Cytomegalovirus infections associated with leukocyte transfusions

Patients randomized to receive or not to receive prophylactic leukocyte transfusions were evaluated prospectively for serologic, histologic, and cultural evidence of cytomegalovirus infection. Recipients of prophylactic leukocyte transfusions and control subjects were similar with regard to age, sex, underlying disease, immunosuppressive therapy, and number of other transfusions. The recipients of prophylactic leukocyte transfusions (mean, 23.1) had significantly more cytomegalovirus infections (19 of 31 versus seven of 27, p = 0.01) than did control patients receiving no leukocytes or only therapeutic leukocyte transfusions (mean, 3.8). Twenty-seven of 66 donors of leukocytes were seropositive for cytomegalovirus complement-fixation antibody, but cytomegalovirus was not isolated from any of 62 leukocyte transfusions cultured for virus. These results are consistent with the hypothesis that latent cytomegalovirus may be present in leukocytes of blood donors with previous cytomegalovirus infection and after transfusion may be activated to produce active cytomegalovirus infection.