Outcome of Allogeneic HSCT after Chemo-Based Conditioning in Infants with Acute Myeloid Leukemia in First Complete Remission: A Multicenter EBMT-PDWP Study

Background: Pediatric patients younger than two years of age with acute myeloid leukemia (AML) commonly receive a chemotherapy-based myeloablative conditioning regimen before allogeneic hematopoietic stem cell transplantation (HSCT). The optimal choice of cytotoxic agents is still controversial.

Methods: A retrospective EBMT-registry based study was conducted to investigate the impact of different chemotherapy-based conditionings on the outcomes in young children. Children younger than two years of age receiving a first HSCT of bone marrow (BM), peripheral blood stem cells (PBSC) or cord blood (CB) from matched siblings (MSD) or unrelated donors (UD) in first complete remission (CR1) between 2000 and 2019 were included. Busulfan/Cyclophosphamide (BuCy) and BuCy/Melphalan (BuCyMel) were the most frequent combinations on which this analysis focused. The primary endpoint was leukemia-free survival (LFS). Multivariate analysis adjusting for differences between the conditioning regimens and risk factors influencing outcome was performed using the Cox's proportional hazards regression model.

Results: 289 patients (56% male) transplanted at a median age of 1.2 years (IQR 0.9-1.6) after BuCy (164, 57%) or BuCyMel (125, 43%) were included. 184 (64%) patients received BM, 71 (24%) CB and 34 (12%) PBSC from UD (201, 70%) and MSD (88, 30%). In-vivo T-cell-depletion (TCD) was performed in 160 (58%, missing data 14) of the HSCTs with anti-thymocyte-globulin (ATG, 153) or alemtuzumab (7). Ex-vivo TCD was performed in 13 (5%, missing data 3) of the HSCTs. Graft-versus-host-disease (GvHD)-prophylaxis was Cyclosporin-A-based in 90% of the HSCTs. Median follow-up (FU) was 4.9 years (95% CI 3.9-5.5).

After a median FU of 4 years, 4-y-LFS after BuCyMel (74.3%, 95% CI 65.1-81.4) was significantly better compared to BuCy (59.7%, 95% CI 51.2-67.2), hazard ratio (HR) 0.56 (95% CI 0.35-0.90, P=0.02). Overall survival (4-y-OS) after BuCyMel (77.2%, 95% CI 68.1-84.0) was significantly better compared to BuCy (66.6%, 95% CI 58.0-73.8), HR=0.58 (95% CI 0.35-0.97, P=0.04).

No significant differences were found in the probability of relapse (4-y-RI (whole cohort) 26.2% (95% CI 21.0-31.7), HR of BuCyMel 0.59 (95% CI 0.34-1.02), P=0.06), non-relapse mortality (4-y-NRM (whole cohort) 7.8% (95% CI 5.0-11.4), HR of BuCyMel 0.49 (95% CI 0.19-1.24), P=0.13) and incidence of acute grade II-IV GvHD at day 100 (day-100-aGvHD II-IV (whole cohort) 36.8% (95% CI 31.2-42.5), HR of BuCyMel 0.59 (95% CI 0.35-1.01), P=0.06). Incidence of chronic GvHD (4-y-cGvHD (whole cohort)) was 9.8% (95%-CI 6.3-14.2). The donor type had no significant influence on the outcome.

Conclusion: Bu-based conditionings of HSCT for infants with AML at high risk of relapse offer a high probability of cure. Conditioning with three alkylators (BuCyMel) resulted in better LFS and OS compared with two alkylators (BuCy) without significantly increasing the risk of both NRM and aGvHD. Future trials will evaluate the impact of the more recently introduced alkylator Treosulfan within the conditioning of HSCT in pediatric AML.