Test performance characteristics of quantitative nucleic acid testing for polyomaviruses in kidney and kidney-pancreas transplant recipients

Abstract

Screening for polyoma BK virus (BK) using nucleic testing (NAT) is recommended for kidney and kidney-pancreas transplant recipients, but the performance characteristics of quantitative BK NAT at different thresholds of plasma BK viral loads are unclear. We aim to evaluate the diagnostic accuracy of quantitative BK NAT as an add-on test to qualitative polyoma NAT for the diagnosis of BK virus-associated nephropathy (BKVAN) in kidney and kidney transplant recipients. We calculated the test sensitivity, specificity, and predictive values at the different thresholds of plasma BK viral load for BKVAN. At the recommended threshold of >1 x 10(3) serum BK copies/mL serum for test positivity, the sensitivity for BKVAN was 92.9% (95% confidence intervals [CI]: 66.1-99.8) and specificity 79.1% (95%: CI 67.4-88.1), with corresponding positive and negative predictive values of 42.0% (95% CI: 24.8-57.7%) and 98.6% (95% CI: 98.3-99.9%), respectively. The overall area under curve for the quantitative BK NAT was 0.92 (95% CI: 0.85-0.97). Quantitative BK NAT displays properties of high sensitivity and specificity that are fit for purpose as an add-on test to qualitative polyomavirus NAT for kidney and kidney-pancreas transplant recipients at risk of BKVAN.