Serum and urine nucleic acid screening tests for BK polyomavirus-associated nephropathy in kidney and kidney-pancreas transplant recipients: Diagnostic test accuracy review
Thida Maung Myint1,2, Chanel CH Chong 2, Amy von Huben 2, John Attia 3, Angela C Webster 2,4, Christopher D Blosser 5, Jonathan C Craig 6,7, Armando Teixeira-Pinto2, Germaine Wong 2,4.
1Nephrology and Transplantation , John Hunter Hospital, Newcastle, Australia; 2Sydney School of Public Health, University of Sydney, Sydney, Australia; 3University of Newcastle, Newcastle, Australia; 4Centre for Transplant and Renal Research, Westmead Hospital, Sydney, Australia; 5Department of Medicine, Nephrology, University of Washington & Seattle Children’s Hospital, Seattle, United States; 6College of Medicine and Public Health, Flinders University, Adelaide, Australia; 7Cochrane Kidney and Transplant, Centre for Kidney Research, The Children's Hospital at Westmead, Sydney, Australia
Introduction: BK polyomavirus-associated nephropathy (BKPyVAN) occurs when BKPyV affects a transplanted kidney leading to an initial injury characterised by cytopathic damage, inflammation and fibrosis. BKPyVAN may cause permanent loss of graft function and premature graft loss. Early detection allows clinicians to intervene by timely reduction in immunosuppression to reduce adverse graft outcomes. Quantitative nucleic acid testing (QNAT) for the detection of BKPyV DNA in blood/serum and urine is increasingly used as a screening test as diagnosis of BKPyVAN by kidney biopsy is invasive and associated with procedural risks.
Method: We performed a systematic review to determine the diagnostic test accuracy of QNAT tests for the diagnosis of BKPyVAN. We searched MEDLINE, EMBASE, Cochrane and other databases until June 2023. We included cohort studies assessing the diagnostic accuracy of blood or urine BKPyV QNAT for the diagnosis of BKPyVAN, as verified by the reference standard (histopathology). We assessed the methodological quality by using QUADAS‐2 assessment criteria. We used the bivariate random-effects model to obtain the summary estimates. We explored possible sources of heterogeneity by adding covariates to meta-regression models.
Results:
Blood QNAT: The diagnostic performance of blood QNAT using a common threshold viral load of 10,000 copies/ml was reported in 18 studies (3434 participants). Summary estimates indicated that the pooled sensitivity and specificity were 0.85 (95% CI 0.85 to 0.86) and 0.95 (95% CI 0.94 to 0.96) respectively. After the exclusion of studies with a high risk of verification bias, the sensitivity was 0.79 (95% CI 0.64 to 0.89) and the specificity was 0.96 (95% CI 0.90 to 0.98). Analysis of all studies including various viral load thresholds indicated that test performance remains robust; pooled sensitivity 0.90 (95% CI 0.85 to 0.94) and specificity 0.93 (95% CI 0.91 to 0.95). However, there were a limited number of studies available to analyse the summary estimates for individual viral load thresholds other than 10,000 copies/ml. Across various thresholds, a cut-off of blood BKPyV at 10,000 copies/mL is the best trade-off between the test sensitivity and specificity.
Urine QNAT: We were unable to identify any specific threshold that might be useful in clinical practice as the thresholds used in these studies were highly variable.
Conclusion: There is insufficient evidence to suggest the use of urine QNAT as the primary screening tool for BKPyVAN. The summary estimates of blood QNAT test at a threshold of 10,000 copies/ml for BKPyVAN are 0.85 (95% CI 0.85 to 0.86) and 0.95 (95% CI 0.94 to 0.96) respectively. While some studies included in this review were of low methodological quality concerning limitations related to verification bias, the current evidence suggests that a cut-off of blood QNAT at 10,000 copies/ml is the best trade-off between the test sensitivity and specificity across all thresholds.
