Detecting microbial engraftment after FMT using placebo sequencing and culture enriched metagenomics to sort signals from noise

Fecal microbiota transplantation (FMT) has shown efficacy for the treatment of ulcerative colitis but with variable response between patients and trials. The mechanisms underlying FMT’s therapeutic effects remains poorly understood but is generally assumed to involve engraftment of donor microbiota into the recipient’s microbiome. Reports of microbial engraftment following FMT have been inconsistent between studies. Here, we investigate microbial engraftment in a previous randomized controlled trial (NCT01545908), in which FMT was sourced from a single donor, using amplicon-based profiling, shotgun metagenomics, and culture-enriched metagenomics. Placebo samples were included to estimate engraftment noise, and a significant level of false-positive engraftment was observed which confounds the prediction of true engraftment. We show that analyzing engraftment across multiple patients from a single donor enhances the accuracy of detection. We identified a unique set of genes engrafted in responders to FMT which supports strain displacement as the primary mechanism of engraftment in our cohort.