Secular trends of bloodstream infections during neutropenia in 15 181 haematopoietic stem cell transplants: 13-year results from a European multicentre surveillance study (ONKO-KISS)

Objectives: Antibacterial resistance is emerging in patients undergoing haematopoietic stem cell transplantation (HSCT), and most data on the epidemiology of bloodstream infections (BSI)-causing pathogens come from retrospective single-centre studies. This study sought to investigate trends in the epidemiology of BSI in HSCT patients from a prospective multicentre cohort.

Methods: We investigated changes in the incidence of causative organisms of BSI during neutropenia among adult HSCT patients for 2002-2014. The data were collected from a prospective cohort for infection surveillance in 20 haematologic cancer centres in Germany, Austria and Switzerland (ONKO-KISS).

Results: A total of 2388 of 15 181 HSCT patients with neutropenia (1471 allogeneic (61.6%) and 917 autologous (38.4%) HSCT) developed BSI (incidence 15.8% per year). The incidence of Gram-negative BSI increased over time both in patients after allogeneic HSCT (allo-HSCT) and autologous HSCT (auto-HSCT). BSI caused by Escherichia coli in allo-HSCT patients increased from 1.1% in 2002 to 3.8% in 2014 (3/279 vs. 31/810 patients, p <0.001), and the incidence of BSI caused by enterococci increased from 1.8% to 3.3% (5 vs. 27 patients, p <0.001). In contrast, the incidence of BSI due to coagulase-negative staphylococci decreased in allo-HSCT patients from 8.2% to 5.1%, (23 vs. 40 patients, p <0.001) and in auto-HSCT patients from 7.7% to 2.0% (13/167 vs. 30/540 patients; p = 0.028 for period 2002-2011). No significant trends were observed for the incidence of BSI due to methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci or extended-spectrum β-lactamase-producing Enterobacteriaceae. The BSI case fatality remained unchanged over the study period (total of 477 fatalities, 3.1%).

Conclusions: The incidence of Gram-negative BSI significantly increased over time in this vulnerable patient population, providing evidence for reevaluating empiric therapy for neutropenic fever in HSCT patients.