Longer use of certain antimicrobial drugs was associated with higher odds of bloodstream infection caused by vancomycin-resistant Enterococcus faecium among patients with blood cancers and stem cell transplant recipients, according to a Canadian study. Patients who developed these infections also had approximately twice the hazard of death from any cause compared with those who did not.
The findings, published in Open Forum Infectious Diseases, highlight the association between antimicrobial treatment duration and infection risk, particularly among patients with acute myeloid leukemia and recipients of allogeneic stem cell transplants.
First author Khuloud Aldhaheri, MD, of United Arab Emirates University in Al Ain, and colleagues reviewed patient records for a matched case-control study at Vancouver General Hospital in British Columbia, Canada.
“While VRE BSI epidemiology is well characterized in certain populations, including patients undergoing stem cell transplant, evidence remains limited in mixed units that include patients with diverse treatment exposure, hematologic malignancies, and transplant histories,” the authors wrote. “Studies examining VRE BSI risk factors and outcomes across these cohorts, within the context of shared stewardship practices and local colonization pressure, remain sparse.”
The researchers reviewed records from 3,301 adults admitted to the hospital's leukemia and bone marrow transplant unit between January 2014 and January 2024. They identified 100 patients who developed VREfm BSI and matched them with 400 controls who did not develop the infection.
Patients were matched by reason for admission and admission period. The researchers examined underlying diagnoses, transplant type, severe neutropenia, intensive care unit admission, antimicrobial exposure and deaths.
Infection rates increased over time
The rate of VREfm BSI more than tripled, increasing from 0.02 cases per 1,000 patient-days during 2014-2018 to 0.07 during 2019-2024.
Patients with AML accounted for 54% of infection cases vs. 36% of controls. Recipients of allogeneic stem cell transplants accounted for 36% of cases vs. 14% of controls, while recipients of autologous transplants accounted for 1% vs. 17%, respectively.
Patients who developed infections also had longer hospital stays, with a median of 42.5 days vs 22 days for those without infections.
Among patients who developed infections, 41% were admitted to the ICU vs. 5% of controls, while 98% had severe neutropenia vs. approximately 80% of controls.
After adjustment for other factors, patients with AML and/or those who received allogeneic stem cell transplants had nearly four times the odds of VREfm BSI compared with other patients (adjusted odds ratio, 3.90). ICU admission was associated with 5.6 times the odds of infection (adjusted odds ratio, 5.63).
Antimicrobial duration associated with infection
Longer use of several antimicrobial drug classes was independently associated with higher odds of infection in the primary adjusted analysis. Each additional day of treatment was associated with a 9% increase in the odds of VREfm BSI with carbapenems, 9% with beta-lactam/beta-lactamase inhibitor combinations and 12% with systemic antifungals.
Neither vancomycin nor fluoroquinolone exposure was independently associated with infection after adjustment for other factors.
Additional analyses showed that the odds of infection generally increased with longer antimicrobial exposure, particularly beyond about 15 days, although the associations were nonlinear. Estimates became less precise with longer treatment, so the researchers could not identify a specific duration at which infection risk increased.
When evaluated separately from AML, allogeneic stem cell transplantation remained associated with infection. However, the association between AML and infection was no longer statistically significant.
In another analysis that standardized antimicrobial exposure by the number of inpatient days before infection or discharge, no individual antimicrobial class remained significantly associated with infection.
Bloodstream infection associated with higher hazard of death
Patients with VREfm BSI had approximately twice the hazard of death from any cause compared with those without infection, after adjustment for age and sex (adjusted hazard ratio, 2.09).
During 2019-2024, death rates were 1.3 per 1,000 patient-days among patients with VREfm BSI vs. 0.5 among those without infection. During 2014-2018, the corresponding rates were 0.7 vs. 0.4.
Patients with VREfm BSI also had lower survival rates at one and three years of follow-up.
The researchers noted that antimicrobial use may disrupt protective bacteria in the gut, allowing resistant bacteria to multiply and potentially enter the bloodstream. However, the study could not establish whether antimicrobial use directly caused the infections.
The study had several limitations, including its reliance on past patient records from a single hospital, incomplete information about reasons for antimicrobial treatment and transplant complications, and limited data on bacterial colonization after routine screening ended in 2018. Differences in illness severity also may have influenced the findings.
The authors concluded that antimicrobial stewardship, particularly “minimizing unnecessary antibiotic duration,” may help reduce VREfm BSI among these patients.
The authors reported receiving no funding or support for the manuscript. Coauthors Ryan J. Stubbins, MD, MSc, and Marthe K. Charles, MD, MSc, reported financial relationships with pharmaceutical and diagnostic companies, including consulting fees, honoraria, grants or contracts, travel support, and equipment or materials. Cesilia Nishi, PharmD, reported an unpaid leadership role with the British Columbia branch of the Canadian Society of Hospital Pharmacists. The remaining authors reported no potential conflicts of interest.