Antibiotic Prophylaxis Therapy for Pediatric Patients Undergoing Hematopoietic Stem Cell Transplant (HSCT): A Tale of 2 Centers

Introduction: Bacterial infections are a leading cause of morbidity and treatment-related mortality in children following hematopoietic stem cell transplant (HSCT). Meta-analysis of studies of bacterial prophylaxis in adult oncology patients with neutropenia showed a significantly decreased risk of death associated with prophylaxis regimens. However, the use of prophylaxis vs. empiric treatment is controversial in the setting of HSCT and data in children is limited. The concerns for using prophylaxis therapy are development of antibiotic resistance and increased drug-related toxicity. An empiric approach to antibiotic therapy prompts the concern that therapy is delayed until an infection has already occurred. Our study aimed to compare the bacteremia rates in two pediatric centers that use contrasting approaches; empiric vs. prophylaxis.

Methods: We compared the incidence of bacterial infections in pediatric HSCT patients in two units; The Hospital for Sick Children (SickKids) where an empiric antibiotic strategy is utilized and Vanderbilt University Medical Center, where prophylaxis antibiotic (Cefipime) is given when ANC<500/ul or on the day of transplant, which ever is earlier. Baseline characteristics were compared between the 2 groups with 2 sample tests, where categorical variables and continuous variables were evaluated using double-sided Fisher exact tests respectively.

Results: 224 pediatric patients from SickKids and 294 pediatric patients from Vanderbilt who underwent autologous and allogeneic HSCTs between 2005-2010 were evaluated. Total bacteremia rate (Gram positive and Gram negative) was significantly higher at SickKids (68/224) vs. Vanderbilt (49/294), p<0.001).  At SickKids, 19% presented with Gram positive infection as their first infection  at a mean of 8.9 d after transplant and 11% with gram negative infection at a mean of 4.0 d after transplant.At Vanderbilt, 14% presented with gram positive infection at a mean of 9.5 d after transplant and 2% with gram negative infection at a mean  of 5.2 d after transplant. No treatment-related mortality in the first 100 days was attributed to bacterial infections in either center. Using a multivariable model including: age, diagnosis (malignant vs. non-malignant) and type of transplant ( allogeneic vs. autologous),  only antibiotic prophylaxis approach  resulted in a  significantly lower gram negative bacteremia (p=<0.001) and a trend toward lower incidence of gram positive bacteremia (p=0.052).  

Conclusions: The use of antibiotic prophylaxis in pediatric HSCT decreased the incidence of bacteremia during transplant.  The use of antibacterial prophylaxis in pediatric patients undergoing HSCT should be considered, and prospective studies are needed to confirm our results.