Analytic validation of a quantitative real-time PCR assay to measure CMV viral load in whole blood

Cytomegalovirus (CMV) is a significant cause of morbidity and mortality in immunocompromised patients. We compared the CMV pp65 antigenemia test with a less labor intensive quantitative polymerase chain reaction (PCR) assay in 109 whole blood samples predominantly from transplant patients and patients with AIDS. DNA was amplified on an Applied Biosystems 7900 instrument using a TaqMan probe targeting the CMV polymerase gene and the APOB human control gene. The DNA assay was linear over a 6-log range from 8 to 800,000 CMV genomes per reaction; coefficient of variation was 20%. CMV DNA was undetectable in 20 blood samples from healthy donors whereas it was detected in 55 of 109 patient samples. Results were concordant in a nonlinear fashion with those of the antigenemia test in 90/109 (83%). Evaluation of the discrepancies suggested that either PCR or antigenemia assays could be falsely negative when virus levels were quite low. A point mutation interfered with probe binding in 1 sample. A second real-time PCR targeting the immediate early gene was even more likely to be false negative. In summary, CMV viral load measurement targeting the polymerase gene is nearly equivalent to the antigenemia assay for detecting and monitoring active CMV infection in whole blood samples.