OBJECTIVE: Non-radiographic axial spondyloarthritis (SpA) is characterised by a lack of definitive radiographic sacroiliitis and is considered an early stage of ankylosing spondylitis. The objective of this study was to develop candidate classification criteria for axial SpA that include patients with but also without radiographic sacroiliitis. METHODS: Seventy-one patients with possible axial SpA, most of whom were lacking definite radiographic sacroiliitis, were reviewed as "paper patients" by 20 experts from the Assessment of SpondyloArthritis international Society (ASAS). Unequivocally classifiable patients were identified based on the aggregate expert opinion in conjunction with the expert-reported level of certainty of their judgement. Draft criteria for axial SpA were formulated and tested using classifiable patients. RESULTS: Active sacroiliitis on magnetic resonance imaging (MRI) (odds ratio 45, 95% CI 5.3 to 383; p<0.001) was strongly associated with the classification of axial SpA. The knowledge of MRI findings led to a change in the classification of 21.1% of patients. According to the first set of candidate criteria (sensitivity 97.1%; specificity 94.7%) a patient with chronic back pain is classified as axial SpA in the presence of sacroiliitis by MRI or x rays in conjunction with one SpA feature or, if sacroilitiis is absent, in the presence of at least three SpA features. In a second set of candidate criteria, inflammatory back pain is obligatory in the clinical arm (sensitivity 86.1%; specificity 94.7%). CONCLUSION: The ASAS group has developed candidate criteria for the classification of axial SpA that include patients without radiographic sacroiliitis. The candidate criteria need to be validated in an independent international study.
OBJECTIVE: Non-radiographic axial spondyloarthritis (SpA) is characterised by a lack of definitive radiographic sacroiliitis and is considered an early stage of ankylosing spondylitis. The objective of this study was to develop candidate classification criteria for axial SpA that include patients with but also without radiographic sacroiliitis. METHODS: Seventy-one patients with possible axial SpA, most of whom were lacking definite radiographic sacroiliitis, were reviewed as "paper patients" by 20 experts from the Assessment of SpondyloArthritis international Society (ASAS). Unequivocally classifiable patients were identified based on the aggregate expert opinion in conjunction with the expert-reported level of certainty of their judgement. Draft criteria for axial SpA were formulated and tested using classifiable patients. RESULTS: Active sacroiliitis on magnetic resonance imaging (MRI) (odds ratio 45, 95% CI 5.3 to 383; p<0.001) was strongly associated with the classification of axial SpA. The knowledge of MRI findings led to a change in the classification of 21.1% of patients. According to the first set of candidate criteria (sensitivity 97.1%; specificity 94.7%) a patient with chronic back pain is classified as axial SpA in the presence of sacroiliitis by MRI or x rays in conjunction with one SpA feature or, if sacroilitiis is absent, in the presence of at least three SpA features. In a second set of candidate criteria, inflammatory back pain is obligatory in the clinical arm (sensitivity 86.1%; specificity 94.7%). CONCLUSION: The ASAS group has developed candidate criteria for the classification of axial SpA that include patients without radiographic sacroiliitis. The candidate criteria need to be validated in an independent international study.