Streptococcus pneumoniae infections in 47 hematopoietic stem cell transplantation recipients: clinical characteristics of infections and vaccine-breakthrough infections, 1989-2005.

Streptococcus pneumoniae infections can cause serious systemic disease in patients following hematopoietic stem cell transplantation (HSCT), and the response to pneumococcal vaccine is inadequate in most HSCT recipients. We evaluated the clinical spectrum of pneumococcal disease and vaccine-breakthrough infections in HSCT recipients at our cancer center in a retrospective analysis of all consecutive episodes of S. pneumoniae infection from 1989 through 2005. During the study period, 7888 patients underwent HSCT at our center; we identified 47 HSCT recipients with 54 S. pneumoniae infections. The overall incidence of S. pneumoniae infection was 7 per 1000 HSCTs. The incidence was higher in recipients of allogeneic grafts than in recipients of autologous grafts (9 vs. 5 per 1000 HSCTs, respectively; p <or= 0.012). Thirty-two of the 47 patients (68%) had leukemia or lymphoma; 24 patients (51%) had a malignancy that was in complete remission. Seventeen patients (36%) had graft-versus-host disease, which was chronic in 16. The 54 episodes of S. pneumoniae infection occurred 433 +/- 669 days after HSCT; 5 patients (11%) had multiple episodes. Four episodes of S. pneumoniae infection occurred within 100 days following transplantation (45 +/- 49 d); 2 of these were during the pre-engraftment period and 3 were nosocomial infections. All 50 late post-transplant episodes (93%), which occurred 473 +/- 671 days following transplantation, were community-acquired infections (p < 0.00016). Thirty-three episodes (61%) presented as bacteremic pneumonia, 10 (19%) as pneumonia, and 8 (15%) as uncomplicated S. pneumoniae bacteremia alone. Logistic regression analysis showed that patients receiving systemic corticosteroids had increased risk for bacteremic pneumonia (odds ratio [OR], 11.7; 95% confidence intervals [CI], 1.371-99.280; p <or= 0.025). Patients with lymphoma (OR, 6.101; 95% CI, 1.106-33.640; p <or= 0.04) were more likely to develop pneumonia alone. In 27 episodes (93%) among 29 in which S. pneumoniae susceptibility testing was performed, the patients received concordant antimicrobials. Among the 6 patients (13%) who died of S. pneumoniae infection, 4 had S. pneumoniae bacteremic pneumonia and only 1 had chronic GVHD. Admission to a critical care unit at the onset of infection (OR, 15.5; 95% CI, 2.116-113.541; p <or= 0.007) and each unit increase in APACHE II score increase the probability of death (OR, 1.9; 95% CI, 1.181-3.054; p <or= 0.008). All 5 (11%) patients who developed vaccine-breakthrough S. pneumoniae infection (546 +/- 732 d following vaccination) had pneumonia, and in 4 patients concurrent bacteremia also occurred. A serious S. pneumoniae infection in HSCT recipients occurred more commonly in patients with lymphoma and patients receiving high-dose systemic corticosteroid therapy. It is noteworthy that there were no cases of extrapulmonary organ infection in HSCT recipients who presented with S. pneumoniae infection at our institution.