Comparison of total culturable virus assay and multiplex integrated cell culture-PCR for reliability of waterborne virus detection.

The total culturable virus assay (TCVA) and an integrated cell culture-PCR (ICC-PCR) were compared in parallel to evaluate their detection reliability. Source, finished, and tap water samples from three drinking water treatment plant systems were analyzed by TCVA, and every cell culture dish was subsequently examined by reverse transcription (RT) multiplex PCR using enterovirus- and adenovirus-specific primers. Twenty-seven of 180 (15%) inoculated dishes exhibited cytopathic effects (CPE). Virus concentrations for source water ranged from 3.3 to 21.0 most probable numbers of infectious units (MPN) per 100 liters. No finished or tap water samples were positive. On the other hand, 38 (21%) of the dishes were positive in multiplex ICC-PCR. Virus concentrations ranged from 4.5 to 10.2 MPN/100 liters for source water and 0 to 0.9 MPN/100 liters for finished and tap water. In spite of its superior sensitivity, the ICC-PCR assay resulted in lower virus concentration values than the TCVA for two of the source water sites. Retest of the CPE-positive dishes using reovirus-specific RT-PCR revealed that 24 of the 27 (89%) dishes were also positive for reoviruses. These observations suggested that the detection reliability of ICC-PCR is restricted by the primer sets that are integrated in the reaction mixture. The observation of an uneven distribution of PCR-positive culture dishes in a given sample raises an additional caution that simple extrapolation of the ICC-PCR result from the analysis of a limited fraction of collected samples should be avoided to minimize possible over- and underestimation of the amount of virus.