• In our cohort, 8.4% of carbapenem-resistant Gram-negative bacilli (CR-GNB) negative patients on baseline rectal screening, were colonized during follow-up, and 23% of vancomycin-resistant Enterococcus (VRE) negative patients were colonized, with prior colonization being associated with re-colonization. • While 43% of patients experienced at least one BSI, only three concordant CR-GNB BSIs among 16 colonized patients (18.7%) and two concordant VRE BSIs among 45 colonized patients (4.4%) were observed. • These concordance rates support antibiotics de-escalation in clinically stable patients empirically receiving carbapenems and vancomycin during febrile neutropenia. Bacterial infections, especially bloodstream infections (BSIs), compromise the overall survival of patients with hematological malignancies undergoing cellular therapies. The interplay between rectal colonization by different types of multidrug-resistant organisms (MDRO) and subsequent BSIs is not fully determined. We aimed to describe the natural history and risk factors for colonization by MDRO, as well as its impact on subsequent BSIs in patients treated with cellular therapies in an environment of antimicrobial resistance. Retrospective, single-center cohort study of consecutive adult HM patients, treated with allogeneic or autologous hematopoietic cell transplantation (HCT) or chimeric antigen receptor T-cell (CAR-T) therapy between 2021-2024 in Attikon University Hospital (Greece), with ≥1 available rectal screening test for carbapenem-resistant Gram-negative bacilli (CR-GNB) and vancomycin-resistant Enterococcus faecium (VRE). Univariate and multivariate Cox regression was implemented for identifying predictors of CR-GNB and VRE colonization. We also studied the concordance between rectal colonization and subsequent BSIs. One-hundred eighty-two patients were included (male: 64%, mean age: 50.2±13 years, acute myeloid leukemia/myelodysplastic syndrome diagnosis: 55%, relapsing/refractory disease: 58%, allogeneic HCT: 85%). Median duration of hospitalization and neutropenia was 35 and 15 days, respectively. At baseline, 95% of patients were negative on rectal screening for both CR-GNB and VRE, and of those, 8.4% and 23% were colonized with CR-GNB and VRE during follow-up, respectively. Prior colonization with the same pathogen was independently associated with subsequent CR-GNB (HR: 13.0, 95% CI: 2.83-59.3) and VRE (HR: 4.83, 95% CI: 2.28-10.2) colonization. Seventy-nine patients had ≥1 BSIs (Gram-positive: 60%, Gram-negative: 38%), with three concordant CR-GNB BSIs among 16 colonized patients (18.7%) and two CR-GNB BSIs among 153 non-colonized patients (1.3%). Regarding VRE, we identified two concordant BSIs among 45 colonized patients (4.4%). Despite the high MDRO prevalence, we observed low rates of CR-GNB colonization in neutropenic patients treated with cellular therapies, with one patient out of five experiencing a concordant CR-GNB BSI. VRE colonization was more frequent, but it rarely resulted in a concordant BSI. These findings highlight the importance of infection control strategies and support de-escalation approaches in empirical antibacterial treatment of febrile neutropenia in this vulnerable population.
Kazakou et al. (Sun,) studied this question.