Decreased Discontinuation and Switching of b/tsDMARDs in RA When Treatment Aligns with Molecular Signature Response Classifier Results: An Analysis from The Study to Accelerate Information of Molecular Signatures (AIMS)

Background: Rheumatoid arthritis (RA) patients with inadequate response to methotrexate often experience trial-and-error treatment selection, due to a lack of guidance from clinical guidelines or validated biomarkers. This frequent therapy discontinuation and switching increases healthcare costs without necessarily improving patient outcomes. A blood-based molecular signature response classifier (MSRC) has been validated to predict RA patients who will be non-responders to tumor necrosis factor-ɑ inhibitors (TNFi), a common RA treatment.

Objectives: This study examined the impact of treatment discontinuation and switching of a biologic or targeted-synthetic therapy (b/tsDMARD) according to treatment selection alignment with MSRC results.

Methods: Analyses were conducted in the Study to Accelerate Information of Molecular Signatures (AIMS), a longitudinal, prospective study of RA patients in the US. Included patients initiated a b/tsDMARD after MSRC testing and had a 6-month follow-up visit. Patients were classified according to whether they initiated a TNFi or an alternative mechanism of action (altMOA), and within the TNFi-treated group, whether the MSRC detected a signature of predicted non-response (Unaligned-TNFi) or not (Aligned-TNFi). Discontinuation was defined as termination of the b/tsDMARD prescribed at treatment decision, and a switch as initiation of a new b/tsDMARD. Kaplan-Meier curves were used to estimate the time to discontinuation or switch. Cox proportional hazard models were used to adjust for age, sex, and baseline clinical disease activity index (CDAI), tender joint count, and TNFi-naïve status.

Results: Analyses included 666 patients: mean age 56.6 (SD=13.3) years, mean baseline CDAI 32.6 (SD=16.1), 80% female, 72% TNFi-naïve. Most (72%) had a treatment decision aligned with the MSRC result, of which 39% were prescribed a TNFi. The Unaligned-TNFi group (TNFi-treated patients with a signature of predicted non-response to TNFi) had a higher proportion of discontinuation and switching than the Aligned group. Adjusted analyses showed the Aligned group was less likely to discontinue treatment (Aligned-TNFi HR=0.54, 95%CI: 0.33 – 0.86; Aligned-altMOA HR=0.63, 95%CI: 0.43 – 0.94) and less likely to switch therapy (Aligned-TNFi HR=0.54, 95%CI: 0.33 – 0.89; Aligned-altMOA HR=0.53, 95%CI: 0.34 – 0.81).

Conclusions: When rheumatologists select a b/tsDMARD therapy aligned with MSRC results, patients experience less therapy discontinuation and switching than those treated with a TNFi despite MSRC prediction of future non-response. This study further demonstrates the value of precision medicine in RA in which a treatment paradigm is informed by a validated biomarker test.