Influence of ESBL colonization status on gut microbiota composition during allogenic hematopoietic stem cell transplantation

Abstract

After allogeneic HSCT (allo-HSCT), the diversity of the intestinal microbiota significantly decreases. The changes can be rapid and are thought to be caused by chemotherapy, antibiotics, or intestinal inflammation. Most patients are exposed to prophylactic and therapeutic antibiotics during neutropenia and several patients are colonized by ESBL bacteria. We investigated the changes in gut microbiota composition in allo-HSCT, aiming at investigating if the acquisition of ESBL colonization may affect gut microbiome diversity during allo-HSCT. This was a single-center prospective pilot study. All patients consecutively admitted to the Haematological Unit of the City of Health and Science, Molinette Hospital in Turin, Italy, and undergoing allo-HSCT between August 2017 to August 2020 were enrolled in the study. Microbiome analysis on fecal samples were collected every 7 days from hospital admission to discharge and until 1 year after HSCT. 48 patients were enrolled in the study. At baseline 14 patients (29.16%) were colonized by MDR bacteria, mostly extended-spectrum beta-lactamase (ESBL)-producing gram negatives (N = 11; 78.57%). During allo-HSCT, one patient had a positive rectal swab for a carbapenemase-producing Klebsiella pneumoniae and eight patients lost the colonization during the hospital stay. Microbiota composition was compared between patients colonized by ESBL at baseline and non-colonized patients. Patients colonized by ESBL had a greater abundances of Bifidobacterium, Blautia, Clostridium, Coprococcus, L-Ruminococcus Mogibacteriaceae, Peptostreptococceae and Oscillospira, while non-colonized ESBL patients had a greater abundance of Actinomycetales, Staphylococcus and Sutterella. Moreover, microbiota composition of colonized by ESBL that retained colonization after HSCT showed an increased in abundances of Akkermansia, Dialister, Erysipelotrichaceae and Methanobrevibacter when compared with patients that become negative at rectal swabs. From a clinical perspective, the evolution of this prospective pilot study will be to investigate markers of gut barrier functions, SCFA productions and to correlate the predictivity of these parameters with risk of invasive infections and clinical outcomes in allo-HSCT population.

Keywords: Allo-HSCT; ESBL; MDR; Microbiome.

Fig. 1

Average-linkage clustering based on the Spearman distance of fecal samples of patients undergoing allo-HSC. Rows and columns are clustered by Ward linkage hierarchical clustering.

Fig. 2

Boxplots of differentially abundant ASVs in fecal samples across patients undergoing allo-HSCT. ASVs frequency is obtained from ASV table summarized at the higher taxonomy resolution and testing procedure based on Wilcoxon matched pairs test (FDR < 0.05).

Fig. 3

Boxplots of differentially abundant ASVs in fecal samples across ESBL carriers at baseline (T0; n = 11) and non-colonized patients. ASVs frequency is obtained from ASV table summarized at the higher taxonomy resolution and testing procedure based on Wilcoxon matched pairs test (FDR < 0.05).

Fig. 4

(A) Boxplots of differentially abundant ASVs in fecal samples across ESBL carriers that persisted after HSCT compared to patients who lost ESBL colonization at follow-up rectal swabs. ASVs frequency is obtained from ASV table summarized at the higher taxonomy resolution and testing procedure based on Wilcoxon matched pairs test (FDR < 0.05). (B) Boxplots of differentially abundant ASVs in fecal samples across ESBL carriers that persisted after HSCT compared to patients who lost ESBL colonization at follow-up rectal swabs the. ASVs frequency is obtained from ASV table summarized at the higher taxonomy resolution and testing procedure based on Wilcoxon matched pairs test (FDR < 0.05).