Abstract
T cell depletion by CD34(+) cell selection of hematopoietic stem cell allografts ex vivo reduces the incidence and severity of GvHD, without increased risk of relapse in patients with acute leukemia in remission or MDS. The optimal candidate for CD34(+)-selected HCT remains unknown, however. OBJECTIVE: To determine outcomes based on both disease- and patient-specific factors, we evaluated a prognostic model combining the Disease Risk Index (DRI) and Hematopoietic Cell Transplantation Comorbidity Index (HCT-CI), an approach recently shown to predicted overall survival in a broad population of allograft recipients (1). METHODS: This was a retrospective analysis of 506 adult recipients of first allogeneic HCT with CD34+ selected PBSCs from 7/8- or 8/8-matched donors for AML (n = 290), ALL (n = 72), or MDS (n = 144). The Kaplan-Meier method estimated OS and RFS. The cumulative incidence method for competing risks estimated relapse and non-relapse mortality (NRM). We evaluated the univariate association between variables of interest and OS and RFS using the log-rank test. Cox regression models assessed the adjusted effect of covariates on OS/RFS. RESULTS: Stratification of patients based on a composite of DRI (low/intermediate vs. high/very high) and HCT-CI (0-2 vs. ≥ 3) revealed differences in OS and RFS between the 4 groups. Compared with reference groups of patients with low/intermediate DRI and low or high HCT-CI, those with high DRI had a greater risk of death (HR 2.30; 95% CI 1.39, 3.81) and relapse or death (HR 2.50; 95% CI 1.55, 4.05) than patients with any HCT-CI but low/intermediate DRI (HR death 1.80; 95% CI 1.34, 2.43; HR relapse/death 1.68; 95% CI 1.26, 2.24). CONCLUSIONS AND CLINICAL IMPLICATIONS: A model combining DRI and HCT-CI predicted survival after CD34(+) cell-selected HCT. Application of this combined model to other cohorts, both in retrospective analyses and prospective trials, will enhance clinical decision making and patient selection for different transplant approaches. DATA AVAILABILITY STATEMENT: The data that support the findings of this study are available on request from the corresponding author, C Cho. In order to protect the privacy of research participants, the data are not publicly available.