Retrospective review of cell doses in haploidentical hematopoietic stem cell transplant and associated outcomes

Abstract

Background & Aim Hematopoietic progenitor cell (HPC) transplant (HCT) products are routinely analyzed for infused dose of CD3+ T cells and CD34+ HPCs. Data on cell dose vs. patient (pt) outcomes in haploidentical (haplo) HCT are incomplete, preventing clear guidelines for dosing and storage of residual product. Some infuse a max 5-10 × 10^6 CD34+ cells/kg based on prior studies, but cutoffs haven't been investigated directly. This affects operations including collection time and freezer capacity. We analyzed cell products of haplo HCT pts at Ohio State University (OSU) to correlate to pt outcomes and identify cutoffs for cell dose. Methods, Results & Conclusion 845 pts underwent allo HCT at OSU from 2011-2019, 87 were haplo HCT. We excluded pts with ex vivo product manipulation, and pts undergoing 2nd HCT. Kaplan Meier method was used to calculate survival rates and Cox proportional hazards models to estimate the effect of cell dose on outcomes. 79 pts met criteria. 47 underwent haplo HCT with an HPC, Marrow (HPC,M) product; 32 with a mobilized HPC, Apheresis (HPC,A) product. HPC,M pts were separated into 2 groups by a median CD34+ cell dose. The low dose group received 1.1-3.6 × 10^6 CD34/kg compared to 3.7-8.1 × 10^6 CD34/kg in high dose. No significant differences in acute GVHD (aGVHD), chronic GVHD (cGVHD), relapse, progression free survival (PFS) nor overall survival (OS) were found. Analysis of HPC,M CD3+ cell dose found no differences. HPC-A pts were separated by Lo (3.7-8.3 × 10^6) vs Hi (8.04-10.9 × 10^6) CD34/kg and Lo (0.8-2.1 × 10^8) vs Hi (2.1-4.4 × 10^8) CD3/kg. We found no statistically significant differences between groups, but noted some interesting trends. The Hi CD34 group trended towards better OS w/ a hazard ratio (HR) of 0.13 compared to the Lo CD34 group (p=0.06), and better PFS (HR 0.27, p=0.07). Analyzed as a continuous variable, 1 × 10^8 CD3/kg increase in HPC,A was associated with reduction in relapse (HR 0.14, p=0.01), and trended towards improved PFS (HR 0.4, p=0.08) but increased risk of aGVHD (HR 1.5, p=0.07). Our analysis, though underpowered, showed recipients of haplo peripheral blood grafts had lower relapse rate with higher CD3+ dose and a trend towards improved OS and PFS with higher CD34+ dose. We will verify and study these findings in a larger cohort via the CIBMTR database. In this way we hope further standardize cell collection, infusion and storage protocols and maximize patient safety.