Busulfan, Cyclophosphamide and Melphalan As Conditioning Regimen for Pediatric Patients with AML in 1st or 2nd CR: A Retrospective Analysis from the AIEOP HSCT Registry

Allogeneic stem cell transplantation (HSCT) is largely adopted as post-remissional therapy in children with acute myeloid leukemia (AML) in first complete remission (CR1) but with high-risk characteristics (including high-risk cytogenetics or high levels of minimal residual disease at the end of induction therapy) or to rescue patients reaching CR2 after a previous relapse.

Busulfan-based regimens represent the standard of care for these patients in association with alkylating agents. One of the most frequent drugs combination used in Europe in pediatric patients is Busulfan, Cyclophosphamide and Melphan (BuCyMel), which provide a potent anti-leukemic effect, despite remarkable extramedullary toxicities, especially in adolescents.

We aimed at analyzing the results of children with AML receiving BuCyMel and reported to the AIEOP registry from 2008 to 2015. A total of 182 patients were reported by 15 transplant centers. Median age at HSCT was 9 years (range 0.3-18); 100 patients (55%) were male. Disease status at HSCT was CR1 in 159 (88%) patients and CR2 in the remaining 23 (12%). All patients received the same myeloablative conditioning regimen with BuCyMel and GVHD prophylaxis was mainly based on calcineurin inhibitors, with the addition of methotrexate in unrelated donors recipients. In vivo T-cell depletion/modulation with ATG was used in 90 cases (49.5%). In almost all cases, pharmacokinetics monitoring of Busulfan was performed, with the drug dosage adjusted according to the systemic exposure evaluated after the first dose.

Donor type was an HLA-matched family donor (MFD) in 82 (45%) patients and an unrelated donor (UD) in 100 (55%); 154 (85%) patients received bone marrow (BM) as stem cell source, while the remaining patients (15%) were transplanted with peripheral blood stem cells (PBSCs).

Median follow up for surviving patients was 39 months (range 1-111).

All patients achieved neutrophil engraftment. The cumulative incidence (CI) of grade II-IV and grade III-IV aGVHD was 35% (95%CI 28-42) and 11% (95% CI 7-16), respectively. The CI of aGVHD was not different according to the type of donor, being 37% (95%CI 28-50) and 32% (95%CI 24-46) in MFD and UD, respectively (p=0.38). The CI of chronic GVHD at 3 years was 17% (95%CI 12-24), while that of extensive cGVHD was 6% (95%CI 3-10). No difference was found in the CI of CGVHD according to the donor employed (MFD 15% and UD 19%, p=0.49).

Overall, the CI of relapse and non-relapse mortality (NRM) at 3 years was 18% (95%CI 12-26) and 15% (95%CI 10-22), respectively. The CI of relapse and NRM was significantly different according to age at HSCT (using 12 years as cut-off): (Relapse age<12y: 21% (95%CI 15-32) and age>12y: 11% (95%CI 3-32), (p=0.003); NRM age<12y: 10% (95%CI 5-20) and age>12y: 24% (95%CI 15-37), (p=0.005).

According to disease status at HSCT the CI of relapse and NRM were as follows: Relapse: CR1: 18% (95%CI 18-26), CR2 15% (95%CI 5-41) p=0.90) and NRM CR1: 14% (95%CI 9-21), CR2 19% (95%CI 8-46) p=0.38).

Also, there was no difference in relapse and NRM by donor type, relapse: MFD 16% (95%CI 9-28), UD 19% (95%CI 11-32) p=0.38) NRM: MFD 19% (95%CI 11-34), UD 11% (95%CI 7-20) p=0.62).

Causes of deaths were disease recurrence (39%), infections (27%), and GVHD (12%).

Three- years overall survival (OS) and disease-free survival (DFS) were 74% (95%CI 67-81) and 68% (95%CI 60-70). DFS was 70% (95%CI 60-77) and 67% (95%CI 47-87) for patients transplanted in CR1 and CR2 respectively, (p=0.39); and was 70% (95%CI 59-81) and 65% (95%CI 53-78), p=0.77, for UD and MFD HSCT recipients, respectively.

In conclusion, our results confirm the efficacy of BuCyMel in preventing relapse in a large series of pediatric patients affected by AML in CR1 and CR2. Adolescents represent a population of more fragile patients at risk of developing transplant-related fatalities. Optimization of toxicity profile and supportive care could further improve outcomes. Prospective randomized clinical trials are warranted to assess the best conditioning regimen for children and adolescents with AML.