Impact of replacing radiographic sacroiliitis by magnetic resonance imaging structural lesions on the classification of patients with axial spondyloarthritis

Impact of replacing radiographic sacroiliitis by magnetic resonance imaging structural lesions on... Abstract Objectives To investigate in patients with chronic back pain of a short duration, the utility of adding structural MRI lesions of the SI joints to the imaging criterion of the Assessment of SpondyloArthritis International Society (ASAS) axial SpA (axSpA) criteria and the utility of replacement of radiographic sacroiliitis by structural MRI lesions. Methods MRI STIR (inflammation, MRI-SI), MRI T1-weighted images (structural lesions, MRI-SI-s) and radiographs of the SI joints of patients in the SPondyloArthritis Caught Early-cohort (chronic back pain: ⩾3 months, ⩽2 years; onset <45 years) were scored by two well-calibrated readers. Previously proposed cut-offs for a positive MRI-SI-s were used (based on <5% prevalence in no-SpA patients): erosions ⩾3, fatty lesions ⩾3, fatty lesions and/or erosions (erosions/fatty lesions) ⩾5. Using the definitions of MRI-SI-s, patients were classified according to the ASAS axSpA criteria. Results Twenty-nine of 294 patients were modified New York (mNY) positive and 32 were MRI-SI-s positive (erosions/fatty lesions ⩾5). Agreement between mNY and MRI-SI-s (erosions/fatty lesions ⩾5) was moderate (κ: 0.58). Using the erosions/fatty lesions ⩾5 cut-off, 3/294 additional patients were classified as axSpA (adding MRI). Using this cut-off instead of mNY (replacing mNY), classification did not change in 286 patients (97.3%), but 5 patients (1.7%) would not be classified as axSpA and 3 previously unclassified patients (1.0%) would be classified as axSpA. Similar results were seen for the other cut-offs (erosions ⩾3 and fatty lesions ⩾3). Conclusion Assessment of structural lesions (fatty lesions and erosions) on MRI-SI instead of or in addition to conventional radiographs does not lead to a different ASAS axSpA classification in most of the patients with early disease onset. This suggests that structural lesions (fatty lesions and erosions) can be reliably used in the ASAS axSpA classification of patients, as both addition and replacement of radiographs of the SI joints. axial spondyloarthritis, sacroiliitis, structural lesions, MRI, radiographs Rheumatology key messages Structural MRI lesions in SI joints can be reliably used to classify patients with axial SpA. Adding or replacing radiographic sacroiliitis by structural MRI lesions led to minor changes in axial SpA classification. Introduction Recently, we have demonstrated that in the DEvenir des Spondyloarthrites Indifférenciées Récentes (DESIR) cohort of patients with inflammatory back pain with a high suspicion of having axial SpA (axSpA), structural lesions on MRI can be used reliably either as an addition to or as a substitute for radiographs in the Assessment of SpondyloArthritis International Society (ASAS) axSpA classification of patients spondyloarthritis [1]. Since this is a new concept with potential consequences for daily clinical practice, replication in other cohorts is highly warranted. The current study investigates the same research question in an entirely independent cohort, the SPondyloArthritis Caught Early (SPACE) cohort of patients with chronic back pain of a short duration. The rationale behind using structural lesions on MRI instead of conventional radiographs is mainly based on the fact that a reliable detection of sacroiliitis on conventional radiographs is notoriously difficult. Sacroiliitis is the hallmark of AS and the presence of sacroiliitis on conventional radiography is obligatory in the modified New York (mNY) criteria for AS [2]. However, it has been shown that substantial observer variation (interreader, intrareader) exists for both radiologists and rheumatologists, and training does not lead to improvement [2]. Recent data from the DESIR cohort have also revealed a poor to moderate agreement between different readers leading to substantial misclassification of patients [3]. With the help of MRI it has become possible to detect active sacroiliitis (bone marrow oedema, BME) in early stages of axSpA. The ASAS definition for a positive MRI is solely based on identification of BME highly suggestive for SpA, but structural lesions can be seen [fatty lesions, erosions, sclerosis and (partial) ankylosis] are visible on MRI as well [4]. The 3D (albeit tomographic) character of MRI may be an advantage as compared with the 2D projection of conventional radiographs [5]. Another advantage of MRI above radiography is that patients are not exposed to ionizing radiation. Both the SPACE and the DESIR are cohorts of early disease. Since it often takes 6–8 years from the onset of symptoms before radiographic sacroiliitis can be detected on plain radiographs, conventional radiography is a suboptimal imaging technique especially in patients with complaints of a short duration (in addition to the earlier described reliability issues). The DESIR cohort only includes patients with inflammatory back pain (IBP) with up to 3 years of symptoms, whereas the SPACE cohort includes patients with chronic back pain (CBP) with up to 2 years of symptoms. Different reader pairs were involved in the scoring process of the two cohorts, to rule out a reader-dependent effect. In this study we aim to evaluate the usefulness of structural lesions with regard to the ASAS axSpA classification criteria in the SPACE cohort. First of all, we determine the agreement between sacroiliitis seen on radiographs (mNY) and structural lesions seen on MRI. Then we aim to evaluate the potential impact of adding structural lesions on MRI to the definition of a positive MRI or replacing radiographic sacroiliitis by MRI structural lesions on the ASAS axSpA classification of patients. Methods Study population For this analysis, baseline data from the SPACE cohort were used. An extensive description of the SPACE cohort has been given elsewhere [6]. In short, SPACE is an inception cohort with on-going inclusion and follow-up of patients with CBP of short duration (⩾3 months but ⩽2 years, with onset <45 years). Patients were recruited from five participating centres in the Netherlands (Leiden, Amsterdam, Gouda), Norway (Oslo) and Italy (Padua). Approval by the local medical ethics committees was obtained, as well as written informed consent from all patients in accordance with the Declaration of Helsinki. A full diagnostic work-up was performed in all patients including: HLA-B27 testing, conventional radiographs and MRI of the SI joints (X-SI and MRI-SI, respectively), and the assessment of all other SpA features, in agreement with the descriptions supplied by the ASAS group [7]. Imaging and scoring methods MRI-SI was performed on a 1.5 Tesla machine (Philips Medical Systems, Best, Netherlands). Coronal oblique T1-weighted TSE (TR 550/TE 10) and STIR (TR 2500/TE 60) with a slice thickness of 4 mm were the acquired sequences used. All available baseline MRI-SI and X-SI were read independently, in different reading sessions, by two trained, well-calibrated readers (P.A.B., M.d.H.). Readers were blinded to the score of the other reader and other modality as well as the clinical information, throughout the scoring process. The mNY criteria were used to assess radiographs of the SI joints; radiographic sacroiliitis was defined as bilateral grade ⩾2 or unilateral grade ⩾3 [8]. In consensus with the ASAS definition, an MRI-SI was defined positive if one BME lesion highly suggestive of SpA was present on two or more consecutive slices, or otherwise if several BME lesions highly suggestive of SpA were visible on a single slice [9]. A third reader served as adjudicator (R.vd.B.) in case of disagreement among the two initial readers regarding a positive MRI (ASAS definition) or the presence of sacroiliitis (mNY criteria). The presence of structural lesions, namely fatty lesions, erosions, sclerosis and (partial) ankylosis, was assessed on MRI T1-weighted images in conjunction with the STIR images. A scoring system with similarities to the methodology outlined in the Spondyloarthritis Research Consortium of Canada (SPARCC) online training module, as described by Weber et al. [10] was used. Weber et al. proposed a scoring system to quantify structural lesions called SPARCC SI structural lesion score. This method is founded on the assessment of lesions (present vs absent) counting them in each quadrant on six consecutive slices through the SI joints. The starting point is the slice on which at least 1 cm of vertical height of the cartilage compartment can be seen, from anterior to posterior, evaluating the cartilaginous compartment of the SI joints and the antero-inferior portion of the SI joint. Each SI joint is split into four quadrants. Structural lesions were taken into account only if present on at least two consecutive slices. This is reflected by a maximum score of 40 per lesion (5 lesions per quadrant × 4 quadrants × 2 SI joints) except for (partial) ankylosis. As an exception, (partial) ankylosis was considered sufficient when seen on a single slice reflected by a maximum score of 24 per patient. As shown earlier by our group, an MRI positive for structural lesions (MRI-SI-s) was determined by the use of different cut-offs [11]. These chosen cut-offs were anchored on ⩽5% false positives whereby the false positives were specified as structural lesions among patients not having axSpA according to the ASAS axSpA criteria. The described cut-offs have pointed out to be: erosions ⩾3, fatty lesions ⩾3, fatty lesions and/or erosions ⩾5. In this early cohort, the prevalence of sclerosis and (partial) ankylosis was so low that there was no cut-off that could clearly differentiate between SpA and no-SpA patients. As a consequence, these types of lesions were not further considered. Classification criteria Patients were classified according to the ASAS axSpA criteria. Subsequently, patients were grouped based on the way they met the criteria: through the imaging arm of the ASAS axSpA criteria alone (either by mNY criteria and/or by positive MRI); through the clinical arm of the ASAS axSpA criteria alone; or through both. If patients fulfilled more than one category, they were classified in that way, reflected by seven possible combinations (Figs 1 and 2). The no-axSpA group is made up of patients not fulfilling the ASAS axSpA criteria. Fig. 1 View largeDownload slide Addition of MRI-SI-s to the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 1 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 1 View largeDownload slide Addition of MRI-SI-s to the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 1 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 2 View largeDownload slide Replacement of mNY criteria by MRI-SI-s in the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 2 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 2 View largeDownload slide Replacement of mNY criteria by MRI-SI-s in the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 2 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Data analysis Descriptive statistics were used to calculate disease characteristics of the included patients. In the analysis, regarding structural lesions the mean score of the two readers in agreement of a positive MRI (ASAS definition) was used. In case of disagreement, the mean of the scores of the adjudicator and the reader in agreement with the adjudicator’s judgement regarding a positive MRI for that particular case were used. Agreement about the absence or presence of structural lesions using both imaging modalities (X-SI and MRI-SI-s) was assessed by cross-tabulation and expressed as Cohen’s κ. Percentage positive agreement was calculated in order to leave out patients labelled as negative by the two readers, which could lead to an artificially high agreement. Subsequently, using the various definitions of MRI-SI-s, patients were classified by the ASAS axSpA criteria. MRI-SI-s was added to the imaging criterion of the ASAS axSpA criteria in the first place, resulting in an additional possibility to fulfil the imaging arm (scenario 1). Second, replacement of the mNY criterion by MRI-SI-s (scenario 2) was applied as if only an MRI was performed. Disease probabilities were calculated by the positive likelihood ratio (LR) product, for those patients changing ASAS axSpA classification in scenario 2. This was done by multiplying the individual LRs of all identified SpA features [12]. In patients with CBP with an assumed disease prevalence of axSpA of 5%, an LR product of 200 results in a positive predictive value of 90%. These patients were further described by their clinical phenotype: gender, age and whether a patient was diagnosed as axSpA according to the treating rheumatologist. The analyses were performed in STATA 12.0 (StataCorp LP, College Station, Texas, USA). Results Included in the analysis are 294 patients with complete imaging data at baseline (both MRI-SI and X-SI present). Table 1 describes patient characteristics. Patients had a mean (s.d.) age of 31.6 years of age (10.7 years) and a mean (s.d.) duration of back pain of 13.1 months (7.3 months). Of these patients, 34.6% were men and 34.6% were HLA-B27 positive. Table 1 Disease characteristics of the included patients Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) ASAS: Assessment of SpondyloArthritis International Society; IBP: inflammatory back pain; SPACE: SPondyloArthritis Caught Early. Table 1 Disease characteristics of the included patients Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) ASAS: Assessment of SpondyloArthritis International Society; IBP: inflammatory back pain; SPACE: SPondyloArthritis Caught Early. One hundred and three out of 294 patients (35.0%) fulfilled the ASAS axSpA criteria using the standard definition. Of these 103 patients, 50 patients fulfilled the imaging arm (48.5%) and 53 patients fulfilled the clinical arm only (51.5%). The most prevalent SpA features were IBP (62.6%), a positive family history for SpA (38.1%) and a good response to NSAIDs (31.6%). Regarding the prevalence of structural lesions, 20/294 patients (6.8%) showed three or more fatty lesions; 16 of these 20 patients (80%) already formally fulfilled the ASAS axSpA criteria. Thirty-four patients (11.6%) had three or more erosions, and 27 of these 34 (79.4%) already formally fulfilled the ASAS axSpA criteria. Thirty-one patients (10.5%) had 5 or more fatty and/or erosive lesions (combination), of which 26 patients (83.9%) already formally fulfilled the ASAS axSpA criteria. The agreement regarding the presence/absence of radiographic sacroiliitis and the presence/absence of structural lesions on MRI was moderate (Table 2). Subtle differences were seen between the various definitions used: κ: 0.51 (erosions ⩾3); κ: 0.45 (fatty lesions ⩾3); and κ: 0.58 (combination of fatty lesions and erosions ⩾5). Table 2 Agreement between sacroiliitis on conventional radiographs (mNY criteria) and a positive MRI-SI based on structural lesions mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 Agreement based on MRI-SI structural lesions using the three different definitions. mNY: modified New York criteria; MRI-SI: MRI of the SI joints. Table 2 Agreement between sacroiliitis on conventional radiographs (mNY criteria) and a positive MRI-SI based on structural lesions mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 Agreement based on MRI-SI structural lesions using the three different definitions. mNY: modified New York criteria; MRI-SI: MRI of the SI joints. Scenario 1 The addition of MRI-SI-s to the imaging criterion of the ASAS axSpA criteria was investigated first (Fig. 1). Classification did not change in the majority of the patients (96.3%) using a cut-off value of 5 for the combination of fatty lesions and/or erosions. If a cut-off value of 3 for erosions or fatty lesions only was used, comparable results were seen: classification did not change in 94.9 and 97.6% of the patients, respectively. Considering the combination of five fatty lesions and/or erosions, three patients would be classified additionally axSpA if structural lesions on MRI were taken into consideration. The positive LR products of the three additionally classified patients were 15.8, 5.1 and 2.5, corresponding to post-test probabilities of 44, 20 and 11%, respectively. The rheumatologist diagnosed only one of these three patients (all female and HLA-B27 negative) with axSpA. Regarding the definition based on erosions, five patients would be additionally classified (all female and HLA-B27 negative). Only two of them were diagnosed as axSpA by the rheumatologist. Regarding the definition based on fatty lesions, three patients would be additionally classified as axSpA (one was described above, the other two were male patients, HLA-B27 negative, of whom one was diagnosed as having axSpA by the rheumatologist). Some patients changed subgroups within the ASAS classification, without a change in ASAS axSpA positivity or negativity. Eight patients would be classified via different arms due to the presence of structural lesions, using the combination of fatty lesions and/or erosions. Five patients also showed structural lesions on MRI (Fig. 1) but were already classified via the imaging arm based on inflammatory lesions on MRI. Three other patients fulfilled the clinical arm only, but fulfilled the imaging arm as well based on a positive MRI-SI-s. Using the cut-offs for fatty lesions or erosions only (both cut-off values were 3), the same trends were seen (data not shown). Scenario 2 Second, the replacement of radiographic sacroillitis by structural lesions on MRI (Fig. 2) and its impact on the ASAS axSpA classification was assessed. Using the same cut-off values of 5 for the combination of fatty lesions and/or erosions, classification did not change in the large majority of the patients (93.5%). A similar result was seen at a cut-off value of 3 for erosions and fatty lesions only: classification did not change in 91.8 and 92.9% of the patients, respectively. The same patients would be additionally classified as axSpA as in scenario 1. But assuming that only an MRI was performed, five patients (1.7%) would not be classified axSpA anymore if radiographic sacroliliitis was replaced by structural lesions on MRI using the combination of fatty lesions and/or erosions. The SpA-features of the patients that are newly classified by the ASAS axSpA criteria are described in Fig. 3, and the SpA-features of the patients that are no longer classified by the ASAS axSpA criteria are described in Fig. 4. One of these patients had four axSpA features, leading to a positive LR product of 192.4 and a corresponding high disease probability of 91%. One other patient had a disease probability of 75%. The disease probabilities of the other three patients turned out to be much lower: between 14 and 32%. Three of these five patients were female and two of the five patients were HLA-B27 positive (one male, one female). Of the five patients who would not be classified as axSpA anymore in this scenario, three patients were diagnosed axSpA by the rheumatologist. Regarding the definition based on erosions only, one additional patient would not be classified as ASAS axSpA anymore. This female patient was HLA-B27 negative, and was not diagnosed axSpA by the treating rheumatologist. Regarding the combination of fatty lesions and/or erosions, 11 patients changed arms within the criteria under this scenario but all stayed ASAS axSpA positive. Similar results were found when using a cut-off for fatty lesions only and erosions only (both cut-off values of 3) (data not shown). Fig. 3 View largeDownload slide ASAS axSpA positive patients (mNY positive) becoming ASAS axSpA negative (MRI-SI-s negativity) in scenario 2 Five patients marked in blue in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 3 View largeDownload slide ASAS axSpA positive patients (mNY positive) becoming ASAS axSpA negative (MRI-SI-s negativity) in scenario 2 Five patients marked in blue in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 4 View largeDownload slide ASAS axSpA negative patients becoming ASAS axSpA positive (due to MRI-SI-s positive) in scenario 2 The patients marked in purple in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 4 View largeDownload slide ASAS axSpA negative patients becoming ASAS axSpA positive (due to MRI-SI-s positive) in scenario 2 The patients marked in purple in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Discussion In this cohort of patients with CBP of short duration we have shown that adding structural lesions to the imaging criterion of the ASAS axSpA criteria has a limited effect on the classification of patients. Also, when sacroiliitis on radiographs was replaced by structural lesions on MRI, only minor changes in the ASAS axSpA classification of patients were seen. In patients with CBP suspicious for axSpA, structural lesions are associated with other SpA features and the majority of patients with structural lesions in our study fulfil the ASAS axSpA criteria anyway. If conventional radiographs were replaced by MRI-SI-s, using a combination of fatty lesions and erosions, only 3/294 (1.0%) patients would be additionally classified as ASAS axSpA. On the other hand, only 5/294 (1.7%) patients would lose their ASAS axSpA classification. Three out of these five patients with a state change had relatively low ratio products and corresponding probabilities of axSpA, suggesting that they did not have classic disease presentations. However, the two other patients have high disease probabilities and therefore could be missed erroneously. Most changes that are seen are a change in a subcategory between the various ASAS axSpA criteria rather than a change in classification per se. In other words, patients do not lose their ASAS classification solely by changing the content of the criteria. This characteristic adds to the credibility of the ASAS criteria. Only a few patients lose ASAS axSpA classification in scenario 2, which may justify replacement of conventional radiographs by MRI. However, it is important to realize that this is purely data-driven and feasibility issues should not be overlooked. MRI is an expensive imaging technique, especially in certain areas of the world. Furthermore, rheumatologists and radiologists are worldwide familiar with the mNY criteria, and evaluation of structural lesions on an MRI of the SI joints is a new concept. Education and time is needed in order to become familiarized with it. At this point in time, we therefore favour the addition of structural lesions to the imaging criterion of the ASAS axSpA criteria above replacement. These results are in line with recent data from the DESIR cohort, investigating the same research question in another cohort. In the SPACE cohort, even more patients do not change classification while replacing radiographic sacroiliitis by structural lesions seen on MRI compared with DESIR (SPACE: 95.3% vs DESIR: 80%). In general, a notable difference between the two cohorts is that in DESIR only patients with IBP are included, whereas in SPACE 62.6% of the patients have IBP. Other SpA features are also more common in the DESIR cohort, among which HLA-B27 positivity and radiographic sacroiliitis on conventional radiography and inflammation on MRI are highlighted. This is reflected in the fact that in the DESIR cohort, 71.8% fulfil the ASAS axSpA classification criteria, compared with 35.0% in the SPACE cohort. In DESIR all patients included are of French origin, whereas the SPACE cohort recruits patients from three European countries (The Netherlands, Italy, Norway). Though there might not be a big disparity between the prevalence of axSpA and CBP in general between these countries, it is important to observe these results in two populations of a different origin. Reliability of structural lesions remains a difficult issue and in a research setting often two or more well-calibrated readers are involved, which may cause difficulties translating this to clinical practice. So we should also be informed on the agreement between evaluation of MRI-SI-s in clinical practice before this can be advocated for use in a clinical setting. A limitation of the study is the absence of a gold standard to assess structural changes in the SI joint, by means of CT. Although the agreement regarding the presence/absence of radiographic sacroiliitis and the presence/absence of structural lesions on MRI in this study is only moderate, it is slightly better compared with the similar study in the DESIR cohort [1]. In general, it is reassuring to see consistent findings in two independent cohorts with different (though well-trained) reader pairs. This increases confidence regarding the generalizability. Although the research question has now been investigated in two cohorts, both cohorts include patients with short-standing back pain complaints and it would be very interesting to see replication of these findings in cohorts with advanced disease before possible far-reaching conclusions can be drawn on potentially changing ASAS axSpA classification criteria. The ideal cut-off could potentially be different in patients with longer symptom duration and more structural lesions in the SI joints. In general, more data are warranted on the prevalence of structural lesions in advanced disease, and it is beyond the scope of this study how lesions develop over time. Although the focus of this study is the impact of structural lesions on the ASAS axSpA classification criteria, we could speculate about possible implications for the diagnostic process. The modified Berlin algorithm advises that all patients suspected of axSpA should have a plain radiograph of the pelvis to check for sacroiliitis as a first step [13]. In patients without evidence of radiological sacroiliitis in whom axSpA still is suspected, an MRI of the SI joints (assessed for inflammation only) may support a diagnosis of non-radiographic axial spondyloarthritis when inflammation is present. Our data suggest that there is no solid indication to change the strategy of first asking a pelvic radiograph, since the classification remains very similar when replacing pelvic X-rays by MRI-s. However, in young patients MRI can be obtained as an alternative to plain radiography. This is in line with the recently published EULAR recommendations [14]. Similarly, if an MRI (STIR and T1 sequence) is present in a clinical setting, but there is no pelvic radiograph, this MRI may suffice and there is no reason to obtain radiographs. A strength of the study is the intensive scoring process by two well-calibrated readers with an adjudication process in place, which adds to the reliability of our findings. Another strength of this study is the SPACE cohort itself. The SPACE cohort is one of early disease, and comprises a control group of (chronic) back pain patients, just as in daily practice where a distinction between axSpA and no-axSpA should be made in every patient presenting with a suspicion of axSpA. In conclusion, our study has confirmed the earlier promising finding that the assessment of structural lesions on MRI instead of or in addition to conventional radiographs does not lead to a different ASAS axSpA classification in most of these patients with symptoms of an early disease onset. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript. Disclosure statement: The authors have declared no conflicts of interest. References 1 Bakker PA , van den Berg R , Lenczner G et al. Can we use structural lesions seen on MRI of the sacroiliac joints reliably for the classification of patients according to the ASAS axial spondyloarthritis criteria? Data from the DESIR cohort . Ann Rheum Dis 2017 ; 76 : 392 – 8 . Google Scholar CrossRef Search ADS PubMed 2 van TA , Heuft-Dorenbosch L , Schulpen G et al. Radiographic assessment of sacroiliitis by radiologists and rheumatologists: does training improve quality? Ann Rheum Dis 2003 ; 62 : 519 – 25 . Google Scholar CrossRef Search ADS PubMed 3 van den Berg R , Lenczner G , Feydy A et al. Agreement between clinical practice and trained central reading in reading of sacroiliac joints on plain pelvic radiographs. Results from the DESIR cohort . Arthritis Rheumatol 2014 ; 66 : 2403 – 11 . Google Scholar CrossRef Search ADS PubMed 4 Rudwaleit M , van der Heijde D , Landewe R et al. The development of Assessment of SpondyloArthritis International Society classification criteria for axial spondyloarthritis (part II): validation and final selection . Ann Rheum Dis 2009 ; 68 : 777 – 83 . Google Scholar CrossRef Search ADS PubMed 5 Poddubnyy D , Gaydukova I , Hermann KG et al. Magnetic resonance imaging compared to conventional radiographs for detection of chronic structural changes in sacroiliac joints in axial spondyloarthritis . J Rheumatol 2013 ; 40 : 1557 – 65 . Google Scholar CrossRef Search ADS PubMed 6 van den Berg R , de HM , van GF et al. Percentage of patients with spondyloarthritis in patients referred because of chronic back pain and performance of classification criteria: experience from the Spondyloarthritis Caught Early (SPACE) cohort . Rheumatology (Oxford) 2013 ; 52 : 1492 – 9 . Google Scholar CrossRef Search ADS PubMed 7 Rudwaleit M , van der Heijde D , Landewe R et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general . Ann Rheum Dis 2011 ; 70 : 25 – 31 . Google Scholar CrossRef Search ADS PubMed 8 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 . Google Scholar CrossRef Search ADS PubMed 9 Rudwaleit M , Jurik AG , Hermann KG et al. Defining active sacroiliitis on magnetic resonance imaging (MRI) for classification of axial spondyloarthritis: a consensual approach by the ASAS/OMERACT MRI group . Ann Rheum Dis 2009 ; 68 : 1520 – 7 . Google Scholar CrossRef Search ADS PubMed 10 Weber U , Lambert RG , Østergaard M et al. The diagnostic utility of magnetic resonance imaging in spondylarthritis: an international multicenter evaluation of one hundred eighty-seven subjects . Arthritis Rheum 2010 ; 62 : 3048 – 58 . Google Scholar CrossRef Search ADS PubMed 11 de Hooge M , van den Berg R , Navarro-Compán V et al. Patients with chronic back pain of short duration from the SPACE cohort: which MRI structural lesions in the sacroiliac joints and inflammatory and structural lesions in the spine are most specific for axial spondyloarthritis? Ann Rheum Dis 2016 ; 75 : 1308 – 14 . Google Scholar CrossRef Search ADS PubMed 12 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 . Google Scholar CrossRef Search ADS PubMed 13 van den Berg R , de HM , Rudwaleit M et al. ASAS modification of the Berlin algorithm for diagnosing axial spondyloarthritis: results from the SPondyloArthritis Caught Early (SPACE)-cohort and from the Assessment of SpondyloArthritis international Society (ASAS)-cohort . Ann Rheum Dis 2013 ; 72 : 1646 – 53 . Google Scholar CrossRef Search ADS PubMed 14 Mandl P , Navarro-Compan V , Terslev L et al. EULAR recommendations for the use of imaging in the diagnosis and management of spondyloarthritis in clinical practice . Ann Rheum Dis 2015 ; 74 : 1327 – 39 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rheumatology Oxford University Press

Impact of replacing radiographic sacroiliitis by magnetic resonance imaging structural lesions on the classification of patients with axial spondyloarthritis

Loading next page...
 
/lp/ou_press/impact-of-replacing-radiographic-sacroiliitis-by-magnetic-resonance-i5m4vzc007
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1462-0324
eISSN
1462-0332
D.O.I.
10.1093/rheumatology/kex532
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To investigate in patients with chronic back pain of a short duration, the utility of adding structural MRI lesions of the SI joints to the imaging criterion of the Assessment of SpondyloArthritis International Society (ASAS) axial SpA (axSpA) criteria and the utility of replacement of radiographic sacroiliitis by structural MRI lesions. Methods MRI STIR (inflammation, MRI-SI), MRI T1-weighted images (structural lesions, MRI-SI-s) and radiographs of the SI joints of patients in the SPondyloArthritis Caught Early-cohort (chronic back pain: ⩾3 months, ⩽2 years; onset <45 years) were scored by two well-calibrated readers. Previously proposed cut-offs for a positive MRI-SI-s were used (based on <5% prevalence in no-SpA patients): erosions ⩾3, fatty lesions ⩾3, fatty lesions and/or erosions (erosions/fatty lesions) ⩾5. Using the definitions of MRI-SI-s, patients were classified according to the ASAS axSpA criteria. Results Twenty-nine of 294 patients were modified New York (mNY) positive and 32 were MRI-SI-s positive (erosions/fatty lesions ⩾5). Agreement between mNY and MRI-SI-s (erosions/fatty lesions ⩾5) was moderate (κ: 0.58). Using the erosions/fatty lesions ⩾5 cut-off, 3/294 additional patients were classified as axSpA (adding MRI). Using this cut-off instead of mNY (replacing mNY), classification did not change in 286 patients (97.3%), but 5 patients (1.7%) would not be classified as axSpA and 3 previously unclassified patients (1.0%) would be classified as axSpA. Similar results were seen for the other cut-offs (erosions ⩾3 and fatty lesions ⩾3). Conclusion Assessment of structural lesions (fatty lesions and erosions) on MRI-SI instead of or in addition to conventional radiographs does not lead to a different ASAS axSpA classification in most of the patients with early disease onset. This suggests that structural lesions (fatty lesions and erosions) can be reliably used in the ASAS axSpA classification of patients, as both addition and replacement of radiographs of the SI joints. axial spondyloarthritis, sacroiliitis, structural lesions, MRI, radiographs Rheumatology key messages Structural MRI lesions in SI joints can be reliably used to classify patients with axial SpA. Adding or replacing radiographic sacroiliitis by structural MRI lesions led to minor changes in axial SpA classification. Introduction Recently, we have demonstrated that in the DEvenir des Spondyloarthrites Indifférenciées Récentes (DESIR) cohort of patients with inflammatory back pain with a high suspicion of having axial SpA (axSpA), structural lesions on MRI can be used reliably either as an addition to or as a substitute for radiographs in the Assessment of SpondyloArthritis International Society (ASAS) axSpA classification of patients spondyloarthritis [1]. Since this is a new concept with potential consequences for daily clinical practice, replication in other cohorts is highly warranted. The current study investigates the same research question in an entirely independent cohort, the SPondyloArthritis Caught Early (SPACE) cohort of patients with chronic back pain of a short duration. The rationale behind using structural lesions on MRI instead of conventional radiographs is mainly based on the fact that a reliable detection of sacroiliitis on conventional radiographs is notoriously difficult. Sacroiliitis is the hallmark of AS and the presence of sacroiliitis on conventional radiography is obligatory in the modified New York (mNY) criteria for AS [2]. However, it has been shown that substantial observer variation (interreader, intrareader) exists for both radiologists and rheumatologists, and training does not lead to improvement [2]. Recent data from the DESIR cohort have also revealed a poor to moderate agreement between different readers leading to substantial misclassification of patients [3]. With the help of MRI it has become possible to detect active sacroiliitis (bone marrow oedema, BME) in early stages of axSpA. The ASAS definition for a positive MRI is solely based on identification of BME highly suggestive for SpA, but structural lesions can be seen [fatty lesions, erosions, sclerosis and (partial) ankylosis] are visible on MRI as well [4]. The 3D (albeit tomographic) character of MRI may be an advantage as compared with the 2D projection of conventional radiographs [5]. Another advantage of MRI above radiography is that patients are not exposed to ionizing radiation. Both the SPACE and the DESIR are cohorts of early disease. Since it often takes 6–8 years from the onset of symptoms before radiographic sacroiliitis can be detected on plain radiographs, conventional radiography is a suboptimal imaging technique especially in patients with complaints of a short duration (in addition to the earlier described reliability issues). The DESIR cohort only includes patients with inflammatory back pain (IBP) with up to 3 years of symptoms, whereas the SPACE cohort includes patients with chronic back pain (CBP) with up to 2 years of symptoms. Different reader pairs were involved in the scoring process of the two cohorts, to rule out a reader-dependent effect. In this study we aim to evaluate the usefulness of structural lesions with regard to the ASAS axSpA classification criteria in the SPACE cohort. First of all, we determine the agreement between sacroiliitis seen on radiographs (mNY) and structural lesions seen on MRI. Then we aim to evaluate the potential impact of adding structural lesions on MRI to the definition of a positive MRI or replacing radiographic sacroiliitis by MRI structural lesions on the ASAS axSpA classification of patients. Methods Study population For this analysis, baseline data from the SPACE cohort were used. An extensive description of the SPACE cohort has been given elsewhere [6]. In short, SPACE is an inception cohort with on-going inclusion and follow-up of patients with CBP of short duration (⩾3 months but ⩽2 years, with onset <45 years). Patients were recruited from five participating centres in the Netherlands (Leiden, Amsterdam, Gouda), Norway (Oslo) and Italy (Padua). Approval by the local medical ethics committees was obtained, as well as written informed consent from all patients in accordance with the Declaration of Helsinki. A full diagnostic work-up was performed in all patients including: HLA-B27 testing, conventional radiographs and MRI of the SI joints (X-SI and MRI-SI, respectively), and the assessment of all other SpA features, in agreement with the descriptions supplied by the ASAS group [7]. Imaging and scoring methods MRI-SI was performed on a 1.5 Tesla machine (Philips Medical Systems, Best, Netherlands). Coronal oblique T1-weighted TSE (TR 550/TE 10) and STIR (TR 2500/TE 60) with a slice thickness of 4 mm were the acquired sequences used. All available baseline MRI-SI and X-SI were read independently, in different reading sessions, by two trained, well-calibrated readers (P.A.B., M.d.H.). Readers were blinded to the score of the other reader and other modality as well as the clinical information, throughout the scoring process. The mNY criteria were used to assess radiographs of the SI joints; radiographic sacroiliitis was defined as bilateral grade ⩾2 or unilateral grade ⩾3 [8]. In consensus with the ASAS definition, an MRI-SI was defined positive if one BME lesion highly suggestive of SpA was present on two or more consecutive slices, or otherwise if several BME lesions highly suggestive of SpA were visible on a single slice [9]. A third reader served as adjudicator (R.vd.B.) in case of disagreement among the two initial readers regarding a positive MRI (ASAS definition) or the presence of sacroiliitis (mNY criteria). The presence of structural lesions, namely fatty lesions, erosions, sclerosis and (partial) ankylosis, was assessed on MRI T1-weighted images in conjunction with the STIR images. A scoring system with similarities to the methodology outlined in the Spondyloarthritis Research Consortium of Canada (SPARCC) online training module, as described by Weber et al. [10] was used. Weber et al. proposed a scoring system to quantify structural lesions called SPARCC SI structural lesion score. This method is founded on the assessment of lesions (present vs absent) counting them in each quadrant on six consecutive slices through the SI joints. The starting point is the slice on which at least 1 cm of vertical height of the cartilage compartment can be seen, from anterior to posterior, evaluating the cartilaginous compartment of the SI joints and the antero-inferior portion of the SI joint. Each SI joint is split into four quadrants. Structural lesions were taken into account only if present on at least two consecutive slices. This is reflected by a maximum score of 40 per lesion (5 lesions per quadrant × 4 quadrants × 2 SI joints) except for (partial) ankylosis. As an exception, (partial) ankylosis was considered sufficient when seen on a single slice reflected by a maximum score of 24 per patient. As shown earlier by our group, an MRI positive for structural lesions (MRI-SI-s) was determined by the use of different cut-offs [11]. These chosen cut-offs were anchored on ⩽5% false positives whereby the false positives were specified as structural lesions among patients not having axSpA according to the ASAS axSpA criteria. The described cut-offs have pointed out to be: erosions ⩾3, fatty lesions ⩾3, fatty lesions and/or erosions ⩾5. In this early cohort, the prevalence of sclerosis and (partial) ankylosis was so low that there was no cut-off that could clearly differentiate between SpA and no-SpA patients. As a consequence, these types of lesions were not further considered. Classification criteria Patients were classified according to the ASAS axSpA criteria. Subsequently, patients were grouped based on the way they met the criteria: through the imaging arm of the ASAS axSpA criteria alone (either by mNY criteria and/or by positive MRI); through the clinical arm of the ASAS axSpA criteria alone; or through both. If patients fulfilled more than one category, they were classified in that way, reflected by seven possible combinations (Figs 1 and 2). The no-axSpA group is made up of patients not fulfilling the ASAS axSpA criteria. Fig. 1 View largeDownload slide Addition of MRI-SI-s to the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 1 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 1 View largeDownload slide Addition of MRI-SI-s to the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 1 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 2 View largeDownload slide Replacement of mNY criteria by MRI-SI-s in the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 2 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 2 View largeDownload slide Replacement of mNY criteria by MRI-SI-s in the ASAS axSpA criteria (combination fatty lesions and/or erosions): scenario 2 axSpA: axial SpA; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Data analysis Descriptive statistics were used to calculate disease characteristics of the included patients. In the analysis, regarding structural lesions the mean score of the two readers in agreement of a positive MRI (ASAS definition) was used. In case of disagreement, the mean of the scores of the adjudicator and the reader in agreement with the adjudicator’s judgement regarding a positive MRI for that particular case were used. Agreement about the absence or presence of structural lesions using both imaging modalities (X-SI and MRI-SI-s) was assessed by cross-tabulation and expressed as Cohen’s κ. Percentage positive agreement was calculated in order to leave out patients labelled as negative by the two readers, which could lead to an artificially high agreement. Subsequently, using the various definitions of MRI-SI-s, patients were classified by the ASAS axSpA criteria. MRI-SI-s was added to the imaging criterion of the ASAS axSpA criteria in the first place, resulting in an additional possibility to fulfil the imaging arm (scenario 1). Second, replacement of the mNY criterion by MRI-SI-s (scenario 2) was applied as if only an MRI was performed. Disease probabilities were calculated by the positive likelihood ratio (LR) product, for those patients changing ASAS axSpA classification in scenario 2. This was done by multiplying the individual LRs of all identified SpA features [12]. In patients with CBP with an assumed disease prevalence of axSpA of 5%, an LR product of 200 results in a positive predictive value of 90%. These patients were further described by their clinical phenotype: gender, age and whether a patient was diagnosed as axSpA according to the treating rheumatologist. The analyses were performed in STATA 12.0 (StataCorp LP, College Station, Texas, USA). Results Included in the analysis are 294 patients with complete imaging data at baseline (both MRI-SI and X-SI present). Table 1 describes patient characteristics. Patients had a mean (s.d.) age of 31.6 years of age (10.7 years) and a mean (s.d.) duration of back pain of 13.1 months (7.3 months). Of these patients, 34.6% were men and 34.6% were HLA-B27 positive. Table 1 Disease characteristics of the included patients Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) ASAS: Assessment of SpondyloArthritis International Society; IBP: inflammatory back pain; SPACE: SPondyloArthritis Caught Early. Table 1 Disease characteristics of the included patients Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) Disease characteristics of the included patients Total number (n = 294) Age at inclusion, mean (s.d.), years 31.6 (10.7) Male, n (%) 102 (34.6) Symptom duration at first visit, mean (s.d.), months 13.1 (7.3) Good response to NSAIDs, n (%) 92 (31.6) IBP, n (%) 184 (62.6) Positive family history of SpA, n (%) 112 (38.1) Peripheral arthritis, n (%) 37 (12.6) Dactylitis, n (%) 10 (3.4) Enthesitis, n (%) 43 (14.6) Uveitis, n (%) 21 (7.1) IBD, n (%) 26 (8.8) Psoriasis, n (%) 29 (9.7) Elevated CRP, n (%) 54 (18.8) HLA-B27 positive, n (%) 100 (34.6) Sacroiliitis present on radiograph, n (%) 29 (9.8) Positive MRI (ASAS definition), n (%) 37 (12.5) ASAS: Assessment of SpondyloArthritis International Society; IBP: inflammatory back pain; SPACE: SPondyloArthritis Caught Early. One hundred and three out of 294 patients (35.0%) fulfilled the ASAS axSpA criteria using the standard definition. Of these 103 patients, 50 patients fulfilled the imaging arm (48.5%) and 53 patients fulfilled the clinical arm only (51.5%). The most prevalent SpA features were IBP (62.6%), a positive family history for SpA (38.1%) and a good response to NSAIDs (31.6%). Regarding the prevalence of structural lesions, 20/294 patients (6.8%) showed three or more fatty lesions; 16 of these 20 patients (80%) already formally fulfilled the ASAS axSpA criteria. Thirty-four patients (11.6%) had three or more erosions, and 27 of these 34 (79.4%) already formally fulfilled the ASAS axSpA criteria. Thirty-one patients (10.5%) had 5 or more fatty and/or erosive lesions (combination), of which 26 patients (83.9%) already formally fulfilled the ASAS axSpA criteria. The agreement regarding the presence/absence of radiographic sacroiliitis and the presence/absence of structural lesions on MRI was moderate (Table 2). Subtle differences were seen between the various definitions used: κ: 0.51 (erosions ⩾3); κ: 0.45 (fatty lesions ⩾3); and κ: 0.58 (combination of fatty lesions and erosions ⩾5). Table 2 Agreement between sacroiliitis on conventional radiographs (mNY criteria) and a positive MRI-SI based on structural lesions mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 Agreement based on MRI-SI structural lesions using the three different definitions. mNY: modified New York criteria; MRI-SI: MRI of the SI joints. Table 2 Agreement between sacroiliitis on conventional radiographs (mNY criteria) and a positive MRI-SI based on structural lesions mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 mNY (adjudicated) κ: 0.51 Erosions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 18 17 35 Negative 11 248 259 Total 29 265 294 mNY (adjudicated) κ: 0.45 Fatty lesions: cut-off ≥3; mean 2 out of 3 readers Positive Negative Total Positive 12 8 20 Negative 17 257 274 Total 29 265 294 mNY (adjudicated) κ: 0.58 Erosions/fatty lesions: cut-off ≥5; mean 2 out of 3 readers Positive Negative Total Positive 19 13 32 Negative 10 252 262 Total 29 265 294 Agreement based on MRI-SI structural lesions using the three different definitions. mNY: modified New York criteria; MRI-SI: MRI of the SI joints. Scenario 1 The addition of MRI-SI-s to the imaging criterion of the ASAS axSpA criteria was investigated first (Fig. 1). Classification did not change in the majority of the patients (96.3%) using a cut-off value of 5 for the combination of fatty lesions and/or erosions. If a cut-off value of 3 for erosions or fatty lesions only was used, comparable results were seen: classification did not change in 94.9 and 97.6% of the patients, respectively. Considering the combination of five fatty lesions and/or erosions, three patients would be classified additionally axSpA if structural lesions on MRI were taken into consideration. The positive LR products of the three additionally classified patients were 15.8, 5.1 and 2.5, corresponding to post-test probabilities of 44, 20 and 11%, respectively. The rheumatologist diagnosed only one of these three patients (all female and HLA-B27 negative) with axSpA. Regarding the definition based on erosions, five patients would be additionally classified (all female and HLA-B27 negative). Only two of them were diagnosed as axSpA by the rheumatologist. Regarding the definition based on fatty lesions, three patients would be additionally classified as axSpA (one was described above, the other two were male patients, HLA-B27 negative, of whom one was diagnosed as having axSpA by the rheumatologist). Some patients changed subgroups within the ASAS classification, without a change in ASAS axSpA positivity or negativity. Eight patients would be classified via different arms due to the presence of structural lesions, using the combination of fatty lesions and/or erosions. Five patients also showed structural lesions on MRI (Fig. 1) but were already classified via the imaging arm based on inflammatory lesions on MRI. Three other patients fulfilled the clinical arm only, but fulfilled the imaging arm as well based on a positive MRI-SI-s. Using the cut-offs for fatty lesions or erosions only (both cut-off values were 3), the same trends were seen (data not shown). Scenario 2 Second, the replacement of radiographic sacroillitis by structural lesions on MRI (Fig. 2) and its impact on the ASAS axSpA classification was assessed. Using the same cut-off values of 5 for the combination of fatty lesions and/or erosions, classification did not change in the large majority of the patients (93.5%). A similar result was seen at a cut-off value of 3 for erosions and fatty lesions only: classification did not change in 91.8 and 92.9% of the patients, respectively. The same patients would be additionally classified as axSpA as in scenario 1. But assuming that only an MRI was performed, five patients (1.7%) would not be classified axSpA anymore if radiographic sacroliliitis was replaced by structural lesions on MRI using the combination of fatty lesions and/or erosions. The SpA-features of the patients that are newly classified by the ASAS axSpA criteria are described in Fig. 3, and the SpA-features of the patients that are no longer classified by the ASAS axSpA criteria are described in Fig. 4. One of these patients had four axSpA features, leading to a positive LR product of 192.4 and a corresponding high disease probability of 91%. One other patient had a disease probability of 75%. The disease probabilities of the other three patients turned out to be much lower: between 14 and 32%. Three of these five patients were female and two of the five patients were HLA-B27 positive (one male, one female). Of the five patients who would not be classified as axSpA anymore in this scenario, three patients were diagnosed axSpA by the rheumatologist. Regarding the definition based on erosions only, one additional patient would not be classified as ASAS axSpA anymore. This female patient was HLA-B27 negative, and was not diagnosed axSpA by the treating rheumatologist. Regarding the combination of fatty lesions and/or erosions, 11 patients changed arms within the criteria under this scenario but all stayed ASAS axSpA positive. Similar results were found when using a cut-off for fatty lesions only and erosions only (both cut-off values of 3) (data not shown). Fig. 3 View largeDownload slide ASAS axSpA positive patients (mNY positive) becoming ASAS axSpA negative (MRI-SI-s negativity) in scenario 2 Five patients marked in blue in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 3 View largeDownload slide ASAS axSpA positive patients (mNY positive) becoming ASAS axSpA negative (MRI-SI-s negativity) in scenario 2 Five patients marked in blue in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 4 View largeDownload slide ASAS axSpA negative patients becoming ASAS axSpA positive (due to MRI-SI-s positive) in scenario 2 The patients marked in purple in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Fig. 4 View largeDownload slide ASAS axSpA negative patients becoming ASAS axSpA positive (due to MRI-SI-s positive) in scenario 2 The patients marked in purple in Fig. 2. aProbability of axSpA is based on the positive likelihood ratio product [13]. IBP: inflammatory back pain; ASAS: Assessment of SpondyloArthritis International Society; mNY: modified New York criteria; MRI-SI-s: MRI of the SI joints assessed for structural lesions. Discussion In this cohort of patients with CBP of short duration we have shown that adding structural lesions to the imaging criterion of the ASAS axSpA criteria has a limited effect on the classification of patients. Also, when sacroiliitis on radiographs was replaced by structural lesions on MRI, only minor changes in the ASAS axSpA classification of patients were seen. In patients with CBP suspicious for axSpA, structural lesions are associated with other SpA features and the majority of patients with structural lesions in our study fulfil the ASAS axSpA criteria anyway. If conventional radiographs were replaced by MRI-SI-s, using a combination of fatty lesions and erosions, only 3/294 (1.0%) patients would be additionally classified as ASAS axSpA. On the other hand, only 5/294 (1.7%) patients would lose their ASAS axSpA classification. Three out of these five patients with a state change had relatively low ratio products and corresponding probabilities of axSpA, suggesting that they did not have classic disease presentations. However, the two other patients have high disease probabilities and therefore could be missed erroneously. Most changes that are seen are a change in a subcategory between the various ASAS axSpA criteria rather than a change in classification per se. In other words, patients do not lose their ASAS classification solely by changing the content of the criteria. This characteristic adds to the credibility of the ASAS criteria. Only a few patients lose ASAS axSpA classification in scenario 2, which may justify replacement of conventional radiographs by MRI. However, it is important to realize that this is purely data-driven and feasibility issues should not be overlooked. MRI is an expensive imaging technique, especially in certain areas of the world. Furthermore, rheumatologists and radiologists are worldwide familiar with the mNY criteria, and evaluation of structural lesions on an MRI of the SI joints is a new concept. Education and time is needed in order to become familiarized with it. At this point in time, we therefore favour the addition of structural lesions to the imaging criterion of the ASAS axSpA criteria above replacement. These results are in line with recent data from the DESIR cohort, investigating the same research question in another cohort. In the SPACE cohort, even more patients do not change classification while replacing radiographic sacroiliitis by structural lesions seen on MRI compared with DESIR (SPACE: 95.3% vs DESIR: 80%). In general, a notable difference between the two cohorts is that in DESIR only patients with IBP are included, whereas in SPACE 62.6% of the patients have IBP. Other SpA features are also more common in the DESIR cohort, among which HLA-B27 positivity and radiographic sacroiliitis on conventional radiography and inflammation on MRI are highlighted. This is reflected in the fact that in the DESIR cohort, 71.8% fulfil the ASAS axSpA classification criteria, compared with 35.0% in the SPACE cohort. In DESIR all patients included are of French origin, whereas the SPACE cohort recruits patients from three European countries (The Netherlands, Italy, Norway). Though there might not be a big disparity between the prevalence of axSpA and CBP in general between these countries, it is important to observe these results in two populations of a different origin. Reliability of structural lesions remains a difficult issue and in a research setting often two or more well-calibrated readers are involved, which may cause difficulties translating this to clinical practice. So we should also be informed on the agreement between evaluation of MRI-SI-s in clinical practice before this can be advocated for use in a clinical setting. A limitation of the study is the absence of a gold standard to assess structural changes in the SI joint, by means of CT. Although the agreement regarding the presence/absence of radiographic sacroiliitis and the presence/absence of structural lesions on MRI in this study is only moderate, it is slightly better compared with the similar study in the DESIR cohort [1]. In general, it is reassuring to see consistent findings in two independent cohorts with different (though well-trained) reader pairs. This increases confidence regarding the generalizability. Although the research question has now been investigated in two cohorts, both cohorts include patients with short-standing back pain complaints and it would be very interesting to see replication of these findings in cohorts with advanced disease before possible far-reaching conclusions can be drawn on potentially changing ASAS axSpA classification criteria. The ideal cut-off could potentially be different in patients with longer symptom duration and more structural lesions in the SI joints. In general, more data are warranted on the prevalence of structural lesions in advanced disease, and it is beyond the scope of this study how lesions develop over time. Although the focus of this study is the impact of structural lesions on the ASAS axSpA classification criteria, we could speculate about possible implications for the diagnostic process. The modified Berlin algorithm advises that all patients suspected of axSpA should have a plain radiograph of the pelvis to check for sacroiliitis as a first step [13]. In patients without evidence of radiological sacroiliitis in whom axSpA still is suspected, an MRI of the SI joints (assessed for inflammation only) may support a diagnosis of non-radiographic axial spondyloarthritis when inflammation is present. Our data suggest that there is no solid indication to change the strategy of first asking a pelvic radiograph, since the classification remains very similar when replacing pelvic X-rays by MRI-s. However, in young patients MRI can be obtained as an alternative to plain radiography. This is in line with the recently published EULAR recommendations [14]. Similarly, if an MRI (STIR and T1 sequence) is present in a clinical setting, but there is no pelvic radiograph, this MRI may suffice and there is no reason to obtain radiographs. A strength of the study is the intensive scoring process by two well-calibrated readers with an adjudication process in place, which adds to the reliability of our findings. Another strength of this study is the SPACE cohort itself. The SPACE cohort is one of early disease, and comprises a control group of (chronic) back pain patients, just as in daily practice where a distinction between axSpA and no-axSpA should be made in every patient presenting with a suspicion of axSpA. In conclusion, our study has confirmed the earlier promising finding that the assessment of structural lesions on MRI instead of or in addition to conventional radiographs does not lead to a different ASAS axSpA classification in most of these patients with symptoms of an early disease onset. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript. Disclosure statement: The authors have declared no conflicts of interest. References 1 Bakker PA , van den Berg R , Lenczner G et al. Can we use structural lesions seen on MRI of the sacroiliac joints reliably for the classification of patients according to the ASAS axial spondyloarthritis criteria? Data from the DESIR cohort . Ann Rheum Dis 2017 ; 76 : 392 – 8 . Google Scholar CrossRef Search ADS PubMed 2 van TA , Heuft-Dorenbosch L , Schulpen G et al. Radiographic assessment of sacroiliitis by radiologists and rheumatologists: does training improve quality? Ann Rheum Dis 2003 ; 62 : 519 – 25 . Google Scholar CrossRef Search ADS PubMed 3 van den Berg R , Lenczner G , Feydy A et al. Agreement between clinical practice and trained central reading in reading of sacroiliac joints on plain pelvic radiographs. Results from the DESIR cohort . Arthritis Rheumatol 2014 ; 66 : 2403 – 11 . Google Scholar CrossRef Search ADS PubMed 4 Rudwaleit M , van der Heijde D , Landewe R et al. The development of Assessment of SpondyloArthritis International Society classification criteria for axial spondyloarthritis (part II): validation and final selection . Ann Rheum Dis 2009 ; 68 : 777 – 83 . Google Scholar CrossRef Search ADS PubMed 5 Poddubnyy D , Gaydukova I , Hermann KG et al. Magnetic resonance imaging compared to conventional radiographs for detection of chronic structural changes in sacroiliac joints in axial spondyloarthritis . J Rheumatol 2013 ; 40 : 1557 – 65 . Google Scholar CrossRef Search ADS PubMed 6 van den Berg R , de HM , van GF et al. Percentage of patients with spondyloarthritis in patients referred because of chronic back pain and performance of classification criteria: experience from the Spondyloarthritis Caught Early (SPACE) cohort . Rheumatology (Oxford) 2013 ; 52 : 1492 – 9 . Google Scholar CrossRef Search ADS PubMed 7 Rudwaleit M , van der Heijde D , Landewe R et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general . Ann Rheum Dis 2011 ; 70 : 25 – 31 . Google Scholar CrossRef Search ADS PubMed 8 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 . Google Scholar CrossRef Search ADS PubMed 9 Rudwaleit M , Jurik AG , Hermann KG et al. Defining active sacroiliitis on magnetic resonance imaging (MRI) for classification of axial spondyloarthritis: a consensual approach by the ASAS/OMERACT MRI group . Ann Rheum Dis 2009 ; 68 : 1520 – 7 . Google Scholar CrossRef Search ADS PubMed 10 Weber U , Lambert RG , Østergaard M et al. The diagnostic utility of magnetic resonance imaging in spondylarthritis: an international multicenter evaluation of one hundred eighty-seven subjects . Arthritis Rheum 2010 ; 62 : 3048 – 58 . Google Scholar CrossRef Search ADS PubMed 11 de Hooge M , van den Berg R , Navarro-Compán V et al. Patients with chronic back pain of short duration from the SPACE cohort: which MRI structural lesions in the sacroiliac joints and inflammatory and structural lesions in the spine are most specific for axial spondyloarthritis? Ann Rheum Dis 2016 ; 75 : 1308 – 14 . Google Scholar CrossRef Search ADS PubMed 12 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 . Google Scholar CrossRef Search ADS PubMed 13 van den Berg R , de HM , Rudwaleit M et al. ASAS modification of the Berlin algorithm for diagnosing axial spondyloarthritis: results from the SPondyloArthritis Caught Early (SPACE)-cohort and from the Assessment of SpondyloArthritis international Society (ASAS)-cohort . Ann Rheum Dis 2013 ; 72 : 1646 – 53 . Google Scholar CrossRef Search ADS PubMed 14 Mandl P , Navarro-Compan V , Terslev L et al. EULAR recommendations for the use of imaging in the diagnosis and management of spondyloarthritis in clinical practice . Ann Rheum Dis 2015 ; 74 : 1327 – 39 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

RheumatologyOxford University Press

Published: Mar 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off