Procedures in Fecal Microbiota Transplantation for Treating Irritable Bowel Syndrome: Systematic Review and Meta-Analysis

Background: Irritable bowel syndrome (IBS) is a prevalent gastrointestinal disease with no effective treatment. Altered microbiota composition seems implicated in disease etiology and therefore fecal microbial transplantation (FMT) has emerged as a possible treatment therapy. To clarify the clinical parameters impacting FMT efficacy, we conducted a systematic review with subgroup analysis. Methods: A literature search was performed identifying randomized controlled trials (RCTs) comparing FMT with placebo in IBS adult patients (8-week follow-up) with a reported improvement in global IBS symptoms. Results: Seven RCTs (489 participants) met the eligibility requirements. Although FMT seems not to be effective in global improvement of IBS symptoms, subgroup analysis shows that FMT through gastroscopy or nasojejunal tube are effective IBS treatments (RR 3.03; 95% CI 1.94–4.73; I2 = 10%, p < 0.00001). When considering non-oral ingestion routes, IBS patients with constipation symptoms are more likely to benefit from FMT administration (p = 0.003 for the difference between IBS subtypes regarding constipation). Fresh fecal transplant and bowel preparation seem also to have impact on FMT efficacy (p = 0.03 and p = 0.01, respectively). Conclusion: Our meta-analysis revealed a set of critical steps that could affect the efficacy of FMT as clinical procedure to treat IBS, nevertheless more RCTs are needed.


Introduction
Irritable bowel syndrome (IBS) is a symptom-based functional bowel disorder characterized by abdominal pain and altered bowel habits in the absence of detectable structural or biochemical abnormalities [1]. With a prevalence of approximately 4-10% worldwide [2,3], IBS is one of the most prevalent gastrointestinal (GI) disorders and a cause of substantial burden to healthcare services and society [4]. Due to its relapsing and chronic nature, this condition impacts patients' social interactions and quality of life (QOL) [5]. Despite this, current treatments for IBS are often inadequate and, unexpectedly, the pipeline for developing new treatments is relatively poor [6].
According to the gold standard symptom-based diagnostic criteria for IBS, the Rome criteria [1], IBS is classified into 4 subtypes: diarrhea-predominant type (IBS-D), constipationpredominant type (IBS-C), mixed type (IBS-M) or unclassified type (IBS-U) (Supplementary Table S1). Since no specific biomarkers are available to distinguish between different IBS subtypes, criteria is based on the abdominal pain and stool form changes, as assessed by the Bristol Stool Form Scale [7] (Supplementary Table S2).
A multifactorial etiology has been associated to IBS, involving a complex interaction between genetic, psychological and environmental factors that lead to GI motility dysfunction and altered visceral sensations [8]. Some studies have shown an altered microbiota composition in patients with IBS, supporting an important role for the intestinal microbiota in IBS etiology [9][10][11][12]. Since intestinal microbiota play a key role in intestinal immunity and inflammation [13,14], manipulation of its composition has been proposed as a treatment strategy for IBS.
Fecal microbial transplantation (FMT) is a technique in which fecal material containing gut microorganisms are transferred from a healthy donor to a patient, with the intention of correcting imbalances in the microbial community of the gut. FMT can be administered either directly to the colon-via colonoscopy, or less frequently via flexible sigmoidoscopy or an enema-or to the upper gastrointestinal tract via nasoenteric tubes, gastroscopy, or capsule ingestion [15].
Based on the concept of repopulating intestinal microbiota, FMT has been proven effective for the treatment of recurring Clostridioides difficile infection (CDI), by inhibiting its colonization and, so far no major differences have been found between the different FMT delivery modes [15]. However, it remains unclear whether FMT efficacy extends to other gastrointestinal disorders such as IBS. To our best knowledge, eight systematic reviews with meta-analysis have been conducted for evaluating the efficacy of FMT on IBS treatment up-to-now [16][17][18][19][20][21][22][23], four of them in 2022. This number shows the clinical relevance of IBS as well as FMT as therapy. Previous systematic reviews have been consistent in unveiling the route of FMT administration as the major factor that impacts FMT efficacy, however they have not explored in detail other clinical and technical conditions influencing FMT on IBS treatment. This exploitation will be critical for designing novel randomized clinical trials addressing FMT as an intervention procedure for treating patients with IBS. Knowing this, we decided to conduct a systematic review with subgroup analysis for assessing the methodological conditions that are more likely to impact FMT efficacy. This knowledge may allow the optimization of the FMT procedure with potential positive impact on its clinical efficacy for IBS treatment.

Protocol and Registration
This study was developed in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement [24] and the Cochrane Handbook for Systematic Reviews of Interventions [25] guidelines. The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) with the registration code CRD42021252141.

Selection Criteria
We defined inclusion and exclusion criteria in accordance to the PICO (Population, Intervention, Comparator and Outcomes) [26] strategy. Inclusion criteria were: (1) prospective, randomized, double-blind, placebo-controlled trials (parallel group or first arm of cross-over); (2) with adult patients older than 16 years with IBS defined by accepted symptom-based criteria including Manning, Kruis, Rome I, Rome II, Rome III, or Rome IV (Population); (3) compared FMT (Intervention) with placebo consisting of only the FMT excipients or an autologous FMT (Comparator); (4) reported improvement in global IBS symptoms (Outcome); and (5) with a minimum duration of 8-week follow-up, according to the recommended duration for the assessment of short-term response to therapy in functional GI disorders [27].
Review articles, systematic reviews, meta-analysis, letters, conference abstracts, case reports, case series, position papers, and author's replies were excluded. Only studies published in English were included.

Search Strategy
To identify eligible reviews, we searched on Cochrane, MEDLINE, Scopus and Web of Science databases on 27 July 2021. Both medical subject headings (MeSH) terms and free text terms referring to fecal microbiota transplantation combined with terms referring to irritable bowel syndrome were used. The PubMed search strategy was converted to search in other databases (Supplementary Table S3).

Study Selection
We used the online tool Rayyan [28] to remove duplicates and to screen articles for eligibility, according to the screening criteria. Two independent reviewers (S.F. and T.R.) screened the titles and abstracts of the articles for relevance, and full-text articles were reviewed when title and abstract did not provide enough information. Once potentially relevant studies were identified, full-text articles were then assessed for eligibility according to previously established criteria. Excluded trials and the reasons for exclusion were recorded and any disagreement between reviewers was resolved through discussion.

Data Extraction
Data items were extracted by two authors for each study; first author, year of publication, country of origin, sample characteristics, methods, and outcomes. Data regarding the global improvement in IBS symptoms, was extracted as intention-to-treat analyses (with dropouts assumed to be non-responders to FMT) and synthesized into tables. When information was missing or incomplete, the corresponding authors were contacted requesting further information.

Risk of Bias Assessment in Individual Studies
Risk of bias in individual studies was assessed using the updated Cochrane Risk of Bias (RoB 2.0) tool recommended by Cochrane Collaboration [29]. The following five domains were assessed: (1) bias due to the randomization process, (2) bias due to deviations from intended interventions, (3) bias due to missing outcome data, (4) bias in the measurement of the outcome, and (5) bias in the selection of the reported result. Regarding the evaluation of the third domain (missing bias), 10% missing and missing above 5% with imbalances between arms, were classified with "some concerns". The overall risk of bias was classified as; high risk, having some concerns, and low risk. Reviewers were blinded to each other's assessment, and disagreements were solved by reaching a consensus.

Quantitative Synthesis
Relative risk (RR) was used as an effect measure for the dichotomous variable "treatment responders". Effect measures were reported along with the 95% confidence interval (CI). The heterogeneity was assessed through the Cochran's Q (significance level of 0.1) and I 2 tests, and when detected, subgroup analysis was performed to explore possible causes. According to the Cochrane guidelines [25], the I 2 values were interpreted as follows: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; 75% to 100% represent considerable heterogeneity.
Pooled estimates were computed and weighted using generic inverse-variance with random-effect. A p-value < 0.05 was considered as statistically significant. Statistical analysis was performed using Review Manager (RevMan), version 5.4, The Cochrane Collaboration, 2020.

Grading the Evidence
Funnel plots were used to assess evidence of publication bias. Quality assessment of the evidence for each outcome was scored using Grading of Recommendations Assessment, Development and Evaluation (GRADE) [30]. The meta-analysis was scored with a maximum of 10 points, according to (1) risk of bias, (2) precision, (3) heterogeneity, (4) directness, (5) publication bias, (6) funding bias, (7) effect-size, and (8) dose-response. Based on the final score, we classified the quality of the evidence as high, moderate, low, or very low.

Study Selection
The literature search identified 5866 citations, of which 267 were reviewed based on eligibility criteria; 243 of the reviewed references were excluded. Of the 24 remaining citations, 17 were excluded after meticulous full-text review, as detailed in Figure 1. In the end, 7 RCTs [31][32][33][34][35][36][37] (full manuscripts) were eligible and included in our meta-analysis.

Grading the Evidence
Funnel plots were used to assess evidence of publication bias. Quality assessm the evidence for each outcome was scored using Grading of Recommendations A ment, Development and Evaluation (GRADE) [30]. The meta-analysis was scored w maximum of 10 points, according to (1) risk of bias, (2) precision, (3) heterogenei directness, (5) publication bias, (6) funding bias, (7) effect-size, and (8) dose-resp Based on the final score, we classified the quality of the evidence as high, moderate or very low.

Study Characteristics
Detailed characteristics of the included RCTs are summarized in Tables 1 and 2. The sample size of each study ranged from 17 to 165 participants, totalling 489 adults. However, since none of the studies reported a true intention-to-treat analysis, only 465 were analyzed with a total of 298 patients allocated (298 to intervention and 140 to control).   The diagnosis criteria were Rome III or Rome IV. One study included IBS-D only [31], two studies included IBS-D and IBS-M [35,36] and four studies included all 4 subtypes of IBS [32][33][34]37]. FMT was administered using colonoscopy [34,36,37], gastroscopy [32], nasojejunal tube [35], and oral capsules [31,33]. The 5 nonoral ingestion route studies performed single-dose administration of donor or autologous fecal microbiota preparation whilst the two oral capsule FMT studies used multiple doses (3 and 12 doses) of donor fecal microbiota or placebo consisting of FMT excipients alone (no microbiota). The follow-up time varied between 4 months [32], 6 months [31,33,34], and 12 months [35][36][37]. As first outcome, all studies aimed to evaluate the improvement in gastrointestinal symptoms after transplantation, identified by a decrease in IBS Severity Symptom Scale (IBS-SSS) of 75 or more points at 12 weeks [36], a decrease of 50 or more points at 12 weeks [31][32][33]37], and by other tools [34,35].

Risk of Bias Assessment
According to the Cochrane Collaboration tool [29] three RCTs presented some concerns and one was classified as having high risk of bias (Supplementary Figure S1). According to funnel plot analysis, there is no evidence of publication bias (Supplementary Figure S2). However, it should be highlighted that considering the few numbers of studies included in this review, there are probably studies that were not published. We found 3 studies registered on Clinicaltrials.gov, completed more than 18 months ago, that have not published their results yet.

GRADE Assessment
Based on the GRADE assessment (Table 3), the current quality of evidence was "very low" mainly due to the serious risk of bias based on the imprecision of effect estimation. The heterogeneity in the methodology of FMT and placebo interventions between studies also affected the quality, especially in studies with capsule administration.

IBS Symptoms Improvement
From the 489 participants allocated, 465 were included in the analysis of the primary outcome with a symptoms response rate of 66% (185/282) in patients assigned to FMT, and 41% in patients assigned to placebo (75/183), at 12 weeks of follow-up (Supplementary Figure S3). Considering an intention-to-treat approach, the clinical response rate at 12 weeks was 62% (185/298) in the FMT group, and 39% in the placebo group (75/191).
No significant difference in global improvement of IBS symptoms was observed between groups (RR 1.35; 95% confidence interval (CI) 0.75-2.43, p = 0.31 from random effects). Moreover, a significant heterogeneity was identified across all studies (I 2 = 82%) (Figure 2). Given these results, intention-to-treat subgroup analyses were performed to further explore possible heterogeneity sources (Tables 3 and 4). 2). Given these results, intention-to-treat subgroup analyses were performed to further explore possible heterogeneity sources (Tables 3 and 4).

IBS Subtype
Two RCTs [32,33] performed subgroup analysis based on IBS subtype and found no differences in the response rate at 12 weeks between the IBS subtypes. Since only two studies grouped efficacy data for different IBS subtypes, we divided them in two groups based on [36] the presence or absence of constipation: with constipation type [32][33][34] and without constipation type [31,[35][36][37].
Subgroup analyses showed that IBS subtype does not significantly influence the efficacy of FMT in IBS treatment (p = 0.77 for the difference between subgroups) ( Table 4). In both groups, with and without constipation type, FMT was not associated with symptoms improvement compared with placebo (RR 1.61; 95% CI 0.30-8.69; I 2 = 93%, and RR 1.25; 95% CI 0.87-1.79; I 2 = 31%, respectively).
However, when studies with capsules were excluded, subgroup analyses found that IBS subtype significantly influences the efficacy of FMT in IBS treatment (p = 0.003 for the difference between subgroups) (Table 5). Indeed, FMT was associated with higher symptom improvement outcomes when administrated to patients with IBS subtypes with constipation (RR 3.50; 95% CI 2.19-5.60; I 2 = 0%; p < 0.001).

Safety of FMT in IBS
Complete adverse events (AEs) data were available for five studies [32][33][34]36,37]. After pooling data from the five studies, 53 (23%) of 231 patients assigned to FMT reported at least one adverse event, compared with 44 (30%) of 147 allocated to placebo. No significant difference in the total number of AEs was observed in patients receiving FMT compared to control patients (RR 0.91; 95% CI 0.58-1.41), with moderate heterogeneity between studies (I 2 = 47%, p = 0.67) (Figure 4). However, when studies with capsules were excluded, subgroup analyses found that IBS subtype significantly influences the efficacy of FMT in IBS treatment (p = 0.003 for the difference between subgroups) (Table 5). Indeed, FMT was associated with higher symptom improvement outcomes when administrated to patients with IBS subtypes with constipation (RR 3.50; 95% CI 2.19-5.60; I 2 = 0%; p < 0.001).

Safety of FMT in IBS
Complete adverse events (AEs) data were available for five studies [32][33][34]36,37]. After pooling data from the five studies, 53 (23%) of 231 patients assigned to FMT reported at least one adverse event, compared with 44 (30%) of 147 allocated to placebo. No significant difference in the total number of AEs was observed in patients receiving FMT compared to control patients (RR 0.91; 95% CI 0.58-1.41), with moderate heterogeneity between studies (I 2 = 47%, p = 0.67) (Figure 4). In total, four participants had serious AEs. Two patients developed diverticulitis 2 months after FMT (both had diverticulosis verified by colonoscopy and experienced several diverticulitis attacks before FMT) [36], one participant had transient vertigo and nausea after the FMT procedure, requiring a few hours of observation in the hospital [36] and one patient died by suicide during the follow-up [35].

Summary of Evidence
We conducted a systematic review and meta-analysis to identify and explore critical steps in the FMT procedure that must be controlled for the efficacy of this therapy in the treatment of IBS patients.
Using the global improvement in IBS symptoms at 12 weeks after FMT as an endpoint, 7 RCTs involving 489 participants were statistically inconclusive mainly due to high heterogeneity. To explore the methodological factors that may have contributed to this heterogeneity, we carried out multiple subgroup analysis targeting the following variables: delivery method, dosage, fresh versus frozen stool, bowel preparation and IBS subtypes.
Regarding the FMT administration method, FMT through multiple-dose oral liquid capsules [31,33] or colonoscopy [34,36,37] showed no benefit, while FMT via gastroscopy and nasojejunal tubes [32,35] demonstrated a clinical benefit in global IBS symptom improvement compared to placebo. The difference observed between FMT capsule and nonoral ingestion could be due to microbial viability disparities in the FMT content after delivery. Indeed, a higher bacterial viability is expected when donor microbiota is directly released in the gastrointestinal tract of the receiver. Nevertheless, other methodological shortcomings may have contributed to the low efficacy of studies with colonoscopy and In total, four participants had serious AEs. Two patients developed diverticulitis 2 months after FMT (both had diverticulosis verified by colonoscopy and experienced several diverticulitis attacks before FMT) [36], one participant had transient vertigo and nausea after the FMT procedure, requiring a few hours of observation in the hospital [36] and one patient died by suicide during the follow-up [35].

Summary of Evidence
We conducted a systematic review and meta-analysis to identify and explore critical steps in the FMT procedure that must be controlled for the efficacy of this therapy in the treatment of IBS patients.
Using the global improvement in IBS symptoms at 12 weeks after FMT as an endpoint, 7 RCTs involving 489 participants were statistically inconclusive mainly due to high heterogeneity. To explore the methodological factors that may have contributed to this heterogeneity, we carried out multiple subgroup analysis targeting the following variables: delivery method, dosage, fresh versus frozen stool, bowel preparation and IBS subtypes.
Regarding the FMT administration method, FMT through multiple-dose oral liquid capsules [31,33] or colonoscopy [34,36,37] showed no benefit, while FMT via gastroscopy and nasojejunal tubes [32,35] demonstrated a clinical benefit in global IBS symptom improvement compared to placebo. The difference observed between FMT capsule and non-oral ingestion could be due to microbial viability disparities in the FMT content after delivery. Indeed, a higher bacterial viability is expected when donor microbiota is directly released in the gastrointestinal tract of the receiver. Nevertheless, other methodological shortcomings may have contributed to the low efficacy of studies with colonoscopy and capsule administration. Namely, considering the FMT administrated by colonoscopy, two of the three studies [34,36] used different cut-offs for treatment response, which may have led to an underestimation in the efficacy of these RCTs. Responses were defined by a decrease of more than 75 points assessed by IBS-SSS in Johnsen et al. [36], and at least 30% in the total GSRS-IBS symptom score in Holster et al. [34]. Considering capsule administration, in Aroniadis et al. [31] less than 30 g was administered despite 30 g being the dose of fecal transplant recommended by the European Committee on Organ Transplantation [38] and the European Consensus [39]. Furthermore, in Halkjaer et al. [33] final fecal suspensions were stored at −20 • C, when the aforementioned guidelines recommend storage at −80 • C to avoid enzyme activity that can lead to degradation of sensitive microbial populations (e.g., Bacteroidetes) [40]. To avoid these disparities, FMT should be produced in a stool banking center following the European and International consensus guidelines for Good Manufacturing Practices (GMP) of FMT donations [38].
Considering the FMT dosage, only three of the seven RCTs analyzed in this study, used a dose of 50 g or more [32,33,36]. Our meta-analysis shows that higher dose (≥50 g) did not result in greater improvement in IBS global symptoms compared with a lower dose per transplant. However, due to the low number of RCTs included in this meta-analysis and the higher heterogeneity yielded (93%), it is difficult to draw definitive conclusions regarding adequate FMT dose.
When comparing fresh versus frozen FMT, our meta-analysis found a significant improvement in IBS symptoms when patients received fresh donor stool. However, as already mentioned by Wu et al. [18] interpretation of this result should be done cautiously since the fresh FMT was exclusively delivered through colonoscopy and nasojejunal tube, and that there was high heterogeneity (I 2 = 90%) among studies using frozen FMT. Thus, the efficacy of frozen FMT for IBS treatment needs to be clarified due to its advantage in terms of implementation in routine clinical practice.
Subgroup analyses revealed that bowel preparation may improve the efficacy of noncapsule FMT. These results are in line with previous findings that suggest bowel preparation can alter the fecal microbiota in healthy individuals [41,42] and with the last European consensus on FMT in clinical practice, that recommends bowel preparation before FMT [34].
Finally, we found that IBS subtype significantly influences the FMT efficiency delivered through colonoscopy, gastroscopy and nasojejunal tube. This may be related to the different characteristics between IBS subtypes, both in terms of clinical manifestations and in terms of microbiota alteration. Thus, some consideration should be given to stratifying randomized controlled trials by IBS subtype, to clarify whether the existence of constipation symptomology influences the efficacy of FMT.

Global Considerations about Randomized Controlled Trials Evaluating the FMT Efficacy on IBS Treatment
In clinical trial design, one methodological consideration that may affect the efficacy of FMT in IBS treatment, and possible cause of the heterogeneity yielded, is the lack of standardization in the recruited patients. Only two of the RCTs included in our review [31,32] considered the diagnosis of small intestinal bacterial overgrowth (SIBO) in the exclusion criteria. SIBO causes GI symptoms such as abdominal pain, bloating, gas, distension, flatulence, and diarrhea, that can lead to an incorrectly diagnosed IBS [43]. Likewise, PI-IBS that frequently occurs after an episode of infectious gastroenteritis [44], and may have a different microbiota signature [45], was only considered as an exclusion criteria in two trials [32,34]. Aroniadis et al. [31] revealed a trend toward greater improvement in PI-IBS patients who received FMT, according to a post-hoc analysis.
Also important, only four studies [31][32][33][34] excluded participants supplemented with probiotics prior to FMT and none gave specific instructions regarding their diet during the follow-up time, other than to keep it stable. Indeed, only one study [36] reported changes in dietary habits during the follow-up. Diet can affect many aspects of gut physiology such as motility, permeability, microbiome, visceral sensation, brain-gut interactions, immune regulation and neuro-endocrine function [46], thus being a relevant confounding variable.
Participants' background diet and change in diet during intervention should be reported to exclude any effect on IBS symptoms.
Regarding medication, only one study [32] excluded patients that were under concomitant IBS medication, such as antimotility, antispasmodic and antidepressant drugs [6]. As medication may mask IBS symptoms and affect gut microbiota composition [47], its intake should be monitored before FMT and during the follow-up.

Strengths and Limitations of This Study
Several limitations in this systematic review should be acknowledged. First, our analyses are limited by the low number of available studies and the quality of the reported data. Second, most studies were performed in Europe, limiting generalizability. Third, in order to perform subgroup analysis, we simply divided FMT dosage and IBS subtypes into two groups, we did not have access to raw data from all studies. Fourth, we did not assess the impact on quality of life. This aspect was already addressed in a previous meta-analysis [18] that found a significant improvement in quality of life in IBS patients 12 weeks after FMT. Fifth, the study populations diversity may also have contributed to the heterogeneity of the results. For instance, some studies included patients with different disease severities and patients had a wide age range.
Despite all limitations, our systematic review brings a comprehensive overview of the methodological limitations in clinical trial design, as well as differences in FMT interventions performed to date. Thus, this review highlights FMT procedure variables that could contribute to the contradictory effects on FMT efficacy.

Future RCTs Assessing FMT Efficacy for IBS Treatment
Future RCTs may benefit from FMT donations prepared in stool banking centers under GMP conditions, but also from a stratification by IBS subtype and disease severity. Thus, the RCTs should exclude participants with the diagnosis of confounding diseases, as well as considering background diet and change in diet during intervention and the use of medication before and during the follow-up period. More well-designed RCTs are needed to firstly assess whether the efficacy of capsules made of materials that allow the delivery of FMT to a specific intestinal location, is comparable with other delivery methods. Secondly, the impact of stool formulation (fresh, liquid frozen or even lyophilized), the FMT doseresponse and lastly, the bowel lavage preparation on FMT efficacy must be investigated. This knowledge will allow optimization of the FMT procedure and thus assess its true clinical potential for IBS patients and beyond.

Conclusions
The present systematic review with meta-analysis shows that IBS patients may benefit from FMT when administered via gastroscopy or nasojejunal tube and that FMT is overall safe for IBS. Furthermore, IBS subtype and bowel lavage may play an important role in the response to FMT.  Funding: This work was sponsored by national funds through FCT, Fundação para a Ciência e a Tecnologia, I.P., within the scope of the projects CINTESIS, R&D Unit (reference UIDB/4255/2020) and "RISE-LA/P/0053/2020".
Institutional Review Board Statement: Not applicable. Systematic reviews and meta-analyses do not involve human subjects and do not require IRB re-view.
Informed Consent Statement: Not applicable. Systematic reviews and meta-analyses do not involve human subjects.
Data Availability Statement: Not applicable. This study uses data already published and available in the cited references.