min p.i. were performed in 3D-OSEM (2mm-isotropic) w(TOF) and w/o (nTOF). All data were co-registered and normalized based on a 43 normal FDG brain database with 90 neuro-anatomic regions created using MIMSoftware and Brain Atlas Mapping (1,620 regions total). SUV and Z-Score were calculated.NEMA phantom (6 spheres with contrast ratio of 4, 0.5-10min) and 3D-Hoffman phantom (0.5-10min, 0.6-2mCi FDG) were scanned.

RESULTS

NEMA TOF PET revealed 1.0-0.7 recovery coefficients for 6 spheres 37-10mm, with an average of 0%, 0%, 6%, 14%, 30% and46% higher SUVmax than nTOF's, across all PET (10-0.5min). Visually, Hoffman phantom and patient data consistently revealedmore robust image quality with improved details and better contrast on TOF than nTOF. No brain tumors were identified onpatients. Quantitatively, an average SUVmean difference (1±3%, 1±3%, 0±2%, 2±4% and 2±5%, TOF vs nTOF) was obtained forPET 10-0.5min. Significant differences in region-based therapy response (p<0.05) were found between TOF and nTOF for PET<=2min. Robust Z-Scores (<=10% variance) were found on TOF PET when reducing from 10 to 1min. TOF PET demonstrated betteradaptability than nTOF to scan time and dose reduction in image quality and quantification (details given at RSNA).

CONCLUSION

Compared to current standard of care FDG brain PET (10-13mCi, 10min), this low dose (3mCi) brain PET study with scan timereduction (10-0.5min) demonstrated advantages of the solid state 325ps DPC PET technology. The new technology advances neuroPET with more precise imaging of the brain enabled by the excellent TOF capability.


Time-of-flight improvement (325ps) enabled by the new generation solid state DPC PET/CT has advantages in improving robustnessof brain neuro PET acquisitions even under low dose and short scan time.

AwardsTrainee Research Prize - Medical Student

ParticipantsXiaoxin Liu, Yinchuan, China (Presenter) Nothing to DiscloseChunhua Wang, MD, Chengdu, China (Abstract Co-Author) Nothing to DiscloseYushu Chen, BSc, Chengdu, China (Abstract Co-Author) Nothing to DiscloseLei Wang, BA, Chengdu, China (Abstract Co-Author) Nothing to DiscloseLi Song, Chengdu, China (Abstract Co-Author) Nothing to DiscloseRuzhi Zhang, BSC, Chengdu, China (Abstract Co-Author) Nothing to DiscloseXuejun Ping, Yinchuan, China (Abstract Co-Author) Nothing to DiscloseFabao Gao, MD, PhD, Chengdu, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

This study aimed to explore the relationship between iron deposition and perihematomal edema after intracerebral hemorrhage (ICH)by diffusion tensor imaging (DTI) and T2* mapping.


14 male SD rats were included in ICH group. 6 rats were included as normal controls. In ICH group, 40 μL of autologous blood wasinjected into the right basal ganglia to mimic spontaneous ICH. All rats underwent T2WI, DTI and T2* mapping on a 7.0 T animalMRI. The time points of scan of DTI and T2* mapping were days 1, 3, and 7. We used T2WI images scanned at 3-6 h after ICH tomeasure the initial hematoma. Abnormal mean diffusivity (MD) value and volume were calculated by MD maps. T2* value andabnormal T2 * value volume were measured using T2* mapping maps.

RESULTS

In ICH group, the initial hematoma was 19.14 ± 8.151 μL. The T2* values were 21.23 ± 2.40 ms, 20.62 ± 2.96 ms, and 17.97 ±2.54 ms at 1d, 3d and 7d after ICH in the ipsilateral side and were 32.47 ± 2.11 ms, 32.51 ± 2.74 ms, and 32.43 ± 3.71 ms in thecontralateral side. In normal control group, T2* value was 32.73 ± 2.55 ms in the ipsilateral side and was 33.07 ± 2.16 ms in thecontralateral side. The T2* value in the ipsilateral side in ICH group were significant lower than normal control group at 1d, 3d and7d after ICH, respectively (all P < 0.001). There were no significant differences in T2* value between the contralateral side andnormal group (all P > 0.05). The volumes of abnormal T2* value were 30.93 ± 18.55 μL, 25.30 ± 9.27 μL, and 31.50 ± 10.58 μL andthe volumes of abnormal MD were 48.86 ± 31.51 μL, 64.30 ± 64.72 μL, and 30.63 ± 24.99 μL at 1d, 3d and 7d after ICH. A positivecorrelation was observed between abnormal T2* volume and abnormal MD volume at 1d after ICH (r = 0.92, P < 0.001).

CONCLUSION

There was a positive correlation between abnormal T2* volume and abnormal MD volume at 1d after ICH. DTI and T2* mapping hasthe potential to explore the relationship between perihematomal edema and iron overload after ICH.


DTI and T2* mapping can not only diagnose the edema and iron overload in hematoma and perihematomal area after ICH, but alsoexplore the relationship between iron deposition and perihematomal edema.

SSA13-04 Imaging Type-Three Diabetes in an Alzheimer's Disease Animal Model: A Preliminary Mouse Study


SSA13-05 Imaging Amyloid Plaques Without Contrast Agent


ParticipantsVal J. Lowe, MD, Rochester, MN (Abstract Co-Author) Research Grant, General Electric Company; Research Grant, Siemens AG;Research Grant, Eli Lilly and Company; Advisory Board, Merck & Co, IncTyler J. Bruinsma, BA, Rochester, MN (Presenter) Nothing to Disclose


[email protected]

PURPOSE

Investigating the brain distribution of iodine-125 labeled insulin (125I-Insulin) by dynamic single photon emission computedtomography/computed tomography (SPECT/CT) in mice with and without metabolic syndrome and/or Alzheimer's disease (AD).


Six-month-old APP/PS1 mice (n=6) and wild type (WT) littermates (n=6) were split into two groups. Half were fed a high fat diet(HFD) while half were fed a regular chow diet (RCD) for four months. Insulin tolerance tests were performed at 1 and 4 monthsafter feeding began. A bolus injection of 125I-Insulin was administered via the femoral vein and each mouse was imaged withSPECT/CT. Regions of interest were drawn around the brain and standard uptake values (SUV) were calculated. Following theimaging protocol, mice were perfused with PBS and individual brain regions and peripheral organs were harvested and assayed for125I activity in a dual channel gamma counter. One-way ANOVA, repeated measures ANOVA, and Student's t-test were used toassess the significance of results.

RESULTS

Blood glucose levels in HFD mice showed diminished response to insulin compared to RCD littermates (p<0.001). APP/PS1 mice onboth HFD and RCD showed attenuated reductions in blood glucose during the insulin tolerance test (p<0.02). The HFD mice hadsignificantly higher brain 125I-Insulin SUV at all time points as compared to RCD mice (p<0.005). In the RCD cohort, WT miceshowed greater brain SUV than the APP/PS1 mice between 30-60 minutes (p<0.005). Post-perfusion gamma counts revealedsignificantly lower retention of 125I-insulin in the eyes and brains of HFD mice as compared to RCD mice (p<0.05).

CONCLUSION

Although, HFD causes peripheral insulin resistance in both WT and APP/PS1 mice, the imaging showed an unexpected increase inthese mice. In contrast, lower insulin retention post-mortem was seen in both WT and APP/S1 mice on HFD. Greater peripheralinsulin resistance and lower brain insulin retention in APP/PS1 mice compared to WT mice suggests the impairment of insulin deliverythat triggers 'type-three diabetes' in the AD brain. Further work is necessary to better understand the brain kinetics of insulin.


Insulin molecular imaging is a promising new frontier for elucidating the underlying connection between AD and insulin resistance.

ParticipantsEshan Dahal, Silver Spring, MD (Abstract Co-Author) Nothing to DiscloseBahaa Ghammraoui, Silver Spring, MD (Abstract Co-Author) Nothing to DiscloseAldo Badano, PhD, Silver Spring, MD (Presenter) Nothing to Disclose

PURPOSE

PET is clinically used to quantify brain amyloid load in vivo in Alzheimer's disease (AD) patients, but requires the use of amyloid-specific radiotracer and provides no information on plaque structure. We study small-angle x-ray scattering (SAXS) imaging forstructural characterization of amyloid plaques in human brains and quantification of the amyloid load without contrast agent.Experimental SAXS images of an amyloid plaque model and Monte Carlo simulations of diagnostic energy x-ray transport in a humanhead digital model are reported to determine SAXS system design choices for amyloid imaging.


SAXS measurements were performed using a point collimation mode and monochromatic x-ray beam. Simulations of a SAXS-CTgeometry in a voxelized human head (MIDA model) were performed using a publicly available GPU-accelerated Monte Carlo radiationtransport tool. SAXS-CT images of the brain with varying amyloid load at relevant q angles were reconstructed using filtered backprojection (FBP).

RESULTS

SAXS measurements of amyloid fibrils pellet show strong scattering with distinct peaks around 6.4 and 13.4 nm-1 due to the ß-sheet fibrillar structure. SAXS is capable of detect amyloid plaques without any contrast agent based on their scattering signature.SAXS-CT simulations performed on a human head digital model with inserted amyloid plaques show feasible detection of plaques assmall as 2 mm. More realistic SAXS-CT simulation requires measured cross-section models of amyloid in the brain with different massfraction of plaques. This allows optimizing the SAXS imaging system to detect micrometer-sized amyloids near 6.4 nm-1.

CONCLUSION

Our results showcase the potential of SAXS imaging method to image amyloid plaques in the human brain and to quantify amyloidload without using contrast agent.


SAXS may surpass the amyloid imaging performance of PET if it can detect and image amyloids in the early stage of plaques

SSA13-06 Features of Corticospinal Tract Using Diffusion Tensor Imaging After Experimental IntracerebralHemorrhage


SSA13-07 The rCBV with Contrast Leakage Correction Improves the Correlation Between MR PerfusionWeighted Imaging and Fluorine-18-Deoxyglucose Positron Emission Tomography in Patients withBrain Tumors


deposition in AD patients with an effective radiation dose of 10 mSv or lower.

ParticipantsChunhua Wang, MD, Chengdu, China (Abstract Co-Author) Nothing to DiscloseXiaoxin Liu, Yinchuan, China (Presenter) Nothing to DiscloseLi Song, Chengdu, China (Abstract Co-Author) Nothing to DiscloseRuzhi Zhang, BSC, Chengdu, China (Abstract Co-Author) Nothing to DiscloseYushu Chen, BSc, Chengdu, China (Abstract Co-Author) Nothing to DiscloseFabao Gao, MD, PhD, Chengdu, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

To dynamically evaluate the effect of hematoma on corticospinal tract (CST) after intracerebral hemorrhage (ICH) by diffusiontensor imaging (DTI).


29 male SD rats were injected with 40uL of autologous blood from tail in the right basal ganglia. DTI sequence was scanned on 7.0TMRI at day 1 (D1), day 3 (D3), day 7 (D7), day 14 (D14), day 21 (D21) and day 28 (D28) after ICH. Sham and normal controlsunderwent the same scan. The initial hematoma volume was obtained from T2WI images (3-6h after ICH). Mean diffusivity (MD),axial diffusivity (AD), radial diffusivity (RD) and fractional anisotropy (FA) were obtained from DTI maps. The regions of interestincluded cerebral peduncle (CP) and pyramidal tract (PY). Modified neurological severity score (mNSS) was used to evaluate theneurological function of rats.

RESULTS

FA values of ipsilateral CP in ICH group were significant lower than in sham group at D1, D3, D7, D14 and D21 (all p < 0.05). FAvalues of ipsilateral CP in ICH group at D3, D7 were significant lower than normal controls (both p < 0.01). MD of ipsilateral CP inICH group at D1 was significant higher than sham group and at D28 was significantly lower than sham group (both p < 0.05). AD ofipsilateral CP at D7 and D28 in ICH group was significantly lower than in normal group (both p < 0.05). RD of ipsilateral CP in ICHgroup was higher than sham group at D1 and D3 and lower at D28 (all p < 0.05). No significant differences were found in DTIparameters in ipsilateral and contralateral PYs between ICH group and sham group, and among different time points and normalcontrols (all p > 0.05). The scores of mNSS in ICH group were significantly greater than in sham group (all p < 0.05). The score ofmNSS in ICH group at D1 was significantly greater than other time points (all p < 0.05).

CONCLUSION

DTI parameters of ipsilateral CP were abnormal, whereas no changes in DTI parameters of PY were found in ICH model induced by40uL autologous blood. DTI has the potential to detect dynamically the effect of hematoma on CST.


DTI reveals the changes of CST in parameters after ICH at basal ganglia.

ParticipantsXiang Liu, MD, Rochester, NY (Presenter) Nothing to DiscloseWei Tian, MD,PhD, Rochester, NY (Abstract Co-Author) Nothing to DiscloseHenry Z. Wang, MD, PhD, Pittsford, NY (Abstract Co-Author) Consultant, VirtualScopics, Inc


[email protected]

PURPOSE

Tumor angiogenesis and tumor metabolite are important for clinical management for patients with brain tumors. Although rCBVwithout contrast leakage-correction is the most widely used imaging parameter of MR dynamic susceptibility contrast perfusionweighted imaging (DSC-PWI), the rCBV with contrast leakage-correction was reported to have better accuracy in the evaluation oftumor hemodynamic abnormality. The purpose of this study is to compare the correction between these two rCBV parameters andfluorine-18-deoxyglucose (FDG) positron emission tomography (PET).


85 paired MR DSC-PWI and FDG-PET examinations in 65 patients with brain tumors, including high grade gliomas, brain metastasesand cerebral lymphomas, were enrolled in this study. The interval between MR DSC-PWI and FDG-PET examinations ranged from 0to 13 days in 72 paired MR DSC-PWI and FDG-PET examinations, another 13 paired stable post-surgical scans were acquired within28 days. The rCBV maps without and with contrast leakage correction were generated using FDA-approved GE BrainStat andNordicICE programs. Two neruoradiologists measured the maximal rCBV ratio and the tumor versus normal tissue count ratio (TNR) inthe "hot" ROIs. The correlation between maximal rCBV ratio of rCBV without and with contrast leakage correction and TNR wasevaluated with Spearman Rank correlation analysis, and the difference between these two correlations was assessed with paired t-test.

RESULTS

The mean maximal rCBV ratio of rCBV with contrast leakage correction (1.65±1.38) were higher than rCBV without contrast leakage

SSA13-08 Peptide Functionalized Nano-Inhibitors Restrain Brain Tumor Growth by Blocking cMET Signaling

Sunday, Nov. 25 11:55AM - 12:05PM Room: S504CD

SSA13-09 Dynamic Cell Tracking with Time-Lapse MRI: The Temporal Window for Detection of InflammatoryDisease

Sunday, Nov. 25 12:05PM - 12:15PM Room: S504CD

correction (1.02 ± 0.876, p=0.03). The rCBV with contrast leakage correction has better correlation with FDG-PET-TNR (Correlationcoefficient =0.618, p<0.001) than rCBV without contrast leakage correction (Correlation coefficient =0.436, p=0.018), p<0.001.

CONCLUSION

The rCBV with contrast leakage correction shows better correlation with FDG-PET-TNR. Combing different MR-DSC-PWI and FDG-PET parameters could provide comprehensive information of tumor hemodynamic change and tumor metabolic abnormality.


The rCBV with contrast leakage correction shows better correlation with FDG-PET-TNR. Combing different MR-DSC-PWI and FDG-PET parameters could provide comprehensive information of tumor hemodynamic change and tumor metabolic abnormality.

ParticipantsYingwei Wu, MD, Shanghai, China (Presenter) Nothing to DiscloseQi Fan, Shanghai, China (Abstract Co-Author) Nothing to DiscloseXiaofeng Tao, Shanghai, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To determine therapeutic effect of peptide functionalized NP in restraining brain tumor growth by blocking cMET Signaling


A dendrimer-based nanoinhibitor(Den-CMBP) has been developed using MET targeted cMBP peptides conjugated on G4 dendrimer.Binding affinity of Den-CMBP and free CMBP was evaluated using Surface Plasmon Resonance(SPR) Technology. Cellular responsesincluding cell apoptosis, viability, proliferation were evaluated by treatment with Den-CMBP at various concentrations.GBM rodentmodels were developed by U87-MG cells implantation. Expression of cMET and downstream signature proteins were tested in U87-MG tumor cells and U87-induced mice models by Western blotting analysis. Mice bearing GBM tumors were treated with Den-CMBP,free CMBP and PF-04217903, a small molecular MET inhibitor, respectively. In-vivo MRI was used to assess tumor volume pre andpost treatment. Immunofluorescence staining was performed to evaluate MET immnuno-activity post treatment. Survival wascalculated for three sub-groups.

RESULTS

Compared to the free cMBP peptide(KD = 3.964 × 10-7 M), the binding affinity of the nanoinhibitor increased three-order ofmultitude to 1.316 × 10-10 M due to the multivalent effect. Nanoinhibitor efficiently blocked MET signaling with remarkably reducedlevels of phosphorylated MET and its downstream signaling proteins in GBM tumor models and U87MG cell culture. In vivo T2-weighted MRI showed significant tumor growth restraint post treatment of nanoinhibitor. The volumes of tumor treated with Den-cMBP1 were recorded as 0.019, 0.408, and 3.659 mm3 at 7, 15 and 21 days, which decreased 63.5%, 80.0% and 78.3%respectively compared to the control group. No obvious therapeutic effect was observed after administration of free cBMP peptide,nanoinhibitor demonstrated remarkable therapeutic responses similar to PF-04217903. Immunoblotting studies verified that thenanoinhibitor attenuated glioma growth by inhibiting MET downstream signaling. Median survival was extended to 35 days withtreatment of nano inhibitors.

CONCLUSION

Overall, this work developed a dendrimer based MET targeted nanoinhibitor that effectively inhibits glioma growth by blocking METdownstream signaling, which would provide an alternative therapeutic strategy for tumor therapy.


The multivalency of dendrimer based NPs would help to decrease dosage and side-effects as well.

AwardsTrainee Research Prize - Resident

ParticipantsMax Masthoff, MD, Muenster, Germany (Presenter) Nothing to DiscloseSandra Gran, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseXueli Zhang, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseLydia Wachsmuth, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseAnne Helfen, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseWalter L. Heindel, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseLydia Sorokin, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseJohannes Roth, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMichel Eisenblaetter, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMoritz Wildgruber, MD, PhD, Munster, Germany (Abstract Co-Author) Nothing to DiscloseCornelius Faber, Muenster, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

While several imaging techniques apply various approaches for cell tracking, the actual dynamic process remained concealed fornon-invasive imaging in deep tissue until recently. With the concept of time lapse MRI, real time tracking of individual cells has

become possible. Here, we investigate which velocity range of cell motility can be resolved and investigate time Lapse MRI to trackimmune cell motility in a model of multiple sclerosis.


A time lapse MRI protocol using a T2*w gradient echo sequence with a single frame scan-time of 8min12s was developed on a 9.4Tsmall-animal MRI. Movies were composed of images from 20 repetitions. Phantom scans were performed on Resovist-labelledmonocytes. In vivo scans were performed in healthy (n=6) and experimental autoimmune encephalomyelitis (EAE, n=14) mice(C57BL/6J) injected i.v. with Resovist 24h prior to MRI. Simulations were performed with a synthetic phantom reproducing theobserved contrast of one labelled cell. Motion was simulated by composing synthetic k space data with different fractions obtainedfrom different positions of the synthetic cell.

RESULTS

Phantom and in vivo scans confirmed that labelled immune cells could be tracked in the brain. Simulations showed that moving cellsup to velocities of 1µm/s were detectable. In EAE mice significantly reduced numbers of in vivo labelled immune cells were observedas compared to naïve mice (253±29, n=6 vs 31±6, n=14). In EAE mice significant differences were observed before (45±9, n=6)versus after onset (21±4, n=8) of symptoms.

CONCLUSION

Time lapse MRI proofed sensitive enough to detect "patrolling" immune cells along the endothelium. With the start of the leucocyte-adhesion cascade in inflammation, cells start "rolling" with a higher velocity, resulting in less detected cells in EAE. Thus, time lapseMRI enables for assessing immune cell dynamics non-invasively and may serve as a tool for detection or monitoring of aninflammatory response prior to onset of clinical symptoms.


Time lapse MRI may be a versatile tool for studying onset and type of innate immune response by non-invasive, real-time imaging ofdynamic immune cells in the brain.

SSA16-01 Association Between Asynchrony and Stenoses in Apparently Normal Coronary Arteries

Sunday, Nov. 25 10:45AM - 10:55AM Room: S505AB

SSA16-02 Evaluation of Role of F-18 FDG Cardiac PET and Tc-99m Sestamibi Myocardial Perfusion Imaging inAssessing the Therapeutic Benefit in Patients with Coronary Artery Disease and Left VentricularSystolic Dysfunction


SSA16

Nuclear Medicine (Chest and Cardiovascular Nuclear Imaging)

Sunday, Nov. 25 10:45AM - 12:15PM Room: S505AB

CA CH CT NM


ParticipantsUkihide Tateishi, MD,PhD, Tokyo, Japan (Moderator) Nothing to DisclosePhillip J. Koo, MD, Phoenix, AZ (Moderator) Advisory Board, Bayer AG; Advisory Board, Johnson & Johnson; ; ; ;

Sub-Events

ParticipantsAndrew Van Tosh, MD, Roslyn, NY (Abstract Co-Author) Grant, Astellas GroupDavid Cooke, Atlanta, GA (Abstract Co-Author) Royalties, Syntermed, IncChristopher J. Palestro, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseKenneth Nichols, PhD, New Hyde Park, NY (Presenter) Royalties, Syntermed, Inc;


[email protected]

PURPOSE

Left ventricular (LV) arteries are considered abnormal if stenosis > 70%, but lesser stenoses may be concerning. Our study wasundertaken to determine the % of cases in which stenoses are < 70% & perfusion images suggest apparently normal (ApNl)arteries, yet myocardial flow reserve (MFR) is abnormally low, & whether PET parameters predict magnitude of stenosis.


Data were analyzed of 105 pts evaluated by Rb-82 rest/regadenoson-stress PET/CT & arteriography, which measured % stenoses.Global ejection fractions (EFs) & regional summed stress score (SSS) & summed rest score (SRS) of relative myocardial perfusionwere assessed. Rest & stress systolic & diastolic asynchrony (Asynch) was assessed by a medical imaging physicist who visuallyscored phase histograms & phase polar maps within a coronary territory using a 5-point scale (0 = normal to 4 = markedlyasynchronous extensive territory), based on phase polar maps being out of phase from expected contraction patterns of normalpts. Absolute myocardial blood flow (MBF) was quantified from rebinned first pass dynamic transit images of the Rb-82 bolusinjection through the heart chambers, with myocardial flow reserve (MFR) computed as stress-MBF/rest-MBF. ApNl arteries weredefined as those with SRS < 4 & SSS < 4 & stenosis < 70%. Following convention, abnormal regional MFR was defined as < 2.0.

RESULTS

Among 315 arteries, 174 had undetectable stenosis, 72 ranged from 25-69% & 69 ranged from 70-100%. Among all arteries, 162were ApNl with higher MFR than the other 153 arteries (2.65±1.34 versus 1.96±1.26, p < 0.0001). Nonetheless, 35% (56/162) ofApNl arteries had abnormally low MFR < 2.0 (mean = 1.50±0.31). For all arteries, magnitude of % stenosis was most stronglyassociated with magnitude of Asynch (r = 0.50, p < 0.0001), & significantly associated with stress MBF (r = -0.25, p < 0.0001),SSS (r = 0.24, p < 0.0001), SRS (r = 0.17, p = 0.002), & MFR (r = -0.18, p = 0.002). For ApNl arteries, % stenosis was associatedwith magnitude of Asynch (r = 0.34, p < 0.0001).

CONCLUSION

In arteries that are apparently normal by relative perfusion assessment & by conventional arteriographic criteria, MFR cannonetheless by abnormally low, with stenoses < 70% associated with regional asynchrony.


It is advisable to measure regional MFR & regional asynchrony for pts with suspected CAD.

ParticipantsAnkur Pruthi, New Delhi, India (Abstract Co-Author) Nothing to DiscloseRitu Verma, New Delhi, India (Presenter) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVidur Mahajan, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseEthel S. Belho, New Delhi, India (Abstract Co-Author) Nothing to Disclose

SSA16-03 A Comparative Analysis of Myocardial Perfusion on Gated SPECT versus Coronary Atherosclerosisand Calcium Score on 64-Slice CT


Rajneesh Jain, New Delhi, India (Abstract Co-Author) Nothing to DiscloseNikhil Seniaray, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVanshika Gupta, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the therapeutic benefit with revascularization and optimal medical treatment (OMT) in patients diagnosed withhibernating myocardium on myocardial perfusion imaging (MPI) using F-18 FDG cardiac PET


59 consecutive patients (43 males, 16 females, Mean Age 60.7 ± 9.4 years) with CAD and LV systolic dysfunction who underwentmyocardial viability imaging for revascularization work-up and were diagnosed with hibernating myocardium were enrolled in thisstudy. Patients were later treated with either revascularization or OMT and were followed for a median duration of 7.7 months forassessing the therapeutic benefit. Therapeutic benefit was assessed under 3 categories (a) Improvement in functional class (b)Adverse cardiac-events and (c) Improvement in LV function and myocardial perfusion on follow-up resting 99mTc-sestamibimyocardial perfusion imaging.

RESULTS

29 patients underwent revascularization (49%) and 25 patients received OMT (42%). Five patients were lost to follow-up. Patientswere matched for baseline characteristics in both treatment arms. On follow-up, significant improvement was noted in NYHAfunctional class and CCS angina class post-revascularization. No such improvement was noted in the OMT group. The cardiac-event rate of patients in OMT group was significantly higher than that of patients in revascularization group (36% vs. 10.3 %; p =0.046). At 1 year of follow-up, event-free survival in revascularization group was significantly superior compared to OMT group(83.8% vs. 50.8%; p= 0.039). On follow-up resting MPI scan, mean improvement in LVEF in revascularization group was significantlyhigher than in OMT group (6.0% vs. 1.4%; p=0.04).

CONCLUSION

Myocardial viability imaging is a sensitive modality to identify hibernating myocardium in patients with CAD and LV dysfunction andpredicting its recovery following revascularization, thereby guiding the optimal treatment strategy for these patients.


Myocardial viability imaging should be performed prior to revascularization in patients with coronary artery disease with left-ventricular dysfunction to help predict recovery post-treatment.

ParticipantsParul Mohan, MBBS, MD, New Delhi, India (Abstract Co-Author) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Presenter) Nothing to DiscloseUpendra Kaul, MBBS, MD, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The aim of the current study was to compare the results of 64-slice CT and gated SPECT on a regional basis (per vesseldistribution territory) in patients with known or suspected CAD.


Three hundred and seventy five patients underwent both gated SPECT for myocardial perfusion imaging and 64-slice CT forcoronary calcium scoring and coronary angiography. The coronary calcium score was determined for each coronary artery. Coronaryarteries on multislice CT angiography were classified as having no CAD, insignificant stenosis (<50% luminal narrowing),significantstenosis, or total or subtotal occlusion (>90% luminal narrowing).Gated SPECT findings were classified as normal or abnormal(reversible or fixed defects) and were allocated to the territory of one of the various coronary arteries.

RESULTS

In coronary arteries with a calcium score of 10 or less, the corresponding myocardial perfusion was normal in 96 %. In coronaryarteries with extensive calcifications (score > 400), the percentage of vascular territories with normal myocardial perfusion waslower, 48%. Similarly, in most of the normal coronary arteries on 64-slice CT angiography, the corresponding myocardial perfusionwas normal on SPECT in >94%. In contrast, the percentage of normal SPECT findings was significantly lower in coronary arterieswith obstructive lesions (<57%) or with total or subtotal occlusions (<10%) (P < 0.01). Nonetheless, only 42% of vascularterritories with normal perfusion corresponded to normal coronary arteries on multislice CTangiography, whereas insignificant andsignificant stenosis were present in, respectively, 40% and 18% of corresponding coronary arteries.

CONCLUSION

Although a relationship exists between the severity of CAD on multislice CT and myocardial perfusion abnormalities on SPECT,analysis on a regional basis showed only moderate agreement between observed atherosclerosis and abnormal perfusion.Accordingly, 64-slice CT and gated SPECT provide complementary rather than competitive information, and further studies shouldaddress how these two modalities can be integrated to optimize patient management.


SSA16-04 The Association of Carotid Plaque 18F-FDG and 18F-Naf Uptake on PET Scan with SymptomaticCarotid Artery Disease: A Systematic Review and Meta-Analysis


SSA16-05 Provider Utilization Trends for Elective Myocardial Perfusion Imaging


Accordingly, 64-slice CT and gated SPECT provide complementary rather than competitive information.

ParticipantsSalama Chaker, New York, NY (Presenter) Nothing to DiscloseKhalid Al-Dasuqi, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseHediyeh Baradaran, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseMichelle Demetres, New York, NY (Abstract Co-Author) Nothing to DiscloseDiana Delgado, MS, New York, NY (Abstract Co-Author) Nothing to DiscloseAjay Gupta, MD, New York, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

We sought to investigate the ability of 18F-FDG and 18F-NaF PET imaging to identify vulnerable carotid plaques and predict strokerecurrence in the setting of recent cerebrovascular accidents by performing a systematic review.


We performed this study according to the Meta-Analysis of Observational Studies in Epidemiology (MOOSE) group and the PreferredReporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines. We performed a systematic review ofOvid MEDLINE, Ovid EMBASE, and the Cochrane Library Databases yielding a total of 4,144 unique articles for screening after de-duplication. These were screened for peer-reviewed journal articles that examined the association between carotid plaque traceruptake and recent or future ischemic events such as strokes, transient ischemic attacks and retinal artery embolisms. Screenedarticles were then adjudicated as meeting inclusion criteria by two independent readers.

RESULTS

Fourteen articles were included for subsequent analysis. Of those, 11 articles analyzed 18F-FDG uptake in recently symptomaticcarotid arteries as compared to asymptomatic carotid arteries. Two of these studies analyzed 18F-NaF uptake as well. Theremaining 3 articles investigated the risk of stroke recurrence associated with 18F-FDG uptake. The existing literature demonstratessignificant heterogeneity in the PET protocols, reported tracer uptake metrics, and thresholds for positive uptake.

CONCLUSION

Our systematic review revealed a growing body of literature supporting 18F-FDG's utility in predicting future stroke recurrence andits modest ability in discerning symptomatic from asymptomatic carotid plaques. Additional studies are needed to elucidate the roleof 18F-NaF as compared to 18F-FDG imaging. Further work is needed to define more standardized approaches for PET imageacquisition and imaging analysis in order to improve the generalizability of this technique to detect high-risk carotid plaques.


Carotid atherosclerosis is responsible for 15% of ischemic strokes. Further work is needed to investigate the utility of 18F-FDG and18F-NaF PET imaging in detecting high-risk carotid plaques.

ParticipantsSarah I. Kamel, MD, Philadelphia, PA (Presenter) Nothing to DiscloseCharles M. Intenzo, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseLaurence Parker, PhD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseDavid C. Levin, MD, Philadelphia, PA (Abstract Co-Author) Consultant, HealthHelp, LLC; Board Member, Outpatient Imaging Affiliates,LLC


[email protected]

PURPOSE

To analyze trends in performance of elective stress nuclear myocardial perfusion imaging (MPI) modalities in the Medicarepopulation.


The nationwide Medicare Part B fee-for-service databases for 2004-2016 were reviewed. CPT codes relevant to stress MPI wereselected: planar and single photon emission computed tomography (SPECT), and positron emission computed tomography (PET).The databases indicate procedure volume for each code, and these were used to calculate utilization rates per 1,000 Medicarebeneficiaries. Elective MPI exams were identified by using place-of-service codes for private offices and hospital outpatientdepartments (HOPDs). The specialty of the performing physician was determined using Medicare physician specialty codes. Becausethe Medicare Part B databases are complete population counts, sample statistics are not required.

RESULTS

Elective standard (STD) MPI (both planar imaging and SPECT) utilization peaked in 2006 at 74 studies per 1,000 beneficiaries andthen progressively decreased to 45 by 2016 (-36%). In 2004, cardiologists' share of elective STD MPI had been 79%, and thissteadily increased in subsequent years to 87% in 2016. Cardiologists perform elective STD MPI mostly in private offices where

SSA16-06 Medium and Large Vessel Vasculitis: Recognizing Patterns on FDG PET-CT


SSA16-07 Assessing the Feasibility of 18F-Naf PET/CT to Detect the Atherosclerotic Calcification of Aortic Wallin Rheumatoid Arthritis Patients


utilization peaked in 2008 at 50 studies per 1,000 and then declined to 22 in 2016 (-56%). In HOPDs, utilization by cardiologists hasincreased over the period of the study from 7 studies to 15 (+120%). Utilization in private offices and HOPDs by radiologists hasdeclined from 13 in 2004 to 6 in 2016 (-58%). Elective PET MPI, less frequently used at 3 studies per 1,000 in 2016, maintained anoverall net upward trend since 2005, and most of this growth reflected increasing use by cardiologists (90% share in 2016).

CONCLUSION

In the Medicare population, the overall use of elective STD MPI is declining, however cardiologists are performing an increasingmarket share in the outpatient setting. A shift in place-of-service has been noted with fewer studies performed in private officesand increasing numbers performed in HOPDs. PET MPI utilization, while still not widespread, has grown over the period of the study,reflecting an increasing use by cardiologists.


Cardiologists maintain an increasing share in utilization of elective standard and PET MPI.

ParticipantsNikhil Seniaray, New Delhi, India (Abstract Co-Author) Nothing to DiscloseRitu Verma, New Delhi, India (Presenter) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVidur Mahajan, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseEthel S. Belho, New Delhi, India (Abstract Co-Author) Nothing to DiscloseP. S. Gupta, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVanshika Gupta, New Delhi, India (Abstract Co-Author) Nothing to DiscloseAnkur Pruthi, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The diagnosis of medium to large-vessel vasculitis and the assessment of its activity and extent remain challenging. We assess theclinical utility of FDG PET CT in patients with suspected medium and large vessel vasculitis to evaluate the pattern and extent ofvessel involvement.


100 consecutive patients (64 males and 36 females) with suspected medium and large-vessel vasculitis were evaluated with FDGPET/CT. FDG uptake in the major vessels was visually graded using a four-point scale and quantified with standardised uptakevalues (SUV max). Patients were further sub-divided into three groups: (a) steroid-naive medium to large-vessel vasculitis (N=34,69% of total positive patients), (b) vasculitis on steroid treatment (N=15, 30.6% of total positive patients) and (c) no evidence ofvasculitis (N=51). Analysis of variance and linear regression were used to investigate the association of FDG uptake with clinicalparameters.

RESULTS

FDG-PET revealed pathological findings in 49 of 100 patients. FDG PET/CT was positive (visual uptake >2; equal to or greater thanliver) in all patients with steroid-naive medium to large-vessel vasculitis. The thoracic aorta, the carotid and the subclavian arterieswere most frequently involved. In these patients, SUVmax values were significantly higher than in the other groups.

CONCLUSION

FDG PET is a sensitive and specific imaging tool for medium and large vessel vasculitis, especially when performed in steroid naivepatients. It increases the overall diagnostic accuracy and has an impact on the clinical management in a significant proportion ofpatients.


FDG-PET should be used in diagnosis and characterisation of medium and large vessel vasculitis to determine optimal treatmentmethodologies.

ParticipantsSiavash Mehdizadeh Seraj, MD, Philadelphia, PA (Presenter) Nothing to DiscloseWilliam Y. Raynor, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseAbdullah Al-Zaghal, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DisclosePegah Jahangiri, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseMohsen Khosravi, MD, MPH, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseThomas J. Werner, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseJoshua Baker, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseAbass Alavi, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseStephen J. Hunt, MD,PhD, Philadelphia, PA (Abstract Co-Author) Consultant, Amgen Inc; Consultant, BTG International Ltd


[email protected]

PURPOSE

SSA16-08 F-18 FLT PET/CT for Therapeutic Monitoring in Patients with Cardiac Sarcoidosis: Comparison withF-18 FDG PET/CT


SSA16-09 Feasibility of Using Global Lung FDG Uptake in COPD Patients on PET/CT to Assess the CorrelationBetween Pulmonary Parenchymal Inflammation and Pulmonary Function Test Indices as well asEmphysema Severity

Sunday, Nov. 25 12:05PM - 12:15PM Room: S505AB

Rheumatoid arthritis (RA) has long been associated with increased risk for atherosclerosis. 18F-sodium fluoride (NaF) is a PET tracerthat detects calcium deposition in the early stages of atherosclerotic plaque formation. We aimed to assess whether NaF-PET/CTcan sensitively discriminate aorta calcification between RA patients and normal subjects.


Fifteen RA patients (11 men, 4 women; mean age 53.8±10.8 y, range 25-64) and fifteen healthy controls (11 men, 4 women; meanage 53.5±11.2 y, range 25-64) were included in this study. Controls were matched to patients by sex and age (±5 years). Allsubjects in this study underwent NaF-PET/CT scanning 90 minutes after NaF tracer administration. Using OsiriX software, regions ofinterest were manually drawn around the abdominal aorta wall starting superiorly with the first axial slice containing the left kidney,ending with the last slice before the aortic bifurcation. The global mean standardized uptake value (global SUVmean) was obtainedand compared between RA patients and healthy subjects. An unpaired t-test assessed the difference in means of RA group andcontrols, and a ROC analysis assessed discrimination.

RESULTS

The global SUVmean of RA patients ranged from 0.88 to 2.35, and from 0.79 to 1.47 in healthy controls. Furthermore, averageglobal SUVmean scores among RA patients was significantly greater than that of healthy controls. (1.62±0.49 and 1.04±0.16,respectively, P<0.01). ROC analysis revealed fair discrimination between the two groups (AUC = 0.77).

CONCLUSION

Our findings indicate that global assessment with NaF-PET/CT is a feasible technique to detect active vascular calcification in theabdominal aorta. Discriminant validity was observed by assessing a known co-morbidity of RA and comparing RA to non-RA. Furtherstudies are needed to validate this technique to diagnose and monitor patients at high risk for atherosclerosis.


Global assessment with NaF-PET/CT can determine the degree of active vascular calcification, which can help diagnose, monitor,and assess treatment response in atherosclerosis.

ParticipantsTakashi Norikane, Kita-gun, Japan (Presenter) Nothing to DiscloseYuka Yamamoto, MD, PhD, Kagawa, Japan (Abstract Co-Author) Nothing to DiscloseTakahisa Noma, Kita, Japan (Abstract Co-Author) Nothing to DiscloseHiroaki Dobashi, MD, Kita-gun, Japan (Abstract Co-Author) Nothing to DiscloseYoshihiro Nishiyama, MD, Kagawa, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

F-18 fluorodeoxyglucose (FDG) PET has been used in sarcoidosis including cardiac involvement for therapeutic monitoring. However,it can be challenging because it accumulates physiologically in normal myocardium. The purpose of this study was to evaluate theability of F-18 fluorothymidine (FLT) PET for therapeutic monitoring in patients with cardiac sarcoidosis, in comparison with FDG.


FLT and FDG PET/CT studies were performed before and after immunosuppressive therapy in 6 patients with newly diagnosedcardiac sarcoidosis. The patients had fasted for at least 18 h before FDG PET/CT, but were given no special dietary instructionsbefore FLT PET/CT. Uptake of FLT and FDG was examined visually and semiquantitatively using maximal standardized uptake value(SUV).

RESULTS

Before therapy, all patients had both cardiac and extra-cardiac thoracic sarcoidosis. Fifteen lesions in cardiac region and 22 lesionsin extra-cardiac region were visually detected on both FLT and FDG PET/CT. After therapy, 10 and 8 lesions in cardiac region and15 and 11 lesions in extra-cardiac region showed no increased uptake on FLT and FDG PET/CT, respectively. On after therapy FLTscan, all SUV for each lesion were lower than those on before therapy FLT scan, and the mean SUVs in cardiac and extra-cardiaclesions decreased significantly (p<0.001 and p<0.001, respectively). On after therapy FDG scan, all SUV for each lesion were alsolower than those on before therapy FDG scan, and the mean SUVs in cardiac and extra-cardiac lesions also decreased significantly(p<0.001 and p<0.001, respectively). The mean SUV reductions in cardiac and extra-cardiac lesions on FLT scan were 53% and57%, respectively. The mean SUV reductions in cardiac and extra-cardiac lesions on FDG scan were 57% and 55%, respectively. Nosignificant difference in SUV reduction was found between FLT and FDG scans.

CONCLUSION

This preliminary study indicates that FLT PET/CT, even without the usually necessary fasting, may have the potential to identifythe therapeutic response in patients with cardiac sarcoidosis as well as FDG PET/CT.


FLT PET/CT, even without the usually necessary fasting, may have the potential to identify the therapeutic response in patientswith cardiac sarcoidosis.

ParticipantsPegah Jahangiri, MD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose

Kamyar Pournazari, MD,MSc, Philadelphia, PA (Presenter) Nothing to DiscloseEsha S. Kothekar, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseAbdullah Al-Zaghal, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseSiavash Mehdizadeh Seraj, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseMohsen Khosravi, MD, MPH, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseThomas J. Werner, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseAbass Alavi, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseDrew A. Torigian, MD, MA, Philadelphia, PA (Abstract Co-Author) Co-founder, Quantitative Radiology Solutions LLC


[email protected]

PURPOSE

The purpose of this study was to determine the relationship between the degree of pulmonary parenchymal inflammation measuredfrom FDG-PET/CT with the degree of emphysema and also with PFT indices in chronic obstructive pulmonary disease (COPD)patients based on image segmentation and partial volume correction.


56 COPD patients (51 men; median age 64) who underwent 18F-fluorodeoxyglucose-positron emission tomography/computedtomography (FDG-PET/CT) were studied. Lung parenchymal volume (L), macroscopic emphysema volume (E) and non-emphysematous lung parenchyma mean attenuation (A) were measured from CT images. Uncorrected maximum standardized uptakevalue of lung (USUVmax) was measured from PET/CT images. A first level of partial volume correction was then applied to accountfor varying amounts of macroscopic emphysema (CSUVmax) followed by a second level of correction to account for the mixture ofair and lung parenchyma at the microscopic level (CCSUVmax). Correlation of fraction of emphysema(F=E/L) with USUVmax,CSUVmax, CCSUVmax were tested using Pearson correlation and linear regression statistical tests. Pearson correlation and linearregression statistical tests were applied to test the correlations of USUVmax, CSUVmax, and CCSUVmax with FEV1/FVC ratio.

RESULTS

Lung USUVmax and CSUVmax were not significantly correlated with fraction of emphysema (r=0.03, p=0.831 and r=0.18, p=0.292,respectively). However, CCSUVmax was significantly positively correlated with fraction of emphysema (r=0.47, p=0.013). LungCSUVmax and CCSUVmax were significantly negatively correlated with FEV1/FVC ratio (r=-0.49, p=0.026 and r=-0.71, p<0.001,respectively), whereas there was no significant correlation between lung USUVmax and FEV1/FVC ratio (r=-0.25, p=0.073).

CONCLUSION

These data demonstrate that the degree of pulmonary inflammation increases with the degree of emphysema severity and thatpatients with lower FEV1/FVC ratios have greater degrees of pulmonary parenchymal inflammation based on FDG-PET/CTquantitative assessment. These correlations are more statistically significant when pulmonary FDG uptake is corrected for thepartial volume effect, which shows the importance of partial volume correction for accurate quantification of lung disease severity.


Measurement of pulmonary FDG uptake on PET/CT may therefore be useful in the diagnostic and response assessment of patientswith COPD.

SSA25-01 Nomogram and Artificial Neural Network for Prognostic Performance on the Albumin-Bilirubin Gradefor HCC Undergoing TACE


SSA25-02 Predicting Patient Survival After TACE for HCC Using a Neural Network: A Promising Tool


SSA25

Vascular Interventional (Chemo-Embolization and Radio-Embolization)


IR NM OI VA


ParticipantsKevin Kim, BA, Arcadia, CA (Moderator)

Sub-Events


ParticipantsBinyan Zhong, MD,PhD, Nanjing, China (Presenter) Nothing to DiscloseCaifang Ni, MD, PhD, Suzhou, China (Abstract Co-Author) Nothing to DiscloseJianSong Ji, MD, Hangzhou, China (Abstract Co-Author) Nothing to DiscloseGuowen Yin, MD, Nanjing, China (Abstract Co-Author) Nothing to DiscloseGao-Jun Teng, MD, Nanjing, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The albumin-bilirubin (ALBI) grade is a newly raised objective liver function assessment tool as well as an emerging alternative ofthe Child-Turcotte-Pugh (CTP) grade in hepatocellular carcinoma (HCC). We aimed to construct ALBI and CTP grades basednomograms as well as develop an artificial neural network (ANN) to compare the prognostic performance of these two grades.


This multicentric retrospective study included all patients with HCC who underwent TACE monotherapy as initial treatment betweenJanuary 2008 and December 2016 at four institutions. In the training cohort, independent risk factors associated with overallsurvival (OS) were identified by univariate and multivariate Cox proportional hazards analyses. The prognostic nomograms and ANNwere then established in the training cohort and validated in the two validation cohorts.

RESULTS

Totally, 838 patients (548, 115, and 175 in the training cohort and validation cohorts 1 and 2, respectively) were included. Themedian OS was 10.4, 15.7, and 9.2 months in the training cohort and validation cohorts 1 and 2, respectively. In the trainingcohort, independent risk factors were identified as: higher Eastern Cooperative Oncology Group (ECOG) grade, portal vein tumorthrombosis (PVTT), extrahepatic spread (exclude PVTT), higher ALBI/CTP grade, a-fetoprotein level greater than 200 ng/mL,multiple tumors, and tumor size larger than 5 cm. The ALBI and CTP grades based nomograms were then established separately,and showed comparable prognostic performance when assessed externally in two independent validation cohorts (C-index invalidation cohort 1: 0.823 vs. 0.802, P =0.417; in validation cohort 2: 0.716 vs. 0.729, P =0.793). ANN showed that ALBI grade hadhigher importance on survival prediction than CTP grade.

CONCLUSION

The ALBI grade outperforms the CTP grade on survival prediction for HCC patients who undergo TACE. Considering the easyapplication, the ALBI grade should be regarded as an alternative to CTP grade.


The ALBI grade should be regarded as an alternative to CTP grade about prognostic prediction for HCC underwent TACE.

ParticipantsRoman Kloeckner, MD, Mainz, Germany (Presenter) Advisory Board, Guerbet SA; Advisory Board, Bristol-Myers Squibb Company;Advisory Board, Sirtex Medical Ltd; Speaker, Guerbet SA; Speaker, Sirtex Medical LtdAline Maehringer-Kunz, MD, Mainz, Germany (Abstract Co-Author) Nothing to DiscloseChristoph Dueber, MD, Mainz, Germany (Abstract Co-Author) Nothing to DiscloseFranziska Wagner, Mainz, Germany (Abstract Co-Author) Nothing to Disclose

SSA25-03 Experimental Study on Transarterial Administration of Bevacizumab Combined with TransarterialChemoembolization in Rats with Hepatocellular Carcinoma


Arndt Weinmann, Mainz, Germany (Abstract Co-Author) Nothing to DiscloseSandra Koch, Mainz, Germany (Abstract Co-Author) Nothing to DiscloseBettina Baessler, MD, Cologne, Germany (Abstract Co-Author) Nothing to DiscloseDaniel Pinto dos Santos, MD, Cologne, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Transarterial chemoembolization (TACE) is the standard of care for intermediate stage hepatocellular carcinoma (HCC).Nevertheless, even for experienced investigators it can be challenging to decide when to repeat and when to stop TACE-treatment.To objectify treatment decisions, several scoring systems have been developed, e.g. ART, ABCR, and SNACOR. However, thepredictive ability was only poor-moderate in several attempts of external validation. Therefore, our aim was to develop a survivalprediction model for patients with intermediate stage HCC undergoing TACE using novel machine learning algorithms.


For this retrospective analysis, we included 792 patients who underwent TACE for HCC at our tertiary referral centre between01/2005-01/2017. Clinical parameters were acquired from the institutions clinical registry and laboratory system. Radiologicalresponse was evaluated in consensus by two board certified radiologists experienced in HCC imaging. Consecutively, we built anartificial neural network, which included all parameters used by the aforementioned clinical risk scoring systems (i.e. ART-, ABCR-,and SNACOR-score): baseline BCLC stage, alpha-fetoprotein level, tumour size and number, as well as change of Child-Pugh stage,AST, and radiological response after the first TACE.

RESULTS

With ten-fold cross validation, the neural network showed a promising performance at predicting one-year survival with an areaunder the ROC curve of 0.69 and a positive predictive value of 78.6%. This outperforms other established risk stratification scoressuch as ART, ABCR, and SNACOR.

CONCLUSION

Neural networks could prove superior in predicting patient survival after TACE for HCC compared to other commonly used scoringsystems. Inclusion of additional parameters in the prediction model is likely to further increase its performance.


Once established, such novel prediction models could easily be deployed in clinical routine and help in determining optimal patientcare.

ParticipantsJun Qian, Wuhan, China (Abstract Co-Author) Nothing to DiscloseKun Qian, Wuhan, China (Presenter) Nothing to DisclosePing Han, MD, Wuhan, China (Abstract Co-Author) Nothing to DiscloseChuansheng Zheng, Wuhan, China (Abstract Co-Author) Nothing to DiscloseThomas J. Vogl, MD, PhD, Frankfurt, Germany (Abstract Co-Author) Nothing to DiscloseElsie Oppermann, Frankfurt, Germany (Abstract Co-Author) Nothing to DiscloseQi Wang, Wuhan, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the effects of transarterial administration of Bevacizumab (Avastin), an inhibitor of vascular endothelial growth factor(VEGF), combined with transarterial chemoembolization (TACE) to treat hepatocellular carcinoma in rats.


Subcapsular implantation of a solid Morris hepatoma 3924A (2 mm3) in the liver was carried out in 20 male ACI rats. Animal subjectswere assigned to group based on which treatment-drugs were injected into the hepatic artery: group A: TACE (mitomycin C +lipiodol + degradable starch microspheres) + bevacizumab; group B: TACE alone. Tumor volumes of the post-treatment (V2) andpre-treatment tumor (V1) were assessed by magnetic resonance imaging (MRI) and the mean ratio (V2/V1) was calculated.Immunohistochemical expression of MMP-9 and VEGF in the tumor was semi-quantified in all rats.

RESULTS

The rate of tumor implantation reached 100 % in all the rats receiving tumor implantation with Morris hepatoma 3924A. None of theanimals died during implantation or interventional therapy. A total of 20 individual HCC tumors were observed with unenhanced MRimaging in the livers of 20 rats (100%) before treatment. After different interventional treatments, intrahepatic metastasesdeveloped in one of the 10 rats in group B. The mean value of the volume ratios [V2 (postreatment /V1 (pretreatment] were 1.6649± 0.1255 in group A, and 3.0061 ± 0.1910 in group B, respectively. Significant differences of mean volume ratio (V2/V1) wereobserved between the two groups using an unpaired t test (P<0.0001) (Fig.1) (Fig.2). The angiogenesis of tumor was evaluatedusing anti-VEGF antibodies, and the metastasis of tumor was assessed using anti-MMP-9 antibody. MMP-9 and VEGF wereexpressed in all specimens. The immunoexpression of these proteins was confirmed by the presence of red cytoplasmic staining intumor cells. Higher expressions of MMP-9 (4.9 ± 0.199) and VEGF (4.2 ± 0.79) in hepatocellular carcinoma were observed in thegroup B (TACE alone) than the MMP-9 (1.9 ± 0.733) and VEGF (2.9 ± 0.678) in group A (TACE + Bevacizumab). Statisticalsignificance was calculated using Wilcoxon signed ranked test (both P<0.0020) (Tab.1) (Fig3).

SSA25-04 Tc-99m-MAA Lung Shunt Fraction Studies Prior to Y-90 Radioembolization Have Limited Utility inNon-Hepatocellular Carcinoma Malignancies


SSA25-05 Post-Radioembolization Lung Shunt Fraction Assessment of Yttrium-90 Microspheres with DigitalPhoton Counting PET/CT: Intra-Individual Comparison with Conventional Photomultiplier Tube-Based PET/CT


CONCLUSION

Transarterial administration of bevacizumab combined with TACE noticeably inhibit the growth of hepatic carcinoma and intrahepaticmetastases in rats.


N/A

ParticipantsMohammad Elsayed, MD, Atlanta, GA (Presenter) Nothing to DiscloseJonathan Martin, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseAlexander Dabrowiecki, MD , Tucker, GA (Abstract Co-Author) Nothing to DiscloseDaryl T. Goldman, MD , New Orleans, LA (Abstract Co-Author) Nothing to DiscloseRazan Faraj, MS, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJames T. Mcmahon, BS, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseRichard Duszak JR, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseMichal Horny, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJanice M. Newsome, MD, Alexandria, VA (Abstract Co-Author) Nothing to DiscloseZachary Bercu, MD, Atlanta, GA (Abstract Co-Author) Speaker, Terumo Medical Corporation; Grant, Coulter Translational Program,Steerable Robotic Guidewire


[email protected]

PURPOSE

Lung shunt fraction (LSF) studies using Technetium-99m macro aggregated albumin (Tc-99m-MAA) are routinely performed prior toYttrium-90 (Y90) radioembolization in all eligible patients regardless of underlying malignancy. This study evaluates the utility ofperforming Tc-99m-MAA lung shunt fraction studies in patients with hepatocellular carcinoma (HCC) compared to those with non-HCC liver tumors.


A multi-hospital retrospective analysis of all pre-Y90 Tc-99m-MAA LSF studies between November 2012 to March 2018 wasperformed. Patient records were evaluated for age, gender, LSF, and underlying malignancy. Tc-99m-MAA studies were evaluatedfor mean LSF and were compared between HCC and non-HCC cases (p=.05).

RESULTS

A total of 734 Tc-99m-MAA studies were identified among 653 patients. Among these cases 368 (50.1%) involved HCC, 112(15.3%) colorectal cancer metastatic to liver, 89 (12.1%) neuroendocrine tumor metastatic to liver, 59 (8.0%) cholangiocarcinoma,27 (3.7%) breast cancer metastatic to liver, and the remaining 79 cases (10.7%) involved other primary malignancies metastatic toliver. The mean LSF of non-HCC cases, 7.4%, was significantly lower than the mean LSF of HCC cases, 11.7% (p=0.0001). Therewas only one non-HCC case in which a Y-90 radioembolization was not pursued due to high LSF (69.3%), a case of metastaticGIST to liver, in which large scale shunting through the 8 cm mass was grossly apparent on the angiogram. This is compared to atleast 37 HCC cases (mean LSF 35.1%) in which LSF was too high to pursue radioembolization.

CONCLUSION

Tc-99m-MAA LSF is low among patients with liver malignancies that are not HCC. This study indicates that pre-Y90 Tc-99m-MAAstudies have limited utility in non-HCC liver malignancies.


Patients with non-HCC liver malignancies may only require a consolidated single procedure selective Y-90 radioembolization withoutprior Tc-99m-MAA nuclear medicine scan.

ParticipantsChadwick L. Wright, MD,PhD, Lewis Center, OH (Presenter) Nothing to DiscloseKatherine Binzel, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseEvan J. Wuthrick, MD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseEric D. Miller, MD, PhD, Columbus, OH (Abstract Co-Author) Nothing to DisclosePiotr J. Maniawski, MSc, Cleveland, OH (Abstract Co-Author) Employee, Koninklijke Philips NVMichael V. Knopp, MD, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseJun Zhang, PhD, Columbus, OH (Abstract Co-Author) Nothing to Disclose

PURPOSE

The purpose of this study is to assess the clinical feasibility and estimate 90Y microsphere lung shunt fraction (LSF) followingradioembolization using both digital photon counting PET detector (dPET) and conventional photomultiplier tube-based PET detector(cPET) technologies and compare to pre-radioembolization 99mTc macro-aggregated albumin (MAA) LSF.


SSA25-06 Hepatocellular Carcinoma Prognostication Survival Based on Scoring System Developed and Validatedon Combined Milieu Intérieur and Milieu Extérieur


In a Phase I intra-individual comparison trial, pre-radioembolization 99mTc MAA was performed in 8 patients who were then treatedwith 90Y glass microspheres for hepatic malignancies/metastases. Investigational 90Y dPET/CT (Vereos, Philips) and cPET/CT(Gemini, Philips) imaging of the lungs and liver was performed in each patient (4 - 50 hrs following radioembolization). Intra-individual comparison of PET image quality and volumetric assessment of intrahepatic radioactivity was performed by a blindedreader panel. 99mTc MAA LSF was routinely calculated using planar scintigraphy. PET estimation of 90Y LSF was performed usingMIMVista (MIM Software).

RESULTS

All patients had evaluable MAA and 90Y PET images for qualitative assessment of radioactivity distribution. Qualitatively, dPETenabled more precise localization of 90Y radioactivity when compared with cPET. Quantitatively, 90Y-treated liver volumes wereconsistently smaller with dPET than cPET (660 +/- 429 mL for dPET and 944 +/- 595 mL for cPET). There were no instances ofsignificant 90Y microspheres shunting outside of the liver or in the lungs. PET estimation of 90Y LSF was consistently andsignificantly less (dPET was 0.1 +/- 0.2 % and cPET was 0.2 +/- 0.3 %) than MAA LSF (7 +/- 5 %, P < 0.001).

CONCLUSION

There remains an unmet clinical need to improve the quantification of 90Y biodistribution following radioembolization. These resultsdemonstrate the feasibility of 90Y LSF assessment using new dPET and cPET approaches that may identify patients with smaller orlarger than anticipated LSFs.


Digital PET technologies may enable new quantitative methodologies for 90Y-dosimetry and improve our understanding of vascularshunting in various liver malignancies/metastases.

ParticipantsRehan Ali, MBBS, Forest Park, IL (Presenter) Nothing to DiscloseAhsun Riaz, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseAhmed Gabr, MD, MBBCh, Chicago, IL (Abstract Co-Author) Nothing to DiscloseRonald A. Mora, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseSamdeep Mouli, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseRobert J. Lewandowski, MD, Chicago, IL (Abstract Co-Author) Consultant, BTG International Ltd; Advisory Board, Boston ScientificCorporation; Consultant, Cook Group Incorporated; Advisory Board, ABK Biomedical Inc; Advisory Board, Accurate Medical;Consultant, C. R. Bard, Inc; Riad Salem, MD, MBA, Chicago, IL (Abstract Co-Author) Research Consultant, BTG International Ltd; Research Grant, BTGInternational Ltd; Consultant, Eisai Co, Ltd; Consultant, Exelixis, Inc; Consultant, Bristol-Myers Squibb Company; Consultant, Dove;;

PURPOSE

To explore role of routine laboratory parameters in prediction of overall survival (OS) for hepatocellular carcinoma (HCC) treatedwith transartrerial radioembolization (TARE) and develop/validate a prognostic scoring system.


With IRB approval, we selected all HCC patients who received TARE and had alpha-fetoprotein (AFP)>100 ng/dl at baseline. Routinelabs [neutrophil-lymphocyte (N/L) (inflammatory-immune response), albumin-bilirubin (ALBI) grade (liver function), and AFP (tumormarker)] were measured at baseline and at 1, 3 and 6 month post-TARE Landmarks. Univariate/multivariate analyses wereperformed to evaluate OS predictability of these parameters. Scoring systems were developed based on baseline imaging (includingPVT, metastases, ascites, portal hypertension); baseline labs and labs at respective landmark (including ALBI and AFP). Cohort wasdivided randomly into two groups: Predicting Group and Validating Group. This scoring was investigated and validated in thepredictability of HCC. Using this score, time-dependent receiver operating characteristics (ROC) were evaluated with survivaloutcomes at each landmark.

RESULTS

345/401, 238/401, and 167/401 patients had laboratory parameters available at the 1, 3, and 6 month Landmarks, respectively.ALBI score and AFP response were significant OS prognosticators at all Landmarks. Laboratory Score [ALBI+(0.3xLnAFP)] wasdeveloped to predict OS from these Landmarks in the Predicting Group and was internally validated. For developing and validatinggroups; at 1-month landmark, 1-year survival was 69% and 71%, at 3-month landmark (p<0.001), 1-year survival was 72% and66%, at 6-month landmark (p<0.001), 1-year survival was 78% and 82% respectively (p<0.001). Area under ROCs were correlativewith significant survival prognostication.

CONCLUSION

Post-therapeutic AFP response and ALBI scores are easily calculable values from routinely performed blood tests. Our proposedLaboratory Score combines post-therapeutic ALBI score with AFP response and is independent of imaging findings. Following TARE,changes in labs can predict survival earlier even in absence of apparent imaging response.


For HCC survival, imaging response is standard but it may take time. As per our study, laboratory values can also surrogate the HCCsurvival with earlier results and outcomes.

SSA25-07 Prospective Randomized FAST I Trial: Evaluation of Tumor Response of Colorectal Liver Metastasesafter Transarterial Chemoembolization with Two Different Protocols Using MRI


SSA25-08 Liver Transplantation Following Yttrium-90 Radioembolization: A Comprehensive Report of Short andLong-Term Outcomes of a 170-Patient Cohort


ParticipantsThomas J. Vogl, MD, PhD, Frankfurt, Germany (Presenter) Nothing to DiscloseChristian Marko, Frankfurt, Germany (Abstract Co-Author) Nothing to DiscloseMarcel C. Langenbach, MD, Frankfurt, Germany (Abstract Co-Author) Nothing to DiscloseMarc Harth, MD, Frankfurt am Main, Germany (Abstract Co-Author) Nothing to DiscloseNour-Eldin A. Nour-Eldin, MD,PhD, Frankfurt Am Main, Germany (Abstract Co-Author) Nothing to DiscloseRenate M. Hammerstingl, MD, Frankfurt Am Main, Germany (Abstract Co-Author) Nothing to DiscloseTatjana Gruber-Rouh, Frankfurt Am Main, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To prospectively evaluate therapy response of third-line transarterial chemoembolization (TACE) for colorectal liver metastases witheither degradable starch microspheres (DSM) or Lipiodol (cTACE) using regular and diffusion MRI.


In total, 50 patients (35 males, 15 females, mean 62 years, range 40-79) underwent TACE. They were randomly assigned into twogroups: group A receiving DSM and group B receiving Lipiodol as embolization agents. Chemotherapy consisted of a combination ofCisplatin, Irinotecan, Mitomycin. Therapy response was evaluated using MRI with diffusion imaging and unenhanced MRI sequences,which were performed before each of the three TACE cycles to obtain tumor volume and apparent diffusion coefficient (ADC). Inaddition, contrast-enhanced MRI sequences were performed before the first and after the last TACE cycle. Local tumor responsewas evaluated using the RECIST criteria and survival data were analyzed using the Kaplan-Meier estimator.

RESULTS

Evaluation using the RECIST 1.1 criteria showed partial response (PR) in 33% of cases, stable disease (SD) in 13% of cases, andprogressive disease (PD) in 53% of cases in the cTACE group while the DSM group showed 23% cases of PR, 59% cases of SD, and18% cases of PD. Over the course of the therapy, the DSM group showed a statistically significant reduction in the average tumorvolume (p=0.006). A statistically significant difference in tumor response was not found between the cTACE and DSM groups(p=0.37). Maximum ADC values after the last MRI session correlated signficantly with therapy response (p=0.005), pre-treatmentADC values, however, did not (p=0.34). Median survival in the cTACE group was 13 months and 16 months in the DSM group.

CONCLUSION

A statistically significant reduction in tumor volume was found in the DSM group. No significant difference in tumor response wasfound comparing the Lipiodol and DSM group. Maximum ADC values may be used to assess therapy response after completion of oneor more TACE cycles.


Both the Lipiodol and DSM group show similar results with greater tumor volume reduction for the DSM group.

ParticipantsAhmed Gabr, MD, MBBCh, Chicago, IL (Presenter) Nothing to DiscloseRehan Ali, MBBS, Forest Park, IL (Abstract Co-Author) Nothing to DiscloseRonald A. Mora, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseAli Al Asadi, BS, Chicago, IL (Abstract Co-Author) Nothing to DiscloseKaren Grace, RN, Chicago, IL (Abstract Co-Author) Nothing to DiscloseNadine Abouchaleh, BA, Chicago, IL (Abstract Co-Author) Nothing to DiscloseSamdeep Mouli, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseAhsun Riaz, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseRobert J. Lewandowski, MD, Chicago, IL (Abstract Co-Author) Consultant, BTG International Ltd; Advisory Board, Boston ScientificCorporation; Consultant, Cook Group Incorporated; Advisory Board, ABK Biomedical Inc; Advisory Board, Accurate Medical;Consultant, C. R. Bard, Inc; Riad Salem, MD, MBA, Chicago, IL (Abstract Co-Author) Research Consultant, BTG International Ltd; Research Grant, BTGInternational Ltd; Consultant, Eisai Co, Ltd; Consultant, Exelixis, Inc; Consultant, Bristol-Myers Squibb Company; Consultant, Dove;;


[email protected]

PURPOSE

To report short and long-term outcomes of liver transplantation (LT) for hepatocellular carcinoma (HCC) patients bridged ordownstaged by Y90 radioembolization


Between 2004 and 2017, 170 HCC patients underwent LT after receiving Y90. Patients were staged using the United Network ofOrgan Sharing (UNOS) staging system at the times of their Y90 and LT. Early post-operative outcomes were recorded. Recurrence-free survival (RFS) and overall survival (OS) were calculated using Kaplan Meier Method.

SSA25-09 Predicting Treatment Response of Primary and Secondary Lung Neoplasms to TranspulmonayChemoembolization (TPCE) and Transarterial Chemoperfusion (TACP) Using Diffusion-Weighted MRI(DWI): Value of Pretreatment Apparent Diffusion Coefficient (ADC)


RESULTS

170 patients underwent LT after Y90, with a median time to LT of 7.4 months (IQR: 4.4-10.3). Only 1 patient grade 3 albumin (<3gm/dl) toxicities and 9 patetients had grade 3 bilirubin toxicities could be attributed to Y90.138 patients were successfully bridgedto LT (maintained within Milan criteria) while 11 patients were down-staged to Milan criteria. 12 (7%) developed post-operativecomplications. Three-month mortality rate after LT was 6/170 (3.5%). 75 (44%), 49 (29%) and 46 (27%) patients showedcomplete, extensive and partial tumor necrosis on explants pathologic assessment. Three, five and ten-year OS rates were 86%,80%, and 56% respectively. 20 patients developed recurrence, with 3, 5 and 10-year RFS rates of 79%, 67% and 40%. Median RFSwas 119 (95%CI: 68-119) months. There were no significant differences in OS or RFS for bridged or downstaged patients. Therewas significant difference in terms of number of recurrences and RFS between patients who had complete or extensive necrosis vspatients who had partial necrosis (p<0.0001).

CONCLUSION

Y90 can be used as a locoregional therapy that permits bridging or downtaging to LT. LT after Y90 represents a curative treatmentfor HCC with excellent OS rates comparable to LT for non-malignant causes. Patients who achieved extensive or complete necrosishave better recurrence free survival outcomes.


Y90 has proven to be not only beneficial for HCC patients with bridging or downstaging to LT but also with good tumor responsepre-LT and prolonged recurrence free survival post-LT.

ParticipantsThomas J. Vogl, MD, PhD, Frankfurt, Germany (Presenter) Nothing to DiscloseA. T. Hoppe, Frankfurt, Germany (Abstract Co-Author) Nothing to DiscloseTatjana Gruber-Rouh, Frankfurt Am Main, Germany (Abstract Co-Author) Nothing to DiscloseNour-Eldin A. Nour-Eldin, MD,PhD, Frankfurt Am Main, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To examine predictive value of apparent diffusion coefficients (ADC) derived from diffusion-weighted MR imaging (DWI) in patientswith unresectable primary and secondary lung cancer undergoing transpulmonary chemoembolization (TPCE) and transarterialchemoperfusion (TACP) treatment.


31 patients with 42 lesions (non-small cell lung cancer (NSCLC) n=13; and lung metastases n=18) underwent examination with DWIprior to first treatment and one month thereafter. Lesion diameter and volume were measured at the beginning and end of eachtreatment cycle (mean 3.1 procedures per patient) in about 4-week intervals between January 2014 and December 2017.Decreases by at least 30% regarding tumor volume were classified as partial response (PR), an increase by 20% or more wasdefined as progressive disease (PD), while remaining lesions were categorized as stable disease (SD).

RESULTS

The PR group contained 9 lesions (NSCLC n=8; metastasis n=1), mean pretreatment ADC was 1.164 x 10-3 mm2/s, increase inmean ADC after first intervention was 32.9%. The PD group had 14 lesions (NSCLC n=1; metastases n=13), mean pretreatment ADCwas 1.418 x 10-3 mm2/s, increase in mean ADC was 5.0%. Difference between ADC changes in the response groups was significantp<0.01). Lesion pretreatment ADC recorded fair diagnostic value for predicting response (AUC 0.774). Applying a threshold ADCincrease of 20.71%, response can be predicted with a sensitivity and specificity of 88% and 78%, respectively (AUC 0.838). Inprimary lung cancer lesions, correlation of ADC changes with changes in tumor diameter and volume of -0.87 and -0.66,respectively. In metastatic lesions, correlation coefficients were -0.18 and -0.35.

CONCLUSION

A correlation was documented between early increases in ADC and tumor size reduction. Furthermore, responding lesions showedlower pretreatment ADC. DWI seems a suitable method for response prediction. The findings correlated better with primary lungcancer than with lung metastases.


DWI provides the interventionalist with information going beyond morphological tumor aspects and response prediction assists inprioritising treatment locations in patients with multiple lesions

NM200-SD-SUA1

Evaluation of a Fast Protocol for Staging Patients with Bronchial Cancer Using PET/MRI

Station #1

NM201-SD-SUA2

Value of Quantitative Parameters and Metabolic Heterogeneity of 18F-FDG PET/CT to PredictPrognosis in Patients with Primary Oropharyngeal Carcinoma

Station #2

NMS-SUA

Nuclear Medicine Sunday Poster Discussions

Sunday, Nov. 25 12:30PM - 1:00PM Room: NM Community, Learning Center

NM



ParticipantsPhillip J. Koo, MD, Phoenix, AZ (Moderator) Advisory Board, Bayer AG; Advisory Board, Johnson & Johnson; ; ; ; Ukihide Tateishi, MD,PhD, Tokyo, Japan (Moderator) Nothing to Disclose

Sub-Events

ParticipantsOle Martin, Duesseldorf, Germany (Presenter) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SALale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AGGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the applicability of a fast MR-protocol for whole-body staging of bronchial carcinoma patients using an integratedPET/MR system.


In this prospective study, 52 patients (16 female, 36 male, mean age 63.9±8.7y) underwent clinically indicated PET/CT andsubsequent PET/MRI. For PET/MRI imaging, a fast whole-body MR-protocol was implemented. MRI and PET/MRI datasets wereanalyzed to identify malignant manifestations. The accuracy for the identification of malignant manifestations was calculated andthe tumor stage for each examination was determined using the actual TNM classification from UICC 8 (union international contre lecancer). Radiation doses derived from administered tracer activities and CT protocol parameters were estimated and the mean scanduration of PET/CT and PET/MRI imaging were determined. In 26 patients, all available histopathological samples as well as resultsof follow-up imaging were used for reference standard. In the other cases, the results of PET/CT imaging used for referencestandard.

RESULTS

Active malignant lesions were present in 47/52 examinations. Both MRI and PET/MRI have a 100% accuracy to correctly identify theextent of the primary (T-stage). In nodal staging, PET/MRI revealed higher values of diagnostic accuracy than MRI alone (93.5%for PET/MRI, 83.9% for MRI, p<0.05). However, the results did not differ significantly. In identification of metastases, PET/MRIshowed higher diagnostic accuracy than MRI alone (90.3% for PET/MRI vs. 77.4% for MRI, p<0.05). In TNM classification, PET/MRIenabled higher values of correct stage than MRI (86.9% for PET/MRI vs. 64.5% for MRI, p<0.05). The values of correct stage didnot differ significantly between PET/CT and PET/MRI (93.5% for PET/CT vs. 86.9% for PET/MRI). Average scan duration of whole-body PET/CT and PET/MRI examinations were 17.5±2.3 min and 29.4±3.9 min, respectively. Estimated mean effective-dose forPET/CT scans were 61.3% higher than for PET/MRI.

CONCLUSION

Using 18F-FDG PET data in addition to whole-body MRI leads to a more accurate evaluation of patients with bronchial cancer. Withregard to patient comfort related to scan duration and reduced radiation exposure, fast PET/MRI may serve as a powerfulalternative to PET/CT for a diagnostic workup of bronchial cancer patients.


In staging lung cancer patients, whole body PET/MRI may serve as a powerful alternative to PET/CT

ParticipantsKeiichiro Tahara, Fukuoka, Japan (Presenter) Nothing to Disclose

NM202-SD-SUA3

Role of Tri-PET in Managing Indolent Lymphomas

Station #3

Shingo Baba, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseTakuro Isoda, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseYoshiyuki Kitamura, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseRyo Somehara, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseMasayuki Sasaki, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseHiroshi Honda, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The aim of this study was to evaluate the impact of quantitative FDG PET/CT imaging parameters and intratumoral metabolicheterogeneity for predicting patient outcomes in primary oropharyngeal cancer.


We retrospectively investigated 30 patients with oropharyngeal carcinoma with ipsilateral LN metastasis. SUVmax and metabolictumor volume (MTV) and Total lesion glycolysis (TLG) were measured for the primary tumors and metastatic lymph nodes. Primarytumor intratumoral metabolic heterogeneity was calculated as the CV (coefficient of variation) of FDG uptake and the area under acumulative SUV volume histograms curve (AUC-CSH). The median follow-up time was 35.4 months (range, 3-120 months). Eachparameters were compared between the good prognosis group and the poor prognosis group.

RESULTS

Of the 30 patients included, 8 patients relapsed and 7 deceased during the study period. No significant difference was seen inSUVmax and MTV between two groups. TLG of primary tumor was significantly higher in poor prognosis group compared with goodprognosis group (334±86 vs 199±53). Week correlation was observed between AUC-CSH indexes and prognosis but no statisticallysignificant difference was found.As for lymph node metastasis, TLG of poor prognosis group was significantly higher than those ofpoor prognosis group (264±86 vs 135±53). In addition, CV of FDG uptake in lymph nodes was significantly higher in poor prognosisgroup than in good group (0.51±0.11 vs 0.41±0.10).

CONCLUSION

In addition to TLG, metabolic heterogeneity of Lymph node metastasis was a prognostic factor for the patients with oropharyngealcancer. The combined predictive effect of metabolic tumor burden, and heterogeneity provided prognostic survival information inthese patients.


Metabolic heterogeneity of Lymph node metastasis was a prognostic factor for the patients with oropharyngeal cancer.

ParticipantsShanker Raja, MD, Bellaire, TX (Presenter) Nothing to DiscloseMusab M. Ahmed, MBBS, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseShahid Iqbal, MBBS,MRCP, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseSharad P. George, MD, Dhahran, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseSyed Z. Zaidi, MD, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseBelal M. Albtoosh, BSN, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseAtta M. Gill, MD, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseAhmad A. Butt, MD, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to DiscloseNawal F. AlShehry, MD, Riyadh, Saudi Arabia (Abstract Co-Author) Nothing to Disclose

PURPOSE

Indolent lymphomas (ind-LYM) are a subset of B-cell lymphomas, characterized by slow growth, protracted, and a tendency toreoccur. The management of ind-LYM is varied and evolving. Reliable tools for prognostication and treatment strategies are limited.We explored the utility of FDG tri-PET (baseline (BAS), interim/6 months (INT), and end of therapy/1 yr (EOT)), in managing ind-LYM.


Retrospective analysis of all pts. diagnosed as ind-LYM and having undergone tri-PET from 2015 to current (no 40;male=2,female=13, mean age at diagnosis= 59.68 ± 14.5. All PET scans were obtained per accepted protocols. SUVmax Deauvillescores (DS) were obtained from five target lesions, the average (cSUV & cDS) were computed for each pt. to obtain compositescores. Statistical analyses were performed with the cSUV and cDS. The following statistics were performed (T-test, mean deltachange (DELch) between each time point (BS, INT, EOT), for both cSUV (cSUV-DELch and cDC(cDC-DELch.

RESULTS

The types of ind-LYM types were CLL/SLL and Follicular, 17 and 15 respectively, 3 each of mantle cell and marginal zone lymphoma, and 1 each MALT and lymphoplasmocytic lymphoma. The data were analyzed for two arms of the study i.e treatment(chemotherapy arm (CHEMO-22/40)arm, and Rituximab+observation arm (OBSER-18/40). At BAS, pts. on CHEMO had significantlyhigher cSUV and cDC compared to those who received Rituximab only (cSUV: 6.99 vs. 4.32 p-value 0.0124) (cDC: 4.37 vs 3.71 p-value 0.0075), respectively. Although there was a DELch in the cSUV and cDC in both arms from INT to EOT, both treatment armsshowed no significant difference in the cSUV and cDC (p-value: 0.754 and 0.572, respectively). A 2-tailed t-test was for DELch ofcSUV BAS to INT after removal of an outlier was statistically significant p-value 0.05; while it was not significant for BAS to EOT(0.98) nor for any of the cDC-DELch..

CONCLUSION

The higher cSUV/cDC at BAS for the treatment arm may be due to disease profile/ selection bias; at EOT no difference in bothestimates were noted. Both arms showed similar trends in DELch from BS-INT-EOT. Our findings suggest that SUV and DC at BAS

NM002-EB-SUA

The Renaissance of Thallium SPECT in Neuro-Oncology: A Pictorial Review of Radiation Necrosisversus Tumour Recurrence

Hardcopy Backboard

may be independent criteria for treatment selection in ind-LYM.


Our findings on tri-PET suggest that metabolic profiles are helpful in selection of treatment strategies at baseline. compositeSUVmax is is potentially a good prognosticator.

ParticipantsLaura Preston, BMBCh,FRCR, Glasgow, United Kingdom (Presenter) Nothing to DiscloseAlice Nicol, PhD,MSc, Glasgow, United Kingdom (Abstract Co-Author) Nothing to DiscloseRavi V. Jampana, MBBS, MRCP, Glasgow, United Kingdom (Abstract Co-Author) Nothing to DiscloseCelestine Santosh, BMBCh, FRCR, Glasgow, United Kingdom (Abstract Co-Author) Nothing to DiscloseNatasha E. Fullerton, MD, PhD, Glasgow, United Kingdom (Abstract Co-Author) Nothing to Disclose


[email protected];[email protected]

TEACHING POINTS

Differentiation between radiation necrosis and tumour is an increasing challenge in neuro-oncology imaging due to a greater varietyof treatment options and longer patient survival. There is large overlap in appearances on CT/MRI with no reliable imaging finding toenable confident discrimination. Thallium-201 SPECT has shown potential as a tool to help solve this dilemma. Whilst its use fortumour grading has decreased in our institution over the last 10 years, its emerging role in radionecrosis has resulted in aresurgence in its use. The aims of this exhibit are: 1) To review the MRI features of radionecrosis in a variety of scenarios and toillustrate the challenge in differentiation from tumour progression/recurrence. 2) To illustrate the use of thallium-SPECT indifferentiating radionecrosis from tumour, using experience from our institution.


Overview of thallium SPECT. •Technical aspects (including fusion) •Variety of uses in neuro-oncology Aetiology/ clinical features ofradionecrosis. Pictorial review. •Cases to illustrate the shortfalls of structural imaging for radionecrosis versus tumour, and thecomplementary use of thallium SPECT. Sensitivity and specificity of Thallium-SPECT for distinguishing radionecrosis from tumour•Experience from our institution versus wider literature.

RC105A Advanced Hippocampal Neuroimaging in Alzheimer's Disease

RC105B Non-Alzheimer's Causes of Dementia: Focus on CJD and NPH

RC105C Neuronuclear Imaging in Alzheimer's with FDG and Amyloid PET

RC105D The Clinical Classifications, Diagnostic Dilemmas, and the Impetus for Imaging Biomarkers inDementia

RC105

Update on Imaging in Dementia

Sunday, Nov. 25 2:00PM - 3:30PM Room: E451A

MR NR NM


ParticipantsCyrus Raji, MD,PhD, St. Louis, MO (Moderator) Consultant, Brainreader ApSJody L. Tanabe, MD, Aurora, CO (Moderator) Nothing to Disclose


[email protected]

[email protected]

LEARNING OBJECTIVES

1) Enhance radiologist understanding in current clinical applications of structural and functional hippocampal imaging in Alzheimer'sdisease. 2) Overview MR imaging findings in non-Alzheimer's causes of dementia such as Creutzfeldt-Jakob disease and normalpressure hydrocephalus. 3) Update radiologists about neuronuclear techniques in Alzheimer's disease such as FDG PET and amyloidPET. 4) Overview correlative structural and functional neuroimaging findings in the neuropsychological characterization of dementiawith a focus on non-Alzheimer's dementia such as frontal temporal dementia.

Sub-Events

ParticipantsMichael M. Zeineh, PhD, MD, Stanford, CA (Presenter) Research funded, General Electric Company; Consultant, Biogen Idec Inc

LEARNING OBJECTIVES

1) Describe the benefits of advanced forms of imaging (e.g. 7T MRI, DTI, fMRI). 2) Identify major issues around neuroimaging in AD.C) Appraise potential benefits of advanced neuroimaging in AD.

ParticipantsLeo P. Sugrue, MD, PhD, San Francisco, CA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) To use clinical history to narrow the dementia differential. 2) To identify imaging findings in CJD. 3) To recognize CJD mimics. 4)To identify imaging findings in NPH. 5) To explain the difference between NPH and communicating hydrocephalus.

ParticipantsJames M. Mountz, MD, PhD, Pittsburgh, PA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) To describe the physiologic characteristics of F-18 FDG and Amyloid imaging binding agents as it's related to imaging findings indementia. 2) To explain the imaging methods and scan findings that are obtained in normal and Alzheimer's disease patients ascompared to normal controls. 3) To show imaging characteristics in other dementias that have cognitive symptomatology which aresimilar to those of patients with Alzheimer's disease.

ParticipantsJohn L. Ulmer, MD, Milwaukee, WI (Presenter) Stockholder, Prism Clinical Imaging, Inc; Medical Advisory Board, General ElectricCompany


[email protected]

LEARNING OBJECTIVES

1) To familiarize with current classification of Alzheimer's Disease and Frontotemporal Lobe Degeneration. 2) Come to know thechallenges faced by Neurologist and Neuropsychologist in diagnosing dementias. 3) Understand potential for imaging biomarkers insupporting diagnosis of Alzheimer's Disease and Frontotemporal lobe dementias.

Active Handout:John L. Ulmer

http://abstract.rsna.org/uploads/2018/18002521/Clinical Classifications, Diagnostic Dilemmas RC105D.pdf

RC111A New Guidelines for I-131 Therapy of Thyroid Cancer

RC111B Lu177-DOTATATE Therapy for Neuroendocrine Tumors

RC111C Hepatic Artery Infusion Therapy with Y90 Microspheres

RC111

Update on Radionuclide Therapies


IR NM OI



Sub-Events

ParticipantsDon C. Yoo, MD, E Greenwich, RI (Presenter) Consultant, Endocyte, Inc


[email protected]

LEARNING OBJECTIVES

1) Describe why thyroid cancer is increasing. 2) Review guidelines for the use of I-131 in the treatment of thyroid cancer. 3)Review the controversies in thyroid cancer treatment.

ABSTRACT

The purpose of this educational activity is to review the reasons why the incidence of thyroid cancer has risen so rapidly over thelast 40 years and discuss the role of radioiodine ablation in patients with thyroid cancer. Issues that will be discussed includecontroversies in the extent of thyroid surgery and the appropriate use of radioiodine ablation in patients with thyroid cancer whichis controversial in low risk and intermediate risk patients. The incidence of thyroid cancer in the United States has almost tripledsince the early 1970s with unchanged mortality principally due to overdiagnosis. The extent of surgery performed for thyroid canceris controversial especially in small cancers but only patients with complete thyroidectomy are candidates for radioiodine ablation.Recently lower doses of I-131 have been shown to be effective for radioiodine ablation of remnant thyroid tissue afterthyroidectomy. High risk patients will benefit from radioiodine ablation with decreased recurrence and improved mortality.Radioiodine ablation in low risk patients is very controversial and has not been shown to improve mortality.

ParticipantsSalvador Borges-Neto, MD, Durham, NC (Presenter) Grant, General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Learn the objectives and indications of the Lu-177 DOTATATE treatment. 2) Learn the short term and long term side effects ofthe Lu-177 treatment. 3) Learn how to set up this treatment modality at one's center.

ParticipantsCharles Y. Kim, MD, Durham, NC (Presenter) Consultant, Merit Medical Systems, Inc; Consultant, Cook Group Incorporated

LEARNING OBJECTIVES

1) Review range of malignancies treated with Y90 microsphere infusion. 2) Discuss the types of Y90 therapy and dosimetricconsiderations. 3) Describe the procedures and technical steps involved in Y90 therapy. 4) Recognize pertinent scintigraphicfindings associated with Y90 therapy.

ABSTRACT

Intra-arterial Yttrium-90 (Y90) therapy is an important treatment modality for a variety of hepatic tumors. While numerous types ofembolotherapies are employed by interventional radiologists for treatment of cancer, Y90 therapy is unique in its multimodality andmulti-procedural nature. Not only does this treatment effect rely on deposited ionizing radiation therapy, but scintigraphic imaging isalso an integral component of treatment. Two types of Y90 therapies are available, made by two different manufacturers. Thedifferences between the two types are subtle, but there are differences in administration and manufacturer-recommendeddosimetric calculation. These various differences will be highlighted. Y90 therapy is comprised of several steps and is frequentlysubclassified into a 'planning' phase and 'treatment' phase. In the planning phase, detailed angiographic imaging is performed todelineate arterial anatomy , determine tumoral distributions, and redistribute vascular flow if indicated. Scintigraphic imaging is anintegral component of this planning phase, in order to help identify angiographically occult arterial anomalies, confirm appropriateinfusion site, and to quantify the hepatopulmonary shunt fraction. From this information, as well as other factors, the appropriatetreatment doses can be determined. In the treatment phase(s), the Y90 dose is administered to the appropriate portions of the

liver with subsequent scintigraphic imaging for confirmation.

ED010-MO

Nuclear Medicine Monday Case of the Day

Monday, Nov. 26 7:00AM - 11:59PM Room: Case of Day, Learning Center



TEACHING POINTS


MSMI21A Molecular Imaging Using Radioactive Tracers

MSMI21B Molecular Imaging with MRI and MRS

MSMI21C Nanoparticles

MSMI21

Molecular Imaging Symposium: Basics of Molecular Imaging

Monday, Nov. 26 8:30AM - 10:00AM Room: S405AB

BQ MR MI NM US



ParticipantsZaver M. Bhujwalla, PhD, Baltimore, MD (Moderator) Nothing to DiscloseJan Grimm, MD, PhD, New York, NY (Moderator) Investor, NortisBio

Sub-Events

ParticipantsJan Grimm, MD, PhD, New York, NY (Presenter) Investor, NortisBio

LEARNING OBJECTIVES

1) Discuss the various radio tracers and their applications in Molecular Imaging studies. 2) Understand in which situations to usewhich radio tracers, what to consider when developing the imaging construct and what controls to obtain for nuclear imagingstudies. 3) Examples will contain imaging with small molecules, with antibodies and nanoparticles as well as with cells in order toprovide the participants with examples how o correctly perform their imaging studies. 4) Most of the examples will be from theoncology field but their underlying principles are universally applicable to other areas as well.

ParticipantsZaver M. Bhujwalla, PhD, Baltimore, MD (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) To list the basic principles of magnetic resonance (MR) molecular imaging. 2) To describe the uses of noninvasive multi-nuclearMRI and magnetic resonance spectroscopic imaging (MRSI) for molecular imaging applications that provide spatial and temporalinformation on vasculature, metabolism and physiology. 3) To identify the applications of targeted MR contrast agents to detectreceptor and gene expression. 4) To describe strategies that combine detection with therapy for theranostic imaging and formetabolotheranostics. 5) To provide examples of translational applications of molecular imaging and theranostics.

ABSTRACT

Noninvasive multi-nuclear magnetic resonance (MR) imaging and spectroscopic imaging (MRSI) provide a wealth of spatial andtemporal information on vasculature, metabolism and physiology. Novel targeted contrast agents have widened the scope of MRtechniques for molecular imaging applications to detect receptor and gene expression. In cancer, molecular imaging can be appliedto identify targets specific to cancer with imaging, design agents against these targets to visualize their delivery, and monitorresponse to treatment, with the overall purpose of minimizing collateral damage. Genomic and proteomic profiling can provide anextensive 'fingerprint' of each tumor. With this cancer fingerprint, theranostic agents can be designed to personalize treatment forprecision medicine of cancer, and minimize damage to normal tissue.

ParticipantsHeike E. Daldrup-Link, MD, Palo Alto, CA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) To understand important safety aspects of ultrasmall superparamagnetic iron oxide nanoparticles (USPIO). 2) To understand thebiodistribution of ferumoxytol nanoparticles and implications for imaging diagnoses. 3) To recognize the value of ferumoxytolnanoparticles for cancer MR imaging and PET/MR imaging.

ABSTRACT

Gadolinium chelates as contrast agents for MRI have been associated with mounting concerns about nephrogenic sclerosis andgadolinium deposition in the brain. Therefore, a search for safe alternatives is currently underway. In North America, the ironsupplement ferumoxytol has gained considerable interest as an MR contrast agent. In Europe, ferumoxtran-10 is re-entering clinicaltrials. Both ferumoxytol and ferumoxtran-10 provide long-lasting blood pool enhancement, which can be used for MR imaging examsthat require detailed and/or long-lasting vessel delineation for MR angiographies, tissue perfusion studies, and whole body tumorstaging. Iron oxide nanoparticles are slowly phagocytosed by macrophages in the reticuloendothelial system, making them ideal for

MSMI21D Molecular Imaging with Ultrasound

MSMI21E Quantitative Imaging Biomarkers

MR imaging detection of tumors in the liver, spleen, lymph nodes, and bone marrow. Similarly, iron oxide nanoparticles are slowlyphagocytosed by tumor-associated macrophages in cancers; which can be used to grade tumor-associated inflammation andmonitor the efficacy of new cancer immunotherapies. This presentation provides an introduction to the use of iron oxidenanoparticles for clinical MR and PET/MR imaging, including safety data acquired in children thus far, recent insights andmechanisms of rare, but potentially severe adverse reactions, applications that impact patient care and comparisons withgadolinium chelates. New developments for image guided therapy and theranostics are under way.

ParticipantsAlexander L. Klibanov, PhD, Charlottesville, VA (Presenter) Co-founder, Targeson, Inc; Shareholder, Targeson, Inc; Institutionalresearch collaboration, AstraZeneca PLC; Contract, SoundPipe Therapeutics;


[email protected]

LEARNING OBJECTIVES

1) Understand the principles of microbubble design-how to prepare fully biocompatible and safe ultrasound contrast agent particlesthat are clinically translatable, stable on storage, provide strong acoustic response and high sensitivity of detection by clinicalultrasound imaging systems, and could be targetable. 2) Understand the principles of selection of disease-specific targeting ligandsusable for contrast ultrasound imaging, based on receptor levels in the vasculature in the disease issues, as well as vascularbiomechanics. 3) Assess the results of early stage clinical trials performed with targeted microbubbles, and opportunities for clinicaltranslation in diagnostic imaging and image-guided interventions.

ABSTRACT

Ultrasound is the most widespread clinical imaging modality. Therefore, enabling molecular imaging potential in an ultrasound settingwill lead to the expanded and improved clinical diagnostic benefit. Ultrasound contrast microbubbles are already used in clinic asblood pool contrast agents, with excellent detection sensitivity: single particles with sub-picogram mass can be observed withclinical imaging systems in real time, at a depth of several cm. To achieve biomarker-selective molecular imaging, microbubble shellsurface is decorated with targeting ligand molecules (antibodies, peptides, carbohydrates) that assure selective binding andretention in the areas of disease. Clinical microbubbles are typically 1-3 um in diameter; they do not extravasate, so targetbiomarker receptors should be located on the luminal surface of vessel wall, e.g., vascular endothelium. Microbubbles are targetedto the biomarkers in the areas of inflammation and ischemia-reperfusion injury (P- and E-selectin, VCAM-1, ICAM-1) or to tumorneovasculature (VEGFR2). The latter, a heterodimeric peptide-targeted contrast microbubble from industry, has successfullycompleted Phase 1-2 clinical trials for imaging of ovarian, breast and prostate cancer lesions. Overall, targeted microbubblesempower molecular ultrasound imaging; they could also be used in conjunction with image-guided interventions, such as targetedbiopsy and therapy.

ParticipantsRobert J. Gillies, PhD, Tampa, FL (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Describe the differences between diagnostic, prognostic, and predictive biomarkers. 2) Describe the analytic pipeline forconventional radiomics. 3) Describe the biological basis and the practice of "habitat imaging" and why it may be useful for predictionand monitoring. 4) Briefly describe the methods and utility of deep learning in radiomics.

ABSTRACT

Quantitative Imaging: RadiomicsThe practice of radiology is undergoing revolutionary changes from describing images using asemantic lexicon to one that is increasingly quantitative and data driven, enhanced by machine learning algorithms. The discipline ofradiomics grew out CAD systems and has been developing over the last decades to extract quantitative image features and analyzethem with an eye to generating diagnostic, prognostic, and predictive models. In conventional radiomics, volumes of interest (VOI)are identified and segmented. In cancer, the VOIs can be the tumor, or the peri-tumor parenchyma. Within these VOI features areextracted describing size, shape, location, texture, and patterns. More than 1,000 of these features are commonly generated fromeach VOI and this has to be reduced to an smaller informative set using correlation analyses, test-retest data, and then classicalstatistical methods such as CART, LASSO, or ROC analyses. In the training phase, these can be cross validated by holding out asmall sample (1/10 to 1/2) and training on the remaining and testing on the held samples to identify the best multi-variate models.These are then locked down and tested against a completely independent test set, preferably from another institution. Well-powered studies exceeding 1,000 patients have been published to generate strongly prognostic, diagnostic, and predictive models.An extension of radiomics combines data from multiple imaging modalities (PET+CT, multiparametric MRI) to generate "hypervoxels"that can be clustered using e.g. fuzzy C-means clustering, Gaussian mixture models, Otsu thresholding, identifying spatiallycontiguous regions with similar phenotypes. These are referred to as "habitats" which can be highly predictive of responses tospecific therapeutics, and which can be longitudinally monitored for adaptive therapy dosing. Importantly, habitat imaging obviatesthe need for explicit segmentation. A further extension of radiomics is the increased use of "deep learning", usually withconvolutional neural networks, CNN. The depth of learning is limited by the size of the training set, and many methods are employedto augment the training data by rotating the original data set to generate multiple representations. The depth of the neuralnetwork can also be increased trough the use of transfer learning, wherein a network trained on similar data can be used to formthe initial layers, and only the final layers are specific to the training data at hand. While CNNs are commonly disparaged becausethey are "black boxes", newer methods are being deployed that can identify the region of the image that contains the mostimportant diagnostic/prognostic information, making the black box a little more transparent.

RC205-01 Multimodal Molecular Imaging Using Advanced MRI and PET: Applications in Clinical Neuro-Oncology

Monday, Nov. 26 8:30AM - 9:00AM Room: S406B

RC205-02 Robust Pre-Operative Language Mapping in Patients with Brain Tumors: A Feasibility Study


RC205

Neuroradiology Series: Brain Tumors

Monday, Nov. 26 8:30AM - 12:00PM Room: S406B

AI MR MI NR NM OI



ParticipantsSoonmee Cha, MD, San Francisco, CA (Moderator) Nothing to DiscloseKei Yamada, MD, Kyoto, Japan (Moderator) Nothing to Disclose

Sub-Events

ParticipantsNorbert Galldiks, Cologne, Germany (Presenter) Research grant, Wilhelm-Sander Stiftung (Munich, Germany); Advisory board, Abbvie

LEARNING OBJECTIVES

1) To give an overview on the most relevant advanced MRI techniques (i.e., PWI MRI, 2-hydroxyglutatate MRS) and PET tracers(i.e., amino acid PET tracers, radiolabeled somatostatin receptor ligands) to improve diagnostics in the field of clinical Neuro-Oncology.

ParticipantsMohammad Fakhri, MD, Saint Louis, MO (Presenter) Nothing to DiscloseManu S. Goyal, MD, MSc, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseJoshua S. Shimony, MD, PhD, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseCarl Hacker, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseAmrita Hari-Raj, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseAbraham Z. Snyder, PhD, Saint Louis, MO (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

We evaluate multilevel perceptron (MLP)-based mapping as a tool for identification of language-related resting-state networks inbrain tumor patients. Currently available alternative resting state fMRI analysis methods are either biased (e.g., seed-basedcorrelation) or not robust at the single subject level (e.g., Independent Component Analysis), and therefore, not ideal for use inpresurgical planning.


Twenty-one patients with a brain tumor in the vicinity of expressive language areas were included (mean age 42 ±16 years; 71%male). The MLP output was compared to seed-based correlation in two different manually defined language regions of nterst (ROIs).The putative language ROIs were defined using meta-analysis of task-fMRI responses, resting state seed based correlation maps,and direct cortical stimulation language maps. MLP performance in patients was also compared to a cohort of 688 normal subjects.

RESULTS

Upon presentation, 62% of the patients exhibited expressive aphasia prior to the surgical resection. Thirty-two percent of thepatients were positive for IDH-1 mutation and 27% had 1p/11q deletion. The MLP was able to reliably map robust language RSNaffiliation in putative language areas in all patients (n=21, 100%). Results were similar to those obtained in a cohort of young,healthy subjects. Fisher z-transformed Pearson correlation maps obtained from seed ROIs were strongly spatially correlated withthe MLP score in both evaluation ROIs (Spearman rho=0.74 and 0.62, p<0.0001 and p<0.003 respectively).

CONCLUSION

MLP-based language maps are comparable to results obtained using conventional seed-based correlation mapping. MLP-basedmapping is reliable in patients with brain tumors. The trained MLP is robust to anatomical shifts owing to mass effects and focal,tumor-related neural dysfunction, hence, is suitable for use in patients with brain tumors.


A trained machine-learning algorithm can reliably identify resting-state language-related networks on an individual basis in patients

RC205-03 Use of Quantitative Blood Oxygen Level Dependent (qBOLD) in Non-Invasively Determining GliomaGrade, with Correlation through Neuropathology


RC205-04 Probabilistic Atlases of Pre-Treatment MRI Reveal Hemispheric and Lobe-Specific Spatial Distributionsacross Molecular Sub-Types of Diffuse Gliomas


with brain tumors in vicinity of the language cortex.

ParticipantsPejman Jabehdar Maralani, MD, FRCPC, Toronto, ON (Presenter) Nothing to DiscloseJulia Keith, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseDavid Munoz, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseTodd Mainprize, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseArjun Sahgal, Toronto, ON (Abstract Co-Author) Speaker, Medtronic plc; Speaker, Elekta AB; Medical Advisory Board, VarianMedical Systems, Inc; Speaker, Accuray Incorporated; Research Grant, Elekta ABSean P. Symons, MPH, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseBradley J. Macintosh, PhD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseAimee Chan, Toronto, ON (Abstract Co-Author) Nothing to DiscloseSunit Das, MD,PhD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseDavid J. Mikulis, MD, Toronto, ON (Abstract Co-Author) Stockholder, Thornhill Research Inc; Research Grant, General ElectricCompany;


[email protected]

PURPOSE

Quantitative blood oxygen level dependent (qBOLD) magnetic resonance imaging (MRI) has been used as a method to gauge thelevel of oxygen saturation (SO2) in the tumors of patients with gliomas. We investigated whether there was a difference in thelevel of oxygenation in different grades of gliomas.


10 patients were recruited for this prospective, multi-institutional study. Patients underwent a preoperative Ferumoxytol-basedqBOLD MRI. Based on visual inspection of SO2 maps from qBOLD imaging, two volumes of interest (VOIs) from the tumor werechosen. Biopsy samples from the VOIs were taken to correlate histopathological measures of hypoxia against qBOLD, and stainingwas scored by a neuropathologist. Patients with glioblatoma (GBM; WHO Grade IV) was compared with lower-grade astrocytomas. 1patients' research samples were inconclusive, and therefore excluded from pathological analysis.

RESULTS

Pathology reports indicated that: 1 patient had diffuse astrocytoma (WHO Grade II), 3 patients had anaplastic astrocytoma (WHOGrade III), and 6 patients GBM. When comparing low-SO2 samples, non-GBM patients had on average higher SO2 (26.0% vs 10.6%,p=0.07). Pathology staining of low-SO2 samples showed significantly higher levels of staining in HIF1a (p=0.048), VEGF (p=0.04)and CAIX (p<0.01) in GBM compared to lower-grade gliomas. No significant differences were detectable in the high SO2 (mean,37.4%) tumors.

CONCLUSION

Levels of oxygenation appear to decrease with increasing glioma grade, and is detectable by qBOLD MRI. Pathological markers ofhypoxia support this notion. The threshold for differentiating GBM from lower-grade gliomas appears to be ~35% SO2. Moresubjects are required to confirm these results.


qBOLD MRI can be an alternative method to assess tumor grade when biopsy is not feasible. In light of new hypoxia-targetingtherapies, it can also be used to monitor oxygenation state of the tumor during treatment.


ParticipantsNiha G. Beig, MS, Cleveland, OH (Presenter) Nothing to DiscloseMarwa Ismail, PhD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAnant Madabhushi, PhD, Cleveland, OH (Abstract Co-Author) Research funded, Koninklijke Philips NVManmeet Ahluwalia, Cleveland, OH (Abstract Co-Author) Nothing to DisclosePallavi Tiwari, PhD, Cleveland, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Recent WHO classification of diffuse gliomas defined 3 subtypes based on their molecular status: Isocitrate dehydrogenase wildtype (IDH-WT), IDH mutant with 1p/19q intact (IDHmut-noncodel), and IDH mutant with 1p/19q co-deletion (IDHmut-codel). Eachcategory represents different prognosis and chemo-sensitivity thus impacting treatment decisions. Previous studies have linkedtumor location with patient outcome. In this feasibility study, we developed population atlases of pre-treatment MRI lesions toevaluate whether IDH-WT, IDHmut-codel, IDHmut-noncodel tumors will have spatial proclivity to hemispheric or lobe-specificlocations based on their frequency of occurrence.

RC205-05 High-Tumor Mutation Burden (Hypermutation) in Gliomas Exhibit a Unique Predictive RadiomicSignature


RC205-06 Brain Tumor Surveillance Imaging in the Era of Genomics and Personalized Medicine



150 pre-operative MRI sequences (1.5T/3T T1w, T2w, FLAIR scans, multi-center) of patients diagnosed with diffuse gliomas (65low grade gliomas and 85 glioblastomas) were considered from TCIA, along with their IDH mutation and 1p/19q co-deletion status.Frequency atlases of tumor occurrence in T2/FLAIR hyper-intensity regions were developed for each sub-type, by averaging voxelintensities across all patients. To compute significant differences (p-value<0.05), voxel-based analysis of differential involvement(ADIFFI) based on two-tailed Fisher's exact test was performed on (a) IDH-WT (n= 91) vs IDH-mut (n=59), and (b) IDHmut-codel(n=13) vs IDHmut-noncodel (n=57) atlases. Prominent clusters were identified and mapped to LONI Probabilistic Brain Atlas(LPBA40) parcellations to provide anatomic localization for each sub-type.

RESULTS

The ADIFFI analysis revealed that IDHmut tumors were predominant in frontal lobe with a frequency of 52.7% occurrence whereasIDH-WT had a multi-centric distribution across parietal and temporal lobes (p<0.005). Prominent cluster of IDHmut-codel was foundto be lateralized to the left hemisphere in the cingulate gyrus region, while IDHmut-noncodel had 60% occurrence in the superiorfrontal gyrus of the right hemisphere (p<0.05).

CONCLUSION

Our analysis suggests spatial proclivity of molecular subtypes to hemispheric and lobe-specific locations in the brain. The spatiallocalization could serve as an imaging marker for differentiating molecular subtypes of gliomas.


IDHmut-codel have a favorable response to chemoradiation, while IDHmut-noncodel have improved prognosis versus IDH-WT.Identifying radiogenomic markers of sub-types could enable personalized treatments in Gliomas.

ParticipantsIslam S. Hassan, MD, Houston, TX (Abstract Co-Author) Nothing to DiscloseAikaterini Kotrotsou, PhD, MEng, Houston, TX (Abstract Co-Author) Nothing to DiscloseCarlos M. Kamiya, Houston, TX (Abstract Co-Author) Nothing to DiscloseNabil A. Elshafeey, MD, Houston, TX (Presenter) Nothing to DiscloseKristin Alfaro-Munoz, MS, Houston, TX (Abstract Co-Author) Nothing to DisclosePascal O. Zinn, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseJohn F. deGroot, MD, Houston, TX (Abstract Co-Author) Nothing to DiscloseRivka R. Colen, MD, Houston, TX (Abstract Co-Author) Research Grant, General Electric Company;


[email protected]

PURPOSE

Increase in tumor mutation burden (TMB) or hypermutation is the excessive accumulation of DNA mutations in cancer cells.Hypermutation was reported in recurrent as well as primary gliomas. Hypermutated gliomas are mostly resistant to alkylatingtherapies and exhibit a more immunologically reactive microenvironment which makes them a good candidate for immune checkpointinhibitors. Herein, we sought to use MRI radiomics for prediction of high TMB (hypermutation) in primary and recurrent gliomas.


In this IRB-approved retrospective study, we analyzed 101 patients with primary gliomas from the University of Texas MD AndersonCancer Center. Next generation sequencing (NGS) platforms (T200 and Foundation 1) were used to determine the Mutation burdenstatus in post-biopsy (stereotactic/excisional). Patients were dichotomized based on their mutation burden; 77 Non-hypermutated(<30 mutations) and 24 hypermutated (>=30 mutations or <30 with MMR gene or POLE/POLD gene mutations). Radiomic analysiswas performed on the conventional MR images (FLAIR and T1 post-contrast) obtained prior to tumor tissue surgical sampling; androtation-invariant radiomic features were extracted using: (i) the first-order histogram and (ii) grey level co-occurrence matrix.Then, we performed Logistic regression modelling using LASSO regularization method (Least Absolute Shrinkage and SelectionOperator) to select best features from the overall features in the dataset. ROC analysis and a 50-50 split for training and testing,were used to assess the performance of logistic regression classifier and AUC, Sensitivity, Specificity, and p-value were obtained.

RESULTS

LASSO regularization (alpha = 1) was performed with all the 4880 features for feature selection and 40 most prominent featureswere selected for logistic regression modelling. Our 50-50 split ROC analysis showed an accuracy of 94%, sensitivity of 75%, andspecificity of 100% and a p-value of 0.0008).

CONCLUSION

An MRI-radiomic phenotype is predictive of the increase in TMB (Hypermutation) in both primary and recurrent gliomas.


Hypermutated gliomas are resistant to alkylating therapies but responsive to immune checkpoint inhibitors. Our proposed radiomicbiomarker can be used to guide therapy and patient selection for immunotherapy clinical trials.

ParticipantsRajan Jain, MD, Hartsdale, NY (Presenter) Consultant, Cancer Panels; Royalties, Thieme Medical Publishers, Inc

RC205-07 Update on Pediatric Brain Tumor Imaging


RC205-08 Selection of Imaging-based Surrogate Endpoints Depending on a Specific Target of Test Treatment inPhase III Randomized Controlled Trials of Glioblastoma


RC205-09 18F-Choline PET/RM in Brain Tumours: Multimodality Imaging in Gliomas Response Assessment



[email protected]

LEARNING OBJECTIVES

1) To discuss currently used response assessment criteria in brain tumor surveillance, such as Macdonald and RANO criteria andtheir limitations. 2) Participants will learn how the emergence of genomic markers has brought a paradigm shift in the managementand surveillance of gliomas. 3) Participants will learn about the complexities of post-treatment imaging appearance of brain tumorsin the new age of targeted immuno-therapies and what functional imaging techniques can add value to conventional surveillanceMRI.

ParticipantsZoltan Patay, MD, PhD, Memphis, TN (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Familiarize with new concepts and entities introduced in the 2016 update of the WHO Classification of Tumours of the CNS andpertinent to pediatric brain neoplasms. 2) Review the most recent developments related to the classification of embryonal andependymal tumors since the publication of the 2016 update. 3) Discuss relevance and implications of the above for the practcingradiologist.

ParticipantsChong Hyun Suh, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseHo Sung Kim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSang Joon Kim, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose

PURPOSE

Phase III randomized controlled trials (RCTs) in glioblastoma have used various potential surrogate endpoints with imaging. Thesurrogacy of imaging-based endpoints remains largely unknown and can be dependent on the type of test treatment. Weinvestigated the surrogacy of imaging-based endpoints as well as their values depending on a specific target of test treatment inpatients with glioblastoma.


A systematic search of phase III RCTs in glioblastoma was performed. Surrogacy between imaging-based endpoints includingprogression-free survival (PFS), 6 month PFS (6moPFS), 12 month PFS (12moPFS), median PFS, and objective response rate (ORR)with overall survival (OS) were explored using weighted linear regression for the hazard ratio for OS and the hazard ratios or oddsratios for imaging-based endpoints. Subgroup analyses according to disease entity, a specific target of test treatment, andresponse assessment criteria were performed. The quality of the reporting of efficacy with these IBEs was also evaluated.

RESULTS

Twenty-three RCTs published between 2000 and 2017, covering 8387 patients, met the inclusion criteria. OS showed significantcorrelations with PFS (standardized ß coefficient [R]=0.719), 6moPFS (R=0.647), and 12moPFS (R=0.638). OS showednonsignificant correlations with median PFS and ORR. The subgroup analyses consistently showed highly significant correlationsbetween OS and PFS. PFS showed the highest correlations with OS in drugs targeting DNA repair-cell cycle control-epigeneticmodifiers (R=0.913) and drugs targeting growth factor receptors-MAPK/PI3K signaling pathways (R=0.962). 12moPFS showed thehighest correlations with OS in antiangiogenic therapy, (R=0.821). Trials using RANO criteria showed higher correlation coefficientsbetween OS and PFS, 6moPFS, and 12moPFS than trials using MacDonald criteria. In terms of quality of reporting, there were highproportions of clearly defined primary endpoints (91%) and intent-to-treat analyses (83%). Compared with trials published between2000 and 2011, those published between 2012 and 2017 were more likely to be supported by industry.

CONCLUSION

Imaging-based endpoints can be surrogates in phase III RCTs of glioblastoma.


The specific target of test treatment and response assessment criteria should be considered in selection of imaging-basedendpoints as a surrogate endpoint.

AwardsStudent Travel Stipend Award

ParticipantsValentina Ferrazzoli, MD, Rome, Italy (Presenter) Nothing to DiscloseHarpreet K. Hyare, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseAnanth Shankar, London, United Kingdom (Abstract Co-Author) Nothing to Disclose

RC205-10 Prediction of Genomic Profiles and Survival in Glioblastoma Patients: Feasibility of Qualitative andQuantitative Analyses of Arterial Spin-Labeling Perfusion-Weighted Imaging


Christine Tang, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseAhmed Al-Khayfawee, MRCP,MD, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseRoberto Floris, MD, Rome, Italy (Abstract Co-Author) Nothing to DiscloseFrancesco Fraioli, MD, London, United Kingdom (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Evaluation of post-treatment glioma burden remains a significant challenge, particularly in Teenage and Young Adult (TYA)population. Although aminoacid PET has impacted on glioma imaging, 18fluoro-Choline Positron Emission Tomography (ChoPET) iscurrently more widely available. Purpose was to evaluate ChoPET/MR for post-treatment TYA glioma burden.


27 TYA (mean age 16 years, 8-22 years) in treatment (radiotherapy e/o chemotherapy) for astrocytic brain tumours (14 WHOIII/IV; 13 WHO I/II) were evaluated with ChoPET/MR. 59 follow up scans were retrospectively reviewed; maximum standardizeduptake values (SUVmax) and MR features (diameters, enhancement) were recorded. In 13 cases dynamic susceptibility contrastperfusion weighted imaging (DSCpwi) was analyzed; relative cerebral blood volume (rCBV) and SUV in enhancing and non-enhancingtumour volumes (Venh, Vne) and in normal appearing white matter (wm) were calculated (rCBVenh, rCBVne, rCBVwm, SUVenh,SUVne, SUVwm). A blinded nuclear medicine and a radiologist scored the images on tumour probability (1:unlikely-5:definitely).Receiver Operating Characteristic (ROC) analysis was used considering as gold standard for diagnosis the histopathology or followup. Pearson correlation coefficient was used for SUV and rCBV and independent T-Test for differences in ROIs.

RESULTS

MR sensitivity for residual tumour was 92.7% (85.7% WHO III/IV,81.5% WHO I/II). PET sensitivity was 78.2% (78.6% WHO III/IV,63.0% WHO I/II). Discrepancy was of 20% (11/12 non enhancing). Significant positive correlation between SUV and rCBV in all ROIswas found (r=0.051, p=0.0016). Tumour component analysis showed significantly higher SUVenh and SUVne than SUVwm(SUVenh:p<0.001, SUVne: p=0.021) and significantly higher rCBVne than rCBVwm (p=0.005). rCBVenh showed only borderlinesignificance (p=0.053).

CONCLUSION

ChoPET is able to detect post-treatment enhancing and non-enhancing tumour but both conventional MR and pwi are superior forevaluating non-enhancing disease. In TYA gliomas follow up a quantitative multimodality evaluation can better identify bothenhancing and non-enhancing residual/recurrent disease being promising in the early assessment of tumour response.


In teenage and young-adults gliomas early assessment of tumour response to therapy can be improved by complementary use ofPET/MR to evaluate different tumour components.

ParticipantsRoh-Eul Yoo, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseTae Jin Yun, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To explore the feasibility of using arterial spin-labeling perfusion-weighted imaging (ASL-PWI) to predict genomic profiles andsurvival in glioblastoma (GBM) patients.


One hundred thirty-two consecutive GBM patients, who had undergone maximal surgical resection or biopsy followed by concurrentchemo- and radiation therapy and adjuvant chemotherapy using temozolomide between January 2011 and November 2015, wereincluded in this retrospective study. CBF at the contrast-enhancing and T2 hyperintense portions on preoperative ASL-PWI wereevaluated both qualitatively (hypo- / iso- / hyperperfusion relative to gray matter) and quantitatively (mean and maximal CBFs oftumors normalized with respect to those of contralateral gray matter [nCBFmean and nCBFmax]). The associations between ASLfindings and major genomic profiles or survival were evaluated using Mann-Whitney U-test, Fisher's exact test, Spearman rankcorrelation, and Kaplan-Meier analysis. Receiver operating characteristics analysis was performed to determine the diagnosticperformance of the imaging parameters for prediction of genetic biomarkers.

RESULTS

nCBFmean and nCBFmax at contrast-enhancing portions were significantly higher in IDH wild-type group (n = 102) than in IDHmutant (IDH1 or IDH2) group (n = 17) (P = .009 and P = .007, respectively). Sensitivity and specificity for prediction of IDHmutation were 59% and 81% at the optimal cutoff value of 1.13 for nCBFmax. nCBFmax at contrast-enhancing portions tended tobe lower in patients with methylated MGMT promoter (n = 65) than in those with unmethylated MGMT promoter (n = 49) (P = .072).No significant associations were found between nCBF and other genetic biomarkers including ATRX, PTEN, p53, and EGFR.Hyperperfusion at contrast-enhancing portions was significantly more common in IDH wild-type group than in IDH mutant group (P =.003). Hyperperfusion was associated with shorter progression-free survival as compared with hypo- or isoperfusion (Median, 6.9vs. 12.5 vs. 19.3 months; P = .029).

CONCLUSION

ASL-PWI may help noninvasively predict IDH mutation status and progression-free survival in glioblastoma patients.

RC205-11 Clinical Data and Vascular Pattern on MRI to Predict Survival in 'De Novo' Glioblastoma


RC205-12 Application of Radiomics & Deep-Learning in Brain Tumor Imaging

Monday, Nov. 26 11:30AM - 12:00PM Room: S406B

ASL-PWI may help noninvasively predict IDH mutation status and progression-free survival in glioblastoma patients.


ASL-PWI may help noninvasively predict IDH mutation status and progression-free survival in glioblastoma patients and isrecommended as part of preoperative tumor evaluation particularly for patients with impaired renal function.


ParticipantsBlanca Domenech, MD, Barcelona, Spain (Presenter) Nothing to DiscloseAlfredo Gimeno Cajal, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseGerard Blasco, RT, Girona, Spain (Abstract Co-Author) Nothing to DisclosePepus Daunis-I-Estadella, Girona, Spain (Abstract Co-Author) Nothing to DiscloseCarmen Balana, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseJaume Capellades, MD, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseAngel Alberich-Bayarri, MD, Valencia, Spain (Abstract Co-Author) Nothing to DiscloseKambiz Nael, MD, New York, NY (Abstract Co-Author) Medical Advisory Board, Canon Medical Systems CorporationCarlos Leiva-Salinas, MD,PhD, Charlottesville, VA (Abstract Co-Author) Nothing to DiscloseRajan Jain, MD, Hartsdale, NY (Abstract Co-Author) Consultant, Cancer Panels; Royalties, Thieme Medical Publishers, IncMarco Essig, MD, Erlangen, Germany (Abstract Co-Author) Nothing to DiscloseSalvador Pedraza, MD, PhD, Girona, Spain (Abstract Co-Author) Nothing to DiscloseJosep Puig Alcantara, MD, Girona, Spain (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

MRI provides information on the physiologic properties of glioblastomas. In addition to established prognostic markers such as age,performance status, and extent of resection, increased vascularity on contrast-enhanced MRI is associated with shortenedsurvival. We investigated whether glioblastoma vascular pattern (GVP-MRI), combined with clinical variables and other imagingfeatures, could improve the predictive power of survival models.


From January 2012 through December 2016, 97 consecutive patients (62 men; mean age, 58±15 years) with histologically provenglioblastoma (GLIOCAT substudy) underwent 1.5T-MRI including anatomical, diffusion-weighted, first-pass DSC, and T1-weightedsequences after 0.1 mmol/kg gadobutrol (1 mm isometric voxel). We used Olea Sphere V.3.0 software (Olea Medical, La Ciotat,France) to analyze rCBV, rCBF, mean delay time, and apparent diffusion coefficient in volumes of interest for contrast-enhancinglesion (CEL), non-CEL, and contralateral tissue. Glioblastomas with >5 vessels seen within the lesion on postcontrast T1-weightedimages were considered hyper-GVP-MRI. Prognostic factors were evaluated by Kaplan-Meier survival, ROC analyses, and hazardratios (HR).

RESULTS

Glioblastomas were considered hyper-GVP-MRI in 58 (60.4%) patients. Patients with hyper-GVP-MRI glioblastomas were older, hadhigher volumeCEL, increased rCBFCEL and poor survival. Combining Stupp protocol (HR: 0.604; 95% CI: 0.459-0.796), age (HR:0.163; 95% CI: 0.090-0.297), and GVP-MRI (HR: 1.481; 95%CI: 0.909-2.414) best predicted survival at 1 year (AUC 0.901, 83.3%sensitivity, 93.3% specificity, 96.2% PPV, 73.7% NPV).

CONCLUSION

Our preliminary data suggest that combining clinical parameters and vascular pattern on MRI improves survival prediction in 'denovo' glioblastoma. Cross-validation studies in other populations are necessary to test the generalizability of our findings.


Information about baseline risk and prognosis is crucial for assigning patients with glioblastomas to optimized treatment regimens inclinical practice or to subgroups in clinical trials.

ParticipantsPhilipp Kickingereder, MD,MBA, Heidelberg, Germany (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) To understand and critically reflect the impact of radiomics and radiogenomics. 2) To understand the impact of deep-learning forguiding treatment decisions and advancing precision and personalized medicine in neuro-oncology.

RC211-01 Logistics: Incorporating New PET Tracers into Practice


RC211-02 Validation of the Use of Image-Derived Input Function for the Quantification of Ga-68 PSMA-11Uptake in Prostate Cancer Patients Using Dynamic PET/MR


RC211

Nuclear Medicine Series: New PET Tracers for Prostate Cancer (Interactive Session)

Monday, Nov. 26 8:30AM - 12:00PM Room: S505AB

CT GI NM



ParticipantsTerence Z. Wong, MD, PhD, Chapel Hill, NC (Moderator) Consultant, Lucerno Dynamics, LLC;

Sub-Events

ParticipantsNancy M. Swanston, RT, Houston, TX (Presenter) Nothing to Disclose

ParticipantsAnna Maria Ringheim, MSc,BEng, Sao Paulo, Brazil (Presenter) Nothing to DiscloseGuilherme Campos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarcelo L. Cunha, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseTaise Vitor, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseKarine M. Martins, MS, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarycel F. de Barboza, MSc, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJairo Wagner, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseAna Claudia Miranda, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseLeonardo L. Fuscaldi, PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseAna Claudia R. Durante, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGustavo C. Lemos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJose Roberto Colombo Jr, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRonaldo H. Baroni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To validate the use of image-derived input function of the common iliac artery as an alternative to arterial blood sampling inpharmacokinetic modeling of Ga-68 PSMA-11 uptake using hybrid positron emission tomography and magnetic resonance (PET/MR)imaging in primary prostate cancer.


This observational prospective study was approved by our Institution's Ethics Committee. Eleven patients with clinically significantprostate cancer underwent a 60-minute dynamic PET/MR scan of the pelvis with an injected dose of Ga-68 HBED-CC-PSMA (Ga-68PSMA-11). Arterial blood activity was measured by an automatic arterial blood sampling device during the first 10 min and manualblood samples were collected for metabolite analysis and for blood to plasma transformation to derive an arterial input function(AIF). One lesion per patient (with the highest uptake) and the common iliac artery were outlined using isocontour volumes on thePET images. Two image-derived input functions (IDIF) were calculated: whole blood curve (IDIF_bl) and plasma curve (IDIF_pl)corrected by the average plasma-to-blood ratio. An irreversible two-tissue compartment model, with rate constants K1, k2 and k3,were fitted to the data using AIF, IDIF_pl and IDID_bl and the net influx rate Ki=K1k3/(k2+k3) was calculated. The agreementbetween K1 and Ki from AIF and IDIF_bl and IDIF_pl were reported by intraclass correlation coefficients (ICC) with 95% confidenceintervals.

RESULTS

Ga-68 PSMA-11 was stable in-vivo, not necessitating metabolite correction. The mean plasma-to-blood ratio was 1.63. IDIFunderestimated the arterial input function by 50% at the bolus peak and by 20% at late times. For K1, ICC between AIF and IDIF_blwas 0.40 (-0.22, 0.80) and between AIF and IDIF_pl 0.60 (0.04, 0.87). For Ki, ICC between AIF and IDIF_bl was 0.77 (0.34, 0.93)and between AIF and IDIF_pl 0.94 (0.78, 0.98).

CONCLUSION

IDIF plasma curve can be used in clinical practice as an alternative to arterial blood sampling to calculate the uptake of Ga-68

RC211-03 Dual-Time 68Ga-PSMA-11 Imaging for Biochemically Recurrent Prostate Cancer Using LYSO andSiPM-Based Detectors PET/CT


RC211-04 Same Day PET/CT and PET/MRI with 68Ga-PSMA in the Evaluation of Biochemical Recurrence ofProstate Cancer


PSMA-11, but the method requires a known mean plasma-to-blood ratio. Agreement between IDIF whole blood curve and arterialinput function was poor to moderate. Underestimation of the IDIF should be explained by partial volume effects.


Image-derived plasma curves can be used in clinical practice as an alternative to arterial blood sampling to quantify uptake of Ga-68 PSMA-11 in prostate cancer.

ParticipantsHeying Duan, Stanford, CA (Presenter) Nothing to DiscloseSonya Y. Park, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseLucia Baratto, Stanford, CA (Abstract Co-Author) Nothing to DiscloseNegin Hatami, Stanford, CA (Abstract Co-Author) Nothing to DiscloseMohamed H. Khalaf, Stanford, CA (Abstract Co-Author) Nothing to DiscloseThomas Yohannan, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseGuido A. Davidzon, MD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric Company


[email protected]

PURPOSE

68Ga-labeled prostate-specific membrane antigen (PSMA-11) is a highly specific tracer for biochemically recurrent prostate cancerat low PSA levels. In this study we aim to compare the diagnostic performance of a new PET/CT scanner (Discovery MolecularInsights - DMI) using silicon photomultipliers (SiPM) detectors vs standard LYSO detectors PET/CT (Discovery 690 - D690) inpatients with biochemical relapse following a single injection of radiopharmaceutical.


Forty-four patients were prospectively recruited to undergo imaging on the D690 and DMI scanners, in randomized order. Imagesfrom the DMI PET/CT were reconstructed using ToF and a Bayesian penalized likelihood algorithm (Q.Clear®) whereas images fromthe D690 PET/CT were reconstructed using time-of-flight (ToF) and an ordered subset expectation maximization (OSEM) protocol.Two experienced nuclear medicine physicians reviewed both scans for each patient in random order, recorded the number andlocation of each lesion, and acquired standardized uptake value (SUV) measurements.

RESULTS

Twenty-three patients underwent imaging on the D690 PET/CT first followed by the DMI scanner, and twenty-one underwentscanning in the reverse order. The median PSA was 4.33 ng/mL with one outlier of 1170 ng/mL. PSMA PET detected sites ofrecurrence in 32/44 (73 %) patients, including 6/12 (50 %) patients with PSA below 1 ng/mL with the lowest PSA and a positivescan at 0.05 ng/mL. The mean lesion SUVmax measurements were higher on DMI than D690 regardless of the timely order of thescan (6.5 vs. 5.7 in D690 scan first and 4.6 vs. 4.2 for DMI performed first). However, the difference in mean lesion SUVmax wasonly significant for patients scanned on the D690 first (p<0.018).

CONCLUSION

These preliminary results suggest that the SiPM-based DMI PET/CT system offers better performance and superior detectortechnology and image quality compared to conventional LYSO-based D690 PET/CT. These results need to be confirmed in largerstudies.


SiPM-based DMI PET/CT system seems to offer better performance and superior detector technology and image quality comparedto conventional LYSO-based D690 PET/CT.

ParticipantsMarcelo L. Cunha, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseAkemi Osawa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGuilherme Campos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseLilian Y. Yamaga, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJulio Cesar Oliveira, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRicardo C. Fonseca, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJairo Wagner, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseTaise Vitor, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFernando I. Yamauchi, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseThais Mussi, MD,PhD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseRonaldo H. Baroni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose

PURPOSE

To compare PET/CT and PET/MRI with 68Ga-PSMA in patients with radical prostatectomy and biochemical recurrence (BCR).


This is a prospective, IRB approved study. 29 patients with prostate cancer were referred to our department to investigate BCR

RC211-05 Evaluation of Whole-body MRI versus 68Ga-PSMA PET/CT for Detection of Biochemical Recurrence inProstate Cancer Patients


RC211-06 The Effect of Various Beta Values on Image Quality and Semi-Quantitative Measurements in 68Ga-Labeled GRPR and PSMA PET/MRI Images Reconstructed With a Block Sequential RegularizedExpectation Maximization Algorithm


This is a prospective, IRB approved study. 29 patients with prostate cancer were referred to our department to investigate BCRand submitted to PET/CT and PET/MR with 68Ga-PSMA in the same day. The first exam started 50 minutes after 68Ga-PSMAinjection (15 started with PET/CT while 14 started with PET/MRI). The second study began right after the first one. Readers ofeach study were aware of clinical data but blinded of eventual findings of the other scan. Patients had age ranging from 52 to 78years old, had submitted to radical prostatectomy up to 12 years before the exams, and Gleason score ranged from 6 to 9. SerumPSA levels ranged from 0.22 to 12.8 ng/mL at time of the scans.

RESULTS

There were 16 positive PET/CT and 15 positive PET/MRI. Negative studies (no abnormal area suspicious for prostatic cancerrecurrence) counted 11 PET/CT and 13 PET/MR scans. PET/CT found 34 suspicious lesions and PET/MR, 27 lesions. Few equivocalstudies were found by both methods: 2 in PET/CT and 1 in PET/MRI, all of them with one uncertain lesion. There were 24/39studies match studies (positive or negative on both methods).

CONCLUSION

PET/CT and PET/MRI with 68Ga-PSMA can both be used in BCR scenario, in spite of PET/CT seems to have a slightly highersensitivity in our population.


PET/CT and PET/MRI are the most important imaging methods in BCR scenario. The comparison among them is essential to performthe right choice in clinical practice.

ParticipantsLino Sawicki, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseCarolin Buddensieck, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseJohannes Boos, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseRobert Rabenalt, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseHubertus Hautzel, MD, Juelich, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Presenter) Nothing to Disclose

PURPOSE

The purpose of our study was to assess whole-body MRI (wb-MRI) for lesion detection of biochemical relapse in prostate cancer(PCa) patients after curative treatment in comparison to 68Ga-PSMA PET/CT.


This is a prospective IRB-approved trial of 30 patients (age: 65.5 ± 9.6 years) with newly documented biochemical relapse of PCa(mean prostate-specific antigen (PSA) 2.11 ± 1.97 ng/ml) after curative therapy. All patients underwent both wb-MRI including adedicated pelvic imaging protocol and PET/CT with 167 ± 35 MBq 68Ga-PSMA within a time window of 10.5 ± 9.5 days. PET/CT andMRI datasets were separately evaluated regarding PCa lesion count, lesion type, localization, and diagnostic confidence (3-pointscale; 1-3) by two physicians. The reference standard was based on histopathology results, changes of PSA after targetedirradiation, follow-up imaging, and clinical data. Lesion-based and patient-based detection rates were compared using chi2 test.Differences in diagnostic confidence were assessed by Welch test.

RESULTS

A total of 58 PCa lesions were found in 22/30 patients in the study cohort. 68Ga-PSMA PET/CT detected 57/58 (98.3 %) lesions in21/30 (70 %) patients, and 15/58 (25.9 %) lesions were detected in 13/30 (43.3 %) patients using wb-MRI. The higher detectionrate of 68Ga-PSMA PET/CT was statistically significant both on a per-lesion (p = 0.001) and per-patient (p = 0.039) basis. In 8/30patients none of the two modalities actually localized the PCa relapse. Except for one local recurrence in the former prostate fossathat was exclusively detected by wb-MRI, all lesions detected by wb-MRI were also detectable on 68Ga-PSMA PET/CT. 68Ga-PSMAPET/CT offered a superior diagnostic confidence in categorization of PCa lesions compared to wb-MRI (2.7 ± 0.6 vs. 2.3 ± 0.6, p =0.011).

CONCLUSION

68Ga-PSMA PET/CT significantly outperforms wb-MRI for detection of biochemical relapse in PCa patients after curative treatment.


Wb-MRI is inferior to 68Ga-PSMA PET/CT for the detection of recurrent PCa. Nevertheless, as one local recurrence in one patientwas only detectable with MRI, it might be useful in selected cases.

ParticipantsLucia Baratto, Stanford, CA (Presenter) Nothing to DiscloseHeying Duan, Stanford, CA (Abstract Co-Author) Nothing to DiscloseHarsh C. Gandhi, BS,MSc, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseMehdi Khalighi, Palo Alto, CA (Abstract Co-Author) Employee, General Electric CompanyPraveen Gulaka, PhD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric Company

PURPOSE

RC211-07 Fluciclovine PET/CT: Practical Approach to Interpretation


RC211-08 Quantitative Comparison of Standardized Uptake Values of Same-Day Randomized Ga-68 PSMA-11PET/CT and PET/MR Scans in Recurrent Prostate Cancer Patients


The block sequential regularized expectation maximization (BSREM) algorithm is a new image reconstruction method that controlsnoise at higher iterations by applying a relative difference penalty built into the objective function. This enables one to employ moreiterations for convergence and better contrast recovery, while mitigating noise amplification. BSREM was recently introduced on theGE SIGNA PET/MRI platform. Here we evaluated how different values influence image quality and SUVmax in a cohort of prostatecancer patients who underwent 68Ga-RM2 or 68Ga-PSMA-11 scans.


We analyzed 36 prostate cancer patients who underwent either 68Ga-RM2 (15) or 68Ga-PSMA-11 (21) PET/MRI. The raw PET datawere retrospectively reconstructed using values of 250, 350, 500, 750 and 1000. Each reconstruction was reviewed independentlyby 3 nuclear medicine physicians and scored using a Likert scale (1 - poor, 5 - excellent quality). SUVmax values were measuredfrom 68Ga-RM2/PSMA PET/MRI for all the lesions identified as compatible with prostate cancer.

RESULTS

The mean±SD scores for 68Ga-RM2 PET images were 2.5±0.5 for =250 reconstructions, 3.2±0.6 for =350, 4.1±0.6 for =500,4.7±0.5 for =750 and 4.8±0.5 for =1000. The mean±SD scores for 68Ga-PSMA-11 PET images were 3.3±0.9 for =250reconstructions, 4.2±0.9 for =350, 4.7±0.6 for =500, 4.9±0.3 for =750 and 4.9±0.4 for =1000. The relative observed agreementamong readers for the values of 250, 350, 500, 750, 1000 were 49%, 50%, 60%, 70% and 74%, respectively. A total of 24 lesions(6 on RM2 and 18 on PSMA-11) were detected and the mean SUVmax measurements were: 13.1, 12.5, 10.4, 9.3 and 8 for the68Ga-RM2 values of 250, 350, 500, 750 and 1000, respectively; 22.6, 21.2, 19.7, 18.9 and 16.8 for the 68Ga-PSMA-11 values of250, 350, 500, 750 and 1000, respectively.

CONCLUSION

Different values should be used for different 68Ga-labeled radiopharmaceuticals such as those targeting GRPR and PSMA receptorsin prostate cancer. Once selected, the same value should be consistently used since SUVmax measurements differ with differentvalues.


BSREM algorithm improves image quality. Different beta values make different image quality and different SUVmax measurements ondifferent 68Ga-labeled radiopharmaceuticals.

ParticipantsEphraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Presenter) Research support, Blue Earth Diagnostics Ltd; Research support,Advanced Accelerator Applications SA


[email protected]

LEARNING OBJECTIVES

1) Assess the appropriate clinical indications for 18F-fluciclovine PET and understand the diagnostic accuracy of 18F-fluciclovinePET for local and metastatic prostate cancer. 2) Develop 18F-fluciclovine PET protocols for image optimization. 3) Apply the correct18F-fluciclovine PET interpretation for local and metastatic prostate cancer and benign physiologic variants.

ABSTRACT

Anti-1-amino-3-[18F]-flurocyclobutane-1-carboxylic acid (18F-fluciclovine) is a non-naturally occurring amino acid PET radiotracerthat is recently United States Food and Drug Administration approved for detection of suspected recurrent prostate cancer. Thetumor imaging features of this radiotracer mirror the upregulation of transmembrane amino acid transport that occurs in prostatecancer due to increased amino acid metabolism for energy and protein synthesis. This refresher course provides an overview of18F-fluciclovine PET diagnostic accuracy for identifying primary and metastatic disease, as well as proper 18F-fluciclovine PETimaging protocols. Correct interpretation criteria will be explored in detail to identify physiologic and pathologic 18F-fluciclovineuptake patterns and potential pitfalls.

ParticipantsAnna Maria Ringheim, MSc,BEng, Sao Paulo, Brazil (Presenter) Nothing to DiscloseGuilherme Campos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseKarine M. Martins, MS, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseTaise Vitor, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarcelo L. Cunha, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRonaldo H. Baroni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To determine the reproducibility and agreement of standardized uptake values from same-day Ga-68 PSMA-11 PET/CT and PET/MRscans, randomized in order to eliminate the influence of Ga-68 PSMA-11 accumulation as a function of time, in patients withrecurrent prostate cancer.


RC211-09 Detection Rate of 18F-FACBC (Fluciclovine) PET/CT Scan as a Function of Prostatic Specific Antigen(PSA) Level: Initial Experience of 76 Patients with Biochemically Recurrent Prostate Cancer


RC211-10 Role of Early Dynamic PET/CT Imaging with 68Ga-PSMA in Staging and Restaging of Prostate Cancer


Eighteen patients with recurrent prostate cancer after radical prostatectomy, all with visible lesions on the PET scan, were includedin this retrospective study, approved by the Institution's Ethics Committee. All patients underwent PET/CT and PET/MR scans inrandomized order after intravenous injection of a single dose of Ga-68 HBED-CC-PSMA (Ga-68 PSMA-11). Volumes of interest ontumor lesions were outlined and maximum standardized uptake value (SUVmax) corrected for lean body mass was calculated.Correlation and agreement between scans were assessed by generalized linear mixed-effects models and Bland Altman analysis. Apredictive model of PET/CT SUVmax was developed based on PET/MR SUVmax, patient characteristics and imaging parameters.

RESULTS

In total, there were 34 visible lesions on the PET scans: 5 local recurrences, 22 lymph node and 7 bone metastases. SUVmax fromPET/CT and PET/MR were significantly correlated, described by the following regression model equation: (PET/CT SUVmax) = 0.75 +1.00* (PET/MR SUVmax), with a coefficient of determination R2 of 0.77. Bland-Altman analysis showed that SUVmax were onaverage 20% higher on PET/CT than on PET/MR, with the largest percentage differences for small SUV's. The full predictive modelof PET/CT SUVmax showed significant association with PET/MR SUVmax (effect 1.15 (p<0.001)), serum prostate specific antigen(effect 0.99 (p=0.053)), scan time post-injection (effect 0.98 (p=0.003)) and acquisition time per bed position (effect 1.49(p=0.021)) with a coefficient of determination R2 of 0.85.

CONCLUSION

SUVmax from PET/CT and PET/MR are comparable and well correlated, but should not be used interchangeably without applying acorrection factor.


Since both Ga-68 PSMA-11 PET/CT and PET/MR are increasingly used in clinical practice, the reproducibility of SUVmaxmeasurements is essential in monitoring of disease progression and response to treatment.


ParticipantsAli Salavati, MD, MPH, Minneapolis , MN (Presenter) Nothing to DiscloseMehmet Gencturk, MD, Istanbul, Turkey (Abstract Co-Author) Nothing to DiscloseJerry W. Froelich, MD, Minneapolis, MN (Abstract Co-Author) Researcher, Siemens AG

PURPOSE

In 2016, synthetic amino acid anti-1-amino-3-[18F]-flurocyclobutane-1-carboxylic acid (FACBC, Fluciclovine, Axumin) wasapproved by U.S. Food and Drug Administration (FDA) as a new PET tracer for the detection and localization of biochemicallyrecurrent prostate cancer. The goal of this study was to determine the impact of PSA level on the detection rate of 18F-FACBCPET/CT.


After obtaining IRB approval, we retrospectively enrolled 76 patients with biochemical recurrence of prostate cancer referred for an18F-FACBC PET/CT scan at our institution. Relevant clinical information including demographic data,PSA level, and Gleason scorewere collected. Images were interpreted by two experienced nuclear radiologists. Receiver operating characteristic (ROC) curve andbootstrap technique with 2000 iterations were performed to determine the optimal cutoff points of PSA as a predictor of positiveand negative 18F-FACBC PET/CT scan.

RESULTS

The detection rate of 18F-FACBC PET/CT was 67.1% (51 of 76 scans). Positive findings were detected in the prostate/bed andpelvic lymph node regions in 88% of positive scans(45 of 51); metastatic lesion outside the pelvis in 21.5% of positive scans(11 of51), and bone metastasis in 11.7% of positive scans (6 of 51).18F-FACBC PET/CT scan detected potential sites of recurrence in23.5% of patients when PSA is < 1.0 ng/ml, 66.7 % of patients when PSA is 1.0-2.0 ng/ml, and 86.2% of patients when PSA is >2.0ng/ml. The ROC curve analysis with bootstrapping demonstrates a PSA >3.38 ng/ml has a likelihood ratio > 10 and positivepredictive value > 95% for a positive 18F-FACBC PET/CT scan. The PSA < 0.36 ng/mL has a likelihood ratio > 14 of having anegative 18F-FACBC PET/CT scan and results in positive imaging findings in <2.5% of patients.

CONCLUSION

Given the high pretest probability of positive imaging findings in patients with PSA > 3.4 ng/ml, they benefit most from 18F-FACBCPET/CT scans. In addition, negative scans of these patients should be interpreted more cautiously. Conversely, considering the lowpretest probability of positive PET/CT findings in patients with PSA < 0.36 ng/ml, these patients may benefit from a short follow-upbefore a diagnostic 18F-FACBC PET/CT scan.


18F-Fluciclovine is a valuable novel clinically-available PET/CT tracer for the detection and localization of biochemically recurrentprostate cancer, particularly when the PSA level is > 2 ng/ml.


ParticipantsMaria El Homsi, MD, Beirut, Lebanon (Presenter) Nothing to Disclose

RC211-11 Change in Salvage Radiotherapy Management Based On Fluciclovine (18F) PET/CT Guidance in Post-Prostatectomy Recurrent Prostate Cancer


Andrew Barakat, MD, Beirut, Lebanon (Abstract Co-Author) Nothing to DiscloseMohamad B. Haidar, MD, Beirut, Lebanon (Abstract Co-Author) Nothing to DiscloseBasel Yacoub, Beirut, Lebanon (Abstract Co-Author) Nothing to Disclose

PURPOSE

Ga-68 Prostate-Specific Membrane Antigen (PSMA)PET/CT is a new tool for the detection of new and recurrent prostatecancer.Standard imaging time is 60 minutes post injection of radiotracer but some lesions may be obscured by physiologicaccumulation of radiotracer in bladder. The aim of the study is to determine if the addition of early imaging at 3 and 6 minutes tostandard imaging at 60 minutes can improve the detection of new and recurrent prostate cancer at Ga-68 PSMA PET/CT.


After obtaining IRB approval, retrospective review of 257 consecutive patients who underwentGa-68 PSMA PET/CT betweenDecember 2016 and July 2017 was conducted. 167 patients underwent early (3 and 6 minute) and late (60 minute) imaging. Tworeaders blinded to the patient's clinical information, independently reviewed the early and late images and qualitatively andquantitatively scored the visibility of prostate lesions on a scale of 1-2 (1: lesion seen, 2: lesion not seen). Qualitatively, focaluptake higher than background that did not correspond to physiologic tracer accumulation was considered cancer. Quantitatively, acut off maximum standardized uptake value (SUVmax) of 2 was indicative of prostate cancer. Detection of prostate cancer wascompared between early and late imaging using McNemar test.

RESULTS

A total 115 patients (68.9%) had prostate cancer on imaging as seen on early (median SUVmax= 6.4) and late (median SUVmax= 8)PET/CT images. In 106/115 (64%), the cancers were seen on early and delayed imaging, in 8/115 (6.9%) the cancer was only seenon early imaging and masked by bladder activity on delayed imaging, and in 1/115(0.6%) only seen on delayed imaging. The additionof early imaging significantly improved the detection rate of prostate cancer (p=0.039).

CONCLUSION

The addition of early imaging at 3 and 6 minutes to the standard 60 minute imaging at Ga-68 PSMA PET/CT improves the detectionof prostate cancer .


Early imaging at Ga-68 PSMAPET/CT can help in the detection of prostate cancer that is obscured by the bladder activity at 60minutes.


ParticipantsOlayinka A. Abiodun-Ojo, MD,MPH, Atlanta, GA (Presenter) Nothing to DiscloseAshesh B. Jani, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseAkinyemi A. Akintayo, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseMehrdad Alemozaffar, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseOladunni O. Akin-Akintayo, MD,MPH, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseOluwaseun Odewole, MD, MPH, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseFunmilayo I. Tade, MD, Atlanta, GA (Abstract Co-Author) Nothing to DisclosePeter Nieh, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseViraj Master, MD, Atlanta, GA (Abstract Co-Author) Nothing to DisclosePretesh Patel, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJoseph W. Shelton, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseOmer Kucuk, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseZhengjia Chen, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseBruce Hershatter, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseBridget Fielder, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseRaghuveer K. Halkar, MD, Atlanta, GA (Abstract Co-Author) Research Grant, General Electric Company Research Grant, GileadSciences, Inc Royalties, General Electric Company Mark M. Goodman, PhD, Atlanta, GA (Abstract Co-Author) Royalties, Nihon Medi-Physics Co, LtdDavid M. Schuster, MD, Decatur, GA (Abstract Co-Author) Institutional Research Grant, Nihon Medi-Physics Co, Ltd; InstitutionalResearch Grant, Blue Earth Diagnostics Ltd; Institutional Research Grant, Advanced Accelerator Applications SA; Consultant,Syncona Ltd; ; ;


[email protected]

PURPOSE

We previously reported a 40.5% post-prostatectomy salvage radiotherapy management decision change based on guidance withfluciclovine PET/CT in a trial with 87/162 accrual (Clin Nucl Med. 2017 Jan;42(1):e22-e28). We set out to determine if this findingcontinued at the current accrual of 145/162 patients.


145 patients with post-prostatectomy biochemical recurrence of prostate cancer and negative bone scan were randomized toundergo treatment planning based on conventional imaging (CT, MRI) or fluciclovine PET/CT in a provider-determined intention-to-treat protocol. Radiotherapy decisions before and after fluciclovine PET/CT were compared and changes in treatment decision andfield were noted. Statistical significance of decision changes was determined using Clopper-Pearson (exact) binomial method withsignificance set at p <0.05.

RC211-12 Impact of 68Ga-PSMA PET on the Management of Patients with Prostate Cancer: A SystematicReview and Meta-Analysis


RC211-13 Clinician's Perspective: Impact and Applications


RC211-14 Fluciclovine (18F) Parameters on Targeted Prostate Biopsy Associated With True Positivity inRecurrent Prostate Cancer

RESULTS

70/145 patients underwent fluciclovine PET. Mean PSA (±SD) at scan was 1.89 (±4.32) ng/ml. 54/70 (77.1%) patients had apositive fluciclovine PET. Radiotherapy decision was changed in 27/70 (38.6%). Four (5.7%) had the decision for radiotherapywithdrawn after the fluciclovine PET findings of extrapelvic uptake. Radiotherapy field decision was changed in 23/66 (34.9%)remaining patients: 13/23 prostate bed only to prostate bed and pelvis; 10/23 prostate bed and pelvis to prostate bed only. Therewas no significant difference in mean age (62.0 vs. 61.6 years) and treatment-recurrence time interval (991.4 vs. 1403.6 days)between those with change in radiotherapy decision and those without. However, the pre-treatment PSA mean was significantlyhigher in those with treatment change (2.76ng/ml vs. 1.35ng/ml, p < 0.05). Changes in overall radiotherapy decision (p = 0.01) andfield (p < 0.05) were statistically significant.

CONCLUSION

Fluciclovine PET/CT had a significant effect on radiotherapy planning in post-prostatectomy patients with biochemical recurrentprostate cancer. Radiotherapy planning decision and field changes in the updated analysis is similar to the prior report. Furtherstudies are required to determine if this change in treatment plan has an effect on clinical outcomes. Research support: NIH(R01CA169188), NCT 01666808.


Use of fluciclovine PET/CT resulted in significant change in management (38.6%) in salvage radiotherapy planning in post-prostatectomy patients with recurrent prostate cancer.

ParticipantsSungmin Woo, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseSangwon Han, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYeon Joo Kim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChong Hyun Suh, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose

PURPOSE

68Gallium prostate-specific membrane antigen positron emission tomography (68Ga-PSMA PET) is a relatively novel imaging modalityfor assessment of patients with prostate cancer. Recent studies have shown promising results, with their ability to detectrecurrent/metastatic prostate cancer foci with superior performance than that of conventional imaging modalities (computedtomography, bone scintigraphy, and choline PET). However, the actual impact that 68Ga-PSMA PET has on management ofprostate cancer patients has not been well-established. Therefore, we aimed to systematically review the literature and to performa meta-analysis on the impact of 68Ga-PSMA PET on management of patients with prostate cancer.


Pubmed and EMBASE databases were systematically searched up to January 20, 2018. Studies reporting the proportion of patientswith prostate cancer that experienced change in management after 68Ga-PSMA PET were included. Quality of the included studieswas evaluated using the GRADE system. The proportion of management changes were pooled using a random-effects model.Subgroup analyses and meta-regression analyses was done to explore potential causes of heterogeneity.

RESULTS

15 studies (1163 patients) were included. The pooled proportion of management changes was 54% (95% CI 47%-60%). Meta-regression analyses revealed that PET-positivity was a significant factor of heterogeneity (p=0.0486). Other variables, includingclinical setting, change type (intended vs implemented), responding entity (referring doctor or multidisciplinary committee), D'Amicorisk, Gleason score, use of androgen deprivation therapy, PSA at initial diagnosis, pre-PET PSA, and PSA-doubling time, were notsignificant (p=0.2802-0.9574). In patients with biochemical failure, proportions of radiotherapy, surgery, focal therapy, andmultimodal treatment increased, while those of systemic treatment and no treatment decreased after performing 68Ga-PSMA PET.

CONCLUSION

This meta-analysis showed that 68Ga-PSMA PET had a large impact on the management of prostate cancer patients. PET-positivityaffected the proportion of management changes.


We found that 68Ga-PSMA PET alters management in approximately half of the patients with prostate cancer. Studies with higherproportion of patients with 68Ga-PSMA PET-positive lesions tend to have their management altered more frequently.

ParticipantsSteve Cho, MD, Madison, WI (Presenter) Imaging Endpoints; General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Review current and emerging PET radiotracers for prostate cancer. 2) Assess how PET imaging can address unmet clinical needsin prostate cancer. 3) Address remaining and new clinical and research questions arising from these new PET radiotracers.


RC211-15 Evaluation of 18F-DCFPyL PSMA-Based PET/CT and mpMRI in Patients with Localized ProstateCancer


ParticipantsOlayinka A. Abiodun-Ojo, MD,MPH, Atlanta, GA (Presenter) Nothing to DiscloseBaowei Fei, PhD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAkinyemi A. Akintayo, MD, Atlanta, GA (Abstract Co-Author) Nothing to DisclosePeter Nieh, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseViraj Master, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseFunmilayo I. Tade, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseOladunni O. Akin-Akintayo, MD,MPH, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseMehrdad Alemozaffar, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseAdeboye Osunkoya, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseMark M. Goodman, PhD, Atlanta, GA (Abstract Co-Author) Royalties, Nihon Medi-Physics Co, LtdDavid M. Schuster, MD, Decatur, GA (Abstract Co-Author) Institutional Research Grant, Nihon Medi-Physics Co, Ltd; InstitutionalResearch Grant, Blue Earth Diagnostics Ltd; Institutional Research Grant, Advanced Accelerator Applications SA; Consultant,Syncona Ltd; ; ;


[email protected]

PURPOSE

Fluciclovine is FDA approved for detection of recurrent prostate cancer. We set out to evaluate fluciclovine uptake parameters thatcorrelate with true positivity for local recurrence in non-prostatectomy treated patients.


21 patients (PSA 7.4±6.8 ng/ml) with nadir PSA+2 after non-prostatectomy local therapy underwent dual time-point fluciclovine(364.1±37.7 MBq) PET/CT (4-15 minutes; 16-25 minutes) from pelvis to diaphragm. Uptake in the prostate over background wasdelineated and co-registered to a previously obtained planning 3-D ultrasound. Fluciclovine uptake (SUVmax) and target-to-background ratios (TBR) (SUVmax/SUVmean) of blood pool (aorta), prostate, and marrow (L3) were recorded. Uptake pattern (focalvs non-focal), subjective suspicion level [3 (equivocal), 4 (moderate), 5 (high)], and lesion location were noted. Targeted biopsiesof the identified lesions with histologic analysis were completed. Statistical significance was determined using univariate regressionanalysis.

RESULTS

17/50 (34.0%) targeted lesions were positive for recurrent cancer. Compared to negative lesions, targeted positive lesions hadsignificantly (p<0.01) higher mean SUVmax of lesion (6.62±1.70 vs 4.92±1.27), TBR (marrow) (2.57±0.81 vs 1.69±0.51), and TBR(blood pool) (4.10±1.17 vs 3.00±1.01) at the first time-point, and remained significant at the later time-point except TBR (bloodpool). Focal uptake (OR 12.07, [95% CI 2.98-48.80], p<0.01) and subjective high (5) suspicion level (OR 10.91, [95% CI 1.19-99.69], p=0.03) correlated with true positivity. All other parameters did not significantly correlate with true positivity. Of the 17targeted lesions with focal uptake and subjective high suspicion, 11/17 (64.7%) were true positive. 16/33 (48.5%) false positivetargeted lesions had evidence of prostatitis or radiation changes.

CONCLUSION

True positivity of fluciclovine targeted prostate biopsy in non-prostatectomy treated patients correlates with focal uptake, higherSUV, target-to-background (blood pool and marrow) ratios and subjective high suspicion level. These parameters may be utilized forfuture modification of interpretative criteria. Research Support: NIH (CA156775, CA204254 and CA176684)


Targeted biopsy of lesions with focal uptake and subjective high suspicion on fluciclovine PET-CT increases the true positivity ratefrom 34.0% to 64.7% in non-prostatectomy treated patients.

ParticipantsMarcin Czarniecki, MD, Bethesda, MD (Presenter) Nothing to DiscloseStephanie A. Harmon, PhD , Bethesda, MD (Abstract Co-Author) Research funded, NCIEsther Mena, MD, Bethesda, MD (Abstract Co-Author) Nothing to DiscloseMaria Lindenberg, MD, Bethesda, MD (Abstract Co-Author) Nothing to DiscloseYolanda Mckinney, Bethesda, MD (Abstract Co-Author) Nothing to DiscloseDeborah Citrin, MD, Bethesda, MD (Abstract Co-Author) Nothing to DiscloseBradford J. Wood, MD, Bethesda, MD (Abstract Co-Author) Researcher, Koninklijke Philips NV; Researcher, Celsion Corporation;Researcher, BTG International Ltd; Researcher, Siemens AG; Researcher, XAct Robotics; Researcher, NVIDIA Corporation;Intellectual property, Koninklijke Philips NV; Intellectual property, BTG International Ltd; Royalties, Invivo Corporation; Royalties,Koninklijke Philips NV; ; ; Peter Pinto, Bethesda, MD (Abstract Co-Author) Nothing to DiscloseRonnie Mease, PhD, Baltimore, MD (Abstract Co-Author) Inventor, Progenics Pharmaceuticals, Inc; Inventor, Advanced AcceleratorApplications SAMartin G. Pomper, MD, PhD, Baltimore, MD (Abstract Co-Author) Researcher, Progenics Pharmaceuticals, Inc; License agreement,Progenics Pharmaceuticals, Inc; Researcher, Advanced Accelerator Applications SA; License agreement, Advanced AcceleratorApplications SA; Co-founder, Cancer Targeting Systems, Inc; Board Member, Cancer Targeting Systems, Inc; Researcher, CelgeneCorporation, Inc; License agreement, Celgene Corporation, Inc; Co-founder, Neurly; Board Member, Neurly; Co-founder, TheralyPharmaceuticals, Inc; Board Member, Theraly Pharmaceuticals, IncPeter L. Choyke, MD, Rockville, MD (Abstract Co-Author) Nothing to DiscloseBaris Turkbey, MD, Bethesda, MD (Abstract Co-Author) Nothing to Disclose

RC211-16 Quantification of the Pharmacokinetics of Ga-68 PSMA-11 in Prostate Cancer Patients using HybridPositron Emission Tomography and Magnetic Resonance Imaging



[email protected]

PURPOSE

To assess the ability of 18F-DCFPyL PET/CT to predict prostate cancer Gleason grade and its role as a potential adjunct tomultiparametric prostate MRI.


Patients with prostate cancer (PCa) underwent multi-parametric MRI (mpMRI) and 18F-DCFPyL PET/CT imaging. MpMRI lesioncharacteristics (PIRADS, ADCmean, ADCmin, ADC10, MRIvol) were derived manually and 18F-DCFPyL PET/CT metrics SUVmax,SUVmean, MTV and TLG (SUVmean*MTV) were extracted from tracer-specific regions of interest (ROI). Lesion metrics werecorrelated with each other using the Spearman rank test, and their ability to differentiate tumor pathology Gleason grade (GG) <3vs. GG 3-5 was performed using the Wilcoxon rank sum test.

RESULTS

Thirteen patients with high-risk PCa were included in the study, with 25 (12 GG<3, 12 GG 3-5) lesions found across the twomodalities (median age 70.6, MRI volume 62mL, and PSA 18.87ng/mL). Seven patients did not have findings suspicious formetastatic disease on 18F-DCFPyL-PET/CT. MpMRI and 18F-DCFPyL PET/CT both detected 22/25 lesions, while 3 were notdetected by any of the two modalities. 18F-DCFPyL PET/CT (SUVmax, SUVmean and TLG) as well as mpMRI metrics (ADCmin andMRI size) were significantly associated with GG 3-5 pathology (p=0.017, 0.020, 0.004, 0.028, 0.008, respectively). Additionally, MRIsize correlated significantly with SUVmax, MTV and TLG (p=0.01, 0.02, 0.02, respectively) and SUVmax, SUVmean correlatedsignificantly to ADCmin. All SUV metrics correlated positively with PIRADS 5 lesions vs. the remaining categories. In some cases, MRIfindings did not entirely colocalize with PET avidity indicating MRI underestimation of functional burden.

CONCLUSION

18F-DCFPyL PET-identified prostatic lesions and mpMRI findings correlate with each other, Gleason grades and PIRADS. MpMRI mayunderestimate the tumor burden seen on 18F-DCFPyL PET/CT imaging, which may aid focal therapy decisions.


Quantitative metrics of 18F-DCFPyL PET/CT and mpMRI correlate well with each other and with PCa Gleason grade and PIRADS.18F-DCFPyL PET/CT adds additional functional metrics which may be meaningful in determining tumour heterogeneity and burden.

ParticipantsAnna Maria Ringheim, MSc,BEng, Sao Paulo, Brazil (Presenter) Nothing to DiscloseGuilherme Campos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarcelo L. Cunha, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseKarine M. Martins, MS, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseTaise Vitor, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarycel F. de Barboza, MSc, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJairo Wagner, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseAna Claudia Miranda, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseLeonardo L. Fuscaldi, PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseAna Claudia R. Durante, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGustavo C. Lemos, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJose Roberto Colombo Jr, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRonaldo H. Baroni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To quantify Ga-68 PSMA-11 uptake by pharmacokinetic modeling using arterial blood sampling and hybrid positron emissiontomography and magnetic resonance (PET/MR) imaging in primary prostate cancer.


This observational prospective study was approved by our Institution's Ethics Committee. Eleven patients with clinically significantprostate cancer underwent a 60-minute dynamic PET/MR scan of the pelvis with an injected dose of Ga-68 HBED-CC-PSMA (Ga-68PSMA-11). Simultaneously, axial T1 Dixon, T2 and diffusion-weighted MR images were acquired. Arterial blood activity wasmeasured by an automatic arterial blood sampling device during the first 10 min. Manual blood samples at time points 3, 7, 15, 25,40 and 60 min were collected for metabolite analysis and for blood to plasma transformation to derive an arterial input function.Time-activity curves of lesion, prostate and muscle were generated and mean standardized uptake values (SUVmean) calculated.An irreversible two-tissue compartment model, with rate constants K1, k2 and k3, were fitted to the data and the net influx rateKi=K1k3/(k2+k3) was calculated. Ki was correlated to SUVmean, patient data and MR parameters.

RESULTS

In total 13 lesions located in the prostate were identified. Ga-68 PSMA-11 was stable in-vivo, not necessitating metabolitecorrection. The mean plasma-to-blood ratio was 1.63, stable over time. The kinetics could be described by an irreversible two-tissue compartment model. K1, k3 and Ki were all significantly higher in lesion compared to normal tissue (p<0.05): mean K1 inlesion, prostate and muscle 0.086, 0.063 and 0.018 mL/min/mL, mean k3 in lesion, prostate and muscle 0.075, 0.033 and 0.034 min-1 and mean Ki in lesion, prostate and muscle 0.031, 0.011 and 0.003 min-1. Ki showed strong correlation with SUVmean (Spearmanrho 0.92, p<0.001). There was no significant correlation between Ki and patient data and MR parameters (p>0.060).

RC211-17 Localization and Restaging of Carcinoma Prostate by 68Ga PSMA PET/CT in Patients with BiochemicalRecurrence: A Descriptive Study


RC211-18 68Ga-PSMA PET/PSMA for Prostate Cancer Staging Correlated to Prostatectomy Specimen


CONCLUSION

The kinetics of Ga-68 PSMA-11 can be described by an irreversible two-tissue compartment model. SUVmean showed strongcorrelation with Ki and can be used in clinical practice to quantify Ga-68 PSMA-11 uptake.


Pharmacokinetic modeling is the gold standard for PET quantification. SUV strongly correlates with net influx rate Ki and cantherefore be used in clinical practice to quantify Ga-68 PSMA-11 uptake.

ParticipantsNikhil Seniaray, New Delhi, India (Abstract Co-Author) Nothing to DiscloseRitu Verma, New Delhi, India (Presenter) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVidur Mahajan, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseSudhir Khanna, New Delhi, India (Abstract Co-Author) Nothing to DiscloseEthel S. Belho, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVanshika Gupta, New Delhi, India (Abstract Co-Author) Nothing to DiscloseAnkur Pruthi, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Prostate cancer is the most common solid cancer in men. Following definitive treatment of prostate cancer by radicalprostatectomy (RP) or radiotherapy, cancer recurrence is heralded by an increase in serum prostate-specific antigen (PSA) which iscalled biochemical recurrence. We investigate the relationship between prostate specific antigen (PSA) level and detection ofsuspected cancer recurrence using 68 Ga-PSMA PET/CT in patients with biochemical recurrence after radical prostatectomy (RP) orradiotherapy.


We analyzed retrospective data of 150 men with carcinoma prostate post RP and post radiotherapy with biochemical recurrencefrom May 2014 to Jan 2018 by 68 Ga-PSMA PET/CT We included men with suspected recurrent prostate cancer based on anelevated post treatment PSA level. The data collected analyzed the relationship of the pre-scan PSA level to the probability of apositive scan finding for recurrent prostate cancer.

RESULTS

Our cohort included 150 men, all had adenocarcinoma of prostate, 126/150 had a previous RP and 24/150 had prior radiotherapy.The mean PSA of the RP group was 4.8 ng/mL and 22.8 ng/mL in the radiotherapy group. In the post RP cohort, the detection rateof 68 Ga-PSMA PET/ CT was 39.3% for PSA 0.2 to <0.5 ng/mL, 45.3% for PSA 0.5 to <1 ng/mL, 88.2% for PSA 1 to <2 ng/mL and95.5% for PSA >=2. Lymph node metastasis post RP was identified in 52% of men with suspected disease recurrence. In the postradiotherapy cohort the detection rate was 96.1 % for PSA 2 to 4 ng/mL, 99.2% for PSA 4 to 6 ng/ mL and 100% for PSA >=6.Local recurrence after radiotherapy was present in 62 % of the cohort and 58 % had lymph node metastasis.

CONCLUSION

68Ga-PSMA PET/CT provides a novel imaging modality for the detection of prostate cancer recurrence and metastasis. SuspectedPSMA avid metastatic lesions are common and are identified at low post treatment PSA levels, which if detected will help directappropriate salvage treatments.


PSMA PET/CT should be considered a routine part of follow-up of treated prostate cancer patients since metastasis may presentwith low PSA levels leading to delay in addressing relapses.

ParticipantsThais Mussi, MD,PhD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseMarcelo L. Cunha, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRonaldo H. Baroni, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the accuracy of 68Ga-PSMA PET/PSMA for prostate cancer staging using prostatectomy specimen as gold standard forlocal staging.


IRB approved retrospective study. We reviewed our database from February 2016 to February 2018 and found 214 patients whohad submitted to a 68Ga-PSMA PET/PSMA and had no prior prostatectomy. 162 patients were excluded because had noprostatectomy in our hospital and eight had not performed lymphadenectomy. A total of 44 patients were included in the study.

RC211-19 Panel Discussion


RESULTS

Histology in 16, 17, 5 and 6 patients were ISUP 2, 3, 4 and 5, respectively. From the 44 patients, six had lymph node uptake andten had uptake in bone lesions suggestive of benignity/fibrous dysplasia. On pathology, tree patient had lymph nodes metastasis,not seen on PET/PSMA. Sensitivity for lymph node was 62% and specificity was 95%.

CONCLUSION

68Ga-PSMA PET/PSMA showed high specificity for lymph node metastasis and low specificity for bone lesions.


68Ga-PSMA PET/PSMA is an advance in a new imaging method for prostate cancer staging, with high level of specificity for lymphnode metastasis.

ParticipantsNancy M. Swanston, RT, Houston, TX (Presenter) Nothing to DiscloseEphraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Presenter) Research support, Blue Earth Diagnostics Ltd; Research support,Advanced Accelerator Applications SASteve Cho, MD, Madison, WI (Presenter) Imaging Endpoints; General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Assess the appropriate clinical indications for 18F-fluciclovine PET and understand the diagnostic accuracy of 18F-fluciclovinePET for local and metastatic prostate cancer. 2) Develop 18F-fluciclovine PET protocols for image optimization. 3) Apply the correct18F-fluciclovine PET interpretation for local and metastatic prostate cancer and benign physiologic variants.

ABSTRACT

Anti-1-amino-3-[18F]-flurocyclobutane-1-carboxylic acid (18F-fluciclovine) is a non-naturally occurring amino acid PET radiotracerthat is recently United States Food and Drug Administration approved for detection of suspected recurrent prostate cancer. Thetumor imaging features of this radiotracer mirror the upregulation of transmembrane amino acid transport that occurs in prostatecancer due to increased amino acid metabolism for energy and protein synthesis. This refresher course provides an overview of18F-fluciclovine PET diagnostic accuracy for identifying primary and metastatic disease, as well as proper 18F-fluciclovine PETimaging protocols. Correct interpretation criteria will be explored in detail to identify physiologic and pathologic 18F-fluciclovineuptake patterns and potential pitfalls.

MSAS22

Evolving Imaging Methods for the Cancer Patient - Part 2 (Sponsored by the Associated Sciences Consortium)(Interactive Session)


NM OI


ParticipantsNancy McDonald, MS, Chicago, IL (Moderator) Nothing to DiscloseWilliam A. Undie, PhD, RT, Houston, TX (Moderator) Nothing to DiscloseBernie McKay, BS, Chicago, IL (Presenter) Employee, Triad IsotopesKatie Tucker, BS, Chicago, IL (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Improve basic knowledge and skills relevant to new imaging procedures and treatments for the cancer patient. 2) Assess thepotential applications to clinical practice and the treatments to these patients. 3) Assess the potential of new radiopharmaceuticalsto imaging and the treatment of the cancer patient. 4) How Nuclear Medicine is integrated with other modalities in the treatment ofthe cancer patient.

NM208-SD-MOA1

Is there a Correlation Between Levels of PSA and SUVmax Values in Patients Evaluated with[68Ga]PSMA PET/CT to be Applied in Prostatic Cancer Patient Follow Up and as a Prognosis Factor?

Station #1

NM209-SD-MOA2

18F-Alfatide PET/CT in Assessment of Locoregional Lymph Nodes in Thoracic Esophageal SquamousCell Cancer: Comparison with 18F-FDG PET/CT

Station #2

NMS-MOA

Nuclear Medicine Monday Poster Discussions

Monday, Nov. 26 12:15PM - 12:45PM Room: NM Community, Learning Center

NM



ParticipantsPeter S. Conti, MD, PhD, Los Angeles, CA (Moderator) Nothing to DiscloseJacob G. Dubroff, MD, PhD, Philadelphia, PA (Moderator) Nothing to Disclose

Sub-Events


ParticipantsErika S. Fajardo, MD, Mexico City, Mexico (Presenter) Nothing to DiscloseDigna Pachuca Gonzalez, MD, Tlalpan, Mexico (Abstract Co-Author) Nothing to DiscloseJuan Pablo Chavez-Torres, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseBlanca K. Gonzalez, MD, Mexico City, Mexico (Abstract Co-Author) Nothing to DiscloseRodrigo Hernandez Ramirez, Leon, Mexico (Abstract Co-Author) Nothing to DiscloseLuis Felipe Alva Lopez, MD, Mexico City, Mexico (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The purpose of this study was to find out if a correlation exist between levels of PSA and SUVmax values in patients evaluated with[68Ga]PSMA PET/CT with prostate cancer, for the application of new methods in the follow-up of these patients.


We evaluated fifty five patients with diagnosis of prostate cancer. These patients were scanned using Ga68-PSMA for eitherstaging or biochemical recurrence. A retrospective, cross-sectional and descriptive study was carried out with the objective ofcorrelating levels of PSA with SUVmax values.

RESULTS

55 patients with a known diagnosis of prostatic carcinoma were evaluated using Ga68-PSMAPET Imaging. The patients enrolledbetween years 2015-2018 were analyzed. The mean age of the study population was 69 years (range 49-84) with a meanprostate-specific antigen (PSA) level of 85.69 ng/ml (range 0.01-3850.90). SUVmax mean found was 12.41 (range 1.9-56.5). Theindication for the study was initial staging in 10 patients (18.2%) and to confirm biochemical relapses in 45 (81.8%). The diseasesites at the moment of the study was locoregional in 22(40%), distant metastasis 12 (21.8%), and both 21 (38.2%). It seemsthere is a correlation between PSA/SUVmax values given in our statistical results. Further studies needed. Values of PSA wereresume in groups, One (0-5ng/ml), two (5.1-10ng/ml), and three (>10ng/ml), as the same for SUVmax in three groups, one(0-10),two(10-20), three (>20). A correlation was made between PSA groups and SUV groups finding that low values of SUVmax didcorrelate with low levels of PSA, with 23 patients with PSA in group one, had a SUVMAX between 0-10. We also correlate the PSAgroups with the extension of the disease, finding that the PSA group one had only locoregional disease. However, PSA group threehad either loco regional as distant disease in this study.

CONCLUSION

The [68Ga] PSMA PET / CT has had a boom in the evaluation of biochemical recurrence in prostate cancer, with our results, theremay be a correlation of PSA levels with SUVmax values, which could have a prognostic application. No literature specificallymentioning this correlation was found, so studies are needed to support it.


[68Ga] PSMA PET / CT have a high sensitivity and specificity to locate disease sites in prostate cancer, with the use of SUVmax,which being correlated with PSA values have great relevance in the prognosis of the disease.

Participants

NM210-SD-MOA3

Correlation of Simultaneously-Acquired SUV of 18FDG and Apparent Diffusion Coefficient in SoftTissue and Bone Tumors Using Voxel-By-Voxel Analysis

Station #3

Yinjun Dong, Jinan, China (Presenter) Nothing to DiscloseShuanghu Yuan, PhD, Jinan, China (Abstract Co-Author) Nothing to DiscloseYuchun Wei JR, Jinan, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

18F-FDG PET/CT has gained acceptance for staging of esophageal cancer. However, FDG is not tumor specific and false-positiveresults may occur by accumulation of FDG in benign tissue. The tracer RGD might not have these drawbacks. The aim of this studywas to investigate the feasibility of novel RGD tracer 18F-Alfatide (18F-ALF-NOTA-PRGD2) PET/CT for the detection of locoregionallymph node (LN) metastases of esophageal squamous cell cancer (ESCC) and to compare 18F-Alfatide PET/CT with 18F-FDGPET/CT


62 patients with thoracic ESCC underwent 18F-Alfatide PET/CT (n=29) and 18F-FDG PET/CT (n=33) scans before surgery. Thetotal number of removed LNs and total number of positive LNs (verified by pathologic finding), including their location, wererecorded and standard uptake values (SUVs) of the resected LNs were measured on the 18F-Alfatide PET/CT and 18F-FDG PET/CTimages, respectively

RESULTS

52 patients underwent successful surgery, and pathologic examination confirmed nodes positive for metastasis and 99 of 1018(51/436 for 18F-Alfatide group, 48/582 for 18F-FDG group) excised nodal groups. The sensitivity, specificity, accuracy, positivepredictive value (PPV), and negative predictive value (NPV) of 18F-Alfatide were 92.2% (47/51 nodal groups), 89.9% (346/385),90.1% (393/436), 54.7% (47/86), and 98.9% (346/350), respectively, whereas those of 18F-FDG were 93.8% (45/48), 79.0%(422/534), 80.2% (467/582), 28.7% (45/157), and 99.3% (422/425), respectively. P values were 0.757, <0.001, <0.001, <0.001,and 0.522, respectively.

CONCLUSION

In general, 18F-Alfatide PET/CT improves the specificity, accuracy and PPV than 18F-FDG PET/CT in the assessment of locoregionalLNs in thoracic ESCC, and these results could provide an important impact to the field of precision medicine.


18F-Alfatide PET/CT can be used for detecting locoregional LNs in thoracic ESCC and provide an important impact to the field ofprecision medicine

ParticipantsYuji Watanabe, MD,PhD, Kobe, Japan (Presenter) Nothing to DiscloseDaiki Shinyama, RT, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseSungtak Hong, PhD, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseKoji Sagiyama, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseKeisuke Ishimatsu, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseHiroshi Honda, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseRyotaro Kamei, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the relationship between simultaneously-acquired SUV of 18FDG and ADC in soft tissue and bone tumors using voxel-by-voxel analysis.


The consecutive 125 patients with tumor of soft tissue and bone were included in this study, and examined with 18FDG-PET/MRhybrid imaging. Diffusion-weighted image (b = 0 and 800) and 18FDG-PET were simultaneously acquired with 3D-T2-weighted TSEimages. After image co-registration and reslice of 3D-T2WI and PET image based on ADC map on a workstation (Intellispace Portal6.0), voxel-based scatter plots of SUV vs ADC and SUV/ADC vs log-ADC were generated for each tumor. Cluster analysis using k-means clustering algorithm was also applied when a tumor consists of multiple components such as liposarcoma, schwannoma andother tumors with large necrotic portions by using our in-house program (MathWorks). Pearson correlation coefficient was comparedbetween SUV vs ADC and SUV/ADC vs log-ADC. The slope of regression line was also compared among malignant, intermediate andbenign groups. Statistical analysis was performed with Wilcoxon test, ANOVA and Mann-Whitney test.

RESULTS

According to the WHO classification 2013, malignancy grades of the 125 tumors were histologically classified into high-gradesarcomas (n=76), intermediate-grade (n=12) and benign tumor (n=37). The relationship between SUV/ADC and log-ADC showed asignificant inverse linear correlation (mean r=-0.60, 95%CI: -0.63~-0.57), while that between SUV and ADC showed no significantcorrelation (mean r=-0.05, 95%CI: -0.10~0.00). All the tumors showed significant higher correlation coefficient for SUV/ADC vslogADC than SUV vs ADC (p<0.001). The slope was much steeper for malignant than benign tumors (p<0.001). In 24 patients withliposarcoma (n=2), myxofibrosarcoma (n=3), schwannoma (n=2), other sarcomas (n=17), cluster analysis demonstrated steep slopeof the two clusters showed higher inverse correlation coefficients than that of the whole tumor and represented the tumor grades.

CONCLUSION

Voxel-by-voxel SUV-ADC analyses could demonstrate significant inverse linear correlation between SUV/ADC and log-ADC, and

NM211-SD-MOA4

Elucidating Mechanisms of Acute Sports Concussion with PET/MRI

Station #4

NM142-ED-MOA6

Clinical Pediatric PET/MR Program: A Guide to Successful Implementation

Station #6

Voxel-by-voxel SUV-ADC analyses could demonstrate significant inverse linear correlation between SUV/ADC and log-ADC, andmalignant tumors showed steeper slope of regression line than benign tumors.


The slopes derived from SUV/ADC vs log-ADC scatter plots could be a new quantitative imaging biomarker for the differentiation ofmalignant and benign soft-tissue and bone tumors.

ParticipantsPaul Vaska, Stony Brook, NY (Presenter) Nothing to DiscloseMichael J. Salerno, BS,MBA, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseKenneth T. Wengler, MS, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseDavid Ouellette, MS, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseShouyi Wei, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseBrian Cruickshank, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseDavid E. Komatsu, PhD, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseDinko Franceschi, MD, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseLev Bangiyev, DO, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseJanet Oseni, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseEmmanuel Thomas, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseSara Yang, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseMark E. Schweitzer, MD, Stony Brook, NY (Abstract Co-Author) Consultant, MMI Munich Medical International GmbH; Data SafetyMonitoring Board, Histogenics Corporation Kristen Dams-O'Connor, New York, NY (Abstract Co-Author) Nothing to DiscloseJames M. Paci, MD, Setauket- East Setauket, NY (Abstract Co-Author) Nothing to DiscloseXiang He, PhD, Stony Brook, NY (Abstract Co-Author) Consultant, Endo International plc


[email protected]

PURPOSE

Animal studies have defined a neuropathological cascade following concussion consisting of large and transient disruptions ofhomeostasis in the acute phase (hours) but this has never been observed in humans. We aim to establish the existence,magnitude, and regional distribution of the most prominent feature of this cascade - a dramatic yet temporary increase in glucosemetabolism and related hemodynamic effects.


Athletes from a single NCAA Division I university consented to undergo PET/MRI brain imaging in the event of concussion, using awithin-subjects (paired) design in which subjects are imaged as soon as possible after injury and again when symptoms haveresolved (>3 months later). We also recruited a non-injured control group from the same population. A 20 min PET/MRI scancommenced 40 min after i.v. injection of 3-5 mCi of FDG. MRI included standard structural scans and quantitative measures ofcerebral blood flow (CBF), oxygen extraction fraction (OEF), and cerebral metabolic rate of oxygen (CMRO2). Images wereregistered to standard brain atlas and regions of interest extracted. Clinical data regarding severity of injury and duration ofrecovery was collected as well.

RESULTS

Thus far we have completed acute phase scans on 4 subjects within 54 hours of concussion injury and a scan of a control athlete.Clinically, there were no functional or structural abnormalities. We observed that gray matter SUV values in the acute phase werehigher than control in all cases except for a subject with extensive history of concussions (including one within the last year).Gray-to-white matter ratios were elevated in all cases. MRI measures demonstrated that acute concussion subjects generally hadlower CMRO2, CBF, and OEF in gray matter vs control. Due to the limited number of subjects scanned so far, statistical significancehas not yet been established.

CONCLUSION

Initial results are consistent with the hypothesis that cortical glucose metabolism is elevated and oxidative metabolism is depressedin the acute phase following sports concussion. Ongoing imaging of study participants will determine the statistical significance andclinical relevance of these findings.


A better understanding of the early pathology of concussion may lead to improved strategies for objective diagnosis and treatment.

ParticipantsSandra Saade-Lemus, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseHansel J. Otero, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseSabah Servaes, MD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseJeffrey P. Schmall, PhD, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseLisa J. States, MD, Plymouth Mtng, PA (Presenter) Nothing to Disclose


[email protected]

TEACHING POINTS

1. To provide a guide for planning and implementation of a pediatric PET/MR program 2. To describe our experience, steps

necessary and timeline for designing a PET/MR program including staffing, protocol optimization, clinical implementation and referringphysicians outreach 3. To provide practical tips and examples for protocol design, imaging acquisition and interpretation learnedover the first year and over 100 cases of our PET/MR program 4. To highlight advantages of PET/MR over PET/CT and Whole BodyMRI - After reviewing this exhibit, readers will be able to recognize the necessary steps for providing PET/MR services to children.


1. Why PET/MR? Background 2. Clinical pediatric PET/MR experience at out institution a. PET/MR Whole Body Workflow b. PET/CT vsPET/MR pilot study (N=9) c. Patient sedation (N=30) 3. Clinical Pediatric PET/MR Program a. Phase 1 (Planning): Staffing, protocoldesign, workflow setup b. Phase 2 (Preclinical): Pilot studies, order sets, billing codes c. Phase 3 (Clinical): Scheduling, imageacquisition, interpretation d. Phase 4 (post implementation): Expanded applications, additional staff training 4. Challenges ofPET/MR 5. Will PET/MR replace PET/CT? 6. Clinical examples 7. Conclusion

NM214-SD-MOB2

Treatment of Castration-Resistant Prostate Cancer with Radium-223 Dichloride: A Clinical Experience

Station #2

NM215-SD-MOB3

Comparing F18-Fluorodeoxyglucose Positron Emission Tomography with MRI (F18-FDG PET-MRI),F18-FDG PET-CT and MRI Imaging for Staging and Restaging of Head and Neck Cancer (HNC)

Station #3

NMS-MOB

Nuclear Medicine Monday Poster Discussions

Monday, Nov. 26 12:45PM - 1:15PM Room: NM Community, Learning Center

NM


ParticipantsPeter S. Conti, MD, PhD, Los Angeles, CA (Moderator) Nothing to DiscloseJacob G. Dubroff, MD, PhD, Philadelphia, PA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsCatherine M. Doyle, BS, Tampa, FL (Presenter) Nothing to DiscloseMatthew Mills, BS, Tampa, FL (Abstract Co-Author) Nothing to DiscloseSultan Damgaci, MSc, Tampa, FL (Abstract Co-Author) Nothing to DiscloseJingsong Zhang, MD,PhD, Tampa, FL (Abstract Co-Author) Nothing to DiscloseMayer Fishman, MD,PhD, Tampa, FL (Abstract Co-Author) Nothing to DiscloseGhassan El-Haddad, MD, Tampa, FL (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The benefits of incorporating Radium-223 dichloride (Ra-223) in the treatment of castration-resistant prostate cancer (CRPC) hasnot yet been fully delineated in real life setting. Therefore, we evaluated the outcome of patients with CRPC who were treated withRa-223 in terms of side effects, pain improvement, bone scan response (BSR), overall survival (OS) and progression free survival(PFS).


A total of 114 patients with CRPC and bone metastases referred for treatment with Ra-223 between March 2010 and February 2018were identified for retrospective analysis. Clinical characteristics, treatments, and outcomes were extracted from clinical chartreview and radiologic bone scans. Survival rates and survival curves were generated using Kaplan-Meier analysis. Chi-square andindependent student t test were used for comparison of categorical variables. Multivariate analysis (MVA) Cox proportional hazardratios (HR) model was used to assess OS and PFS.

RESULTS

Of the 114 patients referred for treatment, 107 received at least one dose of Ra-223, 56 receiving full treatment (6 doses). MedianOS was 12.6 months for the entire cohort. In multivariate analysis, improved OS was significantly associated with the following:completion of treatment (p<0.001), baseline PSA<30 ng/mL (0.013), ALP decline after Ra-223 (p=0.003), and no priorchemotherapy (p=0.014). In addition to the previously listed factors, univariate analysis demonstrated significance for baselinealkaline phosphatase (ALP)<200 U/L (p=0.016) and prior use of Sipuleucel-T (p=0.019).

CONCLUSION

The key factors notable for significant prolongation in OS for patients with CRPC who received Ra-223 may be helpful when decidingwhich patients would benefit most from this therapy. The potential synergistic effect between cellular immunotherapy and Ra-223warrant further evaluation in a prospective study.


Use of Ra-223 for treatment of CRPC has potential to significantly increase OS in patients with certain key pre-treatment factors,including a possible synergistic effect with cellular immunotherapy.

ParticipantsHossein Mehdikhani, MD , New York, NY (Presenter) Nothing to DiscloseMaria Habib, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseSomali Gavane, MBBS, New York, NY (Abstract Co-Author) Nothing to DiscloseBrett Miles, MD, New York, NY (Abstract Co-Author) Nothing to DisclosePeter M. Som, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseLale Kostakoglu, MD, MPH, New York, NY (Abstract Co-Author) Research Consultant, F. Hoffmann-La Roche Ltd


NM216-SD-MOB4

Whole-body Staging Concept of Patients with Temporal Lobe Epilepsy Using Hybrid [(18) FDG] -PET/MRI

Station #4

[email protected]

PURPOSE

Our objective was to compare the sensitivity and specificity of F18-FDG PET/MRI (Siemens, Biograph mMR) with F18-FDG PET/CT(Siemens, Biograph mCT) and dedicated MRI for detection of HNC at both staging and restaging.


A total of 94 HNC patients, 24 for initial staging and 70 for restaging, underwent sequential whole body PET/CT and integratedPET/MRI of head and neck after a single injection of FDG as well as a dedicated head and neck MRI (79 with and 15 withoutgadolinium contrast). Using a 5-point scale, all none-lymph node (nLN) and lymph node (LN) lesions were classified into likelybenign, probably benign, indeterminate, probably malignant and likely malignant on each modality. The first 2 and the last 3categories were combined as negative and positive groups for malignancy, respectively. Histopathology or follow-up imaging resultswere used for final diagnosis of malignant versus benign etiology. Sensitivity and specificity of each modality were calculated. Inaddition, beam hardening artifacts and the SUVmax of lesions were compared between PET/CT and PET/MRI.

RESULTS

A total of 223 lesions (104 nLNs and 119 LNs) were evaluated in 94 HNC patients (57 squamous cell carcinoma and 37 othermalignancies). Sensitivity and specificity of distinguishing malignant from benign were 89% and 47% for PET/CT, 70% and 82% forMRI and 85% and 84% for PET/MRI, respectively. Beam-hardening and metal artifact was significantly less on PET/MRI compared toPET/CT. There was high correlation of SUVmax between PET/CT and PET/MRI (spearman correlation coefficient 0.96, P<0.001) withsystematically higher values on PET/CT by an average of 2.05 units (95% CI 1.63 - 2.47). The optimal SUVmax cutoff todifferentiate malignant vs benign lesions were 9.6 for PET/CT and 7.2 for PET/MRI at initial staging. The corresponding value in thefollow-up cases was 5.8 for PET/CT and 4 for PET/MRI.

CONCLUSION

This study suggest that integrated PET/MRI has the advantage of combining the sensitivity of PET with specificity of MRI and isassociated with less beam-hardening artifact compared to PET/CT. However, more studies with larger patient populations arerequired to prove the accurate clinical utility of PET/MRI.


Integrated PET/MRI is potentially superior to PET/CT and MRI alone and can be considered as a one-stop shopping for staging andrestaging of HNC.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Lale Kostakoglu, MD, MPH - 2012 Honored Educator

ParticipantsCornelius Deuschl, Essen, Germany (Presenter) Nothing to DiscloseTheodor Ruber, Bonn, Germany (Abstract Co-Author) Nothing to DiscloseLeon Ernst, Bonn, Germany (Abstract Co-Author) Nothing to DiscloseChristoph Moenninghoff, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseThorsten D. Poeppel, Essen, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseMichael Forsting, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SAChristian Elger, Bonn, Germany (Abstract Co-Author) Nothing to DiscloseLale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AG


[email protected]

PURPOSE

Simultaneous PET/MRI enables morphologic and metabolic imaging with an excellent co-registration in a single examination,rendering this hybrid imaging modality available for an easy and broad clinical application 1. The purpose of this study is to evaluatethe diagnostic impact of hybrid 18F-FDG PET/MRI in the diagnostic work up of patients with temporal lobe epilepsy using thesimultaneous imaging modality for a cerebral and whole-Body imaging concept.


Nineteen patients with temporal lobe epilepsy were enrolled in this prospective study (mean age: 39,1 years, range: 19-76 years,14 female). All patients underwent a dedicated PET/MRI protocol of the brain (T1-MPRAGE, 3D-FLAIR, STIR cor, T2 ax and SWI)and the whole body (T2 ax., DWI, T1 VIBE ax. with and without ce). Image analyses were performed by a radiologist and a nuclearmedicine physician in consensus reading for MRI (1), and fused PET/MR datasets (2) for potential epileptogenic lesions of the brainand potential neoplasms in whole-body staging. Diagnostic confidence was evaluated on a modified Likert scale.

RESULTS

All examinations were obtained successfully without any relevant artifacts. Based on morphologic MR readings, 15 out of the 19patients were found to show subtle changes of the temporal lobe, whereas PET/MRI showed changes in 17/19 patients in PET/MRI.

NM143-ED-MOB5

The Many Faces of Parathyroid Adenoma - Approach to Multimodality Imaging in Challenging Cases

Station #5

NM003-EB-MOB

Signs and Artifacts in Amyloid PET Imaging

Hardcopy Backboard

Whole-Body staging revealed neoplasms in 2 of the 19 patients, that were identified by MRI and PET/MRI (Figure 1). Based on thefused image analysis the diagnostic confidence was rated higher (mean 2.6) when compared to sole morphologic reading (2.3).

CONCLUSION

This study shows hybrid 18-F-FDG PET/MRI feasible for brain and whole body staging of patients with temporal lobe epilepsy


Hybrid PET/MRI might have the potential as a valuable imaging tool for the diagnostic work-up for patients with temporal lobeepilepsy. Literatur1. Jadvar H, Colletti PM. Competitive advantage of PET/MRI. Eur J Radiol 2014;83:84-94

ParticipantsMichael N. Pakdaman, MD, Los Angeles, CA (Abstract Co-Author) Nothing to DisclosePeyman Kangavari, MD, Los Angeles, CA (Presenter) Nothing to DiscloseCatherine Evans, MD, Los Angeles, CA (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

Obtain a diagnostic approach to evaluation of parathyroid lesions, including adenoma, using mutlimodality imaging including nuclearstudies with correlation to MRI and CT. Gain insight on special approaches to identification of parathyroid lesions on marechallenging and equivocal cases, including algorithms for next step imaging.


Overview of parathyroid anatomy and physiology as it relates to pathogenesis parathyroid adenoma Review of cross-modalityimaging characteristics of Parathyroid Adenoma Correlate imaging characteristics of parathyroid adenoma with diseasepathophysiology Imaging modalities to be reviewed include neck CT, neck MRI, PET/CT, US, Nuclear Thyroid Scan, and NuclearParathyroid Scan Present approaches to differentiating thyroid versus parathyroid adenoma using these modalities Review ofinteresting cases from our institution


ParticipantsTamara Lundeen, MD, Tucson, AZ (Presenter) Nothing to DisclosePhillip Kuo, MD,PhD, Tucson, AZ (Abstract Co-Author) Author, MD Training at Home; Research Grant, Astellas Group; Consultant,Endocyte, Inc; Consultant, General Electric Company; Education Grant, General Electric Company; Speakers Bureau, Eli Lilly andCompany; Consultant, inviCRO, LLC; Consultant, Imaging Endpoints; Consultant, Progenics Pharmaceuticals, IncMatthew Covington, MD, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseNaghmehossadat Eshghi, MD, PhD, Tucson, AZ (Abstract Co-Author) Nothing to DiscloseJohn Seibyl, MD, New Haven, CT (Abstract Co-Author) Researcher, inviCRO, LLC

TEACHING POINTS

1. Discuss the appropriate use criteria for amyloid PET imaging. 2. Explain the algorithmic approach to scan interpretation. 3.Introduce the common amyloid PET imaging signs and artifacts.


Introduction to amyloid PET imaging. -Discuss radiopharmaceutical properties of the three FDA approved PET amyloid agents -Review the current existing appropriate use criteria for amyloid PET imaging - Review manufacture-specific recommendations foramyloid PET interpretation. Present a detailed algorithmic approach to amyloid PET scan interpretation. - Discuss a novel region-based search pattern to aid with amyloid PET interpretation - Provide examples of normal and abnormal amyloid PET scans. Providean atlas of regional amyloid PET imaging signs in normal and abnormal brains. - Example amyloid PET scans with MRI correlation ofthe temporal/occipital, frontal, parietal and striatal regions. Illustrate common imaging artifacts with example images. Conclusionand summary table with the composite atlas of imaging signs.

SSE17-01 18F-Fluciclovine PET Evaluation of Recurrent High-Grade Glioma

Monday, Nov. 26 3:00PM - 3:10PM Room: S505AB

SSE17

Nuclear Medicine (Central Nervous System Nuclear Imaging)


NR NM

AMA PRA Category 1 Credit ™: 1.00ARRT Category A+ Credit: 1.00


ParticipantsAmir H. Khandani, MD, Chapel Hill, NC (Moderator) Consultant, Progenics Pharmaceuticals, Inc; Consultant, F. Hoffmann-La RocheLtd; Yonglin Pu, MD, PhD, Chicago, IL (Moderator) Nothing to Disclose

Sub-Events

ParticipantsAkinyemi A. Akintayo, MD, Atlanta, GA (Presenter) Nothing to DiscloseEphraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Abstract Co-Author) Research support, Blue Earth Diagnostics Ltd; Researchsupport, Advanced Accelerator Applications SAMarc D. Benayoun, MD, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseIjeoma Ibeanu, MD,MPH, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJeffrey Olson, PHD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseConstantinos G. Hadjipanayis, MD, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseDaniel J. Brat, MD, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseVikram Adhikarla, PhD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJonathon Nye, PhD, Atlanta, GA (Abstract Co-Author) Consultant, Lantheus Medical Imaging, Inc David M. Schuster, MD, Decatur, GA (Abstract Co-Author) Institutional Research Grant, Nihon Medi-Physics Co, Ltd; InstitutionalResearch Grant, Blue Earth Diagnostics Ltd; Institutional Research Grant, Advanced Accelerator Applications SA; Consultant,Syncona Ltd; ; ; Mark M. Goodman, PhD, Atlanta, GA (Abstract Co-Author) Royalties, Nihon Medi-Physics Co, Ltd


[email protected]

PURPOSE

18F-Fluciclovine has shown promise in the detection and diagnosis of high-grade gliomas (HGG) due to minimal uptake in normalbrain parenchyma and increased uptake in neoplastic tissue. The goal of this study is to evaluate 18F-fluciclovine PET uptake inpatients with suspected recurrent HGG previously treated with chemotherapy and/or radiotherapy and correlate with overallsurvival.


Nine patients with suspected recurrent HGG (WHO grade III, IV) previously treated with surgical resection followed by eitherchemotherapy and/or radiotherapy underwent dynamic 18F-fluciclovine brain PET. Average 18F-fluciclovine dose was 10.5 ± 0.41mCi (390 ± 15 MBq). Semi-quantitative PET analysis (SUVmax, SUVmean) was performed for each lesion identified on standard ofcare MRI and compared to normal brain parenchyma and venous blood at all time points to obtain time activity curves. Metabolictumor volume of each lesion was measured at each time point using a threshold of 1.6* normal brain parenchyma. True recurrencewas confirmed with histopathological confirmation in 5 patients and subsequent serial MRI examinations.

RESULTS

18F-Fluciclovine uptake greater than background was visually identified in 6 of 9 patients. Average SUVmax and SUVmean ofidentifiable lesions were 5.7±1.2 and 1.4±0.5 at 60 minutes respectively. Average SUVmax and SUVmean for normal brain uptakewere 1.8±0.6 and 0.5±0.2 respectively. Range of glioma SUVmax/normal SUVmean was 32.5 - 6.6 with an average ratio of15.1±10.4. 18F-Fluciclovine PET identified a new distinct local metastasis in a patient that was confirmed on subsequent MRIexaminations. Recurrent gliomas that were not visually identified by 18F-fluciclovine PET had an average SUVmax of 2.1±0.6 with aglioma SUVmax/normal SUVmean average of 6.1±1.7. The mean time of survival from the 18F-fluciclovine PET with identifiablerecurrent HGG was 14.9 months vs 24.1 months for those lesions not visually identified on PET.

CONCLUSION

18F-fluciclovine PET/CT is a promising diagnostic tool for identifying recurrent HGG and may be a valuable tool for survivalprognostication although more work is needed to verify these results.


18F-fluciclovine PET/CT is a promising diagnostic tool for identifying recurrent HGG and may be a valuable tool for survival

SSE17-02 PET-MR Imaging Biomarkers Improving Differential Diagnosis Between Progression andRadionecrosis of Brain Tumors


SSE17-03 Detection of Abnormal Brain FDG-PET Images With Deep Learning


prognostication.

ParticipantsNadya Pyatigorskaya, MA, Paris, France (Presenter) Nothing to DiscloseMarc Bertaux, Paris, France (Abstract Co-Author) Nothing to DiscloseBrian Sgard, Paris, France (Abstract Co-Author) Nothing to DiscloseLydia Yahia-Cherif, Paris, France (Abstract Co-Author) Nothing to DiscloseDamien P. Galanaud, MD, PhD, Paris, France (Abstract Co-Author) Research Consultant, Olea MedicalMarie-Odile Habert, MD, Paris, France (Abstract Co-Author) Nothing to DiscloseMarine Soret, Paris, France (Abstract Co-Author) Nothing to DiscloseDidier Dormont, MD, Paris, France (Abstract Co-Author) Nothing to DiscloseAurelie Kas, Paris, France (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Follow-up under treatment of patients with high grade glioma is essential, the MRI being the modality of choice. However,anatomical MRI may not be always reliable after radiation or chemotherapy. Advanced MRI techniques as well as PET were proposedfor improving the diagnostic accuracy, that remains still moderate and variable across studies. Our purpose was to evaluate thediagnostic accuracy of PET-MRI for differential diagnosis between tumor progression and radionecrosis.


Between December 2015 and September 2017, patients followed for primary malignant brain tumors underwent FDOPA PET-MRI. Theacquisitions were performed with a 3T PET-MR system (SIGNA, GE Healthcare). The MRI acquisition included SE 3D T1-weightedimages without and after contrast injection, 3D FLAIR imaging, DWI, pseudo-continuous arterial spin labeling (pCASL) and dynamicsusceptibility-contrast (DSC) perfusion. The SUVmax, SUVmean, and SUVpeak were measured in each lesion with Volume ofInterest (VOI). A region of interest was drawn in each lesion and the mean rADC, rCBV, and rCBF for both DSC and pCASL perfusionwere calculated. In addition, the visual analysis was performed.

RESULTS

Forty-four patients were included. ROC analysis showed good discrimination between progression and radionecrosis with a gooddiagnostic accuracy for SUVmax (0.82), for SUVpeak (0.9) and for ASL rCBF (0.86). It was fair for rADC (0.63) and rCBV (0.75). Alogistic regression model found among predictor variables, the combination of these SUVpeak and pCASL rCBF variables improvedsensitivity (0.94), specificity (0.83), the AUC (0.97, 95% CI=[0.93,0.99]) and the accuracy (0.94). Visual analysis allowed adiagnostic accuracy of 0.77 for PET reading only, of 0.89 for PET reading with morphological MRI and of 0.98 for PET-MRI combinedreading.

CONCLUSION

We have observed an increase in diagnostic accuracy for combined analysis of PET and MRI biomarkers for both qualititative andqualitative biomarkers, SUVpeak and pCASL rCBF being the most significant quantitative biomarkers. Combined PET-MR imagingallows to increase the diagnostic accuracy for differential diagnosis between tumor progression and radionecrosis in neuro-oncology.


The combined analysis of imaging morphological, functional and metabolic markers on PET-MRI is helpful in differential diagnosisbetween tumor progression and radionecrosis in neuro-oncology.

ParticipantsTomomi Nobashi, MD,PhD, Palo Alto, CA (Abstract Co-Author) Nothing to DiscloseClaudia Zacharias, MD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric CompanyGuido A. Davidzon, MD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseHeying Duan, Stanford, CA (Presenter) Nothing to Disclose

PURPOSE

FDG-PET/CT is widely used in routine clinical practice and its utilization is projected to increase. The brain is often included as partof the study; however, due to its background physiological FDG uptake, the sensitivity for abnormality detection in this region isusually low. By utilizing a deep learning algorithm to aid radiologists detect abnormalities, clinical management and outcomes couldbe improved. The aim of this study was to evaluate the ability of a new deep learning framework to discern between normal andabnormal FDG uptake in the brain.


285 FDG-PETs acquired between 2007 and 2017 were retrospectively reviewed. A deep learning framework was trained using 110normal and 110 abnormal brain studies, including 10 studies for testing in each category. The remaining 36 normal and 29 abnormalstudies were used for validation of the resulting inference model and its sensitivity and specificity were analyzed. DICOM studieswere anonymized, appropriately windowed and converted into portable network graphics format. A network architecture that uses atime distributed 2D convolutional neural network with 100 epochs was generated. A classification was performed based on theprobability of an individual FDG-PET scan being normal or abnormal. Various models were derived.

SSE17-04 Posterior Cortical Variant Alzheimer's Disease and Lew Body Dementia: Similarities and Differences onFDG PET Scan


SSE17-05 Correlation of 4´-[methyl-11C]-thiothymidine Uptake with Human Nucleoside Transporter andThymidine Kinase-1 Expressions in Patients with Newly Diagnosed Gliomas


RESULTS

Accuracy and loss function of the optimal trained model were calculated at 0.761 and 0.462, respectively. Receiver operatingcharacteristic (ROC) curve demonstrated an area under the curve of 0.832 (Figure 1). According to ROC curve, the optimalprobability threshold to detect abnormal Brain FDG-PET scans was 0.661. Validation test characteristics resulted in sensitivity of80.6% and specificity of 75.9%.

CONCLUSION

Preliminary results of a novel deep learning model showed promising capability in detecting brain abnormalities on FDG-PET imageswhich could aid radiologists and improve clinical outcomes.


Improving detection of brain FDG-PET abnormalities in daily clinical practice with the aid of a deep learning method that could helpimprove clinical management and outcomes.

ParticipantsRitu Verma, New Delhi, India (Presenter) Nothing to DiscloseRajeev Ranjan, New Delhi, India (Abstract Co-Author) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVidur Mahajan, MBBS, New Delhi, India (Abstract Co-Author) Nothing to DiscloseVanshika Gupta, New Delhi, India (Abstract Co-Author) Nothing to DiscloseEthel S. Belho, New Delhi, India (Abstract Co-Author) Nothing to DiscloseNikhil Seniaray, New Delhi, India (Abstract Co-Author) Nothing to DiscloseAnkur Pruthi, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Posterior cortical atrophy (PCA) is a form of dementia considered to be an atypical variant of Alzheimer's disease (AD) and Dementiawith Lewy bodies (DLB) is a type of posterior dementia characterized by fluctuating levels of cognition, changes in behavior, visualhallucinations with accompanying extrapyamidal motor symptoms. We attempt to identify specific core areas on FDG PET imagingwhich are common to both and also establish the differences which may be helpful to differentiating the two.


We retrospectively analysed of 30 patients with clinically suspected posterior dementia. All the subjects underwent F-18 FDG PETCT scan of the brain and the studies were analyzed both qualitatively (visually) and semi-quantitatively. The subjects hadundergone dopamine transporter imaging with Tc 99 m TRODAT 1 on a prior date. The subjects were divided into possible PCA withTRODAT scan normal (n=10) and possible DLB with abnormal (n=20). The FDG uptake patterns were recorded and areas of corticalhypometabolism in the cerebral cortex that were two standard deviations from the mean were considered as abnormal.

RESULTS

All the subjects had an abnormal pattern of F-18 FDG uptake on PET scan, both on visual inspection and semiquantitative analysis.Bilateral parieto-occipital hypometabolism was consistently found in all the subjects. Hypometabolism in precuneus, posteriorcingulate and the cortex around the angular gyrus was present in all the subjects of PCA with relative sparing of the medialoccipital cortices. DLB subjects showed variable degrees of involvement of the medial occipital cortices with relative sparing ofposterior cingulate and precuneus.

CONCLUSION

FDG PET scan can act as a non-invasive diagnostic modality in differentiating the two posterior cortical dementias despitesignificant clinical and imaging overlap.


We present features based on which FDG-PET can be used to diagnose and differentiate Posterior Cortical Atrophy (PCA) andDementia with Lewy Bodies (DLB) in the clinical practice.

ParticipantsYasukage Takami, Mikicho, Japan (Presenter) Nothing to DiscloseYuka Yamamoto, MD, PhD, Kagawa, Japan (Abstract Co-Author) Nothing to DiscloseMasaki Ueno, Mikicho, Japan (Abstract Co-Author) Nothing to DiscloseYoichi Chiba, Mikicho, Japan (Abstract Co-Author) Nothing to DiscloseTakashi Norikane, Kita-gun, Japan (Abstract Co-Author) Nothing to DiscloseTetsuhiro Hatakeyama, Kagawa, Japan (Abstract Co-Author) Nothing to DiscloseKeisuke Miyake, Mikicho, Japan (Abstract Co-Author) Nothing to DiscloseJun Toyohara, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseYoshihiro Nishiyama, MD, Kagawa, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

SSE17-06 A Novel Clustering Approach in Brain Tumors Using Dynamic 18F-FET PET/MRI


We examined expressions of four human nucleoside transporters including: human equilibrative nucleoside transporters (hENT1 andhENT2); and human concentrative nucleoside transporters (hCNT1 and hCNT3), and thymidine kinase-1 (TK1), the key enzyme in4'-[methyl-11C]-thiothymidine (4DST) phosphorylation, to elucidate the mechanism of 4DST uptake in patients with newlydiagnosed gliomas.


A total of 19 patients with newly diagnosed gliomas were examined with 4DST PET. Tumor lesions were identified as areas of focallyincreased uptake, exceeding that of normal brain background. For semi-quantitative analysis, tumor-to-contralateral normal braintissue (T/N) ratio was determined by dividing the maximal standardized uptake value (SUV) for tumor by that of the mean SUV forreference tissue. The expressions of hENT1, hENT2, hCNT1, hCNT3, and TK1 in tumor specimens were examined byimmunohistochemistry and compared with 4DST T/N ratio.

RESULTS

All but two gliomas showed focally increased 4DST uptake. All but one grade II glioma that was not visualized with 4DST PETshowed hENT1 staining. Of the gliomas, hENT2 and hCNT3 staining was observed in 11 and 16 gliomas, respectively. No hCNT1staining was observed in any of the gliomas. All but two gliomas that were not visualized with 4DST PET showed TK1 staining. Asignificant correlation was observed between T/N ratio and hENT1 score (r = 0.75, p < 0.001). There was no significant correlationbetween T/N ratio and hENT2 score or hCNT3 score. There was a significant strong correlation between T/N ratio and TK1 score (r= 0.92, p < 0.001). It is likely that expression of TK1 might be more important than expression of hENT1 for uptake of 4DST.

CONCLUSION

Results of this preliminary study indicate that expressions of hENT1 and TK1 appear to be important determinants of 4DST uptake innewly diagnosed gliomas.


Expressions of hENT1 and TK1 appear to be important determinants of 4DST uptake in newly diagnosed gliomas.

ParticipantsCristina Campi, PhD, Padova, Italy (Abstract Co-Author) Nothing to DiscloseStefano De Marchi, Padova, Italy (Abstract Co-Author) Nothing to DiscloseMariagiulia Anglani, Padova, Italy (Abstract Co-Author) Nothing to DiscloseValentina Bodanza, Padova, Italy (Abstract Co-Author) Nothing to DiscloseGiancarlo Gorgoni, Negrar, Italy (Abstract Co-Author) Nothing to DiscloseMatteo Salgarello, Vicenza, Italy (Abstract Co-Author) Nothing to DiscloseGiuseppe Lombardi, Padova, Italy (Abstract Co-Author) Nothing to DiscloseFranco Bui, MD, Padova, Italy (Abstract Co-Author) Nothing to DiscloseDiego Cecchin, MD, Padova, Italy (Presenter) Nothing to Disclose


[email protected]

PURPOSE

Purpose of the present study is to present a user-independent, quantitative approach to easily classify brain areas on the base ofuptake temporal evolution using dynamic 18F-FET PET/MRI.


O-(2-18F-fluoroethyl)-L-tyrosine (18F-FET) PET/MRI performed at out Institution in brain tumors were retrospectively analyzedusing a clustering technique in order to differentiate areas using uptake dynamics: time activity curves (TACs) were considered foreach voxel in the brain volume and an unsupervised clustering algorithm was applied. This algorithm provides as output anautomatic grouping of TACs and hence of voxels they are associated to. Therefore, it is possible to create parametric images,representing the different behavior over time of the tissues.

RESULTS

We were able to automatically identify brain areas grouped by dynamic similarities using 18F-FET uptake and produce parametricimages (see Figure 1, panel (a)) and associated mean time-activity curves obtained averaging the TACs of voxels belonging to thesame group (see Figure 1, panel (b)).

CONCLUSION

The proposed approach allows to exploit dynamic 18F-FET PET data: the automatic nature of the method removes the user-dependent ROI drawing step, the clustered data can be further analyzed to extract representative features of the average TACsand intuitive parametric images are produced. Moreover the method could be easily employed with other tracers.


Validation of the method against pathological data is ongoing and seems to provide encouraging results. The method could then beused to plan areas for radiotherapy clustering areas with similar dynamic characteristics.

SSE21-01 Whole-Body Magnetic Resonance Imaging in Newly Diagnosed Langerhans Cell Histiocytosis Patients:Lesion Detectability and Risk Stratification

Monday, Nov. 26 3:00PM - 3:10PM Room: E353B

SSE21-02 The Complementary Role of Early Interim PET Semi-Quantitative Metrics to Lugano Criteria inPredicting Treatment Response in Children with Hodgkin Lymphoma


SSE21

Pediatrics (Oncology and Nuclear Medicine)


MR NM OI PD



ParticipantsHeike E. Daldrup-Link, MD, Palo Alto, CA (Moderator) Nothing to DiscloseEthan A. Smith, MD, Cincinnati, OH (Moderator) Travel support, Koninklijke Philips NV

Sub-Events

ParticipantsJisun Hwang, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseHee Mang Yoon, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJeong Rye Kim, MD, Cheonan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseAh Young Jung, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYoung Ah Cho, MD,PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJin Seong Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To compare lesion detectability and the accuracy of risk stratification of skeletal survey, bone scan, and whole-body magneticresonance imaging (WB-MRI) in patients with newly-diagnosed Langerhans cell histiocytosis (LCH).


Patients who presented with newly-diagnosed LCH and who underwent skeletal survey, bone scan, and WB-MRI (n=46) betweenJune 2011 and April 2017 were retrospectively included. The sensitivity and mean number of false-positives per patient in the threewhole-body imaging modalities (skeletal survey, bone scan, and WB-MRI) were assessed. Risk stratification was performed in eachpatient for each whole-body imaging modality. The reference standard for LCH lesions was histopathologic findings or clinical andimaging follow-up. The ability to detect LCH lesions and the accuracy of the initial risk stratification were compared between thethree whole-body imaging modalities.

RESULTS

WB-MRI had significantly higher sensitivity (99.0%; 95% confidence interval [CI], 93.2-99.9%) than skeletal survey (56.6%; 95%CI, 46.7-66.0%; p<0.0001) and bone scan (38.4%; 95% CI, 29.4-48.3%; p<0.0001) in the detection of LCH lesions, and therewere no significant differences in the number of false-positives per patient (p>0.017 for all comparisons). WB-MRI tended to havehigher accuracy for the risk stratification than skeletal survey and bone scan (concordance rate of 0.98, 0.91, and 0.83,respectively), although the differences between imaging modalities were not statistically significant (overall p-value 0.066).

CONCLUSION

WB-MRI had higher detectability for LCH lesions than skeletal survey and bone scan, while the three whole-body imaging modalitieshad comparable accuracy in the initial risk stratification of LCH.


WB-MRI might serve as a primary imaging modality for risk stratification in patients with initially diagnosed LCH.


ParticipantsEman E. Marie, MD, MSc, Toronto, ON (Presenter) Nothing to DiscloseAmer Shammas, MD, Toronto, ON (Abstract Co-Author) Nothing to Disclose

SSE21-03 Value of Superb Microvascular Imaging in Biopsy Site Selection of Pediatric Solid Tumor


Oscar M. Navarro, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseAngela Punnett, MD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseReza Vali, MD, Toronto, ON (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate if adding iPET semiquantitative metrics to Lugano criteria will increase its predictability for response to therapy.


112 children with newly diagnosed Hodgkin lymphoma between January 1, 2007 and January 1, 2017 underwent PET beforetreatment, after 2 cycles, and at the end of first line treatment. The diagnostic test accuracy of semi-quantitative metrics resultswere compared to Lugano criteria for assessment of treatment response at iPET time point to predict the final response totreatment. The standard of reference for treatment response was based on clinical assessment, imaging results at end of first lineof treatment, and histopathological confirmation of residual masses. The group of patients with positive iPET as per Lugano criteriahad been re-classified into positive and negative iPET as per semiquantitative metrics cut offs.

RESULTS

Using the Lugano criteria, 41 patients were positive and 71 were negative at iPET with accuracy of 70 %.Using semi-quantitativemetrics, the number of positive cases decreased to 34 cases with a higher accuracy of 75%. On further analysis of the 41 positiveiPET patients detected by Lugano classification, the number of positive cases decreased to 23 after applying our proposed cut offsfor each of the semi-quantitative metrics. Thus,18 patients were reclassified as good responders according to semiquantitativeanalysis.

CONCLUSION

Semiquantitative metrics proved to be more accurate than Lugano criteria in prediction of treatment response at interim timepoint.Patients with positive iPET results according to Lugano classification may require additional evaluation using semi-quantitativemetrics that classify the patients into groups with different outcomes. Larger prospective cohort and longer duration follow up isneeded to support our claim.


Therapeutic protocols for Hodgkin lymphoma have recently employed a response adapted approach with de-escalation of therapyfor those patients with adequate interim response to avoid late effects of therapy. Definitions of adequate response have tendedto be conservative to avoid undertreatment for a disease with excellent cure rates. The addition of semi-quantitative analysis maybe helpful in therapy planning by identifying a larger pool of patients that may be eligible for de-escalation compared to the Luganocriteria.

ParticipantsYaqing Chen, PhD, Shanghai, China (Presenter) Nothing to DiscloseYunkai Zhu, MD, Shanghai, China (Abstract Co-Author) Nothing to DiscloseJun Jiang, Shanghai, China (Abstract Co-Author) Nothing to DiscloseShengli Gu, Shanghai, China (Abstract Co-Author) Nothing to DiscloseWenbin Guan, Shanghai, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To investigate the value of superb microvascular imaging (SMI) in biopsy site selection in children with solid tumor.


A total of 48 children with solid tumors scheduled for biopsy were recruited in this retrospective study with mean age 4.58±3.62y(0.08-15y). All patients underwent CDFI/ CDE and SMI to image tumor vascularity using Adler criteria for biopsy site selection.Ultrasound guided biopsy was performed targeting hyper-vascularity region on SMI within 3 days following ultrasound evaluation.Another 35 patients with mean age 5.6±3.2y (1.6-15y) underwent CDFI/ CDE targeted biopsy between July 2012 and March 2015were served as historical control. The sample adequacy of these two groups was compared retrospectively.

RESULTS

The Adler grade of 48 solid tumors using CDFI/ CDE was as follows: 2 tumors with grade 0 (4.2%, 2/48); 16 with grade I (33.3%,16/48); 16 with grade II (33.3%, 16/48) and 14 with grade III(29.2%,14/48). By applying SMI technology, no tumor was scoredwith grade 0, one patient with grade I (2.1%, 1/48), 8 with grade II (16.7%, 8/48) and 39 with grade III (81.3%, 39/48). SMI wasmore sensitive in imaging tumor vascularity than CDFI/ CDE (P<0.001). The pathologic diagnosis was confirmed in all 48 tumorsevaluated with SMI technology with sample adequacy of 100% (48/48). In comparison, the pathologic diagnosis was confirmed in31 tumors in control group with sample adequacy of 88.6% (31/35, P=0.025). The biopsy was sampled in necrosis regions in 4tumors in control group, and the final pathologic diagnosis was then confirmed by further surgery.

CONCLUSION

By increasing micro vessel display, SMI technology provides a practical solution to the problem of biopsy site selection in childrenwith solid tumor. The biopsy site selected by SMI can avoid sampling necrosis or fibrotic regions and thus improve sampleadequacy.

SSE21-04 Does Ferumoxytol Detect Joint Infiltration of Pediatric Cancer Patients?


SSE21-05 Ferumoxytol Does Not Impact SUV Values on PET/MR Scans



For children with unresectable solid tumor, biopsy pathology is the standard approach for further treatment decision. The biopsyguided by SMI can avoid sampling necrosis or fibrotic regions, and thus reduce false negative results.

ParticipantsAshok Joseph Theruvath, MD, Mainz, CA (Presenter) Nothing to DiscloseAnuj Pareek, MD, Aarhus C, Denmark (Abstract Co-Author) Nothing to DiscloseAnne M. Muehe, MD , Stanford, CA (Abstract Co-Author) Nothing to DiscloseSheri Spunt, Palo Alto, CA (Abstract Co-Author) Nothing to DiscloseHeike E. Daldrup-Link, MD, Palo Alto, CA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

During our MR imaging evaluations of pediatric cancer patients with the iron oxide nanoparticle compound ferumoxytol (Feraheme),we observed a surprising marked T1-enhancement of the joint effusion in some patients and not others. The purpose of this studywas to elucidate the underlying pathological mechanisms that lead to the observed ferumoxytol-induced indirect arthrographyeffect.


We retrospectively identified 13 pediatric cancer patients and young adults (mean age 17.1 ± 4.1, 8 males, 5 females) with bonesarcomas (n=12) and desmoid tumor (n=1) who had undergone MRI scans at 1 hour (n=8), 24 hours (n=6) or 48-120 hours (n=1)after intravenous injection of ferumoxytol at a dose of 5 mg Fe/kg body weight. The patients received a whole-body MRI with T1-weighted LAVA sequences, followed by a local MRI with T1-SE, T2-FSE and LAVA sequences. 6 of the 13 patients had alsoreceived a gadobutrol (Gadavist)-enhanced MRI. 8 of the 13 tumors were resected and underwent surgical and/or histopathologicalevaluation of joint invasion. The signal-to-noise ratio (SNR) of tumors with and without joint invasion was compared using a t-testand a p<0.05.

RESULTS

At 1 hour after ferumoxytol infusion, we did not observe enhancement of joint effusions and no significant difference in SNR(p>0.05), regardless of tumor infiltration status. At >24 hours post-contrast, four patients showed significantly increased SNRvalues (p=0.002) of the effusion compared to muscle as an internal standard on T1-weighted images. Two of these patients werediagnosed with joint infiltration histologically while two others who did not undergo resection showed signs of joint infiltration onimaging. The other three > 24-hour scans did not show significant enhancement of the effusion and showed no joint infiltration onhistology. Standard gadobutrol-enhanced MR images did not show this differential signal effect.

CONCLUSION

This pilot study suggests that ferumoxytol leak into a joint effusion might serve as an indirect indicator for joint infiltration inpediatric cancer patients. Further studies have to prove this finding in a larger cohort of patients.


Ferumoxytol-enhanced MRI can detect joint infiltration in pediatric cancer patients.


ParticipantsAnne M. Muehe, MD , Stanford, CA (Presenter) Nothing to DiscloseAshok Joseph Theruvath, MD, Mainz, CA (Abstract Co-Author) Nothing to DiscloseKetan Yerneni, Stanford, CA (Abstract Co-Author) Nothing to DisclosePraveen Gulaka, PhD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseAvnesh S. Thakor, MBBCHIR, PhD, Menlo Park, CA (Abstract Co-Author) Nothing to DiscloseHeike E. Daldrup-Link, MD, Palo Alto, CA (Abstract Co-Author) Nothing to Disclose

PURPOSE

Accurate measurement of standardized uptake values (SUV) of tumors and normal organs in positron emission tomography (PET) iscrucial for treatment response assessment of pediatric cancer patients. The iron oxide nanoparticle compound ferumoxytol can beused "off label" as a contrast agent for integrated PET/MR scans. However, ferumoxytol accumulation in the reticuloendethelialsystem (RES) could affect MR-based attenuation corrections of PET data. The purpose of our study was to compare SUV values ofnormal organs on ferumoxytol-enhanced and unenhanced 18F-FDG PET/MR scans.


In this IRB approved prospective study, 16 children (2-18 years) with malignant tumors underwent 18F-FDG PET/MR scans (dose 3MBq/kg) before chemotherapy with (n=8) or without (n=8) intravenous injection of ferumoxytol (5 mg Fe/kg). Patients whoreceived ferumoxytol were age- and sex-matched with patients who received unenhanced scans. MR attenuation correction wasobtained by a four-point Dixon LAVA sequence accounting for fat, air, water, and soft tissue (TR 4.2 ms, TE 1.1, 1.7, 2.3 ms, FA5). Anatomical correlation was obtained with a high-resolution LAVA (TR 4.4 ms, TE 1.1, 1.7, 2.2 ms, FA 15). For SUV mean valuesof normal organs three dimensional spherical region of interests were placed with MIM software over specified regions of the brain,

SSE21-06 Voxel-Based Volumetric Analysis of Cortical Asymmetry in Glucose Metabolism in the DevelopingHuman Brain


parotid gland, larynx, mediastinal blood pool, thymus, myocardium, liver, spleen, bone marrow, kidney and muscle and comparedwith a t-test.

RESULTS

The SUV mean values of patients with and without ferumoxytol-enhanced PET/MR scans were for brain (6.5 vs 6.4, p=0.96),parotid gland (1.3 vs 1.5, p=0.49), larynx (1.4 vs 1.6, p=0.14), mediastinal blood pool (1.4 vs 1.2, p=0.33), thymus (2.0 vs 2.0,p=0.76), myocardium (1.8 vs 1.7, p=0.44), liver (1.4 vs 1.6, p=0.67), spleen (1.7 vs 1.7, p=0.92), bone marrow (1.5 vs 1.8,p=0.12), kidney (2.0 vs 1.7, p=0.34), and muscle (0.5 vs 0.7, p=0.23). None of the organs or tissues showed a statisticalsignificant difference.

CONCLUSION

Ferumoxytol at a dose of 5 mg Fe/kg has no impact on the SUV mean values of normal organs on PET/MR.


Ferumoxytol can be used as a contrast agent in pediatric whole body PET/MR scans without impacting the SUV values of normaltissues.


ParticipantsAjay Kumar, MD, PhD, Detroit, MI (Presenter) Nothing to DiscloseV Pilli, Detroit, MI (Abstract Co-Author) Nothing to DiscloseJeong-Won Jeong, PhD, Detroit, MI (Abstract Co-Author) Nothing to DisclosePraneetha Konka, MBBS, Detroit, MI (Abstract Co-Author) Nothing to DiscloseHarry T. Chugani, MD, Bloomfield Hills, MI (Abstract Co-Author) Nothing to DiscloseCsaba Juhasz, Detroit, MI (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Clinical interpretation of cerebral positron emission tomography with 2-deoxy-2[F-18] fluoro-D-glucose (FDG-PET) images oftenrelies on evaluation of regional asymmetries. This study was designed to establish age-related variations in regional cortical glucosemetabolism asymmetries in the developing human brain.


FDG-PET scans of 58 children (age 1-18 years) were selected from a tertiary care single center pediatric PET database. All childrenhad a history of epilepsy, normal MRI, and normal pattern of PET glucose metabolism on careful visual evaluation. PET images wereanalyzed objectively by statistical parametric mapping with the use of age-specific FDG-PET templates. Regional FDG uptake wasmeasured in 35 cortical regions in both hemispheres using an automated anatomical labeling atlas, and left-to-right metabolic ratiosof homotopic regions were correlated with age, gender, and epilepsy variables.

RESULTS

Regional cortical glucose metabolism was mostly symmetric in young children and became increasingly asymmetric in older subjects.Specifically, several frontal cortical regions showed an age-related increase of left>right asymmetries, while right>left asymmetriesemerged in posterior cortex (including portions of the occipital, parietal and temporal lobe) in older children. Similar trends wereseen in a subgroup of 39 children with known right-handedness. Age-related correlations of regional metabolic asymmetries showedno robust gender differences and were not affected by epilepsy variables.

CONCLUSION

These data demonstrate a region-specific emergence of cortical metabolic asymmetries between age 1-18 years, with left>rightasymmetry in frontal and right>left asymmetry in posterior regions. These findings can facilitate correct interpretation of corticalregional asymmetries on pediatric FDG-PET images across a wide age range.


Our findings will help in understanding the regional neurometabolic evolution and its relationship with neurocognitive developmentand its impairment in various neurological conditions.

MSMI24A Assessing Tumor Therapeutic Response with FDG PET and CT: Practical Aspects

MSMI24B What You See May Not Be What You Think

MSMI24

Molecular Imaging Symposium: MI Case-based Discussion


MI NM


ParticipantsMunir Ghesani, MD, New York, NY (Moderator) Nothing to DiscloseHubert J. Vesselle, MD, PhD, Seattle, WA (Moderator) Consultant, MIM Software Inc

ABSTRACT

Molecular imaging is increasingly being utilized in the initial staging as well as in the follow up of various disease conditions. Inaddition, several new imaging tracers are either recently approved by FDA. As a result, this modality, while clinically useful, isbecoming increasingly complex.

Sub-Events

ParticipantsHubert J. Vesselle, MD, PhD, Seattle, WA (Presenter) Consultant, MIM Software Inc

LEARNING OBJECTIVES

1) Use the case-based approach to illustrate utility of molecular imaging in clinical applications and demonstrate how to avoid falsepositive interpretations.

ParticipantsMunir Ghesani, MD, New York, NY (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Recognize that molecular imaging in general is a very sensitive modality but its specificity is relatively lower. 2) Recognize thatthe specificity can be improved by various measures, such as obtaining detailed patient information, multimodality correlation withrecently performed imaging studies and careful comparison with prior imaging studies. 3) Recognize that improved specificity willimprove overall patient care by avoiding unnecessary procedures and treatments resulting from false positive interpretations. It willalso help in building referring physicians' confidence in the molecular imaging modality.

SPSI22A Addiction in America 2018

SPSI22B PET as a Tool in Investigating Addiction

SPSI22C fMRI in Addiction

SPSI22D MRS in Craving with a Focus on Alcoholism

SPSI22

Special Interest Session: Imaging Cognition 2018: Addiction

Monday, Nov. 26 4:30PM - 6:00PM Room: E353C

MR NR NM OT


ParticipantsDavid B. Hackney, MD, Boston, MA (Moderator) Nothing to DiscloseJody L. Tanabe, MD, Aurora, CO (Moderator) Nothing to Disclose

Sub-Events

ParticipantsDiana M. Martinez, MD, New York, NY (Presenter) Nothing to Disclose

ParticipantsDiana M. Martinez, MD, New York, NY (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Learn the basis of PET neurochemical imaging. 2) Learn the association between dopamine signaling and drug seeking behavior.3) Understand the correlation between striatal dopamine signaling and treatment response. 4) Understand the status of dopamineimaging across different types of addictions. 5) Understand the theory behind addiction as a habitual behavior and its imagingcorrelates.

ABSTRACT

The involvement of dopamine in addiction has its origins in studies investigating reward and reinforced behavior. Much of thisresearch has been explored in the human brain using Positron Emission Tomography (PET) imaging of striatal dopamine transmission.These studies show that addiction is associated with a decrease in dopamine D2/3 receptors and a decrease in pre-synapticdopamine release, and that this decrease occurs across different types of addiction, including cocaine, alcohol, and heroindependence. However, these imaging studies also show that, in cocaine abuse, blunted dopamine transmission is predictive ofcocaine seeking behavior. Low D2/3 receptor binding and low dopamine release are associated with the choice to self-administercocaine over alternative reinforcers, which can be viewed as a failure to shift between competing rewards. It is striking thataddiction to different substances of abuse are accompanied by the same alteration in neurobiology, independent of their primaryimpact on the dopaminergic system. Moreover, similar alterations of the dopaminergic transmission and D2-like receptor systemhave been described in psychiatric diseases other than addiction. Although these psychiatric disorders differ in theirphenomenology, they share a common deficit in reward-related behavior, particularly with respect to impulsivity and motivation.This presentation will describe the animal and human studies that link alterations in dopamine transmission and the D2 receptorswith impulsive and motivated behavior. The hypothesis that these alterations in dopamine transmission represent the neurobiologicalunderpinnings that facilitate impulsivity and undermine motivation, rather than the only the consequences of addiction itself, will bediscussed.

ParticipantsJody L. Tanabe, MD, Aurora, CO (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Describe the neural circuit associated with cravings in addictions. 2) Name 3 interventions shown to modulate craving and the"craving circuit.' 3) Identify 3 metrics that would be needed to implement an fMRI craving marker into addiction management.

ParticipantsJohn D. Port, MD, PhD, Rochester, MN (Presenter) Research Consultant, Biomedical Systems; Research Consultant, Neuronetics

LEARNING OBJECTIVES

1) Describe the theoretical reward circuitry of the human brain. 2) Explain how the reward circuitry homeostasis is altered in thesetting of addiction/alcoholism. 3) Discuss current MR spectroscopy findings in alcoholism/craving.

ABSTRACT

The human brain has a complex reward system that impacts much of human behavior. Over the last 60 years, numerous brain

SPSI22E Panel Discussion

The human brain has a complex reward system that impacts much of human behavior. Over the last 60 years, numerous brainregions have been identified in humans and other animals that together explain the different aspects of behavior and reward. Thecortico-basal ganglia-thalamo-cortical loop is perhaps the best known model of the human reward system, but many other areasparticipate in reward. Addiction is the process whereby a drug (alcohol, cocaine, etc) alters the homeostasis of the reward systemin a predictable way, thus creating addictive behaviors that are difficult to manage and reverse. This talk will first introduce thecurrent best model of the human reward system, with examples of its normal operation. Next, I will discuss the process of addition,focusing on how reward system homeostasis is disrupted in addiction, using alcoholism as an example. Finally, I will review the MRspectroscopic studies of alcoholism from the perspective of this altered reward system homeostasis, exploring the metabolicabnormalities in craving.

ParticipantsDavid B. Hackney, MD, Boston, MA (Moderator) Nothing to DiscloseJody L. Tanabe, MD, Aurora, CO (Moderator) Nothing to Disclose

ED010-TU

Nuclear Medicine Tuesday Case of the Day

Tuesday, Nov. 27 7:00AM - 11:59PM Room: Case of Day, Learning Center



TEACHING POINTS


MSCC31A Imaging of Dementia

MSCC31B Head and Neck PET/CT

MSCC31

Case-based Review of Nuclear Medicine: PET/CT Workshop-Head and Neck PET/CT (In Conjunction withSNMMI) (Interactive Session)

Tuesday, Nov. 27 8:30AM - 10:00AM Room: E450B

CT NR NM



ParticipantsSamuel E. Almodovar-Reteguis, MD, Orlando, FL (Director) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Director) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Moderator) Nothing to Disclose

Sub-Events

ParticipantsPhillip Kuo, MD,PhD, Tucson, AZ (Presenter) Author, MD Training at Home; Research Grant, Astellas Group; Consultant, Endocyte,Inc; Consultant, General Electric Company; Education Grant, General Electric Company; Speakers Bureau, Eli Lilly and Company;Consultant, inviCRO, LLC; Consultant, Imaging Endpoints; Consultant, Progenics Pharmaceuticals, Inc

LEARNING OBJECTIVES

1) Discuss appropriate use criteria for PET in the work-up of dementia. 2) Review imaging signs of amyloid deposition and pitfalls ofimage interpretation. 3) Illustrate a systematic approach to interpretation of PET imaging in dementia.

ParticipantsRathan M. Subramaniam, MD,PhD, Dallas, TX (Presenter) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue Earth DiagnosticsLtd


[email protected]

LEARNING OBJECTIVES

1) To review the best clinical practices of head and neck PET/CT in oncologic imaging. 2) To review the common sites of tumorlocation in head and neck and patterns of tumor spread. 3) To review pitfalls of head and neck PET/CT interpretations.

ABSTRACT

This talk will review the best clinical practices of head and neck PET/CT in oncology, patterns of tumor spread and common andunusual intepretation pitfalls.

RC311A SPECT/CT in Endocrine/Oncology

RC311B SPECT/CT Technology: State of the Art

RC311

Advances and Updates in SPECT/CT

Tuesday, Nov. 27 8:30AM - 10:00AM Room: S504CD

CT NM


Sub-Events

ParticipantsEsma A. Akin, MD, Washington, DC (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Through clinical case examples, this activity aims to refresh knowledge of SPECT-CT applications with emphasis onneuroendocrine disorders as well as parathyroid imaging.

ParticipantsTimothy Turkington, PhD, Durham, NC (Presenter) Consultant, Data Spectrum Corporation


[email protected]

LEARNING OBJECTIVES

1. After this course, the attendee should be able to describe the basic functioning of a conventional gamma camera. 2. After thiscourse, the attendee should be able describe three recent innovations of currently-available gamma cameras and SPECT/CTsystems.

RC318A Magnetic Resonance Imaging

RC318B Ultrasound

RC318C PET/CT

RC318

Challenging Cases in Body Oncologic Imaging (Interactive Session)

Tuesday, Nov. 27 8:30AM - 10:00AM Room: E353A

CT MR NM OI US


ParticipantsGary A. Ulaner, MD, PhD, New York, NY (Moderator) Research support, General Electric Company; Research support, F. Hoffmann-La Roche Ltd; Research support, Novartis AG


[email protected]

LEARNING OBJECTIVES

1) Learn how to correlate CT and FDG PET findings to optimize diagnosis. 2) Identify iatrogenic effects which mimic malignancy onFDG PET/CT. 3) Learn histologies of breast cancer which may not be appreciably FDG-avid.

GENERAL INFORMATION

This interactive session will use RSNA Diagnosis Live™. Please bring your charged mobile wireless device (phone, tablet or laptop) toparticipate.

Sub-Events

ParticipantsAlexander R. Guimaraes, MD, PhD, Portland, OR (Presenter) Consultant, Agfa-Gevaert Group

LEARNING OBJECTIVES

1) Updated understanding of soft tissue contrast mechanisms inherent in MRI including T1rho, diffusion weighted imaging, DCE-MRI.2) Updated protocols for each organ site. 3) Potential benefits of PET/MRI in diagnosing disease.

ABSTRACT

This course is designed to update the attendee on novel MRI techniques and the benefits of MRI in diagnosing challenging caseswithin the abdomen and pelvis. Multiparametric MRI offers the unique ability to monitor the tumor microenvironment. Increasingly,multiparametric MRI is used for diagnosis and grading of malignancy in various organ systems (e.g. prostate cancer).

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Alexander R. Guimaraes, MD, PhD - 2018 Honored Educator

ParticipantsDeborah J. Rubens, MD, Rochester, NY (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Understand the technical parameters to optimize to improve ultrasound diagnosis. 2) Identify discrete ultrasound features todiscriminate between various pathologic entities. 3) Characterize disease processes in solid organs, vessels and soft tissues usingthe unique features of ultrasound and appreciate how ultrasound is complementary to CT, MRI and PET in the oncology patient.

ABSTRACT

This session will highlight a variety of disease processes in the oncology patient using grayscale, color and spectral Dopplerultrasound. Technique and potential pitfalls will be highlighted as they contribute to diagnostic acumen of the sonologist. Cases willinclude neoplastic, infectious and vascular processes in multiple organs. Differential diagnosis will be stressed with companion caseexamples, as well as when to use comparative imaging such as CT, MRI or PET/CT

ParticipantsGary A. Ulaner, MD, PhD, New York, NY (Presenter) Research support, General Electric Company; Research support, F. Hoffmann-LaRoche Ltd; Research support, Novartis AG


[email protected]

LEARNING OBJECTIVES

1) Learn where CT findings can improve FDG PET interpretation and where FDG PET findings can improve CT interpretation.

ABSTRACT

FDG PET/CT has become an indispensible modality in the treatment of cancer. While proven to be of great clinical benefit in themanagement of a wide array of malignancies, there are many potential pitfalls which may be detrimental if not properly identifiedand explained. In particular, FDG-avidity may be incorrectly ascribed to malignancy when corresponding CT findings demonstratethe FDG-avidity to be benign. In other cases, the presence of FDG avidity correctly deterimes the presence of malignancy despiteto lack of correlate findings on CT. In this presentation, challenging FDG PET/CT cases will be used to demonstrate how correlationof FDG PET and CT findings leads to optimal FDG PET/CT interpretation.

MSCC32A Chest

MSCC32B Musculoskeletal

MSCC32

Case-based Review of Nuclear Medicine: PET/CT Workshop-Chest and Musculoskeletal PET/CT (InConjunction with SNMMI) (Interactive Session)

Tuesday, Nov. 27 10:30AM - 12:00PM Room: E450B

CH CT MK NM


ParticipantsSamuel E. Almodovar-Reteguis, MD, Orlando, FL (Director) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Director) Nothing to DiscloseDelphine L. Chen, MD, Saint Louis, MO (Moderator) Nothing to Disclose

Sub-Events

ParticipantsDavid M. Naeger, MD, San Francisco, CA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Describe the classic PET/CT appearance of various types of lung cancer. 2) Compare the imaging features that are similarbetween different types of lung cancer.

ABSTRACT

The classic PET/CT appearance of various types of lung cancers will be reviewed. Similaries and differences between differenttypes of lung cancer will be presented in an effort to help attendees interpret thoracic imaging with more confidence and be morehelpful in interdisciplinary settings.



[email protected]

LEARNING OBJECTIVES

1) Demonstrate how to integrate the FDG PET and CT components of an FDG PET/CT exam to distinguish benign and malignantosseous lesions. 2) Identify common benign causes of FDG-avidity in the musculoskeletal system.

SSG09-01 Dynamic PET Perfusion Imaging (DPPI) of Esophageal Cancer to Characterize Angiogenicity: A PhaseI Study to Explore this Potential Imaging Biomarker Enabled by Ultra-Fast Digital PET

Tuesday, Nov. 27 10:30AM - 10:40AM Room: S505AB

SSG09-02 Focal Liver Uptake on 18F-FDG PET/CT without CT Correlate: Utility of MRI Evaluation in Patientswith Known Malignancy

SSG09

Nuclear Medicine (Gastrointestinal Oncology Nuclear Imaging)

Tuesday, Nov. 27 10:30AM - 12:00PM Room: S505AB

CT GI MR NM OI



ParticipantsPeter L. Choyke, MD, Rockville, MD (Moderator) Nothing to DiscloseJacob G. Dubroff, MD, PhD, Philadelphia, PA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsMichael V. Knopp, MD, PhD, Columbus, OH (Presenter) Nothing to DiscloseEric D. Miller, MD, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseKatherine Binzel, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseTerence M. Williams, MD, PhD, Columbus, OH (Abstract Co-Author) Nothing to DisclosePeter Kneuertz, Columbus, OH (Abstract Co-Author) Nothing to DiscloseMichelle I. Knopp, Columbus, OH (Abstract Co-Author) Nothing to DiscloseChadwick L. Wright, MD,PhD, Lewis Center, OH (Abstract Co-Author) Nothing to DisclosePiotr J. Maniawski, MSc, Cleveland, OH (Abstract Co-Author) Employee, Koninklijke Philips NVJun Zhang, PhD, Columbus, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

While PET perfusion imaging using dynamic acquisitions has a long history, its use in clinical practice outside of cardiovascularapplications remains negligible. This Phase I study assesses the feasibility of dynamic PET perfusion imaging as an alternativemethodology to contrast enhanced MRI in esophageal cancer using next-generation digital PET technology with substantiallyimproved TOF timing resolution and reduced dead time.


FDG PET/CT was performed prior to radiation therapy of advanced esophageal cancer in 12 patients using a next-generation digitalphoton counting system (Vereos Philips, dPET). Prior phantom and preclinical experiments established the following protocol.Dynamic PET Perfusion Imaging (DPPI) was performed at the time of bolus injection of 5 mCi FDG over a volume of interest for 10min. PET events were recorded in continuous list mode acquisition and reconstructed using frame rates from 1 sec/fr to 15 sec/fr. Acount density adaptive reconstruction approach was previously developed. Data analysis were performed using IntellispaceDiscovery (Philips) workspace. The reconstructed dynamic PET images were assessed in blinded review by three experienced PETreaders. Descriptive statistics were calculated for data analysis.

RESULTS

All dynamic PET acquisitions were successfully acquired and reconstructed according to protocol. Frame rates of 10 sec led to highquality uptake time curves of the well delineated esophageal cancer. Count density adaptive reconstruction was essential tominimize noise. 2mm voxel length (HD) reconstruction was found to be beneficial. The first 2 min duration of the dynamic serieswere found to be the most relevant and also sufficient for the perfusion assessment. Uptake time curves were found analogous tocharacteristic findings in DCE-MRI.

CONCLUSION

Dynamic PET Perfusion Imaging (DPPI) of esophageal cancer was achievable using frame rates of 10 sec with acceptable quality forquantitative and visual assessment even at the low FDG dose of 5 mCi. A short table time of 5 min appears to be sufficientcombined with the dPET's low dose capabilities making clinical utilization feasible.


Dynamic PET perfusion imaging has the potential to be an alternative methodology to DCE-MRI or CT for assessment of perfusion ofesophageal cancer for therapy planning and response assessment.


SSG09-03 Focal Colonic Tracer Uptake in 18F-FDG PET/CT Scans: Does the Combined Analysis of MorphologicalChanges and PET Improve Lesion Characterization?


ParticipantsTetsuro Araki, MD, PhD, Boston, MA (Presenter) Nothing to DiscloseTarek H. Mouhieddine, Boston, MA (Abstract Co-Author) Nothing to DiscloseSreeharsha Tirumani, MBBS, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseAtul B. Shinagare, MD, Boston, MA (Abstract Co-Author) Advisory Board, Arog Pharmaceuticals, Inc; Research Grant, GTx, IncNikhil H. Ramaiya, MD, Jamaica Plain, MA (Abstract Co-Author) Nothing to DiscloseJyothi Priya Jagannathan, MD, Boston, MA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

There is insufficient evidence to guide management of focal FDG uptake in the liver without CT correlate. This study aims to assessthe utility of MRI for evaluation of focal FDG uptake on PET/CT without CT correlate in patients with known malignancy.


In this IRB-approved, HIPAA compliant retrospective study, between 2005 and 2012, out of 1851 patients who underwent FDG-PET/CT, we identified 36 patients with known malignancy (19 women, 17 men; mean age, 56.1 years) who had focal hepatic uptakeon PET/CT without CT correlate and had follow-up MRI within 100 days for assessment of the uptake. Two radiologists reviewedthe PET/CT images in consensus blinded to the area of uptake noting SUVmax of the lesion and background liver. MR images werethen reviewed to look for the presence of focal lesion corresponding to the uptake. When a focal lesion was present, the size,signal intensity and enhancement characteristics and follow-up imaging were documented. Statistical analysis was performed todetermine correlation between intensity of FDG uptake and presence of focal lesion.

RESULTS

A total of 50 sites of focal hepatic uptake without CT correlate were identified. Median SUVmax was 4.1 (range 2.1-10.1) andmedian SUVmax ratio of hepatic lesion/normal parenchyma was 1.3 (range 0.98-2.6). MRI confirmed focal lesion in 26/50 sites(52%). Median lesion size was 13 mm (range 3-30 mm). Among 26 hepatic lesions noted on MR imaging, 77% (20/26) werediagnosed as metastatic disease (6 with pathological confirmation, 14 based on follow-up image findings). Other 6 lesions werediagnosed as benign, including 3 with pathological diagnosis of nodular regenerative hyperplasia (n=2) and heterogenous hepaticsteatosis (n=1). There was no significant difference in the SUVmax of hepatic lesions (3.85 vs. 4.2, p=0.5) and the SUVmax ratio(1.32 vs. 1.31, p=0.97) between the groups with and without MRI correlate.

CONCLUSION

More than half of focal areas of uptake on PET/CT without CT correlate had a focal lesion on MRI in our study, and more thanthree-quarters of these MRI lesions were hepatic metastases, regardless of SUVmax values.


In patients with known malignancy, focal FDG uptake in the liver without CT correlate warrants further assessment with MRIregardless of SUVmax as it is likely to be metastasis.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Sreeharsha Tirumani, MBBS, MD - 2016 Honored EducatorNikhil H. Ramaiya, MD -2017 Honored EducatorAtul B. Shinagare, MD - 2017 Honored Educator

ParticipantsBenedikt M. Schaarschmidt, MD, Essen, Germany (Presenter) Stockholder, Bayer AG; Stockholder, General Electric Company;Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries LtdJulian Kirchner, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseOle Martin, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseLino Sawicki, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

To evaluate the impact of morphological information derived from contrast enhanced CT on the characterization of incidental, focalcolonic uptake in 18F-fluordesoxyglucose positron emission tomography / computed tomography (18F-FDG PET/CT) examinations.


In this retrospective study, 125 patients (female: n=53, male: n=72, mean age 62.9y) that underwent 18F-FDG PET/CT andcolonoscopy (mean time between colonoscopy and PET: 52±58d) were included. By two readers, PET/CT examinations wereassessed for incidental, focal colonic tracer uptake in comparison to the background. Focal tracer uptake was correlated withmorphological changes of the colonic wall such as focal thickening, polypous shape or focal contrast media uptake. Discrepancieswere resolved in a consensus reading. Then, colonoscopy reports were evaluated for precancerous (oligo-tubular adenoma,serrated adenoma, sessile adenoma and tubulo-villous adenoma) and cancerous lesions verified by histopathology, serving as areference standard. Imaging findings were then compared to the reference standard to calculate sensitivity, specificity as well aspositive (PPV) and negative predictive values (NPV) for focal tracer uptake in PET alone as well as focal tracer uptake in PET and

SSG09-04 Clinical Workflow of PET/MR for Primary Staging of Rectal Cancer


SSG09-05 PET/MR for Staging Rectal Cancer: A Comparison to Conventional Staging with Pelvic MR andThoracoabdominal CT


morphological changes in contrast enhanced CT in the detection of precancerous and cancerous lesions.

RESULTS

In 38.4% (48/125) of all patients, focal tracer uptake was observed in 67 lesions. In the corresponding colonoscopy reports, a totalof 10 cancerous and 26 precancerous lesions were found. In PET, two cancerous lesions and 10 precancerous lesions were missed,resulting in a sensitivity, specificity, PPV and NPV of 51%, 68%, 29% and 84% for focal tracer uptake in PET alone. By correlationfocal tracer uptake with morphological changes in contrast enhanced CT, a sensitivity, specificity, PPV and NPV of 48%; 79%;37,5%; 86% for precancerous and cancerous lesions could be observed.

CONCLUSION

Focal colonic tracer uptake has a low specificity and sensitivity for precancerous and cancerous lesions. By analyzing additionalmorphological changes in contrast enhanced CT imaging in 18F-FDG PET/CT examinations, the specificity can be increased withoutsacrificing sensitivity.


Morphological changes in 18F-FDG-PET/CT examinations help to decide, which patients with focal tracer uptake will profit fromadditional colonoscopy.

ParticipantsMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseCinthia D. Ortega, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFelipe R. Ferreira, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMaira M. Bezerra, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRoberto Blasbalg, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFernanda C. Capareli, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseSergio Nahas, MD, PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGiovanni G. Cerri, MD,PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

analyze the clinical workflow of a PET/MR for rectal cancer staging, regarding MR imaging protocol, total scan time, prevalence anddistribution of metastasis.


55 patients with rectal adenocarcinoma were submitted to whole-body PET/MR for primary staging. MR sequences and total scantime were recorded. One reader analyzed the MR (blinded to PET findings) for locoregional staging. Another reader (blinded to thehigh-resolution MR of the pelvis) analyzed the PET/MR according to T, N and M staging. A third blinded reader analyzed only theliver MRI. Standard of reference was biopsy when feasible or imaging follow-up.

RESULTS

PET/MR workflow consisted of a dedicated pelvic MR, including high-resolution T2w and DWI, followed by whole-body PET/MR fromhead to mid thigh, ending with a T2w and DWI of the liver. Any dedicated MR sequence was used for the thorax. The mean totalacquisition time of PET/MR was 64 minutes, ranging from 53 to 90 minutes. The PET acquisition time was on average 17 minutes,ranging from 12 to 25 minutes and the only-MR time was 52 minutes on average, including the dedicated pelvic and the liversequences. One patient has not completed the exam due to claustrophoby and one patient was recalled due to low imaging quality(movement artifacts). For primary lesion detection, PET and DWI were positive in 54 out of 55 patients (one patient was negativeeven at MR). Metastatic disease was observed by PET/MR in 21 (38.2%) patients, mainly in the liver (12/21, 57.1%) non-regionallymph node (10/21, 47.6%), lungs (7/21, 33.3%) and peritoneum (1/21, 4.8%). Liver DWI detected 64 lesions in 12 patients andPET detected 62 lesions in the same 12 patients. Any brain metastasis was found.

CONCLUSION

PET/MR for rectal cancer staging is feasible, with a tolerable scan time and showing a high incidence of synchronic metastasis. DWIof the liver and of the rectum could be omitted in a staging/detection setting. PET coverage could start from skull base instead ofvertex. Optimization of MR protocol would decrease scan time and allow the inclusion of other MR sequences.


PET/MR for rectal cancer may be used clinically as a one-stop-shop imaging tool for whole-body staging.

ParticipantsMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseCinthia D. Ortega, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFelipe R. Ferreira, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMaira M. Bezerra, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRoberto Blasbalg, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFernanda C. Capareli, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose

SSG09-06 Diagnostic Performance of PET/MR in Identifying High-Risk Primary Rectal Cancer Patients


Sergio Nahas, MD, PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGiovanni G. Cerri, MD,PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To compare the detection rate of metastatic lesions on PET/MR versus conventional staging (pelvic MR and thoracoabdominal CT)in rectal cancer patients referred for primary staging.


Ninety-five patients with biopsy proven rectal adenocarcinoma were submitted to whole-body PET/MR for primary staging inaddition to conventional staging with pelvic MR and thoracoabdominal contrast-enhanced CT (ceCT). One reader analyzed the MR(blinded to PET findings) regarding locoregional staging. Another reader (also blinded to the high-resolution part of pelvic MR)analyzed the PET/MR about T-, N- and M-stage. A third blinded reader assessed the M-stage by ceCT. Standard of reference wasbiopsy when feasible or imaging follow-up. Lesions were cateogrized as positive, negative or indeterminate.

RESULTS

On a lesion-based analysis, PET/MR detected 24.5% more positive metastatic lesions than conventional imaging (305 vs. 254lesions). The higher detection rate on PET/MR was observed for non-regional lymph nodes (73 vs. 37 lesions). On a patient-based,PET/MR was positive in 38 out of 95 patients (40%), while conventional staging in 24 out of 95 patients (25.3%). PET/MR wasindeterminate in 7 patients (7.4%), which 1 was positive on CT (non-regional lymph nodes, later confirmed to be sarcoidosis), whileconventional imaging was in 30 patients (31.6%), which 10 were positive on PET/MR, mainly for non-regional lymph nodes and liver.

CONCLUSION

For rectal cancer staging, PET/MR presents not only a higher detection rate of metastatic lesions than conventional imaging (pelvicMR and thoracoabdominal ceCT), but also clarifies the etiology indeterminate lesions.


PET/MR might be preferred for whole-body staging of rectal cancer patients, with a higher detection rate and potential clinicalimpact.

ParticipantsMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Presenter) Nothing to DiscloseCinthia D. Ortega, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFelipe R. Ferreira, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMaira M. Bezerra, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRoberto Blasbalg, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFernanda C. Capareli, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseSergio Nahas, MD, PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGiovanni G. Cerri, MD,PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

to assess the diagnostic performance of PET and MR parameters to identify high-risk rectal cancer patients.


fifty-three patients with biopsy proven rectal adenocarcinoma were submitted to whole-body PET/MR for primary staging. Onereader analyzed the MR (blinded to PET findings) regarding T stage, mesorectal fascia involvement, extramural vascular invasion,and presence of locoregiona nodal disease. Another reader (also blinded to the high-resolution part of pelvic MR) measured thesemiquantitative PET parameters using a dedicated software (PETVCAR, GE Healthcare) and assessed the M stage. Standard ofreference was biopsy when feasible or imaging follow-up. Mann-Whitney test was performed to compare PET parameters betweenhigh and low-risk patients. ROC analysis was also used to define the diagnostic perfomance of PET parameters in identifying thetumor risk.

RESULTS

PET-volumetric parameters, namely TLG and MTV, performed better than SUVmax and mean to distinguish low and high-riskpatients. MTV (48.9 vs. 18.4, p = 0.006) and TLG (629.4 vs. 190.4, p = 0.008), but not SUVmax (23.4 vs. 19.2, p = 0.306) orSUVmean (11.7 vs. 9.9, p = 0.285) were significantly higher in high-risk patients. Patients with advanced T stage, positive regionalnode and metastatic disease also presented significantly higher MTV and TLG values. No difference was found in any PETparameters in patients according to the involvement of mesorectal fascia and presence of EMVI. The AUC to distinguish low andhigh-risk patients was 77.9% (p = 0.006) for MTV and 77.0% (p = 0.008) for TLG and the best cut-off value (sensitivity of 83.7%and specificity of 70%) of MTV and TLG to detect high-risk patients was 16.1 and 172.8, respectively.

CONCLUSION

In addition to MR adverse risk factors, the volumetric-based PET parameters (MTV and TLG) allow the identification of high-riskpatients, which could tailor therapy management.

SSG09-07 PET/MR Characterization of Mucinous versus Nonmucinous Components of Rectal Adenocarcinoma:A Comparison of Tumor Metabolism and Cellularity


SSG09-08 Comparative Accuracy of Qualitative and Quantitative 18F-FDG PET/CT Analysis in Detection ofLymph Node Metastasis from Anal Cancer



PET/MR provides morphologic and metabolic parameters to identify high-risk rectal cancer patients that need more detailed stagingin a whole-body setting. The volumetric-based PET parameters (MTV and/or TLG) might be included in clinical reports instead ofSUVmax.

ParticipantsAnarosa N. Costa, MD, Sao Paulo, Brazil (Presenter) Nothing to DisclosePriscilla R. Dreyer, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJose F. Marin, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCinthia D. Ortega, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGiovanni G. Cerri, MD,PhD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

to analyze the relation of mucinous component of rectal adenocarcinoma to FDG avidity and diffusion restriction.


ninety-four patients with biopsy proven rectal adenocarcinoma were submitted to whole-body PET/MR for primary staging. Onereader analyzed the MR (blinded to PET findings) regarding locoregional staging, including the presence of mucinous component.Another reader (also blinded to the high-resolution part of pelvic MR) analyzed the PET/MR according to T, N and M staging. Twodifferent readers drawn in consensus the volume of interest (VOI) larger than 0.5 cmCubed of the mucinous component (MC) andthe non-mucinous component (NMC) of the primary tumor on high-resolution T2w sequence and propagated the VOI to PET, DWIand ADC. SUVmax, SUVmean, TLG, MTV, ADCmax, ADCmean and ADCmin values were recorded and compared using Mann-Whitneytest.

RESULTS

Seventeen patients (18.1%) presented MC on MRI. The SUVmax and SUVmean of the NMC were significantly higher than of the MC(16.7 vs. 7.4, p = 0.002 and 13.4 vs. 5.4, p = 0.001). Any of the ADC values was significantly different between MC and NMCgroups.. Among the 17 patients with MC, 16 (94.1%) were at least mrT3b, 15 (88.2%) presented positive extramural vascularinvasion, 14 (82.4%) had involvement of mesorectal fascia, 13 (76.5%) were N-positive and 8 (47.1%) presented distantmetastasis, reinforcing the correlation of MC and tumor risk.

CONCLUSION

The MC identified on PET/MR presents lower glycolytic metabolism than the NMC and is associated with high tumor risk. Tumorcellularity was not different between MC and NMC.


PET/MR enables a conspicuous correlation of mucinous component and tumor metabolism. Imaging readers should be aware of thispattern, recognizing that both primary and metastatic disease of mucinous tumors present low FDG uptake.


ParticipantsThitinan Chulroek, MD, Bangkok, Thailand (Presenter) Nothing to DiscloseHamed Kordbacheh, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseDearada Wangcharoenrung, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseKamonwon Cattapan, MD, Hatyai, Thailand (Abstract Co-Author) Nothing to DisclosePedram Heidari, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseMukesh G. Harisinghani, MD, Boston, MA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the diagnostic performance of qualitative and quantitative FDG PET combined with CT in detection of regional anddistant lymph node metastases in patients with anal cancer


In this IRB approved retrospective study between January 2000 and December 2017, a total of 29 patients (F/M;17/12), mean age59.9±13.7 (15-95) with anal cancer who had staging PET/CT and pathological analysis of suspicious lymph nodes were included forqualitative and quantitative analysis. For qualitative analysis, the positive lymph nodes were defined as uptake close to or higherthan the liver background. For quantitative study, lymph nodes were contoured using the 3D region of interest (ROI) to determine

SSG09-09 68Ga-DOTATATE PET/CT as a Predictive Marker for 177Lu-DOTATATE Therapy

Tuesday, Nov. 27 11:50AM - 12:00PM Room: S505AB

the maximum standard uptake value (SUVmax) and metabolic tumor volume (MTV). Receiver operating characteristic (ROC) curveswere analyzed to extract the optimal cut-off values of SUVmax, lesion to background (L/B) ratio, short axis diameter (SAD) andMTV of the lymph nodes. Histopathologic analysis was the reference standard.

RESULTS

A total of 29 lymph nodes (25 inguinal, 2 external iliac, 1 internal iliac and 1 paraaortic nodes) in 29 patients on PET/CT wereincluded for analysis. For qualitative visual analysis, PET/CT interpreted 27 patients as positive for the presence of nodalmetastases with sensitivity, specificity and accuracy of 1, 0.25 and 0.79. The optimal cut-off values of SUVmax and L/B ratio were2.51 and 1.13 with sensitivity and specificity of 0.91, 0.75 and area under the ROC curve (AUC) of 0.863 (95%CI:0.685,0.962) forSUVmax and 0.887 (95%CI:0.714,0.974) for L/B ratio. Using a best discriminative cut-off value 1.5 cm for SAD and 3.53 cm3 forMTV, the sensitivity and specificity were 0.81 and 1 with AUC of 0.952 (95%CI:0.803,0.997) for SAD and 0.935 (95%CI:0.777,0.993) for MTV. If we used the optimal cut-off value of either SUVmax or SAD, the sensitivity will increase to 0.95(95%CI:0.76,0.99) with a specificity of 0.75 (95%CI:0.35,0.97).

CONCLUSION

Quantitative analysis of lymph node metastases in patients with anal cancer obtained with SUVmax, L/B ratio, nodal size and MTVare more specific and accurate than those performed with qualitative analysis but slightly less sensitive.


The presence of nodal metastasis is crucial for treatment planning. Inguinal nodes with gross nodal involvement will receive abooster dose. Distant metastasis will change the chemotherapy regimen.

ParticipantsHong Song, MD, Sunnyvale, CA (Presenter) Nothing to DisclosePamela Kunz, MD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseThomas Yohannan, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseErik S. Mittra, MD, PhD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric CompanyGuido A. Davidzon, MD, Stanford, CA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Lutetium-177 (177Lu)-DOTATATE has shown significant promise in patients with somatostatin receptor positive neuroendocrinetumors (NETs). Here we report our preliminary experience with the use of 68Ga-DOTATATE PET/CT as a predictive marker foroutcomes with 177Lu-DOTATATE therapy.


Fifteen patients with progressive metastatic NETs were enrolled at our institution on the Expanded Access Program for 177Lu-DOTATATE. Pre-treatment 68Ga-DOTATATE PET/CT scans were performed in all patients. A retrospective analysis of these PET/CTscans was performed for tumor burden. Indices measured include maximal SUV uptake (SUVmax) and total SUV uptake (SUVtotal),which is the sum of 68Ga-DOTATATE uptake by all the lesions in one patient's PET scan. Tumor burden is the sum of all tumorvolumes. These measurements were correlated with progression-free survival.

RESULTS

Of the 15 enrolled patients, six completed all four 177Lu-DOTATATE infusions, and 9 patients completed 3 or fewer infusions.Median progression-free survival is 80% at 12 months from the date of enrollment. The pre-treatment 68Ga-DOTATATE PET/CTdemonstrated well-differentiated NETs and wide-spread metastases in all 15 patients, with liver, lymph nodes and bone being themost common sites. Our data showed a larger tumor burden (1824 ml vs. 748 ml) and higher SUVtotal (636 thousand vs. 186thousand) in deceased patients vs. patients with stable disease, while SUVmax had the opposite trend (36.5 vs. 51.0).

CONCLUSION

It may be possible to metrics from the pre-therapy 68Ga-DOTATATE PET/CT to predict outcomes of 177Lu-DOTATATE in patientswith metastatic NET. While the results are encouraging, larger cohorts are needed to establish 68Ga-DOTATATE PET/CT basedpredictive factors for progression-free survival.


68Ga-DOTATATE PET/CT could serve as a clinical predictor for response to 177Lu-DOTATATE.

NM217-SD-TUA1

Radiation Dose Reduction of PET-CT Scans by Adjusting F18-FDG Dose Based On Patient BMI andModifying Low-Dose CT Protocol

Station #1

NM218-SD-TUA2

Controlled Technique of F18 FDG Dose Reduction for Digital Photon Counting PET/CT Using ListMode Data Truncation

Station #2

NMS-TUA

Nuclear Medicine Tuesday Poster Discussions

Tuesday, Nov. 27 12:15PM - 12:45PM Room: NM Community, Learning Center

NM


ParticipantsAmir H. Khandani, MD, Chapel Hill, NC (Moderator) Consultant, Progenics Pharmaceuticals, Inc; Consultant, F. Hoffmann-La RocheLtd; Ephraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Moderator) Research support, Blue Earth Diagnostics Ltd; Research support,Advanced Accelerator Applications SA

Sub-Events

ParticipantsHasnain Hasham, MD, Kansas City, KS (Presenter) Nothing to DiscloseDaniel L. Kirkpatrick, MD, Kansas City, KS (Abstract Co-Author) Nothing to DiscloseWendell Y. Yap, MD, Shawnee, KS (Abstract Co-Author) Nothing to DiscloseCarissa Walter, Kansas City, KS (Abstract Co-Author) Nothing to DiscloseShelby James Cullom, PhD, Kansas City, KS (Abstract Co-Author) Nothing to DiscloseYu Wang, Kansas City, KS (Abstract Co-Author) Nothing to DiscloseSuzanne L. Hunt, Kansas City, KS (Abstract Co-Author) Nothing to DiscloseMark A. Perry, MD, Spring Hill, KS (Abstract Co-Author) Nothing to Disclose

PURPOSE

To evaluate a BMI-based radiation dose reduction technique for patients undergoing F18-FDG PET-CT scans combining TOF andpenalized likelihood reconstruction for noise optimization.


This retrospective study included 36 patients who underwent PET-CT scans on a GE Discovery MI DR from 1/1/2017-5/1/2017.Quantitative measurements of the mean and maximum SUVs for hepatic, blood pool, and largest lesions were obtained from originaland reformatted list-mode data to simulate the administration of different FDG doses, 0.15 mCi and 0.20 mCi per BMI. Tworadiologists qualitatively assessed each study independently for diagnostic quality.

RESULTS

There was a statistically significant decrease in median radiation dose for both 0.15 mCi (-3.26 mSv, p<0.01) and 0.20 mCi (-1.79mSv, p<0.01) images. Radiologist agreement of baseline and reconstructed exams ranged from 67-81%, suggesting thereconstructed images were of equivalent quality. Metabolic tumor volume was not significantly different between the baseline anddose reduction exams. However, there were varied results when comparing maximum SUV, mean SUV, and total lesion glycolysis(TLG) between studies.

CONCLUSION

This study demonstrates a decrease in patient radiation dose using a dose reduction model, while maintaining diagnostic imagequality. There were statistically significant differences in hepatic and blood pool SUV values when comparing the baseline SUV's tothe reconstructed exams. However, the SUV values of the lesions did not significantly change. This should be taken intoconsideration when using lesion to liver and/or blood pool comparisons, such as Deauville and Lugano scores in lymphoma.


Evaluation of new imaging technology is important to decrease patient radiation dose while maintaining diagnostic quality andquantitative accuracy in PET/CT.


ParticipantsPatrick Silveira, MS, Burlington, VT (Presenter) Nothing to DiscloseJanusz K. Kikut, MD, Burlington, VT (Abstract Co-Author) Nothing to DiscloseNorman V. Sturtevant, MD, Burlington, VT (Abstract Co-Author) Nothing to Disclose


[email protected]

NM219-SD-TUA3

Correlation of Immunohistochemical Parameters, Apparent Diffusion Coefficient (ADC) andStandardized Uptake Value (SUV) in Lung Cancer Using Hybrid PET/MRI

Station #3

[email protected]

PURPOSE

Digital Photon Counting (DPC) PET/CT system with silicon photomultiplier detectors is estimated to have double the clinicalsensitivity, opening opportunity to lower injected activity. The process of injection-dose selection and time per table positionoptimization has traditionally been an empiric one. Software solutions on DPC PET platform allow simulating scans with reducedimaging time by truncation of the list mode data. The aim of this study is to evaluate if this solution can be used as a guide in theprocess of injected activity reduction.


48 consecutive F18FDG DPC PET/CT scans were evaluated. All patients were administered 18F-FDG 0.165 mCi/kg per lean bodyweight. In addition to routine clinical image reconstruction a second reconstruction was performed with simulated decreased timeas follows (BMI<20: 27% from 90s to 75s; BMI 20-40: 25% from 120s to 90s; BMI>40: 20% from 150s to 120s). Additionalreconstruction and image processing added less than 10 minutes of technologist time. Nuclear Radiology reader with 12 years ofexperience in PET/CT compared the image quality of the simulated reduced time vs the standard time on a dedicated NM consoleallowing simultaneous display of both. If the reader determined a difference between the images, the difference was scored 1 to 3(1- negligible, 2-degraded but acceptable for interpretation, 3- non interpretable).

RESULTS

The reader identified 5 out of the 48 cases as showing a difference in quality between the reduced time vs standard time. All 5cases were rated 1 (negligible) and both images would render the same clinical interpretation. All 5 images images that were notedto be better qualitatively were the standard scan time. The results allowed reduction of the prescribed F18FDG activity by 20%without adjustment to acquisition time per table position.

CONCLUSION

Qualitative analysis of simulated shorter acquisition scans allows an expedited, informed radiopharmaceutical dose reduction in DPCPET and can be used without an impact on daily workflow. The procedure allows a controlled optimization of radiopharmaceuticaldose and acquisition time minimizing the risk of non-diagnostics exams.


Analysis of simulated shorter acquisition scans allows for an informed radiopharmaceutical dose reduction in DPC PET. The processenables one to optimize dose and acquisition time, thus minimizing the risk of non-diagnostics exams.

ParticipantsOle Martin, Duesseldorf, Germany (Presenter) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SALale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AGGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseLino Sawicki, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To correlate various prognostically relevant immunohistochemical parameters of primary lung cancer with simultaneously acquiredstandardized uptake values (SUV) and apparent diffusion coefficient (ADC) derived from hybrid PET/MRI.


55 consecutive patients with histologically proven lung cancer (mean age 62.5±9.1y) underwent PET/MRI. Diffusion-weightedimaging (DWI, b values: 0,500,1000s/mm²) was performed simultaneously with PET acquisition. A region of interest (ROI)encompassing the entire primary tumor was drawn into each patient's PET/MRI images to determine the glucose metabolismrepresented by maximum and mean SUV and also into ADC maps to assess tumor cellularity represented by mean and minimum ADCvalues. Histopathological tumor grading was available in 43/55 patients. In 15/55 patients, additional prognostically relevantimmunohistochemical markers, i.e. phospho-AKT Ser473 (pAKTS473), extracellular signal-regulated kinase (pERK), phosphatase andtensin homolog (PTEN), and human epidermal growth factor receptor 2 (erbB2) were determined. Pearson's correlation coefficientswere calculated to compare SUV and ADC values, while Spearman's were used for the immunohistochemical markers.

RESULTS

The average SUVmax, SUVmean, ADCmin, and ADCmean in lung cancer primaries was 12.6±5.9, 7.7±4.6, 569.9±96.1 s/mm², and825.8±93.2 s/mm², respectively. We found a significant inverse correlation between the ADCmin and SUVmax (r=-0.58, p<0.001) aswell as between the ADCmin and SUVmean (r=-0.44, p<0.001). Tumor grading showed a significant positive correlation withSUVmax and SUVmean (r=0.34 and r=0.31, both p<0.05) and a significant inverse correlation with ADCmin and ADCmean (r=-0.30and r=-0.40, both p<0.05). In addition, erbB2 showed a significant inverse correlation with SUVmax and SUVmean (r=-0.50 and r=-0.49, both p<0.05). There was no significant correlation with other immunohistochemical markers.

CONCLUSION

The present data show a correlation between increased glucose-metabolism, cellularity, degree of differentiation as well as erbB2expression of lung cancer primaries. 18F-FDG-PET and DWI from hybrid PET/MRI may offer complementary information for evaluationof lung cancer aggressiveness in initial staging and treatment response.

NM220-SD-TUA4

PET/3-T MRI in Locally Advanced Rectal Cancer: Role of SUV and ADC Volumetric Histograms-BasedAnalyses in Predicting the Tumor Regression Grade After Preoperative Chemoradiotherapy

Station #4

NM221-SD-TUA5

The Role of 18F-FDG SPECT/CT in Predicting EGFR Gene Mutation Status in Lung Adenocarcinoma

Station #5


SUV in PET/MRI, ADC and tumor grading show a correlation in lung cancer patients leading to complementary information for stagingworkup

ParticipantsGiovanna Orsatti, MD, Padova, Italy (Presenter) Nothing to DiscloseFilippo Crimi, MD, Padova, Italy (Abstract Co-Author) Nothing to DiscloseAlessia Varotto, MD, Padova, Italy (Abstract Co-Author) Nothing to DisclosePietro Zucchetta, Padova, Italy (Abstract Co-Author) Nothing to DiscloseChiara Giraudo, MD,PhD, Padova, Italy (Abstract Co-Author) Nothing to DiscloseMichael Weber, Vienna, Austria (Abstract Co-Author) Nothing to DiscloseFabio Pomerri, MD, Padua, Italy (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the relationship among simultaneously acquired SUVs, ADCs, and histopathologic tumor regression grade (TRG) inpatients with locally advanced rectal cancer (LARC) by using volumetric histogram analyses.


Patients with LARC referring to our local tertiary care center were enrolled in this prospective study and underwent a whole body[18F]-FDG-PET/MR scan (integrated hybrid PET/MR), including DWI, at staging and after preoperative chemoradiotherapy (pCRT).PET images were resliced and resampled according to the ADC maps (i.e.,Osirix Software). For each LARC, a region of interest (ROI)was manually drawn on the T2weighted images along the boundaries of the lesion in each slice containing visible mass; each ROIwas then copied on the corresponding PET and ADC datasets. Pixel-based SUVs and ADCs were collected from the entire volume ofthe lesions. Mean, median, skewness, kurtosis of SUVs and ADCs values as well as tumor volume ratio (pre-pCRT volume/post-pCRTvolume) were computed. ADCmean and ADCmin as well as SUVmean and SUVmax were also collected from a ROI manually drawn atthe level of maximum diameter of each lesion. Pearson correlation coefficient was applied to evaluate the correlation among all theinvestigated variables and TRG.

RESULTS

Eleven patients met the inclusion criteria (7M; mean age 64.3±5.1 years). Five LARC showed a complete regression (TRG 4). Asignificant negative correlation emerged between TRG and post-pCRT SUVs' kurtosis (r=-0.644; p=0.033), skewness (r=-0.634;p=0.036), mean (r=-0.631; p=0.037), median (r=-0.624; p=0.040) and SD (r=-0.629; p=0.019). The single-ROI-based SUVmaxnegatively correlated with the TRG (r=-0.637, p=0.035). A positive correlation emerged between volume ratio and TRG (r=0.688,p=0.019). No significant correlation was found between TRG and ADCs, neither pre nor post-pCRT, or between TRG and pre-pCRTSUVs.

CONCLUSION

Our preliminary results demonstrated that post-pCRT volumetric histogram analyses of SUV values are predictor of response topCRT. Additional studies on a larger population are necessary to further assess this evidence and the predictive role of volume-based ADCs' in LARC.


SUVs' volumetric histograms-based analyses seem to be a useful tool to predict the response to pCRT, allowing the selection ofpatients candidate to a wait-and-see approach instead of surgery.

ParticipantsZhao Long, MD, Shanghai, China (Presenter) Nothing to DiscloseWang Huoqiang, MD, Shanghai, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

In most 18F-fluorodoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) studies, tumor FDG uptakeis generally reported to be decreased in lung cancers with EGFR mutation. In the literatures, 18F-FDG single photon emissioncomputed tomography/computed tomography (18F-FDG SPECT/CT) scan was a powerful and reliable tool for evaluating patientswith lung neoplasms, and with which results obtained are in most cases concordant with those obtained with 18F-FDG PET/CT. Thepurpose of this study is to investigate the predictive value of 18F-FDG SPECT /CT in evaluating EGFR mutation status in primarypulmonary adenocarcinoma (ADA).


ADAs were retrospectively identified in 276 consecutive patients who underwent 18F-FDG SPECT/CT scan from July 2014 toDecember 2015. The histopathological results were confirmed by resected, aspirated or biopsied samples. Tumor-to-normal tissue(T/NT) uptake ratios of 18F-FDG were calculated for the primary lesion. We analyzed the association between T/NT value and EGFRmutation status in ADA. Associations between quantitative continuous variables and EGFR mutation status were investigated byusing the Mann-Whitney U test.

NM144-ED-TUA6

The Utility of PET/MR in Head and Neck Cancer

Station #6

NM145-ED-TUA7

Lymphoscintigraphy in Cutaneous Melanoma: Usual and Unexpected Pathways of Lymphatic drainage

Station #7

RESULTS

EGFR mutations were identified in 152 patients (55.3%). EGFR mutations occurred more frequently in females (p < 0.001), in non-smokers (p < 0.001), in those with smaller lesions (p < 0.001), and in those with lower T/NT value (p < 0.001). In multivariateanalysis, sex, T/NT, and tumor size were significantly associated with EGFR mutation. The receiver operating characteristic (ROC)curve yielded area under the curve (AUC) values of 0.636 (95%CI, 0.568-0.704, p < 0.001) and 0.719 (95%CI, 0.657-0.782, p <0.001) for low T/NT alone and the combination of the three factors, respectively.

CONCLUSION

We demonstrated that T/NT value of FDG uptake may be helpful in predicting the EGFR mutation status, which is consistent withresults of 18F-FDG PET/CT. Especially in China, 18F-FDG SPECT/CT scan is included in the Medicare program, while 18F-FDGPET/CT is not, which enhances the clinical value of 18F-FDG SPECT/CT scan for cost reasons.


For the first time, we demonstrated that EGFR mutations were more frequent in lung adenocarcinomas with lower T/NT value, whichis a semi quantitative parameter of FDG uptake in 18F-FDG SPECT/CT scan.

ParticipantsTetsuro Sekine, MD, PhD, Tokyo, Japan (Presenter) Nothing to DiscloseFelipe D. Barbosa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseGaspar Delso, PhD, Cambridge, United Kingdom (Abstract Co-Author) Employee, General Electric Company; Edwin E. ter Voert, PhD, Zurich, Switzerland (Abstract Co-Author) Nothing to DisclosePatrick Veit-Haibach, MD, Zurich, Switzerland (Abstract Co-Author) Research Grant, Bayer AG Resaarch Grant, F. Hoffmann-LaRoche Ltd Research Grant, General Electric CompanyMartin W. Huellner, MD, Zurich, Switzerland (Abstract Co-Author) Institutional Grant, General Electric Company


[email protected]

TEACHING POINTS

The combined MRI components on PET/MR system serve detailed anatomy which could impact on the diagnosis of head and neckcancer. The aim of this presentation is to understand the role of PET/MR in the assessment of head and neck cancer by comparingit with PET/CT.


All presentations will be performed as a manner of the comparison of PET/MR with PET/CT. The contents of the presentation are asbelow. 1. T staging 2. N staging 3. Follow up study 4. Resectability defining findings 5. Perineural spreading 6. Artifact 7. Unknownprimary tumor

ParticipantsAmit Gupta, MD, Cleveland, OH (Presenter) Nothing to DiscloseJoseph M. Curcio, DO, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseElias Kikano, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseMaharshi A. Rajdev, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseNils Grosse Hokamp, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseKai Roman Laukamp, Cleveland, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

To Understand the basis behind the procedure, the actual technique and different radiotracers used for this study. To know theusual pathways of lymphatic drainage in different regions of the body. To have understanding of unexpected pathways of lymphaticdraingae and additional nodal basins which may be important in treatment planning. To emphasize the complementary role of SPECTimaging and SPECT/CT fusion


- History of Lymphoscintigraphy - Various radionuclides that are commonly used - Technique - Different cutaneous malignanciesthat are currently being evaluated by lymphoscintigraphy - Examples of usual drainage pathways: Head and Neck Anterior trunkBack Upper extremities Lower extremities - Examples of unusual pathways of lymphatic drainage: Epitrochlear lymph node in upperextremity lesoin Popliteal lymph node drainage in lower extremity lesion Retroperitoneal drainage from a back melanoma

NM222-SD-TUB1

To Evaluate the Unknown Primary Site of Tumour with 68Ga-DOTATOC PET/CT in Patients withKnown Metastatic Neuroendocrine Tumor

Station #1

NM223-SD-TUB2

Effects of New Block Sequential Regularized Expectation Maximization (BSREM) ReconstructionAlgorithm on SUVmax, MTV, and Changes of SUVmax on FDG PET-CT Before and After NeoadjuvantChemotherapy in Patients with Esophageal Cancer

Station #2

NMS-TUB

Nuclear Medicine Tuesday Poster Discussions

Tuesday, Nov. 27 12:45PM - 1:15PM Room: NM Community, Learning Center

NM


ParticipantsEphraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Moderator) Research support, Blue Earth Diagnostics Ltd; Research support,Advanced Accelerator Applications SAAmir H. Khandani, MD, Chapel Hill, NC (Moderator) Consultant, Progenics Pharmaceuticals, Inc; Consultant, F. Hoffmann-La RocheLtd;

Sub-Events

ParticipantsSikandar M. Shaikh, DMRD, Hyderabad, India (Presenter) Nothing to Disclose

PURPOSE

It is very difficult to evaluate the Localization of the primary site of the unknown primary tumor by various modalities .This is criticalfor management of patients presenting with neuroendocrine tumor (NET) with metastases.


This was the retrospective study which was performed to evaluate the efficacy of 68Ga-DOTATOC PET/CT in patients withdiagnosis of or suspected NET with metastases. A subgroup of patients with metastases and unknown primary after initial work-upwas analyzed. The 68GaGa-DOTATOC Whole body PET/CT was done .This was evaluated by the SUV max uptake values.The studywas considered true positive if the positive primary site has significant uptake and which was confirmed by histology or follow-upimaging. The scan was considered false positive if no primary lesion was found corresponding to 68aGa-DOTATOC positive site. Allnegative scans for primary tumor were considered false negative. A scan was classified unconfirmed if 68Ga-DOTATOC PET/CTsuggested a primary, however, no histology was obtained and imaging follow-up was not confirmatory.

RESULTS

Forty patients with known metastatic NET and unknown primary underwent 68Ga-DOTATOC PET/CT. After evaluation andquantification the study was true positive, false positive, false negative and unconfirmed rates for unknown primary tumor were38%, 7%, 50% and 5% respectively.

CONCLUSION

The efficacy of 68Ga-DOTATOC PET/CT is an effective modality in localization of unknown primary in patients with metastatic NETas compared to other modalities including F FDG PET-CT.


Thus 68Ga-DOTATOC PET/CT is an effective modality for evaluation of Neuroendocrine tumours.

ParticipantsMitsuaki Tatsumi, MD, PhD, Suita, Japan (Presenter) Nothing to DiscloseTakashi Kamiya, Suita, Japan (Abstract Co-Author) Nothing to DiscloseKayako Isohashi, Suita, Japan (Abstract Co-Author) Nothing to DiscloseHiroki Kato, Suita Osaka, Japan (Abstract Co-Author) Nothing to DiscloseNoriyuki Tomiyama, MD,PhD, Suita, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationJun Hatazawa, MD, PhD, Osaka, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

BSREM reconstruction algorithm, or so called "Q. Clear", was recently introduced to improve image quality and quantification in PETexaminations. The purpose of this study was to evaluate the effects of this new algorithm on SUVmax, metabolic tumor volume(MTV), and changes of SUVmax on FDG PET-CT before and after neoadjuvant chemotherapy (NAC) in patients (pts) withesophageal cancer (EC), comparing the results to those by an ordered subset expectation maximization (OSEM) reconstructionalgorithm.


FDG PET-CT examinations were performed before and after 2 cycles of NAC in 39 EC pts before surgery. PET images acquired with a

NM224-SD-TUB3

Time Course Change and Reproducibility of 18F-FAZA in the Patient with Head and Neck SquamousCell Carcinoma

Station #3

NM225-SD-TUB4

Are Staging F-18-FDG PET/MRI Radiomic Features Associated with Metastases in Cancer of theGastro-Esophageal Junction?

Station #4

FDG PET-CT examinations were performed before and after 2 cycles of NAC in 39 EC pts before surgery. PET images acquired with aGE Discovery 710 scanner were reconstructed using BSREM beta 700 and OSEM (subset 8, iteration 3, and Gaussian filter 4mm;regular setting in our hospital) algorithms. SUVmax and MTV were compared between BSREM and OSEM images in all 39 primary EClesions before and after NAC. Changes of SUVmax after NAC were also compared. Statistical analysis was performed with a pairedt-test and a Spearman's correlation method.

RESULTS

SUVmax ranged from 3.2 to 40.9 (mean+/-SD: 16.0+/-8.8) and MTV from 1.1 to 149.5 ml (24.8+/-30.0 ml) in OSEM before NAC.MTV in BSREM was significantly lower than that in OSEM (7.1% decrease), while SUVmax was almost the same. A moderatenegative correlation was observed between [%increase of SUVmax or MTV with BSREM] and [SUVmax in OSEM] (|Rho|=0.45-0.52)as well as [%increase of SUVmax with BSREM] and [MTV in OSEM] (|Rho|=0.58). After NAC, SUVmax ranged from 2.3 to 22.6(6.5+/-4.5) and MTV from 0 to 70.2 ml (7.0+/-11.7 ml) in OSEM. SUVmax in BSREM was significantly higher than that in OSEM(3.7% increase), while MTV did not differ. A moderate negative correlation was observed between [%increase of MTV with BSREM]and [SUVmax or MTV in OSEM] (|Rho|=0.35-0.44). Changes of SUVmax after NAC did not differ between BSREM and OSEM (52.9%vs. 53.2%).

CONCLUSION

This study demonstrated that BSREM decreased MTV and increased SUVmax before and after NAC, respectively, in EC patients.However, changes of SUVmax after NAC did not differ between BSREM and OSEM. Further studies are required to confirm theseresults in a larger population.


The effect of BSREM on PET was different before and after NAC in EC patients.

ParticipantsShingo Baba, Fukuoka, Japan (Presenter) Nothing to DiscloseTakuro Isoda, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseYoshiyuki Kitamura, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseRyo Somehara, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseKeiichiro Tahara, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseMasayuki Sasaki, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseHiroshi Honda, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

18F- fluoroazomycin arabinoside (18F-FAZA) PET has been recently introduced for noninvasive evaluation of hypoxia. Since itsclearance is faster than the conventional hypoxia tracer, the possibility of early phase imaging is expected. However, the timecourse change and the reproducibility of 18F-FAZA uptake has remained unclarified. We therefore investigated the change of serialtwo 18F-FAZA PET scans by using quantitative analysis of uptake.


Nine patients with untreated head and neck cancer underwent serial 18F-FAZA PET/CT scans (18F-FAZA(1) and 18F-FAZA(2)). Allimages were acquired at 2h and 4 h after intravenous 18F-FAZA injection. The maximum standardized uptake (SUVmax), andtumor-to-muscle ratio (TMR) of 18F-FAZA uptake were statistically compared between the two 18F-FAZA scans by use ofintraclass correlation coefficients (ICCs). The hypoxic volume was calculated as the area with a TMR of greater than or equal to1.25 to assess differences in hypoxic volume between the two 18F-FAZA scans.

RESULTS

The SUVmax (mean ± SD) for 18F-FAZA(1) and 18F-FAZA(2) was 2.14 ± 0.65 and 2.11 ± 0.84, respectively, with the differencebetween the 2 scans being -2.2% ± 7.8%. Uptake of the muscle significantly decreased with time (p<0.05). TMRs for 18F-FAZA(1)and 18F-FAZAZ(2) were 1.35 ± 0.37 and 1.55 ± 0.48, respectively, with a difference of 13.4% ± 7.4%. The ICCs for SUVmax, andTMR were 0.971, and 0.923, respectively. The hypoxic volume based on TMR was increased with time (0.9 ± 1.3 mL and 2.19 ±0.67), with ICCs of 0.910.

CONCLUSION

The values for 18F-FAZA PET uptake in head and neck tumors between the two 18F-FAZA scans were highly reproducible.However, TMRs increase with time because of the washout from the muscle. As the result, defined hypoxic area was increased.Therefore, scan after 4 h cannot be omitted or replaced by 2h scan.


The values for 18F-FAZA PET uptake in head and neck tumors between the two scans were highly reproducible. But to impriove theTMR, relative long tilme uptake time (4 hrs) after tracer injection is still needed in this tracer.

ParticipantsSerena Baiocco, Bologna, Italy (Abstract Co-Author) Nothing to DiscloseBert-Ram Sah, MD, London, United Kingdom (Presenter) Nothing to DiscloseAndrew Mallia, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseJames Stirling, Middlesex, United Kingdom (Abstract Co-Author) Nothing to DiscloseSami Jeljeli, BSc, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseAlessandro Bevilacqua, PhD,MSc, Bologna, Italy (Abstract Co-Author) Nothing to DiscloseGary Cook, MD, FRCR, London, United Kingdom (Abstract Co-Author) Research support, General Electric Company Research support,Alliance Medical Limited Research support, Siemens AG Research Consultant, Blue Earth Diagnostics Ltd Speakers Bureau, Bayer AG

NM146-ED-TUB5

Our First Year with Gallium-68 DOTATATE PET/CT: Lessons Learned Through RAD-PATH Correlationand Potential Pitfalls

Station #5

Vicky J. Goh, MBBCh, London, United Kingdom (Abstract Co-Author) Research Grant, Siemens AG Speaker, Siemens AG

PURPOSE

To identify quantitative imaging biomarkers at staging F-18-Fluorodeoxyglucose (FDG)-positron-emission-tomography/magnetic-resonance-imaging (PET/MRI) for predicting distant metastases in patients with gastro-esophageal junction (GEJ) cancer.


Following IRB approval and informed consent, 24 patients with histologically proven GEJ cancer were prospectively recruited; 4patients were excluded for technical reasons. Finally, 19 male and 1 female (68.3±9.1 years) were considered. Patients wereinjected with 326±28 MBq FDG intravenously. Uptake time was 90 minutes. Two experienced radiologists and nuclear physiciansreviewed the images in consensus. Maximum standardized uptake value (SUVmax) and tumor size were analyzed. First-order andsecond-order statistical texture features were computed on SUV values of the whole tumor volume. k-means clustering algorithmwas used to assess the correlation of feature-pairs with the presence of distant metastases. Sensitivity (SE), specificity (SP),positive predictive value (PPV), negative predictive value (NPV) and accuracy (ACC) were calculated to quantify the discriminationability of features.

RESULTS

Second-order entropy and maximum probability, linked to texture irregularity and homogeneity respectively, were the best feature-pair in discriminating patients with and without metastatic disease (SE=80%, SP=70%, PPV=73%, NPV=78%, ACC=75%). SUVmax(SE=80%, SP=30%, PPV=53%, NPV=60%, ACC=55%) and tumor size (SE=90%, SP=10%, PPV=50%, NPV=50%, ACC=50%)performed worse, particularly for specificity.

CONCLUSION

These results confirm the common expectation that greater intra-tumor heterogeneity correlates with metastatic potential. Theextraction of advanced quantitative PET imaging features from the primary lesion may help prognostication.


Radiomics may help in improving prognostication at staging.

AwardsMagna Cum Laude

ParticipantsBrian L. Bones, MD, Winston-Salem, NC (Presenter) Nothing to DiscloseJennifer A. Schroeder, MD, Winston-Salem, NC (Abstract Co-Author) Nothing to DiscloseAngela Niehaus, MD, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseWilliam J. Beuerlein, MD, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseWencheng Li, MD, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseAlexei V. Mikhailov, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseKimberly Stogner-Underwood, MD, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseAnita J. Thomas, MD, Winston-Salem, NC (Abstract Co-Author) Nothing to DiscloseShane C. Masters, MD,PhD, Winston Salem, NC (Abstract Co-Author) Nothing to DiscloseJean-Luc C. Urbain, MD,PhD, Lebanon, PA (Abstract Co-Author) Software support, General Electric CompanyPaige Bennett, MD, Winston-Salem, NC (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

This exhibit provides a clinical overview of Gallium-68 DOTATATE PET/CT. The learner should be able to describe theradiopharmacology of Ga-68 DOTATATE, imaging characteristics for different pathologies and their radiologic-pathologic correlation,and potential pitfalls when interpreting and communicating Ga-68 DOTATATE findings.


1. Radiopharmacy of Ga-68 DOTATATE.2. Normal Biodistribution.3. Radiological-pathological correlation of Ga-68 DOTATATE PET/CTincluding: Well-differentiated Neuroendocrine (NET), Poorly-differentiated NET of the rectum with comparison of Ga-68 DOTATATEand F-18 FDG PET/CT, Pheochromocytoma in the setting of Multiple endocrine neoplasia type 2, Paraganglioma, SynchronousPancreatic NET and Renal Cell Carcinoma.4. Metastatic disease workups without pathology including: Small bowel NET with livermetastasis, metastatic medullary thyroid, metastatic VIPoma, NET of unknown primary with osseous metastatic disease.5. Pitfallsand Incidentals including: Metastatic NET to the liver with incidental meningioma, Breast cancer uptake from somatostatinexpression, uptake with osteoclastic activity and osteoarthritis, importance of communicating with referrers the differencesbetween Ga-68 DOTATATE PET/CT and F-18 FDG PET/CT.

MSCC33A Abdomen/Pelvis I

MSCC33B Abdomen/Pelvis II

MSCC33C Pediatrics

MSCC33

Case-based Review of Nuclear Medicine: PET/CT Workshop-Abdomen/Pelvis & Pediatrics PET/CT (InConjunction with SNMMI) (Interactive Session)

Tuesday, Nov. 27 1:30PM - 3:00PM Room: E450B

CT GI GU NM PD



ParticipantsSamuel E. Almodovar-Reteguis, MD, Orlando, FL (Director) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Director) Nothing to DiscloseTerence Z. Wong, MD, PhD, Chapel Hill, NC (Moderator) Consultant, Lucerno Dynamics, LLC;

Sub-Events



[email protected]

LEARNING OBJECTIVES

1) Review Challenging and Instructive cases which will help with interpretation of PET/CT scans in the abdomen and pelvis.

ABSTRACT

For oncologic studies, F18-FDG is an outstanding tracer with wide applications. However, there are many normal variants andpitfalls which can make interpretation challenging. Appropriate clinical history can help avoid mistakes. Having patients fill out aquestionnaire form can provide helpful information and if possible directly interviewing the patients for relevant pertinent clinicalinformation can also be valuable. Discussing a complicated PET/CT scan with the referring physician can also yield valuableinformation and improve interpretation. This review will discuss challenging cases, physiologic variants and pitfalls seen in theabdomen and pelvis when interpreting PET/CT scans concentrating on genitourinary and gastrointestinal cases.



[email protected]

LEARNING OBJECTIVES

1) Review the current applications of PET-CT in gynecologic disease and bladder cancer.

ABSTRACT

FDG PET-CT is an effective tool in initial assesment andtreatment follow up of various genitourinary and gynecological tumors. Thissession aims to highlight the utility of PET-CT in gynecological disroders as well as tumors of the bladder. An updated review of thecurrent applications of PET-CT in gynecologic disease as well as bladder cancer will be presented.Pitfalls and variants that maychallenge image interpretation will also be highlighted.

ParticipantsHelen R. Nadel, MD, Palo Alto, CA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) To be able to identify optimal PET/CT imaging parameters for children. 2) To be able to describe PET/CT patterns of disease inpediatrics. 3) To be able to compare PET/CT with PET/MR imaging in children.

MSRO37-01 Dynamic Perfusion Area-Detector CT versus Dynamic Perfusion MR Imaging versus FDG-PET/CT:Capability for Therapeutic Outcome Prediction in Small Cell Lung Cancer Patients with Limited Disease

Tuesday, Nov. 27 1:30PM - 1:40PM Room: S103CD

MSRO37

BOOST: Lung, Mediastinum and Pleura


BQ CH CT NM OI RO



ParticipantsMeng X. Welliver, MD, Columbus, OH (Moderator) Nothing to DiscloseTracy M. Sherertz, MD, San Francisco, CA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsYoshiharu Ohno, MD, PhD, Kobe, Japan (Presenter) Research Grant, Canon Medical Systems Corporation; Research Grant,Koninklijke Philips NV; Research Grant, Bayer AG; Research Grant, DAIICHI SANKYO Group; Research Grant, Fuji Pharma Co, Ltd;Research Grant, Guerbet SA; Yasuko Fujisawa, MS, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationMasao Yui, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationYuji Kishida, MD,PhD, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseShinichiro Seki, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationTakeshi Yoshikawa, MD, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationNoriyuki Negi, RT, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseKatsusuke Kyotani, RT,MSc, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseTakamichi Murakami, MD, PhD, Osakasayama, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To directly compare the capability for therapeutic outcome prediction among dynamics contrast-enhanced (CE-) perfusion area-detector CT (ADCT) and CE-perfusion MR imaging (MRI) assessed by same mathematical method and FDG-PET/CT in small cell lungcancer (SCLC) patients assessed as limited disease (LD).


Forty-three consecutive pathologically diagnosed SCLC patients assessed as LD (25 male, 18 female; mean age 67 year old)underwent FDG-PET/CT, dynamic CE-perfusion ADCT and MRI, chemoradiotherapy, and follow-up examination. In each patient,therapeutic outcomes were assessed as therapeutic effect based on RECIST guideline, disease free interval and overall survival.Then, all patients were divided into two groups as follows: 1) responder (CR+PR: n=33) and 2) non-responder (SD+PD: n=10)groups. In each patient, total perfusion (TP) and tumor perfusions from pulmonary (TPP) and systemic (TPS) circulations calculatedby dual-input maximum slope method from dynamic CE-perfusion ADCT and MRI data and SUVmax on PET/CT were assessed attargeted lesions. Then, final values were determined as average values from all targeted lesion, and compared between two groupsby Student's t-test. To compare the capability for distinguishing two groups, all indexes as having significant difference wereassessed by ROC analysis. Finally, disease free and overall survivals between responders and non-responders assessed by eachindex were compared by Kaplan-Meier method followed by log-rank test.

RESULTS

There were significant difference of all indexes except TPP determined by each method (p<0.05). Area under the curves (Azs) ofTPS (ADCT: Az=0.92, MRI: Az=0.92) were significantly larger than that of SUVmax (Az=0.73, p<0.05). Disease free survivals ofresponder were significantly longer than that of non-responder by TP (ADCT: p=0.006, MRI: p=0.02) and TPS (ADCT: p=0.001,MRI: p=0.02). Overall survivals of responder were also significantly longer than that of non-responder by TP (ADCT: p<0.0001, MRI:p=0.0003,) and TPS (ADCT: p=0.001, MRI: p=0.001).

CONCLUSION

Dynamic CE-perfusion ADCT and MRI have better potential for predicting therapeutic outcome than FDG-PET/CT in small cell lungcancer patients with limited disease.


Dynamic CE-perfusion ADCT and MRI have better potential to predict therapeutic outcome than FDG-PET/CT in small cell lungcancer patients with limited disease.

MSRO37-02 Modern Treatment Patterns and Overall Survival of Non-Small Cell Lung Cancer Patients ReceivingPalliative Radiation Therapy for Brain Metastases at Diagnosis


MSRO37-03 Quantification of Radiation Pneumonitis in Lung Cancer Patients Receiving Proton or PhotonRadiotherapy Using FDG-PET/CT



ParticipantsPamela Samson, MD, Saint Louis, MO (Presenter) Nothing to DiscloseBenjamin W. Fischer-Valuck, MD, MS, St. Louis, MO (Abstract Co-Author) Nothing to DisclosePrashant Gabani, MD, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseMichael C. Roach, MD, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseJeffrey D. Bradley, MD, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseCliff G. Robinson, MD, Saint Louis, MO (Abstract Co-Author) Investigator, Varian Medical Systems, Inc; Research funded, VarianMedical Systems, Inc; Speakers Bureau, Varian Medical Systems, Inc; Research funded, Elekta AB; Travel support, Merit MedicalSystems, Inc; Speakers Bureau, ViewRay, Inc; Stockholder, Radialogica, LLCChristopher D. Abraham, MD, Manchester, MO (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Non-small cell lung cancer (NSCLC) is one of the most common malignancies associated with brain metastases (BM) at diagnosis.Recent randomized trials have shown equivalent survival outcomes and improved neurocognitive outcomes with stereotacticradiosurgery (SRS) compared to whole brain radiation therapy (WBRT). We reviewed the NCDB to identify trends of RT for NSCLCpatients with BM.


11,299 NSCLC patients with BM at diagnosis and treated with palliative brain RT between 2010 and 2014 were identified in theNCDB. Patients receiving "stereotactic radiosurgery, NOS," "LINAC radiosurgery," or "gamma knife radiosurgery," or receivedexternal-beam RT with fraction size >=6 Gy were included in the SRS cohort. The WBRT cohort included all patients receiving RT tothe brain in >=5 fractions. Patient characteristics were correlated with treatment received using multivariable logistic regression.Kaplan-Meier was used to compare overall survival (OS) between these two groups and Cox Proportional Hazards modeling (CPHM)to identify variables associated with OS.

RESULTS

9,680 (85.7%) patients were included in the WBRT group and 1,619 (14.3%) patients in the SRS group. Median dose in the WBRTwas 3000 cGy and 2200 cGy in the SRS group. The frequency of SRS increased from 9.9% in 2010 to 19.6% in 2014. On MVA,variables associated with increased likelihood of receiving SRS included: increasing age (OR 1.01, 95% CI 1.01-1.02; P<0.0001),most recent year (2014) of diagnosis (OR 2.14, 1.78-2.56; P<0.0001), treatment at an academic facility (OR 3.21, 2.51-4.10;P<0.0001), private insurance (OR 2.25, 1.62-3.11), income in zip code >$63,000 (OR 1.33, 1.13-1.56; P=0.001), living >20 milesfrom treatment facility (OR 1.19, 1.03-1.37; P=0.016), and receipt of chemotherapy (OR: 2.48, 2.12-2.88; P<0.0001). WBRTpatients had median OS of 4.1 months (95% CI, 4.0-4.3) vs. 8.9 months (8.2-9.7) for SRS patients (P<0.0001). On CPHM, factorsindependently associated with improved OS included receipt of SRS, chemotherapy, treatment at an academic center, and privateinsurance (P<0.02 for all).

CONCLUSION

Our analysis reveals that WBRT remains the most common palliative treatment for BM in NSCLC. SRS use is increasing and hasnearly doubled between 2010 and 2014. In this study, SRS was associated with increased OS although there are biases in theselection of patients who receive SRS.


SRS use for NSCLC patients with BM at the time of diagnosis is increasing, and is independently associated with improved OS.

AwardsTrainee Research Prize - Fellow

ParticipantsPegah Jahangiri, MD, Philadelphia, PA (Presenter) Nothing to DiscloseKamyar Pournazari, MD,MSc, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseCharles B. Simone II, MD, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseDrew A. Torigian, MD, MA, Philadelphia, PA (Abstract Co-Author) Co-founder, Quantitative Radiology Solutions LLCAbass Alavi, MD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

This study assessed the feasibility of FDG-PET/CT to quantify radiation-induced pneumonitis in ipsilateral and contralateral lungs ofpatients with locally advanced non-small cell lung cancer (NSCLC) who received proton, photon, or combined proton-photonradiotherapy (RT).

MSRO37-04 Differentiating EGFR Mutation Status in Non-Small Cell Lung Cancer Using Imaging Features FromPET/CT


MSRO37-05 Stereotactic Body Radiotherapy for Centrally Located Non-Small Cell Lung Cancer: Single CenterExperience



39 consecutive patients (53.8% female, median age 67y) with predominantly stage IIIA (62%) or IIIB (31%) NSCLC underwentFDG-PET/CT before and after proton or photon RT. Regions of interest (ROIs) were drawn manually along the margins of the lungparenchyma on PET/CT images. Lung mean standardized uptake value (SUVmean), global lung glycolysis (GLG), and lung volumewere measured. Partial volume correction (PVC) of PET-based parameters was then performed. To quantify tumor metabolicresponse to RT, metabolically active tumor volume (MTV), tumor SUV, and total lesion glycolysis (TLG) were measured. Total lesionglycolysis was then subtracted from GLG to calculate global lung parenchymal glycolysis (GLPG). Parameters of FDG-PET/CT scansbefore and after RT were compared using two-tailed paired t-tests.

RESULTS

Among the 9 patients who received photon RT, there was a significant increase in PVC-GLPG of ipsilateral (p <0.001) and in GLG ofcontralateral (p =0.036) lungs. Also, in the subset of 9 patients who received combination of proton-photon RT, there was astatistically significant increase in PVC-GLPG in only the ipsilateral lung (p <0.001). In contrast, among the 21 patients treatedexclusively with proton RT, no significant increase in PVC-SUVmean (p=0.114) or in PVC-GLPG (p=0.453) were observed in ipsilaterallungs. Also, there were no significant increase in SUVmean (p=0.841) or in GLG (p=0.241) of contralateral lungs of patients whoreceived exclusively proton RT.

CONCLUSION

We identified significant increases in PVC-SUV and PVC-GLPG in patients who received photon RT (either alone or in combinationwith proton RT) that were not identified in patients who received only proton RT. These observations suggest less induction ofinflammatory response in both ipsilateral and contralateral lungs of patients treated with proton compared to photon or combinedproton-photon RT, suggesting a mechanism by which proton therapy reduces radiation-induced pneumonitis.


Proton RT induces less inflammatory response in both the ipsilateral and contralateral lungs of patients compared to photon orcombined proton-photon RT.

ParticipantsMengmeng Jiang, Shanghai, China (Presenter) Nothing to DiscloseYiqian Zhang, MS, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging Healthcare Co, LtdMin Ji, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging Healthcare Co, LtdJunshen Xu, Beijing, China (Abstract Co-Author) Nothing to DiscloseJiyong Wang, MSc, Jiading, China (Abstract Co-Author) Researcher, Shanghai United Imaging Healthcare Co, LtdNan-Jie Gong, Houston, TX (Abstract Co-Author) Employee, UIH America, IncQiang Li, PhD, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging Healthcare Co, LtdXiuzhong Yao, Shanghai, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

In this study, we investigated whether epidermal growth factor receptor (EGFR) mutation status in non-small cell lung cancer(NSCLC) can be assessed from quantitative as well as qualitative features extracted from both CT and PET.


Eighty patients with stage II and III NSCLC and a confirmed EGFR mutation status (30 patients were positive and 50 were negativefor EGFR mutation), who underwent PET/CT between January 2017 to December 2017, were included in this study. We extracted514 quantitative features from PET/CT (257 for PET and 257 for CT) and 12 qualitative features from CT. Principal componentanalysis (PCA) was applied for feature selection. We selected principal components retaining 95% of the variability from all features.We then rebuilt the original features using the selected principal components and the original features were selected thatcorrelated by at least 99% to the rebuilt features. Finally, 5 qualitative features, 24 quantitative features for CT as well as 10quantitative features for PET were selected. A predictive model of EGFR mutation in terms of selected features using generalizedlinear regression with lasso regularization. The regularization parameter was selected through a 10-fold cross validation. Allstatistical analysis were performed in R software version 3.4.4.

RESULTS

With the total of 39 features selected which were significantly associated with EGFR mutation status, a predictive model forassociating image features with EGFR positive/negative was built. We estimated the performance of the model using the area underthe receiver operating characteristic curve (AUC). The result revealed an AUC=0.74.

CONCLUSION

By combing the PET-CT images together with first generation gene testing data, we investigate the relationship between imagefeatures and EGFR mutation status and built a radiogenomics model which can predict whether the patients have EGFR mutation ornot from a certain number of qualitative features as well as quantitative features.


A non-invasive method from image features to predict gene mutation status correlated with NSCLC and further advancing the roleof imaging in precision medicine.

Participants

MSRO37-06 Evaluation of the Tumor Response Using FDG-PET/CT Scans in Non-Small Cell Lung Cancer PatientsTreated with Proton or Photon Radiotherapy


Lorenzo Livi, Florence, Italy (Abstract Co-Author) Nothing to DisclosePierluigi Bonomo, Florence, Italy (Presenter) Nothing to DiscloseVieri Scotti, Florence, Italy (Abstract Co-Author) Nothing to DiscloseMarco Perna, Florence, Italy (Abstract Co-Author) Nothing to DiscloseLuca Visani, Florence, Italy (Abstract Co-Author) Nothing to DiscloseGabriele Simontacchi, Florence, Italy (Abstract Co-Author) Nothing to DiscloseVanessa Di Cataldo, Florence, Italy (Abstract Co-Author) Nothing to DiscloseIsacco Desideri, Florence, Italy (Abstract Co-Author) Nothing to DiscloseDaniela Greto, Florence, Italy (Abstract Co-Author) Nothing to DiscloseGiulio Francolini, Florence, Italy (Abstract Co-Author) Nothing to DiscloseLaura Masi, Florence, Italy (Abstract Co-Author) Nothing to DiscloseStefania Pallotta, MS, Florence, Italy (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

By definition, centrally located lung tumors are identified as a lesion located within 2 cm or touching the zone of the proximalbronchial tree or tumors immediately adjacent to the mediastinal or pericardial pleura. In these cases, the use of stereotactic bodyradiotherapy (SBRT) is debated due to the potential risk of severe toxicity. Currently, no high-level evidence is available to supportits use.


Between 2010 an 2015, 40 patients were treated with SBRT for 45 centrally located lesions. SBRT was delivered through either aLINAC-based intensity modulated radiotherapy (IMRT) technique or a robotic technique with Cyberknife. The prescribed total dosevaried between 26 and 60 Gy delivered in 1 or 8 fractions, respectively, with median BED10 of 69 Gy (range 37,5-105 Gy). OverallSurvival (OS) and Progression Free Survival (PFS) were reported using Kaplan-Meier method. Treatment-related toxicity wasevaluated according to CTCAE version 4.0

RESULTS

The median age of the cohort was 62 years (48-86). The majority of treated lesions were secondary hilar or mediastinallymphadenopathies (31/45, 69%), while unresectable primary tumors represented the remaining 14 cases (14/45, 31%). The mostcommonly used technique was VMAT for 21 lesions (47%), followed by Cyberknife for 14 (31%) and step and shoot IMRT for 10targets (22%), respectively. The predominant NSCLC histology was adenocarcinoma (32/45, 71%). The median longest tumordiameter was 31 mm (range 10-60 mm). At a median follow-up of 14.5 months, OS and PFS were 86.5%, 55.6%, 49.4% and 48.6%,24.1% and 12% at 1, 2 and 3 years, respectively. According to RECIST 1.1 criteria, a clinical benefit was achieved for 23 patients(57.5%) with a complete or partial response or stable disease in 4 (10%), 15 (37.5%) and 4 (10%) patients, respectively.Consistent with previous experiences using the same fractionation regimens, SBRT was well tolerated, with no G3/G4 toxicities: themost severe side effect was G2 esophagitis in 5/40 patients (12.5%).

CONCLUSION

In accordance with standardized risk-dose prescriptions, the use of SBRT for centrally located NSCLC was confirmed to be a safeand effective strategy. Prospective studies are warranted to support its use with high level evidence.


Our single-center experience adds to the limited available evidence on the feasibility and clinical benefit of SBRT for centrallylocated NSCLC

ParticipantsPegah Jahangiri, MD, Philadelphia, PA (Presenter) Nothing to DiscloseKamyar Pournazari, MD,MSc, Philadelphia, PA (Abstract Co-Author) Nothing to DiscloseCharles B. Simone II, MD, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseDrew A. Torigian, MD, MA, Philadelphia, PA (Abstract Co-Author) Co-founder, Quantitative Radiology Solutions LLCAbass Alavi, MD, Philadelphia, PA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Lung cancer is one the leading causes of death worldwide. Radiation therapy (RT) is a major treatment option for lung cancer,including for unresectable locally advanced non-small cell lung cancer (LA-NSCLC). The aim of this study was to evaluate theresponse of the primary lung tumor to proton versus photon RT using 18F-fluorodeoxyglucose (FDG)-PET/CT in patients with LA-NSCLC.


Thirty-nine consecutive patients who underwent FDG-PET/CT imaging pre- and post- proton or photon RT were assessed. Patientswere predominantly female (53.8%) with a median age of 67 years and with predominantly stage IIIA (62%). An adaptive contrast-oriented thresholding algorithm was applied to measure metabolically active tumor volumes, uncorrected SUV, partial volumecorrected SUV and total lesion glycolysis. Parameters of FDG-PET/CT scans before and after RT were compared using two-tailedpaired t-tests.

RESULTS

Among the 9 patients who received photon RT and the 9 patients who received a combination of proton-photon RT, there was asignificant decrease in PVC-TLG. Interestingly, among the 21 patients treated exclusively with proton RT, all tumor parametersincluding MTV, SUVmax, uncorrected SUVmean, PVC-SUVmean, uncorrected TLG, and PVC-TLG after treatment decreasedsignificantly (all p <0.001). The decreases in PVC-TLG and tumor PVC-SUVmean were more obvious than non-PVC ones (ΔPVC-TLG-357.26 cc versus ΔTLG -252.92 cc; ΔPVC-SUVmean -16.2 versus ΔSUVmean -10.19).

CONCLUSION

Adaptive contrast-oriented thresholding algorithm is a promising method to quantify whole tumor glycolysis in LA-NSCLC, and ourfindings demonstrates that proton RT is as effective as photon RT metabolically in inducing tumor response of LA-NSCLC.


Proton RT, which is much safer, is as effective as photon RT in treatment of LA-NSCLC.

SSJ03-01 Utility of Dual Energy CT for Assessment of Myocardial Fibrosis: Comparison of Single-Phase IodineMap Extracellular Volume Fraction (ECV) With Standard Multiphase ECV Technique

Tuesday, Nov. 27 3:00PM - 3:10PM Room: E353A

SSJ03

Cardiac (Myocardial Disease)


CA MR NM



ParticipantsPrachi P. Agarwal, MD, Canton, MI (Moderator) Nothing to DiscloseDavid A. Bluemke, MD,PhD, Bethesda, MD (Moderator) Nothing to DiscloseJacobo Kirsch, MD, Weston, FL (Moderator) Nothing to Disclose

Sub-Events


ParticipantsJamie L. Schroeder, MD,DPhil, Baltimore, MD (Presenter) Nothing to DiscloseAnkur Pandey, MD, Baltimore, MD (Abstract Co-Author) Nothing to DisclosePallavi Pandey, MD, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseMatthew Czarny, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseRani K. Hasan, MD, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseCheng Ting Lin, MD, Baltimore, MD (Abstract Co-Author) Nothing to DiscloseIhab R. Kamel, MD, PhD, Baltimore, MD (Abstract Co-Author) Research Grant, Siemens AGElliot K. Fishman, MD, Baltimore, MD (Abstract Co-Author) Institutional Grant support, Siemens AG; Institutional Grant support,General Electric Company; Co-founder, HipGraphics, IncStefan L. Zimmerman, MD, Ellicott City, MD (Abstract Co-Author) Project consultant, Siemens Healthcare; Research grant, AmericanHeart Association;


[email protected]

PURPOSE

Diffuse myocardial fibrosis is associated with chronic heart diseases, including aortic stenosis, and may be quantified using cardiacCT or MRI calculation of extracellular volume fraction (ECV). This study compared a dual-energy CT approach using an iodine mapfrom a single phase to calculate ECV to the standard multiphase technique to calculate ECV.


61 patients with aortic stenosis (35 male, 26 female; average age = 81 years) undergoing preoperative CT for transcatheter aorticvalve replacement (TAVR) were scanned on a Force (Siemens Inc.) using dual energy mode prior to (non-contrast phase) and 10minutes following (delayed phase) administration of 120-150 cc of iodinated contrast. Matched 1cm² regions of interest wereselected in the interventricular septum and ventricular blood pool in each phase. The ROI densities obtained from blended virtual120 kV images in each phase were used to compute the ECV (standard technique), while the dual energy data from the delayedphase only was used to calculate ECV using iodine map technique. Bias between ECV computation methods was visualized usingBland-Altman plot. Linear regression with Pearson technique was also performed.

RESULTS

The mean ECV calculated using standard method was 28%, while mean ECV measured using the iodine map method was 30%; for amean bias of +2% [95% CI, -5% to +9%]. The Pearson correlation coefficient between the two ECV measurements was R² = 0.48.

CONCLUSION

There was good correlation between ECV measured by standard multiphase technique and ECV measured by dual energy techniqueusing iodine maps created from the delayed phase alone. The dual energy technique trended towards mild overestimation of fibrosis,possibly attributable to beam hardening, however this difference was not significant.


Prior studies using histopathology and cardiac MRI have shown that the greater degree of myocardial fibrosis can predict worseoutcomes after surgery. ECV is an accepted way to measure myocardial fibrosis. This study demonstrates a dual energy methodthat simplifies ECV measurement and reduces radiation dose without significant difference relative to standard measurement of ECV.

SSJ03-02 Occurrence of Myocardial Oedema After Sporting Event? Quantification by T1 and T2 Mapping


SSJ03-03 Focal and Interstitial Fibrosis in Highly Trained Endurance Athletes


Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Stefan L. Zimmerman, MD - 2012 Honored EducatorStefan L. Zimmerman, MD -2015 Honored EducatorIhab R. Kamel, MD, PhD - 2015 Honored EducatorElliot K. Fishman, MD - 2012 Honored EducatorElliot K.Fishman, MD - 2014 Honored EducatorElliot K. Fishman, MD - 2016 Honored EducatorElliot K. Fishman, MD - 2018 Honored Educator

ParticipantsJitka Starekova, MD, Hamburg, Germany (Presenter) Nothing to DiscloseEnver G. Tahir, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseMonica Patten, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseMaxim Avanesov, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseJulius M. Weinrich, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseSebastian Bohnen, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseUlf K. Radunski, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseKai Mullerleile, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseGerhard B. Adam, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseGunnar K. Lund, MD, Hamburg, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Purpose of this study was to analyse the occurrence of myocardial oedema in triathletes after sporting events using T2 and T1mapping.


29 competitive asymptomatic triathletes (45 ±10 years) underwent a CMR study performed on a 1.5T Achieva (Philips) before(baseline) and after a sporting event (follow-up). CMR protocol included SSFP cine, T2w-GraSE and T1, T2 mapping using MOLLI5(3)3 sequence. Additionally, LGE Imaging was performed in the CMR baseline study. T1 and T2 quantification was performed usingthe OsiriX Software.

RESULTS

CMR performed before the sporting event revealed a normal global T1 time (with 989 ±28ms) of the left ventricular (LV)myocardium. There was no significant change in the global T1 time after the sporting event (988 ±21ms; p=0.926). Furthermore, T2time was in the normal range before the sporting event (54 ±3ms) without relevant change after the sporting event (53 ±3ms,p=0.797). In 10 of the 29 Triathletes (34%) a focal myocardial fibrosis with a non-ischemic pattern was detected (LGE+triathletes). Likewise, in the subgroups (LGE+ und LGE-) no significant changes in T1, T2 times before and after the sporting eventwere detected.

CONCLUSION

In contrast to the previously published T2 STIR data our results using T1, T2 mapping have not revealed any myocardial oedemaafter sporting events. Exercise-induced myocardial oedema previously detected using T2 STIR sequences might be falselyinterpreted due to signal inhomogeneities.


No myocardial oedema indicating acute myocardial injury was detected using mapping methods after sporting events, suggestingthat competitive endurance events are safe.

ParticipantsBlanca Domenech, MD, Barcelona, Spain (Presenter) Nothing to DiscloseMaria Sanz-de-la-Garza, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseAlvaro Sepulveda, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseFatima Crispi, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseRosario Jesus Perea, Barcelon, Spain (Abstract Co-Author) Nothing to DiscloseAna Garcia-Alvarez, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseSusanna Prat-Gonzalez, Barcelona, Spain (Abstract Co-Author) Nothing to DiscloseMarta Sitges, Barcelona, Spain (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

There is evolving evidence that cumulative effects of intensive endurance exercise may induce a broad spectrum of rightventricular (RV) adaptation/remodelling patterns. Thus, our aim was to assess the prevalence of myocardial fibrosis (MF), focal andinterstitial, in the myocardium and its relationship with cardiac remodelling/adaptation, among a cohort of highly trained enduranceathletes as compared to control subjects.

SSJ03-04 Evaluation of a Shortened Cardiac MRI Protocol of Left Ventricular Examinations: DiagnosticPerformance of T1-Mapping and Myocardial Function Analysis as a Screening Method



93 highly trained endurance athletes (>12 hours training/week at least during the last 5 years; age: 36 ± 6 years; 52.7% male) and72 age and gender-matched controls underwent a resting cardiac magnetic resonance to assess biatrial and biventriculardimensions and function. The presence of focal MF was assessed by late gadolinium enhancement (LGE). In a subgroup of 28athletes, T1 mapping sequence was added and extracellular volume (ECV) measurements were performed in remote myocardium toasses interstitial MF.

RESULTS

High endurance training load was associated with larger bi-ventricular and bi-atrial sizes, mildly reduced systolic ventricularfunction, as compared to controls in both genders (p < 0,05). LGE was significantly more prevalent in athletes (n=35, 37.6% vs2.8%; p < 0,001), with a constant pattern in the RV insertion points (Figure 1). Among men population, those athletes with LGEtended to have trained for more hours per week (14.55±3.6 vs 12.21±3.4, P = 0.07). In T1 mapping sequences, abnormal ECVvalues (>28%) were only found in 2 of 28 subjects. Those athletes who had focal fibrosis had higher ECV at remote myocardiumthan those without LGE (27,3±1,8 vs 25,1±2,2; P = 0,01).

CONCLUSION

Highly trained endurance athletes showed ten times higher prevalence of LGE than control subjects; always confined to the hingepoint. Although this pattern of LGE may be another feature of the athlete's heart, our results suggest that those with focal fibrosismight have globally higher myocardial ECV values. Its clinical impact is currently uncertain, and it still warrants further investigation.


Myocardial fibrosis, which is a predictive factor for adverse cardiac outcome, has been also described in some endurance athletesand its clinical meaning remains controversial.

ParticipantsJonathan Nadjiri, MD, Munich, Germany (Presenter) Nothing to DiscloseAnna-Lena Zaschka, Munich, Germany (Abstract Co-Author) Nothing to DiscloseDaniela Pfeiffer, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseAlexandra S. Straeter, Munich, Germany (Abstract Co-Author) Nothing to DiscloseMaximilian Englmaier, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseFlorian Weis, Munich, Germany (Abstract Co-Author) Nothing to DiscloseKarl-Ludwig Laugwitz, Munich, Germany (Abstract Co-Author) Nothing to DiscloseErnst J. Rummeny, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseMichael Rasper, Munich, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

In this study we sought to retrospective evaluate whether a very brief CMR protocol comprising only left ventricular(LV) functionanalysis and T1 mapping sufficiently distinguishes patients with relevant myocardial changes with need for further examination fromhealthy patients.


From October 2015 until October 2017 all patients with clinical indication for CMR for any myocardial characterisation (n = 160)were included. The scanner was a Philips Ingina 3T. The full CMR protocol comprised sBTFE Cine-Imaging, T1 and T2 mapping, T2wDark-Blood as well as Early- and Late-Gadolinium-Enhancement. Patients were categorized into two groups depending on presenceof LV dysfunction. ROC-analysis was done for results of T1-, T2- mapping and extracellular volume(ECV) in patients without LVdysfunction. Reference was depicted pathology in the conventional CMR techniques and report.

RESULTS

In the patient's cohort without LV dysfunction (n = 78 [49%]) ROC for T1 mapping was 81% with p < 0.001, 65 % for T2 mappingwith p= 0.4 and 82% for ECV with p < 0.001. T1 mapping was superior to T2 mapping by trend; p = 0.057. ECV was significantlysuperior to T1 mapping, p = 0.026. For maximum T1 relaxation times of 1300ms sensitivity was 83 % and specificity was 55%; thenegative predictive value was 91%. In patients with no LV dysfunction 31 (40%) patients did not exceed a maximum T1 of 1300ms;out of those none had significant myocardial alterations but 3 patients were diagnosed with chronic myocarditis. In general, out ofthe daily routine study population 111 (70%) patients had a pathological finding and in 49 cases (30%) CMR did not provideadditional information. In that group T1 mapping detected 57% of the patients who would not benefit from additional CMR.

CONCLUSION

A shortened CMR protocol comprising T1 mapping and LV-function analysis seems to rule out clinically relevant myocardialalterations. However, 3 cases of chronic myocarditis with normal LV-function were overlooked; yet therapeutic consequencesremain uncertain in this entity. These results need to be prospectively confirmed in a lager study to increase confidence in thisshortened protocol in clinical use.


The proposed protocol might allow for an improvement of efficiency of CMR examinations in the future.

SSJ03-05 Ferumoxytol-Enhanced MRI for Intramyocardial Blood Volume Mapping: Early Pre-Clinical Results


SSJ03-06 Simultaneous 18F-FDG PET/MR Study for Assessment of Different Stages of Cardiac Impairment inPatients with Anderson-Fabry Disease


ParticipantsKim-Lien Nguyen, MD, Los Angeles, CA (Presenter) Nothing to DiscloseJiaxin Shao, PhD, Los Angeles, CA (Abstract Co-Author) Nothing to DiscloseVahid K. Ghodrati, PhD, Los Angeles, CA (Abstract Co-Author) Nothing to DiscloseJ. Paul Finn, MD, Los Angeles, CA (Abstract Co-Author) Speakers Bureau, Bayer AG; Scientific Advisory Board, AMAGPharmaceuticals, IncPeng Hu, PhD, Los Angeles, CA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Myocardial perfusion is important for organ vitality. The intramyocardial blood volume (MBV) represents the fractional blood volumeof the intravascular space within a unit volume of myocardial tissue. Because ferumoxytol has high r1 relaxivity and a longintravascular half-life, we hypothesize that ferumoxytol-enhanced (FE) MRI may enable mapping of the MBV. We aim to evaluatethe vasodilator-induced variation in myocardial T1 signal in normal swine.


In this ARC-approved study, four healthy Yorkshire swine (33-52 kg) underwent FE-MRI at 3.0T under general anesthesia. Weacquired myocardial T1 maps using 5-(3)-3-(3)-3 MOLLI and FLASH-MOLLI pre- and post-ferumoxytol infusion (4mg/kg). Weinduced coronary vasodilation with 4-minute cycles of adenosine infusion (200-400 mcg/kg/min). Using in-house T1 fittingalgorithms, we generated myocardial T1 maps and derived T1 values from regions of interest drawn in the mid interventricularseptum of short axis FE T1 maps.

RESULTS

No adverse events occurred and vital signs were stable throughout the adenosine infusion and FE-MRI exam. Myocardial T1 signaldifferential between pre- and post-ferumoxytol was -48.5%±6.4%. The adenosine-induced native T1 response as reflected by theslope between baseline and peak adenosine was less robust when compared to FE T1 response (3.0±0.6ms/min vs -35.3±16.1ms/min, p=0.03). During the two adenosine-on cycles post-ferumoxytol, the FE T1 values steadily shortened due toadenosine-induced vasodilation and increasing MBV. During adenosine-off cycles, FE T1 values increased towards values at rest.For adenosine 200-300 mcg/kg/min, the FE T1 shortened (decreased) 10.2±5.4% from baseline. For 400 mcg/kg/min dose, the FET1 shortened >15% (swine #4). Of note, the increase in native T1 from baseline to peak adenosine was 0.7±0.2%.

CONCLUSION

Ferumoxytol, as a potent intravascular contrast agent, sensitizes the T1 signal to changes in the MBV and substantially amplifiesthe intramyocardial vascular T1 estimate. Additional work in models of varying myocardial perfusion is needed to better characterizethe T1 response in normal vs pathologic states.


Vasodilator-induced changes in ferumoxytol-enhanced myocardial T1 reflect dynamic changes in the intravascular myocardialcompartment and has implications for a panoply of myocardial disease states.


ParticipantsAndrea Ponsiglione, MD, Naples, Italy (Presenter) Nothing to DiscloseMassimo Imbriaco, MD, Napoli, Italy (Abstract Co-Author) Nothing to DiscloseCarmela Nappi, MD, Naples, Italy (Abstract Co-Author) Nothing to DiscloseEmanuele Nicolai, Napoli, Italy (Abstract Co-Author) Nothing to DiscloseMarco Aiello, Napoli, Italy (Abstract Co-Author) Nothing to DiscloseSerena Dell'Aversana, MD, San Marcellino, Italy (Abstract Co-Author) Nothing to DiscloseAndreas Greiser, PhD, Erlangen, Germany (Abstract Co-Author) Employee, Siemens AGKelvin Chow, PhD, Chicago, IL (Abstract Co-Author) Employee, Siemens AGAlberto Cuocolo, Napoli, Italy (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

We evaluated the relationship between T1 mapping and 18F-FDG PET by cardiac PET/MR imaging in AFD female patients.


Seventeen AFD female patients with normal left ventricular (LV) function underwent simultaneous cardiac PET/MR (Biograph mMR;Siemens Healthcare, Erlangen, Germany) imaging after administration of 18F-FDG and gadolinium-DPTA for assessment of lategadolinium enhancement (LGE). In all patients and in 7 female controls T1 mapping was performed using pre-contrast T1 ModifiedLook-Locker Inversion-recovery prototype sequence. Mean T1 values were measured by drawing 6-pixel regions of interest in theseptal and lateral segments of LV apical, mid-ventricular and basal short-axis slices. Cardiac FDG uptake was quantified by

measuring the standardized uptake value in 17 myocardial segments in each subject. The coefficient of variation (COV, i.e. thestandard deviation divided by the average) of the uptake of the 17 segments was calculated as an index of heterogeneity in theheart and values >0.17 were considered abnormal.

RESULTS

Five patients showed focal LGE indicating intra-myocardial fibrosis and were excluded from the final analysis. Compared withcontrols, mean T1 values of AFD female patients were significantly lower (1238±51.1 vs. 1334.32±26.6, p<0.001). At PET, 7 out ofthe remaining 12 patients showed abnormal COV values suggesting inflammation pattern and the other 5 demonstrated normal COVvalues (0.32±0.1 vs. 0.12±0.03, p<0.005) with homogeneous FDG uptake. Patients with abnormal COV showed higher mean T1values of lateral segments of the mid-LV wall (1219.16±23.4 vs. 1154±62.1, p<0.05), suggesting a potential relationship betweenprogressive myocyte sphingolipid accumulation and inflammation.

CONCLUSION

This study highlights the role of 18F-FDG PET/MR imaging for early detection of cardiac involvement in AFD patients allowing toidentify different stages of disease progression. In particular, pseudo-normalization of T1 mapping values, associated with abnormalCOV values, may represent an intermediate "inflammatory" stage before the development of myocardial fibrosis.


Simultaneous cardiac 18F-FDG PET/MR imaging may allow early detection of cardiac involvement in AFD patients identifyingdifferent stages of disease progression.

SSJ04-01 Cardiac Keynote Speaker: Cardiac Complications of Oncology Therapy

Tuesday, Nov. 27 3:00PM - 3:20PM Room: E260

SSJ04-03 Myocardial Tissue Characterization in Rat Models of Anthracycline-Induced Cardiotoxicity: HistologicChange and Correlation with T1 Mapping Parameters


SSJ04

Science Session with Keynote: Cardiac (Oncology)


CA MR NM OI



ParticipantsGautham P. Reddy, MD, Seattle, WA (Moderator) Researcher, Koninklijke Philips NV; Daniel Ocazionez, MD, Houston, TX (Moderator) Nothing to DiscloseTina D. Tailor, MD, Durham, NC (Moderator) Nothing to Disclose

Sub-Events

ParticipantsGautham P. Reddy, MD, Seattle, WA (Presenter) Researcher, Koninklijke Philips NV;

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Gautham P. Reddy, MD - 2014 Honored Educator

ParticipantsYoo Jin Hong, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHeae Surng Park, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChul Hwan Park, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DisclosePan Ki Kim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseKyungsun Nam, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseDong Jin Im, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYoung Joo Suh, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYoung Jin Kim, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHye-Jeong Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseByoung Wook Choi, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJin Hur, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To examine the pathologic changes in rat models of cardiotoxicity and to determine correlation with quantitative magneticresonance imaging (MRI) parameters.


All experiments were approved by our institutional animal care and use committee. Cardiotoxicity models were induced by injectingadult male Sprague-Dawley rats with doxorubicin (1mg/kg, twice a week). Cardiac MRI was performed with a 9.4-T scanner (BrukerBiospin Co., Billerica, MA, USA) using cine, pre- and post-T1 mapping sequences. T1 mapping sequences were performed by using asaturation recovery Look-Locker sequence. Left ventricular ejection fraction (LVEF) was evaluated using cine imaging. Native T1and extracellular volume (ECV) were measured at the mid ventricle. All rats were sacrificed after MRI. Pathologic changes weregraded according to their degree and then scored (0:absent, 1:minimal, 2:mild, 3:moderate, 4:severe) and correlated with MRIparameters (native T1 (ms), ECV (%)) and LVEF(%))

RESULTS

A total of 10 control and 14 cardiotoxicity models were included. Rats were classified into two groups, the early (n=9) and late(n=5) group, by 6 weeks after modeling. In cardiotoxicity models, LVEF decreased (control vs. cardiotoxicity subjects: 74%,63.2%), native T1 and ECV increased (1,186 ms, 15.5% vs. 1,232.44 ms, 18.68%). The main histologic findings were vacuolarchanges, inflammation, interstitial edema, expansion of interstitial space, and fibrosis. In subgroup analysis, myocardial fibrosis,expansion of interstitial space scores were significantly different between the two groups (p=0.007, p=0.002). Other histologic

SSJ04-04 Myocardial Tissue Phase Mapping Detects Cardiac Dysfunction in a Mouse Model of Doxorubicin-Induced Cardiotoxicity


SSJ04-05 68Ga-Galmydar: A PET Imaging Tracer for Noninvasive Detection of Doxorubicin-InducedCardiotoxicity


factors (e.g. vacuolar changes, inflammation, interstitial edema), native T1, ECV, and LVEF were not significantly different betweenthe two groups. ECV was correlated with fibrosis (r=0.632, p=0.015), vacuolar change (r=0.705, p=0.005), and the sum ofhistologic scores (r=0.694, p=0.006). Native T1 was correlated with myocardial inflammation (r=0.638, p=0.012) and expansion ofinterstitial space (r=0.656, p=0.011)

CONCLUSION

Prominent changes in MRI parameters and pathologic findings were noted in both early and late cardiotoxicity models, even insubjects with preserved LVEF. ECV and native T1 values showed good correlations with histologic scores.


T1 mapping MRI is a useful quantitative method to detect anthracycline-induced cardiotoxicity.

ParticipantsNivedita Naresh, PhD, Chicago, IL (Presenter) Nothing to DiscloseBradley D. Allen, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseSol Misener, Chicago, IL (Abstract Co-Author) Nothing to DiscloseCynthia Yang, Chicag0, IL (Abstract Co-Author) Nothing to DiscloseAlexander Ruh, PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseNicola Bertolino, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJeremy D. Collins, MD, Chicago, IL (Abstract Co-Author) Consultant, Guerbet SA Grant, Siemens AG Grant, C. R. Bard, IncMichael Markl, PhD, Chicago, IL (Abstract Co-Author) Institutional research support, Siemens AG; Consultant, Circle CardiovascularImaging Inc; Zhuoli Zhang, MD,PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJames C. Carr, MD, Chicago, IL (Abstract Co-Author) Research Grant, Astellas Group; Research support, Siemens AG; Speaker,Siemens AG; Advisory Board, Guerbet SADaniele Procissi, MS, PhD, Pasadena, CA (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Cardiotoxicity following chemotherapy is common in cancer patients treated with anthracyclines such as doxorubicin (Dox).Conventional imaging approaches to identify anthracycline associated cardiac dysfunction have targeted reductions in ejectionfraction (EF) with newer approaches focusing on detecting alterations in myocardial strain. The purpose of this study was to usemyocardial tissue phase mapping (TPM) to study early functional changes in a mouse model of Dox-induced cardiotoxicity.


16 week old female C57Bl/6 mice (n = 9) were imaged at 7T. 25 mg/kg Dox was administered over 3 weeks in the form of 5 mg/kgsubcutaneous pellets (Innovative Research of America, Florida, USA). Mice were imaged at baseline, 6 weeks and 10 weeks post-treatment. MRI protocol included multi-slice cine MRI covering the entire LV, and TPM in a single mid-ventricular short-axis slice.The cine images were analyzed using Segment (Medviso, AB) to calculate EF. TPM was performed using 2D cine phase contrast MRIwith prospective ECG and respiratory triggering. Imaging parameters included: image/time resolution = 0.117x0.117x1 mm3x20.8 ms,tri-directional VENC=4 cm/s. TPM images were analyzed in MATLAB to measure global peak radial and longitudinal velocities atsystole and diastole. Following imaging at 6 weeks, 4 mice were sacrificed for histopathologic assessment utilizing terminaldeoxynucleotidyl transferase (TdT) dUTP nick-end labeling (TUNEL) to detect apoptotic cells.

RESULTS

There were no significant differences in EF (72±11% at baseline, 75±12% at 6 weeks and 63±2% at 10 weeks). Global systoliclongitudinal velocity was significantly reduced at 6 weeks (p=0.03) but our power was low to detect significant differences at 10weeks. Histopathologic results demonstrated minimal apoptosis in all mice, suggesting early-stage cardiotoxicity.

CONCLUSION

Using myocardial TPM, we detected cardiac dysfunction prior to reduction in EF and the onset of cardiomyocyte apoptosis in amouse model of Dox-induced cardiotoxicity. Future studies comparing this technique with other myocardial tissue and functionalcharacterization may demonstrate a role for TPM as an early biomarker of cardiotoxicity.


In the present study, we demonstrated that longitudinal systolic velocity quantified using myocardial tissue phase mapping mayrepresent an early imaging biomarker for doxorubicin-induced cardiotoxicity.

ParticipantsJothilingam Sivapackiam, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseShivesh Kabra, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseSylvia Speidel, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseRichard Laforest, PhD, Saint Louis, MO (Presenter) Nothing to DiscloseMichael P. Rettig, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseVijay Sharma, St. Louis, MO (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Doxorubicin (DOX; Adriamycin), an anthracycline analogue is widely used chemotherapeutic drug in cancer. DOX treatments aresusceptible to acute and chronic cardiac anomalies, including aberrant arrhythmias, ventricular dysfunction, and heart failure. PETtracers could also provide noninvasive assessment of early and reversible metabolic changes of the myocardium. Herein, we reporta preliminary investigation of 68Ga-Galmydar potential to monitor DOX-induced cardiomyopathy in vivo, ex vivo, and in celluloemploying both nuclear- and optical imaging.


Galmydar was obtained through a ligand exchange reaction. 68Ga-Galmydar was purified on C-18 column using radio-HPLC.MicroPET imaging was performed 5 d post treatment of rats either with a single dose of DOX (15 mg/kg) or vehicle as a control(saline). For correlation of PET imaging data, post-imaging quantitative biodistribution studies were also performed. In cellulo (H9c2)dose and time dependent doxorubicin treatments were also studied using live cell optical imaging.

RESULTS

68Ga-Galmydar, micro-PET static scan (10 min acquisition; 60 min post tail-vein administration) demonstrated 1.91-fold loweruptake in hearts of DOX-treated (Standard Uptake Value; SUV: 0.92, n=3) rats compared with their vehicle treated (SUV: 1.76, n =3) counterparts. The post imaging pharmacokinetic data demonstrated heart uptake values of 2.02 fold lower for DOX treated ratscompared to control counterparts (%ID/g; DOX: 0.44 ± 0.1, n=3; Control: 0.89 ± 0.03, n=3) thus correlating well with micro-PETimaging data. Employing moderate fluorescent traits of Galmydar, live cell optical imaging indicated a gradual decrease in uptakeand retention of Galmydar within mitochondria of H9c2 cells following DOX-treatment, while indicating also dose-dependentpharmacological response and time-dependent uptake profiles. Furthermore, the decreased uptake in H9c2 cells also correlated withcaspase-3 expression resulting from DOX-induced cardiotoxicity and cell death. Combined data indicate that 68Ga-Galmydar couldprovide a sensitive and specific readout of DOX-induced cytotoxicity attributed to depolarization of mitochondrial potential in heartcells.

CONCLUSION

68Ga-Galmydar could provide a noninvasive assessment of DOX-related early and likely reversible metabolic changes that remains tobe evaluated clinically.


none

SSJ05-01 Chest Invited Speaker: Beyond Morphology-Comprehensive Imaging of Pulmonary Disease with MRI

Tuesday, Nov. 27 3:00PM - 3:10PM Room: S404AB

SSJ05-02 Combination of MR Free-Breathing 3D T1-Weighted Star VIBE and DWI for the Differentiation ofBenign from Malignant Peripheral Pulmonary Lesions: A Comparative Study Using Routine-Dose CT


SSJ05-03 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography (PET/CT) andDiffusion-Weighted Magnetic Resonance Imaging (DWI-MRI) Diagnostic Performance in the

SSJ05

Science Session with Keynote: Chest (Thoracic MRI)


CH CT MR NM



ParticipantsJurgen Biederer, MD, Seeheim-Jugenheim, Germany (Moderator) Nothing to DisclosePatricia J. Mergo, MD, Jacksonville, FL (Moderator) Nothing to Disclose

Sub-Events

ParticipantsJuergen Biederer, MD, Heidelberg, Germany (Presenter) Nothing to Disclose

ParticipantsShan Dang, Xian, China (Presenter) Nothing to DiscloseGuangming Ma, Xianyang, China (Abstract Co-Author) Nothing to DiscloseDong Han, MD, Xianyang, China (Abstract Co-Author) Nothing to DiscloseNan Yu, MD, Xian Yang, China (Abstract Co-Author) Nothing to DiscloseWei Wang, MD, Xi'an, China (Abstract Co-Author) Nothing to DiscloseShaoyu Wang, Shanghai, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

The differentiation of benign pulmonary lesions from malignant pulmonary lesions is very difficult. High resolution CT is the mostcommonly used radiology methods for lung. While it is limited to children, child-bearing women and disorders requiring repeatedexaminations over prolonged periods because of its radiation dose.The result of our previous studies showed that: comparing withroutine-dose CT, the MR T1-weighted 3D Star VIBE sequence was slightly lower in differentiating the peripheral pulmonarylesions(PPL) with morphological features. MR can provide not only the morphological information, but also functional information.Theapparent diffusion coefficient (ADC) value was used widely in whole body, but this research of combining MR-DWI ADC value withmorphological characteristics of PPL was rare. The purpose of this study was to evaluate the no radiation-dose MR (ADC value andT1-weighted 3D Star VIBE sequence) diagnostic efficiency in differentiating the malignant PPL from benign, with the routine-doseCT as a reference standard.


Fourty-seven patients(30 males and 17 females,mean age 64.1 years old; age range 48-83 years) were enrolled in this study, allthe patients had undergone routine-dose CT,MR T1 Star VIBE and DWI with 3.0T MR sanner. These lesions were all diagnosed bytransthoracic needle biopsy or surgery. Two radiologists observed the morphological signs of MR and CT images independently. Theorder of observation was MRI first, and followed by CT. Then the ADC value of lesions were measured. The logistic regressionanalysis was used to calculate the probability. The ROC curve was used to analyze the capabilities of morphological characteristicsand DWI in distinguishing malignant PPL from benign.

RESULTS

There was significant difference of the ADC value between benign and malignant groups(p=0.001). The cut-off ADC value was1197×10-6 mm2/s. Combined MR T1 Star VIBE with ADC value can distinguish PPL better than only routine-dose CT, ADC value andT1 Star VIBE alone.

CONCLUSION

The ADC value could differentiate the peripheral pulmonary lesions initially, but the distinguishability was better if combing MR T1Star VIBE morphological characteristics with ADC value.


We can use MR T1-weighted 3D Star VIBE and DWI to replace routine-dose CT to distinguish PSPLs in order to avoid radiationexposure.

Evaluation of Pulmonary Lesions: A Systematic Review and Meta-Analysis


SSJ05-04 Chemical Exchange Saturation Transfer (CEST) Imaging versus Diffusion-Weighted Imaging (DWI)versus FDG-PET/CT: Capability for Diagnosis of Solitary Pulmonary Nodule



ParticipantsAdriano Basso Dias, MD, Porto Alegre, Brazil (Presenter) Nothing to DiscloseMatheus Zanon, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseStephan Altmayer, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseGabriel Sartori Pacini, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseNatalia H. Concatto, MD, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseTassia M. Medeiros, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseAnderson Garcez, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseRicardo H. Do Amaral, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseIrai L. Giacomelli, MD, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseEdson Marchiori, MD, PhD, Rio de Janeiro, Brazil (Abstract Co-Author) Nothing to DiscloseKlaus L. Irion, MD, PhD, Liverpool, United Kingdom (Abstract Co-Author) Nothing to DiscloseNupur Verma, MD, Gainesville, FL (Abstract Co-Author) Nothing to DiscloseTan-Lucien H. Mohammed, MD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseGuilherme Watte, Porto Alegre, Brazil (Abstract Co-Author) Nothing to DiscloseBruno Hochhegger, MD, PhD, Porto Alegre, Brazil (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To perform a systematic review and meta-analysis of the diagnostic performance of DWI-MRI and 18F-FDG PET/CT in theevaluation of pulmonary lesions.


Databases of MEDLINE and Embase were searched through December 2017. Studies published in English were included when thediagnostic performances of 18F-FDG PET/CT or DWI in detecting malignant pulmonary lesions were clearly identified in the articles.Two reviewers evaluated the study quality of all selected articles using QUADAS-2 and only those that met a minimum quality scorewere included. Parameters of lesion quantification were analyzed separately for each imaging modality (e.g., lesion-to-spine ratio(LSR), and apparent diffusion coefficient (ADC) for DWI-MRI). Meta-analysis using a random-effects model were conducted tocalculate the pooled sensitivities, specificities, positive and negative likelihood ratios (PLR and NLR), diagnostic odds ratios (DOR)and area under the curve (AUC) for PET/CT and DWI with 95% confidence intervals (95% CI).

RESULTS

The literature search yielded 1280 results, and the inclusion criteria were met by 37 studies (23 FDG PET/CT studies, 8 MRI studiesand 6 studies including both methods), with a total of 4224 participants and 4463 lesions (malignant, n = 3090, 69.2%; benign,n=1362, 30.8%). Pooled sensitivity and specificity of SUVmax (n = 25) were 0.86 (95%CI, 0.82-0.90) and 0.73 (95%CI, 0.62-0.82),respectively. For DWI-MRI, LSR studies (n = 4) showed a sensitivity of 0.81 (0.71-0.88) and a specificity of 0.90 (0.79-0.95),whereas studies utilizing ADC (n = 12) had a sensitivity and specificity of 0.83 (0.77-0.88) and 0.86 (0.76-0.92), respectively.DWI-LSR yielded the greatest diagnostic odds ratio (DOR = 38, 95%CI 12-115) compared to DWI-ADC (DOR = 30, 95%CI 14-66)and SUVmax (DOR = 17, 95%CI 10-28).

CONCLUSION

Diagnostic performance of DWI-MRI is comparable to 18F-FDG PET/CT for the evaluation of potentially malignant pulmonary lesions.


This is the first meta-analysis to compare the diagnostic performance of DWI-MRI and 18F-FDG PET/CT in the evaluation ofpulmonary lesions. Our study demonstrated that the diagnostic performance of DWI-MRI is comparable or even superior to that ofPET/CT, which supports the inclusion of MRI as a low-cost and radiation-free option to the diagnostic work-up of pulmonarylesions.

ParticipantsYoshiharu Ohno, MD, PhD, Kobe, Japan (Presenter) Research Grant, Canon Medical Systems Corporation; Research Grant,Koninklijke Philips NV; Research Grant, Bayer AG; Research Grant, DAIICHI SANKYO Group; Research Grant, Fuji Pharma Co, Ltd;Research Grant, Guerbet SA; Masao Yui, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationYuji Kishida, MD,PhD, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseShinichiro Seki, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationKatsusuke Kyotani, RT,MSc, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseTakeshi Yoshikawa, MD, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationTakamichi Murakami, MD, PhD, Osakasayama, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To directly and prospectively compare the capability for diagnosis of solitary pulmonary nodules (SPNs) among chemical exchange

SSJ05-05 2D-ROI and 3D-VOI Histogram Analysis of Apparent Diffusion Coefficient in Assessing SolitaryPulmonary Lesions


SSJ05-06 Utilization of 19F MRI for Identification of Iraqi-Afghanistan War Lung Disease


To directly and prospectively compare the capability for diagnosis of solitary pulmonary nodules (SPNs) among chemical exchangesaturation transfer (CEST) imaging, diffusion-weighted imaging (DWI), and FDG-PET/CT.


113 consecutive patients (69 male and 44 female; mean age 71 year old) with 122 SPNs underwent CEST imaging and DWI at a 3TMR system, FDG-PET/CT, and pathological and/or follow-up examinations. According to final diagnoses, all SPNs were divided intomalignant (n=76) and benign (n=46) SPNs. In each patient, magnetization transfer ratio asymmetry (MTRasym) was calculated fromz-spectra at 3.5ppm in each pixel, and MTRasym map was computationally generated from CEST data. In each lesion, MTRasym,apparent diffusion coefficient (ADC) and SUVmax were assessed by ROI measurements. To compare all indexes between twogroups, Student's t-test was performed. Then, multivariate logistic regression analyses were performed to investigate thediscriminating factors of two groups. In addition, ROC analyses were performed to compare diagnostic performance among allindexes as well as combined methods. Finally, sensitivity, specificity and accuracy were compared among all methods by McNemar'stest.

RESULTS

MTRasym, ADC and SUVmax had significant difference between malignant and benign SPNs (p<0.05). Multivariate regressionanalyses identified MTRasym (Odds ratio [OR]: 0.54), ADC (OR: 47.6) and SUVmax (OR: 0.47) as significant differentiators(p<0.05). ROC analyses showed area under the curve (Az) of MTRasym (Az=0.88) and combined all indexes (Az=0.92) weresignificantly larger than that of SUVmax (Az=0.77, p<0.05). Sensitivity (SE) and accuracy (AC) of MTRasym (SE: 81.6 [62/76] %,AC: 82.8 [101/122] %) and combined methods (SE: 85.5 [65/76] %, AC: 85.2 [104/122] %) were significantly higher than those ofADC (SE: 69.7 [53/76] %, p<0.05; AC: 77.9 [95/122] %, p<0.05) and SUVmax (SE: 64.5 [49/76] %, p<0.05; AC: 71.3 [87/122] %,p<0.05).

CONCLUSION

CEST imaging has a better potential and can improve diagnostic performance of SPNs, when compared with DWI and FDG-PET/CT.


CEST imaging has a better potential and can improve diagnostic performance of SPNs, when compared with DWI and FDG-PET/CT.

ParticipantsShi Y. Deng, Guangzhou, China (Presenter) Nothing to DiscloseXinhua Wei, MD, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseXinchun Li, Guangzhou, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To determine the diagnostic performance of apparent diffusion coefficient (ADC) histogram analysis derived from both largest cross-sectional (2D-ROI) and whole tumor region of interest(3D-VOI) in differentiating benign from malignant solitary pulmonarylesions(SPLs).


Sixty-nine patients with pathologically confirmed lung lesions (benign: malignant = 23:46) were included in the study. All patientsunderwent 3.0T diffusion-weighted imaging (DWI) with 2 b values of 0 and 600s/mm2. The histogram metrics including minimum,mean, maximum, 10th, 25th, 50th, 75th and 90th percentiles, skewness, and kurtosis were calculated from the largest cross-section and whole tumor ROI, respectively. Inter-class correlation characteristic(ICC) was used to assess inter-observer reliability.Histogram metrics were analyzed using Mann-Whitney U-test. The diagnostic performance was evaluated using receiver-operatingcharacteristic (ROC) analysis.

RESULTS

Minimum, mean, maximum and 10th,25th,50th,75th,90th percentile ADCs with two groups were significantly lower (all P<0.05),except for the maximum and 25th percentile ADCs in whole-volume group (P=0.128, P=0.221) in malignant lesions compared withbenign ones. The 75th and 50th percentile ADCs in two ROI setting group respectively achieved the highest AUC (single-slice:whole-volume=0.891:0.894) with cutoff value of 1.57×10-3 mm²/s and 1.41×10-3 mm²/s in differentiating solitary pulmonarylesions. ICC for the whole-volume ROIs(0.76~0.97) was better the largest-slice ROIs(0.59~0.91) .

CONCLUSION

Both single-slice and whole-volume ADCs are helpful for distinguishing malignant from benign lung lesions. Whole-volume ADChistogram analysis could have greater diagnostic properties and repeatability.


In conclusion, ADC histogram is helpful for distinguishing malignant from benign lung lesions, and the 75th percentile ADC derivedfrom 2D-ROI and 3D-VOI could provide better imformation in characterizing in SPLs, with no statistical difference. Moreover, placingROIs in the largest lesion would be more suitable for clinic practice considering about saving time.

ParticipantsJoseph G. Mammarappallil, MD,PhD, Durham, NC (Presenter) Nothing to Disclose

Scott Shofer, MD, Durham, NC (Abstract Co-Author) Nothing to DiscloseJesse L. Tucker, MD, Durham, NC (Abstract Co-Author) Nothing to DiscloseMartha Carraway, MD, Durham, NC (Abstract Co-Author) Nothing to DiscloseSamantha Womack, Durham, NC (Abstract Co-Author) Nothing to DiscloseCecil Charles, PhD, Durham, NC (Abstract Co-Author) Research Grant, Merck & Co, Inc


[email protected]

PURPOSE

Iraq-Afghanistan War Lung Injury (WLI) describes new onset respiratory symptoms occurring in deployed soldiers to the Middle Eastthat can ultimately lead to constrictive bronchiolitis. This study sough to determine if 19F MRI could identify patients with WLI.


Three soldiers who presented to local VA clinics were evaluated for reactive airways disease, post deployment dyspnea, anddecreased respiratory fitness. All subjects had a full pulmonary function evaluation. Inspiratory and expiratory imaging with HRCTwas also obtained for each subject. For 19F MRI studies, each subject inhaled a gaseous mixture of 79% PFP mixed with 21% O2mixture over the course of several minutes in a protocol consisting of three tidal breaths followed by a 6 second breath-hold attotal lung capacity during which a 3D imaging of the lung airspaces was obtained. The image data was analyzed to generateregional wash-in and wash-out time constants (seconds) throughout the lung airspaces. Finally, the fraction of slow fillingcompartments was calculated as the number of volume elements exhibiting a wash-in time constant of > 100 seconds divided bythe total number of lung airspace elements in the imaged [FVR1>100].

RESULTS

Two subjects were non-smokers and the third is a current smoker (11 pack-years). All three subjects were exposed to aerosolizedcontaminants during deployment in Iraq/Afghanistan. Spirometry for all subjects were normal FEV1% predicted (81, 109, 86),FEV1/FVC (72, 75, 73). Expiratory CT imaging was normal for subject 1, while the 2nd and 3rd subjects had mild basilar or lobularareas of air trapping. Imaging with 19F for each subject was as follows: FVR1>100 = 17.5%, 37.8%, and 24.5% for the threesubjects respectfully. While the first subject's FVR1>100 was close to values seen in subjects with normal lung function the secondand third subjects more resembled patients diagnosed with COPD (25->60%).

CONCLUSION

19F MRI demonstrated delayed regional filling of PFP gas in two subjects with suspected WLI when compared to data from normalsubjects.


19F MRI has the potential to detect airway abnormalities at earlier time points than current techniques. This may ultimately lead tobetter diagnosis of challenging airway abnormalities such as WLI and perhaps a tool for evaluation of interventions.

SSJ14-01 Multispectral Optoacoustic Tomography of Systemic Sclerosis: A Feasibility Study Using a HybridApproach at the Patient's Bedside


SSJ14-02 Optimization of Chest Acquisition on Whole Body PET/MRI for HCC Tumor Staging Using Non-Cartesian K Space Reconstructions


SSJ14

Molecular Imaging (Novel Multi-Modal Applications)


CT MR MI NM



ParticipantsCiprian Catana, MD, PhD, Charlestown, MA (Moderator) Nothing to DiscloseSabah Servaes, MD, Philadelphia, PA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsAnne Helfen, MD, Muenster, Germany (Presenter) Nothing to DiscloseMax Masthoff, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseUlrich Gerth, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseWalter L. Heindel, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMichel Eisenblaetter, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMoritz Wildgruber, MD, PhD, Munster, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

Systemic sclerosis is an autoimmune disease with dysfunctional connective tissue repair inducing skin and internal organ fibrosis aswell as vasculopathy, leading to serious complications such as digital ulcers. Microvascular dysfunction is a prognostic marker formorbidity and mortality in systemic sclerosis. However, due to a lack of functional tissue information using established imagingtechnologies, risk stratification remains challenging.We aimed to evaluate the clinical feasibility of a hybrid multispectraloptoacoustic tomography (MSOT)/ultrasound (US) approach in systemic sclerosis.


We used a combined handheld MSOT/US imaging system (iThera Medical, Munich) for imaging fingers D2 to D5 of systemic sclerosispatients (n=7, n=56 fingers) and healthy volunteers (n=8, n=64 fingers). In subcutaneous tissue, tissue levels of hemoglobin(deoxygenated (HbR), oxygenated (HbO2) and total (HbT)) were calculated after spectral unmixing. Furthermore, MSOT valueswere analysed to stratify patients with either progressive or stable disease activity. Statistical analysis was performed usingunpaired t test.

RESULTS

In systemic sclerosis, MSOT values for HbR, HbO2 and HbT were significantly lower compared with healthy volunteers. On anindividual basis, all patients had - as compared to healthy volunteers - a reduced ratio (<1.0) for HbO2 (ratio 0.53, p<0.0001) andHbT (ratio 0.49, p<0.0001). Additionally, in systemic sclerosis patients with a progressive disease significantly lower MSOT valueswere detected than in patients with a stable disease activity (HbR: 27.31 vs. 33.03 AU, HbO2: 23.87 vs. 29.22 AU, HbT: 51.18 vs.62.24 AU; p<0.001).

CONCLUSION

This preliminary study demonstrates the feasibility of MSOT imaging to detect and quantify microvascular dysfunction in systemicsclerosis and to distinguish between progressive and stable disease activity. Therefore, MSOT may serve as a valuable non-invasivetool for disease stratification and therapy monitoring.


As an easy-to-use, non-invasive and contrast-free imaging technique hybrid MSOT/US may facilitate both diagnosis and monitoringof treatment response in systemic sclerosis at the patient's bedside.

ParticipantsMathilde Vermersch, Lille, France (Presenter) Nothing to DiscloseSebastien Mule, MD,PhD, Creteil, France (Abstract Co-Author) Nothing to DiscloseBerivan Emsen, Creteil, France (Abstract Co-Author) Nothing to DiscloseAurelien Monnet, Creteil, France (Abstract Co-Author) Employee, Siemens AGJulia Chalaye, MD, Creteil, France (Abstract Co-Author) Nothing to Disclose

SSJ14-03 Whole-Body Functional PET-MR for Multiple Myeloma Staging: Impact of Sequence Design on BoneMarrow Infiltration and Focal Lesion Assessment


Laurence Baranes, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseEdouard J. Herin, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseFrederic Pigneur, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseAlain Rahmouni, MD, Nogent Sur Marne, France (Abstract Co-Author) Nothing to DiscloseEmmanuel Itti, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseAlain Luciani, MD,PhD, Creteil, France (Abstract Co-Author) Research Consultant, Bracco Group; Research Grant, Bracco Group;Research Consultant, General Electric Company; Research Consultant, Siemens AG

PURPOSE

To compare diagnostic performance and image quality for chest imaging of stack-of-stars T1 EG acquisition (STARVIBE) with noncartesian reconstructions of k-space, and Caipirinha 3D T1 EG acquisition, during whole-body PET-MR of HCC patients.


33 consecutive patients referred for HCC staging in PET MRI (Biograph mMR, Siemens erlangen) were retrospectively included. Allbenefited from Caipirinha Dixon 3D T1 (T1EG Dixon post gadolinium) and Starvibe (post-gadolinium T1EG Dixon with radial filling ofthe Fourrier plan). A subjective evaluation of the image quality was performed. Detection of pulmonary nodular lesions with bothsequences was compared by taking the chest CT scan as the standard of reference (Exact Fisher test). The signal-to-noise ratioof lung parenchyma (SNR) and contrast-to-noise ratio (CNR) of nodules, muscles and pulmonary vessels were compared betweenthe two sequences (Student T-test).

RESULTS

The Starvibe sequence was judged subjectively better in 42% of cases and at least equivalent to Vibe in 88% of cases. It improvedSNR (375% vs 89%, p <0.001) and CNR for muscles (53% vs 41%, p <0.001), vessels (84% vs 73%, p <0.001) and pulmonarynodules (34% vs 24%, p = 0.03). The detection of nodules was improved (sensitivity = 55% vs 41%) with a decrease in falsepositives (VPP = 92% vs 35%), without significant difference.

CONCLUSION

K space non cartesian reconstruction improves MRI's diagnostic performance at the chest stage by elevation of SNR and CNR.


This new type of sequence without apnea could allow an improvement of the MRI diganostic performances especially at thethoracic stage in patients followed in oncology.

ParticipantsRobert E. Burns, MD, Creteil, France (Presenter) Nothing to DiscloseSebastien Mule, MD,PhD, Creteil, France (Abstract Co-Author) Nothing to DisclosePierre Zerbib, Creteil, France (Abstract Co-Author) Nothing to DiscloseAurelien Monnet, Creteil, France (Abstract Co-Author) Employee, Siemens AGLaurence Baranes, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseDominik Nickel, Erlangen, Germany (Abstract Co-Author) Employee, Siemens AGAlto Stemmer, PhD, Erlangen, Germany (Abstract Co-Author) Employee, Siemens AGKarim Belhadj, Creteil, France (Abstract Co-Author) Nothing to DiscloseCorinne Haioun, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseAlain Rahmouni, MD, Nogent Sur Marne, France (Abstract Co-Author) Nothing to DiscloseEmmanuel Itti, MD, Creteil, France (Abstract Co-Author) Nothing to DiscloseAlain Luciani, MD,PhD, Creteil, France (Abstract Co-Author) Research Consultant, Bracco Group; Research Grant, Bracco Group;Research Consultant, General Electric Company; Research Consultant, Siemens AG

PURPOSE

To study the impact of functional simultaneous 18FDG-PET-MR on bone marrow infiltration evaluation and focal lesion detection ininitial staging of Multiple Myeloma (MM).


25 consecutive patients referred to our centre for initial staging of a monoclonal gammopathy were included. PET-MR sequencedesign included whole-body T1-weighted spin-echo and T2-fat supressed DIXON weighted images (CWB-MRI) combined to whole-body Simultaneous Multi-Slice diffusion-weighted imaging (WB-DWI), 3D whole-body isotropic dynamic contrast-enhanced images(WB-DCE) and 3D-T1 weighted images combined with PET (T1-PET) acquisition using 4-5 MBq/kg of 18FDG-glucose. In the firstpart of the evaluation, the mean apparent diffusion coefficient (mADC), the mean enhancement (mE) and mean maximal standarduptake value (mSUV) of the bone marrow measured at the posterior part of the ischium, the lowest lumbar vertebral, and lowestthoracic vertebral possible outside of focal lesions were measured. The second part of the evaluation comprised the counting offocal lesions on CWB-MRI, CWB-MRI combined with WB-DWI, CWB-MRI combined with WB-DCE, CWB-MRI combined with PET, andPET-MR. Based on these findings, each patient was allocated a specific staging for bone marrow infiltration (BMI) and the numberof focal lesions.

RESULTS

Functional imaging significantly impacted assessment of BMI; Compared to CWB-MRI, WB-DWI upstaged 7 and down staged 8patients, WB-DCE upstaged 17 and down-staged 5 patients, PET upstaged 6 and down-staged 4 patients. Correlation betweenfunctional imaging and percentage of plasma cell infiltration were [r=0.7655, p<10^-6], [r=0.549 p<0.005], [r=0.678, p=0.0001] formADC, mE, and mSUV. Compared to CWB-MRI, PET-MRI detected 30 additional focal lesions among 8 patients and upstaged 6 ofthem. 4 patients were upstaged from MGUS to stage 1 Multiple Myeloma (MM), 1 patient from stage 1 to stage 2 MM, and 1 patientfrom stage 1 to stage 3 MM. The detection of additional focal lesions did not modify staging for 2 patients.

CONCLUSION

SSJ14-04 Multispectral Optoacoustic Tomography of Vascular Malformations


SSJ14-05 Ultra-Fast Whole-body PET/CT Enabled by Digital Photon Counting PET Detector Technology:Findings from a Phase II Clinical Trial


PET-MRI modifies initial staging of patients with a monoclonal gammopathy. Further studies are necessary to improve medullaryinfiltration evaluation as well as prospective longitudinal studies to evaluate the prognostic value of these modifications.


Functional PET-MRI significantly impacts initial staging of patients with suspected MM.

ParticipantsMax Masthoff, MD, Muenster, Germany (Presenter) Nothing to DiscloseAnne Helfen, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseJing Claussen, Munich, Germany (Abstract Co-Author) Employee, iThera Medical GmbHAngelos Karlas, Neuherberg, Germany (Abstract Co-Author) Nothing to DiscloseWalter L. Heindel, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseVasilis Ntziachristos, PhD, Munich, Germany (Abstract Co-Author) Stockholder, iThera Medical GmbHMichel Eisenblaetter, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMoritz Wildgruber, MD, PhD, Munster, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

Differential diagnosis and treatment of congenital vascular anomalies is challenging and misdiagnosis is frequent. A novel non-invasive imaging approach combining not only visualization of anatomical features but also quantitative assessment of moleculartissue biomarkers would aid both diagnosis and monitoring of treatment response of vascular anomalies. We aimed to evaluate thefeasibility of hybrid ultrasound (US) and multispectral optoacoustic tomography (MSOT) for non-invasive, real-time imaging ofvascular malformations.


In this pilot study 6 patients with arteriovenous (AVM) and 6 patients with venous (VM) malformations were investigated with aclinical hybrid MSOT/US system before and after either endovascular embolization (AVM) or percutaneous sclerotherapy (VM). Allpatients were diagnosed with AVM or VM according to the classification system of the International Society for the Study ofVascular Anomalies (ISSVA). Region of interest analysis of the lesion and contralateral healthy tissue revealed quantitative valuesfor oxygenated (HbO2) and deoxygenated hemoglobin (HbR). Ratios of HbO2 over HbR were calculated for all vascular malformationsand healthy tissue before and after treatment.

RESULTS

HbO2/HbR ratio was significantly higher for AVM versus VM and compared with healthy tissue (1.82±0.08 vs. 1.12±0.04 vs.0.89±0.03, all p-values<=0.001). Therefore MSOT provided intrinsic biomarker patterns to distinguish arteriovenous from venousmalformations. After therapy the HbO2/HbR ratio decreased in correlation to treatment success validated by MRI and angiography.

CONCLUSION

Different types of vascular malformations are clearly distinguished by MSOT-based, non-invasive assessment of hemoglobin levels invascular malformations. Therapy effects could be instantly visualized and quantified.


As an easy-to-use, non-invasive and contrast-free imaging technique hybrid MSOT/US may facilitate both diagnosis and monitoringof treatment response of vascular malformations at the patient's bedside.

ParticipantsMichael V. Knopp, MD, PhD, Columbus, OH (Presenter) Nothing to DiscloseMichelle I. Knopp, Columbus, OH (Abstract Co-Author) Nothing to DiscloseTaylor Porter, Columbus, OH (Abstract Co-Author) Nothing to DiscloseKatherine Binzel, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseJun Zhang, PhD, Columbus, OH (Abstract Co-Author) Nothing to DiscloseFrederik L. Giesel, MD, MBA, Heidelberg, Germany (Abstract Co-Author) Patent application, F18-PSMA-1007Chadwick L. Wright, MD,PhD, Lewis Center, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To assess in a Phase II study the clinical feasibility of ultrafast whole body PET imaging using the new generation digital photoncounting PET and to compare by intraindividual comparison the diagnostic and quantitative findings to current clinical whole bodyPET acquisition.


52 patients scheduled for FDG whole body PET/CT were imaged using three separate acquisitions as part of an intra-individualcomparison study to compare a new genertaion digital system, dPET/CT (Vereos) with a current generation conventional sytem,cPET/CT (Gemini). Standard cPET imaging was performed at ~75 min p.i. of 13mCi FDG with investigational dPET imaged at ~55 minp.i. The first dPET acquisition was performed using 90s/bed position, immediately followed by a 9s/bed position acquisition lead toaverage table times of ~15 and ~2 min and compared with 90s/bed position cPET ~20 min . The 9s/bed dPET were reconstructedusing a previously optimized methodology. All other aspects of image acquisition were kept identical. Three blinded reviewers

SSJ14-06 SiPM-Based versus LYSO-based 68Ga-DOTA-TATE PET/CT: Comparison of Semi-QuantitativeMeasurements in Normal Tissues and Lesions


evaluated the data sets regarding visual characteristics, diagnostic confidence and semi-quantitative readouts.

RESULTS

All ultrafast scans were classified to be assessible. As expected, visual assessment scores were significant higher for 90s/bed dPETwhole body (p<.01), while no significant between the ultra-fast wholebidy and the cPET scans were reported. The ultra-fast scanpresented with slightly increased background noise levels. The ultra-fast scans also presented with substantially less motionartefacts including bowl movements. A county density regularized reconstruction approach is essential to achieve the acceptableimage quality on a consistent basis.

CONCLUSION

Next generation digital photon counting detector technology enabled consisten accepatable image quality even for ultra-fastwholebody imaging with a whole body acquistion time of 2 min. The concept of ultra-fast whole-body acquisition is feasible,however requires count density adaptive, regularized reconstruction, New PET workflow concepts, improved patient comfort,minimized patient motion and whole-body pseudo dynamic imaging of tracer were demonstated as feasible.


Ultra-Fast Wholebody PET/CT with 2 min acquisition time was shown to be feasible using a new generation digital photon countingPET/CT system

ParticipantsLucia Baratto, Stanford, CA (Presenter) Nothing to DiscloseAkira Toriihara, Stanford, CA (Abstract Co-Author) Nothing to DiscloseNegin Hatami, Stanford, CA (Abstract Co-Author) Nothing to DiscloseGuido A. Davidzon, MD, Stanford, CA (Abstract Co-Author) Nothing to DiscloseShyam Srinivas, MD, PhD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseSanjiv S. Gambhir, MD, PhD, Stanford, CA (Abstract Co-Author) Board Member, Enlight Biosciences Board Member, ImaginAb, IncBoard Member, FUJIFILM Holdings Corporation Board Member, ClickDiagnostics, Inc Consultant, FUJIFILM Holdings CorporationConsultant, Gamma Medica, Inc Speaker, ImaginAb, Inc Stock, Enlight Biosciences Stock options, Enlight Biosciences Travelsupport, Gamma Medica, IncAndrei Iagaru, MD, Emerald Hills, CA (Abstract Co-Author) Research Grant, General Electric Company


[email protected]

PURPOSE

To compare the performance of a silicone photomultiplier (SiPM)-based PET/CT scanner (DMI) and a LYSO-based PET/CT scanner(D690) used in our clinic and to determine any differences in semi-quantitative measurements.


We prospectively enrolled patients referred for 68Ga-DOTA-TATE PET/CT. All patients underwent a single 68Ga-DOTA-TATEinjection dual imaging protocol: they were randomly scanned first on the D690 or the DMI scanner. SUVmax of detectable lesionsand SUVmean of different normal tissues were measured independently by two Nuclear Medicine Physicians, from data acquiredusing both scanners.

RESULTS

Data from 61 patients (35-80 year-old; 36 women, 25 men) was analyzed. Thirty-one patients out of 61 (51%) underwent DMI asfirst scan and a total of 95 lesions were detected on both scanners; the average SUVmax measurements for all 95 lesions werehigher on DMI than D690 (28.1 and 25.2, respectively, r = 0.944, p<0.001, 95% CI, 1.382 to 4.375). Thirty patients out of 61(49%) performed D690 PET/CT as first scan and a total of 84 lesions were seen on both scanners; the average SUVmaxmeasurements for all 84 lesions were higher on DMI than D690 (38.3 and 33.8, respectively, r = 0.991, p<0.001, 95% CI, -6.139 to-2.879). Mean lesion:aortic arch ratios were higher on DMI than D690 (47.1 vs 42.7, and 26.6 vs 26.4 when D690 was performedfirst and when DMI was done first, respectively), but differences were not statistically significant (p<0.006 and p<0.877,respectively). The agreement analysis for different background organs indicated that scanners are similar in normal tissues uptakeregardless of scan order.

CONCLUSION

We observed higher SUVmax values for lesions measured from DMI compared to D690 regardless of the order of the scan, while themeasurements were similar for normal tissues. While delayed imaging can lead to higher SUVmax values in cancer lesions, in theseries of lesions identified when DMI was performed as first scan this was not seen; therefore, the data suggests superiorperformance of the DMI scanner.


DMI PET/CT new generation scanner have better performance on standard of care PET/CT scanners. This would allow to reducethe injected dose or the scan time without missing images quality.

SSJ17-01 Nuclear Medicine Keynote Speaker: Radiomics in Lung Cancer


SSJ17-02 An Updated and Validated PET/CT Volumetric Prognostic Index for Non-Small Cell Lung Cancer


SSJ17-03 Prospective Comparison of 18F-FDG PET/MRI and 18F-FDG PET/CT for Thoracic Staging of Non-Small Cell Lung Cancer


SSJ17

Science Session with Keynote: Nuclear Medicine (Chest/Breast Oncology Nuclear Imaging)


BR CH CT MR NM OI


ParticipantsPeter S. Conti, MD, PhD, Los Angeles, CA (Moderator) Nothing to DiscloseAndrew C. Homb, MD, Rochester, MN (Moderator) Nothing to Disclose

Sub-Events

ParticipantsLawrence H. Schwartz, MD, New York, NY (Presenter) Committee member, Celgene Corporation Committee member, Novartis AGCommittee member, ICON plc Committee member, BioClinica, Inc

ParticipantsJoshua H. Finkle, MD, Chicago, IL (Presenter) Nothing to DiscloseBill C. Penney, PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseYonglin Pu, MD, PhD, Chicago, IL (Abstract Co-Author) Nothing to Disclose

PURPOSE

Whole-body metabolic tumor volume (MTVWB) and TNM staging are independent prognostic factors for overall survival (OS) in non-small cell lung cancer (NSCLC). We aimed to update and validate the PET/CT volumetric prognostic index (PVP index) using the new8th edition TNM staging system to evaluate its prognostic power versus TNM staging and MTVWB alone.


This study was a retrospective analysis of 949 non-small cell lung cancer (NSCLC) patients diagnosed between 2004 and 2014.Clinical TNM stage, MTVWB, age and gender, tumor histology type at the initial staging PET/CT exam, as well as treatment historyand long-term survival data were obtained. Patients were randomly assigned to modeling or validation group. Univariate andmultivariate Cox regression analyses were performed to compare PVP index, TNM stage, and MTVWB in the validation group.

RESULTS

The updated PVP index included the 3 variables TNM stage, and MTVWB and age. Univariate Cox models showed significantassociation of PVP index with overall survival (OS) in patients with NSCLC (with Hazard ratio HR= 2.88 in the validation group,p<0.001). The C-statistic of the PVP index (C-statistic = 0.71 in the validation group) was significantly greater than that of 8thedition TNM staging (C-statistic = 0.68, p=0.029 ), MTVWB (C-statistic = 0.68, p=0.001), and patient age (C-statistic = 0.53,p<0.001). Multivariate Cox regression analyses demonstrated significant association of PVP index with OS (with HR= 2.80, p<0.001)after adjusting patient's gender and tumor histology.

CONCLUSION

The updated PVP index provides a quantitative risk assessment for NSCLC patients using 8th edition TNM staging, MTVWB, and age.The index provides a simple and practical way for the care team to incorporate the independent prognostic value of both the TNMstage and MTVWB. This approach can further improve the accuracy of overall survival prognosis.


The PVP index combines the prognostic power of the TNM stage, whole-body metabolic tumor volume and age, offering prognosticaccuracy superior to whole-body metabolic tumor volume or TNM stage alone.

ParticipantsLino Sawicki, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Presenter) Nothing to DiscloseBenedikt M. Schaarschmidt, MD, Essen, Germany (Abstract Co-Author) Stockholder, Bayer AG; Stockholder, General ElectricCompany; Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries LtdKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, Sofie

SSJ17-04 The Relationship Between PET/CT Imaging Features and Pathological Types and Gene Mutations ofPrimary Lung Cancer: A Study of 213 Untreated Lung Cancer Patients with Bone Metastases


Biosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SAGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

To compare the diagnostic performance of 18F-FDG PET/MRI and 18F-FDG PET/CT for primary and locoregional lymph node stagingin non-small cell lung cancer (NSCLC).


In this prospective study a total of 84 patients (51 men, 33 women, mean age 62.5 ± 9.1 years) with histopathologically confirmedNSCLC underwent 18F-FDG PET/CT followed by 18F-FDG PET/MRI in a single injection protocol. Two readers independentlyassessed T and N staging in separate sessions according to the seventh edition of the American Joint Committee on Cancer stagingmanual for 18FFDG PET/CT and 18F-FDG PET/MRI, respectively. Histopathology as reference standard was available for N staging inall 84 patients and for T staging in 39 patients. Differences in staging accuracy were assessed by McNemars chi2 test. Themaximum standardized uptake value (SUVmax) and longitudinal diameters of primary tumors were correlated using Pearson'scoefficients.

RESULTS

T stage was categorized concordantly in 18F-FDG PET/MRI and 18F-FDG PET/CT in 38 of 39 (97.4%) patients. Herein, 18F-FDGPET/CT and 18F-FDG PET/MRI correctly determined the T-stage in 92.3% and 89.7% of patients, respectively. N-stage wascategorized concordantly in 83 of 84 patients (98.8%). 18F-FDG PET/CT correctly determined the N stage in 78 of 84 patients(92.9%), while 18F-FDG PET/MRI correctly determined the N stage in 77 of 84 patients (91.7%). Differences between 18F-FDGPET/CT and 18F-FDG PET/MRI in T and N staging accuracy were not statistically significant (p > 0.5, each). Tumor size andSUVmax measurements derived from both imaging modalities exhibited excellent correlation (r=0.963 and r=0.901, respectively).

CONCLUSION

18F-FDG PET/MRI and 18F-FDG PET/CT showed an equivalently high diagnostic performance for T and N staging in patientssuffering from NSCLC.


PET/MRI as a dose-saving alternative to PET/CT proved coequal to the current gold standard for thoracic staging of NSCLC. Thus,clinicians might use PET/MRI instead of PET/CT for this purpose. However, considering the longer examination times and higherexpenses of PET/MRI, a general recommendation in favor of PET/MRI cannot be drawn from this study.

ParticipantsXiaomeng Li, MD, Beijing, China (Presenter) Nothing to DiscloseNing Wu, MD, Beijing, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the relationship between 18F-FDG PET/CT image characteristics and pathological types and gene mutations of primarylung cancer in untreated lung cancer patients with bone metastases.


A total of 213 untreated lung cancer patients with bone metastases were enrolled in this study. All patients underwent 18F-FDGPET/CT examination, pathological and gene mutation examination of primary lung cancer. Spearman's correlation test was performedto evaluate the association between primary tumors and bone metastases. Single factor analysis of variance was used to comparegroups.

RESULTS

(1)A total of 213 cases were evaluated. The mean SUVmax of primary lung cancer was 7.9±4.7; that of bone metastases was8.2±4.3. The SUVmax of primary lesions had a significantly positive correlation with the SUVmax of bone metastases (r = 0.622; p =0.000). Osteolytic metastasis was the most common type. (2)The SUVmax of primary lung lesions with different pathological typeswere statistically different (all P = 0.000): squamous cell carcinoma > small cell carcinoma > adenocarcinoma. Their SUVmax were11.7±3.7, 9.3±3.1, and 6.7±4.6, respectively. (3)In non-small cell lung cancer (NSCLC), the gene mutation rates of epidermalgrowth factor receptor (EGFR), K-ras and anaplastic lymphoma kinase (ALK) were 35.7%, 10.1% and 3.8%, respectively. There wasno statistical difference in SUVmax of primary lung cancer between gene mutation type and wild type (P>0.05).

CONCLUSION

The SUVmax of primary lung lesions with different pathological types were statistically different. Squamous cell carcinoma was thehighest, and adenocarcinoma was the lowest. The SUVmax of primary lung cancer had a significantly positive correlation with theSUVmax of bone metastases. In NSCLC, the mutation rate of EGFR is the highest. There was no statistical difference in SUVmax ofprimary lung cancer between gene mutation type and wild type.


The SUVmax of primary lung cancer is suggestive of its pathological type. But the SUVmax of primary lung cancer is not helpful topredict the gene mutations in NSCLC.

SSJ17-05 Local and Whole-Body Staging in Patients with Primary Breast Cancer: A Comparison of One-Step toTwo-Step-Staging-Algorithms Utilizing PET/MRI


SSJ17-06 Multimodal Radiomic Imaging: Comparison of PET and MRI-pCM Heterogeneity in Breast Cancer


ParticipantsLale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AGJohannes Grueneisen, Essen, Germany (Abstract Co-Author) Nothing to DiscloseMark Oehmigen, Essen, Germany (Abstract Co-Author) Nothing to DiscloseHarald H. Quick, PhD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseAnn-Kathrin Bittner, Essen, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SAChristian Buchbender, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Presenter) Nothing to Disclose


[email protected]

PURPOSE

To compare the diagnostic value of a one-step to a two-step staging algorithm for local and whole-body staging utilizing 18F-FDGPET/MRI in breast cancer patients.


A total of 38 patients (37 females and one male, mean age 57 ± 10 years; range 31-78 years) with newly diagnosed,histopathologically proven breast cancer were prospectively enrolled in this trial. All PET/MRI examinations were assessed for localtumor burden and metastatic spread in two separate reading sessions: (1) One-step algorithm comprising supine whole-body 18F-FDG PET/MRI, (2) Two-step algorithm comprising a dedicated prone 18F-FDG breast PET/MRI and supine whole-body 18F-FDGPET/MRI.

RESULTS

On a patient based analysis the two-step algorithm correctly identified 37 out of 38 patients with breast carcinoma (97%), while 5patients were missed by the one-step 18F-FDG PET/MRI algorithm (33/38; 87% correct identification; p=0.37). On a lesion-basedanalysis 56 breast cancer lesions were detected in the two-step algorithm and 44 breast cancer lesions could be correctlyidentified in the one-step 18F-FDG PET/MRI (79%), resulting in statistically significant differences between the two algorithms(p=0.0015). For axillary lymph node evaluation sensitivity, specificity and accuracy was 93%, 95 % and 94%, respectively.Furthermore, distant metastases could be detected in 7 patients with both modalities.

CONCLUSION

The results demonstrate the necessity and superiority of a two-step 18F-FDG PET/MRI algorithm, comprising dedicated pronebreast imaging and supine whole-body imaging, when compared to the one-step algorithm for local and whole-body staging inbreast cancer patients.


Two-step 18F-FDG PET/MRI comprising dedicated breast and whole-body imaging enables high-quality local and whole-body stagingin patients with breast cancer.

ParticipantsBert-Ram Sah, MD, London, United Kingdom (Presenter) Nothing to DiscloseMarta Bogowicz, Zurich, Switzerland (Abstract Co-Author) Nothing to DiscloseChristian Leissing, Zurich, Switzerland (Abstract Co-Author) Nothing to DiscloseStephanie Tanadini-Lang, Zurich, Switzerland (Abstract Co-Author) Nothing to DisclosePatrick Veit-Haibach, MD, Zurich, Switzerland (Abstract Co-Author) Research Grant, Bayer AG Resaarch Grant, F. Hoffmann-LaRoche Ltd Research Grant, General Electric Company

PURPOSE

This study investigated the value of pre-treatment F-18-Fluorodeoxyglucose (FDG)-positron-emission-tomography (PET) radiomicsin comparison to T1-weighted-post-contrast-magnetic-resonance-imaging (MRI-pCM) radiomics in patients with breast cancer.


Following IRB approval and informed consent, a total of 30 patients with histologically proven breast cancer were prospectivelyrecruited. Patients were injected 225+/-55 MBq FDG intravenously. PET and MRI-pCM were acquired on the same machine. 154radiomic features of first, second, and higher order were extracted from the primary tumor. Dimensionality of features was reducedwith a Principal Component Analysis. The relationship of selected features to staging and histological parameters was determined.Association of features between the different modalities was compared (Spearman "ρ").

RESULTS

Selected radiomic features of PET showed moderate correlation to T-stage (-0.52 < ρ < 0.54) and weak correlation to N-Stage (-0.35 < ρ < 0.38). Selected radiomic features of MRI-pCM showed moderate correlation to T-stage (-0.64 < ρ < 0.57) and to N-stage (-0.52 < ρ < 0.54). Correlation of radiomic features of both modalities to hormone receptor status is shown in Table 1.Comparison between PET and MRI-pCM showed moderate to strong correlation for the comparison of all radiomic features (-0.66 <

ρ < 0.68) (Figure 1), whereas the correlation for the comparison of a respective radiomic parameter was only weak to moderate(0.22 < ρ < 0.56) (1st diagonal in Figure 1).

CONCLUSION

Radiomics in a multimodality approach might be a complementary tool for non-invasive pre-therapeutic characterization of breastcancer.


Combining radiomic features from different imaging modalities may help in non-invasive specification of breast cancer.

SSJ22-01 Modified Model Based Iterative Reconstruction Method to Improve CT Number Accuracy in Low-DoseCT

Tuesday, Nov. 27 3:00PM - 3:10PM Room: N227B

SSJ22-02 Joint Reconstruction of Low-Count PET and Undersampled MR in PET/MR Using Deep Learning


SSJ22

Physics (Image Reconstruction)


AI CT MR NM PH



ParticipantsXiaochuan Pan, PhD, Chicago, IL (Moderator) Research Grant, Bondent Imaging; Research Grant, Canon Medical SystemsCorporation; Stockholder, Clarix ImagingXiangyang Tang, PhD, Atlanta, GA (Moderator) Research Grant, SINOVISION Technology Co, Ltd

Sub-Events

ParticipantsJohn W. Hayes, MS, Madison, WI (Presenter) Nothing to DiscloseRan Zhang, PhD, Madison, WI (Abstract Co-Author) Nothing to DiscloseDaniel Gomez-Cardona, PhD, Rochester , MN (Abstract Co-Author) Nothing to DiscloseJuan Pablo Cruz Bastida, Madison, WI (Abstract Co-Author) Nothing to DiscloseGuang-Hong Chen, PhD, Madison, WI (Abstract Co-Author) Research funded, General Electric Company Research funded, SiemensAG


[email protected]

PURPOSE

Recent theoretical and experimental studies about CT number accuracy in low dose CT have shown that filtered backprojection(FBP) and conventional model based iterative reconstruction (MBIR) methods are biased by an amount that is dose and contrastdependent. This work validates that the data weighting scheme in the MBIR framework is the culprit for both these biases and thata more optimal weighting scheme has been found to eliminate bias across all dose and contrast levels.


Raw CT data was acquired for two phantoms on a benchtop CT system using a photon counting detector (XC-HYDRA FX50,XCounter AB, Sweden). The Catphan phantom (Catphan 600, Phantom Laboratory, Salem, New York) was scanned at several doselevels in the range 69-367 mAs, with 50 repeated scans for each dose level. A customized head phantom was also scanned in therange 40-400 mAs, with 30 repeated scans for each dose level. Reference images for each phantom were obtained by averagingthe pre-log projections for the highest dose level across all repetitions, and then performing FBP reconstruction. These referencesserved as the experimental ground truth. For each dose level and reconstruction method (FBP, MBIR, proposed MBIR), bias imageswere calculated by subtracting the reference image from the mean of the reconstructions of each repeated scan. Bias wasmeasured in 4 small inserts of varying contrast in the Catphan phantom. Bias images of the Catphan and head phantoms were alsoassessed across the image field of view (FOV).

RESULTS

There are three main results: 1) The theoretical relationship bias=±a/mAs*(1+ßΔHU) was validated experimentally for both FBP(positive polarity) and conventional MBIR (negative polarity). 2) The proposed MBIR method, which uses a modified weightingscheme, eliminates bias for each contrast and dose level in the Catphan phantom. 3) The proposed MBIR method demonstratespromising preliminary results for reducing bias across the FOV in a more complex anthropomorphic head phantom.

CONCLUSION

The proposed MBIR method maintains CT number accuracy of varying contrast across dose levels by using a theoretically basedmodified data weighting scheme.


Certain tasks, e.g. detection of acute cerebral venous sinus thrombosis (CVST), rely on CT number estimation. It is critical the CTreconstruction method maintains accurate HU values.

Participants

SSJ22-03 Implementation of a CT Reference Library Containing Manufacturer-Neutral Projection Data, Images,and Clinical Metadata


Junshen Xu, Beijing, China (Presenter) Nothing to DiscloseTuoyu Cao, PhD, Houston, TX (Abstract Co-Author) Employee, Medical Device ManufactorZheng Zhang, MD, Shanghai, China (Abstract Co-Author) Employee, Healthcare Device ManufactorLingzhi Hu, PhD, Houston, TX (Abstract Co-Author) Employee, UIH America, IncNan-Jie Gong, Houston, TX (Abstract Co-Author) Employee, UIH America, IncHongcheng Shi, Shanghai, China (Abstract Co-Author) Nothing to DiscloseKui Ying, PhD, Beijing, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

Acquiring low-count PET and undersampled MR can shorten PET/MR scan time, which, however, may also lead to noisy PET imagesand MR images with artifacts. The goal of this report is to evaluate whether deep learning method can reconstruct high-qualityPET/MR images from its low-quality counterpart, potentially enabling shorter scan time in PET/MR. We also compared the proposedmodel with single modality models to investigate whether the resulting image quality can benefit from sharing features of the twomodalities in the network.


We developed a fully convolutional encoder-decoder network to predict high quality PET and MR images from low-count PET andundersampled MR. Concatenate skip connections and strategy of residual learning is adopted to restore high resolution details. BrainPET/MR data are acquired with a simultaneous PET/MR system (uPMR790, United Imaging Healthcare) from 50 patients whoreceived 0.12 mCi/kg FDG. To generate low-count PET, the PET list-mode data was randomly undersampled for 10% events. Bothstandard-count and low-count PET images were reconstructed with OSEM (4 iterations, 20 subsets). Undersampled T1 weight MR isgenerated using radial sampling in k-space with sampling rate equal to 10%. Standard-count PET and fully sampled MR were takenas ground-truth in network training.

RESULTS

Models were trained on 40 patients and evaluated on the other 10 patients. The proposed joint model gains 4.5/7.9dB in peaksignal-to-noise ratio (PSNR) and 0.036/0.42 in structural similarity index (SSIM) compared with low-count PET/undersampled MR.When compared with the single modality model of PET/MR, results shows that our joint model has an improvement of 0.97/0.15dB inPSNR and 0.006/0.0012 in SSIM.

CONCLUSION

Using a deep learning algorithm, we can estimate high-quality PET and MR images from low-count PET and undersampled MRimages. Results also showed that joint reconstruction of PET and MR by sharing features in network can improve image quality oftwo modalities compared with single modality model.


This method was demonstrated promising in greatly reducing the scan time in PET/MR imaging by up to 90%.

ParticipantsTaylor Moen, Rochester, MN (Presenter) Nothing to DiscloseJayse Weaver, Rochester, MN (Abstract Co-Author) Nothing to DiscloseCynthia H. McCollough, PhD, Rochester, MN (Abstract Co-Author) Research Grant, Siemens AGPhillip Edwards, Rochester, MN (Abstract Co-Author) Nothing to DiscloseDavid R. Holmes Iii, PhD, Rochester, MN (Abstract Co-Author) Nothing to DiscloseLifeng Yu, PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseBaiyu Chen, PhD, New York, NY (Abstract Co-Author) Nothing to DiscloseJoel G. Fletcher, MD, Rochester, MN (Abstract Co-Author) Grant, Siemens AG; Consultant, Medtronic plc; ;


[email protected]

PURPOSE

A manufacturer-neutral CT projection data (PD) format (DICOM-CT-PD) has been previously developed and used to allow access toCT PD and the scanner information required for image reconstruction. Access to such data was not previously possible, limiting theability of reconstruction scientists to work with patient data. In this work, we aim to construct a reference DICOM-CT-PD librarycontaining patient PD with corresponding images and clinically relevant metadata, and to publish this library for public access.


CT images and PD were acquired from three different manufacturers for three clinical scanners at routine dose levels for head,chest and abdomen exams. The PD were converted to the DICOM-CT-PD format and a lower dose exam was simulated for each PDset using a validated noise-insertion method. Radiologists reviewed each case and marked lesion locations and diagnosis. Referencetruth was obtained from the patient medical record, either from histology or subsequent imaging. Metadata such as lesion location,diagnosis, and source of truth were acquired for each case and formatted into a reference report. Each case was anonymized toremove protected health information for transfer to an NCI-hosted public data archive, The Cancer Imaging Archive (TCIA).

RESULTS

450 total cases from Siemens (n=150), GE (n=150), and Philips (n=150) scanners were obtained, including both negative andpositive patient cases. PD are available for two dose levels, routine full dose and simulated low dose (25% of routine dose for headand abdomen cases and 10% of routine dose for chest cases). Routine dose image series are available for all of the cases, andreduced dose images are additionally available for exams acquired on a Siemens scanner. Clinical metadata are organized in an easyto use spreadsheet. The assembled projection, image and clinical data provide a rich data library with which CT imagereconstruction scientists can validate their algorithms.

SSJ22-04 Motion Compensation in Liver SPECT using Simultaneous X-Ray and Nuclear Imaging


SSJ22-05 Motion Compensated Reconstruction of the Aortic Valve for Non-Gated Helical CT Scans


CONCLUSION

A large patient library containing manufacturer-neutral PD, the corresponding full dose images, and clinical reference information hasbeen developed and is being made available through the TCIA.


The successful implementation of this library will provide open source CT PD with correlated images and clinical information toinvestigators for reconstruction research and development.

ParticipantsMartijn Dietze, Utrecht, Netherlands (Presenter) Nothing to DiscloseRemco Bastiaannet, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseBritt Kunnen, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseSandra v. Velden, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseMarnix G. Lam, MD, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseMax A. Viergever, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseHugo W. de Jong, PhD, Utrecht, Netherlands (Abstract Co-Author) Nothing to Disclose

PURPOSE

Quantitative accuracy of liver SPECT/CT is crucial for e.g. dosimetry in radioembolization, but due to respiratory motion limited inprecision. Motion can be compensated for in the reconstruction, but in clinical practice this requires an external device for thetracking of the motion signal and a prior motion vector field estimate to link the motion signal to organ movements, complicating theacquisition. A device under development, which simultaneously measures x-ray and nuclear projections, could be used to retrieveboth measures intrinsically. Such a data-driven approach eliminates the need for external devices and provides a real-time vectorfield. The purpose of this work is to evaluate the performance of the proposed motion compensation technique using simulations.


Nuclear and x-ray projections of a realistic digital phantom with respiratory motion were generated using Monte Carlo simulations forseveral breathing patterns. X-ray projections were sampled at 1 to 5 Hz; nuclear projections were acquired continuously. Total x-ray imaging dose was varied from 1 to 1000 µGy. The motion signal was extracted from x-ray projections by calculation of thecenter of mass and then used to bin the projections into gates. The x-ray gates were individually reconstructed and registeredonto each other, resulting in the vector field to be included in the nuclear reconstruction.

RESULTS

The respiratory motion signal was accurately extracted from the x-ray projections, provided the x-ray sampling rate was greaterthan 2 Hz and the motion was stable in amplitude. The total minimally required dose for x-ray sampling was 10 µGy for a 5 minutescan. The inclusion of motion correction into the SPECT reconstruction improved contrast-to-noise ratio, in comparison with nomotion correction, from 11.9 ± 0.5 to 19.1 ± 0.7.

CONCLUSION

The proposed motion compensation technique has the potential to improve quantitative SPECT reconstructions. Additionally, theneed for external devices and a prior vector field estimate are eliminated. Only a limited amount of dose is required to obtainsignificantly improved results, paving the way for clinical use.


Liver radioembolization requires quantitative SPECT to study the activity distribution. In order to improve accuracy and personalizedosimetry, motion should be accounted for in reconstructions.

ParticipantsClemens Spink, Hamburg, Germany (Presenter) Nothing to DiscloseTanja Elss, Hamburg, Germany (Abstract Co-Author) Doctorate student, Koninklijke Philips NVRolf Bippus, Hamburg, Germany (Abstract Co-Author) Employee, Koninklijke Philips NVMichael Morlock, PhD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseGerhard B. Adam, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseGunnar K. Lund, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseMichael Grass, PhD, Hamburg, Germany (Abstract Co-Author) Employee, Koninklijke Philips NV

PURPOSE

Precise CT imaging is prerequisite for reliable planning of transcatheter aortic valve implantation (TAVI). Especially in non-gated CTscans, cardiac motion leads to severe artifacts in the reconstructed CT images. Blurring of the valve and the neighboring vascularanatomy potentially result in incorrect device sizing. A second pass motion correction method for non-gated helical CT scans with apitch <1 is introduced here.


The new post-processing method was applied to five non-gated clinical datasets acquired with a 256-slice CT scanner (BrillianceiCT, Philips Healthcare). Redundancy in the helical projection data was used to generate three image volumes at identical spatialpositions, but different time points. During each reconstruction a subset of detector rows was selected which may be eitheroverlapping or fully separated depending on the pitch size. The 3D edge-filtering scheme included Gaussian smoothing for noisereduction, gradient calculation for edge enhancement, non-maximum-suppression and hysteresis thresholding for reduction ofincoherent edges. The sparse filter results were taken as input for an elastic registration to estimate the displacement of each

SSJ22-06 Multi-Channel GAN: A Machine Learning Approach to Parallel MRI Reconstruction


voxel between the given time points. Reconstructed datasets were evaluated with a TAVI planning software (IntelliSpace Portal,Philips Healthcare) by two blinded readers.

RESULTS

The method achieved significant motion artifact reduction in CT aortic valve reconstructions. A removal of doubled structures atthe aortic boundaries could be observed, as well as reduced blurring compared to the uncompensated reconstructions.

CONCLUSION

Motion compensated reconstruction is feasible for non-gated helical CT scans using edge filtering and image based registration formotion estimation. Reconstructed CT image datasets may improve planning and device selection for TAVI procedures.


Motion compensated reconstruction yields reduced artifact levels at the aortic valve in non-gated helical CT scans with a pitch <1.

ParticipantsPengyue Zhang, Stony Brook, NY (Presenter) Nothing to DiscloseFusheng Wang, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseYulee Li, Greenvale, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Magnetic resonance imaging (MRI) has a low imaging speed. MRI acceleration relies on undersampling that may introduce aliasingartifacts in image reconstruction. Here we propose a machine learning approach that can automatically learn parallel MRImechanisms underlying multi-channel k-space data and reconstruct high-quality MR images from undersampled data.


Parallel MRI is a standard approach to imaging acceleration on clinical MRI scanners. This approach can effectively suppress aliasingartifacts associated with undersampling, but requires an additional calibration procedure that limits the overall imaging speed. Herea deep learning based neural network model, multi-channel generative adversarial network (multi-channel GAN), is developed toprocess multi-channel raw MRI data. This model can learn parallel MRI reconstruction mechanisms underlying a large amount ofmulti-channel k-space data. The trained model can be used to reconstruct images from undersampled data without calibration,thereby providing a higher imaging speed than conventional parallel MRI. In our approach, the basic unit of multi-channel GAN hastwo sub-networks: a generator network which learns the relationship between undersampled and fully-sampled data and adiscriminator network which justifies if the generated data are real. The whole model consists of the same number of basic unitnetworks as that of radiofrequency channels on the MRI scanner for parallel MRI reconstruction. The training process uses astochastic gradient descent and back-propagation algorithm. The trained multi-channel generator network is used to perform imagereconstruction.

RESULTS

We evaluate the proposed method with a total of 170 sets of 2D multi-channel brain MRI images. Figure 1 shows an example ofreconstruction results with an undersampling factor of 5. It is found that the machine learning method outperforms other state-of-the-art parallel MRI reconstruction methods.

CONCLUSION

We demonstrate a machine learning approach to parallel MRI reconstruction.This approach can generate high-quality images fromundersampled data without calibration, providing a higher imaging speed than conventional parallel MRI.


The machine learning approach to parallel MRI reconstruction can enhance diagnostic MRI quality, shorten clinical MRI proceduresand improve clinical MRI throughput.

MSCC34A Somatostatin Receptor PET & Therapy

MSCC34B Fluciclovine/PSMA PET

MSCC34C Response Assessment

MSCC34

Case-based Review of Nuclear Medicine: PET/CT Workshop-Advances in PET (In Conjunction with SNMMI)(Interactive Session)

Tuesday, Nov. 27 3:30PM - 5:00PM Room: E450B

BQ CT NM



ParticipantsSamuel E. Almodovar-Reteguis, MD, Orlando, FL (Director) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Director) Nothing to DiscloseChadwick L. Wright, MD,PhD, Lewis Center, OH (Moderator) Nothing to Disclose

Sub-Events

ParticipantsThomas A. Hope, MD, San Francisco, CA (Presenter) Research support, General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Define the role of somatostatin receptor (SSTR) PET in patients with neuroendocrine tumors (NETs). 2) Compare the use ofconventional imaging and SSTR PET in staging NETs. 3) Explain the mechanism of SSTR based peptide receptor radionuclide therapy(PRRT).

ABSTRACT

Neuroendocrine tumors (NET) are unique in that they overexposes the somatostatin receptor (SSTR). This can be leveraged inimaging by labelling somatostatin analogs with radiation to image the location of tumors. DOTATATE is a SSTR analog, that whenlabeled with Gallium-68 can be used to image neuroendocrine tumors with very high sensitivity and specificity. It is important toremember that although SSTR PET using Ga68 DOTATATE is very effective, conventional imaging using either CT or MRI will remainthe most common imaging modality for NET patients over time. Beyond imaging, SSTR analogs can be labeled with beta emittersthan can be used therapeutically. Most commonly DOTATATE is labeled with Lutetium-177. This was studied prospectively in arandomized controlled trial (NETTER-1 trial), which demonstrated significant improvement in radiographic progression free survival.These results led to the FDA approval of this therapy in 2018.

ParticipantsAndrei Iagaru, MD, Emerald Hills, CA (Presenter) Research Grant, General Electric Company

LEARNING OBJECTIVES

1) List some of the molecular imaging targets that are used in prostate cancer. 2) Understand underlying biology and mechanism ofaction for some of the new PET radiopharmaceuticals in prostate cancer. 3) Discuss patterns of prostate cancer appearance whenusing some of the new PET radiopharmaceuticals.

ABSTRACT

Data from the American Cancer Society suggests that prostate cancer will continue to be the leading cancer diagnosis in men with164,690 estimated new cases and will have the second highest mortality (after lung cancer) with 29,430 estimated deaths for 2018in the United States. Initial and subsequent treatment of prostate cancer may involve surgery, radiation therapy, hormonal therapy,chemotherapy, or a combination of these. Additional molecular pathways in prostate cancer lead to the identification of newtargets that may be amenable to diagnostic and therapeutic intervention with novel agents. Areas of interest for the NuclearMedicine and Molecular Imaging community include mainly aminoacid analogues (Fluciclovine) and the prostate specific membraneantigen (PSMA), but also gastrin releasing peptide receptors (GRPR).

ParticipantsDavid A. Mankoff, MD, PhD, Philadelphia, PA (Presenter) Speaker, Koninklijke Philips NV; Consultant, General Electric Company;Advisory Board, RefleXion Medical Inc; Consultant, Blue Earth Diagnostics Ltd; Research Funded, Siemens AG; Advisory Board,ImaginAb, Inc; Spouse, Owner, Trevarx


[email protected]

LEARNING OBJECTIVES

1) List applications of molecular imaging as a cancer biomarker. 2) Describe clinical setting for which molecular imaging responseapproaches are applicable. 3) Discuss investigational agents being investigated for response assessment and early results.

ABSTRACT

This talk will review molecular imaigng approaches for cancer, considering molecular imaging as a cancer biomarker to guidetreatment decisions and evaluate therpaeutic repsonse. Examples form recent or ongoing multi-center trials will be presented asexmaples of possible future clinical role for molecular imaigng cancer biomarkers.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ David A. Mankoff, MD, PhD - 2013 Honored EducatorDavid A. Mankoff, MD, PhD -2018 Honored Educator

RC411A Dopamine Transporter Scans and Movement Disorders

RC411B Imaging for Epilepsy

RC411

PET/CT and SPECT/CT in Movement Disorders, Epilepsy, and Dementia


NR CT NM



Sub-Events

ParticipantsVani Vijayakumar, MD, Ridgeland, MS (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Apply basic knowledge and skills relevant to clinical practice of Movement Disorders. 2) Assess the potential of emergingtechnological innovations and advances to enhance clinical practice and problem-solving. 3) Develop new ideas from experts andpeers in the nuclear imaging sciences. 4) Differentiate Essential Tremor and Presynaptic Parkinson Diseases on DATscans. 5)Compare different image findings for interpretation of Movement Disorders.

ABSTRACT

Introduction: Parkinson Disease ( PD) is the most common movement disorder affecting 1-2 % of the general population over theage of 65 years and the second most common neurodegenerative disorder after Alzheimer´s disease (AD) PD presents with 3 mostcommon symptoms. 1. Resting tremor: Most common first symptom, usually asymmetric and most evident in one hand with the armat rest. 2.Bradykinesia: Difficulty with daily activities such as writing, shaving, using a knife and fork, and opening buttons;decreased blinking, masked facies, slowed chewing and swallowing. 3.Rigidity: Muscle tone increased in both flexor and extensormuscles providing a constant resistance to passive movements of the joints; stooped posture, anteroflexed head, and flexed kneesand elbows. Nuclear Imaging Diagnosis: Datscan: (123I-ioflupane) Patient preparation: Thyroid blockade with Lugols- 3 drops onehour before Stop medicines that bind to the dopamine transporter 7 days prior to study, e.g. SSRIs, amphetamine, benzotropine,cocaine, mazindol,methylphenidate and phentermine and sertraline Radiopharmaceutical: (123I-ioflupane) is a molecular imagingagent 3-5 mCi IV and Brain SPECT in 3 hours Used to demonstrate the location and concentration of dopamine transporters (DaTs)in the synapses of striatal dopaminergic neurons. Interpretation: Normal: comma shaped striatum Abnormal: dot, asymmetriccaudate or putamen, high background Summary: A highly sensitive marker for accurate assessment of striatal dopaminergic functionto differentiate EssentialTremor from PD Early diagnosis of presynaptic Parkinsonian syndromes Differentiation of presynapticParkinsonian syndromes from parkinsonism without presynaptic dopaminergic loss, such as drug-induced parkinsonism orpsychogenic parkinsonism A straightforward one-day protocol An objective adjunct to the differentiation of PD syndromes from ETin clinically uncertain patients A diagnostic tool helping differentiate between probable DLB and AD Visualizing DaT distribution isuseful as a novel diagnostic adjunct in movement disorders and dementia

ParticipantsAnson L. Thaggard, MD, Jackson, MS (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Define the components of a multidisciplinary evaluation for the surgical treatment of epilepsy. 2) Compare brain SPECT with FDGPET for evaluation of an epileptogenic focus. 3) Discuss barriers to the use of ictal SPECT imaging and functional MRI. 4) Appraisethe added value of fusion imaging in epilepsy evaluation.

ABSTRACT

Medically refractory epilepsy is now often treated surgically. A holistic multidisciplinary review of the patient preoperatively helps tooptimize outcome. FDG PET and perfusion SPECT imaging are an integral part of the evaluation. Both imaging techniques arereviewed in context of the multidisciplinary evaluation. Imaging findings, pearls, and pitfalls of each are reviewed using caseexamples.

URL

http://abstract.rsna.org/uploads/2018/16002150/RC411B%20Imaging%20for%20Epilepsy%20Thaggard.pdf

Active Handout:Anson Lee Thaggard

http://abstract.rsna.org/uploads/2018/16002150/RC411B Imaging for Epilepsy Thaggard.pdf

https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fabstract.rsna.org%2Fuploads%2F2018%2F16002150%2FRC411B%2520Imaging%2520for%2520Epilepsy%2520Thaggard.pdf&data=02%7C01%7Ccricchio%40rsna.org%7C16dd138594cd456081aa08d654933eb3%7Cfb5fefcd7ca642f3b443917c2a68cecd%7C1%7C1%7C636789388838885141&sdata=V1Fs7NrZ85Ss8bCwn9496aXp57PJZdCu%2BkCmV6BrXJE%3D&reserved=0

RC411C PET Imaging for Dementia

http://abstract.rsna.org/uploads/2018/16002150/RC411B Imaging for Epilepsy Thaggard.pdf

ParticipantsPhillip Kuo, MD,PhD, Tucson, AZ (Presenter) Author, MD Training at Home; Research Grant, Astellas Group; Consultant, Endocyte,Inc; Consultant, General Electric Company; Education Grant, General Electric Company; Speakers Bureau, Eli Lilly and Company;Consultant, inviCRO, LLC; Consultant, Imaging Endpoints; Consultant, Progenics Pharmaceuticals, Inc

LEARNING OBJECTIVES

1) Understand the basic pathophysiology of Alzheimer's dementia. 2) Distinguish the different roles of PET imaging with FDG,amyloid, and tau tracers for evaluating dementia.

ABSTRACT

Alzheimer's disease is the most common form of dementia affecting the aging population, and is currently the 6th leading cause ofdeath. Clinical diagnosis is difficult, and there is currently no cure. Functional imaging biomarkers may detect early stages ofdisease prior to the onset of symptoms, and may improve diagnostic accuracy. This in turn may improve evaluation of therapeuticinterventions and provide a roadmap toward developing a cure.

RC420A The Role of Molecular/Functional Imaging for Radiotherapy in Lymphoma

RC420B The Role of Molecular/Functional Imaging for Radiotherapy in Pediatric Cancers

RC420C The Role of Molecular/Functional Imaging for Radiotherapy in Head and Neck Cancer

RC420D The Role of Molecular/Functional Imaging for Radiotherapy in CNS Tumors

RC420

The Role of Molecular and Functional Imaging in Radiation Oncology

Tuesday, Nov. 27 4:30PM - 6:00PM Room: S403B

HN MI NR NM PD RO


ParticipantsNina A. Mayr, MD, Seattle, WA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsStephanie A. Terezakis, MD, Baltimore, MD (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Understand the significance of molecular/functional imaging in guiding the management of both Hodgkin's and non-Hodgkin'slymphomas. 2) Interpret functional imaging as it relates to treatment response and radiation planning. 3) Determine how toIncorporate PET and CT imaging in delineating radiation treatment volumes utilizing ISRT principles.

ParticipantsRalph P. Ermoian, MD, Seattle, WA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) List 3 non-central nervous system pediatric diseases in which functional imaging is standard of care for staging. 2) Describe howfunctional imaging plays a role in assessing response to therapy in two non-central nervous system disease. 3) List two emerginguses for functional imaging in pediatric tumor treatment and response assessment.

ParticipantsMinh T. Truong, MBBS, Boston, MA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Role of Molecular/Functional Imaging in the diagnosis and staging of Head and Neck Cancer (HNC). 2) IntegratingMolecular/Functional Imaging into Radiotherapy Simulation and Planning. 3) Interpretation of Treatment Response toChemoradiotherapy. 4) Molecular/Functional Imaging as a Biomarker for Patient Quality of Life and Survival.

ParticipantsAnca L. Grosu, MD, Freiburg, Germany (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Amino-acids PET (AA-PET) for tumor detection and differentiation between tumor and treatment-related changes in braingliomas. 2) Comparison between AA-PET and mpMRI for radiation treatment planning in brain tumors. 3) New concepts for targetvolume delineation in brain tumors.

ED010-WE

Nuclear Medicine Wednesday Case of the Day

Wednesday, Nov. 28 7:00AM - 11:59PM Room: Case of Day, Learning Center



TEACHING POINTS


RC511A Oropharyngeal Cancer: Evolving Challenges-Clinician's Perspective

RC511B CT and MRI Anatomy and Interpretation

RC511C FDG-PET/CT: Applications and Interpretation

RC511D Panel Discussion: Q&A

RC511

Head and Neck PET/CT: Clinical Approach

Wednesday, Nov. 28 8:30AM - 10:00AM Room: S504CD

CT HN MR NR NM


Sub-Events

ParticipantsBhishamjit Chera, MD, Chapel Hill, NC (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) To understand how radiological interpretation of pre-treatment and post-treatment imaging studies influences the managementof patients with head and neck cancer.

ParticipantsValerie L. Jewells, DO, Chapel Hill, NC (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Provide radiologists with the tools to access CT and MRI imaging for head and neck cancer. 2) Teach attendees how to addressthe images in a manner that will assist the ENT surgeon for staging and surgical planning. 3) Address the principles for criticalthinking and analysis as well as preparation and skill development for a head and neck tumor board.

ABSTRACT

A successful multidisciplinary head and neck tumor board requires coordination and imaging review on the part of radiology to assistthe surgeon, radiation oncolcogist and medical oncologist. The goal is to reach the best option for each individual patient dependingupon tumor type, staging and underlying medical conditions. Appropriate imaging and interpretation is key to this endeavor. Thesetopics will be addressed through discussion of selective CT and MRI cases from our weekly tumor board. References: 1. HeinemanT, St John MA, Wein RO and Weber RS. It takes a village: The importance of multidisicplinary care. Otoloaryngol Clin North Am 2017Aug;50(4):679-687. 2. Liao CT, Kang CJ, Lee LY et al. Association between multidisciplinary team care approach and survival ratesin patients with oral cavity squamous cell carcinoma. Head Neck 2016 Apr;38 Suppl 1:E5444-53. 3. Shah BA, Qureshi MM, Jalisi etal. Analysis of decision making at a multidisciplinary head and neck tumor board incorporating evidence-based National CancerComprehensive Network (NCCN) guidelines. Pract Radiat Oncol 2016 Jul-Aug;6(4):248-54.

ParticipantsTerence Z. Wong, MD, PhD, Chapel Hill, NC (Presenter) Consultant, Lucerno Dynamics, LLC;

LEARNING OBJECTIVES

1) Describe applications for FDG-PET/CT for initial evaluation and follow up of patients with head and neck cancer. 2) Learn thevalue of combining metabolic findings on FDG-PET findings with morphology on CT and endoscopic appearance. 3) Understandpotential etiologies of false positive and false negative studies.

ABSTRACT

Optimal evaluation of patients with head and neck malignancies requires a multidisciplinary approach. Correlation of FDG-PET, CT,direct visualization, and clinical examination is important to provide the best management of these patients.

ParticipantsBhishamjit Chera, MD, Chapel Hill, NC (Presenter) Nothing to DiscloseValerie L. Jewells, DO, Chapel Hill, NC (Presenter) Nothing to DiscloseTerence Z. Wong, MD, PhD, Chapel Hill, NC (Presenter) Consultant, Lucerno Dynamics, LLC;

LEARNING OBJECTIVES

Case examples which highlight the value of multidisciplinary approaches for managing patients with head and neck cancer.

RC517A PET/MRI: Update 2018

RC517B PET/MRI Update 2018: Clinical Practice Implementation - Pearls

RC517C PET/MRI Update 2018: Clinical Applications - Brain and Head and Neck

RC517D PET/MRI Update 2018: Clinical Applications - Body

RC517

Emerging Technology: PET/MRI Update 2018

Wednesday, Nov. 28 8:30AM - 10:00AM Room: S505AB

MR NM



ParticipantsRathan M. Subramaniam, MD,PhD, Dallas, TX (Moderator) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue Earth DiagnosticsLtd


[email protected]

LEARNING OBJECTIVES

1) To discuss opportunities of PET/MRI in clinical practice and research. 2) To discuss challenges of PET/MRI in clinical practice andresearch.

Sub-Events



[email protected]

LEARNING OBJECTIVES

1) To discuss the status of PET/MRI in clinical practice in 2018 and the opportunities and challenges in implementation.

ParticipantsGeoffrey B. Johnson, MD,PhD, Rochester, MN (Presenter) Research Grant, General Electric Company; Research Grant, Pfizer Inc

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Geoffrey B. Johnson, MD,PhD - 2015 Honored EducatorGeoffrey B. Johnson,MD,PhD - 2017 Honored Educator

ParticipantsAlexander Drzezga, MD, Cologne, Germany (Presenter) Consultant, Siemens AG; Consultant, Bayer AG; Consultant, General ElectricCompany; Consultant, Eli Lilly and Company; Consultant, The Piramal Group; Speakers Bureau, Siemens AG; Speakers Bureau, BayerAG; Speakers Bureau, General Electric Company; Speakers Bureau, Eli Lilly and Company; Speakers Bureau, The Piramal Group

LEARNING OBJECTIVES

1) Review relevant clinical applications for PET/MR in the diagnostic work-up of disorders of the brain. 2) Review strengths ofPET/MR for disorders of the head and neck. 3) Understand the value of different currently available tracers for neuroimaging andoncological applications. 4) Review challenges and limitations of PET/MR in brain/head&neck and expected future developments.

ParticipantsThomas A. Hope, MD, San Francisco, CA (Presenter) Research support, General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Review common current applications for abdominopelvic oncologic PET/MRI, including hepatic malignancies, rectal cancer, and

RC517E PET/MRI Update 2018: Clinical Applications - Cardiac

RC517F PET/MRI Update 2018: Physics

1) Review common current applications for abdominopelvic oncologic PET/MRI, including hepatic malignancies, rectal cancer, andcervical cancer. 2) Understand the role of novel tracers in prostate cancer (PSMA PET) and neuroendocrine tumors (somatostatinreceptor PET). The presentation will focus on prostate cancer as an application. 3) Present the current limitations and futureadvances in PET/MRI that will help increase the clinical acceptance and applicability of body PET/MRI.

ParticipantsPamela K. Woodard, MD, Saint Louis, MO (Presenter) Research agreement, Siemens AG; Research, Eli Lilly and Company; Research,F. Hoffmann-La Roche Ltd; ; ; ; ; ;


[email protected]

LEARNING OBJECTIVES

Discuss clinical cardiac PET/MR imaging applications; applications will include myocardial perfusion and viability, nonischemiccardiomyopathy, and tumor assessment.

ParticipantsGeorges El Fakhri, PhD, Boston, MA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Understand the challenges and opportunities afforded by simultaneous PET/MR. 2) Understand the role of PET/MR in imagingmyocardial membrane potential.

RC518A PET Tracers: Which Ones Will Be Next to Make it to Clinical Practice?

RC518B PET/MRI: The Added Value in Oncology

RC518C Hyperpolarized MRI: Current and Future Applications

RC518

Metabolic Tumor Imaging: Current and Beyond

Wednesday, Nov. 28 8:30AM - 10:00AM Room: N229

MR NM OI



ParticipantsEvis Sala, MD, PhD, Cambridge, United Kingdom (Moderator) Nothing to Disclose

LEARNING OBJECTIVES

1) Learn about he new PET tracers and they new potential clinical applications. 2) Review the added value of PET/MRI in oncology.3) Learn about the current and future applications of hyperpolarised MRI.

Sub-Events

ParticipantsJason S. Lewis, PhD, New York, NY (Presenter) Shareholder, pHLIP, Inc; Research support, MabVax Therapeutics, Inc; Researchsupport, Eli Lilly and Company; Research support, Sapience Therapeutics; Research support, Telix Pharmaceuticals

LEARNING OBJECTIVES

1) To have an appreciation for some of the latest PET tracers in clinical research in oncology. 2) Understand the PET andradiotherapy agents currently FDA approved and those undergoing the approval process. 3) Understand the next generation of PETtracers and molecular imaging agents that could be the next standard-of-care imaging probes.

ParticipantsHebert Alberto Vargas, MD, Cambridge, United Kingdom (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) To understand the concept of value in imaging and how it relates to PET/MR technology. 2) To discuss the need for researchspecifically geared toward assessing the value of PET/MRI in oncology.

ParticipantsFerdia A. Gallagher, PhD, FRCR, Cambridge, United Kingdom (Presenter) Research support, General Electric Company; Researchsupport, GlaxoSmithKline plc


[email protected]

LEARNING OBJECTIVES

1) To explore the role of metabolism in cancer development. 2) To understand how these changes in metabolism can be exploitedusing hyperpolarised 13C-pyruvate. 3) To review the current evidence for hyperpolarised carbon-13 imaging in oncology. 4) Tounderstand potential clinical applications for hyperpolarised carbon-13 imaging. 5) To consider the role of new hyperpolarisedmolecules in oncology.

ABSTRACT

There is increasing evidence to support a role for metabolism in tumor development; for example, deregulation of cellular energeticsis now considered to be one of the key hallmarks of cancer. Changes in tumor metabolism over time are now known to be earlybiomarkers of successful response to chemotherapy and radiotherapy. There are a number of imaging methods that have been usedto probe cancer metabolism: the most widely available is 18F-fluorodeoxyglucose (FDG), an analogue of glucose, used in PET.Hyperpolarized carbon-13 MRI (13C-MRI) is an emerging molecular imaging technique for studying cellular metabolism, particularly inthe field of oncology. This method allows non-invasive measurements of tissue metabolism in real-time. To date, the most promisingprobe used in conjunction with hyperpolarized MRI has been 13C-labelled pyruvate: pyruvate is metabolized into lactate in normaltissue in the absence of oxygen, but in tumors this occurs very rapidly even in the presence of oxygen. Results from many animalmodels have shown that there is a reduction in the metabolism of pyruvate following successful treatment with chemotherapy.Tumor lactate labelling has also been shown to correlate with the grade of some tumor types. There are now a small number ofsites performing human hyperpolarized carbon-13 MRI imaging. This talk will discuss the progress that has been made in this fieldwithin the area of oncology and potential clinical applications.

RC522A State of the Art in PET Imaging

RC522B Technical Challenges in the Integration of PET Imaging into Radiotherapy Treatment Planning

RC522

Advanced PET Imaging for Radiotherapy Planning and Response Assessment

Wednesday, Nov. 28 8:30AM - 10:00AM Room: N226

BQ NM PH RO


ParticipantsPaul E. Kinahan, PhD, Seattle, WA (Moderator) Research Grant, General Electric Company; Co-founder, PET/X LLC

Sub-Events

ParticipantsPaul E. Kinahan, PhD, Seattle, WA (Presenter) Research Grant, General Electric Company; Co-founder, PET/X LLC

LEARNING OBJECTIVES

1) Understand the connections between the capabilities of PET imaging and clinical and research uses. 2) Become familiar withrecent technical advances in PET imaging and tradeoffs. 3) Gain awareness of initiative in quantitative imaging for clinical trials.

ParticipantsStephen R. Bowen, PhD, Seattle, WA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Understand the differences between diagnostic and treatment planning PET/CT imaging technical requirements. 2) Becomefamiliar with the source and propagation of technical errors in PET/CT-guided radiation therapy. 3) Gain awareness of technicaldesign elements in PET/CT-guided radiation therapy clinical trials.

MSRO42

BOOST: Lymphoma-Oncology Anatomy and Case-based Multidisciplinary Review (Interactive Session)

Wednesday, Nov. 28 10:30AM - 12:00PM Room: S103CD

CT MR NM OI RO



ParticipantsChelsea C. Pinnix, MD, PhD, Houston, TX (Moderator) Research Grant, Merck & Co, Inc; Consultant, Global One Inc; Speaker,International Journal of Radiation Oncology, Biology & PhysicsJurgen Rademaker, MD, New York, NY (Presenter) Nothing to DiscloseBradford Hoppe, MD, Jacksonville, FL (Presenter) Nothing to DiscloseAlison M. Friedmann, MD, Boston, MA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Case-based review of staging and treatment response in lymphoma (CT, PET, MRI).

SSK03-01 Predictive Value of Cardiac CT, Cardiac MR, and Transthoracic Echocardiography for CardioembolicStroke Recurrence

SSK03-02 Quantitative Assessment of Myocardial Infarction with Computed Tomography (CT) Using a BolusContrast Injection Scheme: Comparison Between Extravascular Contrast Distribution Volume (ECDV)and Extracellular Volume Fraction (ECV)


SSK03

Cardiac (CT, MRI and PET: General Topics)


CA CT MR NM



ParticipantsSatinder P. Singh, MD, Birmingham, AL (Moderator) Nothing to DiscloseJens Bremerich, MD, Basel, Switzerland (Moderator) Nothing to DiscloseJadranka Stojanovska, MD, Ann Arbor, MI (Moderator) Nothing to Disclose

Sub-Events

ParticipantsSimon S. Martin, MD, Frankfurt, Germany (Presenter) Nothing to DiscloseFrancesco Lavra, MD, Cagliari, Italy (Abstract Co-Author) Nothing to DiscloseCarlo N. De Cecco, MD, PhD, Atlanta, GA (Abstract Co-Author) Research Grant, Siemens AGAkos Varga-Szemes, MD, PhD, Charleston, SC (Abstract Co-Author) Research Grant, Siemens AGLuca Saba, MD, Cagliari, Italy (Abstract Co-Author) Nothing to DiscloseU. Joseph Schoepf, MD, Charleston, SC (Abstract Co-Author) Research Grant, Astellas Group; Research Grant, Bayer AG; ResearchGrant, Siemens AG; Research support, Bayer AG; Consultant, Guerbet SA; Consultant, General Electric Company; Consultant,HeartFlow, Inc; Consultant, Bayer AG; Consultant, Siemens AG; ; ; Marly van Assen, MSc, Charleston, SC (Abstract Co-Author) Nothing to DiscloseTaylor M. Duguay, Charleston, SC (Abstract Co-Author) Nothing to DiscloseBrian E. Jacobs, BS, Charleston, SC (Abstract Co-Author) Nothing to DiscloseMarco Scarabello, MD, Milan, Italy (Abstract Co-Author) Nothing to DiscloseParkwood Griffith, Charleston, SC (Abstract Co-Author) Nothing to DiscloseMarwen Eid, MD, Charleston, SC (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To determine the sensitivity, specificity, and predictive value of cardiac CT angiography (cCTA), cardiac MR (CMR), andtransthoracic echocardiography (TTE) for stroke recurrence in patients with suspected cardioembolic stroke.


163 patients (55% men, 61.9±16.9 years) with suspected cardioembolic stroke who underwent TTE, CMR, or cCTA betweenJanuary 2013 and May 2017 were retrospectively analyzed. The presence of left atrial thrombus, left ventricular thrombus, complexaortic plaque, cardiac tumors, and valvular vegetation was evaluated. The patient electronic medical records were used todetermine if the patient suffered a recurrent stroke. The sensitivity, specificity, positive predictive value (PPV), and negativepredictive value (NPV) were calculated for each imaging modality and the diagnostic accuracy was compared using receiveroperating characteristic analysis.

RESULTS

cCTA was performed in 82 patients, CMR in 81 patients, and TTE in 151 patients. 28 recurrent strokes occurred (cCTA- n=14;CMR- n=14; TTE- n=26). The sensitivity, specificity, PPV and NPV were: 14%, 79%, 12.5%, and 81.5% for CMR; 28.5%, 88.2%,33.3%, and 85.7% for cCTA; and 11.5%, 88.8%, 17.6%, and 82.8% for TTE. There was no significant difference in diagnosticaccuracy between CMR (0.53, 95% CI [0.42, 0.64]), cCTA (0.56, 95% CI [0.43, 0.69]), and TTE (0.50, 95% CI [0.43, 0.57]).

CONCLUSION

cCTA, CMR, and TTE demonstrated comparably high specificity and NPV for the exclusion of cardioembolic stroke recurrence.


The comparable performance of cCTA, CMR, and TTE in predicting recurrent cardioembolic stroke allows physicians to choose apreferred imaging modality for patients with suspected cardioembolic stroke.

SSK03-03 Feasibility of Myocardial Extracellular Volume Fraction Quantification Using Dual-Energy CT


ParticipantsJie Yu, Wuhan, China (Presenter) Nothing to DisclosePing Han, MD, Wuhan, China (Abstract Co-Author) Nothing to DiscloseTing-Yim Lee, MSc, PhD, London, ON (Abstract Co-Author) License agreement, General Electric CompanyAaron So, PhD, London, ON (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Myocardial viability can be assessed with CT by delineating the infarcted tissue with a higher degree of contrast retention in thelate phase after bolus injection of iodinated contrast. In this study, we investigated the effectiveness of two metrics for thequantification of late iodine enhancement in myocardium for viability assessment.


Reperfused acute myocardial infarction was induced in four farm pigs with 1-hr occlusion of the left anterior ascending artery (LAD)with a balloon catheter, and CT studies were performed within 2 weeks after the interventional procedure. After bolus injection ofcontrast at 3 mL/s and 0.7 mg/kg followed by saline flush, a 4-phase dynamic acquisition covering 10 min was performed with a GEHealthcare Revolution CT scanner at 100 kV, 100 mA, 280 ms/rot: 1st phase: 22 axial scans every 1-2 diastoles; 2nd: 6 scansevery 15 s; 3rd: 4 scans every 30 s; 4th: 6 scans every 60 s. Dynamic images were analyzed with a model-based deconvolutionapproach, with the modified Johnson-Wilson-Lee tracer kinetic model used to describe the contrast exchange among the cellularand interstitial and vascular spaces in myocardium to derive ECDV in ml/g. The difference images were also generated bysubtracting the images acquired at 10 min post contrast injection by the baseline images to obtain enhancement in the myocardium(ΔHUmyo) and left ventricular blood pool (ΔHUblood). ECV was then calculated as (1-Hematocrit)·(ΔHUmyo/ΔHUblood). ECDV andECV in normal (LCx territory) and infarcted myocardium (LAD territory) were compared.

RESULTS

Mean ECV in normal and infarcted myocardium were 0.27±0.11 and 0.52±0.11 respectively. The corresponding mean ECDVcalculated from the dynamic images covered up to 3 min post contrast injection were 0.28±0.07 ml/g and 0.60±0.10 ml/grespectively. Infarcted myocardium exhibited a higher percentage increase in ECDV from normal myocardium (114%) compared toECV (93%).

CONCLUSION

ECDV may be a more sensitive marker of myocardial viability compared to ECV due to the larger difference exhibited between thenormal and infarcted tissues, and can be measured with only 1/3 of the time required for ECV (3 min vs. 10 min post contrastinjection).


With bolus contrast injection, ECDV measurement could provide a faster and more reliable assessment of myocardial viability afteracute myocardial infarction compared to conventional ECV measurement.

ParticipantsMarly van Assen, MSc, Charleston, SC (Abstract Co-Author) Nothing to DiscloseParkwood Griffith, Charleston, SC (Presenter) Nothing to DiscloseCarlo N. De Cecco, MD, PhD, Atlanta, GA (Abstract Co-Author) Research Grant, Siemens AGPooyan Sahbaee, Malvern, PA (Abstract Co-Author) Employee, Siemens AGMatthijs Oudkerk, MD, PhD, Groningen, Netherlands (Abstract Co-Author) Nothing to DiscloseU. Joseph Schoepf, MD, Charleston, SC (Abstract Co-Author) Research Grant, Astellas Group; Research Grant, Bayer AG; ResearchGrant, Siemens AG; Research support, Bayer AG; Consultant, Guerbet SA; Consultant, General Electric Company; Consultant,HeartFlow, Inc; Consultant, Bayer AG; Consultant, Siemens AG; ; ; Marwen Eid, MD, Charleston, SC (Abstract Co-Author) Nothing to DiscloseRozemarijn Vliegenthart, MD, PhD, Groningen, Netherlands (Abstract Co-Author) Instutional Research Grant, Siemens AGBrian E. Jacobs, BS, Charleston, SC (Abstract Co-Author) Nothing to DiscloseAkos Varga-Szemes, MD, PhD, Charleston, SC (Abstract Co-Author) Research Grant, Siemens AG


[email protected]

PURPOSE

To assess the feasibility of Dual Energy CT (DECT) to derive myocardial extracellular volume (ECV) and detect ECV differenceswithout the need for a true non-contrast scan compared to Single Energy CT (SECT) results.


A total of 35 patients were included in this IRB-approved, HIPAA-compliant study; 8 control patients, 17 infarct patients (focalfibrosis), and 10 cardiomyopathy patients (diffuse fibrosis). All scans were acquired using a 2nd or 3rd generation dual source CTsystem. A true non-contrast and delayed acquisition were used to calculate SECT-ECV, while only the delayed acquisition in dualenergy mode and derived virtual non-contrast images were used to calculate DECT-ECV. In the control and diffuse fibrotic groups,a region of interest (ROI) encompassing the entire left ventricular myocardium was used to calculate ECV. Two ROIs were placed inthe focal fibrotic group; one in normal myocardium and one in fibrotic myocardium.

RESULTS

The median ECV was 33.4% (IQR, 30.1-37.4) for the SECT approach and 34.9% (IQR, 31.2-39.2) for the DECT approach (p =

SSK03-04 Accuracy of Myocardial Blood Flow Quantification with Dual-source CT: Validation in Human Using15O-Water PET


SSK03-05 Relationship Between Epicardial Adipose Tissue and Coronary Vascular Function in Patients withNormal Myocardial Perfusion by 82Rb PET/TC


0.401). For both SECT-ECV and DECT-ECV, focal fibrotic and diffuse fibrotic tissue had significantly higher ECV values compared tonormal myocardium (all p < 0.021). No systematic bias was observed between SECT and DECT measurements, with limits ofagreement calculated at ± 9.4% (p = 0.348). The DECT acquisition had a lower radiation dose than the SECT scan by 1.1 mSv (p <0.001), which was likely caused by the absence of the true non-contrast acquisition in the DECT approach.

CONCLUSION

Measurement of ECV with only a delayed acquisition is feasible using the DECT approach. The DECT approach provides similarresults at a lower radiation dose compared to a SECT protocol.


This study demonstrates the feasibility of DECT for myocardial ECV measurements using only a delayed acquisition, thus eliminatingthe need for a true non-contrast scan and consequently reducing radiation dose.

ParticipantsMasafumi Takafuji, Tsu, Japan (Presenter) Nothing to DiscloseKakuya Kitagawa, MD, PhD, Tsu, Japan (Abstract Co-Author) Nothing to DiscloseYasutaka Ichikawa, MD, Tsu, Japan (Abstract Co-Author) Nothing to DiscloseMasaki Ishida, MD, PhD, Tsu, Japan (Abstract Co-Author) Nothing to DiscloseYoshitaka Goto, MD, Tsu, Japan (Abstract Co-Author) Nothing to DiscloseSatoshi Nakamura, MD, Tsu, Japan (Abstract Co-Author) Nothing to DiscloseHajime Sakuma, MD, Tsu, Japan (Abstract Co-Author) Research Grant, Fuji Pharma Co, Ltd; Research Grant, DAIICHI SANKYOGroup; Research Grant, FUJIFILM Holdings Corporation; Research Grant, Siemens AG; Research Grant, Nihon Medi-Physics Co, Ltd;Speakers Bureau, Bayer AG


[email protected]

PURPOSE

Myocardial CT perfusion has emerged as a potential method for absolute quantification of myocardial blood flow (MBF). However,there is no standardized technique for CT MBF quantification, and dual-source CT MBF values have never been compared againstpositron emission tomography (PET), which is an established technique for non-invasive quantification of MBF. The aim of this studywas to assess the accuracy and usefulness of absolute MBF values quantified with dual-source CT by comparing them with thosequantified with 15O-water PET.


Dynamic CT perfusion and 15O-water PET were performed in 26 patients (70+/-9 years, 22 male) with known/suspected coronaryartery disease with a median interval of 48 days (interquartile range: 29-73 days). Hyperemic MBF in AHA 16 segments werequantified with a dual-source CT and its dedicated software (Force/VPCT body, Siemens). For the quantification of hyperemic MBFusing 15O-water PET, non-commercial software (Carimas) was used. Comparison of hyperemic MBF quantified by CT and PET wasperformed on segment (n=377), vessel (n=77), and patient (n=26) levels after exclusion of 7 segments out of FOV and 32segments with transmural myocardial infarction.

RESULTS

CT results showed excellent linear correlation with PET results at segment (r=0.87, p<0.0001), vessel (r=0.91, p<0.0001), andpatient level (r=0.93, p<0.0001). Area under the receiver-operating characteristics curve for detecting reduced MBF (<2.3mL/min/g) on 15O-water PET was 0.88, 0.91, and 0.92 at segment, vessel, and patient level, respectively. Although CTdemonstrated significantly lower hyperemic MBF than PET (1.16 ± 0.29 mL/min/g vs 2.46 ± 1.56 mL/min/g, p<0.0001), there wasgood per-vessel sensitivity (79.5%), specificity (92.1%), negative predictive value (81.4%) and positive predictive value (91.2%)for diagnosing reduced PET-derived MBF with a CT-derived MBF cutoff value of 1.09 mL/min/g.

CONCLUSION

Hyperemic MBF quantified by CT demonstrated excellent correlation with MBF estimated by 15O-water PET, and yielded highdiagnostic accuracy for detecting abnormal perfusion.


CT MBF quantification has potential to provide objective assessment of perfusion abnormality in patients with known or suspectedCAD with high accuracy comparable to 15O-water PET.

ParticipantsMassimo Imbriaco, MD, Napoli, Italy (Presenter) Nothing to DiscloseAndrea Ponsiglione, MD, Naples, Italy (Abstract Co-Author) Nothing to DiscloseCarmela Nappi, MD, Naples, Italy (Abstract Co-Author) Nothing to DiscloseSerena Dell'Aversana, MD, San Marcellino, Italy (Abstract Co-Author) Nothing to DiscloseMarta Puglia, Napoli, Italy (Abstract Co-Author) Nothing to DiscloseLudovica D'Acierno, Napoli, Italy (Abstract Co-Author) Nothing to DiscloseAlberto Cuocolo, Napoli, Italy (Abstract Co-Author) Nothing to Disclose


SSK03-06 Beyond CT-Fractional Flow Reserve (FFR): Non-Invasive Assessment of Instantaneous Wave FreeRatio (iFR), Coronary Flow Reserve (CFR) and Hyperemic Stenosis Resistance Index (HSR) from CTA


[email protected]

PURPOSE

We assessed the relationship between epicardial adipose tissue (EAT) and coronary flow reserve (CFR) in patients with suspectedor known coronary artery disease (CAD) and normal myocardial perfusion imaging (MPI).


The overall population consisted of 272 subjects referred for the evaluation of suspected or known CAD to stress-rest 82Rb PET/CTand showing normal MPI. CAC score was measured according to the Agatston method. Using unenhanced CT images for CAC, EATvolume was measured (cm3). The ln(CAC+1) score and lnEAT transformation were used to reduce heteroscedasticity. Myocardialperfusion was assessed using standardized segmentation of 17 myocardial regions. The summed stress, summed rest and summeddifference scores were automatically calculated. Myocardial perfusion was considered normal when the summed stress score was<3. Absolute myocardial blood flow (MBF) was computed (in milliliters per minute per gram) from the dynamic rest and stressimaging series. CFR was defined as the ratio of hyperemic to baseline MBF; CFR 2 was considered reduced.

RESULTS

In the overall population, 95 (35%) patients showed reduced and 177 (65%) normal CFR. Compared to patients with normal CFR,those with reduced CFR were older (60±11 vs. 67±9, P<0.05) and showed higher values of ln(CAC+1) (3.9±3 vs. 4.7±3, P<0.05)and lnEAT volume (4.5±1 vs. 4.7±.1, P<0.05). At univariable logistic regression analysis age, ln(CAC+1) and lnEAT resultedsignificant predictors of reduced CFR. At multivariable analysis, only age and lnEAT volume were independently associated withreduced CFR (hazard ratio 1.05 and 1.89 and 95% confidence interval 1.02-1.08 and 1.01-3.54, P<0.005). The addition of lnEAT toclinical data significantly increased the global chi-square of the model (from 23.8 to 28.6, P<0.05) in predicting reduced CFR.

CONCLUSION

In patients with suspected and known CAD and normal myocardial perfusion, age and EAT are strongly associated with reduced CFRconfirming that visceral fat depot may directly influence coronary vascular function. Thus, EAT evaluation may play a major role inthe identification of coronary vascular dysfunction in patients with normal perfusion.


In patients with suspected and known CAD and normal myocardial perfusion, age and EAT are associated with impaired CFRconfirming that visceral fat may directly influence coronary vascular function.

ParticipantsAndreas Giannopoulos, MD, Zurich, Switzerland (Abstract Co-Author) Nothing to DiscloseAnji Tang, Boston, MA (Abstract Co-Author) Nothing to DiscloseFrank J. Rybicki III, MD, PhD, Ottawa, ON (Abstract Co-Author) Medical Director, Imagia Cybernetics IncDimitris Mitsouras, PhD, Boston, MA (Presenter) Research Grant, Canon Medical Systems Corporation;


[email protected]

PURPOSE

Although FFR can be assessed by CTA, patient management in the cathlab uses eg, iFR to avoid pharmacologic stress, or CFR andHSR to assess other factors associated with increased risk of major adverse cardiac events, such as microvascular disease orendothelial dysfunction. Estimating these metrics from CTA requires matched baseline and hyperemic simulations (ie, stresssimulated using results obtained from rest simulation), which no technology to date provides. We sought to determine if couplingrest/stress computational fluid dynamics (CFD) simulations from CTA is feasible, and whether resulting metrics agree with knownrelationships of those metrics to reference-standard FFR.


Rest-stress hemodynamics from CTA were performed for 50 patients with invasive FFR in intermediate lesions in <90d of CTA. RestCFD was performed using only CTA data (myocardial mass, Murray's law). Stress CFD was then performed by coupling the epicardialarteries to a microvascular resistance model using the resistances estimated by the rest CFD for each myocardial territory.Hemodynamic metrics were calculated at the same location as invasive FFR as follows: CT-FFR=stress P/aortic stress P; CT-iFR=rest P/aortic rest P; CT-CFR from average peak velocity (APV), CFR=stress APV/rest APV; and CT-HSR=(aortic stress P-stressP)/stress APV. Correlation to FFR and receiver operating characteristic area-under-the-curve (AUC) to predict FFR<=0.8 wasdetermined for each metric.

RESULTS

Target lesion DS was 46.8±8.7%, and 40% had FFR<=0.8. Pearson correlations against invasive FFR of CT-FFR, CT-IFR, CT-CFRand CT-HSR were r=0.70, 0.69, 0.35, and -0.71, respectively (all p<0.01). Diagnostic accuracy to detect FFR<=0.8 was 0.87(95%CI:0.77-0.98), 0.86 (95%CI:0.75-0.97), 0.72 (95%CI:0.56-0.88), and 0.9 (95%CI:0.8-0.99), respectively. These matchreported relationships between invasive FFR and iFR, CFR and HSR (eg, Pearson r~0.75 for iFR and r~0.34 for CFR compared to FFR,and AUC of FFR to predict significant HSR of ~0.94).

CONCLUSION

Coupling baseline and hyperemic simulations enables key physiologic parameters dependent on both pressure and flow to beestimated non-invasively from standard retrospective CTA.


Evaluation of coronary artery disease by CTA can non-invasively assess coronary physiology beyond FFR, delivering key physiologic

SSK03-07 Feasibility of Coronary Flow and Velocity Measurement using 4D CTA Reconstruction


SSK03-08 Deep Learning Reconstruction of Non-Contrast Magnetic Resonance Coronary Angiography at 3TMachine


information that is used in clinical decision making for patients with angina

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Frank J. Rybicki III, MD, PhD - 2016 Honored Educator

ParticipantsMartin Wagner, PhD, Madison, WI (Presenter) Owner, LiteRay Medical LLCPaul F. Laeseke, MD, PhD, Madison, WI (Abstract Co-Author) Consultant, NeuWave Medical, Inc; Shareholder, Elucent Medical;Shareholder, HistoSononics; Shareholder, McGinley OrthopaedicsFred T. Lee JR, MD, Madison, WI (Abstract Co-Author) Consultant, NeuWave Medical, Inc Stockholder, HistoSonics, IncMichael Speidel, PhD, Madison, WI (Abstract Co-Author) Nothing to DiscloseCharles M. Strother, MD, Madison, WI (Abstract Co-Author) Research Consultant, Siemens AG Research support, Siemens AGLicense agreement, Siemens AGCharles A. Mistretta, PhD, Madison, WI (Abstract Co-Author) Founder, Mistretta Medical Intellectual Property Licensing Activities;Research, Siemens AG; Co-Founder, LiteRay Medical LLC


[email protected]

PURPOSE

Commercially available CT scanners can achieve gantry rotation times of 0.3 s (~3 frames per second). However, the temporalresolution might not be sufficient to calculate blood flow and velocity in the coronary arteries, which are important for the diagnosisof coronary artery disease (CAD). Currently, ~1000 projection images are acquired during each gantry rotation. The purpose of thisstudy was to determine the feasibility of a new reconstruction technique called 4D CTA, which calculates a 3D time frame for eachprojection image and therefore provides high temporal resolution for flow calculations in the coronary arteries.


The previously described 4D DSA technique (Davis, 2013) was extended for the time-resolved 3D reconstruction of coronaryarteries (CA). A pig study was retrospectively analyzed, where continuous axial CT acquisitions (64 slices) were performed with a0.4 s gantry rotation time over a period of 50 s during intravenous contrast injection. A 3D image of the vasculature wasreconstructed using a short scan (235 degrees) during diastole and the CA were manually segmented. A constrained back-projection was then performed for each projection image to create a 3D time frame. The reconstructed time attenuation curveswere used to calculate the blood flow and velocity in the CA based on the mean transit time. The velocity and flow values werecompared to values from literature and the flow conservation was determined.

RESULTS

In the first order branches, the average diameter, velocity and flow were 4.12 mm (3.28 mm), 110.43 mm/s (128 mm/s), and 1.44ml/s (1.07 ml/s) respectively. The same measurements for the second order branches were 1.31 mm (1.70 mm), 63.34 mm/s (46.10mm/s) and, 0.32 ml/s (0.10 ml/s) respectively. Values given in brackets are from literature as reported in Kassab et al. (1997) Theflow conservation in the measured branches of the CA was 96.71 %.

CONCLUSION

Calculated coronary arterial velocity and flow correlated well with previously reported values from the literature suggesting thatflow determination from 4D CTA is feasible. Additionally, the high flow conservation shows that the calculated values areconsistent.


The presented technique could provide both anatomical and functional information in diagnostic settings as well as cath labs usingexisting CT systems to detect pathologies of the coronary arteries.

ParticipantsDaisuke Utsunomiya, MD, Kumamoto, Japan (Presenter) Nothing to DiscloseMasafumi Kidoh, Kumamoto, Japan (Abstract Co-Author) Nothing to DiscloseKosuke Morita, Kumamoto, Japan (Abstract Co-Author) Nothing to DiscloseSeitaro Oda, MD, Kumamoto, Japan (Abstract Co-Author) Nothing to DiscloseTakeshi Nakaura, MD, Kumamoto, Japan (Abstract Co-Author) Nothing to DiscloseYasuyuki Yamashita, MD, Kumamoto, Japan (Abstract Co-Author) Consultant, DAIICHI SANKYO GroupYuichi Yamashita, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems Corporation

PURPOSE

Dedicated T2 preparation pulse have enabled non-contrast magnetic resonance coronary angiography (MRCA) at 3T system;however, the vascular contrast-to-noise ratio (CNR) is still inadequate for clinical use. The deep learning reconstruction (DLR) is anovel technique to improve the image quality. The purpose of this study was to investigate the effects of DLR on the image qualityof 3T non-contrast MRCA.


SSK03-09 Clinical and Transthoracic Echocardiography Predictors of Non-Detectable Left Ventricular Thrombus:When Is Cardiac Magnetic Resonance Necessary


We enrolled 10 volunteers (2 female, mean age 48 years) with no known coronary artery disease. Non-contrast MRCA wasperformed on a 3T MR scanner (Galan 3T ZGO, Canon medical) with following parameters: 3D fast FE, TR/TE =5.3/1.9ms, flip angle= 12°, slice thickness = 1.7mm with ECG trigger and real time motion correction. DLR images at moderate level and high level weregenerated by using dedicated workstation. In the quantitative evaluation, we measured signal-to-noise ratio of 3 coronary vessels(proximal and distal segments). In the qualitative evaluation, the 2 observers graded the vessel visualization and artifacts on a 4-point scale (worst, 1; best, 4).

RESULTS

The CNR (original MRCA) was 31 ± 7 and 16 ± 5 in the proximal and distal vessel, respectively. The corresponding CNR (moderate-level DLR) was 46 ± 9 and 24 ± 10; and the CNR (high-level DLR) was 85 ± 20 and 45 ± 14. The visual scores for overall imagequality and image noise were significantly better in DLR images than original images. The vessel sharpness scores were comparableamong 3 reconstructions (3.4, 3.8, and 3.8 for original, moderate DLR, and high DLR, respectively). The visual scores for imagenoise/graininess was significantly better in DLR (2.4, 3.8, and 4.0 for original, moderate DLR, and high DLR, respectively).

CONCLUSION

Non-contrast MRCA at 3T using DLR provides higher CNR without degrading the vessel sharpness.


The deep learning reconstruction technique contributes in improved visualization of coronary arteries in non-contrast MR coronaryangiography, enabling noninvasive scrutiny of the heart.

ParticipantsRolf Symons, MD, Leuven, Belgium (Presenter) Nothing to DiscloseJan Verduyckt, Leuven, Belgium (Abstract Co-Author) Nothing to DiscloseDavide Curione, Roma, Italy (Abstract Co-Author) Nothing to DiscloseSteven Dymarkowski, MD, Leuven, Belgium (Abstract Co-Author) Nothing to DiscloseJan G. Bogaert, MD, Leuven, Belgium (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Currently, transthoracic echocardiography (TTE) remains the most commonly used technique for the identification of LV thrombi.However, not all thrombi are visualized with TTE. Therefore, the purpose of our study was to identify predictors of unsuccessfulTTE thrombus visualization and to develop a risk score to stratify which patients may benefit from cardiac magnetic resonance(CMR) to reliably detect or exclude LV thrombus.


We performed a retrospective search of our CMR database including 10300 patients and identified 118 patients with LV thrombusand a time interval between CMR and TTE of <72h. Univariate logistic regression analysis was used to assess the associationbetween baseline characteristics and TTE parameters with the primary endpoint (i.e. unsuccessful LV thrombus visualization onTTE). Variables with P<0.10 at univariate analysis were included as covariates in the multivariate logistic regression analysis.Receiver-operating characteristic (ROC) curve analysis was performed to examine differences in performance of each variable forprediction of the primary endpoint. A two-sided P-value<0.05 was considered to represent a significant difference.

RESULTS

In multivariate analysis, body mass index (BMI), LV end-diastolic diameter (EDD), and mitral valve regurgitation (MVR) wereidentified as significant predictors of unsuccessful LV thrombus visualization by TTE (all P<0.001). ROC analysis showed BMI >=26.9kg/m2, LVEDD >=52 mm, and MVR >=2/4 to be the optimal cutoff points for prediction of the primary endpoint. The combination ofthe independent predictors allowed generation of a gradient response risk score of unsuccessful LV thrombus visualization by TTE(0/3 present: 0% missed; 1/3 present: 33.3% missed; 2/3 present: 79.5% missed; 3/3 present: 100% missed) (P<0.001).

CONCLUSION

Individual clinical and TTE parameters can predict the sensitivity of TTE for the successful detection of LV thrombus in heartdisease. By using the presented risk score, a cost-effective strategy may be implemented by selectively referring patients to CMRwhen these risk factors are present.


Clinical and TTE parameters can predict unsuccessful detection of LV thrombus in heart disease. These findings may lead to a cost-effective referral of certain patients to CMR to rule out LV thrombus.

SSK14-01 Incremental Value of Interim 18F-FDG PET/CT over CT-Scan for Early Response Evaluation inPatients with Hodgkin Lymphoma Treated with Immune Checkpoint Inhibitors

SSK14-02 Impact of FDG-PET/CT on the Staging, Management and Outcomes of Patients with PresumedLimited Stage Hodgkin's Lymphoma and Aggressive Non-Hodgkin's Lymphoma


SSK14

Nuclear Medicine (Lymphoma, Sarcoma and Melanoma)

Wednesday, Nov. 28 10:30AM - 12:00PM Room: S505AB

BQ CT NM


ParticipantsHelen R. Nadel, MD, Palo Alto, CA (Moderator) Nothing to DiscloseMatthew S. Robertson, MD, Cleveland, OH (Moderator) Nothing to Disclose

Sub-Events

ParticipantsAi-ping Chen, Nanjing, China (Presenter) Nothing to DiscloseFatima-Zohra Mokrane, MD, Toulouse, France (Abstract Co-Author) Nothing to DiscloseLawrence H. Schwartz, MD, New York, NY (Abstract Co-Author) Committee member, Celgene Corporation Committee member,Novartis AG Committee member, ICON plc Committee member, BioClinica, IncRomain-David Seban, Villejuif, France (Abstract Co-Author) Nothing to DiscloseSamy Ammari, Villejuif, France (Abstract Co-Author) Nothing to DiscloseRandy Yeh, MD, New York, NY (Abstract Co-Author) Nothing to DiscloseAurelien Marabelle, Villejuif, France (Abstract Co-Author) Nothing to DiscloseBinsheng Zhao, DSc, New York, NY (Abstract Co-Author) License agreement, Varian Medical Systems, Inc; Royalties, Varian MedicalSystems, Inc; License agreement, Keosys SAS; License agreement, Hinacom Software and Technology, Ltd; Roch Houot, Rennes, France (Abstract Co-Author) Nothing to DiscloseLaurent Dercle, MD, New York, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Anti-Programmed Death 1 (anti-PD1) antibody triggers new patterns of response and progression in patients with Hodgkin lymphoma(HL). We aimed to evaluate the incremental value of interim 18F-FDG PET/CT over CT scans in patients treated by anti-PD1.


We retrospectively analyzed patients treated by anti-PD1 from 2013 to 2017. Concomitant 18F-FDG PET/CT and CT scans wereacquired at baseline and upon treatment. A pair of radiologists classified blindly and independently patients as immune-respondingor immune-refractory based on the first evaluation, using the International Harmonisation Project Cheson 2014 criteria and theLymphoma Response to Immunomodulatory therapy Criteria (LYRIC) (2016 revised criteria).

RESULTS

Forty-four consecutive HL patients were included. Forty-four interim 18F-FDG PET/CT and CT scans were acquired at a mediantime of 3.7 months after anti-PD1 initiation. Radiologists classified patients as immune-responding or immune-refractory on both18F-FDG PET/CT and CT-scan in 55.7% and 35.2% of cases, respectively. Radiologists experienced a significant incremental valueof 18F-FDG PET/CT in 8.0% (95%CI: 3.3%-15.7%) of patients, whom were reclassified as immune-refractory (2.3%) or immune-responding (5.7%). Additionally, 18.2% (95%CI: 10.8%-27.8%) of patients were reclassified from PR on CT-scan to CR on PET.

CONCLUSION

CT-scan alone can reliably be used for response assessment in patients with HL treated with anti-PD1: radiologists correctlyclassified patients as immune-responding or immune-refractory in 92% of cases. However, interim 18F-FDG PET/CT showed clearincremental value to reclassify immune-responding patients from partial response to complete response, which is crucial for risk-adapted strategies.


Interim 18F-FDG-PET/CT in HL patients treated with anti-PD1 supplied incremental value over CT-scan by reclassifying patients toimmune-responding or complete response. This concept is crucial for risk-adapted therapeutic strategy.

ParticipantsUr Metser, MD, FRCPC, Toronto, ON (Presenter) Nothing to DiscloseAnca Prica, Toronto, ON (Abstract Co-Author) Nothing to Disclose

SSK14-03 Altered Liver FDG Uptake in Lymphoma Patients with Chemotherapy Associated Hepatic Steatosis onPET/CT


David C. Hodgson, MD, MPH, Toronto, ON (Abstract Co-Author) Nothing to DiscloseMindaugas Mozuraitis, Toronto, ON (Abstract Co-Author) Nothing to DiscloseMaria Eberg, Toronto, ON (Abstract Co-Author) Nothing to DiscloseVictor Mak, Toronto, ON (Abstract Co-Author) Nothing to DiscloseBo Green, Toronto, ON (Abstract Co-Author) Nothing to DiscloseAmit Singnurkar, MD, Hamilton, ON (Abstract Co-Author) Nothing to DiscloseJill Dudebout, Kingston, ON (Abstract Co-Author) Nothing to DisclosePamela Maccrostie, Toronto, ON (Abstract Co-Author) Nothing to DiscloseNoam Tau, MD, Petah Tikva, Israel (Abstract Co-Author) Nothing to DiscloseNicole Mittmann, PhD, Toronto, ON (Abstract Co-Author) Nothing to DiscloseDeanna L. Langer, PhD, Toronto, AB (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To determine the impact of positron-emission tomography/ computed tomography (PET/CT) on the staging, management andoutcomes of patients with apparent limited-stage (LS) Hodgkin's lymphoma (HL) or aggressive non-Hodgkin's lymphoma (aNHL)being treated with curative intent.


This single arm, prospective multicenter registry included patients with apparent LS HL or aNHL based on clinical data and CT, orwith equivocal CT findings for advanced stage, being considered for curative-intent therapy. Pre-PET/CT treatment plan wascompared to actual treatment received. Outcomes at 1 year post first line therapy included survival and second-line therapyinitiation. These were compared to a historical control pool staged with CT alone. Administrative data sources were used to obtainand control for baseline characteristics using propensity score matching and regression adjustment. Outcomes were assessed usingadjusted Cox proportional hazards regression and propensity score matching.

RESULTS

PET/CT upstaged 58/330 (17.6%) patients with HL and 92/520 (17.7%) patients with aNHL. Change in planned mode of therapywas seen in 119/266 (44.7%) patients with HL and 131/334 (39.2%) with aNHL (p<0.00001 for both). There was a lower 1-yearmortality for aNHL patients with LS on PET compared to those with LS on CT (for propensity score matched cohort: HR, 0.34; 95%CI: 0.15, 0.74; p=0.0072). For patients with HL, no significant difference was found in survival or second-line therapy initiation at 1year.

CONCLUSION

PET/CT upstaged >17% of patients with presumed LS aggressive lymphoma to advanced stage and planned management wasaltered in a significant proportion of patients. Patients with confirmed LS aNHL after PET/CT treated with curative intent hadsignificantly better survival compared to the cohort of LS determined by CT.


1. PET has a significant impact on the management of patients with HL and aNHL. 2. Patients with presumed limited stage aNHL onPET/CT treated with curative intent had a significantly better survival at one year compared to patients with presumed limitedstage as determined by CT. 3. These results support the recent recommendation of the International Conference on MalignantLymphoma in Lugano for the utilization of PET in the staging of patients with aggressive lymphoma.

ParticipantsYi W. Mo, Guangzhou, China (Presenter) Nothing to DiscloseYuan Hua Li, MD, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseWen Long, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseLu Li, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseXu Zhang, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseDan Lei, Guangzhou, China (Abstract Co-Author) Nothing to DiscloseWei Fan, Guangzhou, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the prevalence of hepatic steatosis in lymphoma patients after chemotherapy; to reveal whether lymphoma patientwith chemotherapy associated hepatic steatosis (CAHS) will change liver FDG uptake on PET/CT, compared to their baseline and tofurther study the relation between liver FDG uptake and severity of fatty liver.


88 of 1647 lymphoma patients had been diagnosed fatty liver during or after chemotherapy from December 1, 2014, to June 30,2017. 176 FDG PET/CT scans of the 88 lymphoma patients were reviewed retrospectively. These 88 lymphoma patients all had anormal liver before chemotherapy and then got hepatic steatosis after chemotherapy. Each patient had performed two PET/CTscans: baseline and post-chemotherapy. Paired t test was used to compare BMI, blood glucose (BG), liver SUVmax (SUVmax-l),liver average SUV (SUVave-l), liver SULmax (SULmax-l), aorta SUVmax (SUVmax-a), aorta average SUV (SUVave-a) and aortaSULmax (SULmax-a) between baseline and post-chemotherapy. CAHS was divided into three groups: mild-grade, moderate-grade,and severe-grade. The relationship within or between groups was assessed.

SSK14-04 Is Dual-Time Point 18F-FDG PET/CT Valuable for Differentiating Goitrous Hashimoto's Thyroiditisfrom Primary Thyroid Lymphoma?


SSK14-05 Utility of Integrated 18F-FDG PET/MRI for Response Assessment of Isolated Limp Perfusion inPatients with Soft-Tissue Sarcomas


RESULTS

The prevalence of CAHS in lymphoma patients of our hospital was about 88/1647 (5.3%). After chemotherapy, 28 of 88 (31.8%)patients had increased liver SUVs, whereas 60 of 88 (68.2%) patients showed decreased liver SUVs. There were significantdifferences of mean liver SUVs between baseline and CAHS (baseline versus CAHS; SUVmax-l, 2.84±0.57 vs 2.57±0.64, P<0.00;SUVave-l, 2.17±0.43 vs 1.95±0.51, P=0.001; SULmax-l, 2.24±0.40 vs 2.0±0.41, P<0.001). BG had a slight decrease afterchemotherapy (baseline versus CAHS, 5.5±1.2 vs 5.2±1.0, P=0.01). No difference was identified when the mean aorta SUVs andBMI for baseline were compared with those for CAHS (P>0.05). The patients with severe-grade of CAHS had significant lower liverSUV values, compared to those with mild-grade (P<0.05). And BMI showed no difference among the three groups of CAHS.

CONCLUSION

Increase of liver FDG uptake coexists with decrease. The severer the fatty liver is, the more likely the liver FDG uptake declines.The altered liver SUV due to CAHS might affect the response assessment and prognostic evaluation for lymphoma patients.


(dealing with 18F-FDG PET/CT)'18F-FDG PET/CT is a powerful imaging technique and has become the standard for staging andresponse assessment of FDG-avid lymphomas'.

ParticipantsKunihiro Nakada, Sapporo, Japan (Presenter) Nothing to DiscloseNaoya Hattori, MD, PhD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseHiroki Sugie, MD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseNoriyoshi Katoh, MD,PhD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseMasayuki Sakurai, Sapporo, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Goitrous Hashimoto's thyroiditis (HT) occasionally shows diffusely increased F-18 fluorodeoxyglucose (FDG) uptake in the thyroidthat it mimics primary thyroid lymphoma (PTL) on FDG-PET/CT i,ages. The aim of the study was to determine whether delayedimaging of FDG-PET/CT was valuable in differentiating PTL from HT.


53 patients with HT, who were suspected of PTL due to enlarging goiter, underwent dual-time-point (60 ± 5min and 120± 10minafter FDG injection ) PET/CT scan using FDG combined with neck ultrasound (US) and US-guided core needle biopsy. Specimen ofcore needle biopsy was subjected to immunohistochemical staining (CD20,CD3,CD79a,Ki-67, etc.) along with H-E staining. Re-arrangement of igH was also analyzed by PCR. In addition to visual assessment based on 5-point scale from The Luganoclassification, the maximum standardized uptake value for the thyroid at the early image (SUV-E) and that at the delayed image(SUV-D) for the thyroid were determined. In addition, SUV increment (ΔSUV) was calculated by subtracting SUV-E from SUV-D.Those parameters were compared between patients with PTL and those with HT.

RESULTS

Pathological diagnosis was PTL in 36 (MALT lymphoma 31, DLBCL 5) and was HT in 17. 11 patients with PTL was excluded fromanalysis because they had either nodular FDG uptake in the thyroid or abnormal uptake in extrathyroidal area. The remaining 25patients with PTL and all patients with HT showed diffusely increased thyroid FDG uptake on both the early and the delayed PET/CTimages. There was no statistically significant difference between PTL and HT in 5-PS (early 4.4 vs. 4.2. delayed 4.1 vs. 4.0 ) aswell as SUV-E and SUV-D (9.02 vs. 7.51, 8.28 vs. 6.54 ). 7 of 25 patients (28%) with PTL had plus values of ΔSUV while none butone (6%) with HT had plus value of ΔSUV. When plus value of ΔSUV was considered as a sign for PTL, PPV, NPV, and accuracy forPTL was 88%. 47%, and 55%, respectively.

CONCLUSION

Neither visual nor semi-quantitative analysis of dual tome-point FDG-PET/CT was helpful in differentiation HT from PTL amongenlarging goiter. Needle biopsy may be the best strategy in differential diagnosis of HT with enlarging goiter.


Addition of delayed imaging does not improve diagnostic performance of FDG-PET/CT in diffusely increased thyroidal uptake inenlarging goiter.

ParticipantsJohannes Grueneisen, Essen, Germany (Presenter) Nothing to DiscloseAhmet Oernek, Essen, Germany (Abstract Co-Author) Nothing to DiscloseWolfgang Fendler, Munich, Germany (Abstract Co-Author) Nothing to DiscloseAxel Wetter, Essen, Germany (Abstract Co-Author) Nothing to DiscloseNika Guberina, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseLars Podleska, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; Research

SSK14-06 Evaluation of 18F-FDG-PET/CT for Response Assessment in Patients with Advanced MelanomaTreated with Immune Checkpoint Inhibitors


Grant, Ipsen SALale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AG

PURPOSE

To evaluate the diagnostic potential of simultaneously obtained PET- and MR-datasets for therapy response assessment of isolatedlimp perfusion with TNF-alpha and melphalan (TM-ILP) in patients with soft-tissue sarcomas.


A total of 32 patients with histopathological confirmation of a soft-tissue sarcoma were prospectively enrolled for an integrated18F-FDG PET/MRI examination before (1st scan) and after (2nd scan) neoadjuvant TM-ILP. In each examination morphological(tumor size) and metabolic (SUVmax, SUVpeak) parameters of the tumors were determined. Two readers analysed the datasets andassessed treatment response based on RECIST 1.1 and PERCIST criteria. Results from subsequent tumor resection served asreference standard and therapy response was determined based on the tumour regression grading scale of Salzer-Kuntschik.

RESULTS

Based on the reference standard, a total of 25 patients were classified as partial responder (PR) and 7 patients as stable disease(SD). Calculated mean values of the maximum tumor diameter, SUVmax and SUVpeak in patients with stable disease amounted to62.4 ± 42.4mm, 11.1 ± 7.9 and 9.1 ± 6.2 before and 59.5 ± 50.3mm, 8.4 ± 5.3 and 6.7 ± 5.1 after treatment. The respectivevalues in the responder group were 78.1± 65.6mm, 11.9 ± 7.4 and 9.6 ± 6.1, before and 71.1 ± 65.9mm, 5.1 ± 3.1 and 3.9 ± 2.4after treatment. Based on RECIST criteria, 25 patients were classified as SD and 6 patients as PR, whereas 1 patient showedprogressive disease (PD). PERCIST criteria categorized 11 patients as SD, 20 patients as PR and one patient as PD. In accordancewith PERCIST, a significantly higher number of patients (n = 23, 71.9%) could be correctly categorized as SD/PR, when comparedto RECIST (n = 9, 28.1%; p < 0.005).

CONCLUSION

Our results demonstrate the significant discrepancy in morphological and metabolic response and underline the diagnostic superiorityof 18F-FDG PET data over MRI for response assessment of neoadjuvant ILP in sarcoma patients.


18F-FDG PET/MRI might enable more accurate therapy response assessment of isolated limp perfusion in patients with soft-tissuesarcomas when compared to MRI alone.

ParticipantsDominique Fuser, MD, Saint Louis, MO (Abstract Co-Author) Nothing to DiscloseLeonel Hernandez-Aya, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseJoyce C. Mhlanga, MBBCh, Portland, OR (Abstract Co-Author) Nothing to DiscloseJohn Crandall, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseLauren Ash, St. Louis, MO (Abstract Co-Author) Nothing to DiscloseRichard L. Wahl, MD, Saint Louis, MO (Abstract Co-Author) Research Consultant, Nihon Medi-Physics Co, Ltd; Contract,WhiteRabbit.AI IncDelphine L. Chen, MD, Saint Louis, MO (Presenter) Nothing to Disclose


[email protected]

PURPOSE

Treatment with immune checkpoint inhibitors (ICI) have improved outcomes for patients with advanced melanoma. The aim of thisretrospective study was to characterize the findings on 18F-FDG PET/CT that predict response in these patients.


Ninety-seven patients with advanced melanoma treated with ICI were identified, with 70 having baseline (PET0) and interim 18F-FDG PET/CT studies (PET1, median 84 and range 19 to 181 days after cycle 1). Of these, 34 were treated with ipilimumab alone(N=17) or combined with nivolumab (N=17, analyzed together as the ipi cohort) and 36 with pembrolizumab (pembro cohort) alone.AutoPERCIST software was used to determine the SULpeak and SUVmax of the hottest lesion on each scan. Clinical assessment atthe last office visit determined responders (R- partial response, stable, or no disease) and nonresponders (NR- progression) at amedian 9.5 (range 0 to 61) months after PET1 for the ipi cohort and 5.5 months (range 0 to 20) for the pembro cohort. Repeatedmeasures analysis of variance (SigmaPlot 12.5, Systat Software, Inc.) determined differences between PET0 and PET1 metricsamong response categories by cohort. The Mann-Whitney rank sum test assessed for differences in the % change in each metricbetween response categories.

RESULTS

Thirteen R and 21 NR were identified in the ipi cohort and 26 R and 10 NR in the pembro cohort. Within the ipi R cohort, SULpeak(PET0 6.8±5.1, PET1 1.2±2.4, p=0.002) and SUVmax (PET0 9.9±7.0, PET1 2.0±4.3, p=0.002), were significantly different. The PET1SULpeak (13±12, p=0.007) and SUVmax (19±19, p=0.006) of the NR was also significantly different from the R PET1 values. For thepembro NR cohort, SULpeak (PET0 6.2±5.3, PET1 10±8.7, p=0.017) and SUVmax (PET0 9.5±9.1, PET1 14±12, p=0.033) weresignificantly different, and the pembro R PET1 SULpeak (2.0±3.6, p<0.001) and SUVmax (3.4±6.0, p<0.001) were also significantlydifferent from NR PET1 values. The % change in all metrics was significantly different between R and NR in both treatment cohorts.

CONCLUSION

Changes in the SULpeak and SUVmax of the hottest lesion on interim 18F-FDG PET/CT studies may be useful in predictingresponders to ICI.

SSK14-07 Is 18F-FDG PET/MR Including DWI an Alternative to Sentinel Lymph Node Biopsy in Initial N-Stagingin Patients with Malignant Melanoma?


SSK14-08 Prognostic Value of 18F-FDG PET/CT in Intralesional Interleukin-2 Therapy for Cutaneous MetastaticMelanoma



Higher SULpeak and SUVmax values and lower % decrease in these parameters on interim 18F-FDG PET/CT scans during ICItreatment are predictors of poor treatment response.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Richard L. Wahl, MD - 2013 Honored Educator

ParticipantsBenedikt M. Schaarschmidt, MD, Essen, Germany (Presenter) Stockholder, Bayer AG; Stockholder, General Electric Company;Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries LtdJohannes Grueneisen, Essen, Germany (Abstract Co-Author) Nothing to DiscloseLino Sawicki, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseThorsten D. Poeppel, Essen, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

To compare 18F-fluordesoxyglucose positron emission tomography / computed tomography (18F-FDG PET/CT), 18F-FDG PET /magnetic resonance (18F-FDG PET/MR) as well as 18F-FDG PET/MR and diffusion weighted imaging (DWI) with sentinel lymph nodebiopsy (SLNB) in initial lymph node staging in patients with malignant melanoma.


In this retrospective study, 52 patients with malignant melanoma (female: n=30, male: n=22, mean age 50.5y) that underwent18F-FDG PET/CT and consecutive 18F-FDG PET/MRI including DWI prior to lymphoscintigraphy with single photon emissioncomputed tomography / CT (SPECT/CT) and consecutive SLNB were included. By two readers, the status of the sentinel lymphnodes detected by SPECT/CT (benign/malignant) was assessed on 18F-FDG PET/CT, 18F-FDG PET/MR as well as 18F-FDG PET/MRand DWI images. In all modalities, increased tracer uptake in comparison to the background was considered as a sign of malignancy.In PET/MR, morphological criteria were considered as additional signs of malignancy. In 18F-FDG PET/MR and DWI, all of theaforementioned criteria and diffusion restriction were considered as signs of malignancy. Discrepancies were resolved in aconsensus reading. Histopathologic results served as a reference standard to calculate sensitivity, specificity as well as positive(PPV) and negative predictive values (NPV).

RESULTS

In all patients, a total of 87 sentinel lymph nodes were detected by lymphoscintigraphy and SPECT/CT. According tohistopathology, lymph nodes were metastatic. We found a sensitivity, specificity, PPV and NPV of 17.7%, 95.6%, 50.0% and 82.3%for PET/CT and of 23.5%, 96.9%, 66.7% and 82.3% for PET/MR. Additional DWI was available in 56 lymph nodes and led to twoadditional false positive findings, thus decreasing specificity of PET/MR and DWI.

CONCLUSION

Due to its low sensitivity and specificity, 18F-FDG PET/MR cannot be considered an alternative to SLNB in initial N-Staging inpatients with malignant melanoma even if additional DWI is performed.


18F-FDG PET/MR is inferior to SLNB in N-Staging in patients with malignant melanoma. Therefore, neither 18F-FDG PET/MR nor 18F-FDG PET/MR and DWI will be able to replace SLNB in clinical routine.

ParticipantsRick Wray, MD, Sacramento, CA (Presenter) Nothing to DiscloseJason Kao, BS, Sacramento, CA (Abstract Co-Author) Nothing to DiscloseSarah Bateni, MD, Sacramento, CA (Abstract Co-Author) Nothing to DiscloseEmanuel Maverakis, MD, Sacramento, CA (Abstract Co-Author) Nothing to DiscloseAmanda Kirane, MD, Sacramento, CA (Abstract Co-Author) Nothing to Disclose

PURPOSE

Intralesional interleukin-2 (IL-2) therapy is an effective treatment for cutaneous metastatic melanoma even in cases refractory tomultiple treatment modalities. 18F-FDG PET/CT is a valuable tool for diagnosis, staging and surveillance of melanoma, but its role intherapy assessment is unclear given the strong local inflammatory response in injected sites. This study investigates the prognosticvalue of 18F-FDG PET/CT for assessment of intralesional IL-2 in cutaneous metastatic melanoma.


13 patients (10M/3F, 23 - 96 years) with Stage IIIC/IV cutaneous metastatic melanoma or stage IIB disease not amenable tosurgical intervention had a total of 31 PET/CT scans performed at baseline, interim and completion time points while receivingvariable injections of intralesional IL-2 (range, 1 - 20; mean 7.8) at variable doses (range 5 - 22 million units; mode 7) . PET/CTscans were evaluated using maximum SUV (SUV max) and a 5 point scale (5PS): 1, no uptake; 2, uptake <= mediastinum; 3,uptake > mediastinum <= liver; 4, uptake moderately < liver; 5, uptake markedly > than liver. The 5PS was dichotomized to

SSK14-09 Opportunistic Screening of FDG-PET/CT Reveals Undiagnosed Low Bone Mass in Patients BeingEvaluated for Oncology Purposes

Wednesday, Nov. 28 11:50AM - 12:00PM Room: S505AB

negative (score 1, 2, and 3) or positive (score 4 and 5). The Kaplan-Meier (KM) method with log-rank test and Cox-regressionanalysis were performed.

RESULTS

Of the 31 scans, 17 were positive and 14 were negative. SUV max range, 1.3 -20.6 g/ml. Follow-up range, 8 - 51 months (median15). Baseline scans with higher SUV max had a significantly lower PFS with hazard ratio (HR) of 1.55 (95% CI 1.001-1.33, p=0.048).KM curves demonstrated a trend of improved OS with lower SUVmax at baseline (p=0.11). Positive scans at completion trendedtoward lower OS with HR of 2.74, (p=0.48). The progression-free survival (PFS) was worse for positive groups at completion, HR of12.12 (95%CI 1.22-120.49, p=0.03), with significant separation in KM curves (Log-rank, p=0.008).

CONCLUSION

18F-FDG PET/CT SUV max at baseline and qualitative therapy assessment at completion time points during intralesional IL-2 canpredict PFS and show potential to predict OS, with larger sample size, in patients with cutaneous metastatic melanoma.


A baseline PET/CT prior to intralesional IL-2 adds prognostic value. Additionally, using a qualitative therapy assessment methodsuch as the 5PS increases the prognostic value of 18F-FDG PET/CT for therapy response in intralesional IL-2 treatment forcutaneous metastatic melanoma.

ParticipantsFernando U. Kay, MD, Dallas, TX (Presenter) Nothing to DiscloseEdmund Dosnumu, Dallas, TX (Abstract Co-Author) Nothing to DiscloseVinh Ho, Dallas, TX (Abstract Co-Author) Nothing to DiscloseKeenan Brown, Austin, TX (Abstract Co-Author) Employee, Mindways Software, Inc Stockholder, Mindways Software, IncOrhan K. Oz, MD,PhD, Dallas, TX (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To assess the prevalence of undetected low bone mineral density (BMD) in a cohort of patients undergoing PET/CT withopportunistic quantitative computed tomography (QCT).


A retrospective survey was conducted to identify PET/CT studies obtained between Oct/2015 and Jan/2016 in a Biograph 64scanner (Siemens). CT images were processed with the QCT Pro software (Mindways). A calibration CT scan was obtained in thesame PET/CT scanner using the asynchronous Model 4 QCT Phantom. Two radiologists and a trained research technician performedthe analyses of trabecular BMD at vertebral bodies L1 and L2. The American College of Radiology (ACR) criteria was used fordiagnosing low BMD. Total BMD of femoral necks were measured on DXA-equivalent images (CTXA) of the hips and used to generateFRAX-scores for calculating absolute fracture risks. We obtained clinical data from institutional medical records. Requirement forsigned informed consent was waived by the IRB.

RESULTS

Sixty-nine studies were identified, two studies excluded due to severe scoliosis and one excluded due to Schmörl node affectingthe analysis. The final cohort comprised 66 subjects (20F/46M, mean age: 53.8, SD: 12.1). Mean coefficient of variation (CV) fortrabecular BMD in L1-2 between the 3 readers was 1.2%. Distribution of subjects according to ACR category is shown in Table 1.Thirty-two percent (21/66) of subjects showed low lumbar spine BMD on QCT. Twenty-four percent (5/21) of subjects with lowBMD on QCT had a prior DXA scan, all of which showing low BMD. None of the subjects with normal BMD on QCT had a prior DXAscan. Femoral neck BMD was assessed with CTXA in 20 of 66 subjects by one radiologist and the research technician. Mean CVbetween the readers was 6.1%. Fifteen percent (3/20) of the subjects had at least a 3% risk of hip fracture within 10 years. Only33.3% (1/3) of these subjects had a prior diagnosis of low BMD.

CONCLUSION

Low BMD was an under recognized condition in our sample. Future analysis will correlate metabolic activity (FDGuptake) with bonemass.


PET/CT provides a unique opportunity to screen patients for occult low BMD by leveraging the quantitative capabilities of CT.Identification of subjects at risk for future osteoporotic fractures may not only improve outcomes, but also decrease downstreamcosts.

NM226-SD-WEA1

Impact of Different Iterative Metal Artifact Reduction (iMAR) Algorithms on PET/CT AttenuationCorrection in Artifacts After Port Implantation

Station #1

NM227-SD-WEA2

Essential Information Which Helps to Reduce Negative FDG PET-CT Study in Investigating LargeVessel Vasculitis

Station #2

NMS-WEA

Nuclear Medicine Wednesday Poster Discussions

Wednesday, Nov. 28 12:15PM - 12:45PM Room: NM Community, Learning Center

NM



ParticipantsPeter L. Choyke, MD, Rockville, MD (Moderator) Nothing to DiscloseHelen R. Nadel, MD, Palo Alto, CA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsOle Martin, Duesseldorf, Germany (Presenter) Nothing to DiscloseJohannes Boos, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseChristian Buchbender, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseChristina Antke, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseBenedikt M. Schaarschmidt, MD, Essen, Germany (Abstract Co-Author) Stockholder, Bayer AG; Stockholder, General ElectricCompany; Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries Ltd


[email protected]

PURPOSE

To evaluate the impact of iterative metal artifact reduction (iMAR) algorithms on Hounsfield units (HU) and standardized uptakevalues (SUV) in bright-band artefacts caused by port catheter systems.


In this prospective study, 30 oncological patients (12 female, 18 male, mean age 59.6±10.5y) with port chambers undergoingclinical indicated PET/CT on a Biograph mCT were included. An additional phantom scan was performed, consisted of a 25-Lcontainer containing a standard port catheter system in a solution of 18F-fluorodeoxyglucose (18F-FDG) and water. CT datasetswere reconstructed on a dedicated workstation using standard weighted filtered back projection (WFBP) CT and three differentiMAR algorithms (algorithm for dental filling (DF), hip and pacemaker (PM)). PET attenuation correction was performed with all fourdatasets. SUVmean and HU measurements were performed in fat and muscle tissue in the vicinity of the port chamber at thelocation with the most prominent bright band artifacts. Differences between HU and SUVmean values across all CT and PET-imageswere investigated using paired t-tests. Bonferroni correction was used to prevent alpha-error accumulation (p<0.008).

RESULTS

In comparison to WFBP (fat: 94.2 ± 53.9 HU, muscle: 197.6 ± 49.2 HU) all three iMAR algorithms led to a significant decrease of HUin bright band artifacts. iMAR-DF led to a decrease of 159.2% (fat: -51.9 ± 58.5 HU, muscle: 94.5 ± 55.3 HU), iMAR-hip of 138.3%(fat: -30.3 ± 58.5, muscle: 70.4 ± 28.8) and iMAR-PM of 122.3% (fat: -21.2 ± 47.2 HU, muscle: 72.5 ± 25.1 HU; for all p<0.008).There was no significant effect on SUVmean in bright band artifacts (SUVmean 0.48±0.23 with WFBP, 0.43±0.21 with iMAR-DF,0.46±0.22 with iMAR-hip and 0.47±0.22 with iMAR-PM). Similar results were observed for HU and SUVmean measurements in thephantom scan.

CONCLUSION

Using iMAR-CT images for attenuation correction of PET-datasets led to a significant change of HU values in artifacts caused byport catheter chambers in comparison to WFBP. However, these findings did not translate to significant changes in attenuationcorrection and consecutive differences in SUV.


iMAR reconstructions in PET/CT improve CT image quality by reducing bright band artifacts but do not influence attenuationcorrection.

ParticipantsHeok Cheow, MBBCh, MSc, Cambridge, United Kingdom (Presenter) Nothing to DiscloseBadriya Al-Saqri, MD, Cambridge, United Kingdom (Abstract Co-Author) Nothing to Disclose

NM228-SD-WEA3

Comparing TNM Staging Adequacy and Processing Time of Structured Reports versus FullySegmented and Annotated PET/CT Data of Non-Small Cell Lung Cancer

Station #3

Ines D. Harper, MBChB,FRCR, Cambridge, United Kingdom (Abstract Co-Author) Nothing to Disclose

PURPOSE

In recent year, FDG PET-CT has been advocated as the technique of choice for investigating large vessel vasculitis. The goal ofthe study is to correlate (1) the use of steroid prior PET-CT study, (2) serum inflammatory markers (CRP & ESR) with the results ofPET-CT study in order to reduce the number of negative studies.


Patients referred for FDG PET-CT investigation of large vessel vasculitis over two-year period were recruited; serum inflammatorymarkers (within two weeks window prior to PET-CT study) and the use of steroid were used to compare with the results of thePET-CT.

RESULTS

In total, 74 patients (30 men, 44 women) were identified. None of the 13 patients with positive PET-CT study was on steroid. Ofthe 61 PET-CT negative studies, 24 patients were on steroid ranging from 2mg to 60mg. All except one positive PET-CT have highCRP. 28 patients with negative study have normal CRP whereas 30 of them with raised CRP have a negative study. 9 out of 13patients with high ESR were PET-CT positive. 42 out of 52 patients with normal ESR were PET-CT negative. Combining CRP and ESRresults, 9 out of 13 patients with high CRP & ESR were PET-CT positive compared to only 1 with normal inflammatory markers. 10out of 52 patients with high CRP & ESR were PET-CT negative compared to 27 with normal inflammatory markers.

CONCLUSION

From this study, we would not advocate PET-CT study if patients were on steroid or have normal inflammatory markers. On theother hand, raised inflammatory markers do not predict a positive study.


PET-CT should be done prior starting of steroid therapy.

ParticipantsRaphael Sexauer, Basel, Switzerland (Abstract Co-Author) Nothing to DiscloseThomas Weikert, MD, Basel, Switzerland (Abstract Co-Author) Nothing to DiscloseKevin Mader, DPhil,MSc, Basel, Switzerland (Abstract Co-Author) Employee, 4Quant Ltd; Shareholder, 4Quant LtdAndreas Wicki, MD, Basel, Switzerland (Abstract Co-Author) Nothing to DiscloseBram Stieltjes, MD,PhD, Basel, Switzerland (Presenter) Nothing to DiscloseJens Bremerich, MD, Basel, Switzerland (Abstract Co-Author) Nothing to DiscloseGregor Sommer, Basel, Switzerland (Abstract Co-Author) Nothing to DiscloseAlexander Sauter, Basel, Switzerland (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Results of PET/CT examinations are usually communicated as text-based reports. These are often unstructured and may beincomplete, possibly causing staging errors. We examined to what extent text-based reports differ from a full 3D-segmentationapproach regarding the completeness of TNM information and processing time.


TNM information was extracted retrospectively from 395 text-based reports by a dual-board-certified radiologist and nuclearmedicine physician. Also the RIS time tracking of the reports was analyzed. The corresponding image data were transferred to a 3Dslicer-based software, where 2995 lesions were segmented manually using a set of 41 classification labels (TNM features +location). The time required per lesion was automatically recorded. Information content and processing time of text-based reportsand segmentations were compared using descriptive statistics and modelling.

RESULTS

The TNM/UICC stage was mentioned explicitly in only 6% (n=22) of the reports, but could be extracted in 78% (n=309) of thecases. In the remaining 22% (n=86), information was incomplete, most frequently affecting T-stage (19%, n=74), followed by N-(6%, n=22) and M-stage (2%, n=9). Full lesion segmentation required less time (median: 12.9 min) than the lowest estimator of thereporting time (dictation: median 18.1, p=.01). Tumor stage (UICC I/II: 5.2min, UICC III/IV: 20.3min, p<.001), lesion size (p<.001)and lesion count (n=1: 4.4min, n=12: 37.2min, p<0.001) correlated significantly with the segmentation time, but not with thedictation time.

CONCLUSION

A considerable number of text-based staging reports are lacking important information. A segmentation-based reporting approachtailored to the staging task can improve report quality with manageable efforts in terms of processing time.


Segmentation-based reporting can improve report quality and avoid erroneous therapy decisions based on incomplete reports.Segmented data may be used for multimedia enhancement and automatization.

NM229-SD-WEA4

To Evaluate the Role of FDG PET/CT in Patients with Cutaneous Squamous Cell Carcinoma for InitialStaging

Station #4

NM230-SD-WEA5

Implications of Empiric I-131 Activity Administration to Treat Thyroid Cancer in Patients withAbnormal Renal Function

Station #5



[email protected]

PURPOSE

Cutaneous squamous cell carcinoma (cSCC) comprises of 20-25% of non melanoma skin malignancies. CT and/or MRI is used inevaluating local extent of disease in advanced stages. The role of 2-fludeoxyglucose positron emission tomography/computedtomography (FDG PET/CT) is not clearly defined.


This was retrospective study which was conducted in patients with cutaneous SCC who presented for a period of two years.Patients who received FDG PET/CT scan at diagnosis for initial staging were evaluated for lesion detection by FDG PET/CT. Theseimages were assessed for abnormal sites of increased uptake as compared to normal adjacent background. All suspicious increaseduptake sites of uptake were noted. Standard uptake value (SUV) in primary and secondary sites was measured. Findings werecorrelated with biopsy/histopathological findings and follow up imaging and clinical data

RESULTS

Of the 22 patients screened, 13 patients with primary cSCC who had FDG PET/CT scan at diagnosis were evaluated (mean age: 76;range 60-90 years; M:F = 12:1) (pathological stage II; n=5, stage III; n=3 and stage IV; n=5). Primary sites were located in scalp(n=5), eyebrow (n=1), ear (n=1), cheek (n=3), nose (n=1), neck (n=1) and foot (n=1). All patients had positive FDG scan withareas of increased uptake . A total of 26 lesions were detected. All primary lesions (n=14) showed high FDG uptake (SUV 2.3-22.8;mean 11.6). Additional 12 FDG avid lesions were noted in 7 patients (2 cutaneous lesions in arm and cheek in a patient; 1 bonelesion, 1 lung nodule, and 8 nodes). Pathology correlation was available for 22/26 lesions; 20/22 sites were positive for malignancywhile 2/22 were benign nodes. 2/20 positive lesions were basal cell carcinoma. Remaining 4/26 FDG positive sites included 3 nodesthat were also malignant on follow up. SUV (mean ±SD; (range)) was 9.7±6.8 (2.3-22.8) for malignant lesions (n=20) and 4.5±1.3(3.5-5.4) for benign lesions (n=2). Overall sensitivity and accuracy of FDG PET/CT was 100% and 92% respectively. FDG detectedfour prior unknown secondary lesions and changed management plan in 3 of 13 patients (23%).

CONCLUSION

Thus FDG PET/CT has significant high sensitivity in detection of lesions in cutaneous SCC. It has potential role in initial staging andmanagement of primary cutaneous SCC.


Thus FDG PETCT is very sensitive modality for evaluation of cutaneous squamous cell carcinoma.

ParticipantsKenneth Nichols, PhD, New Hyde Park, NY (Presenter) Royalties, Syntermed, Inc; Maria B. Tomas, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseGene G. Tronco, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseFritzgerald Leveque, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseKuldeep K. Bhargava, PhD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseChristopher J. Palestro, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Accepted post-surgical treatment of pts with thyroid carcinoma is I-131 ablation. Empiric amounts of I-131 are administered basedon the assumption that the administered activity will deliver a dose to blood (D) below 200 cGY. In the setting of renal disease,reduced iodine clearance potentially could result in excessive red bone marrow doses. The objective of this retrospectiveinvestigation was to determine the effects of renal disease on the maximum permissible activity (MPA) of I-131 that can be safelyadministered to pts with thyroid cancer & renal disease.


One hundred pts (age = 59±16 years; 50 female pts; 50 male pts) with thyroid cancer underwent blood sample collection of 4-7days following ingestion of 1-4 mCi I-131, & whole-body I-131 scanning 48 hrs after administration of I-131. Ninteen of the 100 ptshad impaired renal function (IRF), defined as BUN & creatinine values exceeding normal limits. I-131 blood-measurement-onlydosimetry was used to compute I-131 MPA. The % of pts for whom D > 200 cGy for administered activity of 100-200 mCi I-131were tabulated.

RESULTS

MPA ranged from 106-361 mCi for the 19 patients. None would have received > 200 cGy at an empiric activity of 100 mCi, but 1 pt(5%) would have received 237 cGy at 125 mCi, 6 pts (32%) would have received 311±37 cGy at 150 mCi & 12 pts (63%) wouldhave received 277±46 cGy at 200 mCi.

CONCLUSION

Pts with impaired renal function who are candidates for treatment with 125 mCi or more of I-131 should undergo prior dosimetry to

Pts with impaired renal function who are candidates for treatment with 125 mCi or more of I-131 should undergo prior dosimetry toensure that MPA is not exceeded.


I-131 MPA should be assessed prior to thyroid cancer therapy to avoid excessive radiation dose.

NM231-SD-WEB1

Variability of Cold-Activated Brown Adipose Tissue Depends on Thermoregulatory Brain Network

Station #1

NM232-SD-WEB2

Comparing Osseous Lesions on 11C-Choline PET/CT and MRI in Prostate Cancer

Station #2

NMS-WEB

Nuclear Medicine Wednesday Poster Discussions

Wednesday, Nov. 28 12:45PM - 1:15PM Room: NM Community, Learning Center

NM


ParticipantsHelen R. Nadel, MD, Palo Alto, CA (Moderator) Nothing to DisclosePeter L. Choyke, MD, Rockville, MD (Moderator) Nothing to Disclose

Sub-Events

ParticipantsOtto Muzik, PhD, Detroit, MI (Abstract Co-Author) Nothing to DiscloseAjay Kumar, MD, PhD, Detroit, MI (Presenter) Nothing to Disclose


[email protected]

PURPOSE

The presence of cold-activated Brown Adipose Tissue (BAT) depots is highly variable in humans for unknown reasons and thisvariability might be associated with differences in CNS thermoregulatory control. As a result, central thermoregulatory brainpathways could be responsible for physiological differences observed in peripheral tissues.


A total of 20 lean subjects (10M/10F, 23.3 +/- 3.8 years, BMI = 18.5 - 24.9 kg/m2) subjects were studied using both PET/CT andfMRI. All subjects underwent 18F-fluorodeoxyglucose (FDG) PET imaging during mild cold condition in order to assess the presenceof cold-activated supraclavicular BAT. In addition, to assess functional brain responses to cold stress, all subjects underwent afunctional MRI (fMRI) scan paradigm. During fMRI, subjects were exposed to an oscillating whole body temperature challenge (five5min blocks of alternating neutral and cold exposure) applied using a specialized whole-body garment through which temperature-controlled neutral (31-34°C) or cold water (15-17°C) was circulated. The challenge was designed to induce periods of mildhypothermia interspersed by periods of return to basal core body temperature.

RESULTS

Based on FDG PET imaging, a wide variability in cold-activated BAT ranging from 0 to 182g was deterimed, characterized by abimodal distribution. Following cold exposure, 60% of the subjects showed relatively high BAT activation (~100g, High-BAT group orBAT+), whereas 40% of the subjects displayed low BAT activation (< 20g, Low-BAT group or BAT-). According to group selection,the SUVmax determined in supraclavicular BAT of the BAT+ group was significantly higher than that measured in the BAT- group(19.5 +/- 8.0 vs. 7.2 +/- 1.7; p < 0.01). Corresponding fMRI studies showed significant differences in BOLD signal between theBAT+ and BAT- groups in both cortical (right anterior insula) and subcortical (midbrain) regions.

CONCLUSION

Our data suggests that mild exposure to cold elicits differential neural responses in interoceptive regulatory centers (insula) ofsubjects with and without significant amounts of cold-activated BAT mass.


Our data provides an explanation for the observed high variability of cold-activated BAT mass observed in humans, indicating that itmight be, in part, related to different sensitivities of higher-order interoceptive brain regions to skin temperature changes.


ParticipantsStephen J. Nogel, MD, Rochester, MN (Presenter) Nothing to DiscloseAnn Packard, MD, Rochester, MN (Abstract Co-Author) Nothing to DiscloseAnil N. Kurup, MD, Rochester, MN (Abstract Co-Author) Research Grant, Galil Medical Ltd; Research Grant, EDDA Technology, Inc;Royalties, Wolters Kluwer nvMark A. Nathan, MD, Rochester, MN (Abstract Co-Author) Nothing to Disclose


[email protected]

NM233-SD-WEB3

Comparison of Different Metal Artifact Reduction (MAR) Techniques in 18F-FDG PET/CTExaminations: How Do They Impact Attenuation Correction in Metal Implants?

Station #3

PURPOSE

Patients with biochemically recurrent prostate cancer may be evaluated with 11C-Choline PET/CT. For suspicious osseous findings,they frequently undergo image-guided biopsy, which may or may not be preceded by dedicated MRI of the area of interest. Thepurpose of this study was to compare imaging findings of suspected osseous metastases from prostate cancer on 11C-CholinePET/CT and convetional MRI in patients who have underwent image-guided biopsy to determine the value of MRI and biopsy in thissetting.


After obtaining IRB approval, a list of all 11C-Choline PET/CT exams performed at our institution was obtained. Patients wereincluded if they had a 11C-Choline PET/CT and MRI prior to bone biopsy. The PET/CTs and MRIs were reviewed by two radiologistsblinded to pathology results. A third radiologist reviewed all pathology with negative results to evaluate if the biopsy accuratelysampled the suspicious lesion. Comparison was then made to pathology.

RESULTS

A total of 10,815 11C-Choline PET/CT exams were peformed between September 29, 2005 and March 31, 2017. Of those patients,60 diagnostic bone biopsy samples were identified that had correlative 11C-Choline PET/CT and MRI exams. The biopsies wereperformed in the pelvis (34), spine (23), scapula, femur and clavicle.The average pate age was 65.5 (range 47-84), averageGleason score at diagnosis was 7.9 (range 5-10) and PSA level a the time of PET ranged from undetectable to 1642. A total of 38lesions (63%) had pathology findings positive for osseous metastasis from prostate cancer, with 33 lesions (55%) having suspiciousfindings on both PET/CT and MRI. Five lesions (8.3%) positive for metastasis had isolated findings on MRI without PET/CT correlate,but no metastasis at biopsy had isolated findings on PET/CT without MRI correlated.

CONCLUSION

11C-Choline PET/CT and MRI have complementary roles in assessing osseous metastatic disease for patients with a history ofprostate cancer. All biopsy-proven osseous metastases had abnormalities on MRI.


Both 11C-Choline PET/CT and MRI are valuable tools in evaluating patients with biochemically recurrent prostate cancer followingdefinitive treatment. If 11C-Choline PET/CT findings are equivocal, MRI may be helpful for further evaluation prior to biopsy. If theMRI is negative, no biopsy is needed.

ParticipantsOle Martin, Duesseldorf, Germany (Presenter) Nothing to DiscloseJoel Aissa, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Heusch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseGerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseChristina Antke, Duesseldorf, Germany (Abstract Co-Author) Nothing to DiscloseBenedikt M. Schaarschmidt, MD, Essen, Germany (Abstract Co-Author) Stockholder, Bayer AG; Stockholder, General ElectricCompany; Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries Ltd


[email protected]

PURPOSE

To evaluate the impact of metal artifact reduction (MAR) algorithms on Hounsfield units (HU) and standardized uptake values (SUV)in bright band artefacts caused by metal implants.


In this prospective study, 25 oncological patients (13 female, 12 male, mean age 70.3±13.0y) undergoing clinical indicated PET/CTwith 32 different metal implants were included. An additional phantom scan was performed, consisted of a 25-L container containinga hip implant in a solution of 18F-fluorodeoxyglucose (18F-FDG) and water (background radiotracer: 6.5 kBq/mL). CT datasets werereconstructed using standard weighted filtered back projection (WFBP), metal artifact reduction in image space (MARIS) anditerative MAR (iMAR, using hip algorithm). PET attenuation correction was performed with all three datasets. SUVmean and HUmeasurements were performed at the site of the most prominent bright band artifacts. In the phantom scan, a total of 10measurements were performed. HU and SUVmean values across all reconstructions were investigated using paired t-tests.Bonferroni correction was used to prevent alpha-error accumulation (p<0.017).

RESULTS

In the phantom scan, in comparison to WFBP (261.4±98.8 HU), MARIS led to a non-significant mean decrease of 25.8%(197.4±74.1 HU, p>0.017), whereas iMAR-hip led to a significant decrease of 82% (48.0±26.4 HU, p<0.017). For SUVmeanmeasurements in WFBP (1.147±0.035), MARIS showed no significant effect (1.102±0.090), while iMAR-hip led to a significantdecrease of 16.4% (0.990±0.075, p<0.017). Similar results were observed in the patient scans: MARIS led to a non-significantmean decrease of 8% (378.0±331.0 HU) in comparison to WFBP (411.4±309.2 HU), whereas iMAR-hip led to a significant decreaseof 66% (138.4±195.3 HU, p<0.017). For SUVmean, MARIS showed no significant effect in comparison to WFBP (WFBP: 0.965±0.380,MARIS: 0.942±0.372), while iMAR-hip led to a significant decrease of 11.3% (0.856±0.321, p<0.017).

CONCLUSION

iMAR significantly reduces bright band artifacts caused by metal implants in CT and thus leads to a significant decrease of SUVmeanin bright band artifacts comparison to WFBP and MARIS. This has a direct effect on PET quantification adjacent to metal implants.


iMAR significantly reduces the overestimation of attenuation in bright band artifacts in PET/CT and therefore can relevantly improve

NM234-SD-WEB4

Prognostic Utility of Semi-Quantitative Metrics from Pre-Treatment F-18 FDG PET/CT of EwingSarcoma

Station #4

NM235-SD-WEB5

Clinical Features of Secondary Primary Malignancies in Patients with Differentiated Thyroid CancerTreated by Radioactive Iodine

Station #5

PET quantification in adjacent, malignant lesions.


ParticipantsAlexander Marchese, BA, Burlington, VT (Presenter) Nothing to DiscloseJanusz K. Kikut, MD, Burlington, VT (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

This retrospective study aimed to evaluate if any semi-quantitative measurements from pre-treatment Fluorine-18 (F-18)fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) as well as any additional clinical factorscould offer prognostic information and predict survival outcomes in Ewing Sarcoma (EWS) patients.


Thirteen adult and pediatric patients from a single academic medical center with histologically confirmed Ewing sarcoma wereretrospectively chart reviewed. Those who had undergone F-18 FDG PET/CT prior to initial treatment with neoadjuvantchemotherapy between January 2000 and December 2017 were included. Semi-quantitative FDG-PET/CT parameters (SUVmax,SUVpeak, metabolic tumor volume, and pixel count) of the primary lesion of each patient were recorded, as well as clinicalprognostic factors (sex, age, tumor location, tumor size, initial clinical presentation, and disease stage). Then, the overallrelationships between each factor and patients' survival times were evaluated with univariate analysis. Survival time was measuredas the time from pre-treatment FDG PET /CT scan to the date of death or to the date of the last follow-up visit in the electronicmedical record.

RESULTS

Six patients (male, 4; female, 2; mean age, 29.7 ± 23.5) were included. One patient died as a result of disease. The mediansurvival time for all six subjects was 1520 days, the median SUVmax by using PET/CT was 8.9 (range, 3.0-13.5), and the medianSUVpeak was 8.8 (range 4.9-11.70). Univariate analysis showed that patients with a SUVmax < 8.9 survived significantly longerthan those with a SUVmax > 8.9 (median survival time, 2608 vs. 946 days; p = 0.018). Similarly, patients with a SUVpeak < 8.8survived longer than those with a SUVpeak > 8.8 (median survival time, 2402 vs. 946 days; p = 0.062). Increased survival time wasalso found to be significantly related to female sex (median survival time, 2701 vs. 1059 days; p = 0.033).

CONCLUSION

Our small cohort demonstrates that pre-treatment measurement of SUVmax and SUVpeak from 18F-FDG PET/CT are potentialprognostic indicators of Ewing sarcoma.


Pre-treatment F-18 FDG PET/CT offers a non-invasive prognostic tool to identify Ewing sarcoma patients who may benefit frommore intensive therapy.

ParticipantsKunihiro Nakada, Sapporo, Japan (Presenter) Nothing to DiscloseNaoya Hattori, MD, PhD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseHiroki Sugie, MD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseNoriyoshi Katoh, MD,PhD, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseMasayuki Sakurai, Sapporo, Japan (Abstract Co-Author) Nothing to DiscloseKatsuhiko Kato, MD, PhD, Nagoya, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The relationship between secondary primary malignancy (SPM) in thyroid cancer survivors and radioactive iodine therapy (RAI)remains unclear/The aim of the study was to investigate clinical features of patients who had SPM after RAI at our institute.


Between Jun. 2007 and May. 2017, we encountered 19 patients ( Male/Female 6/11, Age 37-82 yrs. average 66.9) who developedSPM during the follow up of thyroid cancer after RAI. The cumulative dose of I-131 ranged from 30 to 450mCi. In addition to posttherapt I-131 scan, all patients underwent either of US, CT, FDG-PET or FDG-PET/CT before RAI and were followed up bycombination of imaging modalities and serum Tg test at regular interval. The findings of imaging diagnosis were reviewed by 3radiologists. SPM was confirmed by surgery or biopsy.

RESULTS

Of 19 patients, 6 had non-small cell lung cancer , 3 had breast cancer, 3 had colorectal cancer , 2 had gastric cancer , 2 had softtissue sarcoma, 2 had hematopoietic malignancies, and 1 had pancreatic cancer. The latency period from the latest RAI anddiagnosis of SPM ranged 16-178 mos. In 15 out of 19 patients (79%), SPM was incidentally detected by follow-up imaging studies

NM148-ED-WEB6

Immunotherapy in Oncology: Image Evaluation of the Adverse and Toxic Effects of CheckpointInhibitors

Station #6

before clinical manifestation. In contrast, 4 patients (21%) had SPM at the time of RAI, which was later confirmed within 3 yearsafter RAI. Eight patients (42%) had specific risk factors for SPM such as diabetes, smoking, drinking alcohol, dietary habits or familyhistory. None of the patients had extremely high I-131 uptake in the organs where SPM occurred on the post therapy I-131scintigraphy. During the follow-up period, 8 (42%) patients died of SPM while 3(16%) died of thyroid cancer. The remaining 8 arealive with or without SPM.

CONCLUSION

It is highly unlikely that SPMs in majority of our subjects were substantially linked with RAI. Detailed information regarding strategiesthat detected SPM, stage of SPM at diagnosis, presence of potential risk factor for SPM, and cumulative radiation dose in thetarget organ from RIA and other imaging modalitirs should be taken into consideration to discuss relationship between RAI and SPMin thyroid cancer survivors.


In assessing long term potential risks of RAI, it is important to identify clinical features of SMP in patients with thyroid cancerunderwent RAI, which cannot be evaluated by retrospective analysis of registered database

ParticipantsDaniela F. Vieira Vendramini, Sao Paulo, Brazil (Presenter) Nothing to DiscloseFelipe D. Barbosa, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMarcelo A. Queiroz, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJose F. Marin, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseMariana Ruiz, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseFelipe L. Costa, MD , Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseNicolau F. Guerreiro, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseJucelio P. Moura Filho, BMedSc, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseCarlos A. Buchpiguel, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to DiscloseRafael F. Nunes, MD, Sao Paulo, Brazil (Abstract Co-Author) Nothing to Disclose

TEACHING POINTS

Throughout this study, we will review the main alterations and adverse effects related to immunotherapy using computedtomography (CT), magnetic resonance imaging (MRI) and PET / CT images.The purpose of this exhibition is:• To briefly review themechanisms of action of checkpoint inhibitors.• To recognize the imaging findings of the main adverse effects related toimmunotherapy.• To review the clinical aspects and the course of action to be taken against these adverse effects.


Schematic drawings and illustrations of the mechanism of action of checkpoint inhibitors.• Cases illustrated by images of the mainadverse effects that will be illustrated:• CT, MRI and PET / CT images demonstrating the side effects and follow up with imagingafter suspension, treatment with immunotherapy.• Tables summarizing clinical, laboratory, and behavioral findings.• Discussion ofnew concepts such as hyperprogression of disease in the presence of immunotherapy

MSRO44

BOOST: Advanced Techniques in Image-guided Therapy (Interactive Session)

Wednesday, Nov. 28 3:00PM - 4:15PM Room: S103CD

CT MR NM RO


ParticipantsTheodore S. Hong, MD, Boston, MA (Presenter) Nothing to DiscloseSusanna I. Lee, MD,PhD, Boston, MA (Presenter) Editor, Wolters Kluwer nvHomer A. Macapinlac, MD, Houston, TX (Presenter) Nothing to DisclosePeter Balter, PhD, Houston, TX (Presenter) Research Grant, Varian Medical Systems, Inc; Research Grant, RaySearch LaboratoriesAB


[email protected]

LEARNING OBJECTIVES

1) Explain and apply modern CT and MR imaging technologies and PET tracers for treatment planning of solid malignancies in thechest, abdomen and pelvis. 2) Explain and apply the modern techniques in radiotherapy safely and effectively in the chest,abdomen and pelvis.

ABSTRACT

The last decade has seen emergence of important advances in locoregional cancer therapy. Use of functional imaging andadvanced radiatiotherapy often integrated with targeted chemotherapy have improved patient outcomes. This course will presentthe underlying principles in diffusion MRI, novel MR contrast agents, PET-MR and dual energy CT. PET tracers to be discussed areF-18 FDG, widely used for most solid tumors; C-11 choline/F-18 Fluciclovine for prostate cancer and Ga-68-DOTATATE forneuroendocrine tumors. Advanced radiotherapy techniques such as Image Guided Radiotherapy (IGRT), Intensity ModulatedRadiation Therapy (IMRT), and Stereotactic Body Radiation Therapy (SBRT) using image guidance with X-ray, CT, MRI and PET willbe described.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Susanna I. Lee, MD,PhD - 2013 Honored Educator

SSM16-01 Complementary Gadoxetic Acid-Enhanced MRI in Addition to 18F-DOPA-PET/CT Improves LiverStaging in Patients with Medullary Thyroid Carcinoma


SSM16-02 Radiotheranostics for Regionally Advanced and Metastatic Differentiated Thyroid Cancer: OutcomesFollowing Initial Treatment Strategy Informed By Diagnostic 131-I Scintigraphy with SPECT/CT

SSM16

Nuclear Medicine (Thyroid/Parathyroid Imaging and Therapy)


CT HN NR NM



ParticipantsDon C. Yoo, MD, E Greenwich, RI (Moderator) Consultant, Endocyte, IncBrian M. Rodgers, MD, New Orleans, LA (Moderator) Nothing to Disclose

Sub-Events

ParticipantsDaniel Puhr-Westerheide, MD, Munich, Germany (Presenter) Nothing to DisclosePhilipp M. Kazmierczak, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseAxel Rominger, Munich, Germany (Abstract Co-Author) Nothing to DiscloseMatthias Brendel, Munich, Germany (Abstract Co-Author) Nothing to DiscloseWolfgang G. Kunz, MD, Munich, Germany (Abstract Co-Author) Grant, Medtronic plcClemens C. Cyran, MD, Munich, Germany (Abstract Co-Author) Research Grant, iThera Medical GmbH Speakers Bureau, Siemens AGRobert Stahl, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseJosef Sargsyan-Bergmann, Munich, Germany (Abstract Co-Author) Nothing to DiscloseChristine Spitzweg, Munich, Germany (Abstract Co-Author) Advisory Board, Swedish Orphan Biovitrum AB; Advisory Board, BayerAG; Advisory Board, Eisai Co, Ltd; Advisory Board, AstraZeneca PLC; Speaker, Swedish Orphan Biovitrum AB; Speaker, Bayer AG;Speaker, Eisai Co, Ltd; Speaker, AstraZeneca PLCMaximilian F. Reiser, MD, Munich, Germany (Abstract Co-Author) Nothing to DiscloseJens Ricke, MD,PhD, Munich, Germany (Abstract Co-Author) Research Grant, Sirtex Medical Ltd Research Grant, Bayer AG


[email protected]

PURPOSE

To investigate the additional diagnostic value of gadoxetic acid-enhanced MRI to 18F-DOPA-PET/CT for liver staging in patientswith medullary thyroid carcinoma (MTC).


41 consecutive patients with histologically confirmed MTC who underwent gadoxetic acid-enhanced MRI and 18F-DOPA-PET/CTwithin one month between 2010 and 2015 were selected for this retrospective study. The 18F-DOPA-PET/CT and multiparametricgadoxetic acid-enhanced MRI data sets were analyzed by two blinded radiologists. A 5-point Likert scale (based on the LI-RADScriteria: 1-definitely benign, 2-probably benign, 3-intermediate risk for metastasis, 4-probably metastasis, 5-definitely metastasis)was used for lesion categorization in both modalities. The additional value of MRI was defined as detection of 18F-DOPA-PET/CT-occult category 5 lesions or a definitive categorization (category 1 or 5) of lesions remaining inconclusive on the 18F-DOPA-PET/CTscan.

RESULTS

We categorized a total of 212 liver lesions (166 lesions on 18F-DOPA-PET/CT, 212 lesions on MRI; 165 metastases, 37 cysts, 18hemangiomas). Out of 165 category 5 lesions on MRI, only 94 were classified as category 5 lesions on 18F-DOPA-PET/CT. In 65 %(30/46) of inconclusively categorized lesions on 18F-DOPA-PET/CT (category 2-4), a definitive lesion classification was possiblewith MRI (change in categorization to categories 1 or 5, respectively). A change in lesion classification by MRI was made in 12patients (lesions with a change in category from 2 to 1: n=10; from 3 to 1: n=3; from 4 to 5: n=12; from 3 to 5: n=5).

CONCLUSION

Complementary liver-specific MRI allows for the detection of 18F-DOPA-PET/CT-occult metastases and optimizes liver lesionclassification in MTC patients.


The definitive categorization of detected liver lesions and timely identification of liver metastases in MTC patients is essential inguiding treatment decisions on early surgical or interventional management.


SSM16-03 I-131 Thyroid Dosimetry in Patients with Lung Metastases


ParticipantsAnca Avram, MD, Ann Arbor, MI (Presenter) Nothing to DiscloseNatalja Rosculet, BS, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseNazanene H. Esfandiari, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DisclosePaul G. Gauger, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseBarbra S. Miller, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseMark Cohen, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseKirk A. Frey, MD, PhD, Ann Arbor, MI (Abstract Co-Author) Consultant, MIM Software Inc; Stockholder, General Electric Company;Stockholder, Johnson & Johnson; Stockholder, Novo Nordisk AS; Stockholder, Bristol-Myers Squibb Company; Stockholder, Merck &Co, Inc; David T. Hughes, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Diagnostic staging 131-I scans with SPECT/CT (Dx scan) guide patient-individualized 131-I therapy for differentiated thyroid cancer(DTC). The objective of this study was to determine dynamic risk stratification outcomes after surgery and activity-adjusted 131-Itherapy informed by diagnostic 131-I scintigraphy with SPECT/CT.


Single-institution retrospective cohort study analysis of clinical outcomes in 350 patients with DTC associated with histopathologicrisk factors, nodal metastases, and/or distant metastases treated at University of Michigan. Post-operatively all patients underwentDx 131-I SPECT/CT scans for completion of staging and risk stratification. 131-I therapy was based on integration of informationfrom histopathology, stimulated thyroglobulin (Tg) and scintigraphy. The patients were followed for 1-5 years (mean 39.6 ±23.4months)

RESULTS

23 patients (6.6%) underwent re-operative neck dissection for removal of unsuspected residual nodal metastases identified on Dxscans. Dynamic risk stratification outcomes were: 84.3% complete response; 1.4% biochemical incomplete response; 2.3%indeterminate response and 12% structural incomplete response. Of the entire cohort only 8 patients (2.3%) had persistent iodine-avid metastatic disease which required repeated 131-I therapy. Of 31 patients with iodine-avid distant metastases identified on Dxscans, 13 patients (42%) achieved complete response with a single 131-I treatment.

CONCLUSION

Detection of regional and distant metastases on postoperative Dx scans permits adjustment of prescribed 131-I activity fortargeted treatment, as compared to fixed-activity ablation. This approach resulted in complete response after a single 131-Itreatment in 88% patients with histopathologic risk factors and regional metastases, and 42% patients with distant metastases.Most patients (81%) with structural incomplete response had negative follow-up 131-I scans and positive PET/CT and/or CT scansconsistent with altered tumor biology (non-iodine avid disease).


Postoperative Diagnostic 131-I scans with SPECT/CT are useful for identification of regional and distant metastases in thyroidcancer, informing 131-I therapy decisions. Elimination of iodine-avid regional and distant metastases with complete therapeuticresponse is achieved in the majority of patients after a single 131-I treatment based on radiotheranostics principles.

ParticipantsKenneth Nichols, PhD, New Hyde Park, NY (Presenter) Royalties, Syntermed, Inc; Maria B. Tomas, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseGene G. Tronco, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseSagine A. Berry-Tony, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseKuldeep K. Bhargava, PhD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseChristopher J. Palestro, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To protect bone marrow from excessive radiation, maximum permissible activity (MPA) of I-131 to treat thyroid cancer is the valuethat limits absorbed dose to blood (as a surrogate of marrow) to < 200 cGy. Pts with thyroid carcinoma pulmonary metastasespotentially are a unique subgroup because I-131 uptake in the lungs could result in apparently accelerated blood clearance, whichin turn could lead to an overestimation of MPA. The objective of this investigation was to test the hypothesis that in pts withdiffuse lung metastases from thyroid carcinoma, MPA based on blood measurements alone is not affected by the presence of diffuselung metastases.


Data were analyzed retrospectively for 87 thyroid cancer pts (60±15 yrs; 45 female; 42 male) referred for determination of MPAprior to I-131 treatment. Method1 for determining MPA computed total absorbed dose to blood (DTotal) as the sum of mean whole-body γ ray dose component (Dγ) from un-collimated gamma-camera measurements, along with dose due to ß emissions (Dß) fromblood samples. Method2 estimated DTotal from Dß alone, using linear regression to associate in-vitro blood sample measurements

SSM16-04 Influence of Age on Multivariate Analysis of Disease Specific Survival in Differentiated Thyroid Cancer


SSM16-05 Feasibility of Parathyroid Adenoma Localization with Fluciclovine (18F) PET-CT


alone to conventional Method1 DTotal. MPA was computed as 200 cGy/DTotal for each DTotal estimate for Method1 & Method2. Allpts also underwent whole body imaging 48 hrs after I-131 administration.

RESULTS

Six pts had iodine avid diffuse lung metastases. MPA values were similar for Method1 & Method2 for all pts (14.3±8.9 versus14.1±8.7 GBq, p = 0.34), pts with lung metastases (12.4±6.9 versus 11.7±6.4 GBq, p = 0.06) & for pts without lung metastases(14.4±9.0 versus 14.3±8.8 GBq, p = 0.52). MPA values were similar for pts with lung metastases & pts without lung metastases(12.4±6.9 versus 14.4±9.0 GBq, p = 0.59). Correlations between Method1 & Method2 MPA values were similar for all pts (r = 0.990,p < 0.0001), those with lung metastases (r = 0.999, p < 0.0001), & those without lung metastases (0.989, p < 0.0001).

CONCLUSION

Our data suggest that in pts with iodine avid diffuse lung metastases from thyroid carcinoma, MPA can be accurately estimated bymeasuring I-131 blood clearance alone without the need to perform un-collimated gamma camera whole body counting.


MPA can be estimated by I-131 blood clearance alone in pts with lung metastases from thyroid carcinoma.

ParticipantsJasna Mihailovic, MD, 21204 Sremska Kamenica, Serbia (Presenter) Nothing to DiscloseEmil Matovina, MD, Sremska Kamenica, Serbia (Abstract Co-Author) Nothing to DiscloseJelena Roganovic, MD, Sremska Kamenica, Serbia (Abstract Co-Author) Nothing to DiscloseVeljko Crnobrnja, MD, Novi Sad, Serbia (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The aim of the study was to analyze the influence of age (<45 years versus >=45 years) on disease-specific survival (DSS) and itsprognostic factors in patients with differentiated thyroid cancer (DTC).


569 DTC patients were treated with I-131 (RAI) in our institution from 2001 to 2010. We analyzed DSS and its predicting factors inall 569 patients as well as in different age groups (Group I<45 years, 237 patients and Group II: >=45 years, 332 patients) byKaplan-Meier's method. Statistical significance of differences was tested by Log rank test.

RESULTS

There were 185 (32.51%) high risk and 378 (66.43%) low-risk patients, while T was not defined in 5 (0.88%) patients; 132 (23.2%)males, 437 (76.8%) females; 57 (10%) follicular and 465 (81.7%) papillary carcinomas, while histology was not defined in 3 (0.53%)patients. Initial regional metastases were present in 202 (35.5%) patients. DSS was 96.5%; 93.5%; 87%, and 69.6% after 5,10,15and 17 years, respectively. Prognostic factors that significantly influenced DSS were: gender (p=0.003), age (p=0.0001), T stage(p=0.02), initial metastases (p=0.0001), histology (p=0.039), type of initial treatment (p=0.01), while number of RAI course did notinlfuence the survival (p=0.087). In Group I, DSS after 5, 10 and 15 years was 99.1%. Prognostic factors that significantlyinfluenced DSS in this group were: initial metastases (p=0.015) and histology (p=0.007), while gender, type of initial therapy, Tstage, and number of RAI courses had no significant influence (p=0.89; p=0.143; p=0.969; p=0.504, respectively). In Group II, DSSafter 5 years was 94%, after 10 years = 89.3% and after 15 years = 78%. Strong predicting factors were: gender (p=0.0001),initial metastases (p=0.0001), type of initial therapy (p=0.028), and number of RAI courses (p=0.031), while histology and T stagehad no influence to DSS (p=0.275; p=0.101, respectively).

CONCLUSION

DSS prognostic factors differ among age groups. Young patients show longer DSS significantly influenced by the presence of initialmetastases and histology. In contrast, elderly patients have shorter DSS with significant influence of gender, type of initialtherapy, presence of initial metastases and number of RAI courses.


In comparison to elderly, young DTC patients have longer DSS that is influenced by different prognostic factors.


ParticipantsAkinyemi A. Akintayo, MD, Atlanta, GA (Presenter) Nothing to DiscloseOlayinka A. Abiodun-Ojo, MD,MPH, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseCollin Weber, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseJoe Sharma, MD, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseCynthia Cohen, Atlanta, GA (Abstract Co-Author) Nothing to DiscloseRaghuveer K. Halkar, MD, Atlanta, GA (Abstract Co-Author) Research Grant, General Electric Company Research Grant, GileadSciences, Inc Royalties, General Electric Company Mark M. Goodman, PhD, Atlanta, GA (Abstract Co-Author) Royalties, Nihon Medi-Physics Co, LtdDavid M. Schuster, MD, Decatur, GA (Abstract Co-Author) Institutional Research Grant, Nihon Medi-Physics Co, Ltd; InstitutionalResearch Grant, Blue Earth Diagnostics Ltd; Institutional Research Grant, Advanced Accelerator Applications SA; Consultant,

SSM16-06 Evaluating the Role of Tc99m Sestamibi Scan in Parathyroid Surgery: A 10-Year InstitutionalExperience


Syncona Ltd; ; ;


[email protected]

PURPOSE

Localization of parathyroid adenomas is essential for surgical planning. Uptake of fluciclovine (18F) by parathyroid cells has beenreported in-vitro. We conducted a feasibility pilot study to evaluate the ability of fluciclovine PET-CT in comparison to 99m Tcsestamibi SPECT-CT for localization of parathyroid adenomas in patients with hyperparathyroidism.


Four patients with hyperparathyroidism underwent PET-CT of the neck and upper mediastinum for 60 mins after IV injection offluciclovine (351.5±28MBq). SUVmax of the adenomas, and target-to-background ratios (TBR) (SUVmax/SUVmean) utilizing bloodpool (aortic arch), muscle (pectoralis) or thyroid gland as the background were compared with TBRs of the 45-minute sestamibi(798.3±26.8MBq)SPECT-CT in the same patient. Surgical confirmation was reference standard for truth.

RESULTS

Both modalities had 100% concordance in the localization of the adenomas at surgery (n=4). Mean SUVmax (±SD) of the adenomaswere 3.5±1.67, 2.6±1.17, 2.1±0.81, 2.1±0.95, and 2.2±1.03 at 5, 10, 20, 35 and 50 minutes, respectively. The highest fluciclovineSUVs and TBRs were at the 5 min time-point with rapid washout. Sestamibi had significantly higher TBRs compared with fluciclovine(5 min) for blood pool (10.9±4.70 vs 1.3±0.57; p<0.01) and pectoralis muscle (5.8±3.01 vs 1.7±0.55; p<0.01) with a non-significant trend for thyroid gland (1.3±0.48 vs 1.1±0.45; p=0.73), respectively. After surgical resection of the adenomas, thepost-operative parathyroid hormone levels decreased to normal. A 2x2 mm adenoma at surgery was not detected by eithermodality.

CONCLUSION

In this small series we found that parathyroid adenomas can be detected on fluciclovine PET-CT at early imaging but with rapidwashout. In addition, adenoma conspicuity (TBR) was better with sestamibi compared with fluciclovine. Fluciclovine PET-CT doesnot seem promising in the detection and localization of parathyroid adenomas compared with sestamibi SPECT-CT. Studies inpatients with negative sestamibi scans are encouraged.


Parathyroid adenomas can be detected on fluciclovine PET-CT at early imaging; however, adenoma conspicuity (TBR) seems betterwith sestamibi compared with fluciclovine.

ParticipantsParul Mohan, MBBS, MD, New Delhi, India (Abstract Co-Author) Nothing to DiscloseHarsh Mahajan, MD, MBBS, New Delhi, India (Presenter) Nothing to DiscloseS Bal SR, MBBS,MS, New Delhi, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

In this study we aim to evaluate the (99mTc) sestamibi parathyroid scan as a single localizing modality, and we also assess itsrelation to the weight of the gland and to the preoperative parathyroid hormone (PTH) levels.


We reviewed 744 patients from our hospital from 2007 to 2017, with a mean age of 56.6 years and a female to male ratio of 3.3:1.With primary hyperparathyroidism, all of them had (99mTc) sestamibi parathyroid scan for the localization of the parathyroidadenoma. Preoperative and postoperative PTH levels were recorded. The histopathology reports confirmed the diagnosis and weightof the diseased gland, which were recorded every time. The results were analyzed and correlated with the sestamibi results, toevaluate its accuracy.

RESULTS

506 patients (68%) of the 744 had an exact match (EM) sestamibi results, 227 (30.5%) had a partial match, and only 11 patientswere reported as mismatch. Analyzing the mean weight of the gland in each group between matching (EM, PM) versus mismatchresulted in a mean difference of 0.823 g (1.05 and 0.247 g, respectively) P = 0.045. Hyperplasia to adenoma ratio was more in thepartial matching group (18.5%) versus the exact matching group (7.6%). Finally the mean PTH level was higher in the EM group(36.8 pmol/L) compared to the mismatch group (10.1 pmol/L) P = 0.02. Overall sensitivity and specificity for the (99 mTc) sestamibiin our data was 98.1 and 97%, respectively.

CONCLUSION

(99mTc) sestamibi is a highly accurate test that can be employed as a single localizing modality for identifying a hypersecretingparathyroid, a parathyroid adenoma, or a parathyroidectomy. The weight of the gland plays an important role in the accuracy ofthe test, as also the preoperative PTH levels.


(99mTc) sestamibi is a highly accurate test that can be employed as a single localizing modality for identifying a parathyroidadenoma.

ED010-TH

Nuclear Medicine Thursday Case of the Day

Thursday, Nov. 29 7:00AM - 11:59PM Room: Case of Day, Learning Center



TEACHING POINTS


SPSH50A New PET Technologies and Acquisition Approaches

SPSH50B Molecular Imaging of Heart Diseases

SPSH50C Non-FDG PET Tracers for Molecular Brain Imaging

SPSH50D Molecular Imaging of Cancer: Where Are We Going?

SPSH50

Hot Topic Session: Beyond FDG: Advancing PET Imaging of the Human Disease

Thursday, Nov. 29 7:15AM - 8:15AM Room: E353A

CA MI NR NM



ParticipantsChadwick L. Wright, MD,PhD, Lewis Center, OH (Moderator) Nothing to DiscloseKatherine A. Zukotynski, MD, Ancaster, ON (Moderator) Nothing to Disclose

LEARNING OBJECTIVES

1) To highlight topics related to advances in Cardiovascular PET, Neuro PET and Oncologic PET with FDA approved radiotracersother than FDG. 2) To address myocardial perfusion and atherosclerosis imaging, amyloid imaging, and oncologic imaging.

Sub-Events

ParticipantsMichael V. Knopp, MD, PhD, Columbus, OH (Presenter) Nothing to Disclose

ParticipantsSharmila Dorbala, MD,MPH, Boston, MA (Presenter) Research Grant, Astellas Group

LEARNING OBJECTIVES

1) List clinically available novel PET radiotracers for imaging cardiovascular diseases. 2) Discuss emerging cardiac applications usingradiotracers targeting amyloid fibrils, somatostatin receptors and microcalification.

URL

ParticipantsSatoshi Minoshima, MD, PhD, Salt Lake City, UT (Presenter) Consultant, Hamamatsu Photonics KK; Research Grant, Hitachi, Ltd;Research Grant, Nihon Medi-Physics Co, Ltd;

ParticipantsPeter L. Choyke, MD, Rockville, MD (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Familiarize participants with several new promising PET agents for cancer imaging. 2) Discuss new technologies that will enablethe development of similar agents in the near future. 3) Describe opportunities and barriers to broader use of Molecular Imaging inCancer.

ABSTRACT

Several new PET agents have been developed that promise to revolutionize the way cancer is diagnosed. Agents targeting thesomatostatin receptor for neuroendocrine tumors and PSMA for prostate cancers are changing the way these diseases aremanaged. However, these agents took a long time to develop and even now are not fully available. New small molecule discoverytechnologies promise to greatly speed up the development of future agents. Many of these are also compatible with targetedradionuclide therapy. The future of this field is exciting and there is much work to be done.

RC611A Advances in Cardiac SPECT

RC611B Advances in Cardiac PET

RC611

Advances in Cardiac Nuclear Imaging: SPECT/CT and PET/CT

Thursday, Nov. 29 8:30AM - 10:00AM Room: S504CD

CA CT NM



LEARNING OBJECTIVES

1) Understand the technical advancements associated with new scintillation cameras and SPECT-CT and PET-CT cameras. 2)Appreciate the benefits of CT attenuation correction. 3) Appreciate the adjunctive benefits of anatomic definition provided with CTand physiologic/function information provided by SPECT and PET. 4) Improve interpretive skills related to SPECT and PET-CT.

ABSTRACT

Camera and software technology recently has rapidly advanced, providing improved SPECT image resolution and increased countingstatistics. These advancements in turn have provided the possibility of reduced-time and reduced radiopharmaceutical dose imageacquisitions. Moreover, increased flexibility in imaging protocols has been realized. Future development of these methods holdpromise in increasing diagnostic accuracy and expanding diagnostic applications. The addition of CT to SPECT and PET has affordedthe ability to perform attenuation correction, thereby minizing attenuation artifacts and increasing diagnostic specificity. With CTacquisitions of sufficient resolution, complementary anatomic diagnostic information is provided. In addition, more precise anatomiclocalization of SPECT and PET abnormalities significantly increases clinical applicability.

Sub-Events

ParticipantsE. Gordon Depuey, MD, New York, NY (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Implement protocols that facilitate patient-centered imaging and that reduce patient radiation exposure. 2) Understand softwaremethods to cope with lower SPECT counting statistics in order to reduce scan acquisition time and/or radiopharmaceutical injectedactivity and their clinical impact. 3) Understand instrumentation advances that allow new cameras to perform SPECT with markedlyreduced acquisition times and/or less radiopharmaceutical activity and their clinical impact. 4) Review myocardial perfusion SPECTscans systematically to avoid artifacts and maximize diagnostic accuracy.

ABSTRACT

There has been an intersocietal effort to promote patient-centered imaging with a focus on appropriateness guidelines, cost-containment, radiation dose reduction, and the selection of the most appropriate imaging test and protocol to suit particular patientneeds. The following technical advancements described facilitate implementation of patient-centered imaging. New softwaremethods and new innovative hardware now allow for significantly shortened SPECT acquisition times without a decrease in imagequality. Advancements include iterative reconstruction, resolution recovery, and noise reduction software, and focused collimationand solid state detectors incorporated into new camera designs. Attenuation correction increases diagnostic specificity andfacilitates stress-only protocols. Software advancements such such as high resolution imaging, scatter correction, and respiratorygating increase diagnostic sensitivity. Even with such technical advancements, however, attention to technical detail is essentialto assure optimal image quality. Camera and radiopharmaceutical quality control deserve the highest priority. A systematic reviewof myocardial perfusion SPECT images is essential to recognize artifacts and optimize diagnostic accuracy. Case examples will bepresented to reinforce this approach.

ParticipantsSharmila Dorbala, MD,MPH, Boston, MA (Presenter) Research Grant, Astellas Group

LEARNING OBJECTIVES

1) Review the advantages and disadvantages of myocardial perfusion PET compared to SPECT for evaluation of coronary arterydisease. 2) Learn the added value of absolute quantitative parameters derived from PET for assessment of coronary artery disease.3) Discuss novel clinical applications of cardiovascular PET imaging in systemic diseases 4) Review Case Examples of Cardiac PETs

ABSTRACT

Advances in PET detectors, radiotracer availability, clinical software, as well as hybrid PET/CT and PET/MR scanners haverevolutionized the clinical and investigative applications of cardiac PET. Cardiac PET myocardial perfusion imaging, in the 1970's,was a predominantly investigative tool, with home-grown software, available at select major academic centers with access to acyclotron. Over the last decade, with easy access to PET scanners, and to positron emitting perfusion tracers, the use of cardiacPET has exploded -well beyond major academic centers to several hospitals and to large office-based practices. Robust clinicalevidence coupled with commercially available software has made quantitative myocardial blood flow assessment, a main-streamclinical application. Hybrid PET/CT scanner applications- calcium score and CT based coronary angiography-have further advanced

RC611C Cases, Clinical Examples-Panel: How to Build Practice (Both PET and SPECT)

the applications of cardiac PET. A growing body of recent literature supports the role of targeted molecular PET to imageinflammatory, infectious and infiltrative heart diseases. PET/MR is an emerging technology with promising cardiovascularapplications. Each of these exciting developments has transformed cardiac PET from a predominantly investigative tool of the1970's to the current advanced clinical tool. The primary goal of this session is to discuss the present-day clinical and emergingapplications of cardiac PET/CT and PET/MR using a practical case-based approach.

ParticipantsE. Gordon Depuey, MD, New York, NY (Presenter) Nothing to DiscloseSharmila Dorbala, MD,MPH, Boston, MA (Presenter) Research Grant, Astellas Group

LEARNING OBJECTIVES

1) Interpret cardiac SPECT and PET scans with optimal sensitivity and specificity. 2) Recognize technical and patient-relatedartifacts. 3) Characterize myocardial perfusion defects whereby patients can be risk stratified with regard to risk of future cardiacevents. 4) Formulate reports in a clinically relevant manner.

RC615A PET and PET/MRI

RC615B Molecular Breast Imaging

RC615C Contrast Enhanced Mammography & Tomosynthesis

RC615

Advanced Breast Imaging Technologies

Thursday, Nov. 29 8:30AM - 10:00AM Room: E451A

BR MR NM



ParticipantsMaxine S. Jochelson, MD, New York, NY (Moderator) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

I will put my learning objectives in under my course itself and presume the other 2 presenters will do the same. Don't think thereneeds to be a separate learning objective for the moderator?

Sub-Events

ParticipantsAmy M. Fowler, MD, PhD, Madison, WI (Presenter) Research support, General Electric Company


[email protected]

LEARNING OBJECTIVES

1) Describe current approaches for performing breast PET imaging. 2) Assess diagnostic performance of breast PET imaging forextent of disease and therapy response evaluation. 3) Examine potential uses of PET/MRI for breast imaging.

ParticipantsCarrie B. Hruska, PhD, Rochester, MN (Presenter) Institutional license agreement, CMR Naviscan Corporation


[email protected]

LEARNING OBJECTIVES

1) Describe MBI instrumentation and clinical protocol for low-dose imaging. 2) Assess performance of MBI in screening of womenwith dense breasts. 3) Examine the potential role of MBI as an imaging biomarker of breast cancer risk.

ParticipantsMaxine S. Jochelson, MD, New York, NY (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) The utility of tomosynthesis in the screening setting. 2) The technique and risks of Contrast Enhanced Mammography. 3)Potential uses for Contrast Enhanced Mammography in the screening and diagnostic setting.

RC617A Introduction to Imaging in Prostate Cancer

RC617B Next Generation Prostate MRI

RC617C Molecular Prostate Imaging: Chemistry to Clinic

RC617D Hyperpolarized C-13 MR Molecular Imaging of Prostate Cancer

RC617

Emerging Technologies: Prostate Cancer Imaging & Management - Update 2018

Thursday, Nov. 29 8:30AM - 10:00AM Room: S505AB

GU MI MR NM



ParticipantsPeter L. Choyke, MD, Rockville, MD (Moderator) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Understand current issues in prostate cancer relevant to imaging. 2) Understand the role of emerging technologies in the imagingand management of prostate cancer.

ABSTRACT

Prostate cancer is a major health issue. Imaging has made great strides in the last decade including the use of multiparametric MRI,MR-ultrasound fusion biopsies and most recently PET scanning. This refresher course explores emerging technolgies in prostatecancer imaging and management.

Sub-Events


LEARNING OBJECTIVES

1) Understand the impact of new screening guidelines on imaging of prostate cancer. 2) Understand the issues facing clinicianstreating prostate cancer.

ABSTRACT

This talk will review the current status of screening for prostate cancer and how stage migration is beginning to be seen. Theproblems of early detection, early recurrence and early metastases will be discussed. This talk will serve as a starting off point forthe subsequent talks on new technologies.

ParticipantsBaris Turkbey, MD, Bethesda, MD (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Understand current status and uses of multi-parametric MRI. 2) Understand role of MRI in assessment of prostate canceraggressiveness and tumor heterogeneity. 3) Understand role of computer aided diagnosis systems in evaluation of prostate canceraggressiveness and tumor heterogeneity.

ParticipantsMartin G. Pomper, MD, PhD, Baltimore, MD (Presenter) Researcher, Progenics Pharmaceuticals, Inc; License agreement, ProgenicsPharmaceuticals, Inc; Researcher, Advanced Accelerator Applications SA; License agreement, Advanced Accelerator ApplicationsSA; Co-founder, Cancer Targeting Systems, Inc; Board Member, Cancer Targeting Systems, Inc; Researcher, Celgene Corporation,Inc; License agreement, Celgene Corporation, Inc; Co-founder, Neurly; Board Member, Neurly; Co-founder, TheralyPharmaceuticals, Inc; Board Member, Theraly Pharmaceuticals, Inc

LEARNING OBJECTIVES

View learning objectives under the main course title.

ParticipantsDaniel B. Vigneron, PhD, San Francisco, CA (Presenter) Research Grant, General Electric Company;

LEARNING OBJECTIVES

RC617E Radionuclide Therapy for Prostate Cancer

1) To describe the basic principles and techniques used in hyperpolarized carbon-13 MRI. 2) Understand the cellular metabolicreprogramming that occurs in prostate cancer. 3) Demonstrate the changes in pyruvate to lactate conversion that are observed inprostate cancer and differences with cancer aggressiveness and response to therapy.


LEARNING OBJECTIVES

View learning objectives under main course title.

SSQ05-01 Assessment of Changes in Regional Xenon Ventilation, Perfusion, and Ventilation-Perfusion MismatchUsing Dual-Energy Computed Tomography after Pharmacological Treatment in Patients with COPD

Thursday, Nov. 29 10:30AM - 10:40AM Room: E353B

SSQ05-02 Quantitative Assessment of Emphysema Heterogeneity in Patients with Lung Cancer Using VolumetricChest CT

SSQ05

Chest (Functional Lung Imaging/Dual-Energy CT/Radiation Dose Reduction)

Thursday, Nov. 29 10:30AM - 12:00PM Room: E353B

BQ CH CT MR NM OI SQ



ParticipantsJeffrey B. Alpert, MD, New York, NY (Moderator) Nothing to DiscloseMannudeep K. Kalra, MD, Boston, MA (Moderator) Research Grant, Siemens AG; Research Grant, Canon Medical SystemsCorporation

Sub-Events

ParticipantsHye Jeon Hwang, MD,PhD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseSang Min Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJoon Beom Seo, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseNamkug Kim, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Stockholder, Coreline Soft, Co LtdHee J. Park, MS, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSang Min Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSang Young Oh, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJae Seung Lee, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To assess the changes of regional ventilation (V) and perfusion (Q) status in COPD patients after pharmacologic treatment usingcombined xenon-enhanced V and iodine-enhanced Q dual-energy CT (DECT).


Combined V and Q DECT were performed at baseline and after 3-month pharmacologic treatment in fifty-three COPD patients.Virtual noncontrast images, V and Q maps were anatomically co-registered with in-house software. Normalization of V and Q valuesof each pixel were performed. For visual analysis, V/Q pattern was determined to be matched, mismatched, or reversed-mismatchedand compared with the regional disease patterns-emphysema, bronchial wall thickening, or normal lung-in each segment in baselineand follow-up. Mean V, Q, and V/Q values, standard deviation of V/Q (V/QSD), and proportions of lung area with reversed-mismatch (Rev), mismatch (Mis) and match (Mat) of each patient were quantified and compared with pulmonary function test (PFT)parameters in baseline and follow-up. Changes of quantified CT parameters and PFT results between baseline and follow-up werecompared.

RESULTS

Most of segments showed a matched V/Q, whereas about thirty percent of segments with bronchial wall thickening showed areversed-mismatched V/Q. On follow-up, V/Q pattern did not change in most of segments with matched and mismatched V/Q. Inabout forty percent of segments with reversed-mismatched V/Q, V/Q pattern changed into matched. Quantified mean V, Q, V/Qand Rev values of baseline and follow-up CTs were positively correlated with PFT parameters, respectively (r = 0.286-0.630, p <0.05), while V/QSD values were negatively correlated with PFT parameters (r = -0.528 and -0.375; p < 0.05). Changes of mean V,V/Q and Mat were positively correlated with change of FEV1 (r = 0.315-344; p < 0.05) and changes of Rev were negativelycorrelated with change of FEV1 (r = -0.353; p = 0.010).

CONCLUSION

Quantitative and visual analysis of combined V and Q DECT showed that the improvement of ventilation and V/Q mismatch may beassociated with the response to pharmacological treatment in COPD patients.


Combined V and Q DECT imaging can be applied to assessment of changes of regional V and Q status after pharmacologictreatment in COPD patients.


SSQ05-03 Hyperpolarized Xenon-129 MRI for Detection of Gas Exchange in Healthy Subjects and Lung CancerPatients


ParticipantsEdgar Karanjah, MBChB, Orange, CA (Abstract Co-Author) Nothing to DiscloseAnastasia Cruz, DO, Orange, CA (Presenter) Nothing to DiscloseEduardo Hernandez-Rangel, MD, Monterrey, Mexico (Abstract Co-Author) Nothing to DiscloseMayil S. Krishnam, MBBS, MRCP, Orange, CA (Abstract Co-Author) Nothing to DiscloseSang Min Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose

PURPOSE

In COPD patients, distribution of emphysema shows various patterns (diffuse, unilateral, or focal), however, there is no reportabout distribution of emphysema in lung cancer patients. The purpose of the research is to compare heterogeneity of emphysemabetween lung cancer patients and lung cancer screening patients.


Total 109 subjects with smoking history and thin section chest CT (51 patients with lung cancer M : F = 29 : 22, age = 68.10 ±9.26, 58 lung cancer screening patients; M : F = 31 : 27, age = 64.03 ± 6.65) were retrospectively enrolled. Using commercialsoftware (AVIEW, Coreline soft, South Korea), volume and low attenuation area under -950 HU were semi-automatically quantifiedin whole lung and each lobe by two radiologists. Emphysema index (EI) and emphysema heterogeneity were calculated. Intra-classcorrelation coefficient (ICC) and independent t-test were performed. ANOVA was performed for subgroup analysis according tocancer pathology.

RESULTS

ICC of each lobe volume among two radiologists were 0.993, 0.987, 0.999, 0.999, and 0.999. EI in RUL, RML, RLL, LUL, and LLL oftwo groups were 6.43 ± 9.94, 6.80 ± 9.28, 3.66 ± 5.54, 5.86 ± 6.60, and 3.83 ± 5.86 in the cancer group, and 6.56 ± 7.82, 8.24 ±8.44, 5.68 ± 7.08, 7.16 ± 7.05, and 5.28 ± 6.66 in the screening group. EI and emphysema heterogeneity in whole lung of twogroups were 5.10 ± 6.56, and 12.20 ± 5.14 respectively in the cancer group, and 6.43 ± 6.95, 8.44 ± 4.92 in the screening group.EI showed no significant difference between two groups. However, emphysema heterogeneity of the cancer group was significantlylarger than that of the screening group (p < 0.001). In subgroup analysis, emphysema heterogeneity of the cancer subtypes ofadenocarcinoma and squamous cell carcinoma were significantly larger than that of screening group (p = 0.006 and 0.042).

CONCLUSION

Semi-automated quantification of emphysema in each lobe was feasible. Smokers with lung cancer showed more heterogeneousdistribution of emphysema than smokers without lung cancer.


Quantification of regional and whole lung heterogeneity of emphysema may potentially help in risk stratification of COPD patients indeveloping lung cancer.

ParticipantsOzkan Doganay, PhD,MSc, Oxford, United Kingdom (Presenter) Nothing to DiscloseMitchell Chen, MD, Oxford, United Kingdom (Abstract Co-Author) Nothing to DiscloseTahreema N. Matin, MBBS, Oxford, United Kingdom (Abstract Co-Author) Nothing to DiscloseAnthony McIntyre, BS, Oxford, United Kingdom (Abstract Co-Author) Nothing to DiscloseGeoff Higgins, Oxford, United Kingdom (Abstract Co-Author) Nothing to DiscloseFergus V. Gleeson, MBBS, Oxford, United Kingdom (Abstract Co-Author) Consultant, Alliance Medical Limited; Consultant, Blue EarthDiagnostics Ltd; Consultant, Polarean, Inc


[email protected]

PURPOSE

To determine whether a novel functional magnetic resonance imaging (MRI) technique using hyperpolarized Xenon-129 (HPX) canquantify the xenon gas transfer dynamics (XGTD) from alveoli into the Pulmonary Tissue and Blood Plasma (PTBP), and Red BloodCell (RBC) compartments of the lungs, and identify XGTD differences in patients with COPD and lung cancer pre and post radiation.


A novel spectroscopic MRI technique was developed using Iterative Decomposition of water and fat with Echo Asymmetry andLeast-square estimation (IDEAL) approach. This technique allowed acquisition of the time-series IDEAL gas, PTBP and RBCcompartment images of lungs with various gas transfer times in a single breath-hold interval. The time-series IDEAL gas, PTBP andRBC compartment images were acquired from five healthy subjects at two different time points. XGTD curves were obtained from 10scans (n=10) that represented the control group. The control group was compared to two lung-cancer patients before radiationtherapy started and after radiation therapy ended.

RESULTS

In the control group, there was no statistical difference in XGTD between the left and right lungs (P-value >0.4). XGTD in thecontrol was statistically different than the lung cancer patients (P-value <0.01) suggesting that the novel time-series IDEALtechnique was sensitive to the gas exchange abnormalities. Additionally, the ratio of XGTD from the irradiated lung to un-irradiatedlungs was compared pre and post radiation therapy. We found that xenon gas in the alveoli diffused into the PTBP compartmentwith a slower rate of 20-35% in the radiated lungs from the lung cancer patients.

CONCLUSION

SSQ05-04 Effect of Aging and Smoking on Regional Air Volume Change Distributions in Normal Chest CT


SSQ05-05 Whole-Lung Dynamic Contrast-Enhanced Perfusion Area-Detector CT: Capability for PulmonaryFunction Assessment and Morphological Change Evaluation in Stage IA Non-Small Cell Lung Cancer


The feasibility of the novel IDEAL MRI technique has been successfully demonstrated in healthy subjects and lung cancer subjects.To our knowledge, this is the first-in-man study showing the time course of arrival of Xenon-129 gas from the alveoli to PTBP andRBC compartments of the lungs and to the pulmonary vasculature and the left ventricle of the heart in healthy subjects andpatients with COPD and lung cancer.


This technique may have potential clinical applications ranging from the detection of regional differences in gas transfer on imagingto the detection of early-stage radiation-induced lung injury.

ParticipantsKum Ju Chae, MD, Jeonju, Korea, Republic Of (Presenter) Nothing to DiscloseGong Yong Jin, MD, PhD, Jeonju, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJiwoong Choi, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChang Hyun Lee, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSanghun Choi, Daegu, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChing-Long Lin, PhD, Iowa City, IA (Abstract Co-Author) Nothing to DiscloseEric A. Hoffman, PhD, Iowa City, IA (Abstract Co-Author) Founder, VIDA Diagnostics, Inc; Shareholder, VIDA Diagnostics, Inc;Advisory Board, Siemens AG


[email protected]

PURPOSE

Image registration has been increasingly used to assess pulmonary dynamics between paired inspiratory and expiratory CT images inpatients with pulmonary disease. However, information of pulmonary dynamics of normal subjects is insufficient. The purpose of thestudy is to describe regional air volume change distributions of subjects with normal CT and to investigate the effects of aging andsmoking.


242 subjects (114 male, 128 female) over the age of 18 years with normal inspiration and expiration CTs were included in the study.VIDA Apollo software (Coralville, IA) and an image registration technique were used to compute regional distribution of air andtissue volumes, air volume fractions, and the relative regional changes between inspiration and expiration, including relative regionalair volume changes (RRAVC). In each lobe, the upper lobes, the lower lobes, and the whole lung, the mean values and standarddeviations were correlated with aging and compared to those of smoking groups. Regional volumetric changes were further analyzedusing 3D visualization of acinar scale parenchymal units.

RESULTS

Inspiratory air volume of the lower/upper lobes decreased with age in both nonsmoking males and females (r=-0.388; p=0.006 andr=-0.258; p=0.004, respectively). RRAVC map demonstrates the increase of air volume change from apico-ventral to dorso-basalregion in non-smokers, representing gravitational dependency in normal pulmonary dynamics. In comparison, the directionality ofgravitational dependency of regional volume change tends to against normality in smokers, and the coefficient of variation (CV) ofRRAVC decreased in the whole lung in the smokers (0.64 and 0.35, p=0.020).

CONCLUSION

The air volume of the lower/upper lobes tends to decrease with aging, and the directionality of gravitational dependency of the airvolume change appeared to be against normality in smokers. Visualization of RRAVC map helped recognize these findings moreeasily.


Regional air volume change distribution helped understand the gravitational volume change of the lung in normal adults, and so it isexpected that the localized functional abnormalities of the lung effected by aging and smoking are easily comprehended.

ParticipantsYoshiharu Ohno, MD, PhD, Kobe, Japan (Presenter) Research Grant, Canon Medical Systems Corporation; Research Grant,Koninklijke Philips NV; Research Grant, Bayer AG; Research Grant, DAIICHI SANKYO Group; Research Grant, Fuji Pharma Co, Ltd;Research Grant, Guerbet SA; Yasuko Fujisawa, MS, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationNaoki Sugihara, MENG, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationYuji Kishida, MD,PhD, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseShinichiro Seki, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationTakeshi Yoshikawa, MD, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationTakamichi Murakami, MD, PhD, Osakasayama, Japan (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To prospectively and directly compare the capability of whole-lung dynamic contrast-enhanced (CE-) perfusion area-detector CT

SSQ05-06 Denoised Ultra Low Dose for Screening Lung Cancer


SSQ05-07 Comparison of SENCEFUL-MRI and Lung Scintigraphy for Detection of Lung Perfusion Defects in

(ADCT) for pulmonary functional loss assessment and morphological change evaluation in Stage IA non-small cell lung cancer(NSCLC) patients.


63 consecutive NSCLC patients (39 males, 24 females; mean age 68 years) underwent dynamic CE-perfusion ADCT performed attwo or three different positions as single examination, pulmonary function test, surgical treatment, and pathological examination.From all perfusion ADCT data in each subject, whole lung total perfusion (TP), pulmonary arterial perfusion (PAP) perfusion,systemic arterial perfusion (SAP) maps were computationally generated based on dual-input maximum slope method by previouslyreported software. In each subject, regional perfusion parameters were assessed by ROI measurements, and averaged to determinemean values. According to pathological examination results, all ROIs within operated lung were divided into following four structuregroups: normal lung, emphysema, non-specific interstitial pneumonia (NSIP) and usual interstitial pneumonia (UIP). To determine thecapability of each perfusion parameter for pulmonary function, Pearson's correlation was performed. To compare each perfusionparameter among all structure groups, Tukey's HSD test was performed. Finally, discrimination accuracy for morphological changeevaluation was compared among all indexes and combined method.

RESULTS

All perfusion parameters except SAP had significant correlation with each pulmonary function parameter (TP: 0.47 CONCLUSION

Whole-lung dynamic first-pass CE-perfusion ADCT is useful for pulmonary functional loss assessment and morphological changeevaluation in stage IA NSCLC patients.


Whole-lung dynamic first-pass CE-perfusion ADCT is useful for pulmonary functional loss assessment and morphological changeevaluation in stage IA NSCLC patients.

ParticipantsEdith M. Marom, MD, Tel Aviv, Israel (Presenter) Speaker, Bristol-Myers Squibb Company; Speaker, Boehringer Ingelheim GmbH; Michael Green, MSc, Tel Aviv, Israel (Abstract Co-Author) Nothing to DiscloseMichal Eifer, MD, Ramat Gan, Israel (Abstract Co-Author) Nothing to DiscloseEli Konen, MD, Ramat Gan, Israel (Abstract Co-Author) Nothing to DiscloseNahum Kiryati, PhD, Tel Aviv, Israel (Abstract Co-Author) Nothing to DiscloseArnaldo Mayer, PhD, Ramat Gan, Israel (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To assess the effect of a denoising method (D) for ultra low dose CT (ULDCT) LungRADS categorization.


36 consented patients, referred for an outpatient chest CT, underwent 2 scans: a normal dose CT (NDCT), 120 kVp and automaticcurrent modulation, with or without contrast media, immediately followed by an ULDCT, 120 kvp and fixed current at 10 mA for bmi<29 and 20 mA for bmi>=29. Reconstruction for lung and soft tissue kernels were performed for each scan. Consecutively, eachULDCT was denoised using a locally-consistent non-local-mean (LCNLM) algorithm to obtain a high signal to noise ratio (SNR)version of the ULDCT. The LCNLM algorithm leverages large databases of image patches extracted from high-SNR chest CT scans todenoise ULDCTs while enforcing local spatial consistency to preserve fine details and structures in the image. Blinded to all clinicalinformation, a chest radiologist separately assessed the NDCT, ULDCT, and denoised ULDCT (D), documented findings, assigned aLungRADS category and a subjective suspicion for highly suspicious lesions for lung cancer (H).

RESULTS

Radiation dose using NDCT reduced the radiation for patients with a BMI > 29 by an average of 93% and for those with a BMI of upto 29 by an average of 96% . For patients with a BMI > 29 the average effective radiation dose for ULDCT was 0.41 mSv, whereasfor those with a BMI of up to 29 it was 0.24mSv. For the three imaging methods, the same score was seen in 63.9% (n=23) and adifferent score in 36.1% (n=13). There was complete agreement on LungRADS 4A (or higher) between NDCT and D, but ULDCTcategorized one of the 4A patients as LungRads 2. One lesion assigned as LungRads 4X by ULDCT was assigned LungRads2 by Dand NDCT. Of the 8 patients highly suspicious for lung cancer by NDCT, D indicated so in all 8 whereas ULDCT indicated so only in4.

CONCLUSION

Interpretation of ULDCT may cause errors in LungRADS categorization but implementation of the LCNLM algorithm for denoisingimproves ULDCT images so that LungRADS categorization is similar to normal dose scans.


Denoising ULDCT with the LCNLM algorithm enables screening for lung cancer with dose reductions of greater than 90%.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Edith M. Marom, MD - 2015 Honored EducatorEdith M. Marom, MD - 2018 HonoredEducator

CTEPH Patients


SSQ05-08 Applicability of Monochromatic Energy with 40 keV for Pulmonary Embolism Detection in thePulmonary Embolism CT Angiography: Experience Using a Dual-Layer Detector Spectral CT


ParticipantsChristian Kestler, MD, Wurzburg, Germany (Presenter) Nothing to DiscloseAndreas Kunz, MD, Wurzburg, Germany (Abstract Co-Author) Nothing to DiscloseMalte Kircher, Wuerzburg, Germany (Abstract Co-Author) Nothing to DiscloseAndreas M. Weng, Wuerzburg, Germany (Abstract Co-Author) Research Grant, Siemens AGTobias Wech, Wurzburg, Germany (Abstract Co-Author) Nothing to DiscloseLenon Mendes Pereira, Wuerzburg, Germany (Abstract Co-Author) Nothing to DiscloseMatthias Held, Wurzburg, Germany (Abstract Co-Author) Nothing to DiscloseHerbert Koestler, PhD, Wuerzburg, Germany (Abstract Co-Author) Nothing to DiscloseThorsten A. Bley, MD, Hamburg, Germany (Abstract Co-Author) Nothing to DiscloseSimon Veldhoen, MD, Wurzburg, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To compare self-gated non-contrast enhanced functional lung MRI (SENCEFUL) and V/Q (ventilation/perfusion) scintigraphyregarding detection of lung perfusion defects in patients with chronic thromboembolic pulmonary hypertension (CTEPH).


Following review board approval and informed consent, 16 patients diagnosed with CTEPH and pathologic findings in V/Q scans wereincluded into this prospective study. Patients were scanned at 3T using the SENCEFUL approach based on a 2D-FLASH sequence.Color-coded maps of the lung perfusion and the local blood arrival time i.e. the pulsation phase were manually segmented and ratedfor perfusion defects in lung quadrants by three independent radiologists using a 6-point Likert scale. Coronal V/Q scan images wererated by a nuclear medicine physician accordingly. Due to variation of slice thickness between both techniques, covered lungvolumes were divided into four sectors in coronal orientation each containing four quadrants to improve comparability. Statisticaltests included intraclass correlation coefficient (ICC) and Mann-Whitney-U-test.

RESULTS

Comparison of quadrant-wise rating between SENCEFUL-MRI and V/Q scans revealed good agreement between all raters when thelung perfusion and pulsation phase maps were rated simultaneously (ICC 0.75, 95% CI 0.69-0.80, p<0.05) and an improvement toperfusion rating alone (ICC 0.61, 95% CI 0.52-0.69, p<0.05). Interrater reliability of the radiologists for combinedperfusion/pulsation phase rating was good (ICC 0.77, 95% CI 0.69-0.82, p<0.05). Analysis of a peak-to-offset ratio of pulsationphase histograms showed a significant difference between lung quadrants rated pathologic in scintigraphy and quadrants ratedhealthy (p<0.05).

CONCLUSION

SENCEFUL-MRI showed good agreement for detection of perfusion defects compared with V/Q scans being the current screeningmethod for CTEPH. Analysis of MRI maps by a peak-to-offset ratio of pulsation phase showed a significant difference betweenquadrants rated pathologic and healthy by V/Q scans suggesting a quantifiable value for future determination of threshold values inSENCEFUL-MRI.


SENCEFUL-MRI could be an alternative screening method for detection of lung perfusion defects in patients with suspected CTEPHwithout the need of contrast agent administration or radiation exposure.

ParticipantsMo In Ha, MD, Ansan, Korea, Republic Of (Presenter) Nothing to DiscloseCherry Kim, MD, Ansan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseKi Yeol Lee, MD, PhD, Ansan , Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSeung Hwa Lee, Ansan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJaehyung Cha, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseBo-Kyung Je, MD, PhD, Ansan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseEun-Young Kang, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYu Whan Oh, MD,PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHwan Seok Yong, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose

PURPOSE

Previous studies have shown that the optimal energy level of virtual monoenergetic images (VMI) for pulmonary artery (PA)enhancement is 40 keV in spectral CT. The aim of this study is to evaluate the applicability of VMI at 40 keV for pulmonaryembolism (PE) detection in the CT angiography (PECT).


A total of 876 consecutive PECT using spectral CT were identified between August 2016 and March 2018. Of these, PE at least 4mm in diameter was detected in 73 PECT. Among these, suboptimal enhancement of PA (<250 HU) was shown in 19 cases.Contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR) of VMI at 50 keV, 60 keV, 70 keV, and conventional 120-kVp images(COV) were compared with VMI at 40 keV in all PECT and suboptimal PECT. Readers' subjective scores for PE detection was alsorecorded. The mean diameters of PE were measured, and they were compared between VMI at 40-70 keV and COV. The frequencyof significant PE diameter reduction (>40%) in VMI compared with COV was also recorded and compared between VMIs. The cut off

SSQ05-09 Fluorine-19 MRI Ventilation Defect Analysis in Cystic Fibrosis

Thursday, Nov. 29 11:50AM - 12:00PM Room: E353B

value of the minimum visible PE diameter at 40 keV was investigated in COV.

RESULTS

There was no significant difference in CNR between 40 keV and 50 keV, although the highest CNR and SNR were obtained at 40keV. In the suboptimal subgroup, there were no significant differences in both CNR and SNR between 40 keV and 50 keV. Thesubjective scores was significantly lower at 40 keV, compared with other algorithms in both all PECT and the suboptimal subgroup(P<0.05). The mean diameters of PE were significantly decreased in 40 keV and 50 keV, compared with those in COV (40 keV,5.6±5.8 mm; 50 keV, 7.2±5.3 mm; COV, 8.9±4.9 mm; all P<0.05). The frequency of significant PE diameter reduction wassignificantly higher in 40 keV than in 50 keV (36.8% vs.12.8%, P<0.001). The cut off value of the minimum visible PE diameter at 40keV was 6.4 mm in COV.

CONCLUSION

VMI at 40 keV was not the best option for PE detection, although the best CNR and SNR were obtained at 40 keV. The diameter ofPE was often decreased and small PE was not even detected at 40 keV.


We propose that not only 40 keV but also other algorithms such as 50 keV should be used for PE detection to ensure that we donot miss small PEs.


ParticipantsTyler Glass, BEng, Chapel Hill, NC (Presenter) Nothing to DiscloseJennifer Goralski, Chapel Hill, NC (Abstract Co-Author) Nothing to DiscloseEsther O. Akinnagbe-Zusterzeel, Chapel Hill, NC (Abstract Co-Author) Nothing to DiscloseScott Donaldson, Chapel Hill, NC (Abstract Co-Author) Nothing to DiscloseRayad Shams, Chapel Hill, NC (Abstract Co-Author) Nothing to DiscloseYueh Z. Lee, MD,PhD, Chapel Hill, NC (Abstract Co-Author) License agreement, XinRay Systems IncAgathe Ceppe, Chapel Hill, NC (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The purpose of this study is to investigate the ability of novel fluorine-19 (19F) based MRI to characterize ventilation in subjectswith cystic fibrosis.


Coronal images of nine healthy controls and twelve subjects with CF were acquired using a multinuclear capable 3.0 T MRI scanner(PRISMA, Siemens) along with spirometry. Subjects inhaled 19F labelled perfluoro-propane (PFP) gas mixed with 21% O2 or room airduring the wash-in phase of the scan. Fifteen second 19F GRE vibe breath hold images were obtained following three breaths ofPFP for five cycles. This was repeated five times after switching PFP gas to room air for the wash-out phase. A 19F maximumintensity projection image over time was created and segmented using a semi-automatic approach with an empirically determinedventilation threshold. Anatomic 1H series taken at full inspiration were then manually segmented for all subjects. After correcting fordifferences in respiratory effort by comparing apex-base measurements of the lung in 19F and 1H series, the ventilation defectvolume (VDV) was computed by subtracting 19F segmentation volume from 1H volume and a ventilation defect percentage (VDP)was also computed relative to 1H volume.

RESULTS

In healthy controls, the mean ventilation defect percentage (VDP) was 10% (SD 11%); for mild CF 13% (SD 25%); and for severeCF 31% (SD 24%). A significant difference was found when comparing all CF patients to normal (p=0.0275 via t-test withSatterthwaite correction). VDP had a negative correlation with FEV1 (-0.56 via Spearman correlation, p=0.011). The rate constantfor gas filling (τ1) was significantly increased in CF patients compared with controls, suggesting delay in filling. No safety concernswere detected throughout the study.

CONCLUSION

This study showed the ability of novel 19F ventilation MRI to rapidly and safely quantify regional ventilation defects and gas wash-in and wash-out dynamics. 19F MRI identified ventilation defects in cystic fibrosis subjects even in the setting of normal spirometrywith some variability in healthy volunteers.


This novel imaging technique has advantages over xenon ventilation MRI including cheaper contrast material and inert compoundallowing functional imaging with multiple image sets. We anticipate applications for many other lung diseases including pediatric lungmalformations, lung resection, COPD monitoring, and bronchiectasis.

SSQ12-01 Histological Validation of Chemical Exchange Saturation Transfer (CEST) imaging for theMeasurement of Metabolism Status in Infarcted Myocardium


SSQ12-02 Molecular Lumbar Intervertebral Disc Alterations in Patients with Leg Length Discrepancy Before andAfter Therapy


SSQ12

Molecular Imaging (Musculoskeletal, Gastrointestinal, Cardio)

Thursday, Nov. 29 10:30AM - 12:00PM Room: S504CD

CA GI MR MI MK NM


ParticipantsChun Yuan, PhD, Seattle, WA (Moderator) Research Grant, Koninklijke Philips NV; ;

Sub-Events

ParticipantsKaiyue Diao, Chengdu, China (Presenter) Nothing to DiscloseZhigang Yang, MD, Chengdu, China (Abstract Co-Author) Nothing to DiscloseYingkun Guo, Chengdu, China (Abstract Co-Author) Nothing to DiscloseWanlin Peng, MS, Chengdu, China (Abstract Co-Author) Nothing to DiscloseChunchao Xia, Chengdu, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The purpose of this study was to test if the distribution of creatine shown on Chemical Exchange Saturation Transfer (CEST) MRIcould differentiate infarct myocardium from the normal on pigs, by referring to LGE images and pathologic results.


We prospectively enrolled 27 Bama miniature pigs. MI model was built by applying a ligation at the remote ending of the left anteriordescending artery. CMR scan was arranged at 3 days and 2 months later for the AMI and CMI group on a 3 T whole-body scanner.A single SAX slice was used for CEST scanning by using Amide proton transfer (APT) sequence before the injection of contrast. 36samples were collected from a saturation frequency offset from -5.0 ppm to + 5.0 ppm. The scanned pigs were humanelyeuthanized under deep anesthesia with KCl and the heart was excised. Triphenyl tetrazolium chloride was used to manifest theinfarcted region. CEST values at the frequency of ~1.8ppm, ~2.5ppm, ~2ppm and ~3ppm were recorded respectively for each pigand Color code map was plotted based on the CEST values at a frequency offset of ~1.8ppm (Matlab). Statistic analysis wasperformed on R project.

RESULTS

A total of 5 AMI pigs (M, 7 months, 16.6 ± 1.2 kg), and 14 CMI pigs (M, 9 months, 27.8 ± 2.1 kg), were finally included. Statisticdifferences were observed for Cr, ATP, and Glu between the infarct myocardium and the normal myocardium for CMI pigs, while onlyCr and ATP for AMI pigs. The color code CEST maps showed a prominent larger abnormal region with a lower concentration ofcreatine than the MI regions recognized on the LGE sequences and the pathology images.

CONCLUSION

This study demonstrated that the metabolic conditions measured on CEST imaging could be used for infarcted myocardiumrecognition and the region of myocardium with a lower creatine concentration was larger than the region confirmed with infarction,which again provided proof of the existence of the injured or stunned myocardial tissue surrounding the infarction region.


CEST MRI provided a promising invasive way to observe metabolism status of infarcted myocardium and further studies on MIpatients would be needed to validate its clinical application.

ParticipantsChristoph Schleich, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseMiriam Frenken, Dusseldorf, Germany (Presenter) Nothing to DiscloseDaniel B. Abrar, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseYan Klosterkemper, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseJohannes Boos, MD, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseJoel Aissa, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to Disclose

SSQ12-03 Blood Oxygen Level-Dependent MRI Can Evaluate the Oxygenation of Visceral Adipose Tissue inZucker Diabetic Fatty Rats


Gerald Antoch, MD, Duesseldorf, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Leg length discrepancy (LLD) is a frequent incidental finding during orthopedic physical examination and can be found in about two-thirds of the population without any physical complaints. According to recent studies, LLD greater than 10 mm could be apredisposing factor for early degenerations of lumbar intervertebral discs or vertebral facet joints. However, the need of itstreatment is still controversial. Previous findings suggest that degeneration of the lumbar disc correlates with a decrease ofglycosaminoglycan content (GAG). The purpose of this study was to elucidate the effect of LLD on GAG content in lumbar discs andto show therapy effects after the usage of shoe inserts and physical therapy.


11 patients (25.6 ± 4.3 years) with LLD greater than 10 mm and 14 control subjects (23.9 ± 3.5 years) without LLD were examinedusing a 3T MR scanner. 8 patients were re-examined 6 months after physical therapy and the usage of shoe inserts. MorphologicalT2-weighted sequences in sagittal and transversal orientation and Glycosaminoglycan chemical exchange saturation transfer(gagCEST) sequence were performed. Subjects with bulged or herniated discs were excluded.

RESULTS

Nucleus pulposus-gagCEST values of L5/S1 disc were significantly lower in patients with LLD compared to control group (p =0.0008). For all other disc levels, no significant difference was found. At follow-up, no significant difference of NP-gagCEST valuesat baseline and 6 months after therapy could be found (p > 0.05).

CONCLUSION

This study supports the hypothesis that LLD greater than 10 mm could be a predisposing factor for early molecular alterations oflumbar discs of L5/S1. Remarkably, we observed lower gagCEST values of the lumbar disc of L5/S1 caused by LLD even before anymorphological pathology could be found. Biochemical disc alterations of patients with LDD could be stopped under therapy.


This study supports the hypothesis that LLD could be a predisposing factor for early molecular alterations of the lumbar disc ofL5/S1. Furthermore, lower gagCEST values of the lumbar disc of L5/S1 caused by LLD were observed before any morphologicalpathologies were detectable. This molecular alterations of L5/S1 of patients with LLD could be delayed under the effect of shoeinserts and physical therapy.

ParticipantsMatteo Figini, Chicago, IL (Presenter) Nothing to DiscloseYaqi Zhang, Chicago, IL (Abstract Co-Author) Nothing to DiscloseSu Hu, Chicago, IL (Abstract Co-Author) Nothing to DiscloseLiang Pan, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJunjie Shangguan, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJia Yang, PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseQuanhong Ma, Chicago, IL (Abstract Co-Author) Nothing to DiscloseYuri Velichko, PhD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseVahid Yaghmai, MD, Chicago, IL (Abstract Co-Author) Nothing to DiscloseZhuoli Zhang, MD,PhD, Chicago, IL (Abstract Co-Author) Nothing to Disclose

PURPOSE

Visceral adipose tissue (VAT) hypoxia is associated with insulin resistance and obesity-related chronic low-grade inflammation(metaflammation). Its evaluation is then of great importance for prevention and therapy, but current methods are invasive andfocus on subcutaneous fat rather than VAT. The purpose of this study is to investigate the feasibility of evaluating VAT hypoxiawith Blood Oxygen Level-Dependent (BOLD) MRI, which is sensitive to hemoglobin oxygenation, in Zucker Diabetic Fatty (ZDF) rats.


Seven-week old ZDF rats (n=18) were provided with water and high-fat diet ad libitum; their body weight and blood glucose weremonitored. At 13 weeks of age they were divided into two subgroups, receiving a daily dose of pioglitazone (ZDF-PGZ, n=9) orsaline (ZDF-VE, n=9) respectively. BOLD MRI was performed at 13 and 23 weeks of age using a multi-echo spoiled gradient-echosequence (5 echo times from 3.75 to 29.07 ms with 6.33 ms echo spacing, TR = 408 ms, voxel size = 0.47x0.38x3 mm3). R2*values were measured in the perirenal VAT. The trygliceride, cholesterol and insulin levels were measured by blood biochemistryanalysis, and insulin resistance was calculated by HOMA-IR = insulin[mU/L] x glucose[mmol/L] / 22.5. Immunofluorescence was usedto evaluate hypoxia by pimonidazole adduct-positive area. The proportion of Th17 and Treg cells, CD34+ and CD34++ monocyteswere evaluated by flow cytometry.

RESULTS

The ZDF-VE group had hyperlipidemia (p < 0.01) and hyperinsulinemia (p < 0.001) and higher HOMA-IR (p < 0.001) compared to theZDF-PGZ group. There was a significant R2* increase between the two scans for ZDF-VE (20.14 ± 0.23 vs. 21.53 ± 0.20, p =0.012) but not for ZDF-PGZ (figure 1A). VAT R2* values showed a positive correlation with pimonidazole adduct-positive area,HOMA-IR, the percentage of Th17 cells and CD43+ monocytes, and a negative correlation with the percentage of Treg cells andCD43++ monocytes (figure 1B).

CONCLUSION

This study showed the feasibility of VAT oxygenation by BOLD MRI in ZDF rats with obesity induced by high-fat diet. The R2*

SSQ12-04 Baseline Pancreatic Beta Cell Imaging Post Pancreatic Transplantation Using Whole Body 68Ga-DOTA-Exendin-4 PET/CT: Our Initial Experience


SSQ12-05 In Vivo Bioluminescence Imaging of Transplanted Mesenchymal Stromal Cells and Their RejectionMediated by Intrahepatic NK Cells


values obtained by BOLD MRI are also associated with insulin resistance and metaflammation.


BOLD-MRI can be a non-invasive tool for the evaluation of visceral adipose tissue hypoxia and obesity-related insulin resistance andsystemic inflammation

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Vahid Yaghmai, MD - 2012 Honored EducatorVahid Yaghmai, MD - 2015 HonoredEducatorVahid Yaghmai, MD - 2017 Honored Educator

ParticipantsMurali K. Logudoss, MBBS, MD, Chennai, India (Presenter) Nothing to DiscloseNatesan Chidambaranathan, MD, PhD, Chennai, India (Abstract Co-Author) Nothing to DiscloseRajasekaran Sivaprakasam, DMRD,PhD, Chennai, India (Abstract Co-Author) Nothing to DiscloseKanimozhi Damu JR, MBBS,MD, Coimbatore, India (Abstract Co-Author) Nothing to DiscloseAnand N. Parimalai, MD, Chennai, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Whole-pancreatic transplant and islet cells transplantation are currently available strategies aiming towards diabetes cure. Beta cellspecific non-invasive functional imaging using novel PET radiotracers are now available.68Ga-DOTA-Exendin-4 PET/CT is used fordetecting localised Insulinomas. However this tracer can also be used for beta cell imaging and quantification. We have attemptedin this study to recognise the pattern of uptake by this tracer in patients with pancreatic transplants.


8 patients who had undergone pancreatic transplant for Diabetes Mellitus were included in the study. After obtaining informedconsent from the patients 4-5 mCi of 68Ga-DOTA-Exendin-4 was injected intravenously. One hour after injection whole body PETCT was performed and the images were analysed.

RESULTS

Among the 8 patients who had pancreatic transplant, 4 patients had Type I Diabetes Mellitus and 4 patients had Type II DiabetesMellitus. The mean age of the patients were 36 yrs. All the 8 patients were male patients. One of the patient had undergonesimultaneous pancreatic and renal transplant. Anterior and lateral MIP images demonstrated diffuse heterogeneous GLP-1Rexpression in vertically oriented transplanted pancreas in 7 out of 8 patients. Three dimensional PET CT imaging along revealedincrease tracer uptake in the transplanted pancreas. There was no uptake in the native pancreas in 6 out of the 8 patients. Therewas atrophy and calcification of the native pancreatic tissue in these 6 patients. Mild tracer uptake was noted in 2 out of the 8patients. In one patient there was very low tracer uptake in the transplanted pancreas. Fat stranding was noted surrounding thetransplant tissue with areas of necrosis within. This patient was later confirmed to have transplant rejection.

CONCLUSION

In our initial study of 8 patients we conclude Exendin-4 PET/CT is very sensitive tracer for beta cell imaging. It can be used forbaseline and flow up of graft imaging. Currently biopsy is the only method to prove graft rejection. However with the use ofExendin-4 PET/CT early graft rejection can be detected non invasively. We further hypothesize the future use of Exendin-4 PET/CTfor quantification of beta cell mass using volumetric analysis.


68Ga-DOTA-Exendin-4 PET/CT can be used as baseline and for follow up pancreatic transplant patients for analysis of beta cellmass.

ParticipantsJingjing Liu, Zhengzhou, China (Presenter) Nothing to Disclose


[email protected]

PURPOSE

Mesenchymal stromal cells (MSCs) hold promise in the treatment of liver disease.However, short survival time of MSCs afterintrahepatic transplantation limits their value; therefore,understanding the basis of MSCs survival and rejection may increase theirutility. This study was aimed at determining the role of intrahepatic natural killer (NK) cells on MSCs survival and their retention inthe liver shortly after transplant.


SSQ12-06 Assessment for NASH-Related Hepatocarcinogenesis Inhibition of Shikonin in a Murine Model UsingDW-MRI


Human MSCs were labeled with the Luc2-mKate2 dual-fusion reporter gene (MSCs-R), and the residence time and survival of MSCs-R xenografts after intrahepatic transplantation were evaluated by in vivo bioluminescence imaging (BLI). Coculture of MSCs and NKcells was performed to assess cytotoxicity. To evaluate the role of NK cells in rejection of the xenografted cells, the fates oftransplanted MSCs-R were then assessed in vivo by BLI after activation of intrahepatic NK cells.

RESULTS

We observed a linear correlation between luciferase activity from live MSCs-R and cell number in vitro (R2 = 0.9956). In vivo, weobserved a gradual decline in bioluminescent signals from transplanted MSCs-R over a region corresponding to the liver in both thecontrol group and the NK-activated group.However, the survival time and retention of intrahepatic MSCs-R decreased more rapidlyin the NKactivated group of mice compared to the control group. This indicated that activated NK cells accelerate the elimination oftransplanted MSCs. Also, we found that the number of hepatic NK cells and the expression of NK activation markers significantlyincreased after intrahepatic delivery of MSCs. This suggested that resident NK cells, in a resting state, were activated byintrahepatic transplantation of human MSCs. Taken together, the data suggests that activated hepatic NK cells mediate, in part,rejection of the MSCs xenografts. Cytotoxicity assays showed that activated NK cells may inhibit the proliferation of MSCs and, toa certain extent, induce MSCs death.

CONCLUSION

Human MSCs could be followed dynamically in vivo by BLI, and the role of murine hepatic NK cells, especially activated NK cells,could be inferred from the loss of signals from MSCs.


This finding may have practical clinical implications in MSCs transplantation in treating liver disease.

ParticipantsHong Young Jun, PhD, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseKwon-Ha Yoon, MD, PhD, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChang Won Jeong, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseTae-Hoon Kim, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSi-Hyeong Noh, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJi-Eon Kim, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSeungjin Kim, Iksan, Korea, Republic Of (Abstract Co-Author) Nothing to DisclosePan Jung Kim, Iksan-si, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseKoeun Lee, MD, Seoul, Korea, Republic Of (Presenter) Nothing to Disclose


[email protected]

PURPOSE

Nonalcoholic steatohepatitis (NASH) is a major risk factor for hepatic carcinogenesis. This study was assess the effect of shikoninusing diffusion-weighted magnetic resonance imaging (DW-MRI) in an NASH-related hepatocarcinogenesis murine model.


On the second day after birth, male pups were subjected to a single subcutaneous injection of 200 μg streptozotocin (STZ) andfed high-fat (45% kcal from fat) diet from the age of 4 weeks. The mice were randomly divided into groups when the tumor areawas about > 0.5 mm² as follows: STZ + high-fat diet (SH; n=6) and STZ + high-fat diet + shikonin (SHS; n=7). For theexperimental group, shikonin (2.0 mg/kg) was injected intraperitoneally daily for 14 days (with diluted PBS). DW-MRI was performedto assess effects of shikonin at pre-and post-treatment. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT),glucose, cholesterol, and triglyceride were determined in plasma. The liver tissues were collected at 14 day post-treatment forhematoxylin and eosin staining.

RESULTS

The mean area of tumors were 2.56 ± 2.12 mm² at the SH and SHS groups, before treatment initiation. The tumor area changes ofthe SH and SHS groups were 326.28 ± 320.81% and 91.58 ± 78.22% after post-treatment. The tumor area change in the SHSgroup significantly lower compared to the SH group (p < 0.05). The mean ADC changes of the SH and SHS groups were 41.97 ±50.48% and -9.24 ± 30.46% after post-treatment. The ADC change in the SHS group significantly decreased compared to the SHgroup (p < 0.01). AST and ALT levels were significantly lower in the SHS group than in the SH group after post-treatment. Plasmaglucose, total cholesterol and triglyceride levels were not significantly different between SH and SHS groups. The SH groupexhibited numerous tumors on the liver surface, whereas the SHS group exhibited fewer and smaller liver tumors. The histologicalfindings at SH and SHS groups revealed that the tumors were hepatocellular carcinoma.

CONCLUSION

In this study, we found that the cancer inhibition effects of shikonin in a NASH-related hepatocarcinogenesis murine model by usingDW-MRI.


Shinkonin might be considered a novel preventive or therapeutic approach for NASH-related hepatocarcinogenesis.

SSQ12-07 Management of Complex Regional Pain Syndrome (CRPS) with Sigma-1 Receptor Radioligand andPET/MRI


SSQ12-08 Inflammation Focus Search with 18F-FDG-PET/MRI: Comparative or Additive Value of PET and MRI


ParticipantsSandip Biswal, MD, Stanford, CA (Presenter) Research Grant, General Electric Company; Peter Cipriano, BA, Stanford, CA (Abstract Co-Author) Nothing to DiscloseDaeHyun Yoon, PhD, Stanford, CA (Abstract Co-Author) Research support, General Electric Company


[email protected]

PURPOSE

Complex regional pain syndrome (CRPS) is a severe chronic pain condition affecting millions worldwide. Unfortunately, there is nospecific diagnostic test to identify the pain generators in CRPS, leading to poor pain management of this disease. Given sigma-1receptors (S1Rs) specific association for pro-nociceptive processes, we determine the clinical impact of a more pain-specificPET/MRI approach for CRPS, adopting a novel high affinity sigma-1 receptor (S1R) PET radioligand ([18F]FTC-146; Ki = 0.0025 nM).


IRB and FDA approval were obtained. Fifteen patients suffering from CRPS were referred directly from specialists in pain medicine.Whole-body (head-to-toe) PET/MR (time-of-flight PET; 3.0T MR bore; GE Healthcare) imaging was perfmored following 10 mCi IVinjection of [18F]FTC-146. MR sequences included 3D axial LAVA-FLEX, high-resolution 3D axial DESS and 2D axial T2-weighted FSEscans. ROI analysis was performed (OsiriX v.6.0 64-bit). Findings from the PET/MR scans were discussed with the referring painspecialists, subsequent alterations in the pain management plan were recorded and, in a subset of cases, new treatments wereapplied to which outcomes were measured.

RESULTS

Fourteen out of 15 patients showed unexpected findings on [18F]FTC-146 PET/MRI, which lead to a change in the patients' painmanagement plans. In one specific case, a CRPS patient had severe (8-10/10) unilateral knee pain despite 2 previous unsuccessfulsurgeries. [18F]FTC-146 PET/MRI showed a high, focal [18F]FTC-146 PET uptake of a lesion which co-localized to an abnormalmass-like lesion in the intercondylar notch on the MRI. Subsequent arthroscopic surgery removed the [18F]FTC-146-avid lesion,which completely relieved the knee pain (0/10 pain). A separate CRPS patient with severe bilateral foreleg pain showed increaseduptake of [18F]FTC-146 in the anterior compartment of both forelegs. Botulinum toxin injection in the areas of high [18F]FTC-146uptake resulted in significant improvement in pain score (9-10/10 down to 2/10). We continue to follow the other patients tofurther evaluate our image findings.

CONCLUSION

A whole-body PET/MRI approach with a novel S1R PET tracer, [18F]FTC-146, can potentially identify pain generators in CRPS andimproves treatment outcomes.


The proposed whole body PET/MRI approach could alter the pain management for CRPS patients to achieve better pain-reliefoutcome.

ParticipantsAhmet Oernek, Essen, Germany (Presenter) Nothing to DiscloseLale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AGYan Li, Essen, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SAVerena Ruhlmann, Essen, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To evaluate the comparative or additive value of 18F-FDG PET and MRI for identifying the etiology of inflammation of unknownorigin.


A total of 24 patients (13 m, 11 w, age 42±23 [8-82] y) with suspicion of an inflammation focus due to laboratory inflammationmarkers (increased CRP, leukocytes) or fever and up to now non-leading conventional imaging underwent a whole-body PET/MRI.Image analyses included the detection and localization of pathologically (focal) increased tracer uptake in PET includingdetermination of SUVmax using VOI technique and evaluation of the contrast enhancement and diffusion restriction (ADC values) ofabnormal lesions in MRI. Descriptive analysis included mean values, standard deviation and range. PET/CT, clinical, and radiologicalfollow-up as well as histopathology served as standards of reference.

RESULTS

In 17/24 patients the PET/MRI contributed to the diagnosis of a (focal) pathological etiology of the inflammatory disease (vasculitisn=5, inflammatory bowel disease n=4, pneumonia n=1, infected vascular prosthesis n=2, (active) retroperitoneal fibrosis n=1,

SSQ12-09 Early Detection and Measurement of Disease Activity in Experimental, Inflammatory Bowel DiseaseUsing Target-Specific Molecular Imaging and Fluorescence Colonoscopy

Thursday, Nov. 29 11:50AM - 12:00PM Room: S504CD

peritonitis and cholecystitis n=1, synovitis n=1, mycotic infection (hepatic candidosis) n=1, bone marrow activation n=1). In PET allpathological foci showed a moderately to significantly increased FDG uptake (SUVmax 5.3 ± 3.5, range 1.4-14.2). The MRIsatisfactorily allows the localization of the findings, but only in 12/17 a corresponding contrast-enhancement and in 13/17 acorresponding diffusion restriction could be found. 3/17 patients showed neither a contrast-enhancement nor a diffusion restriction,but only an increased FDG uptake.

CONCLUSION

Integrated 18F-FDG-PET/MRI shows high potential in identifying the etiology of inflammation of unknown origin. The MRIsatisfactorily allows the localization of the findings, but a significant higher detection rate could be found in PET compared to MRI.Considering the significantly lower dose of ionizing radiation, PET/MRI may serve as a powerful alternative to PET/CT.


Inflammation focus search with 18F-FDG-PET/MRI

ParticipantsMichel Eisenblaetter, MD, Muenster, Germany (Presenter) Nothing to DiscloseTobias Nowacki, Warendorf, Germany (Abstract Co-Author) Nothing to DiscloseAnne Helfen, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseAnnika Schnepel, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseWalter L. Heindel, MD, Muenster, Germany (Abstract Co-Author) Nothing to DiscloseMoritz Wildgruber, MD, PhD, Munster, Germany (Abstract Co-Author) Nothing to DisclosePhilipp Lenz, Muenster, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Pro-inflammatory monocytes comprise the majority of the early inflammatory infiltrate in inflammatory bowel disease (IBD). In mice,these cells are characterised by high expression of Ly6C. Purpose of this study was to evaluate Ly6C-specific imaging forvisualisation and measurement of IBD activity in comparison to perfusion-type contrast agents and assess the performance influorescence mediated tomography (FMT) and fluorescence colonoscopy (FC) for whole body and local application respectively.


IBD was induced in 10 female Balb/c wild type mice by application of DSS with the drinking water. The weight was monitored as amarker of disease activity. FMT was performed before and 5 and 10 days after IBD induction. Mice received a Cy5.5-labelled Ly6Cantibody (2nmol dye) or an equivalently labelled, unspecific IgG to reflect perfusion effects. In parallel, all mice underwent FC fordetection and scoring of local disease activity. Histology served for correlation and validation of in vivo imaging.

RESULTS

On day 5 after IBD induction, weight loss did not allow for safe identification of IBD activity and was only significantly increased atday 10 (2% vs. 15%). Perfusion was elevated on day 5 as compared to baseline already but did not increase significantly towardsday 10 as reflected by the IgG-driven signal (192 vs. 328 vs. 342 pmol tracer). Ly6C-specific tracer accumulation was, in contrast,significantly elevated on day 5 already; a further increase towards day 10 reflected the growing disease activity (110 vs. 700 vs.1166 pmol; p<0.001). In vivo colonoscopy allowed for safe identification of inflammatory foci based on the specific probeaccumulation but not the unspecific control. FC-based disease scoring was clearly reflected by Ly6C-specific imaging.

CONCLUSION

Target-specific imaging of Ly6C as a marker for early infiltrating, pro-inflammatory monocytes allows for sensitive and specificmeasurement of IBD activity in vivo by non-invasive and endoscopic approaches. It is superior over clinical examination andperfusion type contrast.


In IBD, monitoring of disease activity and sub-clinical inflammation e.g. under therapy is a relevant challenge. Specific imaging canaid research and potentially improve multi-modal clinical imaging.

SSQ14-01 Initial Results From the World's First Total-Body Positron Emission Tomograph


SSQ14

Nuclear Medicine (Technical Innovations and Emerging Opportunities)

Thursday, Nov. 29 10:30AM - 12:00PM Room: S505AB

CT MR NM



ParticipantsBital Savir-Baruch, MD, Atlanta, GA (Moderator) Research Grant, Blue Earth Diagnostics Ltd; Consultant, Blue Earth Diagnostics Ltd;Speaker, Koninklijke Philips NVChadwick L. Wright, MD,PhD, Lewis Center, OH (Moderator) Nothing to Disclose

Sub-Events

ParticipantsRamsey Badawi, PhD, Sacramento, CA (Presenter) Investigator, Koninklijke Philips NV; Investigator, United Imaging HealthcareWeiping Liu, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareTianyi Xu, PhD, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareYu Ding, Shanghai, China (Abstract Co-Author) Researcher, UIH America, IncYang Lv, PhD, Shanghai, China (Abstract Co-Author) Researcher, Shanghai United Imaging Healthcare Co, LtdXinyu Lv, PhD, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging Healthcare Co, LtdJun Li, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareAng Dong, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareYun Feng Guo SR, MS, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareMiao Li, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareShaohui An, PhD, Shanghai, China (Abstract Co-Author) Nothing to DiscloseYun Dong, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging Healthcare Co, LtdJun Bao, Shanghai, China (Abstract Co-Author) Employee, Shanghai United Imaging HealthcareHongdi Li, PhD, Houston, TX (Abstract Co-Author) Nothing to DiscloseEric Berg, Davis, CA (Abstract Co-Author) Nothing to DiscloseXuezhu Zhang, PhD, Davis, CA (Abstract Co-Author) Nothing to DiscloseMartin Judenhofer, PhD, Davis, CA (Abstract Co-Author) Nothing to DiscloseEdwin Leung, Davis, CA (Abstract Co-Author) Nothing to DiscloseJinyi Qi, PhD, Davis, CA (Abstract Co-Author) Nothing to DiscloseTerry Jones, Sacramento, CA (Abstract Co-Author) Nothing to DiscloseSimon R. Cherry, PhD, Davis, CA (Abstract Co-Author) Research Collaboration, Shanghai United Imaging Healthcare Co, Ltd


[email protected]

PURPOSE

Positron Emission Tomography (PET) offers the most sensitive method for in vivo imaging assays of physiologically importantcompounds, but it is fundamentally limited by low signal and/or high radiation dose to the subject, which negatively impacts imagequality, scan times and the kinds of diseases that may be investigated. PET also has the capacity to dynamically track the fate ofbiomolecules in vivo, allowing for pharmacokinetic analysis. However, standard clinical scanners have an axial field of view (AFOV)of 15-30 cm, which limits such analysis to single organs. This research program aims to address all these limitations by buildingextended AFOV scanners. Here we report initial results from a 194 cm long device - the first medical tomograph capable ofsimultaneously imaging the entire human body.


The scanner consists of 8 rings of 24 PET detector modules, each containing 5 x 14 detector blocks. Blocks consist of 6 x 7 LYSOcrystals of size 2.76 x 2.76 x 18.1 mm3 (total ~560 kg of LYSO), read out by silicon photomultipliers. The PET component is pairedwith an 80-channel CT scanner. PET detector performance has been characterized and system construction and integration hasbeen completed. Static data from a 200 cm phantom has been acquired and reconstructed to investigate detector responseuniformity. A 30-second dynamic scan of activity moving through a tube has also been acquired to verify dynamic framegeneration.

RESULTS

Detector time-of-flight resolution is 409±39 ps and energy resolution is 11.7%±1.5% at 511 keV. Detector dead-time of 3.5% wasfound at count-rates similar to those expected in clinical operation. Images of the 200 cm phantom show reasonable uniformityeven though not all corrections have been implemented yet. The dynamic dataset shows that frame creation is working asexpected.

SSQ14-02 The Effect of a Novel Bayesian Penalised Likelihood (BPL) PET Reconstruction on the Herder RiskPrediction Model of Malignancy in Solitary Pulmonary Nodules Undergoing Assessment with 18F-FDGPET-CT


SSQ14-03 Impact of Point Spread Function Reconstruction on 68Ga DOTATATE PET/CT Quantitative ImagingParameters


CONCLUSION

The world's first total-body PET/CT scanner has been built. Detector performance is in line with expectations. The system isoperational and producing images. Implementation and validation of corrections for accurate quantification is under way. Furtherperformance characterization is planned.


Total-body PET aims to improve all clinical PET through ultra-fast (<1min) scans; ultra-low-dose (<0.35mSv) scans; improved imagequality; and total-body kinetic modeling for precision medicine.

ParticipantsDavid J. Murphy, MBBCh,FFR(RCSI), London, United Kingdom (Presenter) Nothing to DiscloseLeanne Royle, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseZacharias Chalampalakis, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseLuis Alves, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseNuno Martins, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseRonan Breen, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseArjun Nair, MD, FRCR, London, United Kingdom (Abstract Co-Author) Advisory Board, Aidence BV; Andrea Bille, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSugama Chicklore, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseGary Cook, MD, FRCR, London, United Kingdom (Abstract Co-Author) Research support, General Electric Company Research support,Alliance Medical Limited Research support, Siemens AG Research Consultant, Blue Earth Diagnostics Ltd Speakers Bureau, Bayer AGManil Subesinghe, MRCP,FRCR, Leeds, United Kingdom (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The British Thoracic Society (BTS) guidelines recommend using 18F-FDG PET-CT with the Herder model to assess the risk ofmalignancy in solitary pulmonary nodules (SPNs). Qualitative assessment of FDG uptake in SPNs, using an ordinal scale, integral tothe Herder model, is based on analysis of standard Ordered Subset Expected Maximisation (OSEM) reconstruction PET images. NovelPET reconstructions improve image quality by increasing signal-noise ratio and suppressing image noise. Our aim was to assess theimpact of a novel Bayesian Penalised Likelihood (BPL) PET reconstruction on the Herder risk prediction model of malignancy in SPNsin comparison with standard OSEM images.


Subjects with a SPN who underwent 18F-FDG PET-CT between 2014-2017, with assessable OSEM and BPL reconstructions, andeither histological confirmation of malignancy or histological and/or imaging follow-up confirmation of benignity were included. Tworeaders independently and blindly classified FDG uptake in each SPN on both OSEM and BPL images (BTS score; 1=none; 2=MBP but<2x liver; 4=>2x liver). The BTS score in combination with other clinico-radiological features was used to calculate the Herder riskscore (%) for both OSEM and BPL images.

RESULTS

97 subjects (age 69±10 years, 52% male, 84% current/former smokers, mean nodule size 16±6mm) with 75 (77%) malignant SPNswere included. There was very good inter-observer agreement for the BTS score for both OSEM (κ=0.85) and BPL images (κ=0.87).BPL images increased the BTS score in 25 (26%) SPNs (20 malignant & 5 benign); 9 SPNs (7 malignant) increased from a BTS score2 to 3, and 16 (13 malignant) from a BTS score 3 to 4, with a mean increase of 18±22% in Herder risk score. The mean Herderscore using BPL images was significantly higher than OSEM for all SPNs (73±29 vs 68±32% respectively, p=0.001), and formalignant SPNs (83±19 vs 78±25%, p=0.004), but not for benign SPNs (42±35 vs 37±34%, p=0.07).

CONCLUSION

The use of BPL PET reconstruction increases the Herder score in approximately 25% of SPNs compared to standard OSEM datasetswith the potential to affect subsequent management decisions.


Novel BPL PET reconstruction, compared to standard reconstruction, may increase the estimated risk of malignancy in a SPN, usingthe Herder model, thus potentially affecting management decisions.


ParticipantsHelena You, Dallas, TX (Presenter) Nothing to DiscloseYasemin Sanli, Dallas, TX (Abstract Co-Author) Nothing to DiscloseRathan M. Subramaniam, MD,PhD, Dallas, TX (Abstract Co-Author) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue EarthDiagnostics Ltd

PURPOSE

SSQ14-04 Value of CT Iterative Metal Artifact Reduction in PET/CT: Clinical Evaluation in 103 Patients


68Ga DOTATATE PET/CT has been increasingly used for diagnosis and therapy response assessment of patients withneuroendocrine tumors (NETs). We investigated the impact of point spread function (PSF) reconstruction and lesion size on 68GaDOTATATE PET/CT quantitative parameters.


A total of 38 patients with 42 68Ga DOTATATE PET/CT scans and 125 lesions were included. Scans were reconstructed with andwithout PSF modulation. For each lesion, one reader measured the maximum and peak standardized uptake value (SUVmax andSUVpeak), metabolic tumor volume (MTV), total lesion somatostatin avidity (TLS), and tumor somatostatin receptor expressionheterogeneity (TH) using area under the curve method. Intraclass correlation coefficient (ICC) and Bland-Altman analyses wereused to compare PSF and non-PSF values. Subgroup analysis was performed to determine the impact of lesion size.

RESULTS

Mean age of the patients was 55 ± 15 years. 21 patients were male and 17 were female. Of the 42 scans, 11 were baseline scansand 31 were follow-up scans. Of the 125 lesions, 51 were located in the liver, 31 in lymph nodes, 17 in bone, 8 in pancreas, 4 inlung, and 14 in other sites. Correlation coefficients between PSF and non-PSF values were excellent for SUVmax (ICC=0.97),SUVpeak (ICC=0.99), MTV (ICC=0.98), and TLS (ICC=0.99), and was good for TH (ICC=0.81). Comparison of PSF with non-PSFvalues showed a bias (mean percent change ± SD) of +27.5 ± 14.7% for SUVmax, +15.5 ± 9.5% for SUVpeak, -18.6 ± 37.6% forMTV, +0.8 ± 28.1% for TLS, and -7.1 ± 11.0% for TH. For lesions less than 2 cm in size (n=75), comparison of PSF with non-PSFvalues showed a bias of +32.7 ± 15.8% for SUVmax, +19.3 ± 9.3% for SUVpeak, -27.9 ± 45.4% for MTV, -1.7 ± 35.4% for TLS,and -5.0 ± 12.2% for TH. For lesions 2 cm or more in size (n=50), comparison of PSF with non-PSF values showed a bias of +19.7 ±8.0% for SUVmax, +9.8 ± 6.2% for SUVpeak, +0.01 ± 23.1% for MTV, +4.6 ± 8.8% for TLS, and -10.4 ± 7.9% for TH.

CONCLUSION

PSF and non-PSF values for 68Ga DOTATATE PET/CT quantitative parameters were highly correlated. PSF reconstruction increasedSUVmax and SUVpeak, decreased TH, and had a variable effect on MTV and TLS depending on lesion size.


PSF reconstruction increases SUVmax and SUVpeak and should be considered in evaluating 68Ga DOTATATE PET/CT quantitativeparameters for diagnosis and therapy response assessment of NETs.


ParticipantsChristian P. Reinert, MD, Tuebingen, Germany (Presenter) Nothing to DiscloseChristian la Fougere, Munich, Germany (Abstract Co-Author) Nothing to DiscloseKonstantin Nikolaou, MD, Tuebingen, Germany (Abstract Co-Author) Advisory Panel, Siemens AG; Speakers Bureau, Siemens AG;Speaker Bureau, Bayer AGChristina Pfannenberg, MD, Tuebingen, Germany (Abstract Co-Author) Nothing to DiscloseSergios Gatidis, MD, Tubingen, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To assess the technical feasibility and diagnostic benefit of CT iterative metal artifact reduction (iMAR) in patients with metalimplants undergoing PET/CT.


PET/CT examinations of 103 consecutive patients with metal implants in different localization performed between 10/2017 and03/2018 using a state-of-the-art clinical PET/CT scanner (Siemens Biograph mCT) were included. As PET tracers 18F-FDG(75/103), 68Ga-PSMA (25/103) and 68Ga-DOMITATE (3/103) were used. Diagnostic CT data were reconstructed with iMAR andwithout iMAR (noMAR) and used in comparison for PET attenuation correction, generating iMAR-corrected and noMAR-corrected PETdata. The effect of iMAR on quantitative CT and PET analysis was assessed by HU and SUV measurements in predefined normalanatomical structures and pathological lesions in the vicinity of metal implants. Qualitative diagnostic confidence for lesiondelineation was assessed by 2 radiologists using a 3-point Likert Scale (1=not delineated; 2= fair delineated; 3=good delineated).

RESULTS

For artifact-affected anatomical structures, mean HU of iMAR CT images were significantly different compared to noMAR CT andrespective standard deviations were significantly lower (e.g., M. masseter in case of dental fillings/implants: 105.1 HU, SD 43[noMAR] vs. 72.2 HU, SD 14 [iMAR] P<.01; M. gluteus maximus in case of hip endoprostheses: 79.4 HU, SD 23 vs. 50.0 HU, SD 15;P<.01). However, SUVs did not differ significantly in these artifact-affected anatomical structures (SUVmean 0.90 [iMAR] vs. 0.91[noMAR]; P>.05) and pathological findings (SUVmean 10.65 [iMAR] vs. 10.67 [noMAR]; P>.05) between the iMAR and noMAR PETdata. In the qualitative analysis, a significantly improved delineation of pathologic findings was observed using iMAR in CT for boththe interpretation of physiological (score: 1.23 [noMAR] vs. 2.26 [iMAR]; P<.01) and pathological structures (score: 2.31 [noMAR]to 2.80 [iMAR]; P<.01).

CONCLUSION

The use of iMAR in PET/CT significantly improves delineation of both physiological and pathological structures in the vicinity of metalimplants in CT. The PET quantification and image quality are not significantly affected by the use of iMAR based attenuationcorrection.


SSQ14-05 PET/CT versus PET/MR: Quantitative Accuracy in Y-90 Dosimetry Analysis


SSQ14-06 Evolution of PET/MR Protocols Since 2011: A Single-Center Observational Study Including 1797Examinations


Metal related artifacts impair image quality and increase the risk of missing pathological findings in PET/CT. Lesion delineation isquantitatively and qualitatively improvable by iMAR.

ParticipantsRam Kishore R. Gurajala, MBBS, FRCR, Beachwood, NJ (Presenter) Nothing to DiscloseSara A. Hunter, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseFrank P. DiFilippo, PhD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseKarunakaravel Karuppasamy, MBBS, FRCR, Westlake, OH (Abstract Co-Author) Nothing to DiscloseSankaran Shrikanthan, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseChristopher P. Coppa, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseGaurav Gadodia, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAbraham Levitin, MD, Cleveland, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The purpose of this study is to compare Y-90 dosimetry estimates based on PET/MR versus PET/CT, identify errors in PET/MRdosimetry related to MR-based attenuation correction (AC) and PET detector equipment, and offer methods to avoid these errors.


An IRB-approved prospective study was performed on eight patients receiving Y-90 radioembolization for liver malignancies.Following the intervention, patients were scanned by PET/CT (Siemens Biograph mCT) and PET/MR (Siemens Biograph mMR).PET/CT scans were performed arms-up, while the PET/MR scans were performed either arms-up, arms-down, or both. AC forPET/CT was derived from a low-dose CT scan. AC for PET/MR was performed with three class segmentation using the Dixontechnique. Dosimetry calculations were performed using MIMs 6.5 software (MIM Software Inc.). PET/CT dosimetry was used as thestandard to compare PET/MR dosimetry analysis. Accuracy of PET/MR dosimetry was analyzed in relation to injected activity,background liver and tumor dose, and PET/MR arm location.

RESULTS

PET/MR dosimetry provided accurate dosimetry estimates (within 20% of PET/CT) in the majority of cases. Inaccuracies in PET/MRdosimetry estimates were most pronounced in studies having segmentation errors or truncation errors in the PET/MR AC map,causing inappropriate attenuation correction. Inaccurate PET/MR dosimetry also occurred in cases with high Y-90 injected activity(>3 GBq). Such errors were attributed to the slow characteristics of the Biograph mMR's PET detectors given the high singles ratearising from bremsstrahlung x-rays, leading to inaccurate dead-time correction and increased noise and inaccurate corrections forrandom coincidences. These causes for error can be avoided by ensuring the AC map is accurate, checking for truncation errors,and using Y-90 doses less than 3 GBq.

CONCLUSION

PET/MR can provide accurate Y-90 dosimetry estimates as compared to PET/CT, provided that the injected activity is notexcessive and the MR-based AC map has no major errors. Newer technologies, namely high-speed PET detectors using siliconphotomultipliers and new atlas-based methods for PET/MR AC, are expected to improve accuracy of PET/MR Y-90 dosimetry.


PET/MR can provide accurate Y-90 dosimetry estimates as compared to PET/CT, provided that the injected activity is notexcessive and the MR-based AC map has no major errors.

ParticipantsRegine M. Perl, MD, Tubingen, Germany (Presenter) Nothing to DiscloseChristian P. Reinert, MD, Tuebingen, Germany (Abstract Co-Author) Nothing to DiscloseMareen S. Kraus, MD, Tubingen, Germany (Abstract Co-Author) Nothing to DiscloseJuergen F. Schaefer, MD, Tuebingen, Germany (Abstract Co-Author) Nothing to DiscloseChristina Pfannenberg, MD, Tuebingen, Germany (Abstract Co-Author) Nothing to DiscloseChristian la Fougere, Munich, Germany (Abstract Co-Author) Nothing to DiscloseKonstantin Nikolaou, MD, Tuebingen, Germany (Abstract Co-Author) Advisory Panel, Siemens AG; Speakers Bureau, Siemens AG;Speaker Bureau, Bayer AGSergios Gatidis, MD, Tubingen, Germany (Abstract Co-Author) Nothing to Disclose

PURPOSE

PET/MR is a versatile hybrid imaging modality especially used for oncologic imaging. Since the introduction of combined PET/MRsystems in 2011 this relatively new technology has undergone significant developmental stages and has evolved into a robustclinical modality. The purpose of this study was to record and evaluate the development of clinical PET/MR examinations in ourinstitution with respect to acquisition times, protocol complexity and tracer dosage.


Essential parameters of 1797 clinical PET/MR examinations were recorded in an institutional database between 01/2013 and 12/2017including total examination time, PET acquisition time, number of PET bed positions, number of generated images, injected tracerdose and administration of MR contrast agent. All examinations were conducted on a clinical PET/MR system (Siemens Biograph

SSQ14-07 Feasibility of "Low Dose MR" Dixon Technique for Imaging FDG PET-MR Lymphoma


SSQ14-08 An Unsupervised Dixon-Based Five-Tissue 18F-Sodium Fluoride Synthetic CT Generation for PET/MRAttenuation Correction


mMR, 3 T). PET/MR protocols were iteratively adjusted according to available optimal settings over the observation period. Weevaluated the recorded PET/MR parameters with respect to their development over time and with respect to their variation amongdifferent examination groups (adult patients, pediatric patients and brain studies).

RESULTS

The 1797 examinations included in the final database consisted of 1004 adult patient studies, 278 pediatric patient studies, and 515brain studies. Average examination time decreased significantly between 01/2013 and 12/2017 from 75.7±26.7 to 66.6±23.4 min (P< 0.5). Compared to adult patients, the average pediatric examination time was longer but also significantly shortened between01/2013 and 12/2017 (from 96.8±21.2 min to 84±23.0 min (P < 0.5)). In the same period however, overall examination complexitymeasued by the number of acquired images significantly increased from 2697 to 3696 acquired images per examination (P < 0.01).

CONCLUSION

PET/MR is a complex and time-consuming imaging modality producing a large number of complex image data. Despite increasingprotocol complexity however, examination times were significantly reduced by the introduction of accelerated MR imagingtechniques and protocol optimization.


By optimizing examination protocols PET/MR scan times can be reduced, potentially increasing patient comfort and patientcompliance, which is particularly important when examining children.

ParticipantsMusa A. Mufti, MD, MBBS, Stony Brook, NY (Presenter) Nothing to DiscloseRobert Matthews, MD, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseEzemonye O. Madu, DO, Patchogue, NY (Abstract Co-Author) Nothing to DiscloseKavitha Yaddanapudi, MD, Stony Brook, NY (Abstract Co-Author) Nothing to DiscloseDinko Franceschi, MD, Stony Brook, NY (Abstract Co-Author) Nothing to Disclose

PURPOSE

Clinical whole body PET-MR imaging has wrestled with the problem of acquiring high quality multiplanar MR sequences compared tolower resolution fast MR sequences. "Low dose MRI" is a term used in the nuclear medicine community to describe fast acquiredPET-MR scan protocols that relied heavily on PET images for diagnosis. In this study, we sought to determine if the Dixonsequences obtained for attenuation correction could be used as a diagnostic sequence for interpreting PET-MRI lymphoma cases.


We retrospectively identified 40 patients who underwent 88 FDG PET-MR body imaging studies for staging or restaging lymphoma.Brain images were not reviewed. A radiologist and nuclear medicine physician blindly reviewed PET images, attenuation correctioncoronal Dixon MRI, PET-MR fusion with Dixon, and multisequence (ms) MR, and ms PET-MR images. Lesions were characterizedbased on location, imaging characteristics, size, max SUV, and malignant potency.

RESULTS

All patients were adults with average study age 43.8 y. Studies consisted of 40 females and 48 males with 7 for staging and 81 forrestaging. All patients had systemic lymphoma with 29 being diffuse large B-cell lymphoma. 37 studies had active lymph nodes (LN)on Dixon PET-MR that agreed with ms PET-MR in 33 positive cases (89.1%) having avg SUV 10.2 +/-7.74 SD. 4 Dixon PET-MRcases did not detect lesions, avg SUV 2.3 +/-0.55 SD, read as minimal residual activity. ms MR identified 11 patients with enlargedLN without FDG uptake, not seen on Dixon. All 5 studies with bones lesions were detected by Dixon PET-MR as well as 2 soft tissueorgan lesions. ms MR identified 1 patient with nonactive healed bone lesions. 55 true negative. Compared to ms PET-MR, Dixon had89.2% sensitivity, 100% specificity with no false positive studies.

CONCLUSION

In this retrospective study, Dixon PET-MR was shown to be sensitive and specific compared to ms PET-MR in the detection oflymphoma. Low number of cases not detected had minimally active LN that resolved on subsequent imaging and probably were notclinically important.


Low dose MRI sequences using the Dixon technique for interpretation may a play role in PET-MR imaging when scan time becomesimportant. This may be necessary in patients with comorbidity, claustrophobia, or when multiplanar MR of particular areas benecessary.

ParticipantsJung-Wen Kuo, Cleveland, OH (Presenter) Nothing to DiscloseKuan-Hao Su, Cleveland, OH (Abstract Co-Author) Research Grant, Koninklijke Philips NVAdina N. Crisan, ARRT, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseAtallah Baydoun, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseDaniel J. Bucklan, MD, University Heights, OH (Abstract Co-Author) Nothing to DiscloseBryan J. Traughber, MD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseRobert S. Jones, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseRaymond F. Muzic Jr, PhD, Cleveland, OH (Abstract Co-Author) Research Grant, Koninklijke Philips NV


SSQ14-09 Investigation on PET/MR Image Fusion Mismatch Due to Expanding Bladder: A Pilot Study

Thursday, Nov. 29 11:50AM - 12:00PM Room: S505AB

[email protected]

PURPOSE

To create accurate voxel-wise attenuation correction (AC) maps for PET/MR using Na18F images and obviate the need for bothspecialized MR pulse sequences and conventional methods that typically lack bone information.


Sixteen breast cancer patients received research PET/MRI exams (Philips Ingenuity TF) following clinical Na18F PET/CT exams.Free-breathing 3D T1-weighted (T1W), breath-holding mDixon, and (clinical) low-dose CT images were acquired. Rigid-bodyregistration and local-phase deformable registration were used to transform CT, Dixon-water (Water), and Dixon-fat (Fat) to matchfree-breathing T1W images. A bone feature image was segmented from the non-AC Na18F image. Water, Fat, and bone featureswere classified into five tissue types using fuzzy c-means clustering. CT numbers of -1000, -741, -98, 40, 380 HU were assigned toestimated air, lung, fat, soft tissue, and bone classes, respectively. Synthetic CT (sCT) was generated as a linear combination ofthese. Mean error (ME) and mean absolute error (MAE) were estimated to evaluate the accuracy of the sCT generated by water-filled (WF), three-class T1W-based (3C-T1W), four-class Dixon-based (4C-Dx), four-class deformed Dixon-based (4C-defDx) andfive-class deformed Dixon-Na18F (5C-DxBone) methods. A threshold-based CT bone mask was used to assess the accuracy of thesCT in bone regions.

RESULTS

The MAE of the sCT from WF, 3C-T1W, 4C-Dx, 4C-defDx, 5C-DxBone were 135±8, 133±19, 111±14, 105±11, 103±10 HU,respectively. The 4C-defDx group showed better agreement to measured CT than the conventional 4C-Dx. The ME in the bonemask (MEbone) of the 4C-defDx and 5C-DxBone were -351±27 and -225±29 UH. The 5C-DxBone group presented robust Na18F-derived bone information especially in spine and pelvis; it reduced by 33-37% the MEbone when compared with the other ACmethods and resulted in the lowest ME and MAE.

CONCLUSION

The deformable registration mitigated the mismatch between Na18F and Dixon due to different breathing conditions. The resultssuggest that the Dixon-based sCT can be improved by having a Na18F-derived bone feature to increase the accuracy of PET/MRNa18F quantification.


PET/MR Na18F for bone metastasis detection is not well-established as the lack of bone information for attenuation correction(AC). We propose a feasible five-tissue method for PET/MR Na18F AC.

ParticipantsPengcheng Hu, Shanghai, China (Presenter) Nothing to DiscloseZheng Zhang, MD, Shanghai, China (Abstract Co-Author) Employee, Healthcare Device ManufactorTuoyu Cao, PhD, Houston, TX (Abstract Co-Author) Employee, Medical Device ManufactorLingzhi Hu, PhD, Houston, TX (Abstract Co-Author) Employee, UIH America, IncHongcheng Shi, Shanghai, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

A considerable change of the urinary bladder shape between consecutive bed positions poses a unique challenge on a multi-bedsimultaneous PET/MR scan. Our aim is to report our initial experience on the fusion error caused by expanding bladder.


120 patients (63 males and 57 females, average age = 51.3 years, range 22-70 years) who had been diagnosed with cancer or hadprevious history of cancer were recruited. Each patients were scanned on a simultaneous whole-body PET/MR system with 5 bedpositions (feet in first, 4 bed positions for body and 1 for head). All PET/MR images were visually examined by two independentexperts to evaluate the pelvis fusion accuracy with a Likert scale scoring system (1-5, 5 as the best quality).

RESULTS

The mean and standard deviation of the score is 4.57 and 0.75 correspondingly. 14 patients(11.7%) were rated less or equal to 3by both readers all because of mismatch in the bladder area. This is due to the fact that the bladder area is in the overlap region ofPET images from two consecutive bed positions, so the image of the bladder area is a weighted sum of these two PET imagesbased on the sensitivity curve. Because the bladder expanded significantly between these two bed positions, the average of thetwo very different bladder images cannot match the MRI image from either bed position. This effect is magnified with PET/MRscanning because the scan duration for each bed position is usually significantly longer than that of a PET/CT system due to thelimitation of MRI.

CONCLUSION

Our initial clinical results shows that, in most scenarios PET/MRI can achieve very good image fusion accuracy in the pelvis area.However, it is important to know that expanding bladder might cause mismatch between PET and MRI images when the bladder areais in the PET overlap region of two bed positions. Special care might be needed if there is diagnostic interests of the area nearbladder. This effect can be avoided by arranging bed position accordingly so that bladder is close to the center of one bed position.


This study provides a guideline for simultaneous PET/MR scan protocol to avoid the fusion error in pelvis area due to expandingbladder.

NM236-SD-THA1

Correlation of 11C-4DST Uptake with Isocitrate Dehydrogenase 1 Mutation in Patients with Gliomasin Comparison with 18F- FLT Uptake

Station #1

NM237-SD-THA2

Diagnostic Accuracy of Combined 123I-MIBG Scintigraphy and 123I-FP-CIT SPECT in the DifferentialDiagnosis of Degenerative Parkinsonism

Station #2

NMS-THA

Nuclear Medicine Thursday Poster Discussions

Thursday, Nov. 29 12:15PM - 12:45PM Room: NM Community, Learning Center

NM



ParticipantsDon C. Yoo, MD, E Greenwich, RI (Moderator) Consultant, Endocyte, IncMatthew S. Robertson, MD, Cleveland, OH (Moderator) Nothing to Disclose

Sub-Events

ParticipantsYuka Yamamoto, MD, PhD, Kagawa, Japan (Presenter) Nothing to DiscloseYasukage Takami, Mikicho, Japan (Abstract Co-Author) Nothing to DiscloseTakashi Norikane, Kita-gun, Japan (Abstract Co-Author) Nothing to DiscloseJun Toyohara, Tokyo, Japan (Abstract Co-Author) Nothing to DiscloseYoshihiro Nishiyama, MD, Kagawa, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

Beyond tumor histology, molecular alterations, such as isocitrate dehydrogenase enzyme isoforms 1 (IDH1) and 2 (IDH2) mutationas part of the 2016 world health organization classification of tumors of the central nervous system have been found to provideadditional prognostic value in gliomas. A novel radiopharmaceutical, 4'-[methyl-11C]-thiothymidine (4DST), has been developed asan in vivo cell proliferation marker based on the DNA incorporation method. The purpose of this study was to evaluate 4DST uptakein patients with gliomas and to correlate the results with IDH1 mutation, in comparison with 3'-deoxy-3'-18F-fluorothymidine (FLT).


Investigations of 4DST and FLT PET/CT were performed in 25 patients with glioma. Tumor lesions were identified as areas of focallyincreased uptake, exceeding that of background uptake. For semi-quantitative analysis, the maximal standardized uptake value(SUVmax) for tumor was calculated. The presence of IDH1 mutation in tumor specimens was examined by immunohistochemistryand compared with 4DST SUVmax and FLT SUVmax.

RESULTS

All but one glioma showed focally increased both 4DST and FLT uptake. There was no significant difference between mean (±SD)SUVmax using 4DST (2.61±1.65) and FLT (2.14±1.28). The mean (±SD) 4DST SUVmax of IDH1-mutated tumors (1.72±0.51) wassignificantly lower than that of IDH1-nonmutated tumors (3.21±1.89) (p<0.02). Using FLT, there was no significant differencebetween mean (±SD) SUVmax of IDH1-mutated tumors (1.67±1.16) and IDH1-nonmutated tumors (2.45±1.30).

CONCLUSION

These preliminary results indicate that 4DST PET/CT is feasible for imaging of brain gliomas, as well as FLT PET/CT. 4DST uptake ofIDH1-mutated tumors was significantly lower than that of IDH1-nonmutated tumors.


4DST uptake of IDH1-mutated tumors was significantly lower than that of IDH1-nonmutated tumors. 4DST uptake in gliomas mightbe useful for predicting outcomes.

ParticipantsKaoru Maruyama, Hirakatashi, Japan (Presenter) Nothing to DiscloseKeita Utsunomiya, MD, PhD, Moriguchi, Japan (Abstract Co-Author) Nothing to DiscloseMiho Hatanaka, Osala, Japan (Abstract Co-Author) Nothing to DiscloseHiroshi Kishishita, Moriguchi, Japan (Abstract Co-Author) Nothing to DiscloseNaoki Kan, Hirakata, Japan (Abstract Co-Author) Nothing to DiscloseYumiko Kono, MD, Hirakata, Japan (Abstract Co-Author) Nothing to DiscloseYasuhiro Ueno, Moriguchi, Japan (Abstract Co-Author) Nothing to DiscloseNoboru Tanigawa, MD, Hirakata, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

To investigate the sensitivity and specificity, in the clinically based diagnostic criteria for Parkinson's disease (PD), of planar heart

NM238-SD-THA3

18F-FDG PET/CT Postoperative Changes after Maxillectomy: Findings and Pitfalls in Interpretation

Station #3

NM239-SD- Impact of Patient Comfort on Diagnostic Image Quality during PET/MR Exam: An Objective Survey

imaging with the ligand 123I-Metaiodobenzylguanidine (MIBG), and of single photon emission computed tomography (SPECT) braindopamine transporter imaging with the ligand 123I-Ioflupane (FP-CIT), and to evaluate to which extent diagnostic accuracy can beincreased by their combined use together.


Twenty seven patients who performed both methods for differential diagnosis between PD (n = 21) and other parkinsonism (non-PD,n = 6) were enrolled. Readers, unaware of the clinical diagnosis, classified the images as normal or abnormal by visually inspection.Abnormal MIBG or FP-CIT scan is defined as decreased uptake in the myocardium or in the striatum, respectively. Sensitivity,specificity, accuracy, positive and negative values of both methods were calculated.

RESULTS

Abnormal MIBG scans had a sensitivity of 57 % (12/21) for detecting clinical PD, with specificity of 100 % (6/6) for excluding non-PD. For 123I-MIBG scintigraphy, a value of 67 % (18/27) was achieved for overall diagnostic accuracy, 100 % (12/12) for positivepredictive value, and 40 % for negative predictive value. For 123I-FP-CIT SPECT, the overall sensitivity, specificity, accuracy,positive and negative predictive values in PD were 81 % (17/21), 0 % (0/6), 63 % (17/27), 74 % (17/23), and 0 % (0/4),respectively. For combined use of both normal MIBG and abnormal FP-CIT, the highest test accuracy of 81 % for differentiation ofnon-PD from PD with a sensitivity of 100 %, specificity of 76 %.

CONCLUSION

The diagnostic accuracy is sufficiently high for the combination of cardiac 123I-MIBG scintigraphy and brain 123I-FP-CIT SPECT tobe clinically useful in distinguishing PD from non-PD.


PD usually present both myocardial sympathetic and striatal dopaminergic impairments. The combined use of both 123I-FP-CITSPECT and 123I-MIBG scintigraphy may contribute to have a complementary role in differential diagnosis between PD and non-PD.

ParticipantsTima Davidson, MD, Ramat Gan, Israel (Presenter) Nothing to DiscloseJohnatan Nissan, Haifa, Israel (Abstract Co-Author) Nothing to DiscloseMaria Krichmar, Tel-Hashomer, Ramat Gan, Israel (Abstract Co-Author) Nothing to DiscloseEyal Lotan, MD, New-York, NY (Abstract Co-Author) Nothing to DiscloseShai Shrot, Raanana, Israel (Abstract Co-Author) Nothing to DiscloseElla Nissan, Tel-Hashomer, Ramat Gan, Israel (Abstract Co-Author) Nothing to DiscloseIris Gluck, Tel-Hashomer, Ramat Gan, Israel (Abstract Co-Author) Nothing to DisclosePhilip R. Lawson, MBChB,BSC, Ramat Gan, Israel (Abstract Co-Author) Nothing to DiscloseSimona Ben-Haim, MD, DSc, Ramat Gan, Israel (Abstract Co-Author) Consultant, Spectrum Dynamics Ltd Consultant, MolecularDynamics Spouse, Stockholder, Molecular DynamicsShay Duvdevani, Tel-Hashomer, Ramat Gan, Israel (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

PET-CT is the mainstay of disease surveillance in head and neck cancer following local resection. However, increased FDG-uptake inthe surgical bed may be non-specific and present diagnostic challenges.


This is a review of FDG- PET/CT studies of head and neck cancer patients with a history of maxillary surgery treated during 2008-2016. Findings on postoperative PET/CT were correlated with clinical and imaging follow-up.

RESULTS

PET/CT studies of 17 reconstructive maxillary surgeries of 14 patients (10 males, mean age 57 ± 16 years, range 22-77) werereviewed. Increased FDG uptake in the postoperative bed was demonstrated in 12 (71%; 10 obturator, 2 mesh reconstructions),mean SUV max 2.4 ± 1.4 (range 0.3-4.3). In the remaining 5/17 studies (3 with a fat flap and 2 without any reconstructions), therewas no abnormal FDG uptake at the postoperative bed. CT features in postoperative sites included a combination of: non-homogeneous mixed hyperdense material with multiple air bubbles; mucosal thickening along the postoperative bed wall (in all caseswith obturator implants); rich fat density material in reconstructions with a fat flap and in closures without reconstruction;radiopaque hyperdense elongated structures in mesh reconstructions. In 8/9 patients with more than one postoperative PET/CT,there were no significant differences between the studies. One patient had a new FDG-avid lesion in the nasopharynx adjacent tothe surgical bed, which was confirmed by pathology as recurrence of the disease. No correlation was found of the mean SUV max ininitial scans with the time from the surgery date (10 months; ± 6; r=0.04, P=0.90), or with the mean SUV max in final scans (25months; ± 17; 2.4 ± 1.4 vs. 1.6 ± 1.7; P=0.17). In 4 patients, biopsies obtained from the FDG-avid changes showed fragments offibrotic or granulation tissue, with no evidence of malignancy.

CONCLUSION

All patients in this study who had obturator or mesh reconstructive surgery after maxillectomy demonstrated FDG avidpostoperative changes, which persisted more than four years after surgery. Awareness of variations in postoperative PET/CTappearance is important to avoid false interpretation in cancer patients.


Radiologists should be aware of variations in postoperative PET-CT appearance, including increased FDG uptake, even years aftersurgery, to avoid overcalling these lesions as malignance.

THA4 Study for Clinical Workflow Management

Station #4

NM240-SD-THA5

Evaluating the Role of 68Ga-PSMA PET/CT Imaging in Predicting Intraprostatic Tumor Extent ofTumor Involvement Prior to Various Surgical Protocols in Patients with Known or DiagnosedProstatic Cancer

Station #5

ParticipantsShuguang Chen, MD, Shanghai, China (Abstract Co-Author) Nothing to DiscloseZheng Zhang, MD, Shanghai, China (Abstract Co-Author) Employee, Healthcare Device ManufactorGuo Bin, Shanghai, China (Abstract Co-Author) Employee, Medical Deivce ManufactorTuoyu Cao, PhD, Houston, TX (Abstract Co-Author) Employee, Medical Device ManufactorLingzhi Hu, PhD, Houston, TX (Abstract Co-Author) Employee, UIH America, IncHongcheng Shi, Shanghai, China (Abstract Co-Author) Nothing to Disclose

PURPOSE

Whole body PET/MR screening usually takes 30-50 minutes to finish, during which a few factors might induce patient discomfort andfurther cause degraded image quality. The aim of this report is to investigate the patients' perception of the imaging procedure andits correlation with image quality.


120 patients (63 males and 57 females, average age = 51.3 years, range 22-70 years) who had been diagnosed with cancer or hadprevious history of cancer were recruited and scanned with a simultaneous PET/MR system. A questionnaire was given to allpatients retrospectively after the PET/MR scan, which has 9 questions to assess patients' feeling of the scan on a Likert scalescoring system (1-5, 1 as most satisfied). All PET/MR images were also visually examined by two experts independently to evaluatethe quality of the images. Six body positions were assessed and each position was evaluated also with a Likert scale scoringsystem (1-5, 5 as the best quality). With each section of the patient comfort questionnaire, patients were divided into two groupsbased on the score they filled (Group A: 1,2; Group B: 3, 4 ,5). And statistical analysis using Mann-Whitney U-test were used tocheck if there is significant difference between these two groups regarding PET image artifacts, MR image artifacts and PET/MRcoregistration error.

RESULTS

118 questionnaires were filled and returned for analysis. The average overall perception is 1.36. The statistical test showed thatpatients' comfort has strong correlation with MR image artifacts, especially in certain areas of the body such as head and pelvis.And some factors had a bigger impact than others, such as RF heating and coil pressure. No statistically significant difference wereobserved on PET image artifacts or coregistration error due to patients comfort.

CONCLUSION

With proper preparation and patient training, it is practical to achieve good image qualities from simultaneous PET/MR scan, evenwith the longer scan duration compared to other modalities. However, it is still important to pay special attention to patientscomfort for the best performance.


This study provides a guideline for simultaneous PET/MR scan protocol to achieve the optimum image quality.



[email protected]

PURPOSE

The role of performance of gallium Ga 68 (68Ga)-labeled prostate-specific membrane antigen (PSMA) ligand positron emissiontomography/computed tomography (PET/CT) shows very promising results with known or diagnosed prostatic cancers.Theevaluation in relation to intraprostatic involvement.local infiltration and lymph node detection even at low prostate specific antigen(PSA) levels is highly sensitive.In patients with biochemical recurrence after curative therapy for prostate cancer. In this study weevaluated the usefulness of PSMA-PET/CT for the evaluation of intraprostatic tumor extent prior to surgery


This was the retrospective study with diagnosed prostate cancer.30 patients were included in this study who underwent a 68Ga-PSMA-PET/CT before surgical treatment. All patients underwent either open or laparoscopic radical prostatectomy. Intraprostatictumor extent was assessed on PET/CT by the peak standardized uptake value (SUV) .This was correlated with final whole glandhistopathology including Gleason grade, lobe infiltration, TNM stage and seminal vesicle invasion.

RESULTS

The mean SUV over all patients was 12.64 ±9.19. Median Gleason score was 7. The distribution of TNM stage 2a, 2b, 2c, 3a, 3b, 4was 0, 2, 6, 5, 2 and 0 patients. Mean intraprostatic SUV for patients with organ confined vs. extraprostatic tumor onhistopathology was 6.3 and 15 respectively (p = 0.043). Mean SUV for patients with a Gleason score of <= 7a and > 7a was 6.2and 13.5 (p = 0.046). Sensitivity and specificity for identifying tumor invasion of a prostate lobe was 85.6% and 76% (Positive andnegative predictive value (PPV, NPV) for prostate lobe invasion were 95.7% and 42.9%). Sensitivity and specificity for seminalvesicle invasion were 50% and 100% (PPV 100%, NPV 96.6%).

CONCLUSION

Thus the PSMA PET-CT has very effective value in evaluation of intra and extra prostatic disease involvement.


NM149-ED-THA6

'A Burning Issue:" The Utilization of Tc-99m MDP Bone Scan with SPECT/CT in the Burn Unit

Station #6

PSMA PET-CT is highly sensitive of prostatic cancer in comparison to F FDG PET-CT also

ParticipantsDavid Sadowsky, MD, Larchmont, NY (Presenter) Nothing to DiscloseCharles Lugo, MD, Valhalla, NY (Abstract Co-Author) Nothing to DiscloseJennifer Wu, MD, MPH, Valhalla, NY (Abstract Co-Author) Nothing to DiscloseChaitanya Shilagani, MD, Valhalla, NY (Abstract Co-Author) Nothing to DisclosePerry S. Gerard, MD, Valhalla, NY (Abstract Co-Author) Nothing to Disclose


[email protected]

TEACHING POINTS

Patients with severe burn injuries often endure prolonged hospital courses and undergo multiple debridements and amputations.While SPECT/CT protocols exist for a wide variety of clinical indications and employ multiple radioisotopes, current use in the field ofburn care remains extremely limited. The combined ability of three phase Tc-99m MDP SPECT/CT to distinguish between viable andnonviable tissue and provide precise anatomic localization holds promise for reducing repeat procedures and increasing preservationof function in this population.


1- Introduction to burn injuries and common complications and hospital courses of burn patients 2- Amputation background (tissueviability, preservation of function) and special amputation considerations in the burn patient (multiple sites, disfigurement, activesoft tissue infection, prolonged recovery) 3- Example cases of SPECT/CT imaging applied to heat and cold burns, highlighting thevalue added by this imaging modality compared to traditional methods

NM241-SD-THB1

The Iodine/FDG "Flip-Flop" Phenomenon is Ambiguous in Bone Metastasis from Thyroid Cancer:Significant FDG Uptake is Observed Even in Iodine-Positive Lesions

Station #1

NM242-SD-THB2

Size of Detectable Melanoma Metastases to Brain

Station #2

NMS-THB

Nuclear Medicine Thursday Poster Discussions

Thursday, Nov. 29 12:45PM - 1:15PM Room: NM Community, Learning Center

NM



ParticipantsMatthew S. Robertson, MD, Cleveland, OH (Moderator) Nothing to DiscloseDon C. Yoo, MD, E Greenwich, RI (Moderator) Consultant, Endocyte, Inc

Sub-Events

ParticipantsTakuro Isoda, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseShingo Baba, Fukuoka, Japan (Presenter) Nothing to DiscloseYoshiyuki Kitamura, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseRyo Somehara, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseKeiichiro Tahara, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseMasayuki Sasaki, Fukuoka, Japan (Abstract Co-Author) Nothing to DiscloseHiroshi Honda, MD, Fukuoka, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

Radioiodine therapy is used to treat distant metastases from thyroid cancer. A reverse relationship between iodine and FDGaccumulation is found in thyroid cancer lesions, the so-called "flip-flop" phenomenon. The aim of this study was to assess therelationship between iodine and FDG uptake in bone metastasis lesions from thyroid cancer.


The cases of 20 patients who underwent radioiodine therapy for bone metastasis were studied retrospectively (age, 27-74 yrs;median, 62 yrs; males : females, 6 : 14). Iodine uptake was evaluated visually and the lesions showing iodine uptake more thanbackground were determined as positive. FDG uptake was assessed using SUVmax. We compared FDG uptake between the lesionswith (n=19) and without (n=88) iodine uptake in bone metastasis.

RESULTS

The bone metastasis lesions without iodine showed significantly higher FDG uptake than those with iodine uptake (p < 0.0001).However, the degree of FDG uptake in bone metastasis lesions was also relatively high (Median of SUVmax: 4.64). Thirty-nine outof 88 iodine-positive bone metastasis lesions (44.3 %) showed more than 5 of SUVmax. On the other hand, only one iodine-positivelesion (1.1 %) showed more than 10 of SUVmax, while 9 out of 19 (47.4 %) iodine-negative lesions did.:

CONCLUSION

Bone metastasis lesion from thyroid cancer not showing iodine uptake showed higher FDG uptake compared to those with iodineuptake. However, more than 40 % of bone metastasis lesions with iodine uptake showed relatively high FDG uptake (SUVmax > 5).


The result of this study suggests that the patients with bone metastasis from thyroid cancer have chance to receive benefits fromradioiodine therapy even when their bone metastasis lesions are FDG-avid and that it is recommended to perform radioiodinetherapy whether or not bone metastasis lesions show FDG uptake.

ParticipantsJorge D. Oldan, MD, Cleveland, OH (Presenter) Nothing to DiscloseValerie L. Jewells, DO, Chapel Hill, NC (Abstract Co-Author) Nothing to DiscloseAmir H. Khandani, MD, Chapel Hill, NC (Abstract Co-Author) Consultant, Progenics Pharmaceuticals, Inc; Consultant, F. Hoffmann-LaRoche Ltd;


[email protected]

PURPOSE

The decreased sensitivity of FDG-PET for metastases to the brain, due to high physiologic background uptake, is well documented.We aimed to determine the size at which melanoma metastases become reliably visible on PET. (Melanoma was selected as it is one

NM243-SD-THB3

Evaluating the Role of 18F-FDG PET/CT in Diabetic Foot

Station #3

NM244-SD-THB4

Comparison of TNM Stage Evaluation Capability among Whole-Body MRI at 1.5T and 3T MR Systems,PET/MRI at 1.5T and 3T MR Systems and PET/CT in Non-small Cell Lung Cancer Patients

Station #4

of the few malignancies for which vertex-to-toe imaging is commonly done and which frequently metastasizes to brain.)


A retrospective search over a period of 64 months was performed to identify whole-body PET/CT (including bed position of thecranium with a low-dose head CT) with at least one brain MRI within 1 month of the study. 294 cases were identified and the PETand MRI re-examined by a nuclear medicine physician and neuroradiologist, each blinded to history and the other examination.Results were compared.

RESULTS

Despite the generally reputed low sensitivity of PET for intracranial metastases, a fair number of tumors were found even on a PETprotocol not optimized for brain imaging. Sensitivity was 100% over 3.0 cm, 67% for tumors 2.0-2.9 cm, 42% for tumors 1.0-1.9cm, and 5% for tumors below 1 cm. Specificity was 96%, likely due to the low prevalence. Tumors in locations such as the pituitarywhere they were not surrounded by intensely avid brain were detected when they were present. Sensitivities are somewhat higherfor hot than cold lesions.

CONCLUSION

PET has at least partial sensitivity for brain tumors in the 1-3 cm range even without a dedicated protocol. If there is seriousconcern for smaller brain metastases, a brain MRI should of course be performed, but a whole-body PET/CT protocol for melanomashould at least include the brain.


PET can detect at least some brain metastases from melanoma in the 1-3 cm range. Whole-body melanoma studies should includethe cranium and be correlated with MRI of the brain for optimal detection of metastases.



[email protected]

PURPOSE

Diabetes is one of the important and commonest disease entity in Indain subcontinent Infections superadded with diabetic foot isone of the commonest pathological entity .Osteomyelitis represents upto third of diabetic foot infections, is often due to directcontamination from a soft-tissue lesion, and represents a clinical challenge. Early diagnosis is important since antibiotic therapy canbe curative and may prevent amputation. The present study assessed the role of PET/CT using 18F-FDG for the diagnosis ofdiabetic foot osteomyelitis.


Twenty four diabetic patients (14 men and 10 women; age range, 29-70 y) with 28 clinically suspected sites of infection underwentwhole body and lower limb protocol PET/CT after the injection of 185-370 MBq of 18F-FDG for suspected osteomyelitis complicatingdiabetic foot disease. PET, CT, and fused images were independently evaluated for the diagnosis and localization of an infectiousprocess. Additional data provided by PET/CT for localization of infection in the bone or soft tissues were documented. The finaldiagnosis was based on histopathologic findings and bacteriologic/culture assays obtained at surgery or at clinical and imagingfollow-up.

RESULTS

PET detected 28foci of increased 18F-FDG uptake suspected as infection in 20 patients. PET/CT correctly localized 16 foci in 8patients to bone, indicating osteomyelitis. PET/CT correctly excluded osteomyelitis in 10 foci in 10 patients, with the abnormal 18F-FDG uptake limited to infected soft tissues only. One site of mildly increased focal 18F-FDG uptake was localized by PET/CT todiabetic osteoarthropathy changes demonstrated on CT. Eight patients showed no abnormally increased 18F-FDG uptake and nofurther evidence of an infectious process on clinical and imaging follow-up.

CONCLUSION

Thus 18F-FDG PET can be used for diagnosis of diabetes-related infection. The precise anatomic localization of increased 18F-FDGuptake provided by PET/CT enables accurate differentiation between osteomyelitis and soft-tissue infection.


Thus FDG PET-CT is evolving as better modality for evaluation of Diabetis

ParticipantsYoshiharu Ohno, MD, PhD, Kobe, Japan (Presenter) Research Grant, Canon Medical Systems Corporation; Research Grant,Koninklijke Philips NV; Research Grant, Bayer AG; Research Grant, DAIICHI SANKYO Group; Research Grant, Fuji Pharma Co, Ltd;Research Grant, Guerbet SA; Yuji Kishida, MD,PhD, Kobe, Japan (Abstract Co-Author) Nothing to DiscloseShinichiro Seki, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationTakeshi Yoshikawa, MD, Kobe, Japan (Abstract Co-Author) Research Grant, Canon Medical Systems CorporationKota Aoyagi, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationMasao Yui, Otawara, Japan (Abstract Co-Author) Employee, Canon Medical Systems CorporationHisanobu Koyama, MD,PhD, Osaka, Japan (Abstract Co-Author) Nothing to DiscloseTakamichi Murakami, MD, PhD, Osakasayama, Japan (Abstract Co-Author) Nothing to Disclose

NM245-SD-THB5

Hyperintense Rim Sign on MRI STIR as a Diagnostic Correlate of Dermal Backflow onLymphoscintigraphy in Patients with Upper Extremity Lymphedema

Station #5


[email protected]

PURPOSE

To prospectively and directly compare TNM stage classification capability among whole-body MRI and PET/MRI at 1.5 and 3T MRsystems and PET/CT in non-small cell lung cancer patients.


104 consecutive pathologically diagnosed NSCLC patients (62 men, 42 women; mean age 71 years) prospectively underwent whole-body MRI at1.5T and 3T systems, integrated PET/CT, and surgical, pathological and/ or follow-up examinations. Final diagnoses ofT, N and M factors and clinical stage in each patient were determined according to all examination results. All co-registeredPET/MRIs were generated by means of our proprietary software. Then, each factor and clinical stage were visually assessed onboth whole-body MRIs, PET/MRIs and PET/CT with contrast-enhanced brain MRI. Kappa statistics were used to determineagreements for assessment of all factors and clinical stage with final diagnoses, and McNemar's test was used to compare eachdiagnostic accuracy among all methods.

RESULTS

On T factor assessment, agreement between each method and final diagnosis was almost perfect (0.92<κ<0.98). For N factorevaluation, agreement of each method and final diagnosis was substantial (0.60<κ<0.80). On M factor and clinical stageassessments, agreements of all methods were substantial or almost perfect (0.71<κ<0.88). Diagnostic accuracies of N factor onwhole-body MRI at 1.5T (83.7 [87/104] %, p=0.02) and 3T (86.5 [90/104] %, p=0.002) systems as well as PET/MRI at 1.5T (82.7[86/104] %, p=0.03) and 3T (84.6 [88/104] %, p=0.008) systems were significantly higher than that of PET/CT (76.9 [80/104] %).In addition, diagnostic accuracy of clinical stage on whole-body MRI at 1.5T (83.7 [87/104] %, p=0.008) and 3T (88.5 [92/104] %,p=0.0002) systems as well as PET/MRI at 1.5T (83.7 [87/104] %, p=0.008) and 3T (86.5 [90/104] %, p=0.001) systems weresignificantly higher than that of PET/CT (76.0 [79/104] %).

CONCLUSION

Whole-body MRIs and PET/MRIs at 1.5T and 3T systems have significantly better potential for N factor and clinical stageassessments than PET/CT in NSCLC patients.


Whole-body MRIs and PET/MRIs at 1.5T and 3T systems have significantly better potential for N factor and clinical stageassessments than PET/CT in NSCLC patients.


ParticipantsGeunwon Kim, MD,PhD, Boston, MA (Presenter) Nothing to DiscloseMartin P. Smith, MD, Newton, MA (Abstract Co-Author) Research Grant, Bracco Group Research Grant, Bayer AG Consultant, BayerAG Research Consultant, General Electric CompanyKevin J. Donohoe, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseDhruv Singhal, MD, Boston, MA (Abstract Co-Author) Nothing to DiscloseLeo L. Tsai, MD, PhD, Boston, MA (Abstract Co-Author) Co-founder, Agile Devices Inc Stockholder, Agile Devices Inc ResearchConsultant, Agile Devices Inc


[email protected]

PURPOSE

Accurate diagnostic imaging of lymphedema remains challenging. Here we aim to identify MRI correlates of dermal backflow, afinding seen on lymphoscintigraphy that serves as a specific diagnostic indicator for lymphedema.


21 patients referred from lymphedema clinic underwent nuclear lymphoscintigraphy and upper extremity MRI between March andOctober 2017. Both upper extremities were imaged in two stations. STIR sequence was performed at both stations.Lymphoscintigraphy is performed within 24 hours of MRI and all other clinical measures. Two intradermal Tc-99m labelledtilmanocept injections proximal to second and third webspaces are administered, with a total dose of ~0.4 mCi. Large field of viewdetector with parallel collimators was used at 140 KeV with a 20% window. Cobalt-57 is used as a flood source. Flow images aretaken at 30 s/frame for 20 min. The presence of dermal back flow and lymphatic flow are assessed on the static images of theextremities at 1 hour and a minimum delay of 2 hours and up to 6 hours, along with transmission images for localization. MRI imageswere analyzed using McKesson PACS and lymphoscintigraphy using MIM Software. Agreement between MRI findings andlymphoscintigraphy was analyzed using positive and negative percentage agreement with 95% confidence interval.

RESULTS

Imaging studies were retrospectively reviewed. A thin rim of hyperintensity along the peripheral subcutaneous layer on STIR imageswas noted in 12 out of 14 patients that demonstrated dermal backflow on lymphoscintigraphy, with matching distributions. Theother 2 patients showed hyperintense rim sign extending beyond the area seen on lymphoscintigraphy. 1 patient showedhyperintense rim sign on STIR with no dermal backflow on lymphoscintigraphy. Overall, the distribution of edema was mostly inposterior aspect of the upper and ulnar side of the lower arm centered around the elbow, and circumferential if severe. No STIRhyperintensity was seen in any of the 21 contralateral normal limbs or in the six affected limbs that did not demonstrate dermalbackflow on lymphscintigraphy. Positive and negative percent agreement were 80.0% [51.3%, 94.7%] and 100% [84.5%, 100%],respectively.

NM150-ED-THB6

Theranostic Imaging of Prostate Cancers: 68Ga PSMA PET/CT / 177Lu PSMA Radionuclide Therapy

Station #6

CONCLUSION

Hyperdense rim is an MRI correlate for dermal backflow which is a lymphoscintigraphy finding.


MRI is can detect dermal backflow and would provide anatomic detail to the lymphoscintigraphic finding.

ParticipantsYasemin Sanli, Dallas, TX (Abstract Co-Author) Nothing to DiscloseDuygu Simsek, MD, Istanbul, Turkey (Presenter) Nothing to DiscloseSerkan Kuyumcu, MD, Istanbul, Turkey (Abstract Co-Author) Nothing to DiscloseCuneyt Turkmen, MD, Istanbul, Turkey (Abstract Co-Author) Nothing to DiscloseRathan M. Subramaniam, MD,PhD, Dallas, TX (Abstract Co-Author) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue EarthDiagnostics Ltd


[email protected]

TEACHING POINTS

Prostat cancer (PC) is one of the most prevalent cancers worldwide. Although most of the patients have localized disease atdiagnosis, distant metastasis can occur 35% of PC cases during the follow up. 68Ga PSMA PET/CT is a powerful imaging method forthe imaging of PC, using prostate specific membrane antigen (PSMA), a transmembranous enzyme which significantly overexpressedin the majority of PC. 68Ga PSMA PET/CT has remarkable results in detecting PC compared to other PET radiotracers. 68Ga PSMAPET/CT has the potential to influence the management of patients with PC and help patient selection for 177Lu PSMA, an importanttherapy option with superior response rates for metastatic castration resistant prostate cancer comparing other systemic therapies


Value of 68Ga PSMA PET/CT imaging; - in staging intermediate or high-risk PC - in biochemically recurrent PC - in therapy responseof PC - in 177Lu PSMA therapy planning Physician Learning: Illustrative cases to demonstrate physiological distribution, artifactsand pitfalls in 68Ga PSMA imaging. NonPC tumors with 68Ga PSMA imaging. 177Lu PSMA therapy implementation for PC.Future:Combination of 68Ga PSMA PET/MRI with functional MRI sequences can provide additional information by excellentanatomical resolution of suspicious lesions and be used for targeted biopsy.

SPSH52A Personalized Medicine and Lung Cancer Biomarkers: The Oncologist's Perspective

SPSH52B Imaging Biomarkers in Non-small Cell Lung Cancer

SPSH52C Using Artificial Intelligence to Develop Non-invasive Biomarkers in Lung Cancer

SPSH52

Hot Topic Session: Biomarker and Personalized Medicine in Lung Cancer Imaging

Thursday, Nov. 29 3:00PM - 4:00PM Room: E350

AI BQ CH CT NM


ParticipantsPatricia M. de Groot, MD, Houston, TX (Moderator) Nothing to Disclose

Sub-Events

ParticipantsJohn V. Heymach, MD,PhD, Houston, TX (Presenter) Consultant, AstraZeneca PLC; Consultant, Boehringer Ingelheim GmbH;Consultant, Merck KGaA; Consultant, F. Hoffmann-La Roche Ltd; Consultant, Eli Lilly and Company; Consultant, Merck & Co, Inc;Consultant, Spectrum Dynamics Ltd; Consultant, Guardant; Consultant, Johnson & Johnson; Consultant, Novartis AG

LEARNING OBJECTIVES

1) Describe the goals and current state of personalized therapy for patients with non-small cell lung cancer. 2) Identify the lungcancer biomarkers now in clinical use as well as those in experimental trials. 3) Understand the barriers to optimal selection ofindividual patient therapy from the clinical and basic research perspective.

ParticipantsBrett W. Carter, MD, Houston, TX (Presenter) Editor, Reed Elsevier;


[email protected]

LEARNING OBJECTIVES

1) Identify the imaging manifestations and patterns of disease associated with specific non-small cell lung cancer genetic mutationssuch as EGFR and KRAS and rearrangements such as ALK on computed tomography (CT) and FDG positron emission tomography(PET)/CT. 2) Describe the role of established response criteria and emerging and novel imaging techniques on the assessment oftreatment response in non-small cell lung cancer. 2) Understand the continuously evolving impact of radiogenomics, defined as thelinking of medical images with the genomic properties of neoplasms, in predicting the presence of specific genetic alterations,response to therapy, and survival of patients with non-small cell lung cancer.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Brett W. Carter, MD - 2015 Honored EducatorBrett W. Carter, MD - 2018 HonoredEducator

ParticipantsHugo Aerts, PhD, Boston, MA (Presenter) Stockholder, Sphera Inc

LEARNING OBJECTIVES

1) Learn about the motivation and methodology of AI technologies in lung cancer imaging. 2) Learn about scientific studiesinvestigating the role of radiologic AI with other -omics data for precision medicine. 3) Learn about open-source informaticsdevelopments.

SPSH54A The Unmet Imaging Needs in Immunotheraphy Drug Development

SPSH54B The Role of PET-CT in Immunotherapy

SPSH54C Imaging of Tumor Associated Macrophages with Ferumoxytol-enhanced MRI

SPSH54

Hot Topic Session: Immunotherapy for Cancer-A Demanding New Imaging Frontier

Thursday, Nov. 29 3:00PM - 4:00PM Room: S503AB

CT MR NM



ParticipantsJanet F. Eary, MD, Bethesda, MD (Moderator) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Understand the unique role that quantitative clinical imaging plays as costly, highly effective immunotherapies are increasinglyapproved by FDA for advanced cancer treatments and their dependence on imaging. 2) Learn the pitfalls confronting jointoncologist - imager teams can expect to encounter during the time course of cancer treatments and the need to modify existingintellectual models used by clinical cancer imagers. 3) By showing specific case examples that distinguish immune cancertreatments from conventional cytotoxic chemotherapies, attendees will comprehend the need to carefully exercise and reservejudgment when analyzing treatment care patterns.

Sub-Events

ParticipantsElad Sharon, MD, Bethesda, MD (Presenter) Nothing to Disclose

LEARNING OBJECTIVES


ParticipantsRegis Otaviano Bezerra, MD, Sao Paulo, Brazil (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Brief overview of basic concepts in immunotherapy. 2) To review key points of imaging response patterns according to IRECIST:partial response, complete response, pseudo-progression, and disease progression. 3) To illustrate PET-CT cases with emphasis inresponse patterns. 4) To recognize immune-related adverse events: pneumonitis, hepatitis, pancreatitis, colitis,, and nephritis. 5)Others uncommon events: abscopal effect and sarcoid-like reaction.

ABSTRACT

Evidence is now accumulating on the use of functional imaging for evaluation of immunotherapy, once this is a treatment option fora wide range of FDG-avid:tumors, including: melanoma, colorectal, breast,and lung cancers. New response patterns have arisenwith the use of this treatment, which currently represent a challenge for conventional imaging and standard therapy assessmentcriteria. A delayed response and even a transient tumor enlargement are often seen, and can be misdiagnosed as diseaseprogression.Moreover, unusual side effects may occur and these can be subtle or very difficult to recognize in regular anatomicalimages, such as CT or MRI. Especially in this scenario, PET/CT plays a pivotal role, since adverse events might be detected earlier,even preceding clinical symptoms.The most common immune-related adverse events such as pneumonitis, hepatitis, pancreatitisand nephritis will be discussed. Another out-of-target tumor response is demonstrated, known as the abscopal effect, part ofimmunomodulation effect of check point inhibitors. This presentation will demonstrate initial assessment; response patternsfollowing immunotherapy, and tumor recurrence using PET/CT illustrative cases. Perspectives on new imaging probes will beaddressed as well.

URL

https://www.hospitalsiriolibanes.org.br/Paginas/default.aspx

ParticipantsHeike E. Daldrup-Link, MD, Palo Alto, CA (Presenter) Nothing to Disclose

SPSH54D Assessment of Tumor Dynamics beyond RECIST


[email protected]

LEARNING OBJECTIVES

1) To understand basic principles of tumor immune responses. 2) To learn MR imaging approaches for the detection of tumorassociated macrophages (TAM) in patients. 3) To learn, how to assess tumor response to TAM-modulating cancer immunotherapies.

ABSTRACT

Many malignant tumors, including breast cancer, lung cancer, colon cancer, pancreatic cancer, lymphomas, sarcomas andneuroblastomas (among many others) are associated with an anti-inflammatory tumor microenvironment, which is characterized byinfiltration of leukocytes where increases in some leukocyte subsets parallels disease progression and worse clinical outcomes.Tumor-associated macrophages (TAMs) play a key role in this context. New therapeutic drugs that target TAM are currently beingdeveloped and are starting to enter the clinic. Therefore, it becomes increasingly important to identify patients whose tumors areheavily infiltrated by TAM. To serve this goal, an imaging test is advantageous over invasive biopsy because it is non-invasive, cancover the whole tumor and can interrogate treatment effects repeatedly in vivo, in patients. Unfortunately, existing clinical imagingtests do not provide a good method for evaluating response to immunotherapies because most immune modulating therapeutics donot cause changes in tumor size, at least not in the immediate post-treatment time period. This presentation will show how TAMcan be detected with an immediately clinically applicable imaging approach, using the FDA-approved iron supplement ferumoxytol offlabel as an MR contrast agent. Our team showed that ferumoxytol nanoparticles exert an initial perfusion effect of the tumor tissue,followed by retention in the tumor via the 'enhanced permeability and retention (EPR) effect' and subsequent phagocytosis by TAM,which results in a marked negative (dark) signal effect on delayed T2-weighted MR images. This can be used to accurately andnoninvasively track the degree of macrophage infiltration in a tumor tissue. This new TAM imaging test could represent a significantbreakthrough for clinicians as a new biomarker for risk stratification and for monitoring tumor response to novel TAM-modulatingimmunotherapies. Ferumoxytol is particularly suited for tracking cancer immune responses due to its intrinsic pro-inflammatoryeffects on the cancer microenvironment.

URL

http://daldrup-link-lab.stanford.edu/

ParticipantsSean Khozin, MD, Silver Spring, MD (Presenter) Nothing to Disclose

LEARNING OBJECTIVES


RC711A Interpretive Pitfalls

RC711B Impact of Patient Preparation

RC711C Effective Reporting and Communication

RC711

Improving PET Interpretation (Interactive Session)

Thursday, Nov. 29 4:30PM - 6:00PM Room: S504CD

CT NM


LEARNING OBJECTIVES

1) Understand the patient preparation issues with performing PET/CT. 2) Review recommendations on patient preparation prior toperforming PET/CT. 3) Review the issues in performing PET/CT scans on diabetic patients and learn ways to optimize the glucoselevel. 4) With the aid of challenging case examples, this activity aims improve PET-CT interpretation through recognition of pitfallsand variants. In addition, it aims to review typical as well as unusual examples of commonly encountered oncologic diagnoses. 5)Learn how to discriminate malignancy from benign FDG-avid changes caused by surgery and procedures, radiation, andchemotherapy.

Sub-Events



[email protected]

LEARNING OBJECTIVES

1) Identify FDG-avid lesions caused by surgery, radiation, and chemotherapy that could be mistaken for malignancy. 2)Demonstrate how to integrate FDG PET and CT findings on FDG PET/CT to distinguish benign from malignancy causes of FDG-avidity.



[email protected]

LEARNING OBJECTIVES

1) Understand the patient preparation issues with performing PET/CT. 2) Review recommendations on patient preparation prior toperforming PET/CT. 3) Review the issues in performing PET/CT scans on diabetic patients and learn ways to optimize the glucoselevel.

ABSTRACT

F18-FDG PET/CT is a valuable tool for a variety of oncologic applications. The purpose of this educational activity is to discuss theimportance of appropriate patient preparation prior to performing oncologic F18-FDG PET/CT scans. The recommendations from theAmerican College of Radiology (ACR), the Society of Nuclear Medicine and Molecular Imaging (SNMMI), and the National CancerInstitute (NCI) for patient preparation will be discussed. Issues that will be discussed include fasting, limiting exercise, hydration,sedation, low carbohydrate meals, and diabetic patients. Patients are typically asked to fast for at least 4 hours before tracerinjection for oncologic PET/CT scans. The ACR and SNMMI both recommend checking glucose levels on all patients prior toadministration of F18-FDG. SNMMI guidelines recommend that patients with glucose of greater than 150-200 mg/dL should usuallybe rescheduled. Performing PET/CT scans in poorly controlled diabetic patients can result in a PET/CT scan with an alteredbiodistribution limiting interpretation of the study. In a poorly controlled diabetic patient with a glucose level of greater than 200mg/dl, the study should usually be rescheduled if it does not critically affect patient care. Hyperglycemia will dilute the FDG uptakeby tumors through competitive inhibition. Subcutaneous insulin should not be administered to a diabetic patient with high glucosewithin 4 hours of a PET/CT scan as insulin will stimulate FDG uptake by skeletal muscle resulting in an altered biodistribution whichcan severely limit interpretation.

ParticipantsEric M. Rohren, MD, PhD, Houston, TX (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

RC711D Challenging Case Examples

1) With the aid of challenging case examples, this activity aims improve PET-CT interpretation through recognition of pitfalls andvariants. 2) Aims to review typical as well as unusual examples of commonly encountered oncologic diagnoses.

ABSTRACT

Best practices in reporting are a critical aspect of a successful PET program. In this presentation, effective reporting methods willbe demonstrated, along with common pitfalls in reporting that should be avoided.Emphasis will be placed on the eight 'C's' ofeffective reporting: Correctness, Completeness, Consistency, Communication, Clarity, Confidence, Concision, and Consultation.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Eric M. Rohren, MD, PhD - 2015 Honored Educator



[email protected]

LEARNING OBJECTIVES

1) Review challenging case examples of FDG PET-CT of the abdomen and pelvis with particular emphasis on genitourinary andgynecologic imaging.

ABSTRACT

FDG PET-CT is an effective modality for staging, restaging and treatment follow up of various malignancies. In this session, areview of challenging cases will be presented. Variants and pitfalls that may impact interpretation will be discussed with specialemphasis on genitourinary and gynecologic imaging. Updtaes on imaging and interpretation parameters and guidlines will bereviewed.

RC717A Imaging Dementias 2018

RC717B Imaging Dementias: FDG and Amyloid PET/CT

RC717C Imaging Dementias: MRI

RC717D Imaging Dementias: Tau PET/CT

RC717

Emerging Technology: Imaging of Dementias

Thursday, Nov. 29 4:30PM - 6:00PM Room: S505AB

CT MR NR NM



ParticipantsRathan M. Subramaniam, MD,PhD, Dallas, TX (Moderator) Consultant, Blue Earth Diagnostics Ltd; Speaker, Blue Earth DiagnosticsLtd


[email protected]

LEARNING OBJECTIVES

1) To review the value of FDG and amyloid PET/CT in diganosis of dementia. 2) To review the value of MR imaging in diagnosis ofdementia. 3) To review the value of tau PET/CT in diagnosis of dementia.

ABSTRACT

This session will review the importance and value of FDG PET, Amyloid PET, MRI and Tau PET imaging in diagnosis of dementia.

Sub-Events



[email protected]

LEARNING OBJECTIVES

1) To discuss the value of FDG brain PET/CT in differentiating various dementias in cognitive decline. 2) To discuss the value ofAmyloid brain PET/CT in patients with cognitive decline and Alzheimer's disease.

ParticipantsWilliam A. Moore, MD, Dallas, TX (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Understand which FDA approved MR techniques are currently available for improving differential diagnosis in patients withdementia. 2) Improve basic knowlege of how MR results correspond to clinical dementia phenotypes. 3) Discuss recenttechnological advances including applications of dynamic susceptability contrast (DSC) MR, arterial spin labelling (ASL) and restingstate functional connectivity MRI (rs-fcMRI) in the setting of patients with dementia.

ParticipantsKejal Kantarci, MD, Rochester, MN (Presenter) Data Safety Monitoring Board, Takeda Pharmaceutical Company Limited; Data SafetyMonitoring Board, Pfizer Inc; Data Safety Monitoring Board, Johnson & Johnson; Research funded, Eli Lilly and Company


[email protected]

LEARNING OBJECTIVES

1) Describe the MRI techniques used in the imaging of cognitve impairment and dementia. 2) Understand the clinical utility. 3)Discuss the findings with pathologic confirmation.

ParticipantsVal J. Lowe, MD, Rochester, MN (Presenter) Research Grant, General Electric Company; Research Grant, Siemens AG; ResearchGrant, Eli Lilly and Company; Advisory Board, Merck & Co, Inc

LEARNING OBJECTIVES

1) Describe the basic science principles behind tau PET/CT imaging. 2) Understand the utility of tau PET/CT imaging inneurodegenerative disease. 3) Identify the findings of a positive tau PET/CT scan.

RC811A Pediatric Gastrointestinal

RC811B Pediatric Genitourinary

RC811C Pediatric Musculoskeletal

RC811D Case Presentation/Panel Discussion

RC811

Review of Pediatric Nuclear Medicine

Friday, Nov. 30 8:30AM - 10:00AM Room: E263

GI GU MK NM PD


LEARNING OBJECTIVES

1) Review of Pediatric Nuclear medicine, particularly for radiologists and nuclear medicine physicians who may not specialize inpediatric patients, and for resident and fellow trainees.

Sub-Events

ParticipantsHelen R. Nadel, MD, Palo Alto, CA (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Be able to list indications for GI scintigraphy in children. 2) Be able to describe scintigraphic patterns of disease on GIexaminations in children.

ParticipantsNeha S. Kwatra, MBBS, MD, Boston, MA (Presenter) Nothing to Disclose

LEARNING OBJECTIVES

1) Describe pediatric renal diseases highlighting the complementary role of scintigraphy and other imaging modalities. 2) Explainpediatric-specific imaging considerations. 3) Identify important normal variants/pitfalls in interpretation.

ParticipantsSusan E. Sharp, MD, Cincinnati, OH (Presenter) Nothing to Disclose


[email protected]

LEARNING OBJECTIVES

1) Be able to describe the utilization and performance of nuclear medicine imaging for musculoskeletal indications in pediatricpatients. 2) Be able to identify musculoskeletal findings on Tc-99m-MDP and F-18-FDG scans.

ParticipantsNeha S. Kwatra, MBBS, MD, Boston, MA (Presenter) Nothing to Disclose

SST02-01 Radiologist Variability in the Determination of the T-Size Descriptor Cutpoints in the Eighth Edition ofthe TNM Classification of Lung Cancer (TNM8)


SST02

Chest (Thoracic Malignancy)

Friday, Nov. 30 10:30AM - 12:00PM Room: E350

CH CT MR NM



ParticipantsGirish S. Shroff, MD, Houston, TX (Moderator) Nothing to DiscloseAndrew J. Plodkowski, MD, Brookside, NJ (Moderator) Nothing to Disclose

Sub-Events

ParticipantsJulia Hine, MBBS, London, United Kingdom (Presenter) Nothing to DiscloseAnouchka Goldman, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseJie-Ying Kowa, MBBS, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSarah L. Sheard, MBBS, FRCR, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseKonstantinos Stefanidis, MD, PhD, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseSisa Grubnic, MD, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseJoanna Moser, MBChB,FRCR, London, United Kingdom (Abstract Co-Author) Nothing to DiscloseIoannis Vlahos, MRCP,FRCR, London, United Kingdom (Abstract Co-Author) Research Consultant, Siemens AG Research Consultant,General Electric Company

PURPOSE

Based on pathological size TNM8 introduced additional T-descriptor size cutpoints at 1cm intervals impacting stage groups. Our aimwas to determine whether radiological staging by different radiologists consistently classifies lesion size within this more detailedstaging.


4 thoracic radiologists (4-17yr experience) staged 180 consecutive new lung cancers, recording multiple parameters blinded to thestudy aim. Readers were provided with axial 2.5mm, 1mm, coronal and sagittal 3mm images and asked to stage the primary as perclinical practice. Readers recorded the solid component for subsolid lesions. 2 observers covertly recorded the image series used forreview and measurement. Inter-rater consistency of primary lesion size and T-size determination was evaluated. The impact ofreader recorded lesion characteristics on consistency was assessed.

RESULTS

Readers recorded lesions as solid in 78-87% of cases, part-solid in 11-17% and pure ground glass in 1-2% with a moderate meaninter-rater kappa (0.71). 176 lesions were considered measurable by at least 3 readers (median 38mm, 7-113mm), 95% evaluatedby all 4 readers. Readers varied widely in measurement plane (2.5mm:20-90%, 1mm:2-54%, coronal:7-24%, sagittal:0-26%) andmean number of planes reviewed (1.1-3.0). For lesions the mean range of measurement about the consensus median size was 31%(3-175%). Increased reader range of measurement about the median size was associated with part solid (mean 43% v 29% solid,p<0.01 Mann Whitney U) and cavitary lesions (32 v 19%, p<0.05). Atelectasis and spiculation were not significant. Using mediansize to determine T-descriptors, only 42% of cases had 100% reader concordance (74% concordance for at least 67% of readers).Complete concordance was significantly lower for groups T1c-T3 (20-35%) and higher in the remaining groups (42-67%). Meaninter-rater T assignment kappa was 0.57 (moderate), but higher with weighted kappa (0.80, good).

CONCLUSION

There is considerable variation in tumor size determination by thoracic radiologists, influenced by lesion perceived morphology, andmeasurement choices that result in lower inter-reader concordance for the narrower range TNM8 T-size criteria.


Pathological size data informed increased numbers of cutpoints in TNM8 to better predict survival but increases radiological stageuncertainty and inter-reader variance in clinical practice.

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifyingeducational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-qualityeducational content in their field of study. Learn how you can become an honored educator by visiting the website at:https://www.rsna.org/Honored-Educator-Award/ Ioannis Vlahos, MRCP,FRCR - 2015 Honored Educator

SST02-02 A Novel Algorithm to Approach Multiple Lung Cancers with Multiple Pulmonary Sites of Involvement:Differentiation between Multiple Primary Lung Cancers and Intrapulmonary Metastasis


SST02-03 Risk of Occult Mediastinal Disease in Non-Small Cell Lung Cancer Patients with Radiographic N0Disease according to Tumor Location


ParticipantsYoung Joo Suh, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseHyun-Ju Lee, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJiseon Oh, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DisclosePamela K. Sung, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHeera Yoen, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSewoo Kim, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSeungchul Han, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSungeun Park, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJung Hee Hong, Daegu, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHeekyung Kim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJiyeon Lim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHyungjin Kim, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSoon Ho Yoon, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYoung Tae Kim, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYoon Kyung Jeon, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To develop an differentiation algorithm in patents with multiple lung cancers, using clinical and imaging variables.


We retrospectively included 112 lesions in 55 patients (57 pairs) with multiple lung cancers who received at least two separatesurgeries between January 2007 and December 2016. Each pair of multiple lung cancers was classified into two categories withhistopathologic findings as the standard reference: multiple primary lung cancer (MPLC) and intrapulmonary metastasis (IPM). Weestablished five serial questions for differentiation; 'Is either nodule pure ground-glass nodule on CT?' or 'Are both of the two lesionsground-glass dominant nodules?' (Step1), 'Does either nodule harbor air-bronchogram or irregular shape?' (Step2), 'Do both of thetwo nodules have the same or different grade of maximal standardized uptake values (SUVmax) on PET/CT?' (Step3), and 'Doeseither case harbor mediastinal LN or distant metastasis on preoperative work-up?' (Step4). The SUVmax values were classified intograde 1(<2.5), grade 2(2.5-5.0), and grade 3(>5.0). At each decision step, each pair was classified as MPLC or IPM. Thesensitivity, specificity, and accuracy of the differentiation algorithm were analyzed.

RESULTS

Among 57 pairs, 36 pairs (63.2%) were classified as MPLCs, and the other 21 pairs (26.8%) as IPMs of standard reference. Instep1, 14 pairs were classified as MPLC. In step2, 10 pairs with absence of air-bronchogram or irregular contour on both lesionswere classified as IPM. In step3, 8 pairs showing two grades of separate SUV were classified as MPLC. In step4, 3 pairs withmediastinal LN or distant organ metastasis were classified as IPMs and 22 pairs were considered MPLC. The sensitivity for MPLC(specificity for IPM), specificity for MPLC (sensitivity for IPM), and accuracy were 94.4%, 52.4%, and 78.9%, respectively.Accuracy for each step was 100% for step 1, 90% for step 2, 62.5% for step 3 and 68% for step 4, respectively.

CONCLUSION

Approach algorithm using comprehensive information of clinical and imaging variables can allow differentiation between MPLCs andIPMs in a substantial number of cases of multiple lung cancers with multiple pulmonary sites of involvement.


Our approach algorithm using clinical and imaging information can help differentiation between MPLCs and IPMs in multiple lungcancers.

ParticipantsDongyoung Jeong, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseSun Hye Shin, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYeonu Choi, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJiyeong Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Lung cancer guidelines recommend invasive mediastinal staging for patients with centrally located tumors without evidence of nodaldisease on imaging studies. However, there is no uniform definition of central tumor. This study aims to evaluate the risk of occultmediastinal disease in non-small cell lung cancer (NSCLC) patients with radiographic N0 disease according using several differentdefinitions for central tumor.


Of the patients who underwent curative-intent surgical resection or endobronchial ultrasound-guided transbronchial needle

SST02-04 Comparison of Computed Tomography and Clinical Findings Between Immune-Related Pneumonitisand Pneumonia by Pathogen in Patients Treated with Anti-Programmed Death-1 (PD-1)/ProgrammedDeath Ligand 1 (PD-L1) Therapy


SST02-06 Growth Rates of Thymic Epithelial Tumor and Thymic Cyst: Is Differentiation Feasible?


aspiration between January 2014 and December 2015, 1,337 consecutive patients with radiographic N0 disease were identified.Based on the most proximal part of the tumor in computed tomography (CT) image, tumors were categorized using five differentdefinitions; contact with hilar structure, located within inner one-third or two-thirds of hemithorax according to concentric orsagittal lines.

RESULTS

About 7% (93/1337) of patients had occult N2 disease and they had significantly larger tumor size and more solid tumors in CTimage. All but inner two-thirds of hemithorax by sagittal line were associated with N2 disease. However, only inner one-third ofhemithorax by concentric line remained significant after adjustment for tumor size and density in CT (adjusted odds ratio [95%confidence interval], 2.29 [1.28-4.11]).

CONCLUSION

We suggest using inner one-third of hemithorax by concentric line as indication of EBUS-TBNA in NSCLC with radiographic N0disease.


Using inner one-third of hemithorax by concentric line as indication of EBUS-TBNA in NSCLC with radiographic N0 disease.

ParticipantsCherry Kim, MD, Ansan, Korea, Republic Of (Presenter) Nothing to DiscloseMi Young Kim, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseChang-Min Choi, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYeon Joo Kim, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose

PURPOSE

Immune-related pneumonitis (IRP) is an uncommon but potentially fatal toxicity of anti-programmed death-1 (PD-1)/programmeddeath ligand 1 (PD-L1) therapy for intrathoracic malignancy including non-small cell lung cancer. The purpose of study was tocompare CT and clinical findings between IRP and pneumonia by pathogen.


A total of 154 patients who received anti-PD-1/PD-L1 therapy were identified from 2014 to 2017. Among these patients, IRPdeveloped in 9 (5.8%) and pneumonia in 30 (19.5%), which were confirmed through multidisciplinary approach. CT findings(reticulation, consolidation, ground glass opacity [GGO], interlobular septal thickening, micro- [<10mm] and macro-nodules[>=10mm], bronchial wall thickening, bronchiectasis, pleural effusion, and lesion distribution/bilaterality) and clinical features(symptom, smoking history, cancer staging, laboratory findings, underlying disease, prior radiotherapy history) were comparedbetween IRP and pneumonia. Grade and outcome of IRP were also investigated.

RESULTS

In chest CT, diffuse reticulation (44.4% vs.0%, P=0.02), patchy/diffuse GGO (100% vs. 50%, P=0.01), and interlobular septalthickening (66.7% vs. 10%, P=0.002) were significantly more frequent in IRP than in pneumonia, whereas macronodule (0 vs.36.7%, P=0.033) was significantly more common in pneumonia than IRP. IRP significantly showed peripheral location (77.8% vs.16.7%, P=0.001) and bilateral distribution (44.4% vs. 3.3%, P=0.007). However, there were no significant differences in clinicalfindings between IRP and pneumonia. Among the IRP patients, 66.7% (6 of 9) of cases were grade 3, and 66.7% improved withdrug holding/steroid therapy. The median onset duration of IRP from the first prescription was 126 days (range, 40-669), themedian time for improvement was 43 days (range, 21-45), and the median time to death due to IRP was 18 days (range, 11-55).

CONCLUSION

Several CT findings including diffuse reticulation, patchy/diffuse GGO, and interlobular septal thickening with bilateral and peripheraldistribution were more frequent in IRP than pneumonia by pathogen. Clinical findings were overlapped.


It is crucial to suspect IRP as opposed to pneumonia in routine practice. Radiologists should be familiar with those findings of IRP toavoid delayed diagnosis and serious drug related complication.

ParticipantsHyungjin Kim, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseSoon Ho Yoon, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJihang Kim, MD, Seongnam, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseKyung Won Lee, MD, PhD, Seongnam, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYe Ra Choi, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHyoun Cho, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJin Mo Goo, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Research Grant, Samsung Electronics Co, Ltd; ResearchGrant, Lunit Inc


[email protected]

PURPOSE

SST02-07 Cut-Off Value of MR Enhancement for Differentiating Benign Cysts from Solid Anterior MediastinalLesion: A Preliminary Observation


SST02-08 Primary Tumor Standardized Uptake Value (SUVmax) as Powerful Prognostic Factor for Early

To investigate the growth rate of thymic epithelial tumors (TETs) and thymic cysts to determine whether they can bedifferentiated, and to identify clinico-radiological predictors of interval growth and their differential implications.


This retrospective study included 122 patients (male:female=64:58; mean age, 57.2 years) with pathologically proven thymic cysts(n=56) or TETs (n=66) who underwent 2 serial chest CT scans at least 8 weeks apart. Average diameters were measured, andvolume-doubling times (VDTs) were calculated. Attenuation was also measured and clinical characteristics were recorded. VDTswere compared between the thymic cysts and TETs using the log-rank test. Predictors of growth were analyzed using the log-ranktest and Cox regression analysis.

RESULTS

The frequency of growth did not significantly differ between TETs and thymic cysts (P=0.279). The VDT of the thymic cysts(median, 324 days) was not significantly different from that of the TETs (median, 475 days; P=0.808). Water attenuation (<=20Hounsfield Unit) predicted growth in thymic cysts (P=0.016; HR, 13.2 [95% CI, 1.6-107.3]) and lesion size (>17.2 mm) predictedgrowth in TETs (P=0.008 for size and P=0.029 for size*time; HR=e^(-0.001×time+1.654)). Among the growing lesions, positive andnegative predictive values of water attenuation for the thymic cysts was 93% and 80%, respectively.

CONCLUSION

The frequencies of interval growth and VDTs were indistinguishable between TETs and thymic cysts. Water attenuation and lesionsize predicted growth in thymic cysts and TETs, respectively. Among the growing lesions, the water attenuation was a differentialfeature of thymic cysts.


Water attenuation (<=20 HU) indicates thymic cysts for the growing thymic lesions. Thus, CT follow-up, instead of surgicalresection, can be recommended for the obvious cysts even if they show interval growth.

ParticipantsEui Jin Hwang, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseSoon Ho Yoon, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJihang Kim, MD, Seongnam, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseHo Yun Lee, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJin Mo Goo, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Research Grant, Samsung Electronics Co, Ltd; ResearchGrant, Lunit IncMunyoung Paek, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseJeanne B. Ackman, MD, Boston, MA (Abstract Co-Author) Spouse, Stockholder, Everest Digital Medicine; Spouse, Consultant,Everest Digital Medicine; Spouse, Stockholder, Cynvenio Biosystems, Inc; Spouse, Scientific Advisory Board, Cynvenio Biosystems,Inc; Spouse, Consultant, PAREXEL International Corporation


[email protected]

PURPOSE

To determine the optimal cut-off value of MR enhancement for the differentiation of benign cysts from solid lesions in the anteriormediastinum.


The derivation dataset consisted of 19 consecutive patients with pathologically proven benign cysts (n=7) and solid lesions (n=12)in the anterior mediastinum who underwent a diagnostic contrast-enhanced MR from two institutions. We measured maximumdiameters, T1 and T2 signal intensities (SI), apparent diffusion coefficients (ADCs) from diffusion-weighted images, and relativeenhancement ratios (RERs). T1 and T2 SIs were normalized by SI of cerebrospinal fluid. RERs were obtained from the subtraction ofaxial pre- and post-contrast T1-weighted fat-suppressed images after the precise non-rigid image registration of the two imagesusing a dedicated WIP software (MRI Arithmetics, Siemens Healthcare, Erlangen, Germany). After comparison of image variablesbetween cysts and solid masses, the cutoff value of the most differential MR variable was determined based on a receiver operatingcharacteristic curve. For validation, two separate datasets were utilized: 1) 15 patients with 8 cysts and 7 solid lesions fromanother institution (validation dataset 1); 2) 11 patients with MR-proven stable benign cysts more than 2 years (validation dataset2). Diagnostic accuracies were calculated from validation datasets.

RESULTS

Normalized T2 SI (0.21-0.92 vs. 0.12-0.58; P=.013), ADC (1.76-4.09 vs. 0.66-2.93 10-3 mm2/s; P=.013), and RER (0.41-24.1% vs.28.1-771.7%; P<.001) significantly differed between cysts and solid masses. RER of 26% or less was determined as the cutoff valuefor differentiation of cysts from solid masses. In validation dataset 1, the cutoff value showed sensitivity of 87.5% and specificityof 100%, the sensitivity of 90.9% was observed in validation dataset 2.

CONCLUSION

The assessment of RER with the cutoff value of 26% can appropriately differentiate benign cysts from solid anterior mediastinalmasses.


The differentiation of benign cysts from solid anterior mediastinal masses can be supported by quantitative measurement of RER,potentially reducing a futile thymectomy.

Esophageal Squamous Cell Carcinoma


SST02-09 Surgically Resected T1- and T2-Stage Esophageal Squamous Cell Carcinoma: T and N StagingPerformance of EUS- and PET/CT


ParticipantsDongyoung Jeong, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseKyung Soo Lee, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseYeonu Choi, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseSunggoo Park, MD,DVM, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

We have previously shown that initial PET-SUVmax(Standardized uptake value) of early esophageal cancer helps both discriminatingT1a and T1b stage esophageal squamous cell carcinoma(eSCC) from other eSCCs. In this study, we analyze the impact of PET-SUVmax for patient's survival.


This retrospective study was based on 435 patients with a surgically proven early T- (Tis or T1a [< T1a], T1b and T2) stage eSCC.We performed survival analysis by the Kaplan-Meier method and comparisons of survival using log-rank test.

RESULTS

131 < T1a, 234 T1b, and 70 T2 eSCCs were enrolled. Mean SUVmax value were 2.53 for < T1a eSCCs, 4.02 for T1b eSCCs and 9.69for T2 eSCCs. With ROC curve analysis, cut off value of SUVmax 3.05 (AUC: 0.757; 95% CI, 0.710-0.803; P < .001) at PETprovided sensitivity 74.8% (98/131), specificity 70.1% (213/304), respectively, for differentiating < T1a eSCCs from other cancers.Cut off value of SUVmax 5.65 (AUC: 0.897; 95% CI, 0.857-0.937; P < .001) provided sensitivity 77.1% (54/70), specificity 87.7%(320/365), respectively, for differentiating T1 (< T1b) eSCCs from T2 eSCCs. In multivariate analysis, both SUVmax and pathologicstaging including tumor size and node involvement were significant predictors of survival (p < 0.01). Survival analysis and log-ranktest showed significant difference for overall survival among groups based on proposed cut-off SUVmax values (p =0.008 for cut offvalue 3.05, p <0.001 for cut off value 5.65)

CONCLUSION

In early esophageal squamous cell carcinomas, SUVmax gives us powerful predictor of overall survival after resection.


Pretreatment SUVmax of primary esophageal cancer shows powerful predictive values which can be comparable to pathologic Tstage.

ParticipantsDongyoung Jeong, MD, Seoul, Korea, Republic Of (Presenter) Nothing to DiscloseKyung S. Lee, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseMinYeong Kim, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to DiscloseMyung Jin Chung, MD, Seoul, Korea, Republic Of (Abstract Co-Author) Research Grant, General Electronic Company; ResearchGrant, Samsung Electronics Co, Ltd; Research Grant, Lunit IncJoon Young Choi, MD, PhD, Seoul, Korea, Republic Of (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To demonstrate the frequency of nodal metastases and to disclose the diagnostic performance of endoscopic ultrasonography(EUS) and PET/CT in T and N staging in surgically resected early-stage esophageal squamous cell carcinomas (eSCCs).


IRB approved this retrospective study with waiver of informed consent for reviewing medical record. We included 435 patients withan early T-stage (Tis or T1a [< T1a], T1b and T2) eSCC. The rates of metastatic lymphadenopathy were calculated. Then, theperformance of EUS and PET/CT in subdividing T and N stages was assessed.

RESULTS

131 < T1a, 234 T1b, and 70 T2 eSCCs were identified. In discriminating < T1a from other cancers, the sensitivity, specificity andaccuracy of EUS were 60.3% (79/131), 80.3% (244/304), and 74.3% (323/435), respectively. With ROC curve analysis, cutoffvalue of SUVmax 3.05 at PET provided sensitivity 73.3% (96/131), specificity 70.4% (214/304), and accuracy 71.3% (310/435) fordifferentiating < T1a eSCCs from others. Ten (7.6%) of 131 < T1a cancers had nodal metastasis. In discriminating N0 from node-positive disease, sensitivity, specificity and accuracy of EUS were 89.6% (267/298), 41.6% (57/137) and 74.5% (324/435),respectively, whereas those of PET/CT were 88.9% (265/298), 38.7% (53/137), and 73.1% (318/435), respectively.

CONCLUSION

In > 70% of patients with < T1a eSCCs, the tumor stage can be discriminated from higher stage cancers by using EUS or PET/CT,and substantial percentage (7.6%) of < T1a eSCC patients have nodal metastases, but the nodes are missed in more than half ofthe patients in clinical staging.


Substantial percentage (7.6%) of < T1a eSCC patients have nodal metastases, and nodal metastasis rates increase as T stageincreases (T1b [37.6%] and T2 [55.7%]). Moreover, more than half of nodal metastases were missed on PET/CT or EUS. Thus,after endoscopic surgery or even after curative surgical resection of < T1a eSCCs, adjuvant therapy is needed for those havingnodal metastasis.

SST06-01 Liver/Spleen Scintigraphy: An Old Technique with a current Application


SST06-02 The Diagnostic Accuracy of Brain-Lung Uptake and Whole-Body Uptake Derived from Technetium-99m-Labeled Macroaggregated Albumin (MAA) Lung Perfusion Scan for the Diagnosis ofHepatopulmonary Syndrome


SST06

Nuclear Medicine (Abdomen and Pelvis Nuclear Imaging)


CT GI MR NM


ParticipantsMichael V. Knopp, MD, PhD, Columbus, OH (Moderator) Nothing to DiscloseEphraim E. Parent, MD,PhD, Ponta Vedra Beach, FL (Moderator) Research support, Blue Earth Diagnostics Ltd; Research support,Advanced Accelerator Applications SAAndrew C. Homb, MD, Rochester, MN (Moderator) Nothing to Disclose

Sub-Events

ParticipantsMaxwell P. Cocco, Ann Arbor, MI (Presenter) Nothing to DiscloseDaniel J. Wale, DO, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseTimothy Frankel, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseRichard K. Brown, MD, Ann Arbor, MI (Abstract Co-Author) Nothing to DiscloseJames H. Ellis, MD, Ann Arbor, MI (Abstract Co-Author) Research Consultant, General Electric Company ; Medical Advisory Board,Nuance Communications, IncWilliam J. Weadock, MD, Ann Arbor, MI (Abstract Co-Author) Owner, Weadock Software, LLCBenjamin L. Viglianti, MD,PhD, Ann Arbor, MI (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Cross sectional imaging of the abdomen on occasion can demonstrate incidental pancreatic/peri-pancreatic lesions of unknownetiology. Although these lesions sometimes have the appearance of splenic tissue, situation arise were metastasis and/or primarypancreatic etiology are also on the differential. Scintigraphy with Tc99m Sulfur colloid or heat damaged RBC can be a usefuldiagnostic tool in identifying the presence of splenic tissue, suggesting a benign etiology for these lesions although the accuracy isuncertain.


Retrospective review of all non-PET nuclear medicine studies in which a lesion in the upper abdomen were the cause of imagingfrom 1/2000 to 7/2017. Studies performed for hepatic artery perfusion or liver parenchyma lesion were excluded. Patients chartswere reviewed, the date/results of the index study, subsequent imaging, clinical management, along with the last recordedencounter in our electronic medical record to establish benignity in the absence of pathology.

RESULTS

Initial review obtained 623 studies performed, a majority (74%) for hepatic artery perfusion. Liver lesion evaluation was performedin 7% of the cases (hemangioma or FNH). The remainder of the cases were done for evaluating splenic tissue in ITP (~5% ofcases), ectopic splenic tissue (~9% after trauma or splenectomy) or to evaluate incidental pancreatic/peri-pancreatic lesions withpotential neoplastic etiology (~5% cases, 34 cases total). Of these 34 cases, pathology was obtained in 12/34 patients. Imagingwas correct in identifying nonsplenic tissue in 11/12 cases (92%). One case was splenic tissue on path, but was not identified onimaging. For the patients that had imaging indicating splenic tissue without subsequent pathology (12/22), follow up of the patientoccurred > 3-years.

CONCLUSION

Pancreatic/peri-pancreatic lesions of unknown etiology can present a diagnostic challenge with causes ranging from benign splenictissue to a neoplastic process. Scintigraphy offers a unique ability to identify ectopic splenic tissue with a high degree of diagnosticaccuracy, >90%, yielding a benign diagnosis and limiting the need of further workup.


Liver/Spleen imaging with Tc99m Sulfur colloid or damaged RBC can accurately identify incidental pancreatic/peri-pancreatic lesionsas ectopic splenic tissue not requiring further workup.

SST06-03 Diagnosing Hepatobiliary Disease during Myocardial Perfusion Imaging Using Tc99m MethoxyIsobutyl Isonitrile


ParticipantsHe Zhao, Beijing, China (Abstract Co-Author) Nothing to DiscloseJiaywei Tsauo, Beijing, China (Presenter) Nothing to DiscloseXiaowu Zhang, Beijing, China (Abstract Co-Author) Nothing to DiscloseTao Gong, Beijing, China (Abstract Co-Author) Nothing to DiscloseJingui Li, Beijing, China (Abstract Co-Author) Nothing to DiscloseXiao Li, PhD, Chengdu, China (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Technetium-99m-labeled macroaggregated albumin (MAA) lung perfusion scan is considered as a complementary tool for detectingintrapulmonary vascular dilations (IPVD), which is an essential criterion for diagnosing hepatopulmonary syndrome (HPS). Thepurpose of this study was to compare the diagnostic accuracy of brain-lung uptake and whole-body uptake for detecting IPVD.


From December 2014 to October 2015, all patients with chronic liver disease and/or portal hypertension, undergoing interventionalradiological procedures at our institution were eligible for inclusion in this prospective study. The brain-lung uptake was calculatedusing the geometric mean (GMT) of technetium counts around the brain and lung in the following formula: (GMTbrain / 0.13) /(GMTbrain / 0.13 + GMTlung). The brain-lung uptake was regarded as positive when the MAA shunt fraction was > 6%. The whole-body uptake was calculated using the GMT of technetium counts around the lung and whole body in the following formula: (1 -GMTlung / GMTwhole-body)

RESULTS

A total of 69 patients were included, IPVD was detected in 32 (46%) patients by contrast-enhanced echocardiography. Of thesepatients, 26 (38%) patients with elevated AaO2 were diagnosed as HPS. The brain-lung uptake was similar between those with orwithout IPVD [median, 3.5 (interquartile range (IQR), 2.6-5.8) % vs. 3.1 (IQR, 2.5-4.9) %; P = 0.245]. However, the whole-bodyuptake was significantly higher in the patients with IPVD than those without IPVD (48.0 ± 6.1 % vs. 40.1 ± 8.1 %; P = 0.001).Multivariable logistic regression showed that whole-body uptake was the only independent predictor that associated with thepresence of IPVD [odds ratio (OR), 1.29; 95% CI, 1.07-1.55; P = 0.008]. The AUC values of the whole-body uptake for detectingIPVD were 0.75 (95% CI, 0.60-0.86). The optimal cut-off values of whole-body uptake for detecting IPVD was 42.5%. Thesensitivity, specificity, and accuracy for detecting IPVD were 100%, 52%, and 74%, respectively.

CONCLUSION

Whole-body uptake could be a useful alternative to CEE and brain-lung uptake for detecting IPVD, especially in patients with mild ormoderate HPS.


Whole-body uptake derived from MAA lung perfusion scan could be a useful alternative to contrast-enhanced echocardiography andbrain-lung uptake for detecting intrapulmonary vascular dilations.

ParticipantsHaitham M. Elsamaloty, MD, Holland, OH (Abstract Co-Author) Nothing to DiscloseMohamad F. Bazerbashi, MD, Ann Arbor, MI (Presenter) Nothing to DiscloseEslam Y. Youssef, MD, Toledo, OH (Abstract Co-Author) Nothing to DiscloseLina Elsamaloty, Toledo, OH (Abstract Co-Author) Nothing to DiscloseHassan B. Semaan, MD, Toledo, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Tc99m Methoxy Isobutyl Isonitrile (MIBI) has been used for myocardial perfusion imaging (MPI) for the detection of ischemia. Thisstudy aimed to investigate the feasibility of effectively evaluating cystic duct patency, during routine visual analysis of the raw MPIand/or with the 3-D reconstructed data.


A retrospective study of 91 patients undergoing cardiac Sestamibi scan for acute chest pain, and HIDA scan (within no more than 3months) for suspected gallbladder obstructive disease (GBD). Gallbladder visualization during either the rest or stress portion of theMIBI study was indicative of cystic duct patency. These results were compared to those by the HIDA studies.

RESULTS

Ten patients had the MIBI and HIDA studies 4 days apart, both studies agreed 100% with the diagnosis of cystic duct patency.Sixteen patients had both studies between 4 days and 3 weeks and had an agreement of 87.5% with cystic duct patency. Sixty-one patients had both studies 3 weeks to 3 months apart and had an agreement of 80% with cystic duct patency.

CONCLUSION

The initial results of this study indicate that MPI with Tc99m MIBI is useful in detecting a patent cystic duct and should help ineliminating unnecessary additional Gallbladder testing.

SST06-04 Qualitative and Quantitative Analysis of 68ga-DOTA-Peptide Uptake for Identifying NeuroendocrineTumor in Uncinated Process of Pancreas


SST06-05 Hepatobiliary Scintigraphy versus Ultrasound in the Evaluation of Acute Cholecystitis: An InstitutionalReview



If the interpreting physician was made aware of the benefits of being able to diagnose cystic duct obstruction and gallbladderdisease when using the Tc-99m cardiac sestamibi to evaluate myocardial perfusion, it would lead to earlier diagnosis and moreefficient patient care; thus, decreasing the amount of imaging that patients need to go through to reach a diagnosis of gallbladderobstruction, which will lead to decreased cost.

ParticipantsAmrita Kasat, MBBS,MD, Mumbai, India (Presenter) Nothing to DiscloseSachin K. Gawde, MD, Mumbai, India (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

68Ga-DOTA-peptide is a somatostatin analogue used for imaging neuroendocrine tumors (NET). Various organs demonstratephysiological distribution of which uncinate process of pancreas is a particular concern because uncinate is a common site for NETand great variability of tracer distribution at this location is making the interpretation difficult. Thus aim is to characterize 68Ga-DOTA-peptide distribution in uncinate process and useful hints for distinguishing pathology from physiological distribution.


83 68Ga-DOTApeptide PET CT scans of 25 patients done between May 2009 and Oct 2014 were reviewed retrospectively. 66 scansfrom 20 subjects tumor involvement of uncinate process was excluded based on pathological,clinical,radiological evaluation andatleast 1 year follow-up. 17 scans from 5 subjects neuroendocrine tumor involvement of the uncinate process was confirmed byhistology and/or multimodality imaging. Statistical analyses univariable Generalized Estimating Equations was carried out for normaluncinate uptake features. Comparison of normal vs tumor uptake was carried out using Mann Whitley test.

RESULTS

There are 3 types of normal distribution in uncinate process diffuse,focal and multifocal. Average SUVmax for normal uncinateprocess is 5.88 +/-3.34 with highest to be 21.07. The average SUVmax for uncinate neuroendocrine tumor is 76.28 +/- 44.72 withlowest to be 27. Tumor/spleen ratio is significantly higher than uncinate/spleen ratio (8.98 +/- 3.83 with lowest 3.67 vs 0.36 +/-0.41 with highest 1.44). Strong positive correlation between uptake in normal uncinate process to that in pituitary and spleen (bothP<0.0001) and the uptake is negatively affected by dose of peptide (p=0.0002).

CONCLUSION

Distribution pattern and uptake intensity in uncinate process vary greatly between patients and between scans. Pituitary andspleen uptake serve as references in judging the nature of uptake in uncinate. Low grade NET in uncinate process demonstratesignificantly higher than normal uptake and greater than normal uncinate/spleen ratio.


SUVmax and uncinate/spleen ratio is useful for differentiating normal versus tumor. SUVmax of 25 and uncinate/spleen ratio of 1.5are recommended reasonable cutoff values for this purpose. But tumor involvement in uncinate process should be made by not onlyby presentation on PET scan but correlating with other imaging findings and/or biopsy result.

ParticipantsTimur Kotlyar, MD, New Hyde Park, NY (Presenter) Nothing to DiscloseChristopher Kyriakakos, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseVincent Lau, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseMaria B. Tomas, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseChristopher J. Palestro, MD, New Hyde Park, NY (Abstract Co-Author) Nothing to DiscloseMeera Raghavan, MD, Tampa, FL (Abstract Co-Author) Nothing to Disclose

PURPOSE

Acute cholecystitis is a common entity. Hepatobiliary radionuclide scintigraphy (HIDA) is considered the gold standard for diagnosingcholecystitis, with sensitivity up to 97%. Ultrasound (US) is the preferred modality in the initial evaluation of acute cholecystitis,due to its availability, low cost, and short examination time, despite being less sensitive than HIDA. At our institution, US is theinitial imaging study of choice, and a HIDA is subsequently obtained for equivocal studies. We aim to evaluate the concordance ofUS and HIDA, and to identify clinical and/or laboratory parameters which may correlate with equivocal US in order to better guidepatient management.


Following institutional IRB approval, MONTAGE was used to search our reporting database between October 2016 and October2017. HIDA results were categorized as "positive" or "negative". US results were categorized as "negative", "positive", or "equivocal"for the assessment of acute cholecystitis. Clinical and laboratory data were also collected.

RESULTS

A total of 307 patients underwent both US and HIDA (n=307) with 43% (n=132) with an equivocal US. 35% (n=107) of thesepatients had discordant US and HIDA. Of the cases with an equivocal US, 43% (n=57) underwent cholecystectomy. At pathology,

SST06-06 Identification and Characterization of Myocardial Metastases in Neuroendocrine Tumor Patients Using68Ga-DOTATATE PET-CT


SST06-07 Evaluation of [18F]-FDG PET/MR Enterography in the Assessment of Ileocolonic Inflammation inCrohn's Disease - Which Surrogate Marker is Better? MaRIA, Clermont, PET, or PET-MR index?


51% (n=29) had acute cholecystitis with a positive HIDA and 4% (n=2) had acute cholecystitis with a negative HIDA. 14% (n=8)had chronic cholecystitis with a positive HIDA, and 32% (n=18) had chronic cholecystitis with a HIDA negative for acutecholecystitis. In patients with equivocal US, HIDA had a sensitivity of 93.6% and specificity of 69.2%.

CONCLUSION

Over one-third discordance between US and HIDA can have significant clinical implications. Given the high sensitivity of HIDA inpatients with equivocal US, initial evaluation with HIDA may be more appropriate in patients in whom US is likely to be equivocal,possibly leading to decreased time to surgery and length of stay.


Our results may elucidate factors influencing HIDA/US concordance, and whether US or HIDA is a more appropriate initial test.Imaging utilization may influence length of stay and time to surgery.

ParticipantsWolfgang G. Kunz, MD, Munich, Germany (Presenter) Grant, Medtronic plcRalf S. Eschbach, Munich, Germany (Abstract Co-Author) Nothing to DiscloseRobert Stahl, MD, Munich, Germany (Abstract Co-Author) Nothing to DisclosePhilipp M. Kazmierczak, MD, Munich, Germany (Abstract Co-Author) Nothing to DisclosePeter Bartenstein, Munich, Germany (Abstract Co-Author) Nothing to DiscloseAxel Rominger, Munich, Germany (Abstract Co-Author) Nothing to DiscloseChristoph Auernhammer, MD, PhD, Munich, Germany (Abstract Co-Author) Research Grant, Novartis AG Speaker, Novartis AGResearch Grant, Ipsen SA Speaker, Ipsen SA Advisory Board, Novartis AGChristine Spitzweg, Munich, Germany (Abstract Co-Author) Advisory Board, Swedish Orphan Biovitrum AB; Advisory Board, BayerAG; Advisory Board, Eisai Co, Ltd; Advisory Board, AstraZeneca PLC; Speaker, Swedish Orphan Biovitrum AB; Speaker, Bayer AG;Speaker, Eisai Co, Ltd; Speaker, AstraZeneca PLCJens Ricke, MD,PhD, Munich, Germany (Abstract Co-Author) Research Grant, Sirtex Medical Ltd Research Grant, Bayer AGClemens C. Cyran, MD, Munich, Germany (Abstract Co-Author) Research Grant, iThera Medical GmbH Speakers Bureau, Siemens AG


[email protected]

PURPOSE

Focal 68Ga-DOTATATE PET lesions within the myocardium of neuroendocrine tumor (NET) patients are observed in clinical practice.We determined the frequency and characteristics of lesions that are consistent with cardiac metastasis and assessed the lesiondetection rate of conventional imaging.


629 patients who underwent 68Ga-DOTATATE PET-CT at a supraregional comprehensive cancer center on NET were included froma consecutive registry. Inclusion criteria were: (1) focal 68Ga-DOTATATE tracer uptake within the myocardium in more than twosequential PET exams, and (2) contrast-enhanced CT. To determine the diagnostic accuracy of conventional CT imaging, a case-control cohort with a ratio of 1:3 was used. PET and CT were independently analyzed by two blinded readers. Cohen's k wasassessed for interreader agreement. Descriptive statistics were applied for frequencies and characteristics and group comparisonswere analyzed using the Fisher's exact test.

RESULTS

The prevalence of myocardial metastases related to the registry was 2.4% with 15/629 NET patients fulfilling the inclusion criteriaand a total of 21 focal myocardial 68Ga-DOTATATE tracer uptakes detected. Myocardial lesions were most frequently located in theleft ventricle (43%) and the septum (43%). No patient demonstrated a pericardial effusion. Patients with myocardial metastases didnot differ in demographics, tumor grading, disease stage or circulating tumor markers compared to the overall registry (all p>0.05).The patient characteristics are shown in Table 1. Higher Ki67-Indices were observed (p=0.049) for patients with myocardialmetastases. Interreader agreement for PET assessment was excellent (Cohen's =1.0). CT reading showed a sensitivity of 19%(95% confidence interval: 6%-43%) at a specificity of 100% (95% confidence interval: 90%-100%). A patient example with a CT-detected cardiac metastasis is provided in Figure 1.

CONCLUSION

68Ga-DOTATATE PET enables detection of myocardial metastatic lesions in NET patients. In contrast, standard morphologic CTimaging provides very limited sensitivity.


68Ga-DOTATATE PET imaging provides added diagnostic value in the initial staging of NET patients with cardiac metastasis and mayprovide further guidance during patient follow-up.

ParticipantsLale Umutlu, MD, Essen, Germany (Abstract Co-Author) Consultant, Bayer AGBenedikt M. Schaarschmidt, MD, Essen, Germany (Presenter) Stockholder, Bayer AG; Stockholder, General Electric Company;Stockholder, Siemens AG; Stockholder, Teva Pharmaceutical Industries LtdJost Langhorst, Essen, Germany (Abstract Co-Author) Nothing to DiscloseJohannes Grueneisen, Essen, Germany (Abstract Co-Author) Nothing to Disclose

SST06-08 Dynamic Study by PET/CT: Phantom Study and Clinical Trial on Sequential 26 Cases of MalignantLesions of Uterus


Verena Ruhlmann, Essen, Germany (Abstract Co-Author) Nothing to DiscloseKarsten J. Beiderwellen, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseJulian Kirchner, Dusseldorf, Germany (Abstract Co-Author) Nothing to DiscloseAydin Demircioglu, Essen, Germany (Abstract Co-Author) Nothing to DiscloseFelix Nensa, MD, Essen, Germany (Abstract Co-Author) Nothing to DiscloseKen Herrmann, Essen, Germany (Abstract Co-Author) Co-founder, SurgicEye GmbH; Stockholder, SurgicEye GmbH; Consultant, SofieBiosciences; Consultant, Ipsen SA; Consultant, Siemens AG; Research Grant, Advanced Accelerator Applications SA; ResearchGrant, Ipsen SAYan Li, Essen, Germany (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

To define a PET-MR index as hybrid surrogate marker and evaluate the diagnostic performance of validated MR indices, PET andPET-MR index in detecting ileocolonic inflammation in Crohn's disease (CD) with an integrated whole-body PET/MR .


53 CD patients with recurrent symptoms underwent ileocolonoscopy with biopsy (reference standard) and PET/MR enterography. 7ileocolonic segments were divided. The endoscopic activity of inflammation was determined by Simplified Endoscopic Activity Scorefor CD (SES-CD) and categorized in absent, mild to moderate and severe. Receiver Operating Characteristic (ROC) curves wereperformed and tested against each other with DeLong' test. Correlations between surrogate markers and SES-CD were tested withSpearmann's rank correlation test.

RESULTS

303 ileocolonic segments were analyzed. A simplified PET-MR index was defined as 0.87*wall thickness + 1.97*edema +0.83*ulceration + 0.55*SUVmax ratio + 1.14. In detecting active disease (defined as SES-CD >= 2), MaRIA, Clermont score andPET-MR Index as multiparametric indicies performed significantly better than monoparametric SUVmax ratio (areas under ROC:0.916, 0.914, 0.924 and 0.857, p < 0.05). In predicting severe inflammation with ulcerations, among all the surrogate makers (areasunder ROC fo MaRIA, Clermont, PET-MR index and PET: 0.962, 0.970, 0.971 and 0.935) only a slightly significant difference could beobserved between MaRIA and Clermont score (p = 0.02) in their operating characteristics. All surrogate markers correlatedmoderately with SES-CD both on segmental basis and gobal level (0.4 < ρ < 0.7, all p < 0.001).

CONCLUSION

As hybrid surrogate marker comprised of MR parameters and PET component, PET-MR index yielded a significantly increaseddiagnostic performance compared to PET alone. However, neither in predicting active disease nor in detecting severe ulcerativeinflammation, PET-MR index could outperform MaRIA or Clermont significantly. Nevertheless, PET-MR index showed the bestoperating characteristics among all the surrogate markers.


PET/MR enterography provides valuable surrogate markers for assessment of active disease in Crohn´s disease.

ParticipantsEri O'Uchi, MD, Kamogawa City, Japan (Presenter) Nothing to DiscloseToshihiro O'uchi, MD,PhD, Kamogawa, Japan (Abstract Co-Author) Nothing to Disclose

PURPOSE

Dynamic study of PET/CT with 18F-FDG has not been reported up to now. The purpose of this study was to check the reliability ofSUVmax on GE IQ PET/CT by phantom study, and to compare tumor 18F-FDG uptake between two phases of 30-second acquisitionof dynamic deep-inspiration breath-hold PET/CT (BHPC) before and after 5 to 7 steps of 150-sec free-breathingPET/CT(FBPC).These sequence was named 'Dynamic study by PET/CT'.


The PET/CT scanner was GE IQ with 26cm BGO crystal and PET images were reconstructed by patented new method of successiveapproximation (25-times). Before the clinical study, a phantom study was performed using an International ElectrotechnicalCommission body phantom set corresponding to the NU 2-2001 standard. The phantom set was consisted of a torso cavity and twospheres (inner diameters: 10, 13, 17, 22, 28, and 37 mm). The torso cavity was filled with water, and the 6 spheres were filled with18F-FDG solutions of the same radioactivity concentration (25 kBq/mL). We studied sequential 26 patients, from 40 to 84 years old,with uterine malignant tumor including 11 cases of corpus carcinomas, 13 cases of cervical carcinomas, and 2 cases of uterus originmalignant lymphomas. On the basis of the phantom study, patients with uterine malignant tumors smaller than 13mm wereexcluded. Maximum tumor 18F-FDG SUV (SUVmax) was measured in FBPC and the two phases of BHPC.

RESULTS

Our phantom study revealed that BHPC was also reliable when the size of lesion was bigger than 13mm with the accuracy of one SDwhich was smaller than 2% of SUVmax. This reliability was markedly improved by one tenth in comparison with that of 10 years ago.On clinical study, the mean SUVmax was 18.56 with FBPC, 15.51 with the early phase of BHPC, whereas 18.15 with the delayedphase of BHPC. In dynamic study, 19.2% increase in SUVmax in delayed scan.

CONCLUSION

SUVmax obtained in the early and delayed phase of BHPC which had 15 minutes discrepancy showed a significant difference. BHPCbefore and after FBPC may contribute to distinguish benign lesions from malignant tumors, and this technique is feasible in theclinical setting with minimum increase in examination time.

SST06-09 Multiparametric Evaluation by 18F-Fluorocholine PET-CT and MRI Examinations in Patients withProstate Cancer



The new technique 'Dynamic study of PET/CT' may contribute to distinguish benign lesions from malignant tumors, and thistechnique is feasible in the clinical setting with minimum increase in examination time.

ParticipantsXavier Palard-Novello, 35000, France (Presenter) Nothing to DiscloseAnne-Lise Blin, Rennes, France (Abstract Co-Author) Nothing to DiscloseLuc Beuzit, Rennes, France (Abstract Co-Author) Nothing to DisclosePierre Salaun, MD, PhD, Brest, France (Abstract Co-Author) Nothing to DiscloseAnne Devillers, Rennes, France (Abstract Co-Author) Nothing to DiscloseEtienne Garin, MD, Rennes, France (Abstract Co-Author) Nothing to DiscloseSolene Querellou, Brest, France (Abstract Co-Author) Nothing to DisclosePatrick Bourguet, Rennes, France (Abstract Co-Author) Nothing to DiscloseFlorence Le Jeune, MD, PhD, Rennes, France (Abstract Co-Author) Nothing to DiscloseHerve Saint-Jalmes, PhD, Rennes, France (Abstract Co-Author) Nothing to Disclose

PURPOSE

To evaluate the relationship between metabolic 18F-fluorocholine (FCH) PET-CT and functional parameters derived by MagneticResonance Imaging (MRI) in patients with prostate cancer (PC).


Patients with proven PC who underwent FCH PET/CT and 1.5 T multiparametric MRI were included. FCH PET/CT consisted in a dualphase: early and late whole-body acquisition. A 12x5' - 3x30' time sampling was reconstructed from the first 150 seconds of theearly acquisition. A freehand FCH PET/CT volume-of-interest (VOI) with a threshold of 40% of the maximum signal intensity wasdrawn on the late acquisition and projected onto the early static frame of 10 min and each frame of the dynamic reconstruction.For the kinetic analysis, an imaging-derived plasma input function was estimated from VOI placed within the largest arterial blood-pool structures available on the early PET image. The pharmacokinetic modeling was the reversible one-tissue compartment model.Kinetic parameter (K1 as influx) and static parameters (early SUVmean, late SUVmean and SUVmean retention index) wereextracted. Concerning multiparametric MRI, axial diffusion-weighted imaging was obtained using three b values: 0, 50 and 1300s/mm2. Dynamic contrast-enhanced studies were obtained with intravenous administration of gadolinium-based contrast agent witha 11x13' dynamic time sampling. Using co-registration of diffusion-weighting MRI with late whole-body FCH PET/CT, a freehand VOIwas drawn to obtain the mean Apparent Diffusion Coefficient (ADC). VOI was projected onto the perfusion parametric maps toextract the mean transfer constant (ktrans) and the mean volume of the extracellular space (Ve) using the Tofts pharmakineticmodel. Spearman's correlation coefficients were calculated to compare imaging findings.

RESULTS

Thirteen patients were analysed. The median time interval between PET and MRI was 39 days. Concerning correlation analysisbetween PET and MRI parameters, K1 was significantly correlated with ktrans (r=0.59, p=0.035) and early SUVmean wassignificantly correlated with ADC (r=-0.58, p=0.04).

CONCLUSION

FCH influx using the reversible one-tissue compartment model is significantly correlated with the transfer constant of gadolinium-based contrast agent in prostate cancer.


These results might be useful in the design of future clinical trials involving FCHOLINE-PET/DCE-MR for the assessment of prostatecancer.

SST09-01 Performance Evaluation of inSPira HD: A High Resolution SPECT System for Neuroimaging


SST09-02 A Novel Approach for Dosimetry of 90Y Radioembolization Based on Quantitative 99mTc-MAASPECT/CT Imaging


SST09

Physics (Nuclear Medicine, Quantitative Image Analysis in Imaging and Radiation Therapy)


BQ CT MI NM PH



ParticipantsChin-Tu Chen, PhD, Chicago, IL (Moderator) Board Member, BioMed Global; Board Member, EVO Worldwide, Inc; Board Member,AEPX; Board Member, EnDepth Vision Systems, LLC; Research Grant, DxRay, Inc; Advisor, RefleXion Medical Inc; Shareholder, EDDATechnology, IncRobert Miyaoka, PhD, Seattle, WA (Moderator) Research Consultant, MIM Software Inc; Research Grant, Koninklijke Philips NV;Research Grant, General Electric Company

Sub-Events

ParticipantsRameshwar Prasad, PHD, Chicago, IL (Presenter) Nothing to DiscloseVolodymyr Pylypyuk, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJoseph Ringelstein, Hines, IL (Abstract Co-Author) Nothing to DiscloseSumeet Virmani, MD, Chicago, IL (Abstract Co-Author) Nothing to Disclose

CONCLUSION

The measured performances demonstrated that the inSPira HD SPECT scanner has high sensitivity, high resolution, and acceptableimage quality and, hence, is well suited for high resolution static and dynamic SPECT neuroimaging studies.

Background

High-resolution SPECT imaging is of great interest for studying neurological pathologies. The objective of this study is tocharacterize the performances of high resolution inSPira HD SPECT scanner for neuroimaging applications.

Evaluation

inSPira HD is a dedicated high resolution SPECT scanner based on a rotating dual clamshell design that acquires data in dual-spiralgeometry. The performance characteristics were evaluated in terms of spatial resolution, sensitivity, uniformity, and contrast. Thespatial resolution was measured from images of a line source. System volume sensitivity (SVS) was calculated using large floodphantom filled with Tc-99m. ACR Small SPECT Phantom was used to evaluate the image quality in terms of uniformity and contrast.Brain phantom and patients images were acquired to access the system more realistically.

Discussion

Spatial resolution in terms of FWHM was 4.1 mm, 4.2 mm, and 4.3 mm for X, Y, and, Z plane respectively. SVS was 9914.6cts/sec/uci/ml. Integral uniformity for UFOV and CFOV were 4.8 % and 2.1 % respectively. Percent contrast for the five visiblespheres with attenuation correction was 26%, 58%, 76%, 93%, and 99%. Brain phantom and patients images show fine details ofbrain regions.

ParticipantsSara Ungania, Rome, Italy (Presenter) Nothing to DiscloseAntonino G. Guerrisi, MD, Rome, Italy (Abstract Co-Author) Nothing to DiscloseMarco D'Arienzo, Rome, Italy (Abstract Co-Author) Nothing to DiscloseMarco D'Andrea, Rome, Italy (Abstract Co-Author) Nothing to DiscloseVicente Bruzzaniti, Rome, Italy (Abstract Co-Author) Nothing to DiscloseAldo Morrone, Rome, Italy (Abstract Co-Author) Nothing to DiscloseLidia Strigari, Rome, Italy (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Radioembolization (RE) with 90Y-microspheres is a well-established treatment modality for treating liver malignancies. At a time of

SST09-03 An Investigation Study of Deep Learning Convolutional Neural Network for Whole-Body PETDenoising


SST09-04 Large Scale Assessment of Detectability and Estimability Indices from ACR CT Accreditation Database:Report of an ACR-RSNA Collaboration


increasing evidence for dose-effect relationships in RE with 90Y microspheres, the general consensus is that there is an urgentneed for accurate dosimetry in patients undergoing RE treatment. This work aimed at estimating absorbed doses to lesions andnormal liver in a novel anthropomorphic set-up.

CONCLUSION

In RE treatment planning the dose-kernel method proved to be more accurate with respect to deposition method based on full 3Ddose distributions.


Treatment planning in molecular radiotherapy is mandatory to obtain the most appropriate and effective treatment of patient. It ismandatory to validate dose quantification obtained by SPECT/CT imaging.

ParticipantsChung Chan, PhD, Vernon Hills, IL (Presenter) Nothing to DiscloseJian Zhou, PhD, Vernon Hills, IL (Abstract Co-Author) Principal Scientist, Canon Medical Systems Corporation; Wenyuan Qi, Chicago, IL (Abstract Co-Author) Nothing to DiscloseJeffrey A. Kolthammer, MS, Cleveland Heights, OH (Abstract Co-Author) Nothing to DiscloseEvren Asma, Vernon Hills, IL (Abstract Co-Author) Nothing to DiscloseLi Yang, Vernon Hills, IL (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

The goal of this study is to investigate the quantitative impact of deep residual convolutional neural network (DCNN) configurationsfor PET image denoising.


We first compared deep residual networks constructed with different numbers of layers (5 vs 8 vs 10 layers). For the 8 layersnetwork, we also compared two versions such that one has half the number of filters of the other. We trained our networks using64 clinical datasets representing a range of acquisition and reconstruction protocols. We evaluated the performance of differentnetworks on 4 different human 18F-FDG studies acquired for 4 minutes per bed position. Synthetic lung and liver lesions weregenerated using GATE simulation and inserted into the listmode data. The listmode data were further rebinned into 2-minute and 3-minute lists in order to test the robustness of networks against different noise levels. All images were reconstructed using OSEMwith 3 iterations, 10 ordered subsets and applied with a Gaussian filter (GF) at 6 mm FWHM. Quantifications were assessed bymeasuring the lesion contrast recovery versus variability of the background liver uptake.

RESULTS

The 10 layers and 8 layers networks with the full number of filters resulted in comparable quantitative and qualitative performance.However, the 10 layers network used 2-fold the training time than the 8 layers network did. Reducing the layers to 5 resulted inreduced lesion contrast recovery and robustness to the noise levels in the input images. Reducing the number of filters in the 8layers network also reduced quantitative performance compared to the full 8 layers network.

CONCLUSION

The network architecture plays an important role in the denoising performance of a DCNN network. While fewer layers fail tocapture the full complexity of the noise distributions, too many layers result in over-parameterization and difficulties in trainingwithout substantial performance gain.


A properly chosen denoising neural network architecture can significantly improve noise levels over alternative architectures andhelp improve clinical decisions made based on resulting images.

ParticipantsEhsan Samei, PhD, Durham, NC (Abstract Co-Author) Research Grant, General Electric Company; Research Grant, Siemens AG;Advisory Board, medInt Holdings, LLC; License agreement, 12 Sigma Technologies; License agreement, Gammex, IncYakun Zhang, MS, Durham, NC (Abstract Co-Author) Nothing to DiscloseLaura P. Coombs, PhD, Reston, VA (Abstract Co-Author) Nothing to DiscloseTaylor Smith, Durham, NC (Presenter) Nothing to Disclose


[email protected]

PURPOSE

To assess the variability in a subset of image quality attributes of clinical CT systems using advanced task-based metrics,detectability and estimability, across the US through a collaboration between the ACR and the RSNA (QIBA).


SST09-05 A Comprehensive Platform for Automated Analysis and Reporting of QC of Medical ImagingModalities


SST09-06 An Embedded Pre-Screening Electrochemical Profile in Label-Free and Image-Free Quantified TissueBehavioral Analysis


The ACR CT accreditation program requires institutions to submit images using the Gammex 464 phantom. Through a collaborationbetween the RSNA QIBA and ACR, a set of 804 randomly-selected de-identified phantom data sets were analyzed using anautomated image quality estimator. These data were not filtered with regard to ACR phantom pass/fail outcome. Basic image qualitymetrics, including HU numbers, CNR, and noise magnitude, were automatically extracted, as well as task-specific resolution (TTF)and noise power spectra. Two advanced task-based metrics that incorporate all the aforementioned metrics were also computed:detectability index (d') and estimability index (e') for CT volumetry, both with demonstrated correlation with human and machineperformance. Results from each image set were binned into specific protocols. The data were analyzed in terms of variability acrossprotocols.

RESULTS

For task-specific resolution, despite the wide spread of the frequencies for 0.5 TTF (f50) within each protocol, the median valuewas largely consistent across protocols (0.40-0.41 1/mm). Noise magnitude and CNR values, related to radiation dose, were highlydependent on the protocol. The d' median values and distribution changed considerably with the task definition. Similar trends werealso observed for e'. The d' and e' median values for polyethylene were 142.3, 168.1, 85.8, and 129.1; and 0.026, 0.035, 0.017,and 0.024 for adult abdomen, adult head, pediatric abdomen, and pediatric head, respectively.

CONCLUSION

Assessment of detectability and estimatiblity indices from the ACR CT accreditation database was feasible. Phantom images can beused as a surrogate to ascertain variability across clinical operations. Such data could eventually be used in the development offuture conformance assessment procedures.


Analysis of reference phantom datasets across systems and institutions through the ACR Accreditation process enables granularassessment of multi-parameter variability across our national healthcare system enabling the development of key performanceindicators, qualification of quantitative performance, and potential nation-wide image quality registries.

ParticipantsArnold Schilham, PhD, Utrecht, Netherlands (Presenter) Nothing to DiscloseRalph Berendsen, Heerlen, Netherlands (Abstract Co-Author) Nothing to DiscloseDennis Dickerscheid, PhD, Nieuwegein, Netherlands (Abstract Co-Author) Nothing to DiscloseRob van Rooij, PhD, Utrecht, Netherlands (Abstract Co-Author) Nothing to DiscloseAnne Talsma, MS, Groningen, Netherlands (Abstract Co-Author) Nothing to DiscloseTim C. De Wit, PhD, Amsterdam, Netherlands (Abstract Co-Author) Nothing to DiscloseJoost P. Kuijer, PhD, Amsterdam, Netherlands (Abstract Co-Author) Nothing to Disclose


[email protected]

[email protected]

CONCLUSION

This QC platform provides a tested, comprehensive solution for analysis and reporting of QC for most medical imaging modalities.

Background

Periodic testing of image quality is part of Quality Control (QC) of medical imaging devices. Implementation of QC testing is modalityand model specific. If vendor-supplied QC is available, the details are not disclosed and exporting QC results is rarely possible.Those results are needed to spot drifts in system performance and to inform all users about the time and outcome of the latest QC.To aid implementation of a department-wide multi-modality QC program, the Society for Medical Physics of the Netherlands (NVKF)devised a community-driven, open source platform for automated analysis of medical images for quality control. The platformcomprises a PACS, analysis modules, a database, and web based front-ends for administration and reporting. Analysis modules areavailable for most modalities, including MR, CT, MG, US, DX, RF and NM. In a normal workflow a user performs a QC test and sendsthe images to the QC PACS. Based on DICOM metadata, the platform runs the appropriate analysis modules and stores the resultsin the database. The web-based interface allows immediate access to the QC results.

Evaluation

This platform has been used since 2013 in several Dutch hospitals, successfully analyzing tens of thousands of datasets. Easyaccess to recent reports and trend plots of QC metrics were found beneficial for successful implementation of a QC program. Due tothe open-source nature of the project, analysis modules are reviewed and imporved continually.

Discussion

This platform is a community driven, open source project, with only open source third party software requirements. Its goal is acomprehensive solution for analysis and reporting of QC for all medical imaging modalities. Analysis modules were initially developedfor specific models of modalities. Easy exchange of analysis modules and configurations between institutes resulted in a push fromthe user community towards generalizations to other models and brands. After extensive testing, the platform was publiclyreleased.

ParticipantsAli Zarafshani, Norman, OK (Presenter) Nothing to Disclose

SST09-07 Development and Initial Evaluation of a DaTscan Digital Reference Object


Yunzhi Wang, Norman, OK (Abstract Co-Author) Nothing to DiscloseFaranak Aghaei, Norman, OK (Abstract Co-Author) Nothing to DiscloseSeyedehnafiseh Mirniaharikandehei, MS, Norman, OK (Abstract Co-Author) Nothing to DiscloseLiangzhong Xiang, PhD, Norman, OK (Abstract Co-Author) Nothing to DiscloseBin Zheng, PhD, Pittsburgh, PA (Abstract Co-Author) Nothing to Disclose


[email protected]@ou.edu

PURPOSE

To address the overcharging challenge of improving the efficiency of cancer screening programs, a label-free and image-free pre-screening tool to detect a frequency-dependent electrochemical profile (or features) of body tissues is developed and tested.


The study reports the integration of a unique pre-screening tool to non-invasively detect and analyse electrochemical tissueattributes. This pre-screening tool consists of a unique probe configuration including 8-sub-regional probes that are symmetricaldistributed in a mirror reflection. The tool uses a 4-active-probe sensing (4-AP sensing) technique through which electrochemicalprofile of body tissues are measured with high-contrast bioelectrical properties. Therefore, information regarding the extraction oflocal tissue distinguishability and quantify of tissue behaviour in the frequency domain during asymmetric pathways in theelectrochemical profile can be readily assessed without the use of other invasive techniques. The pre-screening tool is also uniquein that it utilizes the delivery and sensing channels in a single-probe electrochemical analytical measurement attached directly toexternal layers of the body (e.g., breast) skin thereby preventing irreproducibility of results.

RESULTS

The performance of the pre-screening tool has been verified by analysis of electrochemical profile using several breast tissue andphantom studies with different simulated tissue density levels at different sub-regions. The results indicated that the tool enabledto sensitively detect, quantify and distinguish the simulated breast tissue and local asymmetrical distribution. The testing resultsare also quite reproducible and can be offer as a personalized measurement analysis.

CONCLUSION

This research for the first time demonstrates a unique electrochemical sensing device coupled with a quantified behavioural analysisof breast tissue in an innovative pre-screening tool for label-free analysis of breast tissue bioelectrical features. In future humanstudies, it has potential to build a new low-cost and easy-to-use tool for pre-screening several types of cancers (i.e., breastcancer) to increase cancer detection yield and reduce false positive recalls in current screening methods.


this is a pre-screening tool can be used to increase cancer detection yeild and reduce false-positive recalls (over-diagnosis) incurrent mammography screening

ParticipantsRobert Miyaoka, PhD, Seattle, WA (Presenter) Research Consultant, MIM Software Inc; Research Grant, Koninklijke Philips NV;Research Grant, General Electric CompanyLarry Pierce, PhD, Seattle, WA (Abstract Co-Author) Nothing to DiscloseRobert L. Harrison, Ludwigsburg, Germany (Abstract Co-Author) Research Consultant, General Electric Company


[email protected]

PURPOSE

The goal of this work was to design and evaluate 123I-Ioflupane DaTscan digital reference objects (DROs) and to create realisticprojection datasets from a DRO to allow testing of DaTscan image reconstruction methods.


The initial striatal brain DRO was created from an MRI image of an age-appropriate patient. The MRI was segmented to delineatestructures of interest. Each structure was assigned a value according to expected Ioflupane uptake: CSF : background : rightputamen : left putamen : caudate => 0.33 : 1: 5.5 : 3.25 : 5.5. This corresponds to a specific binding ratio (SBR) of 4.5 for thecaudate and right putamen and an SBR of 2.25 for the left putamen. The initial, noise-free DRO was then smoothed using 6 mm and10 mm Gaussian filters. The three versions were analyzed by four analysis packages. Next Monte Carlo simulation was used tomodel SPECT acquisition of the 123I-Ioflupane DRO. The simulation included multiple energy emissions and collimator/detectorresponse modeling. After creation of a high count projection sinogram, Poisson-resampling was used to generate sinograms ofdifferent noise levels. Projection sinograms were reconstructed on a vendor workstation. Images were then analyzed usingDaTquant™ software.

RESULTS

An Ioflupane DRO developed for assessing DaTscan analysis software was tested. The SBR for the striatum varied from 2.53-3.19(right, healthy) to 1.81-2.05 (left, diseased putamen) for the non-blurred DRO for the different analysis packages. The caudateSBRs varied from 2.4-3.6 (right) to 2.33-3.05 (left). The putamen SBRs varied from 2.56-2.7 (right) to 1.7-2.14 (left, diseased).There was similar variability for the blurred DRO images. The SBRs for the Monte Carlo simulated DRO projection imagesreconstructed on a vendor workstation were similar to SBRs from experimentally collected phantom data.

CONCLUSION

DROs with known 123I-Ioflupane activity distributions have been created from a patient MRI. Results reveal the variability in the

SST09-08 Quantitative Iodine Maps from Spectral Detector Computed Tomography in Daily Practice: A RealWorld Study in 60 Patients


SST09-09 Quantifying Liver Iron Content Using Photon-Counting Detector Based Computed Tomography


calculated SBR for different analysis software. Monte Carlo simulation data sets have been created with full models of collimatordetector response for testing DaTscan analysis packages and image reconstruction methodologies.


DROs can be used to test/compare vendor analysis packages. Further, Monte Carlo simulations of DROs can be used totest/validate enhanced image reconstruction methods for 123I-Ioflupane brain imaging.

ParticipantsNils Grosse Hokamp, MD, Cleveland, OH (Presenter) Nothing to DiscloseLeslie Ciancibello, RT, Cleveland , OH (Abstract Co-Author) Consultant, Cassling Group Consultant, Siemens AGOmer Alabar, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseDavid W. Jordan, PhD, Cleveland, OH (Abstract Co-Author) Nothing to DiscloseKarin A. Herrmann, MD, PhD, Cleveland, OH (Abstract Co-Author) Nothing to Disclose


[email protected]

PURPOSE

Iodine maps are available from spectral detector computed tomography (SDCT). They visualize the distribution of iodinated contrastmedia and allow for its quantification. While the accuracy of such maps has been investigated ex vivo in phantoms under optimizedconditions, this study aimed to evaluate their accuracy in routine examinations.


We designed a reference object consisting of 5 tubes filled with potassium iodide solutions of different concentrations (0.0, 0.8,1.6, 3.2, 4.8 mg/ml). In 60 patients (31 male, mean age 65.2±11.4 years) who underwent routine SDCT of the abdomen, thereference object was positioned within the scan region at random positions/orientations. Iodine maps (IM) were reconstructed usingthe vendor provided algorithm. Two circular regions of interest were drawn in each tube, mean and standard deviation within eachROI were determined and averaged. Difference between IM and known iodine concentration was calculated. We performed aconfounder analysis for BMI and radiation dose. In addition, we developed a software that allows for automatic measurement of theanterior-posterior and left-right diameter of each patient and further, computes a three-compartment model of the examined tissue(fat, bone, lean); all parameters were considered as possible confounders in statistic assessment.

RESULTS

Overall, iodine maps had a high accuracy as indicated by an offset of -0.01±0.12 mg/ml despite their random placement. The offsetwas found greater in low concentrations (relative offset in 0.8 and 4.8 mg/ml tubes: +6.5% and +1%, respectively). While smallerBMI tended to overestimate iodine content, greater BMI tended to underestimate iodine concentrations. Besides BMI, volume ofbone was found to be a strong confounder regarding iodine quantification accuracy.

CONCLUSION

Iodine quantification using iodine maps from SDCT is technically feasible in routine examinations. When considering estimatedconcentrations for clinical decision making, increased caution is recommended in low concentrations, in skinny and obese patientsand in images with presence of large amount of bone.


Iodine quantification in daily practice is feasible with high accuracy using iodine maps from SDCT.

ParticipantsKishore Rajendran, PhD, Rochester, MN (Presenter) Nothing to DiscloseShengzhen Tao, Rochester, MN (Abstract Co-Author) Nothing to DiscloseGregory J. Michalak, PhD, Rochester, MN (Abstract Co-Author) Nothing to DiscloseShuai Leng, PHD, Rochester, MN (Abstract Co-Author) License agreement, Bayer AGCynthia H. McCollough, PhD, Rochester, MN (Abstract Co-Author) Research Grant, Siemens AG


[email protected]

PURPOSE

In a phantom. to quantify liver iron content using multi-energy photon-counting detector (PCD) CT and an image-based materialdecomposition technique.


A wet liver analog was synthesized using dry liver extract (1:7 dilution of dry liver in water) to yield a base CT numberenhancement of 50 HU at 140 kV, similar as that of normal human liver. Iron concentrations (iron III nitrate) were prepared at 1.8,2.8 and 5.6 mg/mL with either wet liver or water as a solvent. Cylindrical vials containing approximately 10 mL of pure liver, iron-liver and iron-water solutions were placed inside a 20 cm multi-energy CT phantom (Gammex, Sun Nuclear, WI), and scanned usinga whole-body PCD-CT system (Siemens CounT, Forchheim, Germany) at 120 kV, 178 mAs (CTDIvol = 13 mGy) and energythresholds of 25, 63 keV. Multi-energy PCD-CT images were reconstructed at [25, 120] keV, [63, 120] keV and [25, 63] keV energyranges using a quantitative smooth kernel (D30). Quantitative iron maps were obtained using an in-house material decomposition

technique with the [25, 120] keV and [63-120] keV images assuming 3 base materials: liver, water and iron. The image domainmaterial decomposition (MD) algorithm is based on spectral prior image constrained compressed sensing (MD-SPICCS) whichcombines material decomposition and denoising into a unified framework. The mean density (mg/mL) of iron content was estimatedfrom the iron maps generated using MD-SPICCS. A linear regression analysis was performed to compare measured ironconcentrations in liver background with known true concentrations.

RESULTS

MD-SPICCS was able to successfully detect and quantify iron from liver background in the phantom images. Liver background fromthe iron-liver mixtures was assigned to the liver map, while water background from the iron-water mixtures was assigned to watermap. Linear regression showed excellent correlation between measured and true iron concentrations in the iron-liver mixtures (slope= 1.1, R2 = 0.9997, RMSE =0.6 mg/mL).

CONCLUSION

We have demonstrated accurate iron quantification in a liver phantom scanned using the PCD-CT system and an image-basedmaterial decomposition technique. Measured iron concentrations showed excellent correlation with the ground truth.


Diagnosing hemochromatosis characterized by liver iron overload using imaging methods requires accurate iron quantification tofacilitate and monitor therapy.

Documents

Nuclear Medicine - Radiological Society of North …archive.rsna.org/2018/NuclearMedicine.pdfNM104-ED-X Spectrum of Benign Bone and Soft Tissue Findings on Sodium-Fluoride (NaF) PET