Introduction: Early detection of BK virus replication and prompt reduction of immunosuppressant at earlier stage of disease is the key of success in management of BKVAN. Our group had demonstrated BK virus screening with urine cytology and BKV serum quantitative polymerase chain reaction (qPCR) is a cost saving method to identify early infection. However, prospective data on the test performance of 2-stage screening is lacking.

Materials and Methods: Kidney transplant recipients within first 5 years post-transplant and those with graft dysfunction (including after anti-rejection therapy) were included. Recruited patients had regular urine cytology: every 3 months from the time of transplantation until the end of the first year post-transplantation and then annually till 5-year post-transplant. Additional urine cytology tests were performed in patients after anti-rejection therapy. Quantification of serum BK viral load by qPCR was performed in patient who had urinary decoy cell. BK virus associated nephropathy (BKVAN) was defined as sustained plasma BKV DNA and loads of >4 log10 cp/mL with or without biopsy proven BKVAN.

Results: 509 patients were recruited (1740 patient-years) and were followed up for 3.2 years in average. 33 patients were diagnosed to have biopsy proven or presume BKVAN. 5 patients had sustained BK viraemia of 3 log and 13 patients had transient BK viraemia <3 log with no evidence of BKVAN. The overall incident rate of BKVAN was 0.019 cases per patient-year. 82% of BKVAN (27 cases) were diagnosed in the first 2 years after transplantation. 9 cases were missed by the 2-staged screening protocol. 8 of these patients had no urinary decoy cell but with significant BK viraemia +/- Bx proven BKVAN. Only 2 cases had discordance urine cytology and BKV qPCR results. The remaining 6 cases had BKV qPCR tested outside the screening protocol time point. 1 patent had positive urinary decoy cell and low level BK viraemia (2 log) and subsequently found to have biopsy proven BKVAN and increasing viraemia after graft biopsy. The negative predictive value of 2-stage screening was 98.1%.

Conclusions: The incident of BKVAN is low in Hong Kong and kidney transplantation within first two years has higher risk of BKVAN. Routine BK virus screening with urine cytology and BKV qPCR has low false negative rate and would be the screening protocol of choice in places who cannot afford the high cost of routine BKV with qPCR. Extending the every 3 months screening period to up to 2-year post-transplant may help to improve the performance of this screening protocol.