CASE REPORT

Magnetic resonance imaging of abnormal shoulder pain followinginfluenza vaccination

Gokcan Okur & Kimberly A. Chaney &

Laurie M. Lomasney

Received: 20 November 2013 /Revised: 21 January 2014 /Accepted: 17 March 2014# ISS 2014

Abstract The influenza vaccine is increasingly available tothe general public and mandated by many employers in theUnited States. The prevalence of post-vaccination complica-tions is likely on the rise. Complications are well known togeneral clinicians, but are under-reported in the imaging liter-ature. We present four cases of post-vaccination shoulder painwith magnetic resonance imaging (MRI) findings. Anintrasubstance fluid-like signal in deep muscular and/or ten-dinous structures was the most common finding on MRI ofthese four cases. Focal bonemarrow signal within the humeralhead and inflammatory changes in the subacromial/subdeltoidbursa were also observed. The most likely reason for a hu-meral intraosseous edema-like signal was presumed injectionof vaccine substance directly into osseous structures thatmight lead to focal osteitis. In the published literature, thereis little emphasis on the imaging of local injection site com-plications accompanying influenza vaccination. We intendedto increase familiarity of MRI findings in the setting ofprolonged or severe clinical symptoms following influenzavaccination through the imaging findings of these four cases.

Keywords Influenza vaccination . ShoulderMRI . Bursitis .

Bonemarrow edema

Introduction

Widespread availability of the influenza vaccination to thegeneral public and increasing employer mandates have in-creased the number of people receiving the flu vaccine.Complications such as prolonged and severe shoulder painare rare but are likely to become more prevalent. Requests foradvanced imaging of potential vaccination-related complica-tions are also likely to increase. Since most complications ofinfluenza vaccination are self-limited and do not require im-aging, radiologists may be less likely to consider the relation-ship between local vaccination and abnormalities on MRimaging. Recognizing and correctly interpreting the imagingfindings related to post-vaccination complications by radiolo-gists are essential to efficient clinical management and care ofthe patient.

This paper presents advanced imaging results of four caseswith shoulder complaints after flu vaccine injection.Following IRB approval, the Picture Archival System andelectronic medical records were reviewed for pertinent radio-graphic and MR imaging data, clinical presentation and out-come, and laboratory summary. The following informationwas retrieved:

Cases

Case #1

A 66-year-old female with no past medical history was eval-uated for left arm pain exacerbated by motion that beganimmediately after receiving the flu vaccination. Symptomsincluded cool, numb, heavy sensation in her left hand, withradiation down the arm, neck, and scapular region. Despiteintense clinical presentation, physical examination wasnormal.

G. OkurEtimesgut Military Hospital, Department of Radiology, Ankara,Turkey

K. A. ChaneyPresence St. Joseph Hospital, Department of Radiology,77 North Airlite St., Elgin, IL 60123, USA

L. M. Lomasney (*)Department of Radiology, Loyola University Medical Center,2160 S. First Ave., Maywood, IL 60153, USAe-mail: llomasn@lumc.edu

Skeletal RadiolDOI 10.1007/s00256-014-1875-9

ESR was mildly increased at 43 mm/hr (range of normal,0-30 mm/hr). BMI was recorded as 25.7 kg/m2 (range ofnormal, 18.5 and 24.9 kg/m2) and CBC was normal.

Nine days after presentation, left upper extremity venousduplex ultrasound and MRI cervical spine excluded venousthrombosis and radiculopathy. Degenerative changes werenoted on cervical spine MR images.

Figure 1 demonstrates fluid within subacromial/subdeltoidbursa and physiologic glenohumeral joint fluid. Bone marrowsignal onMRI was normal. Incidentally noted were low-gradeundersurface tears in the posterior margin supraspinatus andanterior margin infraspinatus tendons. The measurement ofsubcutaneous fat thickness was noted to be 6.6 mm on obliquecoronal T1-weighted MR images.

No medical or surgical intervention was initiated. Afollow-up exam revealed a normal musculoskeletal exam.The patient had another normal musculoskeletal exam2.5 years after the event.

Case #2

A 59-year-old female was examined for complaints of wors-ening left upper arm and neck pain 1 week following influenzavaccination. On physical examination, there was tendernessand a palpable 3 cm×3-cm non-fluctuant soft tissue mass atthe injection site within the left upper extremity. Otherwise,the musculoskeletal exam was non-contributory. CBC andESR were normal on laboratory assessment. BMI was record-ed as 37.2 kg/m2. MRI of the shoulder was performed 20 daysafter vaccine injection.

Figure 2 demonstrates minimal increased T2-weighted sig-nal in left deltoid muscle and subcutaneous fat tissue, whichmight suggest focal inflammatory changes. There was noevidence of a dominant mass or abscess/hematoma on MRimages. Neither bone marrow signal changes nor bursitis werefound. The subcutaneous fat thickness was 2.1 cm measuredon the oblique coronal T1-weighted MR images.

The patient was treated conservatively with non-steroidalanti-inflammatory drugs and demonstrated complete resolu-tion of symptoms at 33 days post-vaccination.

Case #3

A 39-year-old male presented with prolonged, increasing leftshoulder pain for 2 months following flu vaccination. Physicalexamination was normal. On laboratory assessment, leukope-nia was noted on CBC (3–4,700 on serial tests for 4 monthswith normal range 4–10,000 per microliter) without definedsource, although counts were otherwise normal. ESR wasnormal. BMI was recorded as 25.7 kg/m2.

Figure 3 demonstrates findings on MRI of the shoulder,performed at 3 months following vaccination. Focal subcorti-cal bone marrow edema-like signal within the greater tuber-osity was observed, without associated obvious cortical de-struction. There were periosseous soft tissue inflammatorychanges and small subacromial/subdeltoid bursal fluid. Thesubcutaneous fat thickness was measured to be 4.7 mm onoblique coronal T1-weighted sequences.

The patient was treated conservatively with non-steroidalanti-inflammatory agents for 2 weeks and had complete reso-lution of symptoms after 5.5 months of observation.

a b

Fig. 1 Case #1, a 66-year-old female with regional pain and localsymptoms immediately following influenza vaccination underwent MRimaging 10 days post-vaccination. a Oblique coronal T2-weighted fat-saturated image of the subacromial/subdeltoid bursal fluid (open arrows)without overlying inflammatory changes [TR/TE: 3,800/60, field of view

14, 4-mm slice thickness with 1-mm inter-slice gap]. b Oblique coronalPD image shows equivocal undersurface partial thickness tear of thesupraspinatus tendon (arrow) [TR/TE: 4,000/40, field of view 14, 3-mm slice thickness with 1-mm inter-slice gap]

Skeletal Radiol

Case #4

A 36-year-old male presented with complaints of severepain localized to the injection site following flu

vaccination. There was a subjective decrease in rangeof motion due to pain. Physical examination was other-wise normal. A normal CBC with ESR was noted. BMIwas recorded as 22 kg/m2.

a b

Fig. 2 Case #2, a 59-year-old female presented for local symptoms1 week following influenza vaccination and underwent MR imaging20 days post-vaccination. a, b Oblique coronal T2 fat-saturated MRimage [TR/TE: 4,000/60, field of view 14, 4-mm slice thickness with 1-mm inter-slice gap] and axial T2 fat-saturated image [TR/TE: 4,000/60,

field of view 14, 4-mm slice thickness with 1-mm inter-slice gap] respec-tively show minimal infiltrative edema within the subcutaneous fat(arrowheads) and striations within the deltoid muscle (arrow). VitaminE capsule marks the patient’s pain and palpable complaint in b

a b

c d

Fig. 3 Case #3, a 39-year-oldmale with symptoms immediatelyafter flu vaccination underwentMR imaging 24 days post-vaccination. a, b Oblique coronalfat-saturated T2-weighted MRimage [TR/TE: 3,600/60, field ofview 14, 4-mm slice thicknesswith 1-mm inter-slice gap] andoblique sagittal fat-saturated T2-weighted image [TR/TE: 3,600/60, field of view 14, 4-mm slicethickness with 1-mm inter-slicegap], respectively, show bonemarrow edema (arrows) in thegreater tuberosity. HyperintenseT2 signal along the periosteum inFig. 3a likely indicates periostealreaction. Mild periosseous softtissue inflammatory changes areseen (large arrowhead). cObliquecoronal fat-saturated T2-weightedimaging [TR/TE: 3,600/60, fieldof view 14, 4-mm slice thicknesswith 1-mm inter-slice gap] showsa small amount of subacromial-subdeltoid bursal fluid (openarrow). d Sagittal T1-weightedimaging [TR/TE: 507/12, field ofview 14, 4-mm slice thicknesswith 1-mm inter-slice gap] showshypointense bone marrow signal(arrowhead)

Skeletal Radiol

Findings on the shoulderMR exam completed 1 week aftervaccination are shown in Fig. 4. Subcortical bone marrowedema involving the greater tuberosity (without associatedcortical destruction), periosseous inflammatory changes, andsmall fluid within the subdeltoid bursa are present. No dom-inant mass lesion, hematoma, or abscess was present.Subcutaneous fat thickness at the proposed site of injectionwas measured to be 4.2 mm on oblique coronal T1-weightedsequence.

Conservative treatment with short-term non-steroidal anti-inflammatory agents was initiated with complete resolution ofsymptoms within 2 months, and normal clinical exam at 2-year follow-up.

The MR findings of all four patients are shown in Table 1.Small fluid within subacromial/subdeltoid bursa was found inthree of the four patients. Intrasubstance edema-like signal indeltoid muscle was also a common finding on MR images.Two of the four patients demonstrated focal bone marrowedema-like signal changes within the humeral head.Important clinical data obtained by the electronic charts aresummarized in Table 2, including BMI, onset of pain, and

physical exam findings. Table 3 shows clinical managementand time to clinical resolution. Similarities between presenta-tions of the four patients in Tables 2 and 3 include a non-contributory past medical history, no systemic symptoms orinflammatory changes on physical exam. BMI was low inthree of the four cases presentations. Laboratory testing didnot support any infectious etiology. None of the four patientswere actively anti-coagulated.

Discussion

The Injection-Related Work Group of the U.S. Department ofHealth and Human Services Health Resources and ServicesAdministration Centers for Disease Control published the2011 Institute of Medicine Report, which generated“Proposals for Updates to the Vaccine Injury Table” [1].According to this report, “Shoulder injury related to vaccina-tion administration” is applied if the recipient had shoulderpain with limited range of motion within 48 h after vaccineand had no prior history of pain, inflammation, or dysfunction

a b

c d

Fig. 4 Case #4, a 36-year-oldmale with local symptoms atvaccination site within hours ofinjection underwent MR 20 daysafter symptom onset. a, bObliquecoronal fat-saturated T2-weightedimage (a) [TR/TE: 3,800/62, fieldof view 14, 4-mm slice thicknesswith 1-mm inter-slice gap] andoblique sagittal fat-saturated T2-weighted image (b) [TR/TE:3,600/62, field of view 14, 4-mmslice thickness with 1-mm inter-slice gap] show subcortical bonemarrow edema at the greatertuberosity without corticaldestruction (solid arrows). Smallfluid was present at thesubdeltoid/subacromial bursa(open arrow), as well as mildperiosteal inflammatory changes(arrowheads) suggestingperiosteal reaction. c, dHypointense signal changes inhumeral head (arrows) on T1-weighted images [TR/TE: 512/12, field of view 14, 4-mm slicethickness with 1-mm inter-slicegap]

Skeletal Radiol

of the affected shoulder prior to vaccine administration. Acausal relationship between vaccine administration and del-toid bursitis has been formally established. It has been report-ed that unintentional injection of vaccine antigen intosubacromial/subdeltoid bursa can trigger an inflammatoryreaction. “Shoulder injury related to vaccination administra-tion” has been detailed in Atanasoff et al.’s paper recountingpersistent and severe shoulder pain in a series of 13 patients,which is the only series related to shoulder injury followingvaccination in medical literature [2]. Asmentioned in the 2011Institute of Medicine Report, Atanasoff et al. implied that“shoulder injury related to vaccination administration” maynot only relate to deltoid bursitis but also may include tendon-itis, rotator cuff tear, frozen shoulder, impingement syndrome,adhesive capsulitis, and shoulder bursitis. Localized intra-substance edema-like signal in deltoid muscle is the mostcommon finding observed in Atanasoff et al.’s study.

There is little literature detailing imaging findings of com-plications following vaccination in general. It is not unexpect-ed that there is no available information to determine if there isa difference between flu vaccination and other types of vacci-nation. For the purposes of this report, post-vaccination imag-ing findings will be considered similar, regardless of entity.

Considering bone marrow changes, focal T2-weightededema-like signal changes in the humeral head with probableperiosteal reaction following flu vaccine injection have beenpreviously described in some reports [3, 4]. Two of ourpatients demonstrated humeral subcortical bone marrow ab-normal signal at the probable injection site as well as perios-teal hyperintense T2 signal changes. Osteonecrosis has been

proposed as a possible source of the bone marrow edemafollowing vaccination by some authors [5, 6]. Kuether et al.reported a case where incapacitating-prolonged shoulder dys-function after vaccination was correlatedwithMRI findings offocal bone marrow edema and incongruity of the humeralhead indicating atraumatic humeral head osteonecrosis [5].They hypothesized that the osteonecrosis might have beentriggered by a focal vasculitis, which may have led to bonedestruction and emphasized that mild forms of osteonecrosisfollowing flu vaccine injection may have been under reportedand they described “focal osteonecrosis” to be a possible sideeffect of adjuvant vaccines. Similar to Kuether’s report,Messerschmitt et al. presented a progressive osteolysis andsurface chondrolysis case following influenza vaccine injec-tion [6]. In these two reports, a casual relationship betweenvaccine injection and osteonecrosis has been theorized, basedon focal inflammatory reactions. In cases 3 and 4 presentedhere, outcomes and management do not support osteonecrosisor chondrolysis as the source of bone marrow edema-likesignal.

More likely, the bone marrow changes in our cases arerelated to injection technique combined with body habitus.Reference markers for shoulder anatomy have been previous-ly described for a proper vaccination technique in relevantliterature [7, 8]. Due to variable body habitus and muscle size,the optimum point for vaccine injection may vary amongindividuals. An improper angle of approach of the needle orincreased needle length may result in bursal or cortical pene-tration, particularly in those with low BMI. Several articlessuggested that an ideal needle length may be determined

Table 1 Summary of MRI findings

Caseno.

Time of MRI followingvaccination (days)

Bursal fluid Rotator cuff tear Soft tissue signal Bone marrow signal Subcut. fatthickness (mm)

#1 10 Present SSP/ISP Normal Normal 6.6

#2 20 Absent Absent Subcutaneous edema Normal 21.3

#3 24 Present Absent Intrasubstance deltoid edema Subcortical high T2 signal 4.7

#4 20 Absent Absent Intrasubstance deltoid edema Subcortical high T2 signal 4.2

Table 2 Summary of clinical data

Case no. Age and gender BMI kg/m2 Onset after injection First visit aftervaccination

Pain in deltoid area ROM WBC/ESR

#1 66 years, F 25.7 3 h 1 day + Subjective decrease Normal/43 mm/h

#2 59 years, F 37.2 2 days 1 week + Full Normal/normal

#3 39 years, M 22.5 Immediately 2 months + Full 3,000/μL/normal

#4 36 years, M 22.73 6-8 h 1 week + Subjective decrease Normal/normal

BMI body mass index (nl 18.5–24.9 kg/m2 ), ROM range of motion,WBC white blood cell count (nl 4–10 k/μl), ESR erythrocyte sedimentation rate (nl0–30 mm/h)

Skeletal Radiol

depending on gender, BMI, and deltoid subcutaneous fatthickness to specifically avoid these events [9, 10]. In orderto give an inactivated influenza vaccine intramuscularly toadults, a 1- to 1.5-inch needle should be used, while someexperts recommend a 5/8-inch needle for adults who weighless than 60 kg. For vaccination with the intradermal vaccine,the specifically designed microinjector has a 3/50-inch needle(Centers for Disease Control and Prevention). Two casespresented showed bone marrow edema-like signal and alsohad a low BMI and shallow subcutaneous deltoid fat.Assuming the standard institutional 1.5” needle was employedfor these patients, improper technique may easily have result-ed in cortical penetration.

Additional soft tissue changes in the cases presented mayalso be partly related to technique, with accidental access tomarginal spaces such as the bursa. Bodor et al. presented twocases of shoulder dysfunction following influenza and pneu-mococcal vaccine injections. Ultrasound evaluation in thosecases suggested improper vaccination leading to direct injec-tion of the bursa causing bursitis, tendinitis, and adhesivecapsulitis [11]. The inflammatory responses may be triggeredby the antigenic or non-antigenic components of the vaccineand are not only specific to flu-vaccine. Three of our fourcases had increased bursal fluid supporting improper needlelength and orientation. Although coincidental bursal inflam-mation is possible, none had prior shoulder symptoms, diag-nosis of bursal or rotator cuff pathology, or trauma.

Additionally, chemical inflammation/reaction should beconsidered as etiologies for soft tissue changes. A search upto November 2013 of Vaccine Adverse Event ReportingSystem data, a national vaccine safety surveillance programco-sponsored by the Centers for Disease Control andPrevention and the Food and Drug Administration, specificallyfor “injection site injury” (joint discomfort, injection site jointeffusion, injection site joint inflammation, injection site jointmovement impairment, injection site joint pain)” followinginfluenza vaccination revealed 72 entries in the U.S. and itsterritories. There was no entry with “bursal fluid accumulation”but 172 entries with “bursitis” , four entries with“osteonecrosis”, and 98 entries with “shoulder musculoskeletaldiscomfort” [database accessed at http://wonder.cdc.gov/vaers.html on Jan 12, 2014]. Injection-site complications may be

accepted to be relatively small numbers comparing with all84,368 entries related to influenza vaccination. According toThe Food and Drug Administration (FDA), different pharma-ceutical companies in theU.S sell 16 different influenza vaccineproducts in intramuscular, intradermal, and nasal spray forms.These vaccine products contain different chemical substancescalled “adjuvants” (to enhance the immune response), “preser-vatives”, and “tissue fixatives” (to preserve and fixate thebiological parts of the vaccine from any chemical reactionsand decomposition in the multi-vial forms). The adjuvants arehighly diverse compounds that could presumably cause adverseimmunologic effects [12]. These diverse, heterogeneous com-pounds may be an explanation for prolonged focal inflamma-tory reactions following vaccination, especially when injectedinto alternative sites. These processes may account for thebursal (cases 1, 3, and 4) and soft tissue signal changes (cases2, 3, and 4), as there were no other definitive explanations suchas evolving contusion, hematoma, or abscess.

It is clear that there is a lack of relevant literature reportingMR imaging findings of vaccination complication. In thisreport, we presented four cases with abnormal soft tissueand bone findings on MR related to recent influenza vaccina-tion. Althoughmost complications from influenza vaccinationare self-limited and do not warrant imaging, severe signs orsymptoms, local or systemic, may necessitate imaging. As aretrospective review without patient or vaccination standard-ization, the bone and soft tissue changes are presumed to bethe consequence of technique (needle length and orientation)and/or inflammation due to chemical constituents. With anincreasing percentage of the population undergoing influenzavaccination, secondary imaging may also become more fre-quent to assess unexpected outcome, warranting this reviewfor imagers.

Conflict of interest None.

References

1. Ryan T. 2011 Institute of Medicine (IOM) Report generatedProposals for Updates to the Vaccine Injury Table (VIT).

Table 3 Clinical outcomes

Case no. Management Differential diagnosis Coincident diagnosis Complete clinicalresolution

#1 NSAIDs and physical therapy Brachial plexus neuropathy,radiculopathy

Partial colectomy and OA 30 days

#2 NSAIDs Vaccine reaction Colon polyp and Achilles tendinitis 33 days

#3 NSAIDs Vaccine reaction Atrial fibrillation and malaria (past) 6 months

#4 NSAIDs Vaccine reaction None 2 months

NSAIDs non-steroidal anti-inflammatory agents

Skeletal Radiol

Department of Health and Human Services Health Resourcesand Services Administration Centers for Disease Control.2011.

2. Atanasoff S, Ryan T, Lightfoot R, Johann-Liang R. Shoulder injuryrelated to vaccine administration (SIRVA). Vaccine. 2010;28(51):8049–52.

3. Barnes MG, Ledford C, Hogan K. A "needling" problem: shoulderinjury related to vaccine administration. J Am Board Fam MedJABFM. 2012;25(6):919–22.

4. Schafer B, Burrough K. Shoulder pain in a 25-year-old femalefollowing an influenza vaccination. American Medical Society forSports Medicine. 2010.

5. Kuether G, Dietrich B, Smith T, Peter C, Gruessner S. Atraumaticosteonecrosis of the humeral head after influenza A-(H1N1) v-2009vaccination. Vaccine. 2011;29(40):6830–3.

6. Messerschmitt PJ, Abdul-Karim FW, Iannotti JP, Gobezie RG.Progressive osteolysis and surface chondrolysis of the proximal

humerus following influenza vaccination. Orthopedics. 2012;35(2):e283–6.

7. Beals TC, Harryman 2nd DT, Lazarus MD. Useful boundaries of thesubacromial bursa. Arthroscopy. 1998;14(5):465–70.

8. Matthews LS, Fadale PD. Subacromial anatomy for the arthroscopist.Arthroscopy. 1989;5(1):36–40.

9. PolandGA,BorrudA, JacobsonRM,McDermottK,Wollan PC,BrakkeD, et al. Determination of deltoid fat pad thickness. Implications forneedle length in adult immunization. JAMA. 1997;277(21):1709–11.

10. Koster MP, Stellato N, Kohn N, Rubin LG. Needle length for immu-nization of early adolescents as determined by ultrasound. Pediatrics.2009;124(2):667–72.

11. Bodor M, Montalvo E. Vaccination-related shoulder dysfunction.Vaccine. 2007;25(4):585–7.

12. About the Vaccine Adverse Event Reporting System (VAERS) atCDCWonder. 25 Sept 2013. <http://wonder.cdc.gov/vaers.html> Jan12, 2014 3:25:36 PM.

Skeletal Radiol