RESUMO
We examined data from the New England Organ Bank to characterize the influence of patient sensitization on allograft survival, and our current crossmatching strategy. To evaluate our recipient eligibility criteria, we compared computer-predicted crossmatch results to 3622 actual crossmatches. A computer-predicted positive crossmatch was highly predictive of an actual positive crossmatch, for patients with a percentage reactive antibody of 40% of more (positive predictive value 91-99%), thus obviating the need to perform the actual crossmatch. Given the high prevalence of sensitized patients on our waiting list, very few individuals are inappropriately excluded from consideration for an available organ. In contrast, a negative computer prediction was never sufficiently predictive of a negative crossmatch to dispense with the actual crossmatching procedure. We also compared graft survival in patients with positive antidonor crossmatches using historical (greater than 6 months old) sera with those with negative historical crossmatches (or with no history of humoral sensitization). One-year actuarial graft survival in the first group was 61.0 +/- 6.0%, compared with 85.2 +/- 1.4% in those without positive historical crossmatches (P less than 0.001). This adverse effect of a positive historical crossmatch was true in both first transplants (n = 41, 1-year graft survival 67.9 +/- 7.4% vs. 86.2 +/- 1.6%, P less than 0.05) and in regrafted individuals (n = 29, 1-year graft survival 50.7 +/- 9.8% vs. 78.9 +/- 3.7%, P less than 0.01). The inability to accurately predict negative crossmatches, and the possible adverse effect of positive historical crossmatches on graft survival, represent potential obstacles to a goal of national organ sharing.
