Timing for the second fecal microbiota transplantation to maintain the long-term benefit from the first treatment for Crohn's disease

Abstract

Increasing evidence has shown that fecal microbiota transplantation (FMT) could be a promising treatment option for Crohn's disease (CD). However, the frequency of FMT for CD treatment remains unclear. This study aimed to evaluate the optimal timing for administering the second course of FMT to maintain the long-term clinical effects from the first FMT for patients with CD. Sixty-nine patients with active CD who underwent FMT twice and benefited from the first FMT were enrolled in this study. Clinical response, stool microbiota, and urine metabolome of patients were assessed during the follow-up. The median time of maintaining clinical response to the first FMT in total 69 patients was 125 days (IQR, 82.5-225.5). The time of maintaining clinical response to the second FMT in 56 of 69 patients was 176.5 days (IQR, 98.5-280). The fecal microbiota composition of each patient post the first FMT was closer to that of his/her donor. Compared to that of the baseline, patients prior to the second course of FMT showed significant differences in urinary metabolic profiles characterized by increased indoxyl sulfate, 4-hydroxyphenylacetate, creatinine, dimethylamine, glycylproline, hippurate, and trimethylamine oxide (TMAO). This study demonstrated that patients with CD could be administered the second course of FMT less than 4 months after the first FMT for maintaining the clinical benefits from the first FMT. This was supported by the host-microbial metabolism changes in patients with active CD. Trial registration: ClinicalTrials.gov , NCT01793831. Registered 18 February 2013. https://ichgcp.net/clinical-trials-registry/NCT01793831?term=NCT01793831&rank=1.

Keywords: Crohn’s disease; Fecal microbiota transplantation; Gut microbiota; Urine metabolomics.

Conflict of interest statement

Conflict of interest

Faming Zhang is the core inventor of GenFMTer and TET, and the founder of nonprofit China fmtBank. Other authors declare that they have no conflict of interest.

Ethical approval

This study was reviewed and approved by the Second Affiliated Hospital of Nanjing Medical University Institutional Review Board. All eligible subjects provided written informed consents prior to participation in this study.

Figures

Fig. 1

Flow chart of the study design. Loss of response referred to the flare after FMT-induced clinical response

Fig. 2

Harvey–Bradshaw Index (HBI) scores and clinical response maintaining time of all the patients (n = 69). a The change of HBI score after FMT. Compared with the baseline before the first FMT (pre-first FMT), the HBI score decreased significantly 4 weeks after the first FMT (4W post-first FMT) and right before the second FMT (pre-second FMT). Compared with the 4W post-first FMT HBI score, it increased significantly pre-second FMT. Significance levels: **p < 0.01, ***p < 0.001. b The comparison of patients’ clinical response maintaining time after the first and the second FMT (p > 0.05)

Fig. 3

Clinical response to the first FMT in the selected nine CD patients. a HBI score at baseline; at 3 days, 4 weeks, and 12 weeks after the initial FMT; and at the time point right before the second FMT (pre-second FMT). b Abdominal pain scores at baseline and at 3 days, 4 weeks, and 12 weeks after the initial FMT. c Frequency of defecation at baseline and at and 3 days, 4 weeks, and 12 weeks after the initial FMT

Fig. 4

Microbial composition changes in the selected nine CD patients after the first FMT. a The microbial α-diversity (OTU richness) of donors and patients (n = 9) before the first FMT (pre-first FMT), 3 days after the first FMT (3D post-first FMT), and right before the second FMT (pre-second FMT). b The microbial α-diversity (Shannon diversity index) of donors and patients (n = 9) pre-first FMT, 3D post-first FMT, and pre-second FMT. The Wilcoxon rank-sum test was used to determine the significance between donors and patients, and the Wilcoxon matched-pairs signed-rank test was used between the samples before and after FMT. *p < 0.05, ***p < 0.001. c Principal coordinate analysis (PCoA) with unweighted UniFrac distance for donor and patient samples pre- and post-FMT. The gray lines showed the trajectory from donor’s to patient’s pre-first FMT, 3D post-first FMT, to pre-second FMT microbiome sample. d1 is the distance between donor’s and patient’s pre-first FMT microbiome sample. d2 is the distance between donor’s and patient’s 3D post-first FMT microbiome sample. d3 is the distance between donor’s and patient’s pre-second FMT microbiome sample. dd1 is always less than d2, implying that the taxa composition of the patient after the first FMT is closer to that of his/her donor. ed3 is typically less than d1, implying that the taxa compositions of most of the patients (with a few exceptions) before the second FMT are still close to those of their donors

Fig. 5

Gut microbial compositions of CD patients (n = 9) and donors (n = 9) at the phylum level. For patients, we sequenced their stool samples right before the first FMT (pre-first FMT), 3 days after the first FMT (3D post-first FMT), and right before the second FMT (pre-second FMT)

Fig. 6

The relationship between microbial composition dissimilarity and clinical response maintaining time. Each dot represents a patient. For dissimilarity measures, we use a, b the root Jensen–Shannon divergence (rJSD) and c, d the unweighted UniFrac. a, c Dissimilarity between the donor’s and patient’s 3D post-first FMT microbiome samples. b, d Dissimilarity between the donor’s and patient’s pre-second FMT microbiome samples

Fig. 7

OPLS-DA score plot of 1H NMR profiles of urine samples obtained from the nine CD patients before the first FMT (pre-first FMT) and at right before the second FMT (pre-second FMT)

Fig. 8

Urinary metabolite analysis. Changes of urinary metabolites between pre-first FMT and pre-second FMT (n = 9). Significance levels: *p < 0.05, **p < 0.01