Kinetics of early immune reconstitution and its exploratory association with post-transplant outcomes in patients undergoing haploidentical HCT with post-transplant cyclophosphamide-based GvHD prophylaxis

Introduction Post-transplant cyclophosphamide (PTCy) is widely used for graft-versus-host disease (GvHD) prophylaxis in haploidentical hematopoietic cell transplantation. However, the kinetics of early immune reconstitution in real-world, heterogeneous cohorts, and their relationship with early outcomes, remain incompletely defined. Methods In this prospective single-center study, 95 consecutive patients undergoing haploidentical transplantation with PTCy (October 2021 to December 2024) underwent serial multicolor flow cytometry before conditioning, at neutrophil engraftment, and on days +21, +28, +60 and +90. T, B and natural killer (NK) cells were enumerated, along with CD4+ and CD8+ naive, central memory, effector memory and terminal effector subsets. Regulatory T cells were assessed in 50 patients. An exploratory propensity-matched comparison was performed against a contemporaneous HLA-matched donor cohort. Results Immune reconstitution followed a sequential pattern: NK cells recovered earliest (from day +21), B-cell recovery was delayed, and T cells expanded gradually, dominated by central and effector memory subsets, while naive and terminal effector subsets remained low through day +90. Grade II to IV acute GvHD occurred in 32.6% of patients, and transplant-related mortality (TRM) was 26.3% by day +100. No lymphocyte subset was significantly associated with acute GvHD, TRM or overall survival, and reconstitution kinetics were similar between myeloablative and reduced-intensity conditioning. In the propensity-matched comparison, day +100 TRM was higher in the haploidentical cohort, whereas acute GvHD and overall survival were comparable. Discussion Early immune reconstitution after haploidentical transplantation with PTCy proceeds in a characteristic sequence, but quantitative recovery of circulating lymphocyte subsets alone did not predict acute GvHD or early mortality. Larger studies incorporating functional immune assays and time-to-event modelling are needed to clarify these relationships.