The influence of various doses of busulfan in conditioning regimes on outcome of allogeneic hematopoietic stem cell transplantation in children with acute myeloid leukemia

Abstract

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a potentially curative therapy for patients with acute myeloid leukemia (AML). The conditioning regimen administered for this patient based on busulfan (Bu) combined with cyclophosphamide (Cy), fludarabine (Flu) or some other agents. Comparisons of myeloablative conditioning (MAC) versus reduced intensity conditioning (RIC) have demonstrated a various results between relapse and toxicity in a few reports. We suppose, that dose intensity of Bu across regimens may affect treatment outcomes. Aim of this retrospective study was to evaluate the impact dose of busulfan to overall survival (OS), transplant-related mortality (TRM), relapse-free survival (RFS), toxicity, the incidence of primary graft failure and acute "graft versus host" disease (GvHD) in transplantation in children and adolescents with AML. The study was approved by the Independent Ethics Committee and the Scientific Council of the I.P. Pavlov First Saint Petersburg State Medical University. We analyzed 110 AML pediatric patients with the median age 9 (range 1–19) y.o., who underwent first allo-HSCT with Bu based conditioning in R.M. Gorbacheva Memorial Institute from 2002 to 2018. Patients were divided into 3 groups: Bu1 – patients, who received Bu at the dose 8–10 mg/kg, n = 34 (31%), in Bu2 – dose of Bu was 12 mg/kg, n = 35 (32%), in Bu3 – dose of Bu was > 12 mg/kg, n = 41 (37%). In Bu1 Bu was combined with Flu in 31 (91%) pts and Cy in 3 (9%); in Bu2 – with Flu in 12 (34%), Cy in 7 (20%) and other agents in 16 (46%); in Bu3 – with Cy in 32 (78%), with Flu in 7 (17%) and other agents in 2 pts (5%) (p < 0.001). Patients in Bu2 received more Cy based GvHD prophylaxis regimens (69% vs 44% in Bu1, vs 29% in Bu3, p = 0.003) and more haplo grafts (51% vs 29% in Bu1, vs 15% in Bu3, p = 0.003). The complete remission at the HSCT was observed in 79 % in Bu1, 49% in Bu2, 61% in Bu3 (p = 0.02). Probabilities of OS, RFS, TRM were estimated by using the Kaplan–Meier method. Incidence of toxicity, acute GvHD and primary graft failure – by using Mann–Whitney U-test. Transplant engraftment was achieved in 95 (86%) of patients. Graft failure occurs in the 5 patients of Bu1 group (15%), in the 6 pts of Bu2 (17%) and in the 4 pts of Bu3 (10%) (p = 0.7). Median follow-up was 2 years for Bu1 and Bu3, 1 year for Bu2. Two-year OS was similar (Bu1 = 59% vs Bu2 = 60% vs Bu3 51%, p = 0.7). Two-year OS of pts with CR before HSCT was 70% in Bu1, 82% in Bu2, 60% in Bu3, p = 0,3 and 14%, 39%, 38% for pts with progression disease (PD), respectively (p = 0.5). Two-year RFS was 74% in Bu1, 82% in Bu2, 64% in Bu3 at CR (p = 0.4); 43%, 39% and 38% in pts with progression, respectively (p = 0.9). Median of RFS were also similar for the pts in PD (4 months in Bu1, 5 months in Bu2 and Bu3, p = 0.9) and not achieved for pts at CR. Drug related toxicity grade III–IV 4 experienced in 35% pts in Bu1, 29% in Bu2, in 54% in Bu3 (p = 0.04). Mucositis and toxic hepatitis were the most common adverse events. Sinusoidal obstruction syndrome (SOS) experienced in 8 pts from different group: 4 from Bu2 (11%), 3 from Bu3 (7%) and only pts from Bu1 (3%) with previously treated of inotuzumab (p = 0.4). The most pts with VOD (3/5) had PD at the HSCT. Cumulative incidence of acute GvHD grade 2 (15% vs 14% vs 10%, p = 0.8) were not different. Acute GvHD grade III–IV was observed a bit more often in Bu3 (34%), than in Bu1 (18%) and Bu2 (17%) (p = 0.09). TRM up to D+100 was also higher in Bu3 (15%), than in Bu2 (6%) and Bu1 (0%) (p = 0.05). The transplant results of children with similar disease status of AML, received MAC or RIC conditioning with various dose of Bu, were not associated with significant differences in overall outcomes. The higher dose Bu may increase incidence of toxicity grade III–IV (p = 0.04) and acute GvHD grade III–IV (p = 0.09) with increasing of early TRM (p = 0.05).