Early CD4 and B Cell Immune Reconstitution As Predictors of Outcomes in T Cell Depleted Allogeneic Hematopoietic Cell Transplantation

Introduction We recently showed in a large pediatric tri-institutional cohort of patients receiving a first allogeneic hematopoietic cell transplantation (allo-HCT) for a hematologic malignancy, that, in addition to recovery of CD4 > 50 cells/uL, also recovery of B cells > 25 cells/uL was found to be a predictor for outcomes. The impact of B cell immune reconstitution (IR) in a purely ex vivo T-cell depleted (TCD) setting has not been described. Objectives We analyzed the impact of B cell IR in a cohort of patients who received an ex vivo TCD allo-HCT for either a non-malignant or malignant indication. Methods We retrospectively analyzed data from consecutive patients receiving their first ex vivo CD34+ TCD allo-HCT for any indication at our institution for the period 2008-2018. Statistical analyses involved linearity evaluation using martingale residuals plots, as well as maximally selected log-rank statistics to identify cutoffs related to outcomes. For statistical analyses cox proportional hazard models and Fine-Gray competing risk methods were used. All statistical analyses were done using R statistical software, version 4.3.1. Results 326 patients with sufficient CD4 and B cell data in the first 100 days after allo-HCT were included. Median age for the full cohort was 40.3 (range 0.1-73.3) years; 84 (25.8%) patients were <18 years old. Diagnoses included acute myeloid leukemia (n=125), acute lymphoblastic leukemia (n=71), myelodysplastic syndrome (n=55), non-malignant (n=34), and other malignant (n=41) conditions. There was non-linearity between maximum B cell count before day 100 after HCT and non-relapse mortality (NRM). Log rank statistics demonstrated a B cell count of 199 cells/uL as the best cut-off to discriminate for risk of NRM (fig 1a); based on which we defined B cell IR as B cell > 200 cells/uL before day 100 for this cohort. Achieving both CD4 and B cell IR was associated with higher 5-year overall survival (OS) compared to only CD4 IR (86.8% versus 72.4%, p=0.036; fig 1b), but with similar 2-year cumulative incidence of NRM (2.3% versus 3.7%, p=0.062; fig 1c) and relapse (17.1% versus 21.3%, p=0.55; fig 1d). Compared to only CD4 IR, patients with only B cell IR had lower 5-year OS (56.6% versus 72.4%, p=0.018; fig 1b) and higher NRM (27.6% versus 3.7%, p<0.001; fig 1c). Conclusion In our analyses we show that B cell IR was not an additional predictor for outcomes in this cohort of mostly adult patients undergoing ex vivo CD34+ TCD allo-HCT. This was in contrast with our previous results in a cohort of pediatric/young adult patients, that included mainly T-replete allo-HCT, highlighting the importance of separately evaluating the impact of IR on outcomes in different allo-HCT platforms.