Invasive diagnostic procedures for pulmonary infiltrates in pediatric hematopoietic stem cell transplant recipients

Abstract

To evaluate the role of BAL, CTB, and OLB in the management of pulmonary infiltrates in pediatric HSCT recipients, we conducted a retrospective review of clinical records of pediatric HSCT recipients. Data were analyzed using Chi-square for dichotomous and anova for continuous variables. Logistic regression was used to adjust confounding variables for diagnostic yield. Forty patients underwent 44 separate procedures. Infections were the prevailing cause of infiltrates with a positive diagnostic yield (96%). CTB and OLB were performed more often in patients with focal infiltrates compared with BAL (100%, 71% vs. 22%; p < 0.01). Adverse events were not significantly different across the three procedures. OLB more often yielded information that led to change in medical management (71% vs. 0%, 34%; p < 0.05) compared with CTB and BAL. Patients who had a positive diagnostic yield had no apparent survival advantage when compared with those in whom a procedure yielded no information. Logistic regression demonstrated that focal infiltrate was the only independently predictive variable for identifying a cause of pulmonary infiltrate. In conclusion, all three invasive diagnostic procedures were safe. Having a focal infiltrate was independently and significantly associated with having a positive diagnostic yield.