What level of matching is appropriate

Initially, the importance of matching for HLA-A, -B, and -DR was established by serologic or antigen-level assignment.126 127 With DNA-based typing it was subsequently shown that allele- or high-resolution-level matching for DRB1* was associated with lower GVHD and TRM,112-114'128'129 being most evident when donors were otherwise matched for class I antigens. Since the mid-1990s most centers have matched unrelated donors for DRB1* at the allele- or high-resolution-typing level whenever possible, and often with preference over matching for class I.

Until the later 1990s, most studies of HLA matching involved antigen- or low-resolution-level HLA class I typing and may not have considered HLA-C.

A number of recent large studies have investigated the role of class I matching using allele-level DNA typing. Petersdorf and colleagues23 107 found that singleantigen-level or multiple-allele-level class I (A, B, C), but not class II, mismatches in the HVG (host versus graft) vector were associated with graft failure in CML patients. The risk of graft failure was particularly high when the mismatch was only in the HVG direction (donor is mismatched at a locus for which the patient is homozygous). In a subsequent study,110 allele- or antigen-level HLA-A, -B, -C mismatches were associated with increased mortality in low risk, but not higher risk CML patients. Two large registry studies of transplants performed at multiple centers and in multiple disease types have also been reported. Morishima and colleagues104 105 found that single

HLA-A, -B, -C, and -DRB1* antigen- or allele-level mismatches were associated with increased acute GVHD, and class I mismatches were risk factors for failed engraftment. Flomenberg and colleagues106 found that antigen or low-resolution mismatching at HLA-A, -B, or -Cw was significantly associated with increased mortality. No significant mortality risk was found for single locus mismatches found only at the allele level (i.e., antigen level matched), although a trend was shown for HLA-A allele mismatches. For severe acute GVHD, only HLA-A antigen-level mismatches were significantly associated although trends for association for HLA-B and -C were present. In sum, these studies show that HLA-A, -B, -C antigen or low-resolution mismatches are clearly associated with reduced overall survival. The effect is at least similar to mismatches for HLA-DRB1* and has called into question the common practice of giving matching preference to DRB1* when choosing partially mismatched unrelated donors. With regard to allele-level versus antigen-level class I mismatches, it would appear that single-antigen-level mismatches are more strongly associated with severe consequences than are allele-level mismatches. However, trends toward allele mismatch effect and indications of cumulative or synergistic effects of multiple allele mismatches suggest that allele-level matching for at least HLA-A, -B, -C, and -DRB1' should be utilized when possible. Similar high-resolution matching for HLA-DQ and -DP is also supported by reports by individual centers.

0 0

Post a comment