Abstract 25: During CB Transplant in MRD-Positive AML, Relapse is Usually Early, is Reduced by Acute GVHD and is Influenced by Ciclosporin Exposure: A Multicenter National Experience

Abstract Introduction We report a multicentric experience of the relationship between ciclosporin exposure and time to relapse in a large homogenous cohort of paediatric AML patients receiving a T-replete cord blood transplant(TRCBT), adjusted to pre-transplant MRD and GVHD. Objectives The primary endpoint was time to relapse, and its relationship between ciclosporin AUC, use of additional immunosuppressive treatment (IST) and its total duration was investigated by multivariate Cox regression analysis. Methods It is a retrospective study, analyzing data from 7 paediatric transplant centers in the UK on children who had a TRCBT for HR AML between 2013 and 2023. Case records were analyzed to get transplant characteristics, GVHD prophylaxis, trough ciclosporin levels in the first 8 weeks and the transplant outcomes including GVHD and IST used, relapse and transplant related mortality (TRM). Results The cohort included 124 patients, 66 were MRD-positive pre-transplant, with a median follow up of 737.1 days (range:24-3540 days). The mortality rate was 40.7% and three-quarters occurred in the MRD-positive group. The median time to relapse was 180 days (range: 32 to 949). TRM was 13% and was the most prominent cause of mortality in MRD-negative group.The RFS is greater in MRD-negative group without severe aGVHD than in those with severe aGVHD, though not statistically significant. The RFS is significantly lower in the MRD-positive group without severe aGVHD than in those with it. The COX proportional hazards model using 95% CI confirmed a significant inverse relationship of AUC of Ciclosporin in the first 8 weeks on the time to relapse (p<0.02) and such a significant relationship maintained even after adjusting for the MRD status (p<0.03) and the presence of severe aGVHD (p<0.016). There was a significant inverse relationship between duration of CNI and time to relapse (p<0.03). The study did not find any effect of IST or its duration on time to relapse (p=0.567 and p=0.241 respectively). Discussion This is the largest homogenous cohort of TRCBT in AML investigating the role of ciclosporin prophylaxis on relapse. It confirms the importance of GVL effect and its relation to GVHD, particularly in MRD-positive patients as RFS is significantly better in those who experience aGVHD.