Comparative analysis of three nucleic acid-based detection systems for hepatitis C virus RNA in plasma from liver transplant recipients.

The early detection of hepatitis C viraemia (HCV) following liver transplantation is important for monitoring disease recurrence and planning antiviral chemotherapy. In the current study, the sensitivity, specificity, and concordance of three HCV RNA assays were compared using a random sample of 84 plasma specimens from 23 transplant recipients. Two of the assays were prototype commercial tests: Roche Molecular Diagnostic's RT-PCR HCV Amplicor system; and Chiron's Quantiplex HCV-RNA assay. The third was a 'home brew' PCR-liquid hybridization/gel retardation assay developed at the University of Pittsburgh Medical Center (UPMC). On all criteria the PCR-based assays out-performed the Quantiplex assay and displayed an overall concordance of 87%. A high percentage of specimens in the Quantiplex assay gave indeterminate results (12%) or high coefficients of variance (13%). The specificities of all RNA assays were determined using HCV serostatus as the gold standard. Both of the PCR-based assays had specificities of 100%, whereas the Chiron Quantiplex HCV assay had a specificity of 88%, if indeterminates were counted as negatives, and a specificity of 64% if indeterminates were counted as positives. The calculated sensitivities of the PCR-based assays were 56% and 48% for the 'home brew' and the Roche assays, respectively. The UPMC HCV assay, however, was determined to be capable of reproducibly detecting four or fewer chimpanzee infectious doses, suggesting that HCV viraemia was not present in the PCR-negative cases. The sensitivity of the Quantiplex assay was 41% counting indeterminates as negatives and 46% counting them as positives. The high cost of the Quantiplex assay combined with the number of uninterpretable results, the lack of sensitivity, and reduced specificity may limit the usefulness of this assay for monitoring HCV recurrence.