CD34-Selected, T Cell-Depleted Alternative Donor Stem Cell Transplantation For Children

Many children who need a hematopoietic stem cell transplant do not have a matched related or unrelated donor. The use of alternative donors including mismatched family members can provide a donor for almost all patients. The use of such donors is associated with a high risk of graft-versus-host disease (GVHD).

We present the results of a prospective study using the CliniMACS® device for CD34-selection of peripheral blood stem cell (PBSC) grafts to prevent GVHD under BB-IND 14045. Patients did not receive immunosuppression after transplant. The approach was also used as a platform for a companion post-transplant immunotherapy study. The study included patients with malignant (MD) and non-malignant (NMD) disorders receiving unrelated donor (UD) and mismatched related donors (MMRD) to evaluate the broad applicability of this approach for GVHD prevention. The primary endpoint was acute GVHD incidence and there was a stopping rule for primary graft failure. Overall survival was a secondary endpoint.

Between 2009-2013, 30 children underwent CD34-selected PBSC transplantation from an alternative donor. Fifteen patients had MD (ALL 4, AML 8, JMML 1, acute mast cell leukemia 1, Ewing’s sarcoma 1) and 15 had NMD (sickle cell 6, immunodeficiency 5, bone marrow failure 4). Twenty-three patients had MMRD and 7 had UD (all 7 had NMD).

The conditioning regimen consisted of TBI 1200 cGy (MD) or melphalan 140mg/m² (NMD), thiotepa 5mg/kg x2, fludarabine 40 mg/m²x5, and rabbit-ATG 1.5 (MD) or 2.5 (NMD) mg/kg x4. Seven patients with NMD received rituximab x1 during conditioning. Twenty-two patients received a planned (P) donor lymphocyte infusion (DLI) between days 30 and 42 with methotrexate GVHD prophylaxis on a companion study. Eight patients received a therapeutic (T) DLI for decreasing donor chimerism or viral infection.

The median age at transplantation was 10 years (range 0.3-17). Median CD34+ dose was 21 x 10^6/kg (range 10-25) and all patients received < 1 x 10^4 CD3/kg. The ANC was > 500 at median 14 days (range 9-16). Acute GVHD prior to DLI occurred in 1 patient (3%; stage I skin which resolved with a brief course of prednisone). Primary graft failure occurred in 1 patient (3%). The patient subsequently engrafted after a CD34-selected transplant from the other parent. Two patients who had active disease at transplant (mast cell leukemia, Ewing’s) and early relapse are excluded from the following analyses. Twenty-one of 28 patients are alive with median follow-up of 2 years (range 1 mo-4 yrs). The Kaplan-Meier estimated 100 day and 1 year survival is 96% and 74%, respectively. Relapse occurred in 1/10 MD patients with at least 100 days follow-up.

Grade II-IV acute GVHD and chronic GVHD occurred only after DLI. Acute GVHD grade II-IV occurred in 21% (3 P, 3 T) and chronic GVHD (3 P) in 11%. Viral reactivation was common, but viral disease was less common - CMV 12% and EBV-related PTLD 11% [at risk patients only] and adenovirus 11%. Invasive fungal infections occurred in 7%. All deaths were due to infection, with thrombotic microangiopathy present in 6/7. All outcomes were similar for MD and NMD and for MMRD and UD except for acute GVHD which occurred in 26% MMRD and 0% UD transplants and for EBV-PTLD which only occurred in NMD patients. EBV-PTLD has been reduced in NMD patients by using rituximab in the conditioning.

The use of a CD34-selected, T cell-depleted alternative donor PBSC transplant successfully prevented acute GVHD without the need for post-transplant immunosuppression. The use of this approach and the conditioning regimens employed provided reliable engraftment and a very low incidence of Day 100 transplant-related mortality and relapse (for patients without active disease). Future efforts will focus on approaches for post-transplant immunotherapy to decrease morbidity and mortality due to viral infections.