Methods
Patients and Identification of Donors
FMT was offered as a therapeutic option in a compassionate use setting to patients of our outpatient clinic with moderately to severely active chronic UC (Mayo scores ≥6; see ref. and who were refractory and/or intolerant to standard IBD therapies including steroids, cyclosporine, thiopurines, and/or biologics (definitions of prior clinical treatment failure are specified in Supplementary Methods online), but who were not at imminent risk of colectomy. Diagnosis of UC was based on clinical, radiological, endoscopic, and histopathological criteria as recommended by the European Crohn's and Colitis Organisation (ECCO). Treatment with infliximab, cyclosporine, thiopurine, methotrexate, and/or steroids was stopped at least 8, 4, 4, 4, and 2 weeks before FMT, respectively. Concomitant therapy with 5-aminosalicylates was allowed.
Patients were able to suggest a possible donor, but first-degree relatives of the patients and hospital or health-care workers were excluded. A complete medical and surgical history of each potential donor was obtained. Donors were considered to be suitable if they had no history of chronic gastrointestinal disease and if they had not been hospitalized for at least 3 months before FMT or received antibiotics or proton pump inhibitors for at least 6 months before FMT. Blood collection including complete blood count, chemistry, and iron profile was performed before FMT. Donor blood was negative for common viruses (hepatitis A, B, and C, HIV-1 and HIV-2, cytomegalovirus, Epstein–Barr, Herpes simplex, and Varicella zoster) and Treponema pallidum. Donor feces were negative for common enteric pathogens (Yersinia spp., Salmonella spp., Shigella spp., Campylobacter jejuni, C. difficile toxin, helminths, ova, parasites, and Helicobacter pylori), as determined with standard screening methods.
Patients and donors of fecal material were informed of the potential risks and benefits of the fecal transplantation and its experimental nature. All patients and donors gave their written informed consent to the protocol. The ethics committee of the Medical University of Vienna approved the retrospective review of patients' file.
Patient Preparation
Before FMT, patient blood was collected for chemical and biological analyses and feces were analyzed to ensure that common enteric pathogens were absent (Supplementary Table S1 online). Patients received antibiotic treatment (metronidazole, 500 mg twice a day) for 5–10 days until the evening before they were admitted to the hospital for FMT (except for the first FMT of patient 1). On the day of admittance, a nasojejunal tube was placed in the patient's jejunum using sedation with midazolam and propofol. The presence of cytomegalovirus colitis was excluded by concomitant diagnostic sigmoidoscopy and by PCR and immunohistochemistry of retrieved biopsy samples. In the afternoon of the same day, the position of the nasojejunal tube was verified by abdominal radiography. Polyethylene glycol (2 l) was subsequently administered via the nasojejunal tube for bowel lavage. Proton pump inhibition therapy (pantoprazole 20 mg), begun the evening before FMT, was applied twice a day during the transplantation procedure. Patients did not eat for 6 h before FMT.
Preparation of Fecal Suspension
Because protocols for donor stool preparation differ between published FMT studies, three preparation protocols were initially assessed (Supplementary Methods). A variation of one of these protocols (protocol B) was finally used as described below. Volunteers who passed the selection criteria donated stool samples on the day of FMT in the laboratories of the Medical University of Vienna (Department of Internal Medicine III, Division of Gastroenterology and Hepatology, Medical University Vienna, Wien, Austria). Fresh stool (60 g) was immediately mixed in a standard household blender with 250 ml of 0.9% sterile saline for several seconds until it developed a smooth consistency. The stool suspension was then filtered through a gauze pad to remove large particles. Of the filtered suspension, 100 ml was filled into an enema bag. The rest of the suspension (~150 ml) was diluted in 0.9% sterile saline to a volume of 400–425 ml. Afterwards, it was filtered several times through an increasing number of gauze pads (4 to >6 to >8) to remove small particles that could clog the nasojejunal tube. This stool suspension was poured into a sterile bottle and administered within 1 h.
Route of Administration
Patients received FMT on three consecutive days (except for FMT1 of patient 1, Figure 1) via both nasojejunal tube and enema. Altogether, a median 24 g (17–25 g in median 250 ml) and 20 g (6–22 g in median 100 ml) stool suspension was administered via nasojejunal tube and enema, respectively. The nasojejunal tube was flushed with 20 ml saline solution after each procedure. Patients received two tablets of loperamide to diminish intestinal motility immediately after FMT and 6 h after procedure. After FMT, patients were encouraged to consume fluid food including a high-fiber diet (Nestlé Nutrition Resource 2.0 Faser, Florham Park, NJ). Blood samples were collected every day. One day after the last FMT, patients were discharged from hospital.
(Enlarge Image)
Figure 1.
Schematic overview of fecal microbiota transplantation (FMT) for ulcerative colitis (UC) patients. Probiotics (in blue) and/or antibiotics (in red) were given for either 5 or 10 days before FMT. The pro/antibiotics given are as follows: metronidazole (Met.; 500 mg twice a day), Saccharomyces boulardii (Sb; Yomogi), and =Omnibiotic 6 (O6). FMT procedures are shown with an arrow and the donor is indicated above the arrow. Patient 1 (P1) received two FMTs with two different donors, patients 2 and 3 shared a donor, and patients 4 and 5 shared a donor. Red circles indicate sampling dates with microbiota data and green circles indicate dates with clinical data but no microbiota data. Patient 5 provided a stool sample at week 6 instead of the prespecified week 8. BL, baseline before FMT, which is a post-antibiotics baseline for all FMTs except for FMT 1 of P1.
Clinical Diagnostics
Blood and stool were collected and analyzed as outlined in Supplementary Table S1 online. Clinical activity was assessed at each visit according to the (partial) Mayo score. Furthermore, at weeks 4 and 12, sigmoidoscopy and the H2-glucose breath test to assess microbial overgrowth were scheduled. Fecal calprotectin was determined by enzyme-linked immunosorbent assay according to the manufacturer's instructions (Bühlmann Laboratories AG, Schönenbuch, Switzerland).
Efficacy End Points
Remission was defined as a total Mayo score of ≤2 with no individual subscore exceeding one point for UC patients, and response as a decrease from baseline in the total Mayo score of at least three points and at least 30%, with an accompanying decrease in the subscore for rectal bleeding of at least one point or an absolute subscore for rectal bleeding of 0–1 for UC patients.
Sampling of Stool Samples and Bacterial Community Analyses
Stool samples of the patients were collected after antibiotic treatment and bowel lavage (post-antibiotics baseline), but before the first FMT, and on several time points after FMT (Figure 1). Donor stool samples, including both native stool and stool suspension for FMT, were collected. Stool samples were either immediately frozen at −80 °C or after transportation in cool bags (for weeks 1, 2, 4, 8, and 12).
DNA was extracted from fecal samples using a standard phenol–chloroform protocol with bead beating. Purified DNA served as the template for a two-step, low-cycle number PCR with primers targeting the 16S rRNA gene of most Bacteria (spanning the V6 to V9 regions, 909F, 5′-ACTCAAAKGAATWGACGG-3′ and 1492R, 5′-NTACCTTGTTACGACT-3′). The two-step PCR produces amplicons tagged with an 8-nt sample-specific barcode sequence, but minimizes barcoded primer-associated bias. PCR products were purified with Agencourt AMPure beads (Beckman Coulter Genomics, Danvers, MA), quantified with PicoGreen (Quant-iT PicoGreen, Invitrogen, Carlsbad, CA), and pooled for sequencing. Pyrosequencing was performed on a GS FLX (Roche, Basel, Switzerland) instrument with Titanium chemistry by Eurofins MWG Operon (Ebersburg, Germany). The two-step PCR pyrosequencing approach was previously shown to reliably monitor relative differences in abundant 16S rRNA gene phylotypes.
Sequence data were quality-filtered using the Pyronoise algorithm in mothur, clustered into species-level phylotypes of 97% sequence identity using UCLUST, and checked for chimeras with Chimera Slayer. Samples with <200 reads (n=3) were excluded from downstream analysis. Reads were taxonomically classified using the Ribosomal Database Project naive Bayesian classifier. Relative comparisons of α-diversity (Chao 1 phylotype richness) and β-diversity (weighted UniFrac distances) were performed using QIIME with re-sampling at less than the size of the smallest library (1,000 subsamples of 200 reads). Community similarity was calculated using the weighted UniFrac distance metric, which incorporates phylogenetic as well as relative abundance information. Statistical significance of shifts in the relative abundance of bacterial taxa as well as Pearson's correlation coefficients were corrected for multiple comparisons using the Holm–Bonferroni method, and adjusted P values are reported.
Pyrosequencing data are archived at National Center for Biotechnology Information (NCBI) Sequence Read Archive under Accession SRS350277.