Rituximab dosing to prevent clinical relapse in immune thrombotic thrombocytopaenic purpura.

Abstract Introduction Pre-emptive rituximab (RTX) is used to treat ADAMTS13 relapse in immune TTP (iTTP) and prevent clinical relapse but the optimal dosing strategy is unknown. MethodsWe conducted a prospective, randomised single-blind UK multicentre, non-inferiority trial of low-dose vs standard dose RTX for prevention of relapses in iTTP. Patients with known iTTP and ADAMTS13 activity ≤15% with no features of acute TTP were randomly assigned in a 1:1 ratio to receive RTX 200mg or standard (375mg/m2) dose weekly for 4 doses. The primary outcome was time to retreatment with RTX/other immunosuppression. The key secondary outcomes were ADAMTS13 response; clinical relapse rate; B cell depletion/return and rituximab-related infusion reactions/delayed adverse effects. Follow up visits were Day 29, 3m, 6m and 6 monthly until primary outcome or end of study (minimum 2 yr foIlow up). Patients who met primary outcome and required retreatment could be re-randomised under a new trial number. Results From Sept 2017-Dec 2022, 70 patient episodes were randomised from 4 UK sites. 35 were assigned to low dose (including n=17 re-randomised episodes) and 35 to standard dose (n=15 re-randomised). 68 eligible patient episodes received at least one RTX dose. Baseline demographic 2/3rds had had prior elective RTX. Median time to retreatment with RTX/other immunosuppression was 19.7 months (90% CI 15.5, 27.2; n=28) with low dose RTX, and 20.1 months (17.2, 24.4; n=31) for standard dose HR = 0.93 (0.56, 1.53); p=0.798 in modified ITT population. The upper bound of the 90% CI of the HR was not below the pre-specified non-inferiority margin of 1.282. The per protocol and sensitivity analysis populations produced the same conclusion. Kaplan-Meier curves for the primary endpoint, plus further statistical testing, showed this was likely due to treatment effect being non-constant over time, with curves diverging initially with more retreatments prior to 12 months for low dose RTX and then crossing. The number of censored patients and re-randomised patients was comparable across trial arms and the 59 evaluable observations which contributed to the primary analysis was greater than the planned sample size (52). Secondary outcomes: Median time to ADAMTS13 normalisation was 31 days (90% CI 21, 79) with low dose RTX n=33 and 21 days with standard (21, 73) n=34, HR = 0.73 (0.48, 1.13) p=0.186. Two patients in each arm achieved ADAMTS13 partial remission only (ADAMTS13 20%); one patient in each arm was withdrawn at investigator decision due to lack of ADAMTS13 response. Median duration of ADAMTS13 response was 17.3 months (90% CI 15.0, 26.8) for low dose n=28 and 19.1 months (15.9, 22.3) for standard n=30 HR = 0.93 (0.56, 1.53) p=0.798. Relapse rate with an acute TTP episode during follow up was 2/32 (6.3%) for low dose RTX with clinical relapses at 5 months and 13 months; and 0 for standard dose (n=34). Re-treatment rate with anti CD20 was 19/28 (67.9%) for low dose n=28 and 24/31 (77.4%) for standard p=0.559. Median time to B cell depletion was 7 days (90% CI 7,7) n=32 both arms, HR = 0.97 (0.63, 1.49) p=0.844. Median time to B cell return (90% CI) was 12.2 months (11.5, 17.3) for low dose and 16.1m (12.7, 21.8) for standard HR = 1.56 (0.93, 2.61) p=0.151. Rates of infusion reactions and delayed RTX-related AEs per patient episode were similar between groups. There were more infusion reactions at day 1 infusion compared to subsequent infusions. Serum sickness occurred in 3 episodes with low dose RTX and 1 with standard. There was no signal for increased infections in either arm. Conclusion This is the first RCT in the prevention of acute episodes in this rare life-threatening disorder, providing valuable data for clinicians.There was no difference in median time to retreatment after low dose RTX compared to standard dose (19.7 vs 20.1 months). However, the treatment effect was not constant over time, with more retreatments observed in the first 12 months in the low dose RTX arm. This may be why the non-inferiority threshold was not met. There was no difference in the secondary efficacy outcomes between RTX doses. Overall clinical relapse rate was low across both trial arms with this protocol of close monitoring and pre-emptive therapy. Pre-emptive RTX is effective, safe and well-tolerated and remains so with repeated dosing over long follow up. (Funded by Answering TTP; CT 2017-001117-86)