P-2128. Intestinal Colonization Screening for Extended-Spectrum Beta-Lactamase-Producing Enterobacterales in Hematopoietic Stem Cell Transplant Recipients: When More Is Not Better

TL;DR

Screening for ESBL-E in HSCT patients increases carbapenem use but does not improve treatment effectiveness or survival.

Contribution

Demonstrates that routine ESBL-E screening in HSCT recipients increases carbapenem use without clinical benefit.

Findings

ESBL-E screening increased carbapenem use (29.6% vs. 14%) but did not improve appropriate antibiotic treatment.

In-hospital mortality was similar between screened and non-screened groups (6.2% vs. 4.5%).

The number needed to screen to predict ESBL-E bacteremia was 125.

Abstract

Several guidelines recommend rectal carriage screening (RCS) for extended-spectrum beta-lactamase-producing Enterobacterales (ESBL-E) to guide empirical antibiotic treatment (EAT) in neutropenic patients. However, this strategy may lead to an unnecessary use of carbapenems (CB). We aimed to evaluate the clinical impact of RCS for ESBL-E on appropriate EAT, CB use, and in-hospital mortality during the first episode of neutropenia in the pre-engraftment period of Hematopoietic Stem Cell Transplantation (HSCT).

Prospective cohort study of adult hospitalized patients who received HSCT between January 2017 and December 2024, with a 30-day follow-up. Systematic weekly RCS for ESBL-E was performed from the pre-transplant evaluation until hospital discharge. This practice was discontinued in 2023. We compared those who underwent RCS and those who did not (N-RCS). Chi2 and Mann–Whitney U tests were performed when appropriate.

318 patients with a median age of 52 years (IQR 30.8–61) were included. Allogeneic HSCT was performed in 33% of cases, of which 34.3% were from unrelated donors and 36.2% were haploidentical. The most common underlying diseases were multiple myeloma (42.5%) and lymphoma (30.2%). Enterobacterales bacteremia was diagnosed in 15.4% of patients, with a median of 16 days (IQR 11–20) from admission. ESBL-E RCS was performed in 86.2% of patients, with a positivity rate of 25.9%. The median time from admission to colonization detection was 6 days (IQR 0–14). The incidence of ESBL-E bacteremia was similar between carriers and non-carriers (5.6% vs. 2.5%, p=0.197). Comparisons between RCS and N-RCS were as follows: 1. CB use: 29.6% vs. 14%, p=0.003; 2. Appropriate EAT for Enterobacterales bacteremia: 81% vs. 83.3%, p=0.88; 3. In-hospital mortality: 6.2% vs. 4.5%, p=0.66. The number required to screen to predict ESBL-E bacteremia was 125.

RCS for ESBL-E significantly increased CB use during the first episode of febrile neutropenia. However, it had no impact on the appropriateness of EAT or mortality. These findings suggest that routine ESBL-E colonization screening may not be required in HSCT patients.

All Authors: No reported disclosures