BK Virus Disease Following Allogeneic Stem Cell Transplantation: A Cohort Analysis

Abstract 1927

BK virus (BKV) reactivation after allogeneic hematopoietic stem cell transplantation (HSCT) affects the genitourinary (GU) tract with manifestations ranging from asymptomatic viruria to severe hemorrhagic cystitis. Knowledge about epidemiology, morbidity and clinical spectrum of BKV disease is limited. We studied a recent HSCT cohort at our institution to assess and quantify the incidence and severity of BKV disease.

Between January 1, 2010 and December 31, 2011, 491 patients underwent first HSCT. BKV disease was defined as detection of BKV by PCR testing in association with GU symptoms. BKV disease was considered severe when patients had at least 1 of the following: hematuria with clot formation, GU tract inflammation on imaging, need for invasive GU interventions, or hospitalization for BKV disease management. Medical records were reviewed for patient and HSCT characteristics. For patients with BKV disease, reason for testing, clinical presentation and course, imaging results, and treatments were captured. Time at risk was censored at time of death, second HSCT or on July 1, 2012. Fisher's exact or Wilcoxon test were used to compare variables. Cox modeling was used to analyze potential risk factors for BKV disease.

168 (34.2%) patients were tested for BKV at our institution during the study period. Median time to testing was 61 days (range, 2–663; IQR 19.5 – 114). 88 patients (17.9%) were diagnosed with BKV disease with an incidence rate of 0.54/1000 patient-days (95% CI, 0.43 – 0.66). Most common symptoms at presentation and during a first BKV disease episode are shown in the Table.

31 patients had severe BK disease (35.2%, 6.3% of the cohort): 17 patients (19.3%) had hematuria with clot formation; 16 patients (18.2%) demonstrated GU tract inflammation on imaging studies, including bladder wall (14, 15.9%) or ureteral (5, 5.7%) thickening. 4 patients were hospitalized for management of BKV associated symptoms. Invasive GU interventions were performed in 16 (18.2%) patients including Foley catheter placement in 12, bladder irrigation in 7, intravesicular cidofovir in 3, cystoscopy in 2, and alum instillation in 1. Additional prescribed treatments during BKV disease episodes included quinolones (39, 44.3%), antispasmodics (29, 33.0%), IVIG (20, 22.7%) and pain medications (18, 20.5%). Cidofovir was used in 7 (8.0%) patients and leflunomide in 8 (9.1%). 3 patients had concomitant bacterial GU infections; 1 patient had GU adenovirus disease. Patients (n=71) were symptomatic for a median of 31 days (range 2–385, IQR 9–67). Additional BKV disease episodes were seen in 26/88 patients (29.5%).

Median initial urine BKV load was 3.4×10⁸ copies/mL (range; 600, >1×10¹⁰). Median peak urine BKV load during first BKV disease episode was 1.5×10⁹ c/mL (range; 600, >1×10¹⁰). Blood BKV loads were obtained in 35 patients, with median initial BKV load of 800 c/mL (range, 0, 1.52×10⁶) and median peak BKV load of 2000 c/mL (range, 0, 1.52×10⁶). Detection of BKV viremia was 74% sensitive and 38% specific for severe BKV disease.

Cohort characteristics associated with increased BKV disease risk included myeloablative conditioning (p=0.01), cyclophosphamide conditioning (p=0.0008), cord blood HSCT (p=0.03), GVHD prophylaxis with mycophenolate (MMF, p=0.0006) and acute GVHD (aGVHD) grades II-IV (p< 0.0001). On multivariate Cox modeling, time-dependent aGVHD (adjusted HR [aHR] 3.65, 95%CI 2.23 – 5.97), MMF use (aHR 3.11, 95%CI 1.76 – 5.48), and cyclophosphamide use (aHR 1.95, 95%CI 1.25 – 3.04) remained significant. Time-dependent aGVHD (aHR 5.02, 95%CI 2.20 – 11.5), cord blood HSCT (aHR 4.60, 95%CI 1.67 – 12.7) and cyclophosphamide use (aHR 2.17, 95%CI 1.04 – 4.56) were independent risk factors for severe BKV disease.

BKV disease is a common complication of HSCT, associated with significant and prolonged morbidity, especially in the setting of aGVHD, cyclophosphamide and MMF use, and after cord blood HSCT. Prospective studies are needed to better define the morbidity of BKV disease and to properly inform the impact of future prophylaxis and treatment trials.