Background: The results of NMA, related HLA-haploidentical (haplo) BMT with PTCy are comparable to matched BMT (Blood 2015;125;3024). However, patients (pts) may lack related donors, often because of familial disease susceptibility or cytotoxic antibodies against donor HLA molecules. Although partially HLA-mMUD BMT could be considered if a MUD is unavailable, mismatches > 1 HLA antigen have historically been associated with unacceptable rates of GVHD, graft failure, and nonrelapse mortality (NRM). We hypothesized that PTCy may overcome this HLA barrier.

Methods: We developed a prospective, regimen-finding study of NMA BMT for hematologic malignancies pts without an acceptable matched or first-degree related haplo donor. The objective was to identify a regimen with ≤ 25% severe acute GVHD and ≤ 20% NRM by day 100, using mMUD's or non-first-degree relatives. Results of mMUD BMT with our standard conditioning are presented. Pt eligibility included age 0.5-75 years, no suitable family donor, ECOG PS ≤ 2, LVEF ≥ 35%, FEV1 and FVC ≥ 40% predicted, and adequate hepatic function. The mismatched donor shared ≥ 5/10 HLA alleles (based on molecular typing of HLA-A, -B, -Cw, -DRB1, and -DQB1) with ≥ 1 allele matched for a HLA class I gene and ≥ 1 allele matched for a class II gene. Donors matched at ≥ 1 allele at each locus were prioritized, followed by donors having the fewest number of mismatched loci. All pts received fludarabine (30 mg/m2 IV days -6 to -2), Cy (14.5 mg/kg IV days -6 and -5), TBI (200 cGy day -1) and T-cell replete bone marrow. GVHD prophylaxis consisted of high-dose PTCy (50 mg/kg IV days 3 and 4), mycophenolate mofetil on days 5-35, and either sirolimus (15 pts) or tacrolimus (1 pt) on days 5-180. GCSF was given from day 5 until neutrophil recovery. Safety was continuously monitored.

Results: Sixteen pts (median age 57, range 37-66) received mMUD BMT on this study from 7/2011-6/2014 (9 AML cases, 4 MDS or MPN, 1 CML, 2 PTCL). Major reasons for utilizing a mMUD included the absence of an adult haplo first-degree relative and the presence of prohibitive donor-specific anti-HLA antibodies (DSA's). The unrelated grafts had a median of 2 HLA mismatches and included three 7/10 matches, one 6/10 match, and one 5/10 match. HLA-C mismatch was present in 10 grafts (63%) and complete DP mismatch in 6 (38%).By revised Disease Risk Index, 25% of pts were high risk, 69% intermediate risk, and 6% low risk; 69% were in CR at BMT. The median graft doses were 3.36 x 108 total nucleated cells/kg and 3.87 x 107 CD3+ cells/kg. All pts engrafted, and there were no non-relapse deaths or prohibitive toxicities. By competing-risk analysis, the estimated cumulative incidence (CuI) of neutrophil recovery was 100% by day 29, with a median of 20 (range 14-29) days. The CuI of platelet recovery ≥ 20,000/µL was 100% by day 46, with a median of 32 (range 12-46) days. By day 60, 14/15 (93%) evaluable pts achieved full donor chimerism in either CD3+ or unsorted cells. Notably, there were no cases of acute grade 3-4 GVHD. Acute grade ≥ 2 GVHD was limited to 2 cases of skin-only grade 2 GVHD and 1 case of grade 2 skin and ungradable visceral GVHD. By competing-risk analysis, the estimated CuI of acute grade 2 GVHD was 12% at day 100 (90% CI, 0-27%) and 19% at day 180 (90% CI, 2-37%)(Fig A). The CuI of any chronic GVHD at 1 and 2 years was 7% (90% CI, 0-18%)(Fig A). Critical illness prior to relapse was limited to 1 case of sepsis which resolved. With a 3-year median follow-up, the probability of PFS was 56% at 1 year and 50% at 3 years (Fig B). The estimated CuI of relapse was 44% at 1 year. Median OS has not been reached, with an estimated 1 year and 3 year OS of 68% (Fig B).

Conclusion: These results suggest that virtually no pt should be denied allogeneic BMT because of lack of an HLA-matched or haplo donor. mMUD searches were successful in pts with DSA's that were too high for desensitization.


Off Label Use: posttransplantation cyclophosphamide for GVHD prophylaxis.

Author notes


Asterisk with author names denotes non-ASH members.