Background The development of the ASAS classification criteria has facilitated the diagnosis of Spondyloarthritis (SpA) at early and non-radiological stages. Despite these non-radiological forms are considered as less aggressive by Rheumatologists, their clinical relevance is still discussed and besides, the accuracy of radiological diagnosis of sacroiliitis at the early stages of the disease has been questioned in some forums.

Objectives The aim of this study is to asses the diagnostic reliability of pelvis x-Rays and sacroiliac magnetic resonance imaging (MRI) in an early axial SpA cohort.

Methods Patients included in the study came from three Spanish early SpA units, as part of the ESPERANZA programme, a nationwide health management programme designed to provide excellence in care for early SpA. One of the inclusion criteria was symptom duration between 3 and 24 months. All patients fulfilled ASAS criteria for axial SpA at baseline. Radiographs of pelvis at baseline and at follow up were done for every patient. These x-Rays were centrally digitized and the SIJ were scored according to the grading system of the modified New York criteria by two independent and blinded readers. MRIs of SIJ were assessed according to ASAS definition for bone marrow oedema (BME) by two independent and blinded readers. BME was also scored using the Leeds MRI scoring system. Fatty lesions within the sacroiliac joints were assessed in the T1 sequences. SPSS program version 17.0 was used for Statistical analysis. Kappa test was used for reliability analysis.

Results 48 patients were included, 25 of them (52.1%) were male. The mean age was 33.9 years (range: 20-45; SD 6.51) and 46-9% were HLA B27 positive. The average evolution time since onset of symptoms until baseline clinic visit was 10.21 months (range: 3-24; SD 6.83). The mean follow up time (between baseline and follow up x-Rays) was 3.25 years (range: 1-5; SD 1.08).The global Kappa coefficient of the x-Rays reading, taking into account both visits together (48 baseline x-Rays and 48 follow-up x-Rays) was 0.699, which is a good result. When we analysed the x-Rays reading of both visits separately, we found that the kappa coefficient of the x-Ray reading corresponding to the baseline visit (48 x-Rays) was 0.5, which is a moderate result, but it increased to 0.869 when we analysed the x-Rays reading corresponding to the follow-up visit, which is an excellent result. Regarding MRI, there were 27 available for analysis and the agreement for the ASAS definition of BME was excellent (Kappa 1.00). The kappa coefficient obtained with the Leeds MRI scoring system was 0.686 and the intraclass correlation coefficient was 0.807. The agreement obtained for the fatty lesions was fair (Kappa 0.34).

Conclusions The inter-reader reliability of sacroiliac x-Ray improves as disease progresses, and, regarding to these results, it has a limited utility at the early stages of SpA. In the other hand, sacroiliac MRI is a reliable tool for diagnosis at early stages and in non-radiological axial SpA.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.5949