Imaging of the sacroiliac joints is important for diagnosing early axial spondyloarthritis but not all-decisive

Imaging of the sacroiliac joints is important for diagnosing early axial spondyloarthritis but... Abstract Objectives To evaluate the contribution of the results of sacroiliac imaging to diagnosis and to the level of confidence in diagnosis in patients presenting with chronic back pain (CBP) and suspected of having axial spondyloarthritis (axSpA). Methods. Data from 513 patients from the SPondyloArthritisCaughtEarly cohort with CBP (⩾3 months, ⩽2 years, onset <45 years) were analysed after full diagnostic work-up. Rheumatologists were asked not only to provide a diagnosis before and after the imaging results had been provided to them, but also to provide the level of confidence of this diagnosis on an 11-point numerical scale. Results. Before imaging, 317/513 patients were diagnosed with axSpA. Of these patients, 178/317 (56%) did not have signs of sacroiliitis on either MRI or radiography, which led to the rheumatologist refuting the initial diagnosis of axSpA in 81/178 (46%) patients. Of the 196/513 patients without axSpA before imaging, 35/196 (18%) had signs of sacroiliitis on imaging. Subsequently, 28/35 (80%) patients were diagnosed with axSpA. Overall, imaging was incongruent with the diagnosis before imaging in 213 patients. This led to a change in diagnosis in 109/213 (51%), which corresponds to 21% (109/513) of all patients in the cohort. In general, diagnostic confidence increased by having imaging results available (from 6.2 to 7.4, P < 0.001). Conclusion In patients with CBP suspected of having axSpA, sacroiliac imaging adds to the confidence in the final diagnosis. However, the number of changes in diagnosis suggests that imaging is important but not all-decisive in early axSpA diagnosis. axial spondyloarthritis, ankylosing spondylitis, magnetic resonance imaging, clinical diagnosis Rheumatology key messages Sacroiliac imaging is widely used in the diagnostic work-up of patients suspected of axial SpA. Sacroliac imaging results increase rheumatologists’ confidence in their final diagnosis. Modest changes in numbers of axial SpA following imaging suggest imaging is important but not crucial. Introduction The diagnosis of axial spondyloarthritis (axSpA) in patients presenting with chronic back pain (CBP) is a known challenge in clinical practice as there is a broad spectrum in clinical presentations [1]. Rheumatologists may use information acquired from several sources such as a patient’s medical history, physical examination, laboratory tests and imaging (radiography and/or magnetic resonance imaging of sacroiliac joints) to make a diagnosis [2–4]. Several decades ago, conventional radiography was introduced as an imaging tool to detect sacroiliitis [5]. However, pelvic radiographs can only detect structural changes such as erosions and sclerosis. Furthermore, in early axSpA radiography of the sacroiliac joints may not show structural abnormalities, which can remain absent for many years after disease onset [6]. In recent years, MRI has become an important instrument in the visualization of both inflammation and structural damage in CBP patients especially without radiographic sacroiliitis [7–11]. Imaging of the sacroiliac joints plays a pivotal role in the early recognition of axSpA [12, 13]. We have recently shown that in CBP patients suspected of axSpA, positive imaging of the sacroiliac joints is highly associated with a diagnosis of axSpA (odds ratio = 34.3; 95% CI: 17.3, 67.7) [14]. Unfortunately, recognition of sacroiliitis on radiographs can be difficult. For example, it has been shown that agreement on radiographic sacroiliitis between readers is low and that this does not improve after training [15]. Moreover, bone marrow oedema on MRI suggestive of axSpA may not always be due to axSpA [16]. Therefore, some axSpA experts have expressed their concerns about relying solely on imaging, which may lead to incorrect diagnoses of axSpA. In turn, this may lead to unnecessary exposure to anti-inflammatory drugs with potentially severe side effects [17, 18]. Given these controversies, surprisingly little is known about how rheumatologists actually integrate sacroiliac imaging results in their diagnostic work-up of CBP patients suspected of having axSpA. Therefore, the main objectives of this study were to study the contribution of sacroiliac imaging to the rheumatologist’s diagnosis, and to quantify the contribution of sacroiliac imaging to diagnostic certainty. Methods Study design and population The data used for the current study were obtained from the SPondyloArthritis Caught Early (SPACE) cohort. The SPACE cohort is a prospective multicentre study, which was initiated in January 2009. A detailed description of the study design has been published previously [19]. Briefly, patients with CBP (⩾3 months and ⩽2 years) of unknown origin and age of onset <45 years were recruited and included from multiple European rheumatology centres in the Netherlands, Norway, Italy and Sweden. The clinical database used for the current study was locked on 11 January 2017. Approval for the study was obtained from the local medical ethical committee of the Leiden University Medical Center (reference number: P08.105); patients provided written informed consent before participation. Clinical assessments and measurements All patients were examined according to a standardized full work-up including the assessment of the presence and history of clinical SpA features according to the Assessment of Spondyloarthritis international Society (ASAS) definitions [3]: CRP/ESR, HLA-B27, inflammatory back pain, good response to NSAIDs, positive family history of SpA, peripheral arthritis, dactylitis, heel enthesitis, acute anterior uveitis, IBD and psoriasis. Imaging of the sacroiliac joints Plain radiographs of the pelvis (X-SI) were performed in anteroposterior view. MRI of sacroiliac joints (MRI-SI) was performed in coronal oblique T1-weighted turbo spin echo and short tau inversion recovery with a slice thickness of 4 mm. Interpretation of the radiographs and MRI of the sacroiliac joints (sacroiliitis yes/no) was done by each participating centre as part of routine clinical practice (local reading). Patients were classified according to the ASAS classification criteria for which data from central reading in the SPACE cohort was used [20]. Diagnosis During the diagnostic work-up, rheumatologists were asked to provide a diagnosis twice (axSpA, yes/no): based only on information available after medical history, physical examination and laboratory testing, but before taking sacroiliac imaging into account (diagnosis before imaging), and based on all previously collected information but after taking sacroiliac imaging into account (diagnosis after imaging). In case of no axSpA, rheumatologists were asked to provide the most likely alternative diagnosis. In addition, rheumatologists were requested to provide a level of confidence (LoC) regarding their diagnosis (axSpA or no axSpA) on an 11-point numerical rating scale ranging from 0 (not confident at all) to 10 (very confident). Data analysis For the present study baseline patients with complete information on sacroiliac imaging (both MRI and radiography) and diagnosis were analysed. Descriptive statistics were used to define patient characteristics for the total patient group, for each diagnosis subgroup (axSpA vs no axSpA), and for each imaging subgroup (i.e. sacroiliitis positive and negative) as means (s.d.) for continuous variables or frequencies (%) for categorical variables. Chi-square and unpaired t tests were used to compare variables between groups. The paired t test was used to compare the level of confidence regarding diagnosis before and after imaging within each diagnosis subgroup according to their imaging status. Total number of SpA features was calculated including HLA-B27 status, but without taking imaging of the sacroiliac joints into account. Any positive imaging was defined as one of a negative X-SI but positive MRI-SI (X-SI−/MRI-SI+), a positive X-SI but negative MRI-SI (X-SI+/MRI-SI−) or sacroiliitis on both modalities (X-SI+/MRI-SI+). The rheumatologist’s diagnosis and the LoC regarding diagnosis were the main outcomes. Data analysis was performed using Stata SE v.14 software (StataCorp LP, College Station, TX, USA). Results Baseline data of 583 CBP patients were available. A total of 70 (12%) patients were excluded because of incomplete or missing data regarding diagnosis or sacroiliac imaging. Baseline characteristics of these patients were similar to the remaining 513 patients with complete data (data not shown). Of these 513 CBP patients suspected of having axSpA, 188 (37%) patients were male, mean age (s.d.) was 31.0 (8.2) years, mean symptom duration was 13.3 (7.0) months and 210 (41%) patients were HLA-B27 positive (Table 1). Table 1 Baseline characteristics of CBP patients suspected of axSpA in the SPACE cohort (n = 513) Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Values are listed as n (%), unless otherwise stated. a n = 510 patients for HLA-B27. b n = 499 patients for good response to NSAIDs. axSpA: axial spondyloarthritis; IBP: inflammatory back pain. Table 1 Baseline characteristics of CBP patients suspected of axSpA in the SPACE cohort (n = 513) Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Values are listed as n (%), unless otherwise stated. a n = 510 patients for HLA-B27. b n = 499 patients for good response to NSAIDs. axSpA: axial spondyloarthritis; IBP: inflammatory back pain. In all patients, rheumatologists provided a diagnosis before and after taking sacroiliac imaging into account (Fig. 1). Before imaging, 317 (62%) patients were diagnosed with axSpA (Table 2). Most common diagnoses in the 196 patients without axSpA were non-specific back pain, degenerative disc disease and mechanical back pain (data not shown). Diagnostic confidence was moderate in patients with and without an axSpA diagnosis before imaging [mean (s.d.) LoC axSpA 6.6 (1.9) and mean LoC no axSpA 5.6 (2.0), respectively]. Patients diagnosed with axSpA before imaging were more often male (41% vs 29%) and were more often HLA-B27 positive (53% vs 22%) compared with patients who were not diagnosed with axSpA before imaging. As expected, the mean (s.d.) number of SpA features was twice as high for axSpA patients compared with the patients not diagnosed with axSpA [3.1 (1.7) and 1.5 (1.0), respectively]. Table 2 Clinical features of CBP patients stratified for diagnosis before taking sacroiliac imaging results into account Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Values are listed as n (%), unless otherwise stated. LoC, level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident). a n = 315 for patients with axSpA diagnosis before imaging. b Past or present condition, either confirmed or diagnosed by a physician. c Back pain not present any more or is much better 24–48 h after a full dose of NSAID. d n = 310 for patients with axSpA diagnosis before imaging, e n = 189 for patients without axSpA diagnosis before imaging. f Presence in first- or second-degree relatives of any of the following: AS, acute anterior uveitis, ReA, psoriasis or IBD. axSpA: axial spondyloarthritis; IBP, inflammatory back pain. Table 2 Clinical features of CBP patients stratified for diagnosis before taking sacroiliac imaging results into account Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Values are listed as n (%), unless otherwise stated. LoC, level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident). a n = 315 for patients with axSpA diagnosis before imaging. b Past or present condition, either confirmed or diagnosed by a physician. c Back pain not present any more or is much better 24–48 h after a full dose of NSAID. d n = 310 for patients with axSpA diagnosis before imaging, e n = 189 for patients without axSpA diagnosis before imaging. f Presence in first- or second-degree relatives of any of the following: AS, acute anterior uveitis, ReA, psoriasis or IBD. axSpA: axial spondyloarthritis; IBP, inflammatory back pain. Fig. 1 View largeDownload slide Rheumatologist’s diagnosis before and after considering imaging of sacroiliac joints (n = 513) Any positive imaging defined as sacroiliitis on MRI and/or radiographic sacroiliitis; imaging negative defined as no abnormalities on MRI and radiographs of sacroiliac joints. Boxes in bold represent patients with a change in diagnosis due to imaging results discordant with the diagnosis before imaging. axSpA: axial spondyloarthritis. Fig. 1 View largeDownload slide Rheumatologist’s diagnosis before and after considering imaging of sacroiliac joints (n = 513) Any positive imaging defined as sacroiliitis on MRI and/or radiographic sacroiliitis; imaging negative defined as no abnormalities on MRI and radiographs of sacroiliac joints. Boxes in bold represent patients with a change in diagnosis due to imaging results discordant with the diagnosis before imaging. axSpA: axial spondyloarthritis. In total, 317/513 (62%) of CBP patients were initially diagnosed with axSpA, and this figure decreased to 269/513 (52%) patients with a final diagnosis of axSpA after imaging. Of these 269 axSpA patients, 55% were male, 59% were HLA-B27 positive, 62% had positive imaging and the mean number of SpA features was 3.2 (1.7). A total of 172/269 (64%) patients fulfilled the ASAS classification criteria [52% male, 87% HLA-B27 positive and mean number of SpA features 3.2 (1.5)]. Overall, the mean diagnostic confidence increased by having imaging results available (from 6.2 to 7.4, P < 0.001). This increase in diagnostic confidence was observed in both axSpA patients [from 7.1 (1.7) to 7.7 (2.1), P < 0.001] and non-axSpA patients [from 5.7 (2.0) to 7.5 (2.3), P < 0.001]. Of the 317 patients who were diagnosed with axSpA before imaging, 139 (44%) had positive imaging (Table 3). In these 139 patients, sacroiliitis was seen in 85 (61%) patients only on MRI-SI, in 10 (7%) patients only on X-SI and in 44 (32%) patients on both modalities (X-SI+/MRI-SI+). After imaging, the axSpA diagnosis was maintained in all of these 139 patients and the mean LoC in the diagnosis of axSpA increased significantly [from 7.4 (1.8) to 8.6 (1.7), P < 0.001]. Table 3 Diagnosis and imaging status after sacroiliac imaging in patients with axSpA diagnosis before sacroiliac imaging   AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)    AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)  Values are listed as n (%), unless otherwise stated. a Mean difference 1.2, 95% CI: 0.9, 1.6, P < 0.001; b mean difference −0.3, 95% CI: −0.5, 0.01, P = 0.06; c Mean difference 1.3, 95% CI: 0.7, 1.8, P < 0.001. d Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. e n = 137 for patients with axSpA diagnosis before and after imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Table 3 Diagnosis and imaging status after sacroiliac imaging in patients with axSpA diagnosis before sacroiliac imaging   AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)    AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)  Values are listed as n (%), unless otherwise stated. a Mean difference 1.2, 95% CI: 0.9, 1.6, P < 0.001; b mean difference −0.3, 95% CI: −0.5, 0.01, P = 0.06; c Mean difference 1.3, 95% CI: 0.7, 1.8, P < 0.001. d Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. e n = 137 for patients with axSpA diagnosis before and after imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Of the 317 patients who were diagnosed with axSpA before imaging, 178 (56%) had negative imaging. In 97/178 patients the diagnosis axSpA was maintained. In these patients, the LoC in the diagnosis somewhat decreased [from 6.7 (1.6) to 6.4 (2.1), P = 0.06]. In 81/178 patients with negative imaging the diagnosis was changed to no axSpA after imaging. In these patients the LoC increased significantly after imaging [from 5.1 (1.7) to 6.4 (2.0), P < 0.001]. By comparison, the 97/178 patients diagnosed with axSpA after imaging had a higher number of SpA features (excluding imaging and HLA-B27) than the 81/178 patients without axSpA [mean 3.5 (1.6) vs 2.2 (1.2), P < 0.001]. Moreover, these 97 axSpA patients were also more often HLA-B27 positive than the 81 patients without axSpA (53% vs 33%, P = 0.010). Of the 196 patients not diagnosed with axSpA before imaging, 35 (18%) had positive imaging (Table 4). In these 35 patients, sacroiliitis was seen in 24 (68%) patients only on MRI-SI, in 2 (6%) patients only on X-SI and in 9 (26%) patients on both modalities (X-SI+/MRI-SI+). In 28 of these 35 patients (80%) with sacroiliitis, the diagnosis was changed to axSpA [18/28 (64%) patients only MRI-SI+, 1/28 (4%) only X-SI+ and 9/28 (32%) MRI-SI+/X-SI+]. The mean LoC in diagnosis increased significantly following imaging [from 4.7 (2.0) to 7.6 (2.3), P < 0.001]. In 7 of the 35 patients with positive imaging (20%), the diagnosis no axSpA remained unchanged [6/7 (86%) patients were MRI-SI+ and one (14%) patient was X-SI+]. In these seven patients, the mean LoC in diagnosis remained the same after imaging [LoC from 5.0 (1.3) to 5.0 (2.3)]. By comparison, the 28/35 patients diagnosed with axSpA after imaging had a higher number of SpA features than the 7/35 patients without axSpA [mean 1.6 (0.9) vs 1.0 (0.8), P = 0.10]. Moreover, these 28 patients diagnosed with axSpA were also more often HLA-B27 positive than the seven patients without axSpA (48% vs 29%, P = 0.35). Table 4 Diagnosis and imaging status after sacroiliac imaging in patients without axSpA diagnosis before sacroiliac imaging   No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)    No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)  Values are listed as n (%), unless otherwise stated. a Mean difference 2.9, 95% CI: 1.6, 4.2, P < 0.001; b mean difference 0; c mean difference 1.4, 95% CI: −3.1, 5.9, P = 0.43; d mean difference 1.8, 95% CI: 1.5, 2.1, P < 0.001. e Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Table 4 Diagnosis and imaging status after sacroiliac imaging in patients without axSpA diagnosis before sacroiliac imaging   No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)    No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)  Values are listed as n (%), unless otherwise stated. a Mean difference 2.9, 95% CI: 1.6, 4.2, P < 0.001; b mean difference 0; c mean difference 1.4, 95% CI: −3.1, 5.9, P = 0.43; d mean difference 1.8, 95% CI: 1.5, 2.1, P < 0.001. e Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Of the 196 patients not diagnosed with axSpA before imaging, 161 (82%) had negative imaging. Despite having negative imaging, the diagnosis changed to axSpA after imaging in 5/196 patients (3%). In these five patients, LoC in diagnosis increased [from 5.0 (1.9) to 6.4 (1.9), P = 0.43]. In 156/161 (97%) patients with negative imaging, the diagnosis no axSpA remained unchanged. The LoC in diagnosis increased after imaging [from 5.8 (2.0) to 7.6 (2.2), P < 0.001]. By comparison, the 5/161 patients diagnosed with axSpA after imaging had a higher number of SpA features than the 156/161 patients without axSpA [mean 1.8 (0.8) vs 1.4 (1.0), P = 0.43]. Moreover, these 5 axSpA patients were also more often HLA-B27 positive than the 156 patients without axSpA (40% vs 16%, P = 0.16). In the entire cohort, the diagnosis changed in 21% (109/513) of the CBP patients following imaging. In patients with imaging results that were discordant with their primary diagnosis [n = 213 (178 + 35)], the diagnosis changed in 109 (51%) patients (Fig. 1, boxes in bold). This change of diagnosis more often pertained to patients in whom a clinical diagnosis was suspected but imaging was negative [81/109 (74%)] than in patients in whom a clinical diagnosis of axSpA was not suspected but imaging was positive [28/109 (26%)]. Discussion This study was performed to investigate the contribution of sacroiliac imaging to the rheumatologist’s diagnosis and its respective confidence. We have shown that the rheumatologist’s confidence in the diagnosis of patients with CBP suspected of having axSpA increases after the results of sacroiliac imaging are taken into account. However, the number of changes in diagnosis after imaging implies that imaging is indeed an important but not the all-decisive factor in axSpA diagnosis. Prior to imaging the physician’s confidence in diagnosis was already moderate to high in most patients, which corroborates the value of medical history taking, physical examination and laboratory tests in the diagnostic work-up of axSpA. Congruent imaging results significantly increased the diagnostic confidence of rheumatologists except in axSpA patients (before and after imaging) without sacroiliitis. In general, these findings are in line with results from an international survey conducted among rheumatologists throughout the world illustrating the value rheumatologists place on imaging in the diagnostic work-up of patients suspected of axSpA [21]. Nevertheless, there were patients diagnosed as no axSpA before imaging in which a positive imaging result did not influence the final diagnosis. A few patients were not diagnosed with axSpA after imaging, even though they had sacroiliitis (n = 6 for MRI-SI and n = 1 for X-SI). These patients had only a few other SpA features and rheumatologists remained uncertain about the diagnosis (mean LoC 5.0). Apparently, the diagnosing rheumatologist considered the observed lesions—in combination with the clinical presentation—not sufficiently specific for an axSpA diagnosis, which is an indication that rheumatologists do not necessarily find that imaging is the dominant feature in establishing a diagnosis of axSpA. An even smaller number of patients (n = 5) without a diagnosis of axSpA before imaging and without signs of sacroiliitis were still diagnosed with axSpA after imaging. Since these five patients have a low number of SpA features, it is difficult to understand these diagnoses. Moreover, our findings also stress the importance of imaging in rejecting an axSpA diagnosis as in 81/178 (46%) patients with axSpA before imaging, the diagnosis was dismissed when imaging turned out to be negative. This study has several limitations. First, rheumatologists were asked to provide CBP patients with a diagnosis before and after taking imaging into account. We cannot, however, rule out the possibility that rheumatologists may have already looked at the imaging results before filling out the diagnosis in the case report form. However, this does not explain why change in diagnosis still occurred. More importantly, the overall diagnostic confidence clearly increased after taking imaging results into account, which most likely would not have happened (or to a lesser extent) if rheumatologists had indeed looked at the imaging results before filling out the diagnosis before imaging. Although in line with clinical practice, the fact that each patient was diagnosed by one rheumatologist and the imaging was read by one radiologist might be regarded as a limitation. Furthermore, the rheumatologist may also have used information from other sources such as spinal imaging in the diagnostic process, which may have (additionally) contributed to the LoC regarding the diagnosis. In addition, due to the high overlap of positive imaging (i.e. X-SI+/MRI-SI+), we cannot elaborate on the individual value of each modality (X-SI or MRI-SI) on the final diagnosis and the diagnostic confidence. In conclusion, in CBP patients suspected of having axSpA, sacroiliac imaging increases diagnostic confidence. However, the number of changes in diagnosis suggests that imaging is important but not all-decisive in the early diagnosis of axSpA. 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: L.T.H.J. has received advisory board fees from Abbvie, Celegen, Novartis and Pfizer. R.R. has received honoraria and speaker fees from AbbVie, MSD, Pfizer, Celgene, Janssen, Bristol-Myers Squibb and Mylan. All other authors have declared no conflicts of interest. References 1 Dougados M, Baeten D. Spondyloarthritis. Lancet  2011; 377: 2127– 37. Google Scholar CrossRef Search ADS PubMed  2 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. Google Scholar CrossRef Search ADS PubMed  3 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  4 van Tubergen A, Weber U. Diagnosis and classification in spondyloarthritis: identifying a chameleon. Nat Rev Rheumatol  2012; 8: 253– 61. Google Scholar CrossRef Search ADS PubMed  5 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. 6 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. Google Scholar PubMed  7 Braun J, Baraliakos X, Golder W et al.   Analysing chronic spinal changes in ankylosing spondylitis: a systematic comparison of conventional x rays with magnetic resonance imaging using established and new scoring systems. Ann Rheum Dis  2004; 63: 1046– 55. Google Scholar CrossRef Search ADS PubMed  8 Landewe RB, Hermann KG, van der Heijde DM et al.   Scoring sacroiliac joints by magnetic resonance imaging. A multiple-reader reliability experiment. J Rheumatol 2005; 32: 2050– 5. 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, Ostergaard 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 Lambert RG, Bakker PA, van der Heijde D et al.   Defining active sacroiliitis on MRI for classification of axial spondyloarthritis: update by the ASAS MRI working group. Ann Rheum Dis  2016; 75: 1958– 63. Google Scholar CrossRef Search ADS PubMed  12 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  13 Baraliakos X, Maksymowych WP. Imaging in the diagnosis and management of axial spondyloarthritis. Best Pract Res Clin Rheumatol  2016; 30: 608– 23. Google Scholar CrossRef Search ADS PubMed  14 Ez-Zaitouni Z, Bakker PA, van Lunteren M et al.   Presence of multiple spondyloarthritis (SpA) features is important but not sufficient for a diagnosis of axial spondyloarthritis: data from the SPondyloArthritis Caught Early (SPACE) cohort. Ann Rheum Dis  2017; 76: 1086– 92. Google Scholar CrossRef Search ADS PubMed  15 van Tubergen A, 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  16 Baraliakos X, Hermann KG, Braun J. Imaging in axial spondyloarthritis: diagnostic problems and pitfalls. Rheum Dis Clin North Am  2012; 38: 513– 22. Google Scholar CrossRef Search ADS PubMed  17 Deodhar A, Reveille JD, van den Bosch F et al.   The concept of axial spondyloarthritis: joint statement of the spondyloarthritis research and treatment network and the Assessment of SpondyloArthritis international Society in response to the US Food and Drug Administration's comments and concerns. Arthritis Rheumatol  2014; 66: 2649– 56. 18 Deodhar A. Sacroiliac joint magnetic resonance imaging in the diagnosis of axial spondyloarthritis: “A Tiny Bit of White on Two Consecutive Slices” may be objective, but not specific. Arthritis Rheumatol  2016; 68: 775– 8. Google Scholar CrossRef Search ADS PubMed  19 van den Berg R, de Hooge M, van Gaalen F 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  2013; 52: 1492– 9. Google Scholar CrossRef Search ADS PubMed  20 Ez-Zaitouni Z, Bakker PA, van Lunteren M et al.   The yield of a positive MRI of the spine as imaging criterion in the ASAS classification criteria for axial spondyloarthritis: results from the SPACE and DESIR cohorts. Ann Rheum Dis 2017; 76: 1731– 6. 21 van der Heijde D, Sieper J, Elewaut D et al.   Referral patterns, diagnosis, and disease management of patients with axial spondyloarthritis: results of an international survey. J Clin Rheumatol  2014; 20: 411– 7. 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

Imaging of the sacroiliac joints is important for diagnosing early axial spondyloarthritis but not all-decisive

Loading next page...
 
/lp/ou_press/imaging-of-the-sacroiliac-joints-is-important-for-diagnosing-early-00aDzz9Xte
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/key035
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To evaluate the contribution of the results of sacroiliac imaging to diagnosis and to the level of confidence in diagnosis in patients presenting with chronic back pain (CBP) and suspected of having axial spondyloarthritis (axSpA). Methods. Data from 513 patients from the SPondyloArthritisCaughtEarly cohort with CBP (⩾3 months, ⩽2 years, onset <45 years) were analysed after full diagnostic work-up. Rheumatologists were asked not only to provide a diagnosis before and after the imaging results had been provided to them, but also to provide the level of confidence of this diagnosis on an 11-point numerical scale. Results. Before imaging, 317/513 patients were diagnosed with axSpA. Of these patients, 178/317 (56%) did not have signs of sacroiliitis on either MRI or radiography, which led to the rheumatologist refuting the initial diagnosis of axSpA in 81/178 (46%) patients. Of the 196/513 patients without axSpA before imaging, 35/196 (18%) had signs of sacroiliitis on imaging. Subsequently, 28/35 (80%) patients were diagnosed with axSpA. Overall, imaging was incongruent with the diagnosis before imaging in 213 patients. This led to a change in diagnosis in 109/213 (51%), which corresponds to 21% (109/513) of all patients in the cohort. In general, diagnostic confidence increased by having imaging results available (from 6.2 to 7.4, P < 0.001). Conclusion In patients with CBP suspected of having axSpA, sacroiliac imaging adds to the confidence in the final diagnosis. However, the number of changes in diagnosis suggests that imaging is important but not all-decisive in early axSpA diagnosis. axial spondyloarthritis, ankylosing spondylitis, magnetic resonance imaging, clinical diagnosis Rheumatology key messages Sacroiliac imaging is widely used in the diagnostic work-up of patients suspected of axial SpA. Sacroliac imaging results increase rheumatologists’ confidence in their final diagnosis. Modest changes in numbers of axial SpA following imaging suggest imaging is important but not crucial. Introduction The diagnosis of axial spondyloarthritis (axSpA) in patients presenting with chronic back pain (CBP) is a known challenge in clinical practice as there is a broad spectrum in clinical presentations [1]. Rheumatologists may use information acquired from several sources such as a patient’s medical history, physical examination, laboratory tests and imaging (radiography and/or magnetic resonance imaging of sacroiliac joints) to make a diagnosis [2–4]. Several decades ago, conventional radiography was introduced as an imaging tool to detect sacroiliitis [5]. However, pelvic radiographs can only detect structural changes such as erosions and sclerosis. Furthermore, in early axSpA radiography of the sacroiliac joints may not show structural abnormalities, which can remain absent for many years after disease onset [6]. In recent years, MRI has become an important instrument in the visualization of both inflammation and structural damage in CBP patients especially without radiographic sacroiliitis [7–11]. Imaging of the sacroiliac joints plays a pivotal role in the early recognition of axSpA [12, 13]. We have recently shown that in CBP patients suspected of axSpA, positive imaging of the sacroiliac joints is highly associated with a diagnosis of axSpA (odds ratio = 34.3; 95% CI: 17.3, 67.7) [14]. Unfortunately, recognition of sacroiliitis on radiographs can be difficult. For example, it has been shown that agreement on radiographic sacroiliitis between readers is low and that this does not improve after training [15]. Moreover, bone marrow oedema on MRI suggestive of axSpA may not always be due to axSpA [16]. Therefore, some axSpA experts have expressed their concerns about relying solely on imaging, which may lead to incorrect diagnoses of axSpA. In turn, this may lead to unnecessary exposure to anti-inflammatory drugs with potentially severe side effects [17, 18]. Given these controversies, surprisingly little is known about how rheumatologists actually integrate sacroiliac imaging results in their diagnostic work-up of CBP patients suspected of having axSpA. Therefore, the main objectives of this study were to study the contribution of sacroiliac imaging to the rheumatologist’s diagnosis, and to quantify the contribution of sacroiliac imaging to diagnostic certainty. Methods Study design and population The data used for the current study were obtained from the SPondyloArthritis Caught Early (SPACE) cohort. The SPACE cohort is a prospective multicentre study, which was initiated in January 2009. A detailed description of the study design has been published previously [19]. Briefly, patients with CBP (⩾3 months and ⩽2 years) of unknown origin and age of onset <45 years were recruited and included from multiple European rheumatology centres in the Netherlands, Norway, Italy and Sweden. The clinical database used for the current study was locked on 11 January 2017. Approval for the study was obtained from the local medical ethical committee of the Leiden University Medical Center (reference number: P08.105); patients provided written informed consent before participation. Clinical assessments and measurements All patients were examined according to a standardized full work-up including the assessment of the presence and history of clinical SpA features according to the Assessment of Spondyloarthritis international Society (ASAS) definitions [3]: CRP/ESR, HLA-B27, inflammatory back pain, good response to NSAIDs, positive family history of SpA, peripheral arthritis, dactylitis, heel enthesitis, acute anterior uveitis, IBD and psoriasis. Imaging of the sacroiliac joints Plain radiographs of the pelvis (X-SI) were performed in anteroposterior view. MRI of sacroiliac joints (MRI-SI) was performed in coronal oblique T1-weighted turbo spin echo and short tau inversion recovery with a slice thickness of 4 mm. Interpretation of the radiographs and MRI of the sacroiliac joints (sacroiliitis yes/no) was done by each participating centre as part of routine clinical practice (local reading). Patients were classified according to the ASAS classification criteria for which data from central reading in the SPACE cohort was used [20]. Diagnosis During the diagnostic work-up, rheumatologists were asked to provide a diagnosis twice (axSpA, yes/no): based only on information available after medical history, physical examination and laboratory testing, but before taking sacroiliac imaging into account (diagnosis before imaging), and based on all previously collected information but after taking sacroiliac imaging into account (diagnosis after imaging). In case of no axSpA, rheumatologists were asked to provide the most likely alternative diagnosis. In addition, rheumatologists were requested to provide a level of confidence (LoC) regarding their diagnosis (axSpA or no axSpA) on an 11-point numerical rating scale ranging from 0 (not confident at all) to 10 (very confident). Data analysis For the present study baseline patients with complete information on sacroiliac imaging (both MRI and radiography) and diagnosis were analysed. Descriptive statistics were used to define patient characteristics for the total patient group, for each diagnosis subgroup (axSpA vs no axSpA), and for each imaging subgroup (i.e. sacroiliitis positive and negative) as means (s.d.) for continuous variables or frequencies (%) for categorical variables. Chi-square and unpaired t tests were used to compare variables between groups. The paired t test was used to compare the level of confidence regarding diagnosis before and after imaging within each diagnosis subgroup according to their imaging status. Total number of SpA features was calculated including HLA-B27 status, but without taking imaging of the sacroiliac joints into account. Any positive imaging was defined as one of a negative X-SI but positive MRI-SI (X-SI−/MRI-SI+), a positive X-SI but negative MRI-SI (X-SI+/MRI-SI−) or sacroiliitis on both modalities (X-SI+/MRI-SI+). The rheumatologist’s diagnosis and the LoC regarding diagnosis were the main outcomes. Data analysis was performed using Stata SE v.14 software (StataCorp LP, College Station, TX, USA). Results Baseline data of 583 CBP patients were available. A total of 70 (12%) patients were excluded because of incomplete or missing data regarding diagnosis or sacroiliac imaging. Baseline characteristics of these patients were similar to the remaining 513 patients with complete data (data not shown). Of these 513 CBP patients suspected of having axSpA, 188 (37%) patients were male, mean age (s.d.) was 31.0 (8.2) years, mean symptom duration was 13.3 (7.0) months and 210 (41%) patients were HLA-B27 positive (Table 1). Table 1 Baseline characteristics of CBP patients suspected of axSpA in the SPACE cohort (n = 513) Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Values are listed as n (%), unless otherwise stated. a n = 510 patients for HLA-B27. b n = 499 patients for good response to NSAIDs. axSpA: axial spondyloarthritis; IBP: inflammatory back pain. Table 1 Baseline characteristics of CBP patients suspected of axSpA in the SPACE cohort (n = 513) Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Characteristic    Male  188 (37)  Age, mean (s.d.), years  31.0 (8.2)  Symptom duration, mean (s.d.), months  13.3 (7.0)  Number of SpA features incl. HLA-B27, mean (s.d.)  2.9 (1.8)  HLA-B27 positivea  210 (41)  IBP  346 (68)  Peripheral arthritis  78 (15)  Dactylitis  28 (5)  Enthesitis  106 (21)  Acute anterior uveitis  38 (7)  IBD  35 (7)  Psoriasis  58 (11)  Good response to NSAIDsb  216 (43)  Positive family history of SpA  222 (43)  Elevated CRP/ESR  147 (29)  Values are listed as n (%), unless otherwise stated. a n = 510 patients for HLA-B27. b n = 499 patients for good response to NSAIDs. axSpA: axial spondyloarthritis; IBP: inflammatory back pain. In all patients, rheumatologists provided a diagnosis before and after taking sacroiliac imaging into account (Fig. 1). Before imaging, 317 (62%) patients were diagnosed with axSpA (Table 2). Most common diagnoses in the 196 patients without axSpA were non-specific back pain, degenerative disc disease and mechanical back pain (data not shown). Diagnostic confidence was moderate in patients with and without an axSpA diagnosis before imaging [mean (s.d.) LoC axSpA 6.6 (1.9) and mean LoC no axSpA 5.6 (2.0), respectively]. Patients diagnosed with axSpA before imaging were more often male (41% vs 29%) and were more often HLA-B27 positive (53% vs 22%) compared with patients who were not diagnosed with axSpA before imaging. As expected, the mean (s.d.) number of SpA features was twice as high for axSpA patients compared with the patients not diagnosed with axSpA [3.1 (1.7) and 1.5 (1.0), respectively]. Table 2 Clinical features of CBP patients stratified for diagnosis before taking sacroiliac imaging results into account Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Values are listed as n (%), unless otherwise stated. LoC, level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident). a n = 315 for patients with axSpA diagnosis before imaging. b Past or present condition, either confirmed or diagnosed by a physician. c Back pain not present any more or is much better 24–48 h after a full dose of NSAID. d n = 310 for patients with axSpA diagnosis before imaging, e n = 189 for patients without axSpA diagnosis before imaging. f Presence in first- or second-degree relatives of any of the following: AS, acute anterior uveitis, ReA, psoriasis or IBD. axSpA: axial spondyloarthritis; IBP, inflammatory back pain. Table 2 Clinical features of CBP patients stratified for diagnosis before taking sacroiliac imaging results into account Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Characteristic  Diagnosis before imaging of sacroiliac joints     AxSpA, n = 317  No axSpA, n = 196  LoC regarding diagnosis, mean (s.d.)  6.6 (1.9)  5.6 (2.0)  Male  131 (41)  57 (29)  Number of SpA features incl. HLA-B27, mean (s.d.)  3.6 (1.7)  1.7 (1.1)  HLA-B27 positive  168 (53)a  42 (22)  IBP  253 (80)  93 (48)  Peripheral arthritisb  68 (22)  10 (5)  Dactylitisb  27 (9)  1 (1)  Enthesitisb  92 (29)  14 (7)  Acute anterior uveitisb  32 (10)  6 (3)  IBDb  26 (8)  9 (5)  Psoriasisb  48 (15)  10 (5)  Good response to NSAIDsc  176 (57)d  40 (21)e  Positive family history of SpAf  160 (50)  62 (32)  Elevated CRP/ESR  107 (34)  40 (20)  Values are listed as n (%), unless otherwise stated. LoC, level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident). a n = 315 for patients with axSpA diagnosis before imaging. b Past or present condition, either confirmed or diagnosed by a physician. c Back pain not present any more or is much better 24–48 h after a full dose of NSAID. d n = 310 for patients with axSpA diagnosis before imaging, e n = 189 for patients without axSpA diagnosis before imaging. f Presence in first- or second-degree relatives of any of the following: AS, acute anterior uveitis, ReA, psoriasis or IBD. axSpA: axial spondyloarthritis; IBP, inflammatory back pain. Fig. 1 View largeDownload slide Rheumatologist’s diagnosis before and after considering imaging of sacroiliac joints (n = 513) Any positive imaging defined as sacroiliitis on MRI and/or radiographic sacroiliitis; imaging negative defined as no abnormalities on MRI and radiographs of sacroiliac joints. Boxes in bold represent patients with a change in diagnosis due to imaging results discordant with the diagnosis before imaging. axSpA: axial spondyloarthritis. Fig. 1 View largeDownload slide Rheumatologist’s diagnosis before and after considering imaging of sacroiliac joints (n = 513) Any positive imaging defined as sacroiliitis on MRI and/or radiographic sacroiliitis; imaging negative defined as no abnormalities on MRI and radiographs of sacroiliac joints. Boxes in bold represent patients with a change in diagnosis due to imaging results discordant with the diagnosis before imaging. axSpA: axial spondyloarthritis. In total, 317/513 (62%) of CBP patients were initially diagnosed with axSpA, and this figure decreased to 269/513 (52%) patients with a final diagnosis of axSpA after imaging. Of these 269 axSpA patients, 55% were male, 59% were HLA-B27 positive, 62% had positive imaging and the mean number of SpA features was 3.2 (1.7). A total of 172/269 (64%) patients fulfilled the ASAS classification criteria [52% male, 87% HLA-B27 positive and mean number of SpA features 3.2 (1.5)]. Overall, the mean diagnostic confidence increased by having imaging results available (from 6.2 to 7.4, P < 0.001). This increase in diagnostic confidence was observed in both axSpA patients [from 7.1 (1.7) to 7.7 (2.1), P < 0.001] and non-axSpA patients [from 5.7 (2.0) to 7.5 (2.3), P < 0.001]. Of the 317 patients who were diagnosed with axSpA before imaging, 139 (44%) had positive imaging (Table 3). In these 139 patients, sacroiliitis was seen in 85 (61%) patients only on MRI-SI, in 10 (7%) patients only on X-SI and in 44 (32%) patients on both modalities (X-SI+/MRI-SI+). After imaging, the axSpA diagnosis was maintained in all of these 139 patients and the mean LoC in the diagnosis of axSpA increased significantly [from 7.4 (1.8) to 8.6 (1.7), P < 0.001]. Table 3 Diagnosis and imaging status after sacroiliac imaging in patients with axSpA diagnosis before sacroiliac imaging   AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)    AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)  Values are listed as n (%), unless otherwise stated. a Mean difference 1.2, 95% CI: 0.9, 1.6, P < 0.001; b mean difference −0.3, 95% CI: −0.5, 0.01, P = 0.06; c Mean difference 1.3, 95% CI: 0.7, 1.8, P < 0.001. d Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. e n = 137 for patients with axSpA diagnosis before and after imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Table 3 Diagnosis and imaging status after sacroiliac imaging in patients with axSpA diagnosis before sacroiliac imaging   AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)    AxSpA diagnosis before imaging, n = 317   Any imaging positive, n = 139  Negative imaging, n = 178  Imaging status       Only MRI-SI+  85 (61)  N/A   Only X-SI+  10 (7)  N/A   MRI-SI+/X-SI+  44 (32)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 139  No axSpA, n = 0  AxSpA, n = 97  No axSpA, n = 81    LoC regarding diagnosis before imaging, mean (s.d.)  7.4 (1.8)  —  6.7 (1.6)  5.1 (1.7)  LoC regarding diagnosis after imaging, mean (s.d.)  8.6 (1.7)a  —  6.4 (2.1)b  6.4 (2.0)c  Imaging status       Only MRI-SI+  85 (61)  —  N/A  N/A   Only X-SI+  10 (7)  —  N/A  N/A   MRI-SI+/X-SI+  44 (32)  —  N/A  N/A  Clinical features       Male  71 (51)  —  32 (33)  28 (35)   Number of SpA featuresd, mean (s.d.)  3.4 (1.7)  —  3.5 (1.6)  2.2 (1.2)  HLA-B27+  90 (66)e  —  51 (53)  27 (33)  Values are listed as n (%), unless otherwise stated. a Mean difference 1.2, 95% CI: 0.9, 1.6, P < 0.001; b mean difference −0.3, 95% CI: −0.5, 0.01, P = 0.06; c Mean difference 1.3, 95% CI: 0.7, 1.8, P < 0.001. d Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. e n = 137 for patients with axSpA diagnosis before and after imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Of the 317 patients who were diagnosed with axSpA before imaging, 178 (56%) had negative imaging. In 97/178 patients the diagnosis axSpA was maintained. In these patients, the LoC in the diagnosis somewhat decreased [from 6.7 (1.6) to 6.4 (2.1), P = 0.06]. In 81/178 patients with negative imaging the diagnosis was changed to no axSpA after imaging. In these patients the LoC increased significantly after imaging [from 5.1 (1.7) to 6.4 (2.0), P < 0.001]. By comparison, the 97/178 patients diagnosed with axSpA after imaging had a higher number of SpA features (excluding imaging and HLA-B27) than the 81/178 patients without axSpA [mean 3.5 (1.6) vs 2.2 (1.2), P < 0.001]. Moreover, these 97 axSpA patients were also more often HLA-B27 positive than the 81 patients without axSpA (53% vs 33%, P = 0.010). Of the 196 patients not diagnosed with axSpA before imaging, 35 (18%) had positive imaging (Table 4). In these 35 patients, sacroiliitis was seen in 24 (68%) patients only on MRI-SI, in 2 (6%) patients only on X-SI and in 9 (26%) patients on both modalities (X-SI+/MRI-SI+). In 28 of these 35 patients (80%) with sacroiliitis, the diagnosis was changed to axSpA [18/28 (64%) patients only MRI-SI+, 1/28 (4%) only X-SI+ and 9/28 (32%) MRI-SI+/X-SI+]. The mean LoC in diagnosis increased significantly following imaging [from 4.7 (2.0) to 7.6 (2.3), P < 0.001]. In 7 of the 35 patients with positive imaging (20%), the diagnosis no axSpA remained unchanged [6/7 (86%) patients were MRI-SI+ and one (14%) patient was X-SI+]. In these seven patients, the mean LoC in diagnosis remained the same after imaging [LoC from 5.0 (1.3) to 5.0 (2.3)]. By comparison, the 28/35 patients diagnosed with axSpA after imaging had a higher number of SpA features than the 7/35 patients without axSpA [mean 1.6 (0.9) vs 1.0 (0.8), P = 0.10]. Moreover, these 28 patients diagnosed with axSpA were also more often HLA-B27 positive than the seven patients without axSpA (48% vs 29%, P = 0.35). Table 4 Diagnosis and imaging status after sacroiliac imaging in patients without axSpA diagnosis before sacroiliac imaging   No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)    No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)  Values are listed as n (%), unless otherwise stated. a Mean difference 2.9, 95% CI: 1.6, 4.2, P < 0.001; b mean difference 0; c mean difference 1.4, 95% CI: −3.1, 5.9, P = 0.43; d mean difference 1.8, 95% CI: 1.5, 2.1, P < 0.001. e Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Table 4 Diagnosis and imaging status after sacroiliac imaging in patients without axSpA diagnosis before sacroiliac imaging   No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)    No axSpA diagnosis before imaging, n = 196   Any imaging positive, n = 35  Negative imaging, n = 161  Imaging status       Only MRI-SI+  24 (68)  N/A   Only X-SI+  2 (6)  N/A   MRI-SI+/X-SI+  9 (26)  N/A      Diagnosis after imaging  Diagnosis after imaging  AxSpA, n = 28  No axSpA, n = 7  AxSpA, n = 5  No axSpA, n = 156    LoC regarding diagnosis before imaging, mean (s.d.)  4.7 (2.0)  5.0 (1.3)  5.0 (1.9)  5.8 (2.0)  LoC regarding diagnosis after imaging, mean (s.d.)  7.6 (2.3)a  5.0 (2.3)b  6.4 (1.9)c  7.6 (2.2)d  Imaging status       Only MRI-SI+  18 (64)  6 (86)  N/A  N/A   Only X-SI+  1 (4)  1 (14)  N/A  N/A   MRI-SI+/X-SI+  9 (32)  —  N/A  N/A  Clinical features       Male  15 (54)  2 (29)  2 (40)  38 (24)   Number of SpA featurese, mean (s.d.)  1.6 (0.9)  1 (0.8)  1.8 (0.8)  1.4 (1.0)  HLA-B27+  13 (48)  2 (29)  2 (40)  25 (16)  Values are listed as n (%), unless otherwise stated. a Mean difference 2.9, 95% CI: 1.6, 4.2, P < 0.001; b mean difference 0; c mean difference 1.4, 95% CI: −3.1, 5.9, P = 0.43; d mean difference 1.8, 95% CI: 1.5, 2.1, P < 0.001. e Total number of SpA features after medical history taking, physical examination and measurement of acute phase reactants but before HLA-B27 testing and imaging. axSpA: axial spondyloarthritis; LoC: level of confidence regarding diagnosis: 0 (not confident at all) through 10 (very confident); MRI-SI: MRI of sacroiliac joints; N/A: not applicable; X-SI: radiographs of sacroiliac joints. Of the 196 patients not diagnosed with axSpA before imaging, 161 (82%) had negative imaging. Despite having negative imaging, the diagnosis changed to axSpA after imaging in 5/196 patients (3%). In these five patients, LoC in diagnosis increased [from 5.0 (1.9) to 6.4 (1.9), P = 0.43]. In 156/161 (97%) patients with negative imaging, the diagnosis no axSpA remained unchanged. The LoC in diagnosis increased after imaging [from 5.8 (2.0) to 7.6 (2.2), P < 0.001]. By comparison, the 5/161 patients diagnosed with axSpA after imaging had a higher number of SpA features than the 156/161 patients without axSpA [mean 1.8 (0.8) vs 1.4 (1.0), P = 0.43]. Moreover, these 5 axSpA patients were also more often HLA-B27 positive than the 156 patients without axSpA (40% vs 16%, P = 0.16). In the entire cohort, the diagnosis changed in 21% (109/513) of the CBP patients following imaging. In patients with imaging results that were discordant with their primary diagnosis [n = 213 (178 + 35)], the diagnosis changed in 109 (51%) patients (Fig. 1, boxes in bold). This change of diagnosis more often pertained to patients in whom a clinical diagnosis was suspected but imaging was negative [81/109 (74%)] than in patients in whom a clinical diagnosis of axSpA was not suspected but imaging was positive [28/109 (26%)]. Discussion This study was performed to investigate the contribution of sacroiliac imaging to the rheumatologist’s diagnosis and its respective confidence. We have shown that the rheumatologist’s confidence in the diagnosis of patients with CBP suspected of having axSpA increases after the results of sacroiliac imaging are taken into account. However, the number of changes in diagnosis after imaging implies that imaging is indeed an important but not the all-decisive factor in axSpA diagnosis. Prior to imaging the physician’s confidence in diagnosis was already moderate to high in most patients, which corroborates the value of medical history taking, physical examination and laboratory tests in the diagnostic work-up of axSpA. Congruent imaging results significantly increased the diagnostic confidence of rheumatologists except in axSpA patients (before and after imaging) without sacroiliitis. In general, these findings are in line with results from an international survey conducted among rheumatologists throughout the world illustrating the value rheumatologists place on imaging in the diagnostic work-up of patients suspected of axSpA [21]. Nevertheless, there were patients diagnosed as no axSpA before imaging in which a positive imaging result did not influence the final diagnosis. A few patients were not diagnosed with axSpA after imaging, even though they had sacroiliitis (n = 6 for MRI-SI and n = 1 for X-SI). These patients had only a few other SpA features and rheumatologists remained uncertain about the diagnosis (mean LoC 5.0). Apparently, the diagnosing rheumatologist considered the observed lesions—in combination with the clinical presentation—not sufficiently specific for an axSpA diagnosis, which is an indication that rheumatologists do not necessarily find that imaging is the dominant feature in establishing a diagnosis of axSpA. An even smaller number of patients (n = 5) without a diagnosis of axSpA before imaging and without signs of sacroiliitis were still diagnosed with axSpA after imaging. Since these five patients have a low number of SpA features, it is difficult to understand these diagnoses. Moreover, our findings also stress the importance of imaging in rejecting an axSpA diagnosis as in 81/178 (46%) patients with axSpA before imaging, the diagnosis was dismissed when imaging turned out to be negative. This study has several limitations. First, rheumatologists were asked to provide CBP patients with a diagnosis before and after taking imaging into account. We cannot, however, rule out the possibility that rheumatologists may have already looked at the imaging results before filling out the diagnosis in the case report form. However, this does not explain why change in diagnosis still occurred. More importantly, the overall diagnostic confidence clearly increased after taking imaging results into account, which most likely would not have happened (or to a lesser extent) if rheumatologists had indeed looked at the imaging results before filling out the diagnosis before imaging. Although in line with clinical practice, the fact that each patient was diagnosed by one rheumatologist and the imaging was read by one radiologist might be regarded as a limitation. Furthermore, the rheumatologist may also have used information from other sources such as spinal imaging in the diagnostic process, which may have (additionally) contributed to the LoC regarding the diagnosis. In addition, due to the high overlap of positive imaging (i.e. X-SI+/MRI-SI+), we cannot elaborate on the individual value of each modality (X-SI or MRI-SI) on the final diagnosis and the diagnostic confidence. In conclusion, in CBP patients suspected of having axSpA, sacroiliac imaging increases diagnostic confidence. However, the number of changes in diagnosis suggests that imaging is important but not all-decisive in the early diagnosis of axSpA. 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: L.T.H.J. has received advisory board fees from Abbvie, Celegen, Novartis and Pfizer. R.R. has received honoraria and speaker fees from AbbVie, MSD, Pfizer, Celgene, Janssen, Bristol-Myers Squibb and Mylan. All other authors have declared no conflicts of interest. References 1 Dougados M, Baeten D. Spondyloarthritis. Lancet  2011; 377: 2127– 37. Google Scholar CrossRef Search ADS PubMed  2 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. Google Scholar CrossRef Search ADS PubMed  3 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  4 van Tubergen A, Weber U. Diagnosis and classification in spondyloarthritis: identifying a chameleon. Nat Rev Rheumatol  2012; 8: 253– 61. Google Scholar CrossRef Search ADS PubMed  5 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. 6 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. Google Scholar PubMed  7 Braun J, Baraliakos X, Golder W et al.   Analysing chronic spinal changes in ankylosing spondylitis: a systematic comparison of conventional x rays with magnetic resonance imaging using established and new scoring systems. Ann Rheum Dis  2004; 63: 1046– 55. Google Scholar CrossRef Search ADS PubMed  8 Landewe RB, Hermann KG, van der Heijde DM et al.   Scoring sacroiliac joints by magnetic resonance imaging. A multiple-reader reliability experiment. J Rheumatol 2005; 32: 2050– 5. 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, Ostergaard 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 Lambert RG, Bakker PA, van der Heijde D et al.   Defining active sacroiliitis on MRI for classification of axial spondyloarthritis: update by the ASAS MRI working group. Ann Rheum Dis  2016; 75: 1958– 63. Google Scholar CrossRef Search ADS PubMed  12 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  13 Baraliakos X, Maksymowych WP. Imaging in the diagnosis and management of axial spondyloarthritis. Best Pract Res Clin Rheumatol  2016; 30: 608– 23. Google Scholar CrossRef Search ADS PubMed  14 Ez-Zaitouni Z, Bakker PA, van Lunteren M et al.   Presence of multiple spondyloarthritis (SpA) features is important but not sufficient for a diagnosis of axial spondyloarthritis: data from the SPondyloArthritis Caught Early (SPACE) cohort. Ann Rheum Dis  2017; 76: 1086– 92. Google Scholar CrossRef Search ADS PubMed  15 van Tubergen A, 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  16 Baraliakos X, Hermann KG, Braun J. Imaging in axial spondyloarthritis: diagnostic problems and pitfalls. Rheum Dis Clin North Am  2012; 38: 513– 22. Google Scholar CrossRef Search ADS PubMed  17 Deodhar A, Reveille JD, van den Bosch F et al.   The concept of axial spondyloarthritis: joint statement of the spondyloarthritis research and treatment network and the Assessment of SpondyloArthritis international Society in response to the US Food and Drug Administration's comments and concerns. Arthritis Rheumatol  2014; 66: 2649– 56. 18 Deodhar A. Sacroiliac joint magnetic resonance imaging in the diagnosis of axial spondyloarthritis: “A Tiny Bit of White on Two Consecutive Slices” may be objective, but not specific. Arthritis Rheumatol  2016; 68: 775– 8. Google Scholar CrossRef Search ADS PubMed  19 van den Berg R, de Hooge M, van Gaalen F 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  2013; 52: 1492– 9. Google Scholar CrossRef Search ADS PubMed  20 Ez-Zaitouni Z, Bakker PA, van Lunteren M et al.   The yield of a positive MRI of the spine as imaging criterion in the ASAS classification criteria for axial spondyloarthritis: results from the SPACE and DESIR cohorts. Ann Rheum Dis 2017; 76: 1731– 6. 21 van der Heijde D, Sieper J, Elewaut D et al.   Referral patterns, diagnosis, and disease management of patients with axial spondyloarthritis: results of an international survey. J Clin Rheumatol  2014; 20: 411– 7. 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 20, 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