POS0636 PATIENT GROUPS IN RHEUMATOID ARTHRITIS IDENTIFIED BY DEEP LEARNING RESPOND DIFFERENTLY TO BIOLOGIC OR TARGETED SYNTHETIC DISEASE MODIFYING ANTIRHEUMATIC DRUGS

Abstract

BackgroundCycling of biologic or targeted synthetic disease modifying antirheumatic drugs (b/tsDMARDs) in rheumatoid arthritis (RA) patients due to non-response is a problem preventing and delaying disease control.ObjectivesTo assess and validate treatment response of b/tsDMARDs among RA patient groups identified by deep learning.MethodsIn the Swiss Clinical Quality Management of Rheumatic Diseases registry (SCQM), between 1998 and 2018, we identified all RA patients with a DAS28-erythrocyte sedimentation rate (esr) record within 6 months before start of the first b/tsDMARD. This first-time b/tsDMARD was the cohort entry at which patients were clustered through several runs of deep embedded clustering. Features, measured at cohort entry, included demographics, RA disease burden/duration, life-style factors, and other RA medication. To increase robustness of the obtained clusters, we grouped similar patient clusters together (further referred to as groups).Our outcomes were b/tsDMARD stop due to non-response, and separately a ≥20% reduction in DAS28-esr (RA disease activity in 28 joints using esr measures) as a proxy for treatment response. We followed all patients from cohort entry until b/tsDMARD stop or a maximum of 15 months follow-up. We assessed comparative effectiveness of b/tsDMARDs (ref. adalimumab) using Cox proportional hazard regression in each patient group by estimating hazard ratios (HR) with 95% confidence intervals (CI).We validated results obtained per patient group through stratified analyses according to most distinctive patient characteristics of the respective group, i.e. the characteristics that led to the respective grouping were also used to stratify the overall population by in this validation analysis.ResultsWe obtained 24 clusters which comprised between 362 and 1481 patients (among 3516 unique patients). These clusters were grouped into 5 groups according to most distinct characteristics at b/tsDMARD initiation: 1) ≥2 csDMARDs and prednisone use, 2) male sex, 3) seronegativity, female sex, and no prednisone use, 4) rather low disease burden, 5) seropositivity, female sex, and a rather high disease burden/duration.Comparative effectiveness results among validation strata confirmed comparative effectiveness results observed among the 5 groups: Patients with ≥2 csDMARDs and prednisone at b/tsDMARD initiation, men, as well as patients with a lower disease burden responded better to tocilizumab than to adalimumab (HRs of reaching ≥20% reduction in DAS28-esr: 5.46, 95% CI [1.76-16.94], HR 8.44 [3.43-20.74], and HR 3.64 [2.04-6.49], respectively). Furthermore, seronegative women without use of prednisone at b/tsDMARD initiation as well as seropositive women with a higher disease burden and longer disease duration had a higher risk of non-response with golimumab (HRs of b/tsDMARD discontinuation: 2.36 [1.03-5.40] and HR 5.27 [2.10-13.21], respectively) than with adalimumab.ConclusionOur results suggest that RA patient groups identified by deep learning may respond differently to individual first-line b/tsDMARDs. Thus, our results can possibly support the decision on the best choice of first-time b/tsDMARD for certain RA patients, which is a step forward towards personalizing treatment. However, further research in other cohorts is needed to verify our results.AcknowledgementsWe thank all patients and rheumatologists contributing to the SCQM registry, as well as the entire SCQM staff. A list of rheumatology offices and hospitals which contribute to the SCQM registry can be found at http://www.scqm.ch/institutions. A list of financial supporters of SCQM can be found at http://www.scqm.ch/sponsors. The professorship of Andrea M Burden is partially supported by PharmaSuisse and the ETH Foundation.Disclosure of InterestsMaria Kalweit: None declared, Andrea Michelle Burden: None declared, Thomas Hügle Speakers bureau: Novartis, Janssen, Eli Lilly, Paid instructor for: Abbvie, Menarini, Consultant of: Janssen, Eli Lilly, Grant/research support from: Abbvie, Menarini, Pfizer, Theresa Burkard: None declared