The Risk of Cytomegalovirus Infection in Seronegative Transfusion Recipients Not Receiving Exogenous Immunosuppression

Westudied 637 transfusion recipients seronegative for cytomegalovirus (CMV) in the fol-lowing categories: neonates; pregnant women; and patients experiencing trauma, burns, cardiovascular surgery (adult or pediatric), major surgery, or gastrointestinal hemorrhages. Cultures and serological tests were used to follow up subjects for evidence of CMV infec-tion for a period of three months after their last transfusion. Six (0.9070) developed CMV infection. No significant differences in risk among patient categories were observed. In-fected patients receiveda significantly larger mean number of units of cellular blood prod-ucts (CBP; 50.0 ± 38.9 vs. 6.2 ± 8.5; P <.001) and plasma (23.7 ± 15.3 vs. 2.6 ± 4.6, P<.001)than did uninfected patients. This result represents a risk per unit of CBP trans-fused of 0.14070, or rv0.380J0 per unit of seropositive CBP transfused. We observed, how-ever, that patients exposed to CBP from>30 donors had a higher risk of acquiring CMV infection than would be predicted if infectious units were randomly distributed among all donors (P <.01). Cytomegalovirus (CMV) is known to be transmit-ted by blood transfusion. As recently reviewed by Tegtmeier [1, 2], a number of prospective studies have attempted to define the risk of acquiring CMV from blood products. There is evidence, however, to sug-gest that this risk is changing with changing blood-banking practices [1, 2]. With one exception [3], numerous attempts to isolate CMV from blood prod-ucts have been unsuccessful [1]. It is generally be-lieved that CMV is latent in an as yet unidentified cell in blood products and is reactivated in the re-Received for publication 9 July 1987, and in revised form 2