Outcomes of Human Adenovirus Infection and Disease in a Retrospective Cohort of Pediatric Hematopoietic Cell Transplant Recipients.

Abstract

Human adenoviruses (HAdVs) are associated with significant morbidity and death after hematopoietic cell transplantation (HCT). In this study, we sought to determine the incidence of HAdV infection among pediatric HCT recipients in the polymerase chain reaction (PCR) testing era, identify risk factors for viremia among patients undergoing HAdV surveillance, and assess the effectiveness of preemptive cidofovir. A single-center retrospective cohort of patients who underwent a transplant within a 10-year period was assembled. The incidence of and outcomes of patients with HAdV infection and disease were determined by PCR results and chart review. A Cox regression model was used for surveilled allogeneic HCT recipients to identify factors associated with viremia. We also used a discrete-time failure model with inverse probability treatment weights to assess the effectiveness of preemptive cidofovir for infection. Among 572 HCT recipients, 76 (13.3%) had ≥1 sample that was HAdV PCR positive (3.5% of autologous HCT recipients and 19.7% of allogeneic HCT recipients). Among 191 allogeneic HCT recipients under surveillance, 58 (30.4%) had HAdV detected from any source, and 50 (26.2%) specifically had viremia. The mortality rate was higher in allogeneic HCT recipients with HAdV infection versus those without infection (25.9% vs 11.3%; P = .01). Factors associated with infection included an age of 6 to 12 years, an absolute lymphocyte count of <200 cells/μL, recent prednisone exposure, and recent bacteremia. Preemptive cidofovir was not associated with a reduced risk of infection progression (odds ratio, 0.96 [95% confidence interval, 0.30-3.05]). HAdV infection is common and associated with an increased rate of death after allogeneic HCT. Using prediction models that incorporate factors associated with HAdV might help target surveillance. Preemptive cidofovir therapy was not protective in a subset of HAdV-positive patients. Larger observational or randomized investigations are necessary, because the utility of surveillance requires effective preemptive therapies.