5F-04 - Treatment Sequencing and Line of Therapy for Biologics and JAK inhibitors in Rheumatoid Arthritis Patients: Implications for Design and Uptake of New Drugs and Diagnostics

Background: Patients with rheumatoid arthritis (RA) have numerous treatment options, and predictive biomarker tests for treatment response are emerging. However, access to RA therapies and tests may be limited by health plans based on prior treatment history (e.g., biologic-naïve vs. experienced).

Objectives: To analyze RA treatment sequences by line of therapy (LoT) and medication switch rates to estimate eligible populations for novel therapies and predictive tests using a population-based RA cohort.

Methods: We used a linked health plan claims and electronic health record (EHR) database from the Excellence Network in Rheumatology to Innovate Care and High-impact research (ENRICH). Inclusion criteria: ≥2 RA diagnoses from rheumatologists ≥30 days apart within 365 days, age ≥18 years, ≥12 months of prior continuous coverage, and ≥1 conventional, biologic, or targeted synthetic DMARD (c/b/tsDMARD) after RA diagnosis. Exclusions: other inflammatory rheumatic diseases or prior b/tsDMARD use. Biologic and JAK inhibitor (JAKi) use was identified through various National Library of Medicine coding systems, generics, and brand names. We assessed cohort characteristics by LoT, RA treatment switching rates, time to switch, treatment sequences, and agreement between LoT classification in claims/EHR vs. self-reported treatment history (weighted Cohen’s Kappa, κ).

Results: Among 37,656 newly diagnosed RA patients, 59,557 new RA treatment initiations were identified. Median [IQR] age at first b/tsDMARD initiation: 54.3 [45.1, 61.4] years. Most patients were commercially insured (78.5%), White (68.5% of those with race data), and had high glucocorticoid use (68.1%) within 6 months. Among b/tsDMARD-naïve patients, 36.5% initiated monotherapy, increasing to ~50% with additional b/tsDMARDs. Most (85.2%) started biologics (mainly TNFi), while 40.2% overall used biologics/JAKi as monotherapy. JAKi initiation peaked in 2019-2020 (18.3%) before declining to 13.7% in 2023-2024. The overall medication switch rate was 25.7/100 patient-years (22.7/100py in biologic-naïve, >30/100py in treatment-experienced patients), with substantial provider variability. Time to switch was significantly longer (p < 0.001) in b/tsDMARD-naïve patients but changed minimally with increasing treatment failures. A 1-year baseline using all prior data showed better agreement with self-reported lifetime treatment history (claims: κ=0.68; EHR: κ=0.59) than a fixed 1-year baseline (claims: κ=0.32; EHR: κ=0.30).

Conclusions: These findings offer valuable insights for stakeholders developing RA therapies or diagnostic tests, highlighting eligible RA populations and treatment patterns by LoT.