Long-term Outcomes of Fecal Microbiota Transplantation in Patients With Cirrhosis

Recurrent hepatic encephalopathy (HE) leads to significant disability and mortality despite standard of care (SOC), that is, lactulose and add-on rifaximin.1Vilstrup H. et al.Hepatology. 2014; 60: 715-735Crossref PubMed Scopus (1107) Google Scholar It is associated with impairment of the gut–liver–brain axis, and fecal microbiota transplantation (FMT) may be useful. In recurrent Clostridium difficile, a single FMT has sustained cure rates of greater than 80% with a tolerable long-term safety profile.2Kelly C.R. et al.Gastroenterology. 2015; 149: 223-237Abstract Full Text Full Text PDF PubMed Scopus (397) Google Scholar In a published safety trial, 10 recurrent HE participants each were randomly assigned to receive pretreatment antibiotics and FMT or to receive SOC, which was safe. As a secondary outcome, there was an improvement in short-term cognitive function and hospitalizations.3Bajaj J.S. et al.Hepatology. 2017; Google Scholar However, the long-term impact of FMT is unclear. Our aim was to determine the long-term impact of FMT on cognition, hospitalizations, and HE by extending the results of this trial.3Bajaj J.S. et al.Hepatology. 2017; Google Scholar For rational donor selection, the microbiome profiles of HE patients and healthy control individuals were assessed. HE patients showed a relative reduction of Lachnospiraceae and Ruminococcaceae species compared with control individuals. Leveraging the cross-sectional microbiome data, a donor was selected from the donor database by using random forest analysis, to complement the relative deficiencies of the patient microbiota.3Bajaj J.S. et al.Hepatology. 2017; Google Scholar The same donor was used for all FMT-assigned participants. These participants underwent 5 days of pre-FMT antibiotics, after which 90 mL (27 g of stool) of the donor material containing approximately 2.7 × 1012 colony-forming units was administered via enema. SOC participants did not undergo any interventions or antibiotic therapy. All participants were followed up in keeping with the visits reported in Figure 1A. There were no FMT-related serious adverse events in the 5 months after the study, as judged by the data safety monitoring board. For the long-term study, all participants were followed up prospectively at least every 2 months through chart review, phone calls, or in-person interview with specific focus on hospitalizations and HE episodes. Eligibility (Supplementary Table 1), cognitive testing, further donor selection and methods are in the Supplementary Methods section. At the long-term visit, we readministered cognitive tests (Psychometric Hepatic Encephalopathy Score [PHES] and EncephalApp Stroop) and obtained stool (Figure 1A).3Bajaj J.S. et al.Hepatology. 2017; Google Scholar Data were compared with pre-FMT, postantibiotics, and early post-FMT values. Microbial composition was studied by using 16S ribosomal RNA sequencing. Principal component and linear discriminant analysis effect size (LEFSe) analyses were performed. All participants were receiving lactulose, rifaximin, and proton pump inhibitors (PPIs) at trial initiation and continued taking these throughout. All participants were followed up for >12 months and up to 15 months (mean ± standard deviation [SD], 12.9 ± 2.9 months for FMT and 12.8 ± 3.7 months for SOC), but 3 were excluded: 2 from the SOC group (1 died, 1 liver transplant) and 1 from the FMT group who died (P = 1.0 between groups), with these events occurring a mean ± SD of 11.3 ± 2.9 months after randomization. Overall, in those assigned to FMT, the intervention was well tolerated and had a nonconcerning long-term safety profile, without infections or the need for new antibiotic initiation. There were significantly more hospitalizations and HE episodes in the SOC arm compared with the FMT arm during the long-term follow-up (Figure 1B and C). In the SOC arm, there were 10 hospitalizations (2 participants had 2 events each, and 6 participants had 1 event each) compared with 1 in the FMT arm (P = 0.05). The majority of the hospitalizations were liver related (SOC group: HE, n = 4, infection, n = 2, ascites, n = 2 vs FMT group: ascites = 1). The 2 SOC participants with infections (1 C. difficile and 1 pneumonia) required short courses of broad-spectrum antibiotics. In the SOC arm, there were 8 HE events (2 participants had 2 events each, and 4 participants had 1 event each) compared with 0 in the FMT arm (P = 0.03). In both arms, the Model for End-Stage Liver Disease (ie, MELD) score changes between preintervention and long-term postintervention for the surviving patients were similar (mean ± SD: SOC, 2.78 ± 4.7 vs FMT, 2.8 ± 4.5; P = 0.9). Cognitive function, which had improved in the FMT arm at day 20 after FMT, remained significantly better in the FMT arm compared with the SOC arm on both tests on long-term testing (Figure 1D and E). Stool samples were obtained from 7 in the FMT arm and 6 in the SOC arm at the long-term visit at the same time as cognitive testing. The remainder of patients had died (n = 2), had received a transplant (n = 1), or refused to provide samples (n = 4). There was an increase in the relative abundance of Burkholderiaceae species and decreased Acidaminococcaceae in the FMT arm long term (Figure 1F) on LEFSe. Lachnospiraceae and Ruminococcaceae species relative abundance remained similar between groups. On principal component analysis, there was a clustering of the long-term microbiota, with microbiota obtained at days 7 and 15 after FMT compared with pre-FMT microbiota (Figure 1G), indicating that the microbiota were similar after FMT regardless of short- or long-term follow-up, compared with pre-FMT microbial composition. The current study extends the experience of the first randomized clinical trial of FMT after antibiotics in cirrhosis and recurrent HE over more than 12 months. The results suggest long-term safety and sustained improvement in clinical and cognitive function parameters among patients who received FMT with pretreatment antibiotics.4Ji S.K. et al.Front Microbiol. 2017; 8: 1208Crossref PubMed Scopus (52) Google Scholar The most striking improvement was in the prevention of HE recurrence and liver-related hospitalizations in this recalcitrant population, which was associated with improved cognition compared with those receiving SOC. This occurred despite the continuation of lactulose, rifaximin, and PPIs across groups. Prior studies have shown cognitive impairment as a predictor for future HE episodes, and the alteration of the gut–liver–brain axis with this intervention could interrupt this cycle of readmissions.5Patidar K.R. et al.Am J Gastroenterol. 2014; 109: 1757-1763Crossref PubMed Scopus (127) Google Scholar The improvement in all-cause hospitalizations reflects prior data in C. difficile and alcoholic hepatitis.2Kelly C.R. et al.Gastroenterology. 2015; 149: 223-237Abstract Full Text Full Text PDF PubMed Scopus (397) Google Scholar, 6Philips C.A. et al.Clin Gastroenterol Hepatol. 2017; 15: 600-602Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar From a microbiome perspective, there were differences in profiles between the SOC and FMT arms at long-term analysis but not in Lachnospiraceae and Ruminococcaceae species relative abundance. These results are unlikely due to mitigation of the PPI microbiome effect, and they extend prior FMT studies showing that the ultimate microbial profile does not need to precisely reflect the donor to result in benefit.7Khanna S. et al.Microbiome. 2017; 5: 55Crossref PubMed Google Scholar, 8Kao D. et al.Hepatology. 2016; 63: 339-340Crossref PubMed Scopus (95) Google Scholar Functional evaluation is therefore needed to understand the potential mechanisms of action of microbiome therapeutics. Our study is limited by a small sample size and the use of pretreatment antibiotics without an FMT-alone arm; however, data suggest that antibiotics promote FMT engraftment.4Ji S.K. et al.Front Microbiol. 2017; 8: 1208Crossref PubMed Scopus (52) Google Scholar Overall, this preliminary experience shows that FMT after antibiotics may be safe and potentially effective in preventing long-term recurrence of HE. However, larger controlled trials with HE as the primary outcome are needed. Author contributions: Jasmohan Singh Bajaj conceptualized the study and was involved in all stages. Andrew Fagan and Edith A. Gavis were involved in research coordination and study conduct. Zain Kasam was involved in provision of donor material and study design. Masoumeh Sikaroodi and Patrick M. Gillevet were involved in microbiota analysis and bioinformatics. All authors contributed to the drafting of the manuscript. To select a donor, data from 174 16s rRNA sequences generated from samples collected in prior studies from the study centers were analyzed. These sequencing data were from HE patients and healthy control individuals. These data were used to train a random forest classifier, a machine learning classification technique used to distinguish HE patients from healthy individuals among the OpenBiome donors. The resulting classifier had an area under the curve of 0.94, which then gave each donor a classification score (formally, the percentage of trees in the random forest classifier that classify the sample as healthy). To corroborate these, the relative abundances of Lachnospiraceae and Ruminococcaceae bacterial families in each donor (taxa found to be depleted in HE patients) were calculated. All of these data were combined to rank the 28 stool donors in OpenBiome. Ultimately, the donor with the highest aggregate ranking was selected as the FMT donor for patients randomly assigned to FMT. This donor was a healthy 37-year-old man whose material had been used for at least 280 patients for the treatment of recurrent C. difficile without serious adverse events. This donor also had the highest relative abundances of Lachnospiraceae and Ruminococcaceae species among the potential donors. The same stool sample from this donor was used for each of the aliquots for the FMT-assigned groups. The PHES score is used to evaluate psychomotor speed, reaction time, and visuomotor coordination. It consists of 5 tests: number connection test A, number connection test B, digit symbol test, serial dotting test, and line tracing test (which has 2 components: time and errors). Population control norms are available for this battery, and the added score of each standard deviation beyond norms is the total PHES score. A low total score indicates poor cognitive performance. The EncephalApp Stroop uses a validated smartphone app version of the Stroop test that has On and Off stages. The Off stage tests psychomotor speed and accuracy, and the On stage also tests for cognitive flexibility. The easier Off stage requires participants to correctly identify the color of the pound signs presented, and the more difficult On stage requires them to correctly identify the color of a discordant word presented. The App has 2 practice runs and requires 5 correct runs in the Off and On stages. The total time required for 5 correct On and 5 correct Off stage runs is the “OffTime + OnTime,” which is of relevance in HE, and is recorded. A higher time required to complete the number of runs required indicates worse performance. The study was a randomized, open-label, phase 1 study with exploratory endpoints and pathophysiological evaluation of the FMT. Two groups of outpatients with cirrhosis (Supplementary Table 1) were randomly assigned with a random sequence generator into SOC and FMT groups. The FMT group received zFMP-30 × 3 at the same time (27 g of stool). The intervention was delivered by enema. Once patients were randomly assigned 1:1 into group 1 (FMT) and group 2 (SOC), both were followed over 35 days, and a 5 month follow-up phone call was included to assess serious adverse events. An optional long-term extension was performed as well for at least 12 months after FMT administration. We obtained written informed consent. After the patients consented and were found to be eligible, we performed a detailed medical history and physical examination at day 0. We then collected blood, urine, and stool for baseline testing, including for C. difficile, at day 0 using quantitative polymerase chain reaction. We excluded patients for whom the C. difficile test result was positive. We performed urine pregnancy tests in women of childbearing age, they were allowed to proceed only if the test result was negative and they agree to use effective contraception for the duration of the study and for 10 days before and 30 days after the study. We also performed cognitive testing. We prescribed an antibiotic regimen for patients (ciprofloxacin, 500 mg orally 2 times per day, amoxicillin 500 mg orally 3 times per day, and metronidazole 500 mg orally 3 times per day) for 5 days to set up a clear baseline. The doses were adjusted for end-stage liver disease (for metronidazole) and for creatinine clearance (for all 3 drugs). Drawing from ecological principles of microbial niche environments and data from recurrent C. difficile, pretreatment antibiotics are likely to increase the probability of disrupting the host’s intrinsic microbiota and creating an opportunity for a “healthy” microbiota from the FMT to engraft. At day 5, we reevaluated patients with a directed interval history and focused physical examination as needed. After ensuring that they were still candidates according to the inclusion/exclusion criteria, we again collected stool, urine and blood for pre-FMT evaluation and urine for pregnancy tests from eligible women. Cognitive testing was performed again. As per OpenBiome guidelines:1.Frozen material was thawed for 4 hours at room temperature. After thawing, material could remain at room temperature for up to 4 hours or be kept refrigerated/on ice for up to 8 hours.2.The standard protocol for handling biohazardous material was be used to avoid contamination and risk to health care handlers. At that point, we will delivered 90 mL of the fecal material, using universal precautions. The procedure was completed by trained study personnel. We ensured that patients were able to hold the enema for at least 30 minutes by positioning patients in the left lateral decubitus position. We was each patient in the clinic the day after FMT (day 6, or FMT + 1), day 12 (FMT + 7), day 20 (FMT + 15), and day 35 (FMT + 30), at which point we took a detailed history regarding abdominal symptoms, evaluation of infectious complications, hospitalizations, or complications of cirrhosis. Visits on days 6, 12, and 35 were purely safety associated, and on the day 20 visit, we repeated the pathophysiological studies. To ensure that we had enough samples in case patients were not able to return on day 20, we also collected all samples at day 12, but we analyzed them only in case the day 20 visit did not occur. These visits, apart from the visit after FMT, were ±2 days for patient convenience. At 5 months after FMT, participants were be followed up with a phone call to evaluate potential serious adverse events, new onset of transmitted infections, new onset or significant worsening of chronic medical conditions, or suspected unexpected serious adverse reactions that occurred between 35 days and 5 months, for reporting purposes. The SOC group had the same sample collections, follow-up, and cognitive testing as the FMT group, but without the 5-day antibiotic therapy or the FMT. In addition, we also did not perform the sample collection that was done after 5 days of antibiotics in this group because no reasonable change in microbiota were expected over 5 days without antibiotics. The follow-up of this group was same as that of the FMT group. The primary endpoints were safety and tolerability, defined by the rate of development of FMT-related serious adverse events and withdrawal from the study in the FMT vs SOC groups. The secondary endpoints were 1) microbiota composition and functional change and 2) cognitive function changes. Specific issues to be captured were as follows:1.Antibiotic-related adverse events/serious adverse events2.FMT procedure-related adverse events3.FMT material–related adverse events (eg, transmissible infection, allergic reaction)4.Short-term safety: Both solicited and unsolicited adverse events were recorded by clinical assessment at FMT, (day 6 or FMT + 1), day 12 (FMT + 7), day 20 (FMT + 15), and day 35 (FMT + 30). Participants had the ability to telephone the study team at any point during the study.5.Long-term safety: phone call at 5 months to evaluate for serious adverse events6.Long-term extension (see next section) We aimed to follow all participants for at least 12 months after enrollment in both groups after their initial study follow-up was completed. During this period, patients were called or seen in person at least every 2 months, their clinical course was monitored through active chart review, and after at least 12 months, they had stool collection, cognitive testing with PHES and EncephalApp, and an in-person evaluation for adverse events and serious adverse events as defined in the original protocol. Each intervening hospitalization was evaluated and recorded for causes, changes in medications, and outcomes. Specifically, infections were also recorded. Development of HE, defined as grade 2 or higher of West–Haven criteria that required admission, emergency department visits, or changes to therapy, were also recorded in detail. Patients had the option to agree to this, to participate in the follow-up via other protocols or ongoing studies, or not to participate at all. They also had the option to provide stool samples or not as they chose. We stopped follow-up after death or liver transplantation or after at least 12 months, based on the availability of the patient. Supplementary Table 1Detailed Eligibility CriteriaInclusion criteria:1.Cirrhosis diagnosed by any 1 of the following in a patient with chronic liver disease:a.Liver biopsyb.Radiologic evidence of varices, cirrhosis, or portal hypertensionc.Laboratory evidence of platelet count <100,000 or AST/ALT ratio > 1d.Endoscopic evidence of varices or portal gastropathy2.At least 2 episodes of HE, 1 within the last year but not within the last month (patient can be taking lactulose and rifaximin)3.Age between 21 and 75 years4.Able to give written informed consent (demonstrated by Mini-Mental Status Examination score > 25 at the time of consent)Exclusion criteria:1.MELD score > 172.White blood cell count < 1000 cells/mm33.Platelet count < 50,000/mm34.TIPS in place for < 1 month5.No HE episode within a month before the study6.Allergic to ciprofloxacin, penicillins, or metronidazole7.Currently taking absorbable antibiotics8.Infection at the time of the FMT (diagnosed by positive blood culture result, urinalysis, and paracentesis as needed)9.Hospitalization for any nonelective cause within the last 1 month10.Patients who are >75 years old11.Patients who are pregnant or nursing (will be checked with a urine pregnancy test)12.Patients who are incarcerated13.Patients who are incapable of giving their own informed consent14.Patients who are immunocompromised because of the following reasons:a.Human immunodeficiency virus infection (any CD4 count)b.Inherited/primary immune disordersc.Current or recent (<3 mo) treatment with antineoplastic agentd.Current or recent (<3 mo) treatment with any immunosuppressant medications (including but not limited to monoclonal antibodies to B and T cells, anti-TNF agents, glucocorticoids, antimetabolites [azathioprine, 6-mercaptopurine], calcineurin inhibitors [tacrolimus, cyclosporine], and mycophenolate mofetil). Patients who are otherwise immunocompetent and have discontinued any immunosuppressant medications 3 or more months before enrollment may be eligible to enroll.15.Patients with a history of severe (anaphylactic) food allergy16.Patients who have previously had FMT17.Patients receiving renal replacement therapy18.Patients who are unwilling or unable to hold the enemas19.Patients with untreated in situ colorectal cancer20.Patients with a history of chronic intrinsic gastrointestinal diseases such as inflammatory bowel disease (ulcerative colitis, Crohn’s disease, or microscopic colitis), eosinophilic gastroenteritis, celiac disease, or irritable bowel syndrome21.Major gastrointestinal or intra-abdominal surgery in the last 3 months22.Unable to comply with protocol requirements23.Patients with American Society of Anesthesiologists physical status classifications IV and V24.Patients with acute illness or fever on the day of planned FMT will be excluded, with the option of inclusion at a future date25.Any conditions for which, in opinion of the physician, the treatment may pose a health risk26.C. difficile in the stool at baseline (qPCR)27.Grade 2–4 or complicated hemorrhoidsALT, alanine aminotransferase; AST, aspartate aminotransferase; FMT, fecal microbial transplant; MELD, model for end-stage liver disease; qPCR, quantitative polymerase chain reaction; TIPS, transjugular intra-hepatic portosytemic shunt; TNF, tumor necrosis factor. Open table in a new tab ALT, alanine aminotransferase; AST, aspartate aminotransferase; FMT, fecal microbial transplant; MELD, model for end-stage liver disease; qPCR, quantitative polymerase chain reaction; TIPS, transjugular intra-hepatic portosytemic shunt; TNF, tumor necrosis factor. The Transplantation of Fecal Microbiota for Cirrhotic PatientsGastroenterologyVol. 157Issue 3PreviewI read with interest regarding the transplantation of fecal microbiota for cirrhotic patients to prevent recurrence of hepatic encephalopathy (HE).1 Although innovative and inspiring, several points warrant further clarification. Full-Text PDF