The effect of the administration form of antibiotic therapy on the gut microbiome in patients with infected diabetic foot ulcers - DFIATIM trial

Abstract

Background: Diabetic foot infections (DFIs) contribute to the global disability burden. Beta-lactams are the most commonly used antibiotics for treating DFIs. However, the use of antibiotics may lead to disruption of the healthy balance of the gut microbiota, causing dysbiosis.

Methods: Patients with infected diabetic foot ulcers (iDFUs) were treated with two kinds of beta-lactams (amoxicillin/clavulanic acid or ceftazidime) according to microbial sensitivity of causative agents via bolus or continuous administration modes. Changes in the gut microbiome of patients were analyzed. Diabetic patients without iDFUs were used as a control group. 16 S ribosomal RNA gene amplicon sequencing was performed on stool samples collected from participants.

Results: Alpha diversity and beta diversity of gut microbiota of treated patients did not show significant differences between bolus and continuous modes. However, significant differences were observed between gut microbiota diversity of treated patients and control group. PCoA plots showed individualized responses of the patient's gut microbiota to antibiotics at different times using both administration forms associated with the pre-treatment state of microbiota composition. Enterococcus, Sellimonas, and Lachnoclostridium were the common bacterial markers differentially abundant in the gut microbiota of antibiotic-treated patients with iDFUs while Roseburia, Dorea, and Monoglobus were mainly abundant in the gut microbiota of patients without iDFUs. Predicted pathways like "Transporters", "ABC transporters" and "Phosphotranspherase system (PTS)" were upregulated in the gut microbiome of patients treated with bolus regime which may lead to increased intestinal barrier permeability.

Conclusion: The present study reported alterations in gut microbiota composition and functionality and provided the bacterial markers as well as potential metabolic signatures associated with each administration mode in patients with iDFUs, which may be used as a reference set for future studies of the effect of antibiotics administration on the gut microbiome of patients with iDFUs. This study shed light on the importance of understanding the effect of antibiotic administration form on gut microbiome in patients with iDFUs.

Trial registration: The DFIATIM Clinical Trial (Full title: "Rationalisation of ATB therapy in diabetic foot infection and its impact on the intestinal microbiota") is submitted to the European Union Clinical Trials Database under the EudraCT Number: 2019-001997-27. The date of registration is July 17th, 2020.

Keywords: Antibiotics; Beta-lactam; Bolus; Continuous; Diabetes; Diabetic foot infection; Diabetic foot ulcers; Gut microbiota.

Fig. 1

Experimental scheme of the study workflow showing patients with iDFUs treated with amoxicillin/clavulanic acid (AMC) and ceftazidime (CTZ) using bolus and continuous administration modes and different times of collection of stool: V0 (hospital day 0– inclusion visit), V2 (hospital day 7) and V5 (End of study– 8weeks after hospital discharge)

Fig. 2

Boxplots illustrating pielou_evenness, chao1, shannon and simpson diversity indices in bacterial community of fecal microbiome of control participants and patients with iDFUs treated with (A) amoxicillin/clavulanic acid (AMC), (B) ceftazidime (CTZ) using bolus and continuous administration modes at different times of collection V0 (before hospitalization), V2 (one week after hospital admission) and V5 (two months after hospital discharge), p-value ≤ 0.05 was considered statistically significant based on the Kruskal-Wallis test

Fig. 3

Principal Coordinate Analysis (PCoA) plots based on Bray Curtis distance of fecal microbiome from control participants and patients with iDFUs treated with: (A) amoxicillin/clavulanic acid (AMC), (B) ceftazidime (CTZ), Label numbers correspond to the patient’s ID. Dissimilarity analysis between the two groups was performed using Adonis with permutation 999. The confidence ellipses were traced in the 95% confidence. p-value ≤ 0.05 was considered statistically significant

Fig. 4

Relative abundance of bacterial populations at (A) family and (B) genus levels of fecal microbiomes of control participants and patients with iDFUs treated with amoxicillin/clavulanic acid (AMC) and ceftazidime (CTZ) using bolus and continuous administration modes at different times of collection V0 (before hospitalization), V2 (one week after hospital admission) and V5 (two months after hospital discharge)

Fig. 5

Linear discriminant analysis effect size (LEfSe) of taxa at genus level in fecal microbiomes from patients with iDFUs treated with (A) amoxicillin/clavulanic acid (AMC), (B) ceftazidime (CTZ), with alpha values of 0.05 and a threshold value of 2.0

Fig. 6

Selected KEGG functional pathways at level 3 predicted in the fecal microbiomes of control participants and patients with iDFUs treated with amoxicillin/clavulanic acid (AMC) and ceftazidime (CTZ) using bolus and continuous administration modes. p-value ≤ 0.05 was considered statistically significant