Poor Graft Function (PGF) Following In Vivo T Cell-Depleted Allogeneic Stem Cell Transplantation: The Role of Delayed...

Poor graft function (PGF) is a severe early complication of allogeneic hematopoietic stem cell transplantation (HSCT), defined by persistent cytopenias despite full donor chimerism. Its characterization has been limited by heterogeneous diagnostic criteria, impairing reliable estimates of incidence, risk factors, and biological correlates. An internationally harmonized definition was introduced in 2024 which provided a standardized framework for clinical and biological investigations, facilitating comparisons across studies. To assess the incidence, risk factors, clinical outcomes, and immune reconstitution associated with PGF using the 2024 harmonized definition in a large adult HSCT cohort. We conducted a retrospective single-center study including 655 consecutive adult HSCT recipients between 2015 and 2025. PGF was defined according to 2024 harmonized criteria. The primary endpoint was cumulative incidence of PGF; secondary endpoints included overall survival (OS), nonrelapse mortality (NRM), cumulative incidence of relapse (CIR), and immune reconstitution. Risk factors were assessed using univariate and multivariate logistic regression. Time-to-event outcomes were analyzed using competing-risk models and Simon-Makuch methodology. Immune reconstitution (CD3⁺, CD4⁺, CD8⁺, CD16⁺, CD19⁺ subsets) was evaluated longitudinally using linear mixed-effects and time-point analyses. PGF occurred in 46 patients (12-month cumulative incidence 7%), with a median onset of 86 days post-HSCT. In multivariate analysis, myelofibrosis (OR: 7.25, P < .001), myelodysplastic syndromes (OR: 4.91, P = .002), major ABO incompatibility (OR: 2.32, P = .019), CD34⁺ cell dose <5 × 10⁶/kg (OR: 2.34, P = .022), and post-transplant cyclophosphamide (PTCy) (OR: 2.39, P = .022) were independently associated with PGF. Five-year NRM was higher in PGF (25% versus 15%, P = .04), predominantly due to infections, prompting evaluation of post-transplant immune reconstitution as a potential underlying mechanism. Indeed, PGF was associated with delayed immune reconstitution, with impaired recovery of CD3⁺, CD4⁺, CD8⁺, and CD16⁺ cells throughout the first year after transplant, independent of PTCy or antithymocyte globulin exposure. B-cell recovery was also impaired, with lower CD19⁺ counts and increased hypogammaglobulinemia rates at 1 year after transplant. PGF is an early, severe post-HSCT complication associated with increased infectious NRM. This study represents the largest analysis of immune reconstitution in PGF, demonstrating delayed recovery across lymphocyte subsets, which may contribute to the observed excess infectious NRM. Distinct disease- and transplant-related factors identify higher-risk patients, supporting improved pretransplant risk stratification, graft optimization, and tailored post-transplant monitoring strategies.