Outcomes of Acute Myelogenous Leukemia Patients Undergoing Haploidentical Hematopoietic Cell Transplantation with Post-Transplant Cyclophosphamide: Impact of Total Body Irradiation Versus Chemotherapy-Based Myeloablative Conditioning

Abstract

<h3>Background</h3> Optimal myeloablative conditioning (MAC) in the setting of haploidentical hematopoietic cell transplantation (haplo-HCT) with post-transplant cyclophosphamide (PTCy) is unknown. We studied the outcomes of total body irradiation (TBI) vs. chemotherapy (CT) based MAC regimens in acute myelogenous leukemia (AML) pts undergoing haplo-HCT and reported to EBMT. <h3>Methods</h3> The study included 1008 AML pts who underwent haplo-HCT during 2010-2018, following TBI (n=89, 9%) or CT (n=919, 91%) based MAC. Regimen intensity was defined by EBMT criteria and the cases with busulfan dose <9 mg/kg or TBI dose ≤6 gray were excluded. Fludarabine-TBI (78%) and Thiotepa-busulfan-fludarabine (56%) were the most common MAC regimens in TBI and CT-cohorts, respectively. Pts in TBI cohort were more likely to be younger (median age, 38 vs. 47 yrs, p<0.01) and receive BM graft (57% vs. 43%, p-0.01) Other pt, disease and transplant characteristics were similar in both groups. Median follow up for TBI and CT cohort were 25.4 and 20.7 months, respectively. <h3>Results</h3> In univariate analysis, day 100 incidence of acute GVHD (aGVHD) II-IV and III-VI was 22% vs. 27% (p-0.44) and 12% vs. 12% (p-0.92) in TBI and CT cohorts, respectively. Two-yr total and severe chronic GVHD (cGVHD) incidence were 42% vs. 27% (p<0.01) and 9% vs. 12% (p-0.33) in TBI and CT cohorts, respectively. Graft failure was reported in 2 (2%) and 65 (7%)(p-0.08) pts who received TBI and CT-based MAC, respectively. Three (1%) pts died due to graft failure in CT-group. Death from veno-occlusive disease was reported in 1(3%) TBI pt and 11(3%) CT pts. Death due interstitial pneumonitis was reported in 1 (3%) TBI and 8(2%) CT pts. One pt died from a second malignancy in each group. In multivariate analysis, TBI was associated with higher incidence of overall cGVHD [HR=1.81, 95% CI:1.14-2.87, p-0.01] compared to CT. There was no difference in other outcomes such as extensive cGVHD[HR=0.34, p-0.07], relapse incidence(RI) [HR=1.32, p-0.22], nonrelapse mortality(NRM) [HR=0.82, p-0.52] and leukemia free survival(LFS) [HR=1.09, p-0.62], overall survival(OS) [HR=1.07, p-0.71] and GVHD free relapse free survival (GRFS) [HR=1.02, p-0.92]. 2-yr LFS and OS of TBI and CT cohort were 45% vs. 49%,p-0.98 and 50% vs. 55%,p-0.79, respectively (figure 1). Factors impacting OS were pt age (per 10yrs, HR=1.1, p-0.02), KPS≥90 (HR=0.55,p<0.01) and advanced disease status before haplo-HCT (HR=2.35, p<0.01). In a subgroup analysis of pts <40 yrs old and pts with pre-HCT disease status CR2 or advance disease, no interaction was observed between type of MAC and RI, NRM, LFS, OS and GRFS. <h3>Conclusions</h3> In this large relatively homogenous cohort of AML patients who received haplo-HCT with PTCy, TBI based MAC was associated with higher incidence of overall cGVHD without impacting other transplant outcomes compared to CT based MAC. A prospective study is needed to validate these findings.