Development of BK Virus-Associated Hemorrhagic Cystitis (HC) Is Associated with Decreased Survival Post-Allogeneic Hematopoietic Stem Cell Transplantation (allo-HSCT)

Infection by BK virus in the allo-HSCT setting is associated with the development of HC, which is a major cause of morbidity. There are few studies evaluating the impact of BK virus-associated HC on survival post allo-HSCT.

To evaluate the incidence of HC by BK virus in patients after allo-HSCT and its impact on survival.

We retrospectively reviewed the medical charts of 107 patients who underwent allo-HSCT at Hospital Israelita Albert Einstein from July 2007 until November 2011. HC was defined by the presence of any degree of unexplained hematuria and positivity for BK virus in a urine sample by quantitative polymerase chain reaction (PCR) assay. CMV reactivation was defined as positivity in the antigenemia assay or greater than 165 copies in a quantitative PCR assay. Three patients were excluded because PCR results for BK virus were inconclusive, with 104 patients being analyzed in the final cohort. Cumulative incidence (CI) of HC, CMV reactivation and acute graft-versus-host disease (GVHD) were estimated taking into account the competing risk of death. Overall survival (OS) was estimated by the Kaplan-Meier method. Gray model was used for regression analysis of factors associated with the development of HC. Hazard ratios (HRs) were estimated by a Cox multivariable proportional hazards model, considering HC, CMV reactivation and acute GVHD as discrete time-varying covariates.

Median age was 28 years (range 6 months–76 years), and 60.5% of patients were male. About 37% had high-risk disease (refractory leukemia/lymphoma or 2^nd-transplantation). Source of HSCs included matched related donors (37%), 10/10 HLA-matched unrelated donors (24%) and cord blood/haploidentical donors in 39%. The conditioning regimen was myeloablative in 81% of cases. The median follow-up of the whole cohort was 450 days (range 9–1624 days). At 1 year, the cumulative incidence of HC was 30.5% (95% confidence interval [CI] 21.8%–39.7%). The 1-year incidence of CMV reactivation and acute GVHD (all grades) was 57.1% and 39.7%, respectively. In a multivariate analysis taking into account age, sex, risk of disease, source of HSCs, intensity of conditioning, CMV reactivation and acute GVHD, only receiving cells from cord blood/haploidentical donors was associated with an increased incidence of HC (subhazard ratio 4.04, 95% CI 1.31–12.49, p = 0.015). The 1 year-OS of the whole cohort was 55% (95% CI 44–62%). Patients who developed HC had an inferior OS (1 year: 18% vs. 70%; HR= 4.40, p <0.0001; 95% CI 2.37–8.14). In the multivariate Cox analysis for OS, after adjusting for age, sex, disease risk, source of HSCs, intensity of conditioning, CMV reactivation and development of acute GVHD, development of HC was associated with an increased mortality (table).

In this cohort, the development of CH was associated with an inferior OS in patients undergoing allogeneic HSCT. Even after adjusting for several variables, including development of acute GVHD and CMV reactivation, HC still remained an important factor associated with decreased survrival. It is possible that HC is not the direct cause of the increased mortality, but is rather a surrogate marker of a state of severe immunosuppression and increased risk of dying in the post-transplant setting. Nonetheless, our results suggest that the development of BK virus-associated HC in allo-HSCT patients is associated with inferior survival and future studies should confirm this finding and seek strategies to prevent this complication.

No relevant conflicts of interest to declare.