(P-515) Factors Predictive of Treatment Success of Novel Disease-modifying Anti-rheumatic Drugs (DMARDS) in Patients with Inflammatory Arthritis

Background: The introduction of novel disease-modifying anti-rheumatic drugs (DMARDs) has dramatically changed the management of inflammatory arthritis (rheumatoid arthritis and psoriatic arthritis). Little is known about the factors associated with treatment success of novel DMARDs in patients with inflammatory arthritis.

Objectives: To explore potential factors associated with treatment failure or success of novel DMARDs.

Methods: Patients initiating methotrexate (MTX) prescribed by a rheumatologist between 2008-2017 were selected from PHARMO’s Out-patient Pharmacy Database. The date of the first MTX dispensing was defined as the cohort entry date (CED). Patients initiating a novel DMARD (biological or targeted synthetic DMARD) for inflammatory arthritis after CED were included in this cohort study. Duration of novel DMARDs use was calculated by converting dispensings of novel DMARDs into treatment episodes of uninterrupted use. Treatment failure was defined as a first treatment episode <6 months and treatment success was defined as a first treatment episode ≥12 months. Patient characteristics, including comedication, infections and chronic disease score (CDS) were compared between patients failing and patients with treatment success on novel DMARDS by using the χ2 test for categorical and ANOVA for continuous characteristics.

Results: A total of 698 patients were included. The majority initiated etanercept (51%), followed by adalimumab (26%) and golimumab (10%). About half of the patients (n=391 (56%)) failed on novel DMARDs, 139 (20%) had a first treatment duration between 6 and 12 months, and 168 (24%) patients had treatment success. Patients with treatment failure initiated novel DMARDs in earlier years than patients with treatment success (54% vs 40% initiated in 2008-2013, respectively). Patients with treatment success were more often initiated by a rheumatologist than patients with treatment failure (82% vs 71%, respectively) and more often initiated infliximab than patients with treatment failure (9% vs 2%, respectively).

Conclusions: Year of initiating novel DMARD, type of novel DMARD and type of prescriber differed significantly between patients failing on novel DMARDs and patients with treatment success. These differences suggest that physicians have a relevant impact on drug (dis)continuation, which may have changed over time. From the current study, comedication, infections and CDS do not seem to be predictive for treatment duration of novel DMARDs. Recent evidence suggests that treatment failure might be explained by genetic differences not yet captured in our database network.