Pos0503 Predictive Validity of Data-Driven Definitions for Active and Structural Lesions in the Si Joints Typical for Axial Spa: A 2-Year Follow-Up in the Spondyloarthritis Caught Early Cohort

Background: De Hooge et al. (2016) and the Assessment of Spondyloarthritis international Society MRI (ASAS MRI) group (2021) have proposed data-driven definitions for active and structural lesions typical for axial spondyloarthritis (axSpA) on MRI in the SI-joints (MRI-SIJ) [1,2]. The definitions of de Hooge et al. have not been tested against the rheumatologist's diagnosis, while the ASAS MRI definitions lack validation in other cohorts and assessment in early disease. Objectives: To analyse the predictive validity of the proposed definitions for active and structural MRI-SIJ lesions in early axSpA with a 2-year diagnosis as outcome. Methods: Patients with chronic back pain (≥3 months; ≤2 years; onset <45 years) from the SPondyloArthritis Caught Early (SPACE) inception cohort were followed-up and diagnosed (axSpA/no axSpA) by a rheumatologist after 2 years [3]. Patients with definite diagnosis (level of confidence ≥7) or most likely diagnosis (level of confidence <7 and a consistent diagnosis in the last 2 available visits) of either axSpA or no-axSpA were included in this analysis. Three central readers scored baseline MRI-SIJ for bone marrow edema (BME), erosions, and fat lesions. When data were available for ≥2 central readers, the MRI-SIJ were analysed for consensus between ≥2 central readers on scored quadrants and slices. We assessed lesions as defined by de Hooge et al. and the ASAS MRI group (Figures 1 and 2). De Hooge et al. defined lesions to exist only if present in at least two consecutive slices (Figure 2A and B). The ASAS MRI group uses the number of quadrants affected (Figure 2A) by a lesion, the presence of a lesion across consecutive slices (Figure 2C), and combinations between type of lesions. We calculated sensitivity, specificity, positive and negative predictive values (PPV and NPV) for each definition (Figure 1). Based on consensus scores, a combination of specificity ≥95% and PPV ≥95% was required for a definition to be considered validated. Results: We analysed 643 patients (age 30 (SD 8) years; 39% males; 52% axSpA). Lesions had low prevalence within the cohort (Figure 1), especially structural lesions. All proposed definitions by de Hooge et al. met the threshold (Figure 1A). Lower cut-off definitions also met the threshold (Figure 1A). All but one of the proposed definitions of the ASAS MRI group met the threshold (Figure 1B). However, erosions in ≥2 consecutive slices did not meet the threshold, but a higher threshold of erosions in ≥4 consecutive slices did. Other definitions also met the threshold (Figure 1B). Combinations of lesions performed well but did not outperform single item lesions. Conclusion: Structural lesions were infrequently present in early axSpA. All proposed ASAS-MRI group definitions for structural MRI-SIJ lesions, except for erosions in ≥2 consecutive slices, have been validated in early axSpA based on high specificity and PPV. Consequently, we propose ≥4 consecutive slices for erosions as it outperformed the ASAS MRI group definition of ≥2 consecutive slices. The definitions of de Hooge et al. showed a similar performance to the ones by the ASAS MRI group, however they are easier to apply and thus have a higher feasibility. REFERENCES: [1] de Hooge M, et al. Ann Rheum Dis 2016; 75:1308–1314. [2] Maksymowych et al. Rheumatology 2021; 60:4778–4789. [3] Marques ML et al, Ann Rheum Dis 2024. (epub ahead of print) Acknowledgements: NIL. Disclosure of Interests: Liese J.E. de Bruin: None declared, Mary Lucy Marques: None declared, Miranda van Lunteren: None declared, Manouk de Hooge UCB, Sofia Exarchou Amgen, Janssen, Novartis and UCB Pharma, Karen Minde Fagerli: None declared, Roberta Ramonda Advisory Boards from Abbvie, Novartis, Lilly, Pfizer, Advisory Boards from Abbvie, Novartis, Janssen, Lilly, MSD, Pfizer, and UCB, Robert Landewé: None declared, F. A. van Gaalen Abbvie, ASAS, BMS, Galapagos, Janssen, Lilly, Novartis, Pfizer, UCB Pharma, Novartis, ASAS, UCB Pharma, Désirée van der Heijde Director of Imaging Rheumatology BV., AbbVie, BMS, Galapagos, Glaxo-Smith-Kline, Janssen, Lilly, Novartis, Pfizer, Takeda, UCB Pharma, Sofia Ramiro AbbVie, Eli Lilly, Janssen, MSD, Novartis, Pfizer, Sanofi, UCB Pharma, AbbVie, Galapagos, MSD, Novartis, Pfizer, UCB Pharma.