The evidence for whole-spine MRI in the assessment of axial spondyloarthropathy

RHEUMATOLOGY Rheumatology 2010;49:426 432 doi:10.1093/rheumatology/kep427 Advance Access publication 11 January 2010 Review The evidence for whole-spine MRI in the assessment of axial spondyloarthropathy Alexander N. Bennett 1, Helena Marzo-Ortega 2, Amer Rehman 3, Paul Emery 2 and Dennis McGonagle 2, on behalf of the Leeds Spondyloarthropathy Group Abstract REVIEW In the past decade, fat-suppression MRI techniques have been increasingly used for the assessment of axial-spa. Indeed, newly proposed classification criteria have suggested the inclusion of fat-suppression MRI for the evaluation of the SI joint in inflammatory back pain (IBP) of suspected axial-spa. However, recent data on the whole spine have identified certain MRI spinal lesions to be highly diagnostic of axial-spa; that the SI joint can be spared in axial-spa; and that IBP may originate in the lumbar spine rather than SI joint. Therefore, it is proposed that MRI of the whole spine and not just the SI joint should now become a routine part of the assessment of axial-spa. Not only is spinal MRI of great diagnostic utility in axial-spa but there is also increasing evidence to suggest that it can play a significant role in the management, in particular directing anti-tnf therapy in AS, and also it may be prognostically useful in axial-spa. With the wider availability, improving technology and falling cost of MRI, and the difficulty that clinical assessment of axial-spa poses, especially in early disease, there is now a strong case for the use of whole-spine MRI in the diagnosis and management of axial-spa. Key words: Spondyloarthropathy, Ankylosing spondylitis, MRI, Diagnosis, Prognosis, Management. Introduction the need to image axial-spa and historical limitations The SpAs are a heterogeneous group of diseases that have a propensity for inflammation at diverse skeletal sites that most typically involve entheses, related structures and adjacent bone [1, 2]. The clinical presentation of SpA varies between peripheral disease, axial disease and a combination of both. The clinical presentations of the peripheral SpA-related group of diseases that include monoarticular synovitis, dactylitis or osteitis are diverse, but for the majority of cases peripheral SpA can be readily diagnosed on the clinical history and examination findings that are supplemented by basic laboratory findings. Unlike peripheral skeleton involvement, diagnosis of isolated 1 Defence Medical Rehabilitation Centre, Headley Court, Epsom, Surrey, 2 Section of Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Chapel Allerton Hospital, Leeds, UK and 3 Department of Medical Imaging, American Hospital Dubai, Oud Metha, Dubai, United Arab Emirates. Submitted 31 July 2009; revised version accepted 16 November 2009. Correspondence to: Dennis McGonagle, Academic Section of Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Chapel Allerton Hospital, Chapel Town Road, Leeds LS7 4SA, UK. E-mail: d.g.mcgonagle@leed.ac.uk axial disease in SpA is more challenging. In this article, we synthesize the MRI evidence that has emerged in recent years to make the case for the use of whole-spine MRI, including fat-suppression sequences, as a useful tool in the diagnosis of axial-spa. We also consider the value of MRI as a prognostic test and as a tool for monitoring disease in axial-spa. Many cases of SpA have inflammation confined to the spine and SI joints, most notable in the younger HLA-B27- positive group with inflammatory back pain (IBP). Assessment can be challenging as the spine is not easy to assess with multiple joints, discs, entheses and ligaments, most of which are located deep to the clinician s hands making clinical recognition much more difficult. Until recently, with the exception of radiographic sacroiliitis [3], there has been no confirmatory diagnostic test for axial-spa or AS and it can take up to 10 years before radiographic sacroiliitis becomes diagnostic [4, 5]. Therefore, the physician has had to formulate a diagnosis of suspected early axial-spa based on a synthesis of clinical, serological, genetic and radiological information. Diagnosis is made even more difficult by the fact that acute phase response is often normal in axial-spa and clinical manifestations are often intermittent or transient [6]. Bone scintigraphy and CT of the SI joints have been used to assess axial-spa,! The Author 2010. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Evidence for whole-spine MRI but have limitations. Both expose the patient to high levels of ionizing radiation; bone scintigraphy has a very low specificity and CT is limited to detecting bone changes in established disease, albeit prior to detection on plain radiography, hence both are of limited use in early axial-spa [7]. The recent demonstration of the efficacy of biologic therapies in pre-radiographic axial-spa [8, 9] is another reason why there is a need to move beyond the existing limitations of traditional diagnostic and assessment tools. For all of these reasons, MRI has become the imaging modality of choice, but herein we argue that it is still under utilized and it should become part of the routine assessment of axial disease. Advent of MRI in axial-spa and lessening of cost- and scientific-based objections Following the original studies from the early 1990s [10, 11], MRI has been increasingly used for the assessment of suspected axial-spa. In particular fat-suppression MRI sequences, which depict osteitis [12, 13] have been favoured by rheumatologists. Previously, one of the arguments against MRI has been that it is expensive and not widely available. However, MRI accessibility is no longer a limiting factor in developed countries, and fat-suppression imaging alone is relatively fast. Furthermore, scanning times are falling, imaging resolution is improving and intravenous contrast is no longer required to identify pathology [7, 14, 15]. This latter development is of particular translational relevance, given the recent identification of Gd-DTPA-induced nephrogenic systemic fibrosis [16]. assessment of early disease in SpA with statistical analysis on clinical, radiological, serological and genetic features and their singular or combined effect on the probability of disease. With knowledge of the background prevalence of axial-spa in chronic back pain in the general population (5%) [23], likelihood ratios (LRs) of other common features of disease in axial-spa were combined to estimate an overall probability of disease. LRs, a formulation of the sensitivity and specificity of a test into one value, which is an indicator of the diagnostic value of that test, are regarded as the best statistical method of measuring and expressing diagnostic accuracy [24]. Multiples of LRs from clinical, radiological and genetic features combined with pre-test odds based on background prevalence [23] result in a post-test probability of disease, in this case axial-spa being present. Therefore, according to the previously proposed theoretical scheme, a patient with IBP (LR = 3.1) who is HLA-B27 positive (LR = 9.0) and has MRI evidence of sacroiliitis (LR = 9.0) has a post-test probability of 93% [5, 21, 22]. A post-test probability of 590% is regarded as definite, between 80 and 90% is probable and 410 20% unlikely [21]. However, the sensitivities and specificities used to calculate the above LRs were taken from different studies, without non-axial-spa control groups, in different populations and were regarded as the best estimates at that FIG. 1MRI appearances of axial-spa. MRI STIR sequence images of SpA with multiple inflammatory RL T3 T5 (arrows). MRI in diagnosis of SI joint disease in SpA It has been said that AS is an outcome rather than a diagnosis, since radiographic joint destruction and/or fusion of the SI joints must be evident before a diagnosis can be accurately entertained [3]. Therefore, an extremely important fact is that available and emerging data have shown that MRI of the SI joints are of definite diagnostic and prognostic utility in axial-spa [14, 17, 18] as it defines cases with SI joint osteitis that will progress to radiographic change and a definite diagnosis of modified New York criteria AS [14, 17]. Recently, new axial-spa criteria have been proposed by the Assessments in Spondyloarthritis International Group (ASAS) [19]. The criteria allow for the combination of clinical, serological, genetic and, importantly, SI joint MRI findings to be used in the diagnosis of axial-spa. This is a vital step in the recognition of early pre-radiographic disease where the imaging, clinical and therapeutic findings indicate that this phase of the disease is virtually identical to established radiographic evident AS and that both can be viewed as a continuum of disease under the same diagnostic umbrella [8, 17, 20]. The new ASAS criteria, as proposed by Rudwaleit and colleagues [5, 21, 22], are based on an easy bedside www.rheumatology.oxfordjournals.org 427

Alexander N. Bennett et al. time with little supportive evidence [5]. It is therefore of note that some of these currently used best estimate LRs have been surpassed by actual data from recent MRI studies [17, 25] that could add significantly to improving the diagnostic accuracy of these classification criteria. One study, a prospective 8-year follow-up of patients with very early IBP shows that severe MRI sacroiliitis predicts future AS with a LR of 8.0 [17]. However, perhaps more importantly, there is now evidence [25] to support the use of whole-spine MRI, and not just SI joint MRI in the assessment of potential axial-spa patients which is further set out below. The need for whole-spine MRI in early SpA diagnosis Although MRI of the SI joints has traditionally been used to assess IBP in suspected axial-spa, this may be inadequate for several reasons. First, inflammatory lesions in the spine have been found most commonly in the thoracic spine [25, 26]; secondly, not only can spinal and SI joint inflammatory lesions occur simultaneously in IBP patients [27], but as many as 23% of AS patients with clinically active disease may only have inflammatory spinal lesions and no evidence of active inflammatory sacroiliitis [28], even in very early disease [27]. Indeed MRI studies have shown that lower lumbar back pain may stem solely from lumbar inflammatory lesions rather than SI joint lesions [27]. Thirdly, cervical spine involvement is common not only in AS but also in psoriatic axial-spa [29 32]. Furthermore, the total load of spinal inflammation on fat-suppressed MRI predicts response to anti-tnf therapy [28]. All the above evidence is a powerful reason to perform MRI of the whole spine. A recent study, assessing the diagnostic utility of whole-spine MRI in 185 patients with chronic back pain, has shown that MRI alone without any clinical information could be virtually diagnostic of axial-spa [25]. This study identified that previously described classic SpA MRI lesions were actually common in other causes of chronic back pain, notably degenerative disease and these lesions were also occasionally evident in normal asymptomatic patients [25, 27]. However, it also noted that certain types or numbers of lesions were highly specific to axial-spa. Multiple (53) inflammatory Romanus lesions (RLs) (Fig. 1) (LR: 12.4) and severe inflammatory RLs (LR: infinite) in subjects under the age of 50 years, or mild to moderate posterior element lesions (Fig. 2) on fat-suppressed sequences (LR: 14.5) were all highly diagnostic (LR > 10 is regarded as large and often conclusive evidence of disease) [33]. To illustrate this: if a subject under the age of 50 years assessed in the rheumatology clinic who meets IBP criteria (LR = 3.1) [5, 22], is HLA-B27 positive (LR = 9.0) [5, 22] and has at least one mild to moderate posterior element FIG. 2Inflammatory posterior element lesions on STIR MRI sequences in axial-spa. (A) Posterior element/costovertebral T12 (arrow) lesion. (B) Superior facet lesion of T6 (arrowhead) and grade 1 endplate lesion (arrow) inferior endplate T3. 428 www.rheumatology.oxfordjournals.org

Evidence for whole-spine MRI lesion (LR = 14.5) or 53 inflammatory RLs (LR = 12.4) on spinal MRI, then that patient has definite axial-spa (post-test probability = 95%) [21]. This emphasizes the substantial diagnostic utility that the addition of whole-spine fat-suppressed MRI has in the diagnosis of axial-spa rather than scanning the SI joints alone. In fact the aforementioned MRI spinal lesions have such high LRs for a diagnosis of axial-spa [34] that a combination of these spinal lesions in a patient <50 years old may be sufficient to confirm diagnosis of axial-spa (post-test probability = 99%) even in the absence of any other clinical features such as HLA-27, raised CRP or even IBP. So, in certain circumstances the power of whole-spine fat-suppression MRI as a single test supersedes the most optimistic expectations. Beyond diagnostics predicting therapy response and prognosis An argument can be made for using whole-spine MRI as part of the selection criteria for the use of anti-tnf therapy, especially when the drug is rationed. Active inflammatory spinal lesions on fat-suppressed MRI are a significant predictor of response (BASDAI 50) to anti-tnf [28]. Indeed the combination of an abnormal FIG. 3FRL: example of multiple FRLs (white arrows) with one end plate lesion (black arrow). The endplate lesion is not a FRL because it extends to cover over 50% of the endplate. fat-suppressed MRI (Berlin MRI spinal score 511) with a raised CRP (>40 mg/l) and a disease duration shorter than 10 years gives a 99% response rate to anti-tnf [28]. In the current socio-economic environment, where anti-tnf drug availability is curtailed, such considerations may become increasingly relevant. The long-term prognosis in AS is linked to new bone formation, with bridging syndesmophytes at multiple levels and subsequent spinal fusion. Given that the natural history of AS is variable with a substantial number of cases not progressing to this stage, a question must be raised can MRI be used to predict which patients will develop fusion at a later date? Recent evidence has been published that suggests that spinal inflammation [35, 36], namely osteitis/enthesitis is associated with later spinal fusion, even after the administration of anti-tnf [35]. More widespread clinical use of wholespine MRI in the assessment of axial-spa is likely to lead to stronger confirmation of this evidence. Another recently described lesion is the inactive MRI RL [37] also called MRI Fatty RL (FRL) [38]. This post-inflammatory fat lesion that occurs at vertebral corners is highly specific to axial-spa, particularly when lesions are multiple (> 5 FRLs give a specificity = 98% and a LR for axial-spa = 12.6) [38] (Fig. 3). Therefore, the FRL might be of diagnostic utility in pre-radiographic axial-spa patients with a more protracted history of IBP where fat-suppression MRI is normal. Normal fatsuppression MRI is recognized to occur in active axial-spa in a small but significant number of patients [25, 28, 37]. For this reason, it would seem reasonable to continue to perform conventional T1W imaging in addition to fat-suppression imaging. Furthermore, preliminary evidence suggests that these lesions may predate syndesmophyte formation and spinal fusion [39]. Recommended MRI protocol This review has focused mainly on the benefits of fat-suppressed MRI of the whole spine; however, the benefits of T1W sequences have also been confirmed. The conspicuity of oedematous lesions can be optimized by removing the surrounding high signal of fat, either by suppression [Short Tau Inversion Recovery sequences (STIR)] or by saturation (application of an additional pulse to saturate the protons of fat), which can be applied to T2W or T1W sequences post-gadolinium. Previous authors have investigated the different techniques for depicting inflammatory lesions in axial-spa and have recommended the use of STIR sequences for this purpose [40] and we concur with this technique. A Philips 1.5 T MRI scanner (Philips Medical Systems, Best, Nederland BV) was used to image the whole spine and SI joints in our studies. The scanning protocol comprised five sequences sagittal T1W (TE 13, TR 525) and STIR (TE 70, TR 2500) cervicothoracic spine, sagittal T1 and STIR thoracolumbar spine and coronal oblique STIR pelvis centred on and angulated for the SI joints. Using a body coil and altering the field of view, acquisition time is 3 min per sequence. Thus, total scan time is 20 min. www.rheumatology.oxfordjournals.org 429

Alexander N. Bennett et al. MRI limitations Although there is evidence to suggest that whole-spine, fat-suppressed MRI is increasingly powerful in diagnosing, particularly, early axial-spa and influencing treatment decisions, it remains clear that MRI is not a panacea. Indeed, it has been widely reported that MRI of the spine, SI joint or both may be negative in active axial-spa [25, 27, 28] and that up to a third of AS patients with negative scans still get a major response to anti-tnf [28]. The reasons for these observations remain unclear, although may be due to inadequate resolution of current scanning techniques and hence, their inability to identify existing inflammation [13] or the natural fluctuation in the course of the disease [6]. Either way a negative scan does not necessarily exclude the disease and the new ASAS criteria recognize this by enabling a diagnosis to be made on clinical ± radiographic criteria alone with the exclusion of MRI. However, the MRI capabilities including resolution and sequence development are improving all the time and it is possible that the full spectrum of inflammatory lesions will be visualized in the future. Furthermore, in equivocal cases, a negative MRI scan is highly reassuring that other serious pathologies including malignancy or pathological fractures from other causes are absent. Conclusion There is now plentiful evidence to suggest that MRI of the whole spine and SI joints is an essential tool in the diagnosis, management and prognosis of axial-spa (Fig. 4). The role of MRI in the diagnosis of axial-spa has been recognized with the inclusion of MRI criteria in the new ASAS classification of axial-spa. However, the true potential of MRI has only been partially recognized as the new classification criteria have the limitation of only acknowledging positive SI joint scans with the exclusion of spinal MRI findings. Indeed, in any suspected SpA case already undergoing an MRI investigation of the SI joints, the additional assessment of sagittal sequences of the whole spine adds little to cost or time to the scan. Additionally, imaging of the whole spine may be of value for the diagnostic clues it may provide in patients with negative or equivocal changes at the most symptomatic spinal region involved as well as excluding important differentials such as degenerative spinal disease and malignancy [25, 41]. FIG. 4Benefits of whole-spine MRI in the assessment of axial-spa. Diagnosis (i) Proven diagnostic utility of MRI lesions on STIR sequences in axial-spa [14, 18, 25] (ii) Diagnostic utility on T1-weighted MRI when there are no inflammatory lesions present on STIR [37] Benefits of whole-spine MRI in the assessment of axial-spa Prognosis (i) STIR predicts long-term AS in early IBP [14, 17, 18] (ii) Spinal inflammatory lesions on STIR MRI predict syndesmophyte formation [35, 36] Treatment decisions (i) Increased spinal inflammatory lesion score on STIR sequence predicts response to anti-tnf [28] (iii) Post-inflammatory corner lesions on T1 MRI predict syndesmophyte formation [40] 430 www.rheumatology.oxfordjournals.org

Evidence for whole-spine MRI Furthermore, the role of fat-suppression MRI in influencing treatment decisions in axial-spa/as by identifying likely responders to anti-tnf [28] has become more established, and fat-suppressed MRI of the whole spine and SI joints has prognostic value in predicting AS in early IBP [17] and may be a predictor of spinal syndesmophyte formation and possible spinal fusion [35, 36]. These are compelling reasons for the routine use of fat-suppression MRI of the whole spine in the assessment of axial-spa. Rheumatology key messages. Whole-spine MRI plays an essential role in the early diagnosis of axial-spa.. Whole-spine MRI helps identify which patients respond best to anti-tnf therapy.. Whole-spine MRI can be used to predict SI joint and spinal new bone formation or fusion. Disclosure statement: P.E. has undertaken clinical trials and provided expert advice to Schering Plough, Centocor, Wyeth and Abbott. All other authors have declared no conflicts of interest. References 1 McGonagle D, Khan MA, Marzo-Ortega H, O Connor P, Gibbon W, Emery P. Enthesitis in spondyloarthropathy. Curr Opin Rheumatol 1999;11:244 50. 2 McGonagle D, Marzo-Ortega H, O Connor P et al. Histological assessment of the early enthesitis lesion in spondyloarthropathy. Ann Rheum Dis 2002;61:534 7. 3 van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27:361 8. 4 Mau W, Zeidler H, Mau R et al. Clinical features and prognosis of patients with possible ankylosing spondylitis. Results of a 10-year follow-up. J Rheumatol 1988;15: 1109 14. 5 Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis: do we need new criteria? Arthritis Rheum 2005;52:1000 8. 6 Stone MA, Pomeroy E, Keat A et al. Assessment of the impact of flares in ankylosing spondylitis disease activity using the Flare Illustration. Rheumatology 2008;47: 1213 8. 7 Braun J, van der Heijde D. Imaging and scoring in ankylosing spondylitis. Best Pract Res Clin Rheumatol 2002; 16:573 604. 8 Barkham N, Keen H, Coates LC et al. A randomised controlled trial of infliximab shows clinical and MRI efficacy in patients with HLA-B27 positive very early ankylosing spondylitis. Arthritis Rheum 2009;60:946 54. 9 Haibel H, Rudwaleit M, Listing J et al. Efficacy of adalimumab in the treatment of axial spondylarthritis without radiographically defined sacroiliitis: results of a twelveweek randomized, double-blind, placebo-controlled trial followed by an open-label extension up to week fifty-two. Arthritis Rheum 2008;58:1981 91. 10 Ahlstrom H, Feltelius N, Nyman R, Hallgren R. Magnetic resonance imaging of sacroiliac joint inflammation. Arthritis Rheum 1990;33:1763 9. 11 Braun J, Bollow M, Eggens U, Konig H, Distler A, Sieper J. Use of dynamic resonance imaging with fast imaging in the detection of early and advanced sacroiliitis in spondyloarthropathy patients. Arthritis Rheum 1994;37: 1039 45. 12 Marzo-Ortega H, O Connor P, Emery P, McGonagle D. Sacroiliac joint biopsies in early sacroiliitis. Rheumatology 2007;46:1210 1. 13 Appel H, Kuhne M, Spiekermann S et al. Immunohistologic analysis of zygapophyseal joints in patients with ankylosing spondylitis. Arthritis Rheum 2006; 54:2845 51. 14 Oostveen J, Prevo R, den Boer J, van de Laar M. Early detection of sacroiliitis on magnetic resonance imaging and subsequent development of sacroiliitis on plain radiography. A prospective, longitudinal study. J Rheumatol 1999;26:1953 8. 15 Wittram C, Whitehouse GH, Bucknall RC. Fat suppressed contrast enhanced MR imaging in the assessment of sacroiliitis. Clin Radiol 1996;51:554 8. 16 Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA. Gadodiamide-associated nephrogenic systemic fibrosis: why radiologists should be concerned. AJR Am J Roentgenol 2007;188:586 92. 17 Bennett AN, McGonagle D, O Connor P et al. Severity of baseline magnetic resonance imaging evident sacroiliitis and HLA-B27 status in early inflammatory back pain predict radiographically evident ankylosing spondylitis at eight years. Arthritis Rheum 2008;58:3413 8. 18 Puhakka KB, Jurik AG, Schiottz-Christensen B et al. Magnetic resonance imaging of sacroiliitis in early seronegative spondylarthropathy. Abnormalities correlated to clinical and laboratory findings. Rheumatology 2004;43:234 7. 19 Rudwaleit M, Landewe R, Van der Heijde D et al. The development of assessment of SpondyloArthritis international Society (ASAS) classification criteria for axial spondyloarthritis (Part 1): classification of paper patients by expert opinion including uncertainty appraisal. Ann Rheum Dis 2009;68:770 6. 20 Rudwaleit M, Haibel H, Baraliakos X et al. The early disease stage in axial spondylarthritis: results from the German spondyloarthritis inception cohort. Arthritis Rheum 2009;60:717 27. 21 Rudwaleit M, Feldtkeller E, Sieper J. Easy assessment of axial spondyloarthritis (early ankylosing spondylitis) at the bedside. Ann Rheum Dis 2006;65:1251 2. 22 Rudwaleit M, van der Heijde D, Khan MA, Braun J, Sieper J. How to diagnose axial spondyloarthritis early. Ann Rheum Dis 2004;63:535 43. 23 Underwood MR, Dawes P. Inflammatory back pain in primary care. Br J Rheumatol 1995;34:1074 7. 24 McGee S. Simplifying likelihood ratios. J Gen Intern Med 2002;17:646 9. 25 Bennett AN, Rehman A, Hensor EM, Marzo-Ortega H, Emery P, McGonagle D. Evaluation of the diagnostic utility www.rheumatology.oxfordjournals.org 431

Alexander N. Bennett et al. of spinal magnetic resonance imaging in axial spondylarthritis. Arthritis Rheum 2009;60:1331 41. 26 Baraliakos X, Landewe R, Hermann KG et al. Inflammation in ankylosing spondylitis: a systematic description of the extent and frequency of acute spinal changes using magnetic resonance imaging. Ann Rheum Dis 2005;64: 730 4. 27 Marzo-Ortega H, McGonagle D, O Connor P et al. Baseline and 1 year magnetic resonance imaging of the sacroiliac joint and lumbar spine in very early inflammatory back pain. Relationships between symptoms, HLA-B27, and disease extent and persistence. Ann Rheum Dis 2008; 68:1721 7. 28 Rudwaleit M, Schwarzlose S, Hilgert ES, Listing J, Braun J, Sieper J. MRI in predicting a major clinical response to anti-tumour necrosis factor treatment in ankylosing spondylitis. Ann Rheum Dis 2008;67:1276 81. 29 Queiro R, Sarasqueta C, Torre JC, Tinture T, Lopez-Lagunas I. Prevalence and predictors of cervical involvement in psoriatic spondyloarthropathy. J Clin Rheumatol 2002;8:23 9. 30 Queiro R, Belzunegui J, Gonzalez C et al. Clinically asymptomatic axial disease in psoriatic spondyloarthropathy. A retrospective study. Clin Rheumatol 2002;21: 10 3. 31 El Maghraoui A, Bensabbah R, Bahiri R, Bezza A, Guedira N, Hajjaj-Hassouni N. Cervical spine involvement in ankylosing spondylitis. Clin Rheumatol 2003;22:94 8. 32 Lee JY, Kim JI, Park JY et al. Cervical spine involvement in longstanding ankylosing spondylitis. Clin Exp Rheumatol 2005;23:331 8. 33 Jaeschke R, Guyatt GH, Sackett DL. Users guides to the medical literature III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA 1994;271:703 7. 34 Bennett AN, Rehman A, Hensor EM, Marzo-Ortega H, Emery P, McGonagle D. Evaluation of the diagnostic utility of spinal magnetic resonance imaging in axial spondylarthritis. Arthritis Rheum 2009;60:1331 41. 35 Baraliakos X, Listing J, Rudwaleit M, Sieper J, Braun J. The relationship between inflammation and new bone formation in patients with ankylosing spondylitis. Arthritis Res Ther 2008;10:R104. 36 Maksymowych WP, Chiowchanwisawakit P, Clare T, Pedersen SJ, Ostergaard M, Lambert RG. Inflammatory lesions of the spine on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis: evidence of a relationship between inflammation and new bone formation. Arthritis Rheum 2009;60:93 102. 37 Herman KG, Althoff CE, Schneider U, Zühlsdorf S, Lembcke A, Hamm B, Bollow M. Spinal changes in patients with spondyloarthritis: comparison of MR imaging and radiographic appearances. HYPERLINK Radiographics 2005;25:559 69, discussion 569 70. 38 Bennett AN, Rehman A, Hensor EMA M-OH, Emery PDM. The fatty Romanus lesion a non-inflammatory spinal MRI lesion specific for axial-spondyloarthropathy. Ann Rheum Dis 2009 Aug 9 [Epub ahead of print]. 39 Chiowchanwisawakit P, Lambert RGW, Clare T, Maksymowych W. Post-inflammatory focal fat lesions in the spine of patients with AS predict the development of new syndesmophytes. Arthritis Rheum 2005;64: 1141 4. 40 Hermann KG, Landewé RB, Braun J, van der Heijde DM. Magnetic resonance imaging of inflammatory lesions in the spine in ankylosing spondylitis clinical trials: is paramagnetic contrast medium necessary? J Rheumatol 2005; 32:2056 60. 41 Baraliakos X, Hermann KG, Landewe R. et al. Assessment of acute spinal inflammation in patients with ankylosing spondylitis by magnetic resonance imaging: a comparison between contrast enhanced T1 and short tau inversion recovery (STIR) sequences. Ann Rheum Dis 2005;64: 1141 4. 432 www.rheumatology.oxfordjournals.org