Impact of HLA Class I Typing-Resolution and Mismatches on Survival and Graft-Versus-Host Disease in Patients after Hematopoietic Stem Cell Transplantations from Unrelated Donors.

We analyzed 150 consecutive patients (pts) transplanted from unrelated donors (URD) at single institution- Silesian Medical Academy in Katowice- with use of the same standard operating procedure from February 1997 until December 2004 for CML (74 pts), AML (28), ALL (27), SAA (9), MDS (7), MM (1), NHL (1), PNH (1), OMF (1), bi-phenotypic AL (1). 92 pts were transplanted from matched donors, including 59 with complete 10 alleles (HLA-A,B,C,DRB1,DQB1) high resolution (HR) DNA-typing; and 33 with first class match established on base of low-resolution (LR) (13), serological (7) or un-complete typing without HLA-C (13) and second class HR typing. 58 pts were transplanted from mismatched donors (first class typing was HR in 43, LR in 14 and serological in 1 pt): 22 with single allelic mismatch (2 HLA-A, 7-B, 6-C, 7-DQB1), 33 with single HLA-C antigen mismatch and 3 with double mismatches (2 B+C, 1 C+C). Survival advantage at 4 years, although without statistical significance (p=0.16), was observed in the group of pts transplanted from 10/10 alleles matched donors (36⁺/₋ 11%) over those with mismatched donors (24⁺/₋ 11%). Oppositely, pts transplanted from matched donors who were not completely typed in HR did not achieve better survival (23⁺/₋ 17%). Poorest survival (13⁺/₋ 12%, p=0.007) was observed in patients transplanted from mismatched homozygous donors (n=10). The risk of acute GVHD grade 3–4 was increased (p=0.007) in pts with mismatched donors (31⁺/₋ 6%) when compared to matched completely typed in HR (10⁺/₋ 4%). Also the rate of graft failure tended to be lower in pts with matched than mismatched donors (5.1% versus 10.2%, p=0.25). In contrast, relapse rate was lower in mismatched (23⁺/₋ 10%) than in HR matched pts (34⁺/₋ 12%, p=0.55) what may reflect better GVL effect in mismatched transplant recipients. Unexpectedly, the rate of chronic GVHD was similar in pts with 10/10 alleles matched and mismatched donors (40⁺/₋ 10% versus 42⁺/₋ 9%, p=0.75). These results indicate that complete high resolution HLA class I typing is necessary for adequate selection of unrelated donors. Class I HLA-B and -C mismatches influence both survival and serious a-GVHD incidence.