Infections diagnosed in the first year after pediatric stem cell transplantation.

BACKGROUND: Cumulative incidence of infections in the first year posttransplantation in adult patients has been well-described. Such description is less than complete for pediatric stem cell transplantation (SCT) patients. Further among those patients who have been infected, analysis of risk factors for infection has not been well-described for a large cohort of pediatric SCT patients.
METHODS: We conducted a retrospective cohort study of infections in the first year after SCT at Duke University Medical Center. We recorded all infections in the first year after transplantation. We determined incidences for 6 categories of infection: gram-negative rods; gram-positive cocci; yeast species; Aspergillus sp.; adenovirus; and cytomegalovirus. We determined incidences based on type of transplant and days post transplantation. We also completed bivariable and multivariable analysis of risk factors [neutropenia, graft vs. host disease (GVHD) and GVHD treatment] for infection type among those children who were infected.
RESULTS: We evaluated 510 transplants in 485 children. There were 584 infections in the first year after transplantation. During the first 30 days posttransplantation, type of transplantation did not predict incidence of infection or type of infection. After 30 days children who received unrelated cord blood transplant and matched unrelated donor transplant were at much higher risk of infection than were patients who received autologous, matched sibling or haploidentical transplant (P < 0.001). Patients who received unrelated cord blood or matched unrelated donor transplantation were at higher risk of aspergillosis (P = 0.002), candidiasis (P = 0.005) and adenovirus (P < 0.0001) but not cytomegalovirus (P = 0.18). In analysis of risk factors among those infected, patients with aspergillosis were more likely to have severe GVHD: multivariable 1 year risk ratio, 7.5; 95% confidence interval, 3.0, 18.4. Neutropenia was more strongly associated with gram-negative rod infection than any other type of infection.
CONCLUSIONS: The incidence of infection immediately after transplantation did not differ significantly by type of transplant in this pediatric population. Type of transplant predicted increased incidence of infection 30 days posttransplantation and increased incidence of infection with several organisms traditionally associated with a high mortality rate in the transplant population.