BK virus (BKV) is a polyomavirus that carries up to a 90% sero-prevalence in the general population. After infection the virus enters a latent phase in the renourinary epithelium. In hematopoietic stem cell transplant patients, BKV reactivation may lead to hemorrhagic cystitis with an incidence of 7%-70%. Aside from BKV, hemorrhagic cystitis (>72 h) has been linked to patient age, conditioning regimen, graft versus host disease (GVHD), and unrelated/mismatch donor. Symptoms range from dysuria to severe hemorrhage leading to clot formation, urinary obstruction and renal impairment. These complications have been shown to impact length of hospitalization, cost of care, and mortality. The incidence and impact of hemorrhagic cystitis in patients receiving fludarabine (Flu), busulfan (Bu)and anti-thymocyte globulin (ATG) conditioning is unknown. We carried out this study to determine the incidence and its effect on patient outcomes including overall survival, relapse, GVHD, and health care resource use.

Patient Selection and Methods

We reviewed the medical records of 661 consecutive patients undergoing allogeneic stem cell transplant from January 1, 2003 to January 1, 2013. Hemorrhagic cystitis was diagnosed with symptoms (cystitis with microscopic or macroscopic hematuria, in the absence of other causes such as bacterial/fungal infection or nephrolithiasis) of at least 1 week following stem cell transplant. Microbiological confirmation was obtained routinely after 2005. Controls were matched to cases in a 2:1 ratio in the following priority: age (+/- 10 years), donor type, disease type, GVHD prophylaxis and disease status. All cases received FluBuATG +/- total body irradiation (400 cGy), and all except two received methotrexate (MTX) and cyclosporine (CsA) for GVHD prophylaxis (only received CsA).


Hemorrhagic cystitis was identified in 117/661 transplant recipients over ten years, for an incidence of 17.7%. After removing cases with a previous transplant, unsuitable matched controls, and incomplete data, 94 evaluable cases were matched to 188 controls. The cohort median age was 50.25 years. Diagnoses included AML (38.3%), ALL (21.3%), MDS (11.7%), CLL (13.8%), NHL (6.4%) and myeloproliferative syndromes (6.4%). Disease statuses included CR1 (31.9%, CR2 (19.1%), untreated (14.9%), PR (13%) and primary refractory (4.3%). Of the cases, 77 of 83 (92.8%) tested positive for BKV in urine by PCR. Median onset was 45 days (range 12 – 279 days). Median BK viral copies was 2242.5 x 106copies/mL (range 0.285 to 4530000). Cases had higher proportions of significant (Grade 2-4) acute GVHD (43.6% vs 27.0%, P=0.0052) and chronic GVHD (cGVHD) requiring steroids (34.9% vs 18.6%, P=0.004). The median onset of significant acute GVHD (aGVHD) was 27 days (range 11 – 103 days), and for cGVHD was 108 days (range 29 – 236 days). Out of 41 aGVHD cases, 8 occurred after BKV hemorrhagic cystitis diagnosis while 23 occurred prior. In 30 cGVHD cases, 23 occurred after BKV hemorrhagic cystitis diagnosis while 7 occurred prior. Cases also had longer hospital stays in the 6 months following transplant (65.5 days vs 40.5 days, P<0.0001). There were no differences in regards to overall survival (HR=1.088, 95% CI 0.7519-1.580), or relapse-free survival (HR=1.057, 95% CI 0.7364-1.519). Evaluating all patients with significant aGVHD, there was no difference in regards to overall survival between controls and cases (HR=0.8042, 95% CI 0.5205-1.205). This was similar for with patients with cGVHD (HR=0.9949, 95% CI 0.6020-1.643)


This retrospective study shows that BKV hemorrhagic cystitis is a common complication of allogeneic stem cell transplant. While it does not impact overall survival even in ATG-conditioned patients, it is associated with GVHD and longer hospitalizations. Future studies are warranted to determine how BKV hemorrhagic cystitis is influenced by GVHD. Additionally, studies into the pathophysiology of BK virus reactivation with the eventual goal of developing prophylaxis, preemptive therapy or treatment are also warranted.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract