Outcomes after Allogeneic Hematopoietic Stem Cell Transplantation for Pediatric Acute Myeloid Leukemia in the Contemporary Era

High risk or relapsed pediatric acute myeloid leukemia (AML) is potentially curable with allogeneic hematopoietic stem cell transplantation (HSCT). The use of a matched related donor (MRD) is the standard of care for these patients, but in the majority of patients a MRD is not available. Significant improvements have been made with supportive care practices, but there are limited outcome data for AML in pediatric patients transplanted in the contemporary era. We report retrospective results of HSCT in eighty-seven pediatric patients with high risk or relapsed AML performed at a single institution between 2008 and 2018.

The median age at the time of HSCT was 8 years old (range: 0.6-20 years). More than 2/3^rd of patients were in remission at time of HSCT with 43 patients (49%) transplanted in CR1 and 18 patients (21%) in CR2. Twenty six patients (30%) had active disease (defined as any evidence of disease) at the time of HSCT. Patients were transplanted using a matched related donor (MRD, n=21), a matched unrelated donor (MUD, n=25), a mismatched unrelated donor (MMUD, n=15), a haplo-identical donor (haplo, n=12) or an umbilical cord blood (UCB, n=14) graft. The majority of patients (78/87, 90%) received myeloablative conditioning while 9 patients (10%) received reduced intensity conditioning regimens. MRD and UCB graft recipients did not receive serotherapy. MUD, MMUD and haplo graft recipients (n=52) received serotherapy, the majority of whom received an alemtuzumab-based regimen, 42/52 (81%). Haplo-identical transplants were performed using CD34+ selected grafts.

Three-year overall survival (OS) and disease free survival (DFS) for the entire cohort were 52% (95% CI: 41-62%) and 49% (95% CI: 38-59%) respectively. OS differed significantly according to disease status at time of transplant (p=0.018), with 3 year OS: 60%, (95% CI: 43-74%) for patients in CR1, 56%, (95% CI: 31-75%) for patients in CR2 and 34%, (95% CI: 17-52%) for patients with active disease. OS differed by donor type (p=0.058). OS was comparable for patients with MRD (71%, 95% CI: 46-86%) and 10/10 MUD (59%, 95% CI: 37-75%) (p=0.67). Patients with alternate donor sources had worse outcomes (MMUD 33%, 95% CI: 10-59%; UCB 43%, 95% CI: 18-66%; Haplo 33%, 95% CI: 10-59%) as compared to MRD (MMUD vs MRD, p=0.042, UCB vs MRD, p=0.046, haplo vs MRD p=0.017). There was a low incidence of Grade III-IV GVHD (8%, n=9) and extensive chronic GVHD (9%, n=7) likely secondary to the use of alemtuzumab-based conditioning regimens for MUD, MMUD and haplo donors. The 3-year non-relapse mortality was low at 11%, (95% CI: 5-19%). The primary cause of death was relapse, with a 3-year cumulative incidence of relapse of 40%, (95% CI: 30-50%). Patients with active disease at transplant had significantly higher cumulative incidence of relapse (sHR: 2.42, 95% CI: 1.18-4.95, p=0.016) compared to patients in remission. There was no difference in cumulative incidence of relapse between patients in CR1 and CR2 (p=0.598).

In the contemporary era, outcomes after allogeneic HSCT are comparable for MRD and 10/10 MUD for pediatric AML. Myeloablative conditioning regimens that incorporated alemtuzumab were well tolerated with low rates of NRM and very low rates of acute and chronic GVHD. Relapse remains the major cause of treatment failure and future trials evaluating use of cellular therapies and targeted agents post-HSCT will be needed to improve current rates of relapse.