Comparison of DNA amplification, mRNA amplification, and DNA hybridization techniques for detection of cytomegalovirus in bone marrow transplant recipients.

A total of 676 specimens from 63 recipients of bone marrow allografts were tested for cytomegalovirus (CMV) by the following assays: CMV pp67 NucliSens (NS), AMPLICOR CMV MONITOR (RA), and the Digene CMV DNA test (DG). In a consensus analysis, the sensitivities and specificities were 60 and 99% (NS), 96 and 98% (RA), and 90 and 76% (DG), respectively; for detection of symptomatic CMV infection, they were 60 and 97% (NS), 65 and 97% (RA), and 95 and 77% (DG), respectively. In multivariate analysis, the major risk factor for symptomatic CMV infection was an increase in the viral load in the DG assay; in univariate analyses, maximum viral loads in both DG and RA assays and a rising viral load in the RA assay were also significant. The earliest detection of CMV replication was provided by the RA assay (mean, 39 days posttransplantation), followed by the DG assay (mean, 48 days) and the NS assay (mean, 58 days).