Third-party fecal microbiota transplantation for high-risk treatment-naïve acute GVHD of the lower GI tract

Abstract

Disruption of the intestinal microbiome is observed with acute graft-versus-host disease (GVHD) of the lower gastrointestinal (LGI) tract, and fecal microbiota transplantation (FMT) has successfully cured steroid-refractory cases. In this open-label, single-arm, pilot study, third-party, single-donor FMT was administered in combination with systemic corticosteroids to participants with high-risk acute LGI GVHD, with a focus on treatment-naïve cases. Participants were scheduled to receive 1 induction dose (15 capsules per day for 2 consecutive days), followed by 3 weekly maintenance doses, consisting of 15 capsules per dose. The primary end point of the study was feasibility, which would be achieved if ≥80% of participants able to swallow ≥40 of the 75 scheduled capsules. Ten participants (9 treatment-naïve; 1 steroid-refractory) were enrolled and treated. The study met the primary end point, with 9 of 10 participants completing all eligible doses. Organ-specific LGI complete response rate at day 28 was 70%. Initial clinical response was observed within 1 week for all responders, and clinical responses were durable without recurrent LGI GVHD in complete responders. Exploratory analyses suggest that alpha diversity increased after FMT. Although recipient microbiome composition never achieved a high degree of donor similarity, expansion of donor-derived species and increases in tryptophan metabolites and short-chain fatty acids were observed within the first 7 days after FMT. Investigation into the use of microbiome-targeted interventions earlier in the treatment paradigm for acute LGI GVHD is warranted. This trial was registered at www.ClinicalTrials.gov as #NCT04139577.

Graphical abstract

Figure 1.

Clinical outcomes after FMT in the treatment of high-risk acute LGI GVHD. Swimmers plot demonstrating clinical responses and duration of response after the administration of FMT in combination with corticosteroids for acute LGI GVHD. NR, no response; SR, steroid-refractory.

Figure 2.

Relationship between alpha diversity and urinary 3-IS. (A) Shannon alpha diversity metric over the first 28 days from FMT for participants with CR (CR, green), participants with no response (NR, orange), and particpants with PR (PR, blue) based on the day 28 GI response. Repeated measure correlation coefficient and P value of the correlation are labeled. Solid lines represent individual participants (numbered 1-10) over days, and participant number labels indicate the last available data point for that participant. (B) Correlation between alpha diversity metric and urine 3-IS. Repeated measure correlation coefficient and P value of the correlation are labeled. Repeated measures for the patients are indicated by the label over the points.

Figure 3.

Compositional and differential changes over time. (A) Principal component ordinate analysis at baseline (day 1), day 7, day 14, and day 28 by donor (blue points) and recipient (red points). Blue square shape represents the median of the Euclidean distances for donor, and the red square shape represents median of the Euclidean distances for recipients. The red circle and blue circle showing 95% confidence interval based on the t-distribution of the Euclidean distance for donor and recipient respectively. R² and P values derived using Adonis2 test are annotated for each day. (B) Bray-Curtis distances over the first 28 days from FMT with respect to average taxonomic composition of the 4 donor samples. Each line indicates individual participant’s beta diversity distance with respect to donor. Participant number labels indicate the last available data point for that participant. (C) Species shown to be significantly differentially abundant between baseline and day 7 or day 14 or day 28 with AncomBC analysis. Circles at day and species indicate the log fold change. Dotted line separates the days before and after FMT. Double dotted line with vertical gap separates the origins of species based on the fecal microbiome survey at baseline for recipient and donors. All species have adjusted P values <.05, minimum prevalence of 50% across all samples, and minimum absolute log fold change >1.

Figure 4.

Quantifying the change in abundance of bacteria classified by source. (A) Percent of donor-only, recipient-only, and donor–recipient-shared species out of total bacterial abundance, based on donor and recipient microbiome profile at baseline. Lines are representative of individual participants, colored-coded according LGI response at day 28; CR (green), PR (blue), or NR (orange). (B) Top 5 influential species on beta diversity by nonparametric mean Bray-Curtis distance change within each sample at baseline and after FMT at day 7, day 14, and day 28. Bray-Curtis distance change was calculated by removing 1 species at a time and measuring the change in Bray-Curtis distance for each sample and adjusted with Shannon alphadiversity for that sample. Green dots show the species identified as donor exclusive, and blue dots show the shared species among donor and patients based on the first sample from donor and recipients. (C) Percent abundance of the top 5 species selected based on the nonparametric mean Bray-Curtis distance change at baseline and after FMT at day 7, day 14, and day 28.

Figure 5.

Significant metabolomics changes in the first week after FMT. Metabolomics changes in (A) carbohydrates, (B) tryptophan metabolites, and (C) short-chain fatty acids between baseline and day 7 after FMT administration among matched CR and non-CR participants, indicated by green and orange lines, respectively. Black bar indicates median values for the metabolite within a group. All unadjusted P values were derived using Wilcox signed rank test.