THU0175 REAL-WORLD EXPERIENCE OF SWITCHING FROM ORIGINATOR TO BIOSIMILAR RITUXIMAB IN RHEUMATOID ARTHRITIS

Abstract

Background:Pre-clinical and trial data confirm biosimilar rituximab (RTX-B) has equivalent physicochemical properties, function, efficacy and safety as originator (RTX-O)1,2. However, real-world data are lacking, including non-medical switch to RTX-B in patients established on RTX-O. Loss of response (LOR) has been observed following open-label switch to other biosimilars3, and it is unclear how patients who develop LOR after switching should be managed.Objectives:To describe our real-world experience of non-medical switch from RTX-O to RTX-B in RA and explore treatment strategies for patients developing LOR to RTX-B (switch back to RTX-O versus other b/tsDMARDs).Methods:We conducted a retrospective observational study of RTX-treated RA patients in Leeds from May 2002–Jan 2020. From Oct 2017, all patients on RTX-O were switched to RTX-B unless specified by treating clinician or informed consent not obtained. Demographic and clinical data were collected from electronic health records. Effectiveness was assessed by DAS28-CRP and RTX retention (using Kaplan-Meier method). Changes in treatment were based on clinician judgment.Results:944 patients with RA received ≥1 cycle of RTX in Leeds over the past 17 years. Of these, 358 required RTX re-treatment during the study period (Oct 2017–Jan 2020); 263 (73.5%) switched to RTX-B, whilst 95 (26.5%) remained on RTX-O (medical reason=46.3%; patient decision=53.7%). 256/263 patients on RTX-B with 4-month data were analysed. 201/256 (78.5%) were female, mean (SD) age 63.1 (12.4), 97.7% seropositive, 81.6% on csDMARD (62.9% methotrexate), 14.8% oral steroid, 58.6% had previous bDMARDs, and mean (SD) RTX-O cycles pre-switch was 6 (3.6). Total follow-up was 378 patient-years. Mean (SD) DAS28-CRP was 3.01 (1.17) 4 months post-last cycle RTX-O and 3.32 (1.31) post-switch to RTX-B (mean diff +0.30, 95% CI 0.004-0.60; p=0.047), with increase in VAS (+5.7, 95% CI 0.09-11.37; p=0.047), but not SJC, TJC or CRP. At last follow-up, 185/256 (72.3%) patients remained on RTX-B [median (range) RTX-B cycles 2 (1-4)]. RTX-B retention rate estimates were 79.4% and 71.3% at 12 and 18 months respectively. Following RTX-B discontinuation, 33/256 patients (12.8%) switched back to RTX-O [LOR=30 (11.7%), adverse effects (AEs)=3 (1.2%)], 13/256 (5.1%) started other b/tsDMARDs, and 25/256 (9.7%) stopped treatment (no longer indicated=7, deaths=7, loss to follow-up=6, patient choice=3, new contraindication=2). Of 46 patients with LOR/AE to RTX-B, 39 discontinued after cycle 1 (C1), 5 after C2 and 2 after C3. 31/33 patients switched back to RTX-O remained on treatment but follow-up was limited [mean (SD) 7.5 (5.2) months; 23/33 had 1 cycle only]. Compared with patients treated with other b/tsDMARDs, those switched back to RTX-O had more previous RTX-O cycles [mean (SD) 6.67 (3.28) versus 2.62 (2.10); p<0.0001]. Compared with patients who remained on RTX-B (n=185), those who discontinued for LOR/AE/death (n=53) had more previous bDMARDs (mean 1.26 vs 0.83; difference 0.43, 95% CI 0.10-0.77, p=0.01), but no differences in other clinical characteristics, treatment or B-cell numbers.Conclusion:Our real-world single-centre experience of non-medical switch from RTX-O to RTX-B in RA showed approx. three quarters of patients remained on RTX-B and maintained stable disease activity. However, apparent LOR occurred in 16%. Of >10% switched back to RTX-O, the majority remained on treatment at short-term follow-up.