AB0717 RELIABILITY OF SACROILIAC JOINT RADIOGRAPHS IN THE EARLY SPONDYLOARTHRITIS ESPERANZA COHORT

Background: X-Ray sacroiliitis is the cornerstone in the diagnosis of the ankylosing spondylitis (AS). It is clear that the presence of sacroiliitis is a specific lesion in long standing aS, but little evidence exists on the reliability of this image technique in early axial spondyloarthritis (axSpA), which associates with reduced structural damage. This is relevant because some rheumatologists and pharmacological authorities feel more confident with the diagnosis of aS rather than with non radiographic axial spondyloarthritis (nr-axSpA) forms. On the other hand, according to various research studies, clinicians are subject to potential bias when interpreting radiographs, influenced by their pretest clinical judgment Objectives: The main objective of this study is to determine the reliability of X-Ray sacroiliitis in the early diagnosis of axSpA. Methods: This study included 290 radiographs of the SI joints from patients of the Esperanza early spondyloarthritis cohort. Nine readers, blinded for the diagnosis, participated in the reliability exercise, all of them experienced rheumatologists and members of the Spanish spondiloarthritis working group (GRESSER). Patients with axSpa were classified as having aS if the radiographic criteria of the modified NY criteria (presence of radiographic changes in the SIJ of at least grade II bilaterally or at least grade III unilaterally) were fulfilled. The gold standard was the categorical opinion of at least five of the expert readers. For the statistical analysis, the Chi-square and Kappa tests were performed. Results: The table shows mean kappa values and the agreement reached by each reader. The mean kappa achieved was fair (0.375, range 0.146 - 0.652) and the mean agreement was 73.7% (range: 58.7% - 90%). When the categorical opinion of X-Ray sacroiliitis of at least 5 readers as gold standard was applied, 61 patients were classified as aS and 229, as nr-axSpa. Nevertheless, this scenario varied from 31 to 138 among the readers when the x-ray evaluation was performed by only one reader. Conclusion: Reliability of X-Ray SIJ in an early SpA cohort was weak among nine experimented readers and the diagnosis of aS was subject to a high variability. At least in doubtful cases, a central evaluation performed by highly qualified readers is advisable. Disclosure of interests: Carolina Tornero: None declared, Claudia Urrego-Laurin: None declared, Maria Luz Garcia-Vivar: None declared, Cristina Fernandez-Carballido: None declared, Xavier Juanola-Roura: None declared, Jose Francisco Garcia LLorente: None declared, Maria del Carmen Castro Villegas Paid instructor for: MSD, abbvie, Pfizer, Janssen, Lilly, Roche, Beatriz Joven-Ibanez Speakers bureau: Celgene, Novartis, MSD, Pfizer, abbVie, and Janssen, Eva Galindez: None declared, Eugenio de Miguel: None declared