Low-Dose Antibiotic Prophylaxis Induces Rapid Modifications of the Gut Microbiota in Infants With Vesicoureteral Reflux

Abstract

Background and Objectives: Maturation of the gut microbiota (GM) in infants is critically affected by environmental factors, with potential long-lasting clinical consequences. Continuous low-dose antibiotic prophylaxis (CAP) is the standard of care for children with vesicoureteral reflux (VUR), in order to prevent recurrent urinary tract infections. We aimed to assess short-term GM modifications induced by CAP in infants. Methods: We analyzed the GM structure in 87 infants (aged 1-5 months) with high-grade VUR, previously exposed or naïve to CAP. Microbial DNA was extracted from stool samples. GM profiling was achieved by 16S rRNA gene-based next-generation sequencing. Fecal levels of short- and branched-chain fatty acids were also assessed. Results: 36/87 patients had been taking daily CAP for a median time of 47 days, while 51/87 had not. In all patients, the GM was predominantly composed by Bifidobacteriaceae and Enterobacteriaceae. Subgroup comparative analysis revealed alterations in the GM composition of CAP-exposed infants at phylum, family and genus level. CAP-exposed GM was enriched in members of Enterobacteriaceae and Bacteroidetes, especially in the genera Bacteroides and Parabacteroides, and showed a trend toward increased Klebsiella, often associated with antibiotic resistance. In contrast, the GM of non-CAP children was mostly enriched in Bifidobacterium. No differences were found in fatty acid levels. Conclusions: In infants with VUR, even a short exposure to CAP definitely alters the GM composition, with increased relative abundance of opportunistic pathogens and decreased proportions of health-promoting taxa. Early low-dose antibiotic exposure might bear potential long-term clinical risks.

Keywords: antibiotic prophylaxis; children; gut microbiota; urinary tract infection; vesicoureteral reflux.

Figure 1

The gut microbiota of infants exposed to CAP segregates from that of non-CAP infants. (A) Alpha diversity estimated according to Faith's Phylogenetic Diversity and the number of observed ASVs. No significant differences were found (p ≥ 0.2, Wilcoxon test). (B) Principal Coordinates Analysis (PCoA) based on weighted UniFrac distances between fecal samples. A significant separation between groups was observed (p = 0.015, permutation test with pseudo-F ratios). Ellipses include 95% confidence area based on the standard error of the weighted average of sample coordinates. Bacterial genera with the largest contribution to the ordination space are indicated with blue arrows (p ≤ 0.05, permutational correlation test, “envfit” function).

Figure 2

The GM dysbiosis in CAP infants is independent of potential confounding factors. Principal Coordinates Analysis (PCoA) based on weighted UniFrac distances between fecal samples from CAP vs. non-CAP infants, stratified by gender (male vs. female, A), age group (1-5 months, B), feeding modality (breastfeeding vs. formula vs. mixed, C), gestational age (term vs. preterm vs. postterm, D), probiotics consumption (yes vs. no, E), delivery mode (cesarean section vs. spontaneous delivery, F) VUR grade (grade III vs. IV vs. V, G), class of antibiotics (amoxicillin or amoxicillin + clavulanic acid, trimethoprim, oral cephalosporins, and nitrofurantoin, H) and geographical origin (Italian, Belgian, French, Lithuanian, Polish, Portuguese, and Turkish, I). No significant separation was found (p ≥ 0.118, permutation test with pseudo-F ratios). Ellipses include 95% confidence area based on the standard error of the weighted average of sample coordinates. See also Figure 1.

Figure 3

Gut microbiota structure in CAP vs. non-CAP infants. (A,B) Bar plots representing the relative abundance of the major phyla (A) and families (B) in the GM of CAP and non-CAP infants. (C,D) Box plots showing the relative abundance distribution of bacterial families (C) and genera (D) with a significant difference (p ≤ 0.05, Wilcoxon test; asterisk) or a trend (0.05 < p ≤ 0.1; hashtag) between the two groups. Only taxa with relative abundance >0.1% in at least two samples were considered.