SAT0177 Real-world interruptions in janus kinase inhibitor therapy observed among biologic-naÏve and biologic-experienced rheumatoid arthritis patients

Background: Orally administered janus kinase inhibitors (JAKi) can be used to treat moderate-to-severe rheumatoid arthritis (RA). Objectives: Examine real-world patterns of use, including interruptions in therapy and switching, in biologic-naive and biologic-experienced patients with RA newly prescribed a JAKi. Methods: A retrospective analysis was performed using healthcare claims data from the Optum Research Database. Commercial and Medicare Advantage patients aged ≥18 years were included with: ≥1 pharmacy claim for a JAKi 01 January 2012–29 February 2016 (index date = date of first claim); continuous enrollment in health plan for ≥12 months before and 13 months post-index; and ≥1 claim with a RA diagnosis (ICD-9 = 714.0x; ICD-10 = M05* or M06*) during baseline or on the index date. A variable length pre-index period to 01 January 2006 was implemented to capture certain measures. Persistence was defined as the time from index prescription fill to earliest gap in fills ≥30 days (interruption) or switching to a biologic disease-modifying antirheumatic drug (bDMARD). Patients were classified as being: persistent with the index therapy for the full year, switching before interruption, switching after interruption, interruption with restart, and interruption without restart (discontinuation). Patients were assigned to two groups: those with and without ≥1 claim for bDMARD treatment during the variable length pre-index period (biologic-experienced and biologic-naive). Data were expressed as % or mean (SD). Sensitivity analyses assessed prescription fill gaps ≥90 days. Results: Of 1059 patients who met the inclusion criteria, 80.2% were biologic-experienced, 81.8% female, mean age 56.2 (11.8) years. More biologic-naive patients (48.6%) had Medicare Advantage coverage than biologic-experienced (22.3%). For the primary analysis, 72.4% of patients were not persistent for the 1-year following initiation of a JAKi, with 39.1% not persistent beyond 90 days (figure 1). Among non-persistent patients, 42.4% interrupted and restarted a JAKi (time from index to restart, 168.9 [83.8] days), 16.8% switched to a bDMARD before interruption, 13.3% switched after interruption and 27.5% discontinued bDMARD and JAKi therapy (time to discontinuation, 148.4 [117.6] days). More biologic-naive than biologic-experienced patients neither initiated a bDMARD nor restarted a JAKi after interruption (44.2% vs 23.3%, respectively; P 0.001). For the sensitivity analysis, 52.6% were non-persistent for the full year. Figure 1 Kaplan–Meier curve of persistence with JAKI therapy Conclusions: Most patients (80.2%) newly initiating JAKi therapy had prior bDMARD experience. Over 70% were non-persistent with JAKi treatment for 1 year, with 39% non-persistent beyond 90 days. For non-persistent patients, the pattern of JAKi use was characterized most as interrupting with restart (42%), followed by switching (30%), and then discontinuation (28%). Reasons for the high non-persistence rate are unknown but may include suboptimal efficacy or intolerance. Further research is needed to elucidate these points. Acknowledgements: Research sponsored by Sanofi and Regeneron Pharmaceuticals, Inc. Disclosure of Interest: B. Chastek Grant/research support from: Sanofi and Regeneron Pharmaceuticals, Inc., E. Koep Grant/research support from: Sanofi and Regeneron Pharmaceuticals, Inc., U. Mallya Shareholder of: Sanofi, Employee of: Sanofi, A. Cockerham Employee of: Sanofi, J. Choi Shareholder of: Sanofi, Employee of: Sanofi, S. Boklage Shareholder of: Regeneron Pharmaceuticals, Inc., Employee of: Regeneron Pharmaceuticals, Inc., D. Furst Grant/research support from: Amgen, Bristol-Myers Squibb, Novartis, Sanofi-Regeneron, Pfizer, Roche/Genentech, and Corbus, Consultant for: AbbVie, Amgen, Bristol-Myers Squibb, Corbus, Cytori, Novartis, Pfizer, and Roche/Genentech, Speakers bureau: Bristol-Myers Squibb and Abbvie