A prospective randomized trial of HLA-matched versus mismatched single-donor platelet transfusions in cancer patients

The use of histocompatability antigen (HLA)-matched platelets has been advocated for the support of thrombocytopenic cancer patients. We randomized 78 newly diagnosed cancer patients prospectively (before thrombocytopenia) to receive either HLA-matched or mismatched single-donor platelet transfusions. Three hundred forty-one platelet transfusions were given for 80 separate episodes of therapy-induced thrombocytopenia in 33 patients. Forty-five patients receiving intensive chemotherapy did not develop significant (less than 20,000 platelets/mm3) thrombocytopenia and did not receive a platelet transfusion. No marked difference was observed between the matched and mismatched groups in regard to number of total platelet transfusions per patient (median, 3 vs. 5, respectively; P = 0.076), number of platelet transfusions per episode (median, 3.0 vs. 3.5, respectively; P = 0.28), or days between transfusions (median, 2 vs. 2, respectively, P greater than 0.4). Bleeding episodes, although rare, tended to be of increased severity in the mismatched group. Febrile patients receiving mismatched platelets tended to have a lower posttransfusion increment increase than their nonfebrile counterparts (P = 0.068), although a similar trend could not be demonstrated between febrile and nonfebrile patients who received matched platelets (P = 0.22). Patients treated as outpatients had significantly higher posttransfusion increments than when transfused as inpatients when they were given mismatched platelets (P less than 0.0005). Development of antiplatelet antibody did not appear to affect response to platelet transfusions. Only one patient developed sustained high-level antibody titers. In patients where thrombocytopenia was significant, the transfusion of HLA-matched platelets did not appear to offer a significant advantage. However, HLA-matched platelet transfusions tended to be associated with higher posttransfusion increments in febrile patients and a trend toward fewer severe bleeding episodes. A multi-institution trial containing a large number of patients is needed to evaluate trends observed in this study.