Effect of graft cellular content on outcomes in HLA mismatched hematopoietic cell transplantation using post-transplant cyclophosphamide

Abstract

Abstract Background Since the introduction of post-transplant cyclophosphamide (PTCY) for graft versus host disease (GVHD) prophylaxis, the available donor pool for each recipient has substantially expanded, and in this context donor age remains an independent prognostic factor for allogeneic hematopoietic cell therapy (HCT) outcomes. We hypothesized that donor age impacts both quantitative and qualitative aspects of the infused grafts that in turn affect their activity both in the early and later periods of HCT. We sought to evaluate the graft characteristics that are impacted by donor age, and to correlate these with both early HCT events as well as later outcomes. Methods This is a single center retrospective study of patients undergoing HLA mismatched HCT using PTCY-based GVHD prophylaxis. Clinical data on baseline patient characteristics and disease characteristics as well as donor characteristics was collected on an IRB approved protocol. A flow cytometry panel evaluating CD3+, CD4+, CD8+, CD19+, CD56+, and CD34+ cell counts was applied to donor grafts prior to their infusion. Peripheral blood chimerism was performed following HCT per standard protocol. A non-parametric method (Spearman correlation, Wilcoxon rank sum test) was used to investigate the associations between graft content, donor characteristics, and incidence of cytokine release syndrome (CRS). Cumulative incidences and cause-specific hazards for relapse and acute/chronic GVHD in the presence of a competing risk were evaluated using the Gray method. Overall (OS) and progression-free survival (PFS) were evaluated with the log-rank test and Cox regression models. All p-values are 2-sided unless otherwise indicated. Results A total of 296 patients with graft content data were included in the study, of which 233 received a peripheral blood stem cell (PBSC) product and 63 received a bone marrow (BM) product. Among patients receiving either a PBSC or BM product, donor age was not independently associated with OS or PFS. On multivariable analysis, among patients receiving HLA mismatched PBSC (haploidentical, n=143, or HLA mismatched unrelated donor, n=90), donor age was independently associated with cumulative incidence of relapse (HR=0.97 (95% CI: 0.95, 0.99), p=0.020). Among PBSC grafts, younger vs older donors (age < 33 vs ≥ 33) had higher CD8+ (p < 0.001), lower CD4+ (p=0.014), and median CD4+:CD8+ ratio 2.0 vs 3.2 (p < 0.001). A similar, though numerically smaller, difference was observed with BM grafts (median CD4+:CD8+ in younger vs older donors: 1.2 vs. 1.5; p = 0.034). There was no significant correlation between donor age and CD34+ content, absolute total nucleated cell count, CD3-CD56+ or CD19+ content of the grafts. To evaluate whether the CD4+ and CD8+ contents of PBSC grafts translated to early post-HCT events, we correlated them with CRS and CD3+ T cell engraftment. The incidence of CRS was most positively correlated with the absolute CD4+ count (p < 0.001) and CD4+:CD8+ (p<0.001), but not CD8+ count (p=0.93) in the HLA mismatched PBSC grafts. Day +30 T cell donor chimerism >90% was not associated with any donor graft cellular content but was associated with the development of CRS (p=0.01). With respect to later HCT outcomes, on univariable analysis we found no association of any graft cellular PBSC graft content with OS, PFS, or incidence of relapse. However, higher PBSC graft CD4+ (1-sided p=0.025) and presence of CRS (1-sided p=0.020) were associated with development of any acute GVHD. Similarly, higher PBSC graft CD4+ (1-sided p=0.029), CD8+ (1-sided p=0.010) and presence of CRS (1-sided p=0.034) were associated with development of any chronic GVHD. Conclusions In a large cohort of HCT recipients receiving PTCY-based GVHD prophylaxis, the CD4⁺:CD8⁺ ratio was lower in grafts from younger donors. However, this age differential in cellular graft content did not account for the association of donor age with relapse in HLA mismatched PBSC transplants. Higher CD4⁺, but not CD34⁺, cell counts in PBSC grafts were associated with increased risk of CRS, and both acute and chronic GVHD, suggesting a potential interplay between graft T-cell composition and the early post-HCT cytokine milieu. Although these associations did not translate into differences in OS or PFS, evaluating graft cell content may help identify patients at higher risk of CRS or GVHD. Further studies characterizing the graft cellular content and immune reconstitution are ongoing.