Analysis of long-term mortality after total body irradiation-based and melphalan-based chemotherapy conditioning for acute myeloid leukemia

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative treatment option for selected patients with acute myeloid leukemia. Yet, the influence of total body irradiation (TBI)-based conditioning as compared to non-TBI-based conditioning on long-term mortality is unclear. We retrospectively evaluated outcomes after TBI-based (n = 91) and non-TBI-based conditioning (melphalan-based, n = 248) for 1st allo-HSCT patients transplanted at the University Hospital Regensburg between 1999 and 2020. TBI was performed with an average dose rate of 4 cGy/min. Median follow-up was 8.3 years (interquartile range, 4.8-12.9 years). Cumulative incidence rates of 5-year non-relapse mortality (NRM) were 17% (95% confidence interval, CI, 10-25) and 33% (95% CI, 27-40) after TBI- and non-TBI-based conditioning (P < 0.001). Five-year cumulative incidences of relapse (CIR) were 42% (95% CI, 32-52) and 29% (95% CI, 23-35) after TBI- and non-TBI-based conditioning (P = 0.030). The 5-year OS was 54% (95% CI, 43-64) and 55% (95% CI, 48-62) after TBI- and non-TBI-based conditioning. Both groups had similar 100-day acute graft-versus-host disease (aGVHD, 43% vs. 40%) and 5-year chronic GVHD (34% vs. 36%). The multivariable regression models found no associations of TBI with the outcomes NRM, CIR, PFS, OS, aGVHD, and cGVHD. TBI was no risk factor for NRM, even including mortality caused by secondary malignancies. NRM was influenced by patient age, advanced disease status, and the use of female donors for male recipients. TBI- and non-TBI-based conditioning appear to be equally effective and tolerable for AML patients eligible for 1st allo-HSCT.

Automated summary

This study compared the outcomes of allogeneic hematopoietic stem cell transplantation (allo-HSCT) patients with acute myeloid leukemia who received total body irradiation (TBI)-based conditioning or non-TBI-based conditioning. The results showed that both TBI and non-TBI conditioning had similar effects on non-relapse mortality, relapse incidence, and overall survival. The study suggests that both conditioning methods are equally effective and well-tolerated for eligible AML patients undergoing their first allo-HSCT.