Abstract
BACKGROUND: RBX2660 is an investigational microbiota restoration therapy in phase 3 clinical development for preventing recurrent Clostridioides difficile infections (CDIs). In a randomized, double-blinded placebo-controlled phase 2B trial, RBX2660 was effective at preventing CDI recurrence. The current study was performed to characterize the fecal bacterial microbiome before and after treatment among RBX2660- or placebo-treated responders in that trial. METHODS: Samples were sequenced using 16S methods, and the resulting relative abundance data were fit to a Dirichlet-multinomial distribution to determine group mean relative taxonomic abundance and overdispersion at the class level. Alpha diversity was determined for all samples. Biostatistical tools, including effect size and repeated-measures analysis, were applied to evaluate the statistical significance of observed changes. RESULTS: At study entry, subjects' microbiomes were dominated by Gammaproteobacteria and Bacilli, with low abundance of Bacteroidia and Clostridia. After treatment, Bacteroidia, Clostridia, and alpha diversity increased among RBX2660 responders, concomitant with a decrease of Gammaproteobacteria and Bacilli. The resulting compositions differed significantly from baseline compositions, and the changes among RBX2660 responders differed significantly from those in placebo responders, in whom Bacteroidia or Gammaproteobacteria abundance did not change as much. Repeated-measures analyses indicated more rapid and extensive microbiome remodeling among RBX2660 responders compared with placebo responders, and effect size analyses revealed that RBX2660 responders' microbiomes became more similar to the RBX2660 composition, also compared with placebo responders. CONCLUSIONS: Prevention of recurrent CDI with RBX2660 was associated with restorative microbiome changes that may help resist C. difficile colonization and recurrence. RBX2660 was more effective than placebo at restoring participant microbiomes.