POS0646 NON-MEDICAL SWITCHING FROM TOCILIZUMAB TO SARILUMAB IN RHEUMATOID ARTHRITIS PATIENTS WITH LOW DISEASE ACTIVITY

Abstract

BackgroundTocilizumab (TCZ) and sarilumab (SRL) are IL6-receptor antagonists registered for the treatment of rheumatoid arthritis (RA). Data from extension of the ASCERTAIN trial showed that patients responding to blinded intravenous (IV) TCZ rarely experienced a loss of response when switching to open label SRL [1]. This within-class, between-drug switch has gained relevance given the recent pandemic-driven shortage of several bDMARDS including TCZ.ObjectivesTo assess the efficacy and persistence of switching to sarilumab in RA patients responding well to tocilizumab.MethodsSAARTOOS (SArilumab Actively Replacing TOcilizumab, an Open label Study) is an open, observational single arm study. RA patients doing well (DAS28-CRP<2.9 or <3.5 with clinical judgement of low disease activity) on a stable dose (>6 months) of TCZ were offered to switched to SRL in clinical care (because of lower injection frequency). Patients who switched and consented were followed for 6 months. SRL was prescribed at a dose of 200mg. The SRL dosing interval was determined by doubling patient’s last TCZ dosing interval. All treatment decisions, including adjusting dose of or stopping SRL, were left to the treating physician.Co-primary outcomes were 1) the 90% confidence interval (CI) of DAS28-CRP change from baseline to month 6 compared to the non-inferiority margin of 0.6 and 2) the 90% CI of the proportion of patients persisting with SRL at month 6, compared to a pre-specified minimum persistence of 70% at month 6.Secondary outcomes included the CDAI, proportion of patients experiencing DAS28-CRP based flare (and post-hoc the proportion of patients remaining on SRL at month 6 without flare), changes in co-medication, expectations both patient and physician on SRL efficacy and tolerability.ResultsOf 50 invited patients, 25 agreed to attempt a switch to SRL, of whom 23 patients switched and were included. Switching and therefore inclusion was halted at this point due to an observed increase in disease activity and lack of SRL persistence. One patient was lost to follow up immediately after inclusion, therefore 22 patients are included in analyses (Table 1).Table 1.Baseline characteristicsBaseline characteristicsN=22Age, mean (SD)66 (11)Female, n (%)17 (77%)Rheumatoid factor positive, n (%)18 (82%)Anti-CCP positive, n (%)18 (82%)Erosive disease, n (%)14 (67%)2010 ACR/EULAR classification criteria, n (%)21 (95%)DAS28-CRP, mean (SD)1.9 (0.6)CDAI, mean (SD) (n=15)4.7 (3.1)Disease duration, years, median (IQR)20 (10-24)Duration of tocilizumab use, years, median (IQR)2.7 (2.1-7.3)Tocilizumab dose, n (%) 162mg per week10 (45%) 162mg per 10 days2 (9%) 162mg per 2 weeks9 (41%) 162mg per 3 weeks1 (5%)Concomittant csDMARD use, n (%)7 (32%)DAS28-CRP changed significantly from baseline to month 6: mean 0.48 (90% CI: 0.1 to 0.9). 15 of 22 patients maintained SRL at month 6, leading to a persistence of 68% (90% CI: 51% to 82%). Both co-primary outcomes therefore failed to meet their non-inferiority criteria.DAS28-CRP change was driven by both objective and subjective components. Median (IQR) CDAI at month 6 was 7 (3-10). During the study, 8 (36%) patients experienced 1 or more DAS28-CRP based flare. Only 11 (50%) patients remained on SRL for 6 months without experiencing flare.Expectations of patients and physicians at baseline were positive, with no patients or physicians expecting worsening of either disease activity or adverse effects prior to switching. Patient preference after using SRL (n=20, 2 missings) was inconsistent, with 5 (25%) patients strongly preferring TCZ, and 5 (25%) patients somewhat or strongly preferring SRL.In terms of (co-)medication, 4 (18%) patients required additional corticosteroids (oral or intramuscular), 5 (23%) patients switched back to TCZ and 2 (9%) switched to baricitinib.ConclusionNon-medical switching from tocilizumab to sarilumab in patients doing well on tocilizumab is associated with a non-negligible risk of flare and non-persistence.