Prediction of crossmatch outcome in highly sensitized dialysis patients based on the identification of serum HLA antibodies.

High levels of allosensitization (greater than 50%), which often occur in dialysis patients awaiting renal transplant, make donor selection difficult. Such patients may be included in elaborate protocols in which they are crossmatched with all available ABO compatible donors, or crossmatching may be deferred until a very-well-matched donor becomes available. The former approach of random crossmatching is costly and inefficient, while the latter approach may overlook crossmatch-compatible donors. We believe that the identification of antibodies present in highly reactive sera and the use of this information in donor selection would increase the frequency of crossmatch-negative donors for these patients. In this study eleven sera, reactive with 70% to 100% of a random cell panel, were obtained from multiply transfused dialysis patients. Sera were analyzed by standard (CDC) and antiglobulin augmented (AHG-CDC) lymphocytotoxicity, and by differential absorption with HLA-typed platelets. All sera contained only one or two antibodies directed against the high frequency public HLA epitopes, accounting for 85% to 100% of each serum's total reactivity. These characterized sera were crossmatched with 114 random normal donors. The frequency of negative crossmatches was 20.5%. However, if the serum antibody data had been used to preselect donors for crossmatch--that is, to exclude donors that were likely to be positive--the negative crossmatch frequency would have increased to 86.4%. The use of the serum analysis data in donor selection would have reduced the total number of required crossmatches by 78%. Serum analysis correctly predicted the outcome of 95.6% of crossmatches performed with an average of 3% false positives and 1.3% false negatives. This approach to donor selection reduces unnecessary crossmatching and increases the likelihood of finding crossmatch-compatible donors for highly reactive patients.