ImageImage--Guided and Adaptive Guided and Adaptive RadiotherapyRadiotherapyRadiotherapy Radiotherapy

PMH 50PMH 50thth ConferenceConference

A.A. FylesFylesyyDept. of Radiation OncologyDept. of Radiation OncologyPrincess Margaret HospitalPrincess Margaret Hospital

IMRT forIMRT forCervix CancerCervix Cancer

Four-field

IMRT

M. Milosevic, PMH-RMP IGRT, 25/1/2008

Challenges in IMRT for Cervix CancerChallenges in IMRT for Cervix Cancergg

Target IdentificationMR CT• MR vs CT

• Nodal targets• Online image qualityOnline image quality

Organ motion• Influence of

– Bladder filling– Rectal filling g– Normal uterus position

WorkloadWorkload

PMH CTV DefinitionPMH CTV Definition

GTVCervix

Vagina

Uterus

Cervix

CTV

g

Parametria

• CTV = GTV (+0.7cm margin) + Cervix + parametria + lower 2cm uterus + upper 2cm vagina• Nodal CTV as per Taylor et alNodal CTV as per Taylor et al• Normal Organs at Risk

• rectum, bladder, sigmoid, bowel, bone etc.

Point of Interest Analysis in 16 PatientsPoint of Interest Analysis in 16 Patients Mean ValuesMean ValuesPoint of Interest Analysis in 16 Patients Point of Interest Analysis in 16 Patients –– Mean ValuesMean Values

14.0 mm 3.8 mm

6.2 mm

9.6 mm

1.7 mm1.4 mm

1.6 mm5.6 mm

INTERFRACTION INTRAFRACTION

+

Internal Uterine FundusInternal Uterine Fundus

Cervical OsCervical Osy +

z

Point of Interest Analysis in 16 Patients Point of Interest Analysis in 16 Patients –– Mean ValuesMean Values

14.0 mm 3.8 mmUB = UB = Uncertainty Uncertainty

B dB d

CranioCranio--CaudalCaudal(mm)(mm)

AntAnt--PostPost(mm)(mm)

6.2 mm

9.6 mm

1.7 mm1.4 mm

1.6 mm5.6 mm

Bounds Bounds (confidence (confidence

limits)limits)

(mm)(mm) (mm)(mm)

INTERFRACTION INTRAFRACTIONTotalTotal IntraIntra TotalTotal IntraIntra

Fundus UBFundus UB 36.936.9 9.69.6 16.416.4 4.34.3

+

Internal Uterine FundusInternal Uterine Fundus

Cervical OsCervical Os

Fundus UBFundus UB 36.936.9 9.69.6 16.416.4 4.34.3

Os UBOs UB 14.814.8 3.63.6 24.824.8 4.04.0y +

z

Adapting to ChangeAdapting to Change

•• Tumors are heterogeneousTumors are heterogeneous•• Position changesPosition changes

+

–– InterInter--fraction, Intrafraction, Intra--fractionfraction

•• Size and shape changesSize and shape changes•• Biology changesBiology changes

+ +

8 GyPre-Tx

Tumour Shrinkage & Deformation

20 Gy 28 Gy

• Influence of:

– Tumour factors• Biology

– Hypoxia

T t t f t

38 Gy 48 Gy

– Treatment factors• Chemo-radiation

Adaptive ImageAdaptive Image--Guided RTGuided RTAdaptive ImageAdaptive Image--Guided RTGuided RT

• Complex dynamics of OAR & t t tiOAR & target motion

• Concerns:M i ?– Margins?

– Geographical miss

– ↑ dose to normal tissues↑ dose to normal tissues

• What you see is not What you get

• Adaptive approach is needed for treatment of intact cervix cancerintact cervix cancer

IMRT Lymph Node PlanIMRT Lymph Node PlanIMRT Lymph Node PlanIMRT Lymph Node Plan

Role for image-guidedguided nodal targeting using gUSPIO MR

Lymph Node MarginsLymph Node Margins

2 cm radial expansionaround iliac vessels

Shih, 2005Shih, 2005 Taylor, 2005Taylor, 2005

LN Movement During RTLN Movement During RT

Coverage of Unified nCTV7

105

100

105

age

Day 1 CTV

Summed CTV

95

enta

ge C

over

a85

90Perc

e85

0 4 8 12 16 20

2-mm Expansions of nCTV7Chan and DinniwellChan and Dinniwell

4 mm PTV expansion encompasses 4 mm PTV expansion encompasses >90% of LN movement>90% of LN movement

M. Milosevic, PMH-RMP IGRT, 25/1/2008

Delivered vs. Planned Dose Delivered vs. Planned Dose Image at T=0 Image at T=1

Mesh-BasedBiomechanical Model

•• MotionMotion•• MotionMotion•• DeformationDeformation•• Dose accumulation Dose accumulation •• Adaptive reAdaptive re--planningplanning

MORFEUSMORFEUSMORFEUSMORFEUS Slide c/o Kristy BrockSlide c/o Kristy Brock

pp p gp g

CT_0

BASELINE WEEK 1 WEEK 2 WEEK 3 WEEK 4 WEEK 5

MRI_0

ContoursContours

MRI_1

ContoursContours ContoursContours

MRI_2

ContoursContours

MRI_3

ContoursContours

MRI_4

ContoursContours

MRI_5

Surface Surface MeshesMeshes

Surface Surface MeshesMeshes

Surface Surface MeshesMeshes

Surface Surface MeshesMeshes

Surface Surface MeshesMeshes

Surface Surface MeshesMeshes

Nominal Nominal PlanPlan

Dose Dose Accumulated Accumulated

PlPlMeshesMeshes MeshesMeshes MeshesMeshes MeshesMeshes MeshesMeshesMeshesMeshes PlansPlans

Acc DoseAcc Dose Acc DoseAcc Dose Acc DoseAcc Dose Acc DoseAcc Dose Acc DoseAcc Dose

Deformable Dose AccumulationDeformable Dose AccumulationDeformable Dose AccumulationDeformable Dose Accumulation Dose Dose Accumulated Accumulated

Acc Dosewk1

Acc Dosewk1

Acc Dosewk2

Acc Dosewk2

Acc Dosewk3

Acc Dosewk3

Acc Dosewk4

Acc Dosewk4

Acc DoseWk5

Acc DoseWk5

PlansPlans

Nominal Nominal PlanPlan

Adaptive Adaptive Adaptive Adaptive ReRe--plan plan

Initial IMRT strategyInitial IMRT strategy –– 10 pts10 ptsInitial IMRT strategy Initial IMRT strategy –– 10 pts10 pts

• LM (large margin) plan • SM (small margin) plan– 1-2 cm PTV margin– Expectation

• little OAR sparing

– 0.5 cm PTV margin– Expectation

• Some OAR sparinglittle OAR sparing• Target always covered

Some OAR sparing• Target missed

Initial IMRT strategyInitial IMRT strategy –– 10 pts10 ptsInitial IMRT strategy Initial IMRT strategy –– 10 pts10 pts

• Adaptive re-plan p p– Union CTV of 1st 3 image sets (iCTV)– Expectation

M OAR i• More OAR sparing• Target coverage better

• Results– LM plan covered most pts– SM plan seemed adequate for more pts than

expectedexpected– Re-plan did not spare as much OAR as hoped

• WHY?WHY?

WHY?WHY?WHY?WHY?

• Large CTV – tumour shrinkage did not influence treatment volume as much as expected

• Organ motion – iCTV – resulted in even larger CTVs than baselinethan baseline

• Estimates of 3D margins too generous based on previous POI data

• Influence of bladder / rectal filling– Heterogeneous despite bowel & bladder prep

Caused large displacements in CTV location in some pts– Caused large displacements in CTV location in some pts

• Influence of uterus position

22ndnd Adaptive strategyAdaptive strategy –– 25 pts25 pts22 Adaptive strategy Adaptive strategy –– 25 pts25 pts

• Use of ultra-small PTV margins (0.3cm)g ( )• Assume perfect set-up daily (CBCT)• Separate out pts into consistent & variableSeparate out pts into consistent & variable

bladders• Adaptive re-plan – after 2nd wk treatment, use p p ,

only imaging at that time point, no union CTV.

25pts25pts98% of the PTV to be covered by the 95% isodose for all plans98% of the PTV to be covered by the 95% isodose for all plans

Plan evaluation criteria:Plan evaluation criteria:98% of GTV received 5000cGy; 98% HRCTV received ≥ 4900cGy98% of GTV received 5000cGy; 98% HRCTV received ≥ 4900cGy

Dose Accumulation with InterDose Accumulation with Inter--Fraction Motion AppliedFraction Motion Applied

Consistent BladdersConsistent Bladders

Dose Accumulation with InterDose Accumulation with Inter Fraction Motion AppliedFraction Motion Applied

Variable BladdersVariable Bladders

Nom SMNom SM Acc SMAcc SM Acc AMAcc AM

GTV GTV (D98)(D98) 50365036 50145014 50315031

HRCTV HRCTV (D98)(D98) 49234923 48584858 49094909

Nom SMNom SM Acc SMAcc SM Acc AMAcc AM

GTV GTV (D98)(D98) 50245024 49874987 50085008

HRCTV HRCTV (D98)(D98) 49114911 47834783 47804780HRCTV HRCTV (D98)(D98) 49234923 48584858 49094909

Rectum Rectum (V45)(V45)

5454 5151 4141

Sigmoid Sigmoid 5959 6161 3939

HRCTV HRCTV (D98)(D98) 49114911 47834783 47804780

Rectum Rectum (V45)(V45) 5252 5050 4141

Sigmoid Sigmoid (V45)(V45) 6161 6767 5151Sigmoid Sigmoid

(V45)(V45)5959 6161 3939

Bladder Bladder (Mean)(Mean)

40024002 41034103 39633963

Bladder Bladder (Mean)(Mean)

40654065 39603960 39303930

P < 0.05P < 0.05 compared to Nominal SM plan

Conclusions of MR StudyConclusions of MR Study• Tumour & organ dynamics within the

pelvis is complex

• Inter-fraction motion is a major issue in cervix cancer, supporting IGRT

S ll i t f ti d LN• Small inter-fraction and LN motion/deformation

• Tumour regression during treatment g gsuggests adaptive IGRT

• Conventional plan ≠ Delivered dose

• Daily image-guidance and adequate bladder prep is essential

• Adaptive strategy may be successful forAdaptive strategy may be successful for some patients

Cervical Stump CarcinomaCervical Stump CarcinomaCervical Stump CarcinomaCervical Stump Carcinoma

• 45 year old woman, y ,previous sub-total hysterectomy for ffibroids

• MR– 6 cm. tumor

replacing cervix, small pelvic nodesp

– sigmoid looped over over tumor

IMRT Boost TreatmentIMRT Boost TreatmentIMRT Boost TreatmentIMRT Boost Treatment

• 5-10% of patients - • Boost to 70 - 75 Gy unsuitable for brachytherapy

• Gross nodal disease may be treated to 65-70 Gy

• Cervical stump

• Sidewall or nodal disease that receives less than 60

• Other means of radiation boost

St d d EBRTthat receives less than 60 Gy (EBRT and IC)

• PMH IMRT dose of 25.2

– Standard EBRT

• Limited by normal tissue tolerance

Gy in 14 fractionstissue tolerance (dose of 60-65 Gy)

IMRT BoostIMRT Boost -- IssuesIssuesIMRT Boost IMRT Boost -- IssuesIssues

• Boost Volume• Cervix alone – simulating IUA brachytherapy• Cervix and pelvic sidewall(s)

• Organ movement.I l di bl dd d t l ti• Including bladder and rectal preparation

• Set-up reproducibilityI bili ti• Immobilization

IMRT Boost Spares Normal OrgansIMRT Boost Spares Normal OrgansBladder/Rectum

Femoral heads

ParaPara--aortic Radiationaortic RadiationParaPara--aortic Radiationaortic Radiation

• Positive pelvicPositive pelvic nodes and/or small PAN or recurrence

• Concurrent cisplatin

• Risk of toxicity GI, hematologic

ParaPara--aortic Radiation with Simultaneous aortic Radiation with Simultaneous Integrated BoostIntegrated BoostIntegrated BoostIntegrated Boost

Dose-Volume HistogramKidneys/Small BowelHistogramBowel

PostPost--Operative Adjuvant Pelvic IMRTOperative Adjuvant Pelvic IMRT

• Recent RTOG Phase II trial of feasibilityy

• Toxicity decreased esp. GI

• Possible benefit if chemo-radiation

• Target motion issuesTarget motion issues minimized with ITV (simulation bladder

t d f ll)empty and full)• Cone-beam imaging

for bone matchingfor bone matching

MR ImageMR Image GuidedGuidedMR ImageMR Image--Guided Guided BrachytherapyBrachytherapy

for Cervix Cancerfor Cervix Cancerfor Cervix Cancer for Cervix Cancer

• Adapt to changing volumeAdapt to changing volumep g gp g g•• IRIR--CTVCTV::

Tumor at diagnosisTumor at diagnosis20 Gy20 Gyyy

•• HRHR--CTVCTV::Residual tumor at BTResidual tumor at BT40 Gy from BT40 Gy from BT40 Gy from BT40 Gy from BT

•• Reduced toxicityReduced toxicity•• Identification ofIdentification of

“non“non responders”responders”“non“non--responders”responders”

M. Milosevic, PMH-RMP IGRT, 25/1/2008

ImageImage--Guided BrachytherapyGuided BrachytherapyStd brachytherapy is not optimal for all patientsStd brachytherapy is not optimal for all patientsy py p py py p p

90

100

70

80

%)

40

50

60

Volu

me

(%

HR-CTV20 patients

10

20

30

V Standard plansMR-dosimetry

(Zwahlen)V100

0

10

0 20 40 60 80 100 120 140 160 180 200Dose (%)

ImageImage--Guided BrachytherapyGuided Brachytherapy

303030

gg y pyy py

20

25

me

(cm

3 )

20

25

me

(cm

3 )

20

25

me

(cm

3 ) HR-CTV20 patientsStandard plans

15

wal

l vol

um Rectum15

wal

l vol

um Rectum15

wal

l vol

um RectumMR-dosimetry(Zwahlen)

5

10

Rec

tal w

D2cm35

10

Rec

tal w

5

10

Rec

tal w

D2cm3

00 1000 2000 3000 4000 5000

D2cm0

0 1000 2000 3000 4000 50000

0 1000 2000 3000 4000 5000

D2cm

Dose (cGy)Dose (cGy)Dose (cGy)

M. Milosevic, PMH-RMP IGRT, 25/1/2008

Optimized IGOptimized IG--BrachytherapyBrachytherapyD2cm3

OAR Tumor size Conventional

(Gy) Optimized

(Gy) Change (%)

AllOptimizationOptimization

All tumors 25 ±8.7 22 ±6.6 -14%*

Small 26 ±3.2 19 ±5.7 -32%* Sigmoid

colon Large 25 ±7.4 24 ±6.7 -3%

20 cases20 casesHRHR--CTV CTV >>40 Gy40 GyIRIR CTVCTV >>20 G20 G Large 25 ±7.4 24 ±6.7 3%

All tumors 16 ±6.2 16 ±6.7 0

Small 16 ±7.8 14 ±7.6 -14%* Rectum

IRIR--CTV CTV >>20 Gy20 GyPoint A Point A <<40 Gy40 Gy

Large 17 ±4.5 19 ±4.5 +14%* All

tumors 31 ±9.2 30 ±9.8 -2%

Small 30 ±8.7 26 ±9.6 -12%* Bladder

Large 32 ±9.9 34 ±8.6 +6%

Daniel Zwahlen, 2007Daniel Zwahlen, 2007M. Milosevic, PMH-RMP IGRT, 25/1/2008

Optimized IGOptimized IG--BrachytherapyBrachytherapypp y pyy py

145 patients145 patientsLocal control

T1b/2a 15%T1b/2a 15%T2b/3a 59%T2b/3a 59%

Local control

CSS

T3b 26%T3b 26%>5 cm 54%>5 cm 54%

OS

LN 39%LN 39%RTRT++CTCT

Late ToxicityLate Toxicity

Gr. 3Gr. 3--4 GI 4%4 GI 4%Optimized BTOptimized BT

8080--85 Gy85 Gy

G 3G 3 G %G %GU 4%GU 4%

Vaginal 5%Vaginal 5%

Potter, 2007Potter, 2007M. Milosevic, PMH-RMP IGRT, 2008

Optimized IGOptimized IG--BrachytherapyBrachytherapypp y pyy py

145 patients145 patientsLocal controlSmall tumors

T1b/2a 15%T1b/2a 15%T2b/3a 59%T2b/3a 59%

Local control

CSSLarge tumors

Small tumors

T3b 26%T3b 26%>5 cm 54%>5 cm 54%

OS

LN 39%LN 39%RTRT++CTCT

Late ToxicityLate Toxicity

Gr. 3Gr. 3--4 GI 4%4 GI 4%

Local Control

Optimized BTOptimized BT8080--85 Gy85 Gy

G 3G 3 G %G %GU 4%GU 4%

Vaginal 5%Vaginal 5%

Potter, 2007Potter, 2007M. Milosevic, PMH-RMP IGRT, 2008

What Next?What Next?What Next?What Next?

• Work by our group & th

• Contouring l t &others

– poor correlation between rectal & bladder volumes and cervix position

nomenclature & guidelines (e.g. GEC-ESTRO and Post-op RTOG)and cervix position

• Look at influence of uterus position on target

ti

RTOG)– International

collaboration – RTOG group in June 08motion

• Simple class solution / margin recipe unlikely

group in June 08• Any adaptive strategy

needs to be i di id li d

g ygiven complexity of target motion & deformation

individualised• At present IMRT not

ready for intact cervixy

What next?What next?What next?What next?

• Validate modeled findings with independent g pcohort of pts with more frequent imaging– 2-3 MR scans per wk– Initial bladder full & bladder empty MRIs– CBCT during treatment

E l f ibili f i CBCT• Explore feasibility of contouring on CBCT imagesUlti t l ti d ti• Ultimate goal – propose prospective adaptive IMRT strategy for cervix patients

Princess Margaret HospitalPrincess Margaret HospitalRadiation Oncology: Radiation Oncology: M. Milosevic, K. Lim, P. Chan, D. Zwahlen, R. Dinniwell, W. M. Milosevic, K. Lim, P. Chan, D. Zwahlen, R. Dinniwell, W.

Levin, L. ManchulLevin, L. ManchulMedical Physics: Medical Physics: YY--B Cho, K. Brock, J. Jezioranski, I. Yeung, T. Craig,B Cho, K. Brock, J. Jezioranski, I. Yeung, T. Craig,

A BeikiA Beiki Ardakani J Stewart J Moseley M Sharpe D JaffrayArdakani J Stewart J Moseley M Sharpe D JaffrayA. BeikiA. Beiki--Ardakani, J. Stewart, J. Moseley, M. Sharpe, D. JaffrayArdakani, J. Stewart, J. Moseley, M. Sharpe, D. JaffrayMedical Imaging: Medical Imaging: M. HaiderM. HaiderRadiotherapy: Radiotherapy: V. Kelly, J. PatersonV. Kelly, J. PatersonBiostatistics:Biostatistics: M Pintilie G LockwoodM Pintilie G LockwoodBiostatistics: Biostatistics: M. Pintilie, G. LockwoodM. Pintilie, G. Lockwood

RaySearch Laboratories ABRaySearch Laboratories ABA. Lundin, H. Rehbinder, J. LA. Lundin, H. Rehbinder, J. Lööff

Supported by:Supported by:National Cancer Institute of Canada through the Terry Fox RunNational Cancer Institute of Canada through the Terry Fox RunGiovanni and Concetta Guglietti Family TrustGiovanni and Concetta Guglietti Family Trust

M. Milosevic, PMH-RMP IGRT, 25/1/2008