Imaging of the SeronegativeSpondyloarthopathies

Kimberly K. Amrami, MD

KEYWORDS

� Seronegative spondyloarthropathy � Sacroiliitis � Magnetic resonance imaging� Inflammatory arthritis � Ankylosing spondylitis

KEY POINTS

� The seronegative spondyloarthropathies can be categorized based on imaging findings in associa-tion with clinical features and laboratory testing.

� Multiple modalities (radiography, computed tomography [CT] andmagnetic resonance [MR] imaging)can be used to assess the axial and appendicular skeleton in patients suspected of seronegativespondyloarthopathies.

� MR imaging is the optimal modality for imaging the seronegative spondyloarthropathies, withimproved sensitivity compared with radiography and CT.

� Contrast-enhanced MR imaging can distinguish between active and inactive disease and alsoassess response to treatment.

.com

INTRODUCTION

Spondyloarthritis refers to a diverse group ofdiseases involving inflammation of the axial skel-eton and peripheral joints.1–3 The individual enti-ties are distinguished by specific clinical andlaboratory features with disease presentationoften on a spectrum that is dynamic and progres-sive rather than static and unchanging. Thesediseases can be grouped based on common clin-ical and imaging features such as inflammatoryback pain, sacroiliitis, spondylitis, and enthesitis.Laboratory studies, with the exception of thestrong association with the genetically determinedhuman leukocyte antigen B27 (HLA-B27), aregenerally nonspecific, with elevated inflammatorymarkers such as C-reactive protein and erythro-cyte sedimentation rate sometimes present. Clin-ical features may allow some differentiation (suchas urethritis in Reiter syndrome or reactive spon-dyloarthropathy), but there remains significantoverlap.1,3

The original concept of a group of interrelatedbut distinctive disorders was developed by Moll

Department of Radiology, Mayo Clinic, 200 First Street SE-mail address: amrami.kimberly@mayo.edu

Radiol Clin N Am 50 (2012) 841–854doi:10.1016/j.rcl.2012.04.0100033-8389/12/$ – see front matter � 2012 Elsevier Inc. All

s

and colleagues4 in 1974 to describe a group ofinflammatory diseases affecting the spine andsacroiliac joints. The term seronegative spondy-loarthropathies was coined to indicate that rheu-matoid factor was not present in these patients,with the individual forms of the disease includingankylosing spondylitis, psoriatic arthritis, reactivearthritis (formerly known as Reiter syndrome),arthritis related to inflammatory bowel disease,and a form of juvenile idiopathic arthritis dis-tinguished only by the age of the patient.5 Undiffer-entiated spondyloarthropathy and late-onsetspondyloarthropathy are also sometimes includedin this grouping.6,7 In addition to the distinction bythe absence of rheumatoid factor, the seronega-tive spondyloarthropathies uniquely affect enthe-ses. The definition and subcategorization of thespondyloarthropathies has evolved over time,and multiple groups have attempted to charac-terize the symptoms and natural history of thespondyloarthropathies, including the New Yorkcriteria8 for sacroiliitis and similar criteria for anky-losing spondylitis, in the 1960s and 1970s.1,3 In the1990s there was a move to reclassify the entire

W, Rochester, MN 55905, USA

rights reserved. radiologic.th

eclinic

Amrami842

disease spectrum; this grouped all patients withinflammatory arthritis involving the axial skeletalas seronegative spondyloarthritis, despite a widevariety of clinical symptoms.9 The Amor crite-ria addressed the difficulty in diagnosing thesedisorders through the creation of a scoringsystem.9 A newer (2009) classification wasproposed by the Assessment of SpondyloarthritisInternational Society (ASAS) after a large cross-sectional study.10 Rather than focusing on specificsubtypes such as ankylosing spondylitis, this clas-sification depends on 2 important clinical features:axial symptoms and peripheral involvement. Theinvestigators proposed the term “axial spondy-loarthritis” for the entire spectrum of diseaseswhereby axial involvement predominates. Thistype can then be broken down into the more tradi-tional subtypes based on clinical features, HLA-B27 positivity, and the presence or absence ofsacroiliac involvement based on the detection ofactive inflammation by advanced imaging tech-niques such as magnetic resonance (MR) imaging.The ASAS criteria for axial spondyloarthritisconsists of active sacroiliitis on imaging plus 1 ormore features of spondyloarthritis or HLA-B27positivity with 2 or more features of spondyloarthri-tis. In comparison, the criteria for peripheral spon-dyloarthritis are more complex and include suchoptions as Crohn disease or ulcerative colitis, priorinfection (as in Reiter syndrome), inflammatoryback pain, or positive family history.9 The diag-nosis of a peripheral spondyloarthritis also candepend on the presence of sacroiliitis on imaging(ie, arthritis with sacroiliitis alone meets thesecriteria).9 Although classification remains of in-terest for these complex disorders, the main chal-lenge at the current time is the development ofstrategies for early diagnosis and treatment aimedat limiting disability and disease progression overtime. The newer classifications all depend onadvanced imaging techniques such as MR im-aging, supplanting the prior use of radiography(and radiographic atlases) in assisting clinical deci-sion making.

IMAGING TECHNIQUES

Although the spondyloarthropathies can involvethe entire axial and appendicular skeleton, in-cluding central and peripheral entheses and joints,the hallmark for all types of spondyloarthritisremains sacroiliitis.1–3 Inflammation of one orboth sacroiliac joints is the most characteristicand consistent feature of these disorders. Involve-ment of the remainder of the axial skeleton is rarein the absence of sacroiliitis, as is peripheralinvolvement even when those symptoms, such

as enthesitis at the heel, dominate the clinicalpicture. Inflammatory back pain is commonlyassociated with sacroiliitis, but is a nonspecificsymptom and may be seen in other disordersunrelated to spondyloarthropathy. HLA-B27 posi-tivity on its own is not an indication of spondyloar-thritis in the absence of symptoms or positiveimaging findings.10,11 The unifying diagnostic toolfor the seronegative spondyloarthropathies isimaging of the sacroiliac joints.Conventional radiography remains the most

common initial imaging study for patients sus-pected of having inflammatory arthritis of all kinds(Fig. 1).3 Plain radiography of the pelvis for assess-ment of the sacroiliac joints has significant limita-tions, including the need for ionizing radiation inyoung patients as well as low sensitivity for detec-tion of early disease.2,3 Experience and knowledgeof the clinical context may improve detection, butit is common for radiologists to miss advancedcases of sacroiliitis on radiography.12 Five stagesof radiographic changes in the sacroiliac jointshave been described, ranging from 0 (normal)through 4.8 Themost difficult stages are 1 (unclear)and 2 (small erosions, sclerosis), with moreadvanced disease (3, definite erosions and 4,ankylosis) less problematic to detect, testifying tothe poor specificity and moderate sensitivity ofradiography in detecting sacroiliitis. The relativelylow utility of radiography for sacroiliitis is exacer-bated by poor interobserver and intraobserver reli-ability for subtle changes in early disease.13–15 Thedetection of the structural changes of sacroiliitis ischallenging enough, but the physiologic param-eter of disease activity is beyond the capabilitiesof radiography, which precludes using radiog-raphy to monitor response to therapy.The complexity of the sacroiliac joints them-

selves and the difficulty in seeing the entire jointin a 2-dimensional projection is part of the chal-lenge, but studies comparing more specializedviews such as the angled anteroposterior obliqueFerguson view have not shown significant im-provement in accuracy.1,3,16 Conventional viewsof the pelvis may actually have some added valueover dedicated imaging of the sacroiliac joints, inthat the hips are usually included in such images.17

Despite these limitations, conventional radio-graphs are important, especially in distinguishingankylosing spondylitis from other types of spondy-loarthritis. Radiographic imaging of the entirespine and symptomatic individual peripheral jointscan help to classify the various types of spondy-loarthropathy and to visualize complications suchas discovertebral fractures in ankylosing spondy-litis, although with lower sensitivity than withcomputed tomography (CT).

Fig. 2. Reconstructed oblique coronal CT imageshowing normal sacroiliac joints.

Fig. 1. Normal radiographs of the sacroiliac joints. (A) Normal anteroposterior (AP) radiograph of the sacroiliacjoints. (B) Normal oblique radiograph of the left sacroiliac joint. Note resolution of some of the joint overlap seenin A.

Imaging of Seronegative Spondyloarthopathies 843

Scintigraphy is another technique that has beentraditionally used to evaluate the skeleton whenspondyloarthritis is suspected. In the past, whenradiography was the primary diagnostic tool, scin-tigraphy enhanced the sensitivity of plain radio-graphs for early-stage disease. Unfortunately,though sensitive, scintigraphy is too nonspecificto be used in isolation in the diagnosis of sacroilii-tis.18,19 In fact, mild activity at the sacroiliac jointscan lead to overdiagnosis of sacroiliitis when onlymild degenerative change is present.20 Scintig-raphy may be helpful in establishing bilateraldisease whereby only unilateral disease has beenseen on radiography, but it is inferior to moreadvanced cross-sectional imaging such as CTand especially MR imaging. The use of colloidalagents for joint imaging is similarly nonspecificwhen evaluating the spondyloarthropathies. Posi-tron emission scanning (PET), with or without CT,has little role in the assessment of spondyloarthr-opathy at this time.

CT can be useful in assessing the spine andsacroiliac joints, having a higher degree of sensi-tivity than radiography and with better specificitythan scintigraphy (Fig. 2).20 Because of the highspatial resolution possible with CT, subtle erosionsand subchondral sclerosis in the sacroiliac jointsmay be seen to better effect than with radiog-raphy; early syndesmophytes and “shiny corners”can be better seen in the spine. Unlike radiographythe images may be obtained in any plane, so thatvisualization of the entire joint and individual discspaces is optimized regardless of orientation. Infact, most modern CT acquisitions are essentially

volumetric because of their isotropic nature,making multiplanar and even 3-dimensional refor-matting easy to perform on a routine basis. CT ispreferred for the detection of very early erosionsof the sacroiliac joints and for early ankylosis.21–23

CT requires the use of ionizing radiation in what areusually young patients; it is not an ideal method forfollowing patients over time. Furthermore, activeinflammation can be difficult to assess, as bonemarrow edema is not generally visible on CT;burnt-out sclerosis without active disease mayhave a similar appearance to very active inflamma-tion.22,23 A final caveat is the normal progression ofage-related subchondral sclerosis at the sacroiliac

Amrami844

joints. Care must be taken not to confuse normalsenescent changes or degenerative osteoarthritiswith the erosions, irregularity, and subchondralsclerosis seen in sacroiliitis.21

Ultrasonography has some utility for the evalua-tion of sacroiliitis when it is very active, by usingDoppler ultrasonography to assess blood flowand synovitis.3,24 It may also be useful in somecases in young children as an initial study, but islimited to the evaluation of soft tissues surroundingthe joint and not the joint itself.17 Ultrasound guid-ance may be used for diagnostic and therapeuticinjections into the sacroiliac joint as an alternativeto fluoroscopy in some cases.25

MR imaging has become the gold standard forthe imaging diagnosis of spondyloarthropathies ofthe spine and sacroiliac joints (Fig. 3).1–3,13,21,26–30

MR imaging is highly sensitive and specific for thepresence of inflammatory changes in and aroundthe spine and sacroiliac joints.31 As with CT, multi-planar acquisitions are routinely performed, whichoptimize visualization of the entire sacroiliac jointwith the added benefit of improved soft-tissueand bone marrow contrast. Serial studies overtime are not problematic in terms of radiation expo-sure. Subtle erosions may be difficult to seebecause of the relatively lower spatial resolutionof MR imaging compared with CT, but T2-weighted sequences with fat suppression areexquisitely sensitive and specific in the detectionof bone marrow edema, joint widening, and jointfluid and, thus, synovitis.22,23 With contrastenhancement, MR imaging can distinguish activefrom inactive disease and can be used to monitortreatment response: a decrease in enhancement

Fig. 3. Normal MR images of the sacroiliac joints. (A) Obljoints with fat suppression showing no abnormal signal aimage with fat suppression after contrast administration

even in the presence of persistent bone marrowedema has been strongly correlated with clinicalresponse to treatment.32 Alternatively, a lack ofresponse by imaging with no change in baselineenhancement suggests treatment failure and aneed to alter the therapeutic regimen. Postcontrastimaging can be performed in a variety of ways usingdelayed enhancement, rapid imaging with dynamicenhancement, and with or without the use of imagesubtraction, which can be especially helpful whenlow-field (<1.5 T) imagers are used.33 Enhancementcan be measured in semiquantitative ways in addi-tion to the qualitative visual assessment duringimage interpretation. All of these techniques canbe very helpful when determining the value ofdrug regimens with significant side effects andhigh cost.33

MR imaging is also the preferred modality forimaging the remainder of the spine and for theperipheral joints (Fig. 4). The soft-tissue contrastis ideal for assessing the structureswithin the spinalcanal (including the cord) and for assessing inflam-matory or destructive changes in the joints. Duralectasia and cauda equina syndrome seen in anky-losing spondylitis (AS) can be best evaluated withMR imaging.34 Direct visualization of the atlantoax-ial interval in the cervical spine can also bewell seenon MR imaging; atlantoaxial subluxation can beevaluated in some upright MR imaging scanners,but is more typically assessed with radiographywith flexion and extension.6 CT may be used asan alternative to MR imaging when patients areunable to undergo MR imaging for any reason,andwhen rapid assessment is desired as in assess-ing discovertebral fractures after trauma.28

ique coronal T2-weighted MR image of the sacroiliacround the joint. (B) Coronal oblique T1-weighted MRshows normal sacroiliac joints with no enhancement.

Fig. 4. Sacroiliitis. (A) AP radiograph of the pelvis showing bilateral sacroiliitis in a patient with inflammatorybowel disease. Note right lower quadrant ostomy and residual oral contrast. (B) Coronal short-tau inversionrecovery (STIR) image of the right sacroiliac joint showing T2 hyperintensity on both sides of the joint consistentwith sacroiliitis. Only the right side was imaged, owing to body habitus. This patient has undifferentiated spon-dyloarthropathy and is HLA-B27 positive. (C) Axial CT image showing irregularity and subchondral sclerosis at thesacroiliac joints bilaterally in a patient with ankylosing spondylitis.

Imaging of Seronegative Spondyloarthopathies 845

IMAGING FINDINGS IN THE SERONEGATIVESPONDYLOARTHROPATHIES

The seronegative arthropathies are a group ofdiverse disorders, which share some commonfeatures (such as an increased incidence of ge-netically determined HLA-B27 histocompatibilityantigen and inflammation in the musculoskeletalsystem) but which can also be subclassified intospecific entities based on imaging findings, clinicalpresentation, and natural history of the disease.Most of the entities described here have juvenileforms in addition to the usual presentation inyoung adulthood; rare patients present in old agewith late-onset spondyloarthropathy, which maybe difficult to characterize because of comorbiddegenerative arthritis.7 A small number of patientswho cannot be categorized into one of the

following subtypes are categorized as havingundifferentiated spondyloarthropathy.6

Ankylosing Spondylitis

AS is the prototypical seronegative spondyloarthr-opathy. AS is primarily a disease of white maleswith the male to female ratio being between 4:1and 10:1, with generally milder presentations inwomen. The normal prevalence of the HLA-B27antigen in the North American population is 6%to 8%, but for patients with in AS it is more than90%.1,2 The disease commonly presents in lateadolescence or early adulthood; presentation afterage 40 years is very unusual. The most commonpresenting features of AS are morning stiffnessand low back pain. Patients presenting at youngerages may experience peripheral symptoms such

Amrami846

as heel pain rather than the usual inflammatoryback pain. Because symptoms are nonspecific,the time to diagnosis from the first symptoms aver-ages 7 years. About 20% of patients progress tosevere debilitation primarily related to systemicmanifestations of AS, such as lung fibrosis.2,35

AS is primarily a disease of the axial skeleton thataffects the spine and sacroiliac joints. Sacroiliitis isa required element for diagnosis of AS.2,3,8–10 Thedisease typically begins in the sacroiliac jointswith small erosions resembling the serrated edgesof a postage stamp, typically beginning on the iliacside of the joint, due to the thinner cartilage therecompared with the sacrum. As the diseaseadvances proliferative changes associated withthe enthesitis dominate, with sclerosis seen in thesubchondral areas progressing finally to completeankylosis of the joint. When the fusion is completethe sclerosis resolves completely.2 The sacroiliitismay appear asymmetric in the early stages, espe-cially if MR imaging is used in early diagnosis, butthe disease inevitably progresses to bilateral,symmetric involvement.Spondylitis occurs in about 50% of AS patients,

with females relatively less affected.2,36 Thechanges in the spine occur first in thoracolumbarand lumbosacral regions with extension to themidlumbar, midthoracic, and cervical regions.Involvement of the cervical spine alone is veryrare. The orderly progression of spine involvementis a unique feature of AS when compared with theother spondyloarthropathies, in which spineinvolvement tends to be more random. In AS the

Fig. 5. Ankylosing spondylitis. (A) AP radiograph of the lumthe typical bamboo spine of AS. (B) Lateral view in a diffrepresenting ossification of Sharpey fibers. Note also theanterior aspect of the lumbar vertebrae.

earliest changes in the spine are due to enthesitisat the insertion of the outer fibers of the annulus fi-brosus on the ring apophysis of the vertebral end-plate. This process results in subtle erosions withreactive sclerosis in the vertebral corners, knownas shiny corners or Romanus lesions.2,3 Thesephenomena are generally very short lived, withprogression to the more commonly seen syndes-mophytes, representing the ossification of theouter fibers of the annulus fibrosus.2 These lesionsare very fine and symmetric, bridging individualvertebral bodies from corner to corner. Thissame process results in the “squaring” of thevertebral bodies as the fusion progresses, and isbest visualized in the lumbar spine where there isloss of the normal concave profile of the vertebrae.Other spinal elements are also ossified and fusedas part of this progressive process, including theapophyseal joints, paraspinous ligaments, andspinous processes. When complete the appear-ance is very characteristic, and is termed thebamboo spine (Fig. 5).2 The changes in the spineitself are very well visualized with conventionalradiography but can also be seen on CT. MRimaging obtained in the active phase of enthesitiswill show the inflammatory changes around thedisc and at the apophyseal joints of the spine.Pseudarthrosis in the spine may develop as a

complication of erosions or because of trauma, re-sulting in fracture through the syndesmophytes inthe fused spine.37 Rarely a pseudarthrosis maydevelop at a level of lesser disease involvementbetween long fused segments.2 Themost common

bar spine showing fusion of both sacroiliac joints anderent patient showing the fine, bridging osteophytesloss of the normal concave contour (squaring) of the

Imaging of Seronegative Spondyloarthopathies 847

site of traumatic lesions is at the thoracolumbar orcervicothoracic junctions, which commonly in-volve all 3 columns of the spine (Fig. 6). Theselesions may be easily missed at the time of initialevaluation of trauma and may progress to a truepseudarthrosis with instability at the fracture site,potentially leading to cord injury and paralysis. ASpatients suffering even modest trauma should beconsidered to harbor a spinal fracture until ex-cluded by advanced imaging, preferably CT.Similarly, atlantoaxial subluxation may be seen inAS, but it is uncommon and is more likely to bepresent when patients have peripheral joint in-volvement.38 Dural ectasia and leptomeningealsacculations are common and may result in caudaequine syndrome, best evaluated with MR imaging(Fig. 7).34,39 Large peripheral joints may beinvolved, but this is more likely when presentingsymptoms occur before age 21 years. Small jointinvolvement is rare, but when present can lead toankylosis and loss of mobility, especially in thehands.2,3

Themost commonextraskeletalmanifestation ofAS is progressive lung disease, including bullousemphysema and fibrosis, often complicated byunusual infections such as intracavitary aspergil-losis.2,3 Pulmonary changes occur late in thedisease but may be the source of significantmorbidity and mortality.2 Inflammatory changes inthe heart and aortamay result in conduction abnor-malities and aortic insufficiency.37 The associationof inflammatory bowel disease and AS is tentative,but both are associated with HLA-B27 positivity.11

Fig. 6. Ankylosing spondylitis. (A) Sagittal reformatted CTfracture (Andersson lesion) at C5-C6 in this patient follopatient showing typical changes of AS with fusion of the

Psoriatic Arthritis

Psoriatic arthritis (PA) is an asymmetric, polyartic-ular disorder included in the spondyloarthropa-thies, as up to 40% of patients with PA willdevelop spondylitis or sacroiliitis. PA affects onlyabout 7% of patients with cutaneous psoriasis.40

The arthritis may antedate the skin changes byseveral years; the severity of the arthritis seemsindependent of the severity of the skin disease.2,3

The presence of pitting nail changes correlateswith the arthritis, especially when distal inter-phalangeal (DIP) joint involvement is severe.2,41

Male/female involvement is nearly equal; thedisease has a later onset than other spondyloar-thropathies, with a typical onset between age 30and 50 years.2

Sacroiliac involvement is not universal in PA.Only 5% of patients have exclusive spinal involve-ment without sacroiliitis.2,3 The sacroiliac diseaseis generally asymmetric and ankylosis of thesacroiliac joints is very rare. MR imaging is mostsensitive for the detection of subtle bilateralchanges, which can be important in distinguishingPA from septic sacroiliitis in the early stages of thedisease (Fig. 8).13 The spondylitic changes in PA(and reactive spondyloarthritis) appear morerandomly than those seen in AS,2,3 and are usuallybut not always associated with sacroiliitis. Large,chunky-appearing paravertebral ossifications arecommonly seen in the thoracolumbar junction.2

These ossifications do not bridge the intervetebraldiscs as seen in AS but rather seem to attach

image of the cervical spine showing a discovertebralwing a face-first fall. (B) AP radiograph in the samesacroiliac joints and early changes of bamboo spine.

Fig. 7. Ankylosing spondylitis. Sagittal (A) and axial (B) T2-weighted MR images of the lumbar spine showingdural ectasia and sacculations in a patient with AS. This patient presented with cauda equina syndrome secondaryto the tethering of the spinal nerve roots seen on these images.

Amrami848

to the lateral aspect of the vertebral bodies(Fig. 9A). Ankylosis of the apophyseal joints,squaring of the vertebral bodies, and spinal fusionare very rare in PA.

Fig. 8. Psoriatic arthritis. (A) AP view of the spine in a patperipheral bridging osteophytes typical of psoriatic and rePA view of the hand in a different patient with psoriatic aative periostitis in the digits. This patient does not have c

More distinctive in PA are the changes in thehands and feet. In the hands erosive changesdevelop within the interphalangeal joints, withoutthe sparing seen in rheumatoid arthritis of the

ient with psoriatic arthritis showing the large, chunkyactive arthritis. Bilateral sacroiliitis is also present. (B)rthritis showing classic marginal erosions and prolifer-utaneous psoriasis.

Fig. 9. (A) Coronal oblique T2-weighted MR image without fat suppression showing bilateral sacroiliitis, moreprominent on the right. (B) Coronal oblique T1-weighted MR image with fat suppression after contrast showspatchy areas of enhancement consistent with active sacroiliitis in a patient with severe cutaneous psoriasis.

Imaging of Seronegative Spondyloarthopathies 849

DIP joints (Fig. 9B).2 The erosions are marginal andare often associated with marked soft-tissueswelling of the digits. Proliferative changesincluding fluffy periostitis are a common feature.Erosion of the distal portion of the phalanx with re-modeling of the joint, the pencil-in-cup deformity,may be seen, as well as resorption or sometimesproliferation at the ungual tufts. The appearanceis occasionally confused with erosive osteoar-thritis (OA) but the presence of erosions at themarginal or bare areas of the phalanges in PAshould be discernible from the central, subchon-dral erosions in erosive OA. Dense, circumferentialperiostitis may result in the so-called ivory phalanxon radiography. The interphalangeal joint of thegreat toe in the foot is the most common site for

Fig. 10. Reactive arthritis. (A) Lateral radiograph of theshowing enthesitis and periostitis at the plantar calcaneugraph showing bilateral sacroiliitis, which is worse on the rhad a history of Chlamydia infection.

PA and may be confused with gout. If this isa serious concern, dual-energy CT may be helpfulin this unique instance to determine whether theurate deposition of gout is present.42

Reactive Spondyloarthritis (Reiter Syndrome)

AswithPA, reactive spondyloarthritis (RS) is a poly-articular, asymmetric arthritis. Unlike PA, RS pref-erentially affects the foot and large, peripheraljoints.2 If the hands are involved the interphalan-geal joints are commonly involved, with relativesparing of themetacarpophalangeal andDIP joints.A common presenting symptom is heel paincaused by the fluffy enthesitis that can be seenwith heel spurs in patients with RS (Fig. 10A).43

hindfoot in a different patient with reactive arthritiss. This patient presented with heel pain. (B) AP radio-ight in a patient with urethritis and uveitis. The patient

Amrami850

Because of the historical association with venerealinfection, this has sometimes been called lover’sheel. Sacroiliitis and spondylitis are seen morecommonly than in PA, with sacroiliitis present inup to 45%of patients.2,3 The asymmetric and bilat-eral imaging appearance of sacroiliitis in RS is iden-tical to that seen in PA, with ankylosis even lesscommon than in PA (Fig. 10B). The radiographicappearance of the spinal changes is the same asin PA, with marginal, coarse syndesmophytesthat appear to be randomly placed in the lowerthoracic and lumbar spine.2 Spine involvement ismore common than in PA, seen in the lumbar spinein about 30% of patients with RS.3 Upper thoracicand cervical spine involvement is very rare. It maysometimes be impossible to differentiate PA andRS on radiographs, but the clinical syndromes areusually distinct. Relatively few findings in the upperextremities, sparing of the DIP joints, and involve-ment of the sacroiliac joints and spine may tilt thediagnosis toward RS.Classically RS has been diagnosed when the

triad of urethritis, uveitis, and arthritis have beenpresent, often in association with Chlamydia infec-tion.2,3 Patients with RS have a higher than averageprevalence of HLA-B27 positivity, which maypredispose them to develop the arthritis in thepresence of a triggering event such as infection.11

The term Reiter syndrome is only used when thespecific triggering organism can be identified andwhen the arthritis appears within about a monthof the original illness.1 In addition to the classicassociation with venereal disease, RS can beseen in enteric infections from organisms such asShigella, Salmonella, and Yersinia.2 Human immu-nodeficiency virus (HIV) infection can of coursecoexist with both RS and PA, and the arthritismay be particularly severe in such cases. HIV canitself be associated with joint symptoms butwithout the typical radiographic findings seen inRS, and generally withmore involvement of periph-eral joints than the axial skeleton. HIV infectionmayexacerbate existing rheumatoid arthritis or PA; ifsevere, unexplained worsening is seen clinically,HIV testing may be warranted.1

Fig. 11. Enteropathic arthritis. AP radiograph of thepelvis showing bilateral sacroiliitis and a right lowerquadrant ostomy in a patient with Crohn disease.

Arthritis Associated with Inflammatory BowelDisease (Enteropathic Arthritis)

Enteropathic arthritis may be seen equally in as-sociation with Crohn disease or ulcerativecolitis.2,3,44 The arthritis may have 1 of 2 formsthat rarely coexist. Up to 20% of patients withinflammatory bowel diseasemay develop a nonde-structive and often transient peripheral arthritis,the severity of which parallels the progress of theunderlying bowel disease and which may actually

resolve with surgery or other successful treat-ment.2 This primarily involves joints such as theknees, ankles, elbows, and wrists, and is mostoften bilateral and symmetric.The other form of enteropathic arthritis can

involve the spine and/or sacroiliac joints. An iso-lated spondylitis that is radiographically identicalto AS with the same fine, marginal syndesmo-phytes proceeding in orderly fashion from the thor-acolumbar junction may be present in up to 6% ofpatients.2,3,45 Shiny corners and squaring of thevertebral bodies may be present, although this isless common. Rarely this spondylitis may bepresent without sacroiliitis, an important distinctionfromAS. Sacroiliitis, when present, is similar to thatseen in AS, with bilateral, symmetric findings andlater ankylosis of the joints (Fig. 11).2 Isolated sac-roiliitis is more common than isolated spondylitis; itmay be seen in up to 18% of patients with inflam-matory bowel disease.2,3,13 In contradistinction tothe peripheral arthritis seen with inflammatorybowel disease, the severity of the spondylitis prog-resses independent of the course of disease ortreatment of the primary intestinal problem.2,3,46

The spondylitis may even precede the develop-ment of the bowel disease by months or years(Fig. 12). As with other forms of spondyloarthrop-athy, there is a predilection for HLA-B27 positivitybut without the high prevalence seen with AS.1,2,11

Undifferentiated Spondyloarthropathy

Undifferentiated spondyloarthropathy is a termused when clinical symptoms are present such

Fig. 12. Enteropathic arthritis. (A) AP radiograph showing asymmetric, bilateral sacroiliitis in a 21-year-old manwith inflammatory back pain. (B) Oblique coronal T2-weighted MR image with fat suppression showing bilateralsacroiliitis. (C) Oblique coronal T1-weighted MR image with fat suppression showing enhancement consistentwith active sacroiliitis. The patient is HLA-B27 negative. (D) Radiograph obtained 4 years after A shows worseningsacroiliitis. At this time the patient began experiencing bloody diarrhea and underwent endoscopy showinginflammatory bowel disease.

Imaging of Seronegative Spondyloarthopathies 851

as peripheral arthritis, sacroiliitis, and enthesitiswith inflammatory low back pain, but without dis-tinguishing clinical or imaging features that wouldallow further subclassification.6 Key features aresacroiliitis and increased HLA-B27 positivity, asin other forms of seronegative spondyloarthritis,but other stigmata such as inflammatory boweldisease or distinguishing patterns of peripheraljoint involvement or inciting infection are absent.Atlantoaxial subluxation as an unusual manifesta-tion of undifferentiated spondyloarthropathy wasrecently described in a case report in a youngpatient, reinforcing the need for a high degree ofclinical suspicion when young patients present

with inflammatory back or spinal pain (Fig. 13).Awareness of this syndrome may allow earlierdiagnosis and treatment in the absence of classicfindings of AS, PA, or RS.

Late-Onset Spondyloarthropathy

Although much more commonly seen in youngpatients, AS and other spondyloarthropathiesmay present in the elderly. The clinical presenta-tions are sometimes confusing, often combiningage-related degenerative osteoarthritis with im-aging features more typical for spondyloarthrop-athy such as sacroiliitis.7,31 HLA-B27 positivity is

Fig. 13. Undifferentiated spondyloarthropathy. (A) Axial CT scan obtained without contrast shows bilateralchanges of sacroiliitis in this 28-year-old man with inflammatory low back pain. He is HLA-B27 positive. (B) Ob-lique coronal T1-weighted MR image with fat suppression obtained after contrast administration. Note the floridenhancement on the right, representing active sacroiliitis. There were subtle positive findings present on the leftas well. A fluoroscopic image from an injection of the right sacroiliac joint showed sacroiliitis in the same patient(not shown). (C) Oblique coronal T1-weighted MR image with fat suppression obtained after 1 year of treatmentwith enteracept (Enbrel). The enhancement has resolved, compatible with successful treatment. The patient’ssymptoms improved but did not completely resolve.

Amrami852

elevated in this group, as in the syndromes pre-senting in younger patients. The presentingdisease is often more severe than in youngerpatients, and treatment options may be morelimited because of the toxicity associated withtreatments such as anti–tumor necrosis factoragents and comorbidities such as cardiovasculardisease. In some older patients the new onset ofsymptoms of inflammatory back pain may beconfused with diseases such polymyalgia rheuma-tica or even chronic pain syndromes such as fibro-myalgia.47 Awareness of this entity and the use ofadvanced cross-sectional imaging is increasing,

allowing more accurate diagnosis and treatmentin this unique group of patients.31

SUMMARY

The seronegative spondyloarthropathies are a di-verse group of disorders affecting the axial skeletonand peripheral joints. Based on clinical presentationand imaging findings the individual entities havesome unique features that can help to distinguishthem, but all share inflammatory changes in thesacroiliac joints and spine and may present withback pain. Plain radiography is the mainstay of

Imaging of Seronegative Spondyloarthopathies 853

imaging, but MR imaging has significantly bettersensitivity and specificity for detecting bonemarrowedema and enhancement indicating inflammation.MR imaging may also be used to monitor therapywith disease-modifying drugs. Early MR imagingcan help to establish the diagnosis in the earlystages of disease, avoiding unnecessary delaysand facilitating earlier treatment.

REFERENCES

1. Dougados M, Baeten D. Spondyloarthritis. Lancet

2011;377:2127–37.

2. El-Khoury GY, Kathol MH, Brandser EA. Seronega-

tive spondyloarthropathies. Radiol Clin North Am

1996;34(2):343.

3. Luong AA, Salonen DC. Imaging of the seronegative

spondyloarthropathies. Curr Rheumatol Rep 2000;2:

288–96.

4. Moll J, Haslock I, Macrae IF, et al. Associations

between ankylosing spondylitis, psoriatic arthritis, Re-

iter’s disease, the intestinal arthropathies, andBehcet’s

syndrome. Medicine (Baltimore) 1974;53:343–64.

5. Azouz E, Duffy C. Juvenile spondyloarthropathies:

clinical manifestations and medical imaging. Skel-

etal Radiol 1995;24(6):399–408.

6. Muscal E, Satyan K, Jea A. Atlantoaxial subluxation

as an early manifestation in an adolescent with undif-

ferentiated spondyloarthritis: a case report and review

of the literature. J Med Case Rep 2011;5(1):275.

7. Toussirot E. Late-onset ankylosing spondylitis and

spondyloarthritis: and update on clinical manifesta-

tions, differential diagnosis and pharmacological

therapies. Drugs Aging 2010;27(7):523–31.

8. Van Der Linden S, Valkenburg H, Cats A. Evaluation

of diagnostic criteria for ankylosing spondylitis. A

proposal for modification of the New York criteria.

Arthritis Rheum 1984;27:361–8.

9. Amor B, Dougados M, Mijiyawa M. Criteria of the

classification of spondyloarthropathies. Rev Rhum

Mal Osteoartic 1991;57:85–9.

10. RudwaleitM, VandderHeijdeD, LandeweR, et al. The

development of Assessment of SpondyloArthritis

international Society classification criteria for axial

spondyloarthritis (part II): validation and final selec-

tion. Ann Rheum Dis 2009;68:777–83.

11. Fong K. The genetics of spondyloarthropathies. Ann

Acad Med Singapore 2000;29(3):370–5.

12. Forrester D, Hollingsworth P, Dawkins RL. Difficulties

in the radiographic diagnosis of sacroiliitis. Clin

Rheum Dis 1983;9:323–32.

13. Braun J, Sieper J, Bollow M. Imaging of sacroiliitis.

Clin Rheumatol 2000;19:51–7.

14. Hollingsworth P, Cheah P, Dawkins RL, et al.

Observer variation in grading sacroiliac radiographs

in HLA-B27 positive individuals. J Rheumatol 1983;

10:247–54.

15. Yacizi H, Turunc M, Ozdo�gan H, et al. Observer vari-

ation in grading sacroiliac radiographs might be

a cause of “sacroiliitis” reported in certain disease

states. Ann Rheum Dis 1987;46:139–45.

16. Lawson T, Foley W, Carrera GF, et al. The sacroiliac

joints: anatomic, plain roentgenographic, and com-

puted tomographic analysis. J Comput Assist

Tomogr 1982;6:307–14.

17. Battison M, Manaster B, Reda DJ, et al. Radio-

graphic diagnosis of sacroiliitis: are sacroiliac views

really better? J Rheumatol 1998;25:2395–401.

18. Chase W, Houk R, Winn RE, et al. The clinical useful-

ness of radionuclide scintigraphy in suspected

sacro-iliitis: a prospective study. Br J Rheumatol

1983;22:67–72.

19. Ho GJ, Sadovnikoff N, Malhotra CM, et al. Quantita-

tive sacroiliac joint scintigraphy: a critical assess-

ment. Arthritis Rheum 1979;22:837–44.

20. Fam A, Rubenstein J, Chin-Sang H, et al. Computed

tomography in the diagnosis of early ankylosing

spondylitis. Arthritis Rheum 1985;28:930–7.

21. Carrera G, Foley W, Kozin F, et al. CT of sacroiliitis.

AJR Am J Roentgenol 1981;136:41–6.

22. Wittram C, Whitehouse G, Bucknall RC. Fat sup-

pressed contrast enhanced MR imaging in the

assessment of sacroiliitis. Clin Radiol 1996;51:554–8.

23. Wittram C, Whitehouse G, Williams JW, et al.

A comparison of MR and CT in suspected sacroilii-

tis. J Comput Assist Tomogr 1996;20:68–72.

24. Arslan H, Emin Sakarya M, Adak B, et al. Duplex and

colorDopplersonographic findings inactivesacroiliitis.

AJR Am J Roentgenol 1999;173:677–80.

25. Klauser A, De Zordo T, Feuchtner G, et al. Feasibility

of ultrasound-guided sacroiliac joint injection

considering sonoanatomic landmarks at two

different levels in cadavers and patients. Arthritis

Rheum 2008;59(11):1618–24.

26. Braun J, Bollow M, Eggens U, et al. Use of dynamic

magnetic resonance imaging with fast imaging in

the detection of early and advanced sacroiliitis in

spondyloarthropathy patients. Arthritis Rheum

1994;37:1039–45.

27. Kurugoglu S, Kanberoglu K, Kanberoglu A, et al.

MRI appearances of inflammatory vertebral osteitis

in early ankylosing spondylitis. Pediatr Radiol

2002;32:191–4.

28. Murphey M, Wetzel L, Bramble JM, et al. Sacroiliitis:

MR imaging findings. Radiology 1991;180:239–44.

29. Weber U, Ostergaard M, Lambert RG, et al. The

impact of MRI on the clinical management of

inflammatory arthritides. Skeletal Radiol 2011;

40(9):1153–73.

30. Yu W, Feng F, Dion E, et al. Comparison of radiog-

raphy, computed tomography and magnetic reso-

nance imaging in the detection of sacroiliitis

accompanying ankylosing spondylitis. Skeletal Ra-

diol 1998;27:311–20.

Amrami854

31. Vanhoenacker F, Eyselbergs M, Cotten A. Spinal

degeneration: beyond degenerative disc disease:

how to discriminate degeneration from spondyloar-

thropathies?Neuroradiology2011;53(Suppl 1):175–9.

32. Blum U, Buitrago-Tellez C, Mundinger A, et al.

Magnetic resonance imaging (MRI) for detection of

active sacroiliitis; a prospective study comparing con-

ventional radiography, scintigraphy, and contrast

enhanced MRI. J Rheumatol 1996;23:2107–15.

33. Xu M, Lin Z, Deng X, et al. The Ankylosing Spondylitis

Disease Activating Score is a high discriminatory

measure of disease activity and efficacy following tu-

mour necrosis factor-a inhibitor therapies in ankylosing

spondylitis and undifferentiated spondyloarthropathies

in China. Rheumatology (Oxford) 2011;50(8):1466–72.

34. Fox M, Onofrio B, Kilgore JE. Neurological complica-

tions of ankylosing spondylitis. J Neurosurg 1993;

78:871–8.

35. Arnett F. Seronegative spondyloarthropathies. Bul-

letin on the Rheumatic Diseases, Arthritis Founda-

tion 1987;37:1–12.

36. Braunstein E, Martel W, Moidel R. Ankylosing spon-

dylitis in men and women: a clinical and radio-

graphic comparison. Radiology 1982;144:91–4.

37. Fang D, Leong J, Ho EK, et al. Spinal pseudarthrosis

in ankylosing spondylitis. J Bone Joint Surg 1988;70:

443–7.

38. Suarez-Almazor ME, Russell AS. Anterior atlantoaxial

subluxation in patients with spondyloarthropathies:

association with peripheral disease. J Rheumatol

1988;15(6):973–5.

39. Mitchell M, Sartoris D, Moody D, et al. Cauda equina

syndrome complicating ankylosing spondylitis. Radi-

ology 1990;175:521–5.

40. Leczinsky C. The incidence of arthropathy in a ten-

year series of psoriasis cases. Acta Derm Venereol

1948;28:483–7.

41. Jones S, Armas J, CohenMG, et al. Psoriatic arthritis:

outcome of disease subsets and relationship of joint

disease to nail and skin disease. Br J Rheumatol

1994;33:834–9.

42. Glazebrook K, Guimaraes L, Murthy NS, et al. Iden-

tification of intraarticular and periarticular uric acid

crystals with dual-energy CT: initial evaluation. Radi-

ology 2011;261(2):516–24.

43. Turlik M. Seronegative arthritis as a cause of heel

pain. Clin Podiatr Med Surg 1990;7(2):369–75.

44. Grandbois L, Lomasney L, Demos TC, et al. Radio-

logic case study. Seronegative spondyloarthropathy

associated with Crohn’s disease. Orthopedics 2005;

28(11):1296, 1375–99.

45. McEwen C, DiTata D, Lingg C, et al. Ankylosing

spondylitis and spondylitis accompanying ulcerative

colitis, regional enteritis, psoriasis and Reiter’s

disease. Arthritis Rheum 1971;14:291–318.

46. Mielants H, Veys E. Enteropathic arthritis. In:

Schumacher HJ, Klippel J, Koopman W, editors.

Primer on the rheumatic diseases, vol. 10. Atlanta

(GA): Arthritis Foundation; 1993. p. 163.

47. Aydeniz A, Altindag O, O�gut E, et al. Late onset

spondyloarthropathy mimicking polymyalgia rheu-

matica. Rheumatol Int 2012;32(5):1357–8.