Full Haplotype Mismatch Stem Cell Transplantation: Myeloablative with T-Cell Depletion Vs. Reduced Intensity Conditioning without T-Cell Depletion

Abstract 3533

Some groups have continuously tried to reveal the possible role of full-haplotype mismatch stem cell transplantation (SCT) in patients who lack a compatible donor. However, the long-term event-free survival (EFS) rate after high-grade mismatch SCT using conventional myeloablative conditioning with T-cell depletion (TCD), even in complete remission (CR), is still only in the range of 10–30%. Most patients who were transplanted had heavily pretreated acute leukemia. The most important obstacle to overcome is the high rate of infection-related mortality associated with delayed immunological recovery in the setting of full-haplotype mismatch SCT, specifically by way of complete depletion of donor T cells in the graft. In contrast, by using unmanipulated donor cells and less aggressive conditioning regimens without TCD, most groups in Asia have reported EFS >40%. We investigated the role of reduced-intensity conditioning without TCD in HLA-mismatched related-donor SCT, specifically in patients with high-risk acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) who received unmanipulated granulocyte colony stimulating factor (G-CSF)-mobilized peripheral blood stem cells (PBSCs). SCT was performed in 32 consecutive patients. We compared two different protocols: one with more intensive conventional conditioning with TCD (group 1), and one with reduced intensity without TCD (group 2). The protocols differed between the two groups with respect to total body irradiation (1200 cGy in group 1 vs. 800 cGy in group 2) as well as busulfex, fludarabine, and anti-thymocyte globulin (ATG; thymoglobulin, Genzyme) treatments. No patient in group 1 received post-transplant graft versus host disease (GvHD) prophlyaxis or G-CSF. In contrast, all patients in group 2 received 1.25 mg/kg/day ATG for 4 consecutive days, together with our standard GvHD prophylaxis regimen of methotrexate (5 mg/m² intravenous bolus on days 1, 3, 6, and 11) and tacrolimus starting at day –1. All patients received G-CSF-mobilized PBSCs. The median number of CD34⁺ cells infused was 14.2 (range, 4.3–20.6) × 10⁶/kg for group 1 and 6.0 (range, 3.6–8.5) × 10⁶/kg for group 2. The median numbers of CD3⁺ cells infused were 0.03 (range, 0.015–0.057) × 10⁶/kg and 870 (range, 84–1260) × 10⁶/kg, respectively. G-CSF was administered subcutaneously to all patients in group 2 at a dose of 5 mg/kg/day, from day 7 after transplantation until neutrophil recovery. GvHD, relapse, non-relapse mortality (NRM), overall survival, EFS, reconstitution of immunity, and natural killer (NK) cell alloreactivity were compared between the groups. The median patient age was 33 (range, 17–48) years for group 1 and 39 (range, 19–63) years for group 2. The median follow-up for surviving patients was 18 months (range, 8–100). The majority of patients had intermediate or unfavorable cytogenetic features. The transplanted patients were all successfully engrafted. The median times to neutrophil (>0.5 × 10⁹/kg) and platelet (>20 × 10⁹/kg) recovery were 12 and 13 days in group 1, and 11 and 10 days in group 2. The overall rates of acute and chronic GvHD were 50% and 33%, and 80% and 92% in groups 1 and 2, respectively. Of note, NRM differed significantly between the groups: 46.2% in group 1 vs. 9.5% in group 2 (P = 0.0014). The estimated probability of EFS at 2 years was 15.3% for group 1 vs. 59.2% for group 2 (P = 0.05). We did not find NK alloreactivity in any of the patients, based on the NK-killer cell immunoglobulin-like receptor (KIR) disparity between donor and recipient. However, we noted a significant difference in EFS between C1/C1 homozygotes and C1/C2 heterozygotes according to the donor NK-KIR ligands (P = 0.0175). Recovery of CD4⁺ cell numbers at 3 months after SCT showed a markedly different pattern between the groups, with a median of 4 cells/μl (range, 2–7) in group 1 vs. 311 cells/μl (range, 56–1226) in group 2. Our findings suggest that full-haplotype mismatch SCT using a Korean-adapted protocol is a feasible therapeutic strategy for patients with high-risk AML/MDS, specifically in CR, provided that there is a further defined plan for the investigation of NK/T-cell alloreactivity.