Introduction: HLA molecules play an important role for immunoreactivity in allogeneic hematopoietic stem cell transplantation (HSCT). To elucidate the effect and the potential identification of "permissive" and "non permissive" I and II class HLA mismatching (mm) loci on the early and long term transplant outcome, we conducted a retrospective/prospective observational analysis on 1789 patients transplanted with unmanipulated haematopoietic stem cells from a volunteer unrelated donor (URD).

Methods Between January 2012 to December 2015, 1789 adult patients with a median age of 49 years (18-70) affected by haematological malignant diseases, performed an unrelated HSCT, coordinated by the Italian Bone Marrow Donor Registry (IBMDR). All patients have been typed by high resolution (HR) HLA typing for HLA-A/B/C/DRB1/DQB1 loci, at the start of the donor's search process. Patient and donor characteristics are shown in Table 1. As conditioning regimen and GVHD prophylaxis, 71% of patients received a myeloablative conditioning and 76% a combination of anti-Thymoglobuline, Cyclosporine and Metotrexate short course. Total Body Irradiation was part of conditioning regimen in 14% of cases and PBSC was used as stem cell source in 80% of transplants. Median follow for survivors was 38 months (1-76). Regarding to the allelic compatibility, 890 (50%) of donor/recipient (D/R) pairs were 10/10 HLA matched, 677 (38%) showed 1 mm for A, B, C, DRB1 or DQB1 in 249, 141, 173, 2 and 112 cases, respectively and 222 (12%) received HSCT from a 8/10 or less HLA matched donor.

Results: Overall 90% and 79% of patients achieved PMN and PLTS engraftment within 30 and 90 days, respectively. Probabilities of 3-y Overall Survival (3-yr OS), Progression Free Survival (3-yr PFS), and Graft Relapse Free Survival (3-yr GRFS) were 52%, 42%, and 30%, respectively. The 3-y CI of Transplant Related Mortality (TRM) was 26%, with a 100-days CI of acute GVHD ≥2 of 26%, whereas the 3-yr CI of chronic GVHD was 30%, of which 10% extensive. Cox multivariate analysis showed that, compared to 10/10 HLA-matched HSCT, 9/10 and ≤8/10 HLA-matched HSCT were associated with worse outcomes in terms of OS (HR 1.16, p=0.04 and HR 1.3, p=0.007, respectively), GRFS (HR 1.2, p=0.005 and HR 1.2, p=0.07, respectively), TRM (HR 1.3, p=0.007 and HR 1.6, p<0.0001, respectively), grade 3-4 aGVHD (HR 1.8, p=0.0001 and HR 1.8, p=0.01, respectively) and cGvHD (HR 1.3, p=0.005 and HR 1.1, p=0.35, respectively). Notably, no significant differences occurred through the comparison between ≤8/10 and 9/10 matching. Univariate comparisons are shown in Figure 1. Moreover, in order to identify permissive and non permissive allelic mismatching, we analyzed the donor/recipient pairs with a single HLA mm with a frequency > 5%: the presence of A02:01 in the patient's HLA, after adjustment for HLA matching at the other loci and other clinical variables known to affect HSCT outcome, was associated with significant higher risk of TRM (HR 1.9, p= 0.03) and worst OS (HR 1.7, p=0.04). Patient's age > 49 years (p<0.0001), advanced disease stage (p<0.0001), presence of 1 or more co- morbidity according to the Sorror Hematopoietic Cell Transplant-Comorbidity Index (p=0.01) were associated with a hazard risk of 1.4, 2, 1.2 for OS and 1.6, 1.75, 1.4 for TRM. Moreover, the Italian origin of recipient and donor resulted in reduced grade 2-4 acute (HR=0.6, p=0.001) and chronic GVHD (p=0.002, HR=0.4). Finally, the Transplant Program expertise (>10 HSCT/year) is associated with reduced TRM (HR 0.8, p=0.0001), HSCT from female donor to male recipient was associated with higher risk of extensive cGvHD (HR 1.4, p=0.03), and CMV negative/negative status versus other combinations had protective effect on development of grade 3-4 aGVHD (HR 0.56, p=0.04).

Conclusions: Our large cohort data of homogeneously treated 1789 URD transplants, show that 10/10 HLA matching remains a significantly favorable prognostic factor for OS, TRM, GRFS and acute/chronic GVHD, whereas there are no significant differences between 8/10 and 9/10 matching transplants. Moreover, the HLA A02:01 as single mm seems to play a "non permissive" role. Finally the Italian origin of recipient and donor is related to a reduced development of GVHD probably due to the matching of the extended MHC haplotypes in individuals of the same geographic origin.

Disclosures

Rambaldi:Amgen Inc.: Consultancy; Pfizer: Consultancy; Celgene: Consultancy; Omeros: Consultancy; Roche: Consultancy; Novartis: Consultancy; Italfarmaco: Consultancy. Vago:Moderna TX: Research Funding; GENDX: Research Funding. Patriarca:Janssen: Other: Advisory role; Jazz: Other: Travel, accommodations, expenses; Medac: Other: Travel, accommodations, expenses; Celgene: Other: Advisory Role; Travel, accommodations, expenses; MSD Italy: Other: Advisory Role.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution