Methods
Patients: This is a single-center retrospective review of consecutive pediatric and young adult patients with myeloid malignancies who underwent allo-HCT on the pediatric service from October 2012 to October 2021. All patients received myeloablative (MAC) conditioning with BU-FLU-MEL, which has been our pediatric hematopoietic transplant program’s standard of care conditioning for myeloid leukemias for the last ten years. Allo-HCT recipients met organ criteria allowing for MAC and had no evidence of active untreated infection. University of Arizona institutional review board approval was obtained to review and report our findings. Four of the eleven patients receiving haploidentical bone marrow transplantation (haplo-BMT) were treated as part of an institutional review board (IRB)–approved Phase Ia single-institution trial12, 13.
Conditioning regimens: Allo-HCT conditioning for matched sibling donor (MSD) and unrelated allo-HCT consisted of BU at 0.8 mg/kg IV every 6 hours for a total of 16 doses (days -9 to -6). BU pharmacokinetics of the first dose were performed at the Seattle Cancer Care Alliance laboratory. The seventh and remaining doses were modified to achieve an average area under the curve (AUC) of 3.9 to 4.4 mg x hr/L for the duration of the course. The estimated median exposure was 4.0 for MSD and 4.2 mg x hr/L for unrelated allo-HCT. FLU was given at 30 mg/m2 on days -5 to -2 and MEL 180 mg/m2 divided into three equal doses and administered on days -4 to -2. One patient with AML who underwent matched unrelated donor (MUD) peripheral blood stem cell transplant (PBSCT) also received rabbit anti-thymocyte globulin (rATG) 1 mg/kg on days -3 to -1. The other patient with AML (evolving from myelodysplastic syndrome [MDS]) who received a MUD PBSCT had 32% myeloblasts prior to conditioning and was therefore given a higher total dose of MEL at 200 mg/m2 split between days -3 and -2 without the addition of ATG11.
All eleven patients undergoing T-replete haplo-BMT received BU at 0.8 mg/kg IV every 6 hours for a total of 12 doses (days -8 to -6). BU pharmacokinetics of the first dose were performed with the seventh and remaining doses modified to achieve average AUC exposure of 4.0 to 4.5 mg x hr/L for the duration of the course (median was 4.3). BU was followed by FLU 30 mg/m2 for our first seven patients and FLU 40 mg/m2, for the later four, given on days -5 to -2 and a single dose of MEL 100 mg/m2 on day -212-15.
Graft-versus-host disease prophylaxis: All MSD and MUD PBSCT patients (n=9) received GvHD prophylaxis with methotrexate (MTX) 10 mg/m2 IV infused on days +1, +3, +6 and +11 and cyclosporine A (CSA) starting on day -2 and targeting levels of 150-200 ng/ml. The three umbilical cord blood (UCB) patients received mycophenolate mofetil (MMF) 15 mg/kg q 8 h IV between day -3 and +28 and CSA also starting on day -3. In all patients and in the absence of GvHD, CSA taper occurred between days + 90 and +180. Seven haplo-BMT patients received PT-CY 50 mg/kg IV on days +3 and +4 and another four patients that were part of an IRB–approved phase I single institution clinical trial through the University of Arizona Cancer Center (NCT02996773) received PT-CY/bendamustine (BEN). One patient (cohort 2) received PT-CY 50 mg/kg on day +3 and PT-CY 20 mg/kg followed by PT-BEN 60 mg/m2 on day +4. Another patient (cohort 3) was treated with PT-CY 50 mg/kg IV on day +3 and PT-BEN 90 mg/m2 on day +4. Two patients (cohort 4) received PT-CY 40 mg/kg on day +3 followed by PT-BEN 20 mg/m2on day +3 and PT-BEN 90 mg/m2 on day +4. All patients were started on MMF on day +5 until day +28 and tacrolimus from day +5, targeting levels of 6-10 ng/ml. In the absence of GvHD, tacrolimus was weaned starting day +70 to +90 and discontinued by day +120 to +180. GvHD was graded according to the consensus criteria for grading acute and chronic GVHD16, 17.
Supportive care: Levofloxacin bacterial prophylaxis was started when the absolute neutrophil count (ANC) dropped below 0.5 x 109/L. Antifungal prophylaxis with voriconazole was initiated following completion of BU administration. ForPneumocystis jirovecii prophylaxis IV pentamidine was given on day -1 and then twice monthly until day +45 and monthly afterwards until one year after BMT. Acyclovir was also started on admission for herpes simplex and varicella virus prophylaxis. Bi-weekly polymerase chain reaction (PCR) monitoring for cytomegalovirus (CMV) and weekly for adenovirus, Epstein-Barr virus (EBV) and human herpes virus-6 (HHV-6) were performed until discharge from hospital and subsequently at least every other week during first 100 days and at least monthly for another 6 months. All patients were transplanted in HEPA filtered rooms on a HEPA filtered unit and encouraged to walk laps on the unit daily.
Donor selection: MSD and MUD PBSC donors were 10/10 HLA-A, -B, -Cw, -DRB1, and -DQB1 antigen matched at high resolution and -DPB1 matched or permissive. The UCB units were a 5/6 HLA-A, -B, -DRB1 antigen match (single) with all units being 4/6 in the two double UCB transplants. Haploidentical donors were first degree relatives who were HLA-haploidentical based on high-resolution typing at HLA-A, -B, -Cw, -DRB1, and -DQB1. Five of the donors were 5 of 10 antigen matches, five were 6 of 10 and one was 7/10 (Table 1) . None of the patients had anti-donor HLA antibodies.
Engraftment and Donor chimerism monitoring. For UCB transplants and haplo-BMT granulocyte-colony stimulating factor (G-CSF) was started on day +1 and +5 respectively at 5 µg/kg/day until an ANC of 2.5 x 109/L was achieved for three consecutive days. Day of myeloid engraftment was defined as the first of three consecutive days with an ANC of 0.5 x 109/L. Day of platelet engraftment was considered the first of three consecutive days with platelet counts of > 20 x 109/L with no platelet transfusions administered in the previous 7 days.
Donor chimerism was evaluated on days +28 from bone marrow (BM) and +100, +180 and +365 from peripheral blood (PB) by short tandem repeats (STRs). Engraftment testing was performed using labeled primers to PCR-amplify STR polymorphic DNA markers followed by capillary electrophoresis to distinguish between the DNA contributed by the recipient versus the donor and estimate the percentage of the contribution. The Promega GenePrint 24 System which includes 24 polymorphic markers was used (Promega Corporation, Madison, WI)18.
Statistical analysis: Time to event endpoints were estimated using cumulative incidence curves and Kaplan-Meier curves, with comparisons using log-rank tests19, 20. Cox proportional hazards models for aGvHD II-IV, aGVHD III-IV, cGVHD and severe cGVHD also were adjusted for competing risks due to disease relapse and graft failure. Analyses and graphs were performed using GraphPad Prism version 9.4.0.