Effect of HLA mismatching and antibody status on "homovital" aortic valve homograft performance

Background. Recipients of "homovital" aortic valve homografts are known to produce specific antibodies to human leukocyte antigen (HLA) determinants p resent on the cellular compartment of the valve tissue; however, the clinic al significance of these antibodies is unknown. Data from 182 patients rece iving homovital aortic valve homografts has been analyzed to determine the impact of HLA disparity and HLA antibody production on survival and functio n of the homograft. Methods. Human leukocyte antigen mismatch data were available for 127 patie nts (mean follow-up, 6.02 +/- 0.26 years). Two patients were considered wel l matched for HLA A+B antigens (zero or one mismatch) compared with 125 poo rly matched (two to four mismatches). Nine patients had a zero HLA-DR misma tch compared with 52 with one mismatch and 59 patients completely mismatche d for DR antigens. Results. There was no significant association between the degree of HLA mis match for either class I or class II antigens whether the loci were conside red alone or in combination (ie, A, B, DR, AB, or ABDR mismatching) with ma rkers of long-term valve function including patient mortality, reoperation, valve degeneration, valve stenosis, presence of regurgitation, and postope rative New York Heart Association class. One hundred thirty-six of 167 (82% ) were found to have produced antibodies after operation (mean time after o peration, 6.42 +/- 0.58 years). In 61 cases both antibody specificity and d onor HLA typing was available. In 92% of these, the antibodies were of the IgG subclass and were specific for the HLA class I molecules of the donor. The presence of HLA antibodies was associated with an increase in the frequ ency of mild valve stenosis (not significant) compared with those patients who did not develop HLA antibodies (antibody negative = 9.7%; panel reactiv e antibodies <50% = 29.1%; and panel reactive antibodies >50% = 22.2%; not significant). There was also an increased prevalence of valve degeneration associated with HLA antibodies. The actuarial freedom from valve degenerati on for the 35 HLA antibody-negative patients was 100% at 1, 5, and 10 years compared with 100% at 1 year, 97% at 5 years, and 92% at 10 years for 55 p atients with panel reactivity less than 50%, and 98% at 1 year, 94% at 5 ye ars, and 88% at 10 years for the 77 patients who were highly sensitized (no t significant). There was no correlation with other markers of long-term va lve function. Conclusions. The influence of the immune response on valve function require s further studies involving large numbers of patients followed for a longer period of time. We believe prospective matching for HLA antigens is warran ted to produce a well-matched cohort of patients for analysis and to reduce antibody sensitization, which would help to clarify this issue. (C) 1998 b y The Society of Thoracic Surgeons.