BACKGROUND: Quantitative commercial assays for cytomegalovirus (CMV) detection have recently been developed. Their role in the management of patients after transplantation needs to be evaluated. Widespread use of these assays will allow for comparison of results between centers and meaningful interpretation of the significance of viral load measurements. METHODS: Sequential samples from 52 patients after renal transplantation were tested in the murex hybrid capture assay (HCA) and the Roche Amplicor CMV DNA assay (QPCR) and correlated with the development of CMV disease. A comparison of viral loads in plasma and whole blood was also made. RESULTS: Both assays were sensitive and detected all cases of CMV disease. The specificity and positive predictive value increased from 0.34 and 0.36 to 0.85 and 0.96 for the HCA and 0.37, 0.37 to 0.72 and 0.63 for the QPCR following a receiver operator curve analysis. Higher viral loads were measured using the HCA compared to the QPCR. Response to ganciclovir was associated with a greater than 80% reduction in viral load by HCA or greater than 70% using the QPCR. CONCLUSIONS: Both assays were highly sensitive. By using a receiver operator curve analysis a cutoff viral load can be determined which maximizes the clinical utility of these assays.
BACKGROUND: Quantitative commercial assays for cytomegalovirus (CMV) detection have recently been developed. Their role in the management of patients after transplantation needs to be evaluated. Widespread use of these assays will allow for comparison of results between centers and meaningful interpretation of the significance of viral load measurements. METHODS: Sequential samples from 52 patients after renal transplantation were tested in the murex hybrid capture assay (HCA) and the Roche Amplicor CMV DNA assay (QPCR) and correlated with the development of CMV disease. A comparison of viral loads in plasma and whole blood was also made. RESULTS: Both assays were sensitive and detected all cases of CMV disease. The specificity and positive predictive value increased from 0.34 and 0.36 to 0.85 and 0.96 for the HCA and 0.37, 0.37 to 0.72 and 0.63 for the QPCR following a receiver operator curve analysis. Higher viral loads were measured using the HCA compared to the QPCR. Response to ganciclovir was associated with a greater than 80% reduction in viral load by HCA or greater than 70% using the QPCR. CONCLUSIONS: Both assays were highly sensitive. By using a receiver operator curve analysis a cutoff viral load can be determined which maximizes the clinical utility of these assays.