Children after Allogeneic Hematopoietic Cell Transplantation Can Develop BK Virus Nephropathy: Pilot Data Using Urinary Polyomavirus-Haufen Testing

Overview: BK viremia (BKV) occurs in >15% of patients after hematopoietic cell transplantation (HCT), primarily in association with cystitis. It is unknown if BK virus-induced renal disease (polyomavirus nephropathy, PVN) is also common after HCT. Kidney biopsy is the gold standard for the diagnosis of PVN but has a high risk of bleeding after HCT. The non-invasive urinary PV-Haufen test is sensitive/specific for PVN in renal transplant recipients with positive/negative predictive values >90%. PV-Haufen are cast-like BK aggregates that form in virally injured kidney tubules, are shed, and can easily be detected in urine samples. PV-Haufen do not form in the bladder and are not seen in cystitis without PVN. In this pilot study, we used the PV-Haufen test for the first time in children after HCT to screen for PVN.

Methods: Subjects with BKV (potential cases) and without BKV (presumed controls) after allogeneic HCT were selected from cohorts at Cincinnati Children’s Hospital and the Children’s Hospital of Philadelphia. Urine samples (10-25 ml) were fixed in a 1:1 ratio with 4% paraformaldehyde, stored at 4C, and tested for urinary PV-Haufen using negative staining electron microscopy (JASN, 2009, 20(2):416-27). Urine samples containing ≥1 PV-Haufen were classified as positive, supporting a diagnosis of PVN.

Results: We tested 20 voided urine samples from 15 subjects (median age 11.5 years, range 5.2-29.8; median HCT follow up time 306 days, range 87-815). Urine Haufen (Fig 1) were detected in 4/15 subjects (27%; 5/20 urine samples), including one subject with autopsy confirmed PVN (Fig 2). The table shows associations between urinary PV-Haufen, peak BKV, and clinical outcomes after HCT.

Conclusions: We provide pilot data potentially indicating a high risk of PVN after HCT. PV-Haufen were associated with kidney injury requiring renal replacement therapy and may thus serve as a non-invasive test to identify patients with BKV who are at high risk for PVN. More research is needed including longitudinal follow-up, careful monitoring of renal function, and correlation of urinary PV-Haufen data with BKV and tissue findings, when available.

	Haufen+(n=4)	Haufen- (n=9)	p-value*
Peak BKV after HCT (copies/mL)	1,255,140 [144,512-1,900,000,000]	1,173,953 [1537-6,000,000]	0.44
Hemorrhagic cystitis	3 (75%)	6 (67%)	1.00
Dialysis	3 (75%)	0 (0%)	0.01
Death	3 (75%)	2 (22%)	0.22

Data shown as median [IQR] or n (%).*Wilcoxon rank-sum or Fisher’s exact. Urine samples from 2/15 tested subjects could not be evaluated.