Risk Factors for Mortality in Children with Hemorrhagic Cystitis after Hematopoietic Transplant

Extended Summary

Introduction

Hemorrhagic cystitis (HC) is a devastating complication of bone marrow (BMT) and stem cell transplant (SCT). Much of the literature has focused on exclusively adult patient populations, with limited evidence regarding risk factors for mortality and morbidity among pediatric HC patients.

Objective

To examine factors associated with all-cause mortality in children with HC after BMT/SCT.

Methods

The Pediatric Health Information System database was queried for patients with ICD-9/10 codes for hematopoietic transplant and gross hematuria, hematuria unspecified, or cystitis with hematuria. Multivariable logistic regression examined association of medical and surgical interventions frequently employed for hemorrhagic cystitis with mortality and genitourinary morbidity, defined as having received instillation of any bladder medication or having undergone any genitourinary procedure.

Results

A total of 811 patients, mean age of 12.4 years and 62% male, were included. Primary diagnosis included 388 (49%) leukemia/lymphoma, 182 (22%) blood dyscrasia, 99 (12%) solid organ tumor, 27 (3%) metabolic disease, 115 (14%) unknown. Transplant type included 377 (46%) bone marrow, 329 (41%) stem cell, 105, and (13%) unknown. Performing any bladder instillation (p<0.0001) or any type of GU procedure (p<0.0001) was significantly associated with mortality. On multivariate analysis, dialysis (OR=10.7, 95% CI=5.7-20.2), genitourinary morbidity (OR=4, 95% CI=2.2-6.8) and intravenous cidofovir (OR=2.0, 95% CI=1.2-3.3) were significantly associated with all cause mortality, Table 1. Having an underlying diagnosis of blood dyscrasia was protective against mortality (OR=0.425, CI=0.205-0.88).

Discussion

In this large retrospective study evaluating factors associated with mortality in children with HC, all cause mortality was found to be 11%. This is probably an underrepresentation of true mortality in this population, as many patients discharged from the hospital likely die outside the hospital at home or hospice care. This study supports the current literature that invasive GU procedures are not associated with increased survival in patients with severe HC. This study is limited by retrospective use of a billing database that has the potential for errors in data entry and missing data. Patients who were discharged from the hospital were not captured by the PHIS which only collects data from inpatient stays.

Conclusions

Patients with HC who received dialysis, intravenous cidofovir, or underwent GU intervention had significantly higher all-cause mortality. High grade HC is a marker of disease severity and efforts should be made by urologists and oncologists to maximize quality of life and limit futile treatments in this patient population.