Methods: The prototype MHI is based on the associations of Bacteroidia and Clostridia with colonization resistance, and Gammaproteobacteria and Bacilli with dysbiosis, and Receiver Operating Characteristic analysis of pooled RBX2660 trial data indicated that rCDI participants before treatment (baseline) are distinguished from the healthier RBX2660 profile with an odds ratio of 121 (AUC=0.99, sensitivity=0.96, specificity=0.99, cutpoint=8.2). MHI data for the published FMT cohort were calculated using publicly available data derived from pre- and post-treatment fecal samples (Khanna S, et al. Microbiome 2017 5:55), and this study included patients with or without a co-diagnosis of inflammatory bowel disease (IBD).
Results: At baseline, 92% of patients in the FMT cohort were below the MHI=8.2 cutpoint, consistent with a rCDI diagnosis. Among FMT responders 7 days after treatment, 91% of patients had shifted to MHI>8.2, (p<.0001 compared to baseline). Likewise, a significant shift was observed from baseline to 30 days (p<.0001), with 83% having MHI>8.2. There were insufficient patients to support a statistical comparison of IBD vs no IBD, but MHIs trended lower at all time points among patients with IBD.
Conclusion: MHI parameters derived from RBX2660 trials were predictive of pre- and post-treatment states for a published cohort of FMT-treated rCDI patients, suggesting that this prototype MHI represents a useful dysbiosis measure beyond RBX2660 trials. Lower MHI among patients co-diagnosed with IBD suggests the potential utility of MHI beyond rCDI. Collectively our results continue to support the utility of MHI and its prospective evaluation in ongoing Phase 3 clinical trials.
E. Deych, Rebiotix, Inc.: Research Contractor , Consulting fee .
B. Shannon, Rebiotix, Inc.: Research Contractor , Consulting fee .