HLA-DP Mismatches Increase the Risk of Acute GVHD after Unrelated Donor Hematopoietic Transplantation (UDT).

The HLA class II DP locus encode for both subunits of DPB1 heterodimers, which have low levels of expression on the cell surface of antigen presenting cells. We hypothesized that donor-recipient HLA-DP mismatch would lead to an increased incidence of acute (a) graft-versus-host disease (GVHD), and that 2 mismatches would likely be even more significant.

Methods: We studied 84 consecutive patients (pts) with myeloid leukemias in complete remission (CR) transplanted from 01/02 to 02/06. Preparative regimens were ablative IV Busulfan-based (n=58) or Cy/TBI (n=2), and reduced intensity (Fludarabine (Flu)/Bu 130 mg/m²/2 doses plus Gleevec (n=8), and Flu/Melphalan 140 mg/m² (n=16). Stem cell (SC) source was bone marrow (n=70) or peripheral blood (n=14). ATG was given in 78 cases. GVHD prophylaxis was tacrolimus and mini-methotrexate in all cases, with additional pentostatin in 31 pts. High-resolution typing was sequence-based for HLA-A, B, DRB1; SSP was used for DRB3/4/5, DQB1 and DPB1, and SBT/SSOP for HLA-C. A Cox proportional hazards regression model was used to study aGVHD-free and relapse-free (RFS) survival. Variables with a p-value <0.25 by univariate analysis were included in the multiple regression analysis (MV). Variables were age, gender, weight, conditioning regimen, GVHD prophylaxis, diagnosis, cytogenetics, SC source, ABO group, infused CD34 and CD3 cell dose, and HLA matching. AGVHD-free survival was calculated from transplant date to date of development of grade II–IV GVHD or completion of 100 days of follow-up.

Results: Median age was 48 yrs (range, 14–72). Diagnoses were MDS (n=5), AML (n=58), and CML (n=21). 54 pts (64%) were beyond 1st CR; all CML pts were in >1st chronic phase (CP). Sixty-one pts were 10/10 HLA match (A, B, C, DRB1, DQB1), and 23 had one or more mismatches. All but one pt engrafted neutrophils at a median of 13 days. 33 pts (39%) and 13 pts (15%) developed grade II–IV and III–IV aGVHD, respectively. Chronic GVHD incidence was 51%. With a median follow-up of 18 mo. (range,1.3–52) 60 pts are alive; 40 pts have relapsed or died. Median survival has not been reached. Number of DP mismatches and incidence of aGVHD is shown in the table. The following covariates influenced aGVHD-free survival by MV analysis: Flu-based regimen (P=0.005; HR 0.25 (95%CI 0.1–0.66), reduced intensity regimens (p=0.02; HR 0.35 (95%CI 0.15–0.83) and presence of 2 DPB1 mismatches (p=0.02; HR 3.07 (95%CI 1.19–7.95). Presence of 1 DPB1 mismatch was not significantly associated with aGVHD. There was no statistically significant correlation between presence of 2 DP mismatches and RFS (P=0.17;HR 0.3 (95%CI 0.06–1.65);HR 0.75 for 1 mismatch) or with cGVHD. Actuarial 2-yr survival for 10/10 matched pts without DP mismatches (12/12) versus those with DP mismatches is 82% versus 71%(P=0.6). In the 10/10 matched group, GVHD was the cause of death only among recipients of 2 DP mismatches transplants (n=4).

Conclusion: Mismatching at HLA-DPB1 may increase the risk of aGVHD following UDT. The role of DP in the development of GVHD and GVL effects merits future study.