Total Body Irradiation Versus Chemotherapy Conditioning before Allogeneic Hematopoietic Stem Cell Transplantation for Pediatric Acute Lymphoblastic Leukemia Patients Aged 2 to 4 Years: A Retrospective Study of the EBMT Pediatric Diseases Working Part

Introduction: Allogeneic hematopoietic stem cell transplantation (HSCT) is the standard of care for pediatric acute lymphoblastic leukemia (pALL) patients with inadequate response to therapy or relapse. Children older than 4 years (yrs) of age, are conditioned with total body irradiation (TBI) and etoposide according to the results of the Forum trial, which demonstrated superior overall survival compared to chemotherapeutic conditioning regimens (CTx) in this age group. Due to several side effects of TBI, children under the age of 2 yrs usually receive CTx conditioning resulting in higher post-HSCT relapse. The optimal conditioning regimen for children between 2 and 4 yrs of age remains controversial. Patients and methods: In this retrospective analysis, we investigated the outcomes of TBI- vs. CTx-based conditioning in children of 2 to 4 yrs, undergoing alloHSCT for pALL in first (CR1) or second complete remission (CR2), transplanted between 2000 and 2012. The primary objective was the incidence of subsequent malignant neoplasms (SMN); secondary objectives included key alloHSCT outcome parameters. Results: Overall, 282 patients were included and the median follow-up (FU) was 13.2 yrs [CI95%: (12.8-13.8)]. Patients who received CTx-based conditioning (n=119) were younger [median age: 2.8 yrs (Q1 2.3; Q3 3.3) vs. 3.3 yrs (Q1 2.8; Q3 3.6); p<0.001] than patients receiving TBI (n=163). Additionally, the CTx group received more frequently peripheral blood stem cells (PBSC; 26.1% vs 10.4%), and cord blood (CB; 21.0% vs 17.2%) and less frequently bone marrow (BM; 52.9% vs. 72.4%) as stem cells source compared to the TBI cohort. Descriptive characteristics, including gender, Lansky score prior to HSCT, yr of transplant, disease remission status at HSCT, and donor type, were not significantly different. Within the TBI group the majority received a radiation dosage of 10-12 Gy (79.7%). CTx-conditioning was either based on busulfan (86.4%), treosulfan (10.1%) or fludarabine/melphalan (3.3%). A total of 20 SMNs (7%) were seen in the study population; 5 SMNs were identified in the CTx group: 1 osteosarcoma, 2 post-transplant lymphoproliferative disorders (PTLDs), 1 Non-Hodgkin lymphoma (NHL) and 1 thyroid carcinoma. One patient succumbed to his PTLD and the NHL patient had a fatal ALL relapse. In the TBI group 15 SMNs were identified: 2 glioblastomas, 1 neurinoma, 1 malignant melanoma, 1 osteosarcoma, 1 intestinal polyposis with signs of malignant transformation and 9 thyroid carcinomas. All, but the 2 glioblastoma patients, are still alive. The cumulative incidence of SMNs (with CI95%) was 2.5% (1.1-4.9) at 5 yrs, 5.3% (3.0-8.5) at 10 yrs, and 7.3% (4.5-11.1) at 15 yrs. Overall survival (OS) was significantly higher in the TBI group (10-yr OS: 68.4% vs 50.7%; HR: 1.79, p=0.004), with an increased relapse incidence (RI) in the CTx group (10-yr RI: 41.5% vs 22.1%; HR: 2.34, p<0.001). In line with these results, leukemia-free survival (LFS) was higher for TBI (64.8% vs. 44.0%; HR: 1.94, p<0.001). Graft-versus-host-disease (GvHD) related outcomes did not differ significantly between the two groups. However, recent transplants have been associated with a lower risk for chronic GvHD (HR {per 3-yr increment} 0.55 (CI95%: 0.41-0.74), p<0.001). Conclusions: Based on these results, TBI should be offered to very high-risk ALL patients above the age of 2 years due to its efficacy and the relatively low rate of SMNs. Despite the higher incidence of SMNs in the TBI group, the incidence remained within the expected range and is acceptable when considering the excellent outcomes regarding relapse and leukemia-free survival.