Fig. 4. Donor engraftment explains recurrent CDI FMT clinical outcomes.
a, PEDS at 8 weeks can predict early relapse of FMT in patients (n = 13 biologically independent samples) with recurrent CDI. A two-sided Wilcoxon test was used to estimate statistical significance. b, The PEDS metric can elucidate the successful outcome of repeat FMT in patients who relapsed with recurrent CDI after the initial FMT. c, Predictive power of our approach on all available FMT samples where clinical evaluation was independently noted. Whenever we reported clinical success we found engraftment to be above the threshold of 17% (n = 19 true positives) with 1 false negative. Clinical relapse was always independently associated with low engraftment (n = 2 true negatives) with no false negatives. d, Bacterial strain engraftment and identification of highly transmissible strains that stably engraft in multiple recipients. The first four columns are weekly metagenomic samples from the donor, while the fifth column is the donor sample from five years later. The next six columns are from the FMT recipients who did not have an early relapse. The last column is from one of the recipients five years later. Strainer was used to find the presence (green) or absence (yellow) of each bacterial strain from the corresponding metagenomics sample.