Microbiology Spectrum (Dec 2023)

Are enterococcal bloodstream infections an independent risk factor for a poorer 5-year survival or just a marker for severity of illness?—The Munich multicentric enterococci cohort

  • Kathrin Rothe,
  • Tobias Bachfischer,
  • Siranush Karapetyan,
  • Alexander Hapfelmeier,
  • Milena Wurst,
  • Sabine Gleich,
  • Karl Dichtl,
  • Roland M. Schmid,
  • Julian Triebelhorn,
  • Laura Wagner,
  • Johanna Erber,
  • Florian Voit,
  • Rainer Burgkart,
  • Andreas Obermeier,
  • Ulrich Seibold,
  • Dirk H. Busch,
  • Patrick C. Rämer,
  • Christoph D. Spinner,
  • Jochen Schneider

DOI
https://doi.org/10.1128/spectrum.02585-23
Journal volume & issue
Vol. 11, no. 6

Abstract

Read online

ABSTRACT To assess the long-term survival of patients with enterococcal bloodstream infections (BSI), encompassing various species and resistance patterns compared to Escherichia coli (E. coli) BSI. Between 2010–2019, 3,290 enterococcal and 3,415 E. coli BSI were retrospectively screened in seven hospitals in Munich, Germany. All vancomycin-resistant (VRE), vancomycin/linezolid-resistant (LVRE), and linezolid-resistant (LRE) Enterococcus faecium (ECFM) BSI were included. Enterococcus faecalis (ECFA), vancomycin/linezolid-susceptible ECFM, and E. coli BSI were randomly assigned. Cox regression analysis was used to assess survival as the primary endpoint and was adjusted for limiting prognostic factors, which were measured for their importance using a random forest model (RFM). We analyzed 952 patients with 916 episodes of enterococcal BSI and 193 episodes of E. coli BSI. RFM identified multimorbidity and markers for disease severity as most indicative of low survival in enterococcal BSI. The 5-year survival was significantly lower for enterococcal BSI than for E. coli BSI (23.9% vs 42.3%; P < 0.001). This difference remained significant in the Cox regression analysis after adjusting for 17 prognostic factors and excluding patients with limited life expectancy (metastatic tumor disease, Charlson-Comorbidity-Index ≥5). Adjusted 5-year survival was similar for E. coli and ECFA but significantly different between ECFA and ECFM (29.2% vs 21.7%; P = 0.002). The analysis conducted on monomicrobial ECFM and VRE BSI indicated that their respective 5-year survival was similar (19.6% vs. 21.2%; P = 0.753). ECFM BSI seems to be an independent risk factor for poor long-term survival. However, additional vancomycin resistance does not appear to be a contributing factor. IMPORTANCE The present study provides a substantial contribution to literature, showing that patients with enterococcal bloodstream infections (BSI) have a lower survival rate than those with Escherichia coli (E. coli) bloodstream infections after adjusting for 17 limiting prognostic factors and excluding patients with a limited life expectancy [metastatic tumor disease, Charlson Comorbidity Index (CCI) (greater than or equal to) 5]. This difference in the 5-year long-term survival was mainly driven by Enterococcus faecium (ECFM) bloodstream infections, with vancomycin resistance not being a significant contributing factor. Our findings imply that E. faecium bloodstream infections seem to be an independent risk factor for poor long-term outcomes. As such, future research should confirm this relationship and prioritize investigating its causality through prospective studies.

Keywords