HLA Disparity Is Associated with Improved Outcomes in Hematological Malignancies After Family HLA-Mismatched /Haploidentical HCT but Not URD

Abstract

Abstract 2014▪▪This icon denotes a clinically relevant abstract Background:Previous reports have evidenced that donor-recipient HLA disparity is not associated with improved outcomes after URD hematopoietic cell transplantation (HCT). Some data indicates that Haploidentical (haplo) HCT may result in lower relapse and better outcomes than HLA-identical sibling transplantation in lymphoma. Unfortunately, HLA disparity coming along with severe GVHD and infectious diseases may counteract any benefits from this potentially intensive graft-versus-leukemia (GVL) effects. Here we investigate the impact of donor types on clinical outcomes in 233 patients with hematological malignancies from our single BMT Center. Patients and methods:The outcomes of 233 patients, median age 23 years (4–62), with leukemia undergoing HLA-matched sibling(MRD) (n=52), unrelated (URD) (n=52) and related HLA-mismatched/haploidentical (haplo) (n =129) HCT, performed during the same time period (2007–2012) in single Med Center were compared. Patients received either BUCY2 or BUFlu in HLA-identical sibling HCT, and BUCY + ATG or BUFlu +ATG in URD and haplo cohorts as conditioning regimens, followed by G-CSF mobilized unmanipulated marrow and/or peripheral blood (PB) transplantation. All HLA typing has been performed using high-resolution DNA techniques. In haplo cohort, a hundred-four patients were 3/6 identical, 4 were 4/6 identical and 21 were 5/6 identical at HLA-A,-B, and -DRB1. In URD cohort, thirty-one were 8/8 identical, 13 were 7/8 identical and 8 were 6/8 identical at HLA-A,-B,-Cw and -DRB1. The median number of infused nucleated cells and CD34+ cells were 7.5×108/kg and 5.8×106/kg, respectively. 100/129 haplo, 36/52 URD and 14/52 sibling HCT received HLA 3/6–5/6 matched unrelated CB or haplo BM co-transplantation on day –1 as “third party” cells. GVHD prophylaxis consisted of CSA, MMF and short-term MTX. Results:The median follow-up was 23 mon. (range, 2.5mon.-59.2mon.). All patients engrafted to ANC exceeding 0.5× 109/L at 13.5 d (range, 10–19 d) in matched HCT and URD vs. 125 patients engrafted in haplo patients at 13.5d (8–25d). All patients achieved platelet engraftments in both matched and URD cohorts, at 13.5 d (10–22 d) vs. 124 patients engrafted at 13.5d (9–36 d) in haplo, respectively. Univariate analyses showed that the probability of grades 1–4 acute GVHD at 100 days in MRD, URD and haplo were 42%((95% confidence interval [CI], 28%-56%) vs. 48% (CI, 34%-62%) vs. 63%(CI, 54%-71%, P=0.006), respectively. But aGVHD3-4 were only 10%(CI, 3%-21%) vs. 25%(CI, 14%-39%) vs., 19%(CI, 12%-26%, P=0.155), respectively. The 3-year relapse and TRM rates were 34.6% vs. 13% vs. 17.3%(P=0.0126) and 12.7% vs. 30.2 vs. 21%(P= 0.0496) for patients who underwent MRD, URD and haplo transplantation, respectively. The 3-year RRM rates were 29.2% vs. 24.7% vs. 11.6%(P=0.018)after matched vs. URD vs. haplo, respectively. The 3-year Over all survival(OS)probabilities were 61.8% (range, 48.5%-78.4%), 52.5% (38.6%-71.9%) and 69.8 % (61.6%-79.1%) after the matched, URD and haplo HCT, respectively. There were no differences in survival probabilities according to donor types (P= 0.2743). But further analyses indicated that the 3-year OS for patients with early/intermediate disease were 71.8%, 57.3% and 80.2%, P=0.0401, while patients with advanced disease were 36%, 31.8% and 44.2% (P=0.9633)after the matched, URD and haplo transplantation, respectively. Tests in the multivariate analyses indicated that diseases in advanced stage, long period between diagnoses and transplantation, as well as acute GVHD 3–4 were all correlated with lower survival. Conclusion:These data suggest that HLA disparity in Haplo HCT results in relatively lower relapse and RRM, accepted TRM and improvement of survival for hematological malignant patients, especially in early and intermediate disease. The GVL effect using family HLA mismatched donors may be superior to HLA-identical siblings HCT. URD HCT also have lower relapse but do not associated with improvement of OS because relatively high TRM. Family –mismatched /haploidentical HCT should not be the last choice for these patients with Hematological Malignancies. Choosing the best haploidentical donor and further reduction of relapse and infectious complications in the future will lead to better clinical outcomes. Studies with more patients enrolled are undertaken. Disclosures:No relevant conflicts of interest to declare.