A Systematic Review and Meta-Analysis of Randomized Controlled Trials of Fecal Microbiota Transplantation for the Treatment of Inflammatory Bowel Disease

Objectives Inflammatory bowel disease (IBD) is a chronic recurrent inflammatory disease of the gastrointestinal tract, and its prevalence is increasing worldwide. Fecal microbiota transplantation (FMT) is an emerging therapy that modifies the patient's gut microbiota by transplanting feces from a healthy donor to achieve disease remission. However, its efficacy and safety need to be further investigated. Methods PubMed, the Cochrane Library, Web of Science, Embase, and Google Scholar databases (up to 8th November 2021) were searched and literature was screened by title and abstract as well as full text. The primary outcome was clinical remission, with the clinical response as a secondary outcome. Risk ratios (RR) with 95% confidence intervals (CI) were reported. Results A total of 14 trials were included in this study. In terms of clinical remission, FMT had a significant effect compared to placebo (RR = 1.44, 95 CI%: 1.03 to 2.02, I2 = 38%, P=0.03), with no significant risk of study heterogeneity. Moreover, FMT led to significant results in clinical response compared to placebo with moderate between-study heterogeneity (RR = 1.34, 95 CI%: 0.92 to 1.94, I2 = 51%, P=0.12). Subgroup analysis showed a higher clinical remission for fresh fecal FMT (40.9%) than that for frozen fecal FMT (32.2%); the efficacy of gastrointestinal (GI) pretreatment, the severity of disease, route of administration, and the donor selection remain unclear and require more extensive study. Safety analysis concluded that most adverse events were mild and self-resolving. The microbiological analysis found that the patient's gut microbiota varied in favor of the donor, with increased flora diversity and species richness. Conclusion FMT is a safe, effective, and well-tolerated therapy. Studies have found that fresh fecal microbiota transplant can increase clinical remission rates. However, more randomized controlled trials and long-term follow-ups are needed to assess its long-term effectiveness and safety.


Introduction
Inflammatory bowel disease (IBD), which includes Crohn's disease (CD) and ulcerative colitis (UC), is a chronic recurrent inflammatory disease of the gastrointestinal tract [1]. e prevalence of IBD is increasing worldwide [2,3]. e annual incidence of UC and CD in North America is similar, 0-19.2 per 100,000 and 0-20.2 per 100,000, respectively [2]. UC is characterized by diffuse mucosal inflammation, with lesions limited to the colonic mucosa and typical symptoms of diarrhea, mucus, and bleeding; CD can cause transmural inflammation that can accumulate anywhere in the gastrointestinal tract, most commonly in the terminal ileum or rectum, with abdominal pain, watery diarrhea, and weight loss as the main clinical features [1,4,5]. e etiology of IBD is uncertain, but it is now believed that genetic susceptibility, environmental exposure, gut microbiota, and the immune system are involved. Current treatment is aimed at controlling the progression of inflammation and achieving disease remission. Commonly used drugs are corticosteroids, antitumor necrosis factor-alpha (TNF-α) drugs, immunomodulators (e.g., methotrexate), biologic therapies, and surgery. IBD not only imposes a heavy financial burden and reduces the quality of life but also has a significant impact on the resources of the healthcare system [6,7].
Fecal microbiota transplantation (FMT) is a therapy in which fecal matter from a healthy donor is infused into the gastrointestinal tract of a recipient patient, changing the patient's intestinal microbiota, leading to remission [8]. It has been suggested that the development of chronic inflammatory diseases such as IBD is associated with alterations in the composition of the intestinal microbiota (known as ecological dysbiosis) [9]. e differences in gut microbiota composition between IBD patients and healthy individuals are mainly in terms of microbial diversity and flora richness [10]. Extensive data confirm FMT as an effective treatment for recurrent Clostridium difficile infection (rCDI) [11,12]. European Microbiology, the European Society for Infectious Diseases, and the American College of Gastroenterology all recommend FMT as a treatment for rCDI and have included it in their treatment guidelines [13,14]. FMT is currently considered an emerging treatment for IBD [15,16]. Studies have confirmed the efficacy of FMT in the treatment of IBD; Fang et al. [17] reported a clinical remission rate of 28.8% and a clinical response rate of 53% during follow-up in patients with IBD, concluding that fresh or frozen donor stool, route of administration, and antibiotic pretreatment or not had no effect on the efficacy of FMT for IBD. Caldeira et al. [18] showed an overall clinical remission rate of 37% and an overall clinical response rate of 54% in patients with IBD, suggesting that frozen fecal material was associated with a higher rate of clinical remission. However, several issues remain unresolved in the FMT protocol, such as type of stool (frozen or fresh), route of administration (lower or upper gastrointestinal tract), donor selection (healthy anonymous, or relative, or partners), and whether to pretreat the gastrointestinal (GI) tract (i.e., using vancomycin and/or polyethylene glycol, or antibiotic, or macrogol solution) prior to treatment with FMT. erefore, the purpose of this systematic review and meta-analysis is to synthesize data from published randomized controlled trials on FMT for the treatment of patients with IBD and to evaluate the efficacy and safety of FMT for the treatment of IBD.

Methods
is study was conducted and reported in accordance with the Preferred Reporting Items for Systematic Evaluation and Meta-Analysis (PRISMA (available (here))) guidelines.

Search Methodology.
English articles of randomized controlled trials were searched in PubMed, the Cochrane Library, Web of Science, Embase, and Google Scholar databases (up to 8th November 2021). e databases were used ("Fecal Microbiota Transplantation" OR "Microbiota Transplantation, Fecal" OR "Transplantation, Fecal Microbiota" OR "Bacteriotherapy" OR "Intestinal Microbiota Transfer") AND ("IBD" OR "inflammatory bowel disease" OR "Crohn (s) disease" OR "CD" OR "ulcerative colitis" OR "UC" OR "colitis") for the keywords combinations and wildcards were thoroughly searched to ensure the completeness of the search results.

Inclusion and Exclusion
Criteria. Studies of randomized controlled trials of patients with IBD treated with FMT were screened by a set of inclusion and exclusion criteria. Inclusion criteria were as follows: (1) patients with IBD aged > 18 years with no gender restriction; (2) randomized controlled trials; (3) studies with clearly described clinical endpoints; (4) journal articles; and (5) FMT by different modalities (i.e., nasogastric tube, nasoduodenal tube, colon/ colonoscopy, enema, or oral) was allowed. Exclusion criteria were as follows: (1) in vitro or animal testing; (2) languages other than English or Chinese; (3) patients with concomitant Clostridium difficile infection (CDI) infection or other pathogens; (4) population noncompliance, e.g., pediatric IBD; (5) duplicate literature; (6) other types of literature: reviews, editorials, case reports, abstracts, meta-analysis, observational studies, surveys, and questionnaires; (7) no clinical results reported, or studies with incomplete data.
We went through the PRISMA flowchart to select studies for inclusion. One reviewer performed the initial screening based on titles and abstracts. Titles and abstracts of potentially eligible records were independently reviewed by a second reviewer. Identified studies were then independently screened in full by two reviewers. Final inclusion was determined by both reviewers based on inclusion and exclusion criteria. e senior author discussed and resolved any discrepancies in the selection process.

Data Extraction and Quality
Assessment. Data were extracted and evaluated independently by two reviewers, recorded on self-designed forms, and any disagreements were discussed and resolved by a third reviewer. Extracted study data included (1) first author, study year, and site; (2) participant characteristics: total number of participants, number of controls, number of experiments, and disease severity; (3) type of intervention: type of stool (fresh or frozen), route of administration, relationship to donor, gastrointestinal (GI) pretreatment, IBD type, and follow-up time; (4) clinical outcomes: primary outcome (clinical remission) and secondary outcome (clinical response).
Study quality was assessed using the Cochrane Bias Tool and risks of bias included: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other sources of bias.

Statistical Analysis.
e efficacy of FMT for IBD will be assessed based on clinical remission or clinical response as defined by the respective study authors. Safety will be assessed by study-reported adverse events and serious adverse events. e analysis will be performed using Review Manager, version 5.1. Results were expressed as risk ratios (RR) and 95% confidence interval (CI). Mantel-Haenszel statistics (M-H) were used to represent the results. e heterogeneity of the study results was evaluated graphically using the inconsistency relative index (I 2 ), and the I 2 statistic (I 2 ≥ as 50% using a random-effects model and I 2 < 50% using a fixed-effects model) was used to evaluate statistical heterogeneity [19], and P value < 0.05 was considered statistically significant. In addition, subgroup analysis was performed according to disease severity, route of administration, type of stool storage (fresh/frozen), GI pretreatment, IBD type, gender, and donor-patient relationship. To assess the stability of the results, sensitivity analysis was performed using STATA 16.0 software. Potential publication bias was assessed using funnel plots and Egger's test.

Literature Search.
A total of 5941 studies, including 5459 foreign language studies, and 482 Chinese language studies, were obtained by searching the database using the search strategy. e journal article title information of these studies was imported into the Endnote software. After removing duplicates, the search in the database yielded 3555 records. During the title and abstract screening process, 3391 records were considered irrelevant, and subsequently, 164 studies were evaluated after full-text assessment, and 150 studies were excluded using predetermined exclusion criteria. Ultimately, 14 studies were obtained, including 12 studies in English and 2 studies in Chinese. All included studies were published as journal articles. e study selection flowchart is shown in Figure 1. e data from these studies are presented in Table 1. e 14 studies included 12 UC studies [20-23, 25, 26, 28-33] and 2 CD studies [24,27] that included a total of 666 patients. FMT was the primary intervention in the RCTs trial group. In contrast, the control group took different measures, with 1 studying different routes of administration by gastroscopy or colonoscopy [27], 7 using placebo controls (including 2 autologous stools [20,23], 2 sham procedures [24,28], and 3 isotonic saline [21,22,25]), 1 using a special UC diet [30], 4 choosing conventional treatment as a control [29,[31][32][33], and 1 adding pectin to the FMT [26]. In 6 cases, the gastrointestinal (GI) tract was pretreated prior to FMT treatment (1 with antibiotic pretreatment [28], 4 with vancomycin and/or polyethylene glycol for bowel preparation [20,23,25,26], and 1 with macrogol solution for rinsing the gastrointestinal tract [27]).

Bias Analysis of Included Studies. Figures 2(a) and 2(b)
summarizes the risk of bias for randomized controlled trials. e risk of bias assessment showed that 6 studies reported the method of randomization, of which 5 studies used computerized randomization as low risk [20-22, 25, 28], 1 study chose randomization according to the order of admission as high risk [33], and the remaining studies did not inform the specific grouping method. Seven studies took participant and researcher allocation concealment [20-22, 24, 25, 28, 30], 6 studies used participant and evaluator blinding [20-23, 25, 28], and the rest were not mentioned. Follow-up was not mentioned in only 1 study [33]. e remaining trials clearly reported the number of missed visits and dropouts and the reasons for them and adequately addressed the issue of selective reporting bias. However, given the poor methodological quality, we noted an unclear risk of bias in most of the included trials.

Safety and Adverse Events.
Adverse events during and after treatment with FMT were reported in the vast majority of studies, with only one study not reporting follow-up. Selflimiting symptoms of gastrointestinal distress were the most common adverse events, including abdominal pain, nausea, diarrhea, abdominal distention, increased stool frequency, loss of appetite, and transient fever. ese adverse events were mild and resolved on their own. Paramsothy et al. [22] found worsening or discomfort in two patients undergoing FMT, but the study found no significant association between individual donors or donor lots and adverse events. erefore, consideration may not be related to FMT. Rossen et al. [23] reported four serious adverse events in their study (including two in the donor FMT group and two in the autologous FMT group), of which one patient was found to have small bowel CD, one required cervical cancer surgery, one was admitted with abdominal pain after 11 weeks of treatment that resolved on its own, and one became severely ill from a primo Cytomegalovirus infection 7 weeks after the first infusion and was found to be in the autologous FMT group, so no serious adverse events occurred related to donor FMT. No adverse events were considered to be associated with FMT. e duration of follow-up ranged from 4 weeks to 12 months.

Frozen or fresh
Total events

Frozen and capsules
Total events

Total (95% CI)
Heterogeneity: Tau 2 = 0.10; Chi 2 = 16.08, df = 10 (P = 0.10); I 2 = 38% Test for overall effect: Z = 2.15 (P = 0.03) Test for subgroup differences: Chi 2 = 0.01, df = 1 (P = 0.94); I 2 = 0% Total events Evidence-Based Complementary and Alternative Medicine sequencing to determine changes in patient stool flora before and after transplantation. FMT significantly changes the gut microbiota of IBD patients, increasing microbial diversity or abundance. Patients treated with donor's stool were found to have a more diverse gut microbiota in favor of the donor. A systematic review was conducted due to the lack of articles on microbiota analysis. Moayeddi et al. [21] found a statistically significant change in microbial composition with more diversity in the FMT group compared to the placebo group. e microbiota of the responders were more similar to those of the donor compared to the nonresponders but did not reach statistical significance. Sokol et al. [24] found that after excluding the two patients who failed FMT at week 6 (the microbiota of the donor was poorly colonized for these two patients), a significant difference was found between the FMT and placebo groups, and the similarity of the microbiota of the FMT group to the donor persisted until the end of follow-up. Crothers et al. [28] found that FMT did not increase Shannon diversity in the subjects but caused changes in gut microbiota levels, which were highly correlated with the donor after 2 weeks of FMT treatment and persisted until 20 weeks. is may be related to the antibiotic pretreatment of subjects prior to FMT treatment in this study. Fang et al. [29] reported that receiving single-donor FMT significantly reconstituted the intestinal microbiota and preserved documentation, with an increase in the proportion of Bacteroidetes and a decrease in the proportion of Aspergillus. e relative abundance of Escherichia coli in the gut of treated subjects decreased significantly; patients treated with FMT who achieved remission also tended to have higher Prevotella abundance.

Sensitivity Analysis
Andpublicationbias. Sensitivity analysis of clinical remission (primary outcome) showed that the effect sizes of the outcome indicators did not change significantly after the exclusion of any of the studies, suggesting that the results of the meta-analysis were more stable and credible. Funnel plots based on publication bias and Egger's test indicated that the conclusions were not affected by publication bias (P � 0.714) (as shown in Figure 12).

Discussion
is article presents an updated systematic review and metaanalysis of randomized controlled trials of FMT for IBD. e   overall study quality was low. Randomization, allocation concealment, and blinding of investigators or subjects or evaluators were the main factors that may have influenced the evaluation of bias, which explains the failure of randomized controlled trial studies to provide high-quality evidence. Most of the studies only mentioned "randomization" or "randomized grouping/assignment" in the text but did not specify the specific methods used and whether allocation concealment or blinding was implemented. In addition, these studies remained heterogeneous in design, with different treatment regimen designs (e.g., route of administration and choice of stool type) and different or poorly defined efficacy endpoints.
After FMT treatment, patients with IBD had a clinical remission rate of 36.9% (168/455) and a clinical response rate of 42.0% (150/357). Positive changes in clinical, endoscopic, or fecal microbiota were present. e clinical remission rate was 30.8% (82/266) for mild-moderate IBD and 45.9% (79/ 172) for active IBD. Our results suggest no significant difference in the degree of disease was found.
In our study, there was no significant difference found between the fresh stool (clinical remission rate: 40.9%) and frozen stool (clinical remission rate: 32.2%). Moreover, some studies have found that fresh FMT has superior clinical efficacy to frozen FMT [34]. Hamilton et al. [35] found that frozen stool microbiota from a healthy donor was more effective in treating recurrent CDI, thereby restoring the structure of the gut microbiota and clearing C. difficile. A randomized controlled trial [36] investigated the effectiveness of frozen FMT and fresh FMT in the treatment of recurrent CDI diarrhea, with clinical remission rates of 75.0% for frozen FMTand 70.3% for fresh FMT in a modified intention-to-treat analysis, showing similar efficacy and no major differences between the two stool types. e results of this study provide evidence for the use of fresh stools.
FMT can be administered via the upper GI tract (nasogastric tube, gastroduodenal tube, oral) or the lower GI tract (colonoscopy, cecum, enema), two different routes. A systematic review has shown that the lower GI tract is a safe and effective route for FMT in the treatment of CDI [37,38]. In our study, the clinical remission rate was 38.2% (151/395) in the lower GI tract and 31.2% (15/48) in the upper GI tract.
Our study suggests that no significant difference in the route of administration was found. Ramai et al. [39] concluded that colonoscopy has a better cure rate than the nasogastric tube for CDI. In addition to efficacy, patient compliance, cost-effectiveness, comfort of administration, level of invasiveness, risk of aspiration and infection, multiplicity of drugs to be administered, and relapse rate appear to be the main factors influencing physicians' choice of route of administration [40]. Some studies suggest that the use of upper GI administration requires attention to adverse events such as patient vomiting [41] and aspiration pneumonia [42], while there are also risks associated with lower GI administration in patients with severe colitis.
erefore, more extensive studies are needed to find low-risk, convenient, and cost-effective routes of administration.
At present, there is no consensus on the optimal donor for FMT. Our study found that fecal donors mainly included healthy adults, relatives or friends, or anonymous volunteers. Clinical remission was 37.6% (153/407) for feces from healthy anonymous donors and 31.2% (15/48) for feces from relatives, close friends, or volunteers. In our study, there was no significant difference found between the efficacy of FMT derived from healthy anonymous donors and the FMT using relatives, close friends, or volunteers. Some studies agree with our results that the ideal donor is a healthy donor [43]. Others have concluded that there is no significant difference between healthy donors and relative or close friend FMT [44]. e advantage of healthy anonymous donors is that they are easy to extract and screen from stool banks, saving time and money for screening stool donors [45]. Considering the possibility that the sample size using relatives or close friends was insufficient, more randomized controlled trials are needed to support our conclusions.
Clinical remission rates were 44.4% (95/214) and 30.3% (73/241) in the group with and without gastrointestinal pretreatment, respectively. In our study, there was no significant difference found between pretreatment of the gastrointestinal tract and treatment with FMT. One study found that bowel cleansing prior to colonoscopy and/or stool disposal led to changes in symptoms and gut microbiota composition in irritable bowel syndrome (IBS) [46]. Bowel cleansing brings immediate changes to the gut microbiota [47,48], which plays an integral role in IBD. However, the amount of relevant trials is insufficient, so a more extensive study in this area is warranted. e clinical response rate was 36.8% (161/438) in UC patients and 41.2% (7/17) in CD patients. In our study, there was no significant difference found between CD patients and UC patients and the number of RCTs in CD is relatively insufficient, so more extensive research in this area is needed to support this conclusion. In our study, no significant differences were found in the gender of the patients.
Currently, the most common adverse events are selflimiting gastrointestinal discomfort, including abdominal pain, bloating, diarrhea, nausea, and other adverse events such as transient fever. ese adverse events usually resolve on their own within 24 hours. e follow-up period is 6 weeks to 36 months. e current systematic review still has some limitations. First, different infusion frequencies (single or multiple FMTs), number of donors, and preparation and storage of donor stools are factors that can affect the results, increasing heterogeneity between studies. Secondly, subgroup analysis (e.g., route of administration, type of stool, gastrointestinal pretreatment, and IBD type) has the limitation of a small number of studies and insufficient sample size. Finally, there is uncertainty regarding the selection of FMT donors, although some studies have shown no significant differences between healthy donors and relatives or close friends. Future larger and more extensive trials are needed to optimize the use of FMT, including donor selection, stool preparation and storage, and route of administration, to obtain better outcomes for patients.

Conclusions
In conclusion, our study found that FMT is an effective, feasible, and safe treatment for IBD. It has a positive effect in terms of clinical remission/clinical response. It can also be done by increasing the diversity and species richness of the gut microbiota. Still, further randomized controlled trials are needed to fully evaluate the long-term efficacy and safety of FMT in patients with IBD.
Data Availability e data supporting this meta-analysis are from previously reported studies and datasets, which have been cited. e processed data are available from the corresponding author upon request.