Gut microbiome and intestinal inflammation in preclinical stages of rheumatoid arthritis

Background Faecal Prevotellaceae, and other microbes, have been associated with rheumatoid arthritis (RA) and preclinical RA. We have performed a quantitative microbiome profiling study in preclinical stages of RA.Methods First-degree relatives of patients with RA (RA-FDRs) from the SCREEN-RA cohort were categorised into four groups: controls, healthy asymptomatic RA-FDRs; high genetic risk, asymptomatic RA-FDRs with two copies of the shared epitope; autoimmunity, asymptomatic RA-FDRs with RA-associated autoimmunity; and symptomatic, clinically suspect arthralgias or untreated new-onset RA. Faecal samples were collected and frozen. 16S sequencing was performed, processed with DADA2 pipeline and Silva database. Cell counts (cytometry) and faecal calprotectin (enzyme-linked immunosorbent assay, ELISA) were also obtained. Microbial community analyses were conducted using non-parametric tests, such as permutational multivariate analysis of variance (PERMANOVA), Wilcoxon and Kruskal-Wallis, or Aldex2.Methods First-degree relatives of patients with RA (RA-FDRs) from the SCREEN-RA cohort were categorised into four groups: controls, healthy asymptomatic RA-FDRs; high genetic risk, asymptomatic RA-FDRs with two copies of the shared epitope; autoimmunity, asymptomatic RA-FDRs with RA-associated autoimmunity; and symptomatic, clinically suspect arthralgias or untreated new-onset RA. Faecal samples were collected and frozen. 16S sequencing was performed, processed with DADA2 pipeline and Silva database. Cell counts (cytometry) and faecal calprotectin (enzyme-linked immunosorbent assay, ELISA) were also obtained. Microbial community analyses were conducted using non-parametric tests, such as permutational multivariate analysis of variance (PERMANOVA), Wilcoxon and Kruskal-Wallis, or Aldex2.Results A total of 371 individuals were included and categorised according to their preclinical stage of the disease. Groups had similar age, gender and body mass index. We found no significant differences in the quantitative microbiome profiles by preclinical stages (PERMANOVA, R2=0.00798, p=0.56) and, in particular, no group differences in Prevotellaceae abundance. Results were similar when using relative microbiome profiling data (PERMANOVA, R2=0.0073, p=0.83) or Aldex2 on 16S sequence counts. Regarding faecal calprotectin, we found no differences between groups (p=0.3).Conclusions We could not identify microbiome profiles associated with preclinical stages of RA. Only in a subgroup of individuals with the most pronounced phenotypes did we modestly retrieve the previously reported associations.