Fecal microbiota profile at diagnosis in the French IMODI cohort of breast cancer
Background: Breast cancer (BC) represents the most common cancer worldwide. Its etiology remains elusive involving many interrelated factors (genetics, environment, lifestyle, microbiota).
Objective: The aim of this study was to investigate the fecal microbiota in patients with a localized BC at diagnosis.
Methods: From a total of 122 women enrolled in 2015-2018 in 2 French referring hospitals, 49 patients were excluded of the microbiota analysis because of deviations in stool collection and processing or concomitant treatments affecting bacteria. 16S rRNA V3-V4 regions from feces were sequenced using MiSeq System (Illumina). The taxonomic classification was obtained using an in-house bioinformatic pipeline. Relative abundances (%), diversity (Shannon index), and richness (rarefied Chao index) were calculated using an in-house analysis platform. Statistical analyses including demographic and clinical data were conducted using non-parametric tests with GraphPad-Prism®.
Results: The median age was 53.5 years of age [27-78]. Based on BMI, patients were at 2% underweight (BMI<18.5kg/m²), 52% normal (18.5≤BMI<24.9), 22% overweight (25<BMI<29.9), and 24% obese (BMI≥30). 59% of patients were in menopause. The clinical stage distribution was: 3% 0, 10% I, 48% IIA, 21% IIB, 14% IIIA, 0% IIIB, and 3% IIIC. In all patients, there was a significant moderate positive correlation between diversity and richness (r²=0.661; p<0.0001). Neither of these 2 markers was affected by BMI or menopause, but they increased between early (I, IIA, IIB) and advanced (IIIA, IIIB, IIIC) disease stages and varied according to molecular subtypes. The relative abundance of bacterial phyla was similar among patients, with a predominance of Bacteroidetes (38%) and Firmicutes (52%). The Tenericutes and Lentisphaerae phyla tended to be higher in patients with a high BMI (p=0.060 and p=0.091 respectively). In addition, Eubacteriaceae family was significantly higher in normal vs obese patients (p=0.005) and conversely Pseudomonadaceae family was significantly lower in normal vs obese patients (p=0.016). Menopaused women tended to have a higher level of Proteobacteria phylum than premenopausal patients (p=0.076). Moreover, Micrococcaceae family tended to be higher in postmenopausal women (p=0.060). According to molecular subtypes, Synergistetes (p=0.0533) and Verrucomicrobia (p=0.072) phyla tended to be higher in triple negative vs Luminal A cancers. In IIB, the proportion of Acidaminococcaceae (p=0.020) and Pseudomonadaceae (p=0.079) were significantly higher than in to IIA. Conversely, Defluvittalaceae (p=0.012), Enterococcaceae (p=0.045), and Propionibacteriaceae (p=0.034) were significantly lower in IIB vs IIA.
Conclusions: In this cohort, fecal microbiota profiles varied with some clinical parameters. The potential interactions between microbiota and diet are under investigation.
Keywords: microbiota, IMODI, breast cancer, targeted metagenomics sequencing.
