Stability of BK polyomavirus DNA in urine over time and analytical evaluation of an automated BKV quantitative nucleic acid test

Abstract

Introduction. BK polyomavirus (BKPyV) quantitative testing is an important screening tool post-transplantation, although interpretation can be challenging due to lack of standardization, assay heterogeneity and variability of BKPyV DNA over time (in urine). Methods. Remnant clinical EDTA plasma and urine samples were tested by the cobas BKV test and a validated laboratory-developed test (LDT). Accuracy [positive and negative percent agreement (PPA and NPA), Pearson’s correlation, Bland–Altman analysis] and reproducibility were evaluated. To assess BKPyV DNA stability in urine, prospective urine samples were maintained at two different storage temperatures and tested in triplicate over 7 days. Results. Overall PPA was 95.6 % (43/45) and NPA was 94.4 % (170/180). For plasma, Pearson’s correlation (0.950) and Bland–Altman analysis (0.113±0.22 log10 IU ml–1) showed high agreement. For neat urine, Pearson’s correlation (0.842) and Bland–Altman analysis (0.326±0.80 log10 IU ml–1) showed somewhat higher variability. Reproducibility was high for the cobas BKV versus the LDT. BKPyV DNA levels in neat urine remained relatively stable over 7 days at both storage temperatures, although outlier results were intermittently detected. Conclusion. The cobas BKV test showed high agreement and reproducibility compared to the reference LDT. BKPyV viral load testing in urine has known limitations, but neat urine can be processed by the cobas BKV.