We employed four crossmatch techniques to select platelet donors for refractory patients. Forty-four donor-recipient pairs were studied in 32 patients. Analysis of effectiveness of platelet transfusions revealed that only 18 percent of transfusions gave a borderline response; the remainder were either effective or not effective at all. The corrected predictive values of three crossmatch tests were as follows: enzyme-linked immuno-specific assay, 81 percent; platelet immunofluorescence test, 73 percent; and lymphocytotoxicity, 70 percent (p greater than 0.05). The predictive value of these tests did not differ in HLA-matched versus unmatched platelet transfusions. Donor selection by lymphocytotoxicity compatibility did not appear to be useful if donors were selected by either of the other two methods. The fourth test, antiglobulin-modified lymphocytotoxicity, offered no advantage over lymphocytotoxicity. Our data suggest that platelet crossmatching assays are a useful adjunct to the selection process for the platelet donor in addition to ABO, Rh, and HLA matching.
