Abstract
One hundred twelve peripheral blood specimens were tested for the presence of cytomegalovirus (CMV) by the tube culture indirect immunoperoxidase (TC‐IPA) procedure, the shell vial assay [shell vials were pre‐ and postinoculation treated with medium containing 2 of 10% fetal bovine serum (FBS) or 100 μg% cortisol] (SV‐IFA), and conventional (MRC‐5) tube cultures (TC‐CPE). CMV was detected in 25 (22%) of the 11 2 specimens tested by at least one of these methods. The detection/isolation of CMV among the 25 positive specimens in shell vials maintained with 2% FBS, 100 μg% cortisol + 2% FBS, and 10% FBS was 36,44, and 52%, respectively. Detection/ isolation of the virus from blood by TC‐IPA and TC‐CPE was 52% and 76%, respectively. A significantly greater CMV detection rate occurred using TC‐CPE compared to SV‐IFA treated with medium supplemented with an FBS concentration of 2% (P = .01321, but not medium containing the higher serum supplement or the glucocorticoid (P > .05). Differences in the identification of a CMV viremia were observed by IPA, SV‐IFA, and TC‐CPE methodologies on a patient‐to‐patient basis, denoting the necessity of incorporating each methodology into the CMV screening panel. Demographic analysis of 82 AIDS patients showed a CMV viremia prevalence of 9% (2128) in intravenous drug users, 57% (27/47) in homosexual patients, and 22% (2/9) in heterosexual and transfusion patients, Overnight (24 hr) storage of whole blood at 4 or 24°C, respectively, reduced CMV recovery by 40% and 6570, when tested by TC‐CPE. Improved culture‐based detectioniisolation of CMV in peripheral blood is contingent upon prompt specimen processing, and the utilization of combined SV‐IFA, TC‐IPA, and TC‐CPE methods. © 1992 Wiley‐Liss, Inc.