Comparative evaluation of laboratory developed real-time PCR assays and RealStar(®) BKV PCR Kit for quantitative detection of BK polyomavirus.

Background: Quantitative, viral load monitoring for BK virus (BKV) by real-time PCR is an important tool in the management of polyomavirus associated nephropathy in renal transplant patients. However, variability in PCR results has been reported because of polymorphisms in viral genes among different subtypes of BKV, and lack of standardization of the PCR assays among different laboratories. In this study we have compared the performance of several laboratory developed PCR assays that target highly conserved regions of BKV genome with a commercially available, RealStar(®) BKV PCR Kit.

Method: Three real-time PCR assays (i) VP1 assay: selected from the literature that targets the major capsid protein (VP1) gene (ii) VP1MOD assay: VP1 assay with a modified probe, and (iii) BKLTA assay: newly designed assay that targets the large T antigen gene were assessed in parallel, using controls and clinical specimens that were previously tested using RealStar(®) BKV PCR Kit (Altona Diagnostics GmbH, Hamburg, Germany). Nucleic acid from all samples were extracted using the QIA symphony virus/bacteria kit on an automated DNA extraction platform QIA symphony SP (Qiagen). Primer and probe concentration, and reaction conditions for laboratory developed assays were optimized and the limit of detection of different assays was determined. Positive control for laboratory developed BK assays was prepared through construction of a plasmid carrying respective amplicon sequences.

Results: The 95% detection limit of VP1, VP1MOD and BKLTA assays were 1.8×10(2), 3×10(3) and 3.5×10(2) genomic copies/ml, respectively, as determined by Probit regression analysis of data obtained by testing a dilution series of a titered patient specimen, using RealStar(®) BKV PCR Kit. The inter-assay and intra-assay, coefficient of variations of these assays using calibrated, plasmid standards were <1%. All assays, including the RealStar(®) BKV PCR assay, were highly specific when tested against a panel of external proficiency specimens containing both BK and JC viruses. All assays, except the VP1MOD assay determined BK viral load in proficiency specimens within the same log values. With reference to results obtained by RealStar(®) BKV PCR assay, the sensitivity and specificity of different assays tested in 116 serum specimens submitted for BK viral load assay were 91% and 97% for VP1 assay, 88% and 97% for VP1MOD assay, and 97% and 98% for BKLTA assay, respectively. BK Viral load in positive specimens determined by various assays was highly correlated (R(2)>0.97), based on linear regression analysis.

Conclusions: The performance characteristics of the newly designed, BKLTA assay were highly comparable to RealStar(®) BKV PCR assay, and can be used for routine detection and viral load monitoring of BKV in a cost-effective manner.