Abstract
Background:
Fecal microbiota transplantation (FMT) is a highly effective treatment for recurrent Clostridioides difficile infection (CDI). However, 10–20% of patients still fail to recover following FMT. There is a need to understand why these failures occur and if there are modifiable factors that can be addressed by clinicians performing FMT.
Aims:
We sought to identify factors related to the FMT procedure itself which could impact FMT outcomes. We also aimed to identify patient demographics which might be associated with FMT outcomes and whether any factors were associated with early FMT failure compared to late CDI recurrence.
Methods:
We performed a retrospective multicenter cohort analysis of FMT procedures between October 2005 and November 2020. We collected data on patient demographics, details of the FMT procedure, and procedure outcomes. Using univariate and multivariate regression, we evaluated whether these factors were associated with long-term FMT success, early FMT failure (less than 60 days following procedure), or late CDI recurrence (more than 60 days following procedure).
Results:
Long-term success of FMT was strongly correlated with any delivery of stool to the terminal ileum (Odds Ratio [OR] 4.83, 95% confidence interval [CI] 1.359–17.167) and underlying neurologic disease (OR 8.012, 95% CI 1.041–61.684). Lower bowel prep quality was significantly associated with both early FMT failure (p=0.034) and late CDI recurrence (p=0.050).
Conclusions:
Delivery of stool to the terminal ileum is significantly associated with long-term success following FMT. This is a relatively safe practice which could easily be incorporated into the standard of care for colonoscopic FMT.
Keywords: Fecal microbiota transplantation, Clostridioides difficile infection, clinical factors in failure of fecal microbiota transplantation
Introduction
Clostridioides difficile infection (CDI) remains a significant burden on the healthcare system in both the US and across the world1,2. It is now well-described that fecal microbiota transplantation (FMT) is a highly successful procedure for treatment of multiply recurrent CDI (rCDI), and is recommended for patients with rCDI by both major Gastroenterology and Infectious Diseases guidelines3,4. Despite the successes of FMT, however, 10–20% of patients will fail to recover following initial FMT5. Therefore, there remains a need to understand why these failures occur and how to improve our treatment of CDI.
To address these gaps in patient care, several recent publications have attempted to identify factors associated with recurrence of CDI following FMT. These studies have demonstrated that inpatient status, poor quality of bowel preparation, and underlying inflammatory bowel disease (IBD), among others, appear to be associated with a higher risk of failure of FMT5–7. However, knowledge gaps on how procedure-related factors such as the use of loperamide before or after the procedure and location of endoscopic delivery of transplant material, can affect outcomes following FMT. Critically, these factors may be leveraged by the endoscopist to improve FMT outcomes, but evaluation of these factors has been limited in metanalyses due to considerable heterogeneity among trials of FMT8.
In this study, we sought to identify patient- and procedure-related factors associated with successful FMT. We also evaluated characteristics associated with early failure of the procedure or later recurrence of CDI following FMT.
Methods
Study design and population
We conducted a retrospective multicenter analysis of FMT procedures for CDI at Stanford Health Care and East Bay Center for Digestive Health (Oakland, CA). Approval was obtained from the Institutional Review Board at Stanford University (IRB 54929, approved 9/1/2020). Our primary aim was to determine clinical factors significantly associated with long-term success of FMT, defined as lack of symptom recurrence or a positive stool test for C. difficile following the procedure for the duration of the study. Our secondary aim was to determine if any factors were associated with early FMT failure (development of symptoms congruent with CDI leading to CDI therapy and/or positive C. difficile stool test in the first 60 days following FMT) or recurrent CDI (development of symptoms congruent with CDI leading to CDI therapy and/or a positive C. difficile stool test more than 60 days following the procedure).
We evaluated all FMT procedures performed for CDI in patients over the age of 18 at Stanford Health Care from July 2015 to July 2020, and from October 2005 to November 2020 at East Bay Gastroenterology. Patients were excluded if no documented follow up after FMT was available for review.
We extracted clinical data on patients before and after FMT, including demographics (age, sex, body mass index [BMI]), past medical history (presence of cardiovascular disease, diabetes, neurologic disease, mood disorder, immune compromise, use of narcotics at time of FMT), past GI history (presence of inflammatory bowel disease [IBD], functional GI disorders, other relevant GI history), CDI history (number and type of prior anti-CDI antibiotic therapy), and details of each FMT procedure including bowel prep quality, location of stool infusion (oral pill or any stool delivery to upper GI tract, TI, right colon, transverse colon, or left colon), volume of stool infusion, and use of loperamide.
Statistical analysis
The rate of primary outcome (long-term success following FMT) and secondary outcomes (early FMT failure, late CDI recurrence), and predictive value of clinical variables on FMT outcomes, odds ratio (OR) with its 95% confidence interval (CI), and P-values were calculated using Stata version 14 (StataCorp LLC, College Station, TX). Dichotomous variables were analyzed for outcomes using the chi-squared test or the Fisher’s exact test where appropriate, and continuous variables were analyzed using Student’s T-tests if normally distributed, or the Wilcoxon signed-rank test for non-normal data. Multivariate analyses were performed with a logistic regression model that incorporated the following variables: age, sex, BMI, inpatient status, delivery of material to the terminal ileum (TI), most distal location of stool infusion, bowel preparation quality, use of loperamide, number of prior antibiotics for CDI, prior GI disease, IBD, immune compromised status, diabetes, cardiovascular disease, neurologic disease, mood disorder. Forward elimination was used to add variables that significantly contributed to the regression model at a significance cutoff of p ≤ 0.15. In all cases p ≤ 0.05 was considered significant.
Results
Clinical characteristics of patients undergoing FMT
We identified a total of 241 FMT procedures performed at our study sites between October 2005 and November 2020. Six were excluded from our study as there was no clinical follow up available. Of the remaining 235 procedures, 168 procedures (82.8%) resulted in long-term success following FMT, with no recurrence of CDI for the study duration. 32 procedures (13.6%) resulted in FMT failure within 60 days of the procedure. Of the 203 procedures which were initially successful, there were 35 (17.2%) recurrences of CDI more than 60 days after FMT.
Overall, 169 (71.9%) of total patients were female. The median patient age was 65 (IQR 50–79). 80 (34.0%) had underlying GI disease (excluding GERD and liver disease), including 19 (8.1%) with IBD and 27 (11.5%) with functional GI diseases. Our cohort surprisingly included 63 patients (26.8%) who were immune compromised, due to either a primary immunodeficiency or treatment with immune suppressing medications (Supplementary Table 1). Cardiovascular disease (83 patients [35.3%]) and mood disorders (86 patients [36.6%]) were also common.
Demographics were similar between patients who experienced long-term success of FMT and those who had either FMT failure or recurrent CDI (Table 1). Patients with long-term post-FMT success were significantly more likely to have had colonoscopic delivery of stool (p = 0.033) and higher quality bowel preparation (p = 0.019). They also significantly differed in the distal-most location of stool delivery (p = 0.035).
Table 1:
Patient characteristics (n = 235)
| Clinical Variables | All Patients (N = 235) | FMT Long-term success1 (N = 168) | FMT Failure2 or CDI Recurrence3 (N = 67) | |
|---|---|---|---|---|
| Age, median (IQR) | 65 (50-79) | 65 (51-79) | 66 (48-83) | 0.773 |
| Body mass index, median (IQR) | 23.46 (20.32-28.155) | 23.79 (20.37-27.46) | 23.33 (19.92-29.3) | 0.878 |
| Sex | 1.000 | |||
| Male, no. (%) | 66 (28.1%) | 47 (28.0%) | 19 (28.4%) | |
| Female, no. (%) | 169 (71.9%) | 121 (72.0%) | 48 (71.6%) | |
| FMT Procedure | ||||
| Colonoscopic delivery, no. (%) | 220 (93.6%) | 161 (95.8%) | 59 (88.1%) | 0.033 |
| Delivery to terminal ileum, no. (%) | 96 (40.9%) | 71 (42.3%) | 25 (37.3%) | 1.000 |
| Loperamide used, no. (%) | 173 (73.6%) | 126 (75.0%) | 47 (70.1%) | 0.512 |
| Stool volume infused, median mL (IQR) | 250 (250-350) | 250 (250-300) | 250 (250-450) | 0.165 |
| Inpatient procedure, no. (%) | 19 (8.09%) | 10 (5.95%) | 9 (13.4%) | 0.067 |
| Most distal location of infusion | 0.035 | |||
| Oral (pill), no. (%) | 4 (1.70%) | 1 (0.60%) | 3 (4.48%) | |
| Upper GI, no. (%) | 6 (2.55%) | 3 (1.79%) | 3 (4.48%) | |
| Terminal ileum, no. (%) | 18 (7.66%) | 15 (8.93%) | 3 (4.48%) | |
| Right colon, no. (%) | 148 (63.0%) | 110 (65.5%) | 38 (56.7%) | |
| Transverse colon, no. (%) | 37 (15.7%) | 23 (13.7%) | 14 (20.9%) | |
| Left colon, no. (%) | 13 (5.53%) | 12 (7.14%) | 1 (1.49%) | |
| Bowel prep | 0.019 | |||
| Poor/inadequate, no. (%) | 11 (4.68%) | 7 (4.17%) | 4 (5.97%) | |
| Adequate, no. (%) | 22 (9.36%) | 11 (6.55%) | 11 (16.4%) | |
| Fair, no. (%) | 30 (12.8%) | 21 (12.5%) | 9 (13.4%) | |
| Good, no. (%) | 138 (58.7%) | 110 (65.5%) | 28 (41.8%) | |
| GI and CDI History | ||||
| GI disease, no. (%) | 80 (34.0%) | 58 (34.5%) | 22 (32.8%) | 0.879 |
| IBD, no. (%) | 19 (8.09%) | 13 (7.74%) | 6 (9.00%) | 0.793 |
| Functional GI disease, no. (%) | 27 (11.5%) | 19 (11.3%) | 8 (11.9%) | 0.801 |
| Prior CDI antibiotics courses, median (IQR) | 4 (3-5) | 4 (3-5) | 4 (3-5) | 0.333 |
| Medical History | ||||
| Immune compromised, no. (%) | 63 (26.8%) | 50 (29.8%) | 13 (19.4%) | 0.142 |
| Diabetes, no. (%) | 32 (13.6%) | 22 (13.1%) | 10 (14.9%) | 0.680 |
| Cardiovascular disease, no. (%) | 83 (35.3%) | 61 (36.3%) | 22 (32.8%) | 0.653 |
| Neurologic disease, no. (%) | 30 (12.8%) | 19 (11.3%) | 11 (16.4%) | 0.287 |
| Mood disorder, no. (%) | 86 (36.6%) | 23 (37.5%) | 23 (34.3%) | 0.764 |
| Baseline narcotics use, no. (%) | 75 (31.9%) | 49 (29.2%) | 26 (38.8%) | 0.165 |
Lack of symptoms or positive stool C. difficile PCR test following FMT for duration of study.
Recurrent symptoms or positive stool C. difficile PCR test within 60 days following FMT.
Recurrent symptoms or positive stool C. difficile PCR test more than 60 days following FMT.
Factors associated with long-term FMT success
In our cohort of patients, delivery of stool to the terminal ileum was significantly associated with an increased likelihood of long-term successful FMT (Odds ratio [OR] 4.83, 95% confidence interval [CI] 1.359–17.167; Table 2) on multivariate analysis. Interestingly, we found that underlying neurologic disease was also associated with increased likelihood of long-term successful FMT (OR 8.012, 95% CI 1.041–61.684). Poor or inadequate bowel prep was associated with a decreased likelihood of long-term post-FMT success, with OR 0.409 (95% CI 0.208–0.803, p = 0.009). Additionally, higher BMI (OR 0.856, 95% CI 0.754–0.970), higher number of prior antibiotics courses for CDI treatment (OR 0.683, 95% CI 0.476–0.981), concurrent GI diseases (OR 0.124, 95% CI 0.026–0.589), and immune compromised status (OR 0.124, 95% CI 0.024–0.642) emerged as risk factors associated with decreased likelihood of FMT success.
Table 2:
Multivariate predictors of long-term success following FMT
| Long-term success1 |
||
|---|---|---|
| Clinical Variables | n = 168 | |
|
| ||
| Odds Ratios (95% CI) | P-Value | |
| Age | 1.015 (0.984 - 1.047) | 0.326 |
| Female sex | 2.256 (0.520 - 9.788) | 0.277 |
| BMI | 0.856 (0.754 - 0.970) | 0.015 |
| Inpatient procedure | 0.338 (0.033 - 3.513) | 0.364 |
| Delivery to terminal ileum | 4.830 (1.359 - 17.167) | 0.015 |
| Most distal location of infusion2 | 1.362 (0.527 - 3.524) | 0.524 |
| Bowel prep (poor or inadequate) | 0.409 (0.208 - 0.803) | 0.009 |
| Stool volume infused | 1.004 (0.997 - 1.011) | 0.254 |
| Loperamide used | 1.263 (0.217 - 7.367) | 0.795 |
| Number of prior CDI antibiotics courses | 0.683 (0.476 - 0.981) | 0.039 |
| Prior GI disease | 0.124 (0.026 - 0.589) | 0.009 |
| Inflammatory bowel disease | 1.746 (0.192 - 15.857) | 0.621 |
| Immune compromised | 0.124 (0.024 - 0.642) | 0.013 |
| Diabetes | 0.879 (0.156 - 4.944) | 0.883 |
| Cardiovascular disease | 1.616 (0.384 - 6.807) | 0.513 |
| Neurologic disease | 8.012 (1.041 - 61.684) | 0.046 |
| Mood disorder | 1.195 (0.371 - 3.854) | 0.681 |
Lack of symptoms or positive stool C. difficile PCR test following FMT for duration of study.
Oral (pill), upper GI, terminal ileum, right colon, transverse colon, left colon.
Early failure versus late CDI recurrence
To identify differences between patients with early FMT failure compared to later recurrence, we subdivided our patient cohort into those who had recurrence within 60 days of the procedure compared to those who did not (Table 3) Baseline characteristics between patients who had a successful procedure and those with early return of CDI were not significantly different. The most distal location of stool infusion was significantly different between these groups (p = 0.04) as was quality of bowel preparation (p = 0.034). Next, we studied patients with later recurrence of CDI more than 60 days following FMT (Table 4) and found that patients with poor bowel prep quality at index FMT were more likely to have late recurrence of CDI (p = 0.050).
Table 3:
Factors associated with early FMT failure compared to initial success (n = 235)
| Clinical Variables | FMT Success1 (N = 203) | FMT Failure2 (N = 32) | P-value |
|---|---|---|---|
| Age (median, IQR) | 65 (52-79) | 67 (49.5-84.5) | 0.382 |
| Body mass index, median (IQR) | 23.515 (20.37-27.53) | 23.45 (19.86-29.535) | 0.955 |
| Sex | 1.000 | ||
| Male, no. (%) | 54 (26.6%) | 12 (37.5%) | |
| Female, no. (%) | 149 (73.4%) | 20 (62.5%) | |
| FMT Procedure | |||
| Colonoscopic delivery, no. (%) | 192 (94.6%) | 28 (87.5%) | 0.138 |
| Delivery to terminal ileum, no. (%) | 88 (43.3%) | 8 (25.0%) | 0.150 |
| Loperamide used, no. (%) | 148 (72.9%) | 25 (78.1%) | 0.668 |
| Stool volume infused, median mL (IQR) | 250 (250-350) | 250 (250-375) | 0.643 |
| Inpatient procedure, no. (%) | 15 (7.39%) | 4 (12.5%) | 0.303 |
| Most distal location of infusion | 0.040 | ||
| Oral (pill), no. (%) | 1 (0.49%) | 3 (9.38%) | |
| Upper GI, no. (%) | 6 (2.96%) | 0 (0%) | |
| Terminal ileum, no. (%) | 16 (7.88%) | 2 (6.25%) | |
| Right colon, no. (%) | 131 (64.5%) | 17 (53.1%) | |
| Transverse colon, no. (%) | 30 (14.8%) | 7 (21.9%) | |
| Left colon, no. (%) | 12 (5.91%) | 1 (3.13%) | |
| Bowel prep | 0.034 | ||
| Poor/inadequate, no. (%) | 11 (5.42%) | 0 (0%) | |
| Adequate, no. (%) | 15 (7.39%) | 7 (21.9%) | |
| Fair, no. (%) | 26 (12.8%) | 4 (12.5%) | |
| Good, no. (%) | 124 (61.1%) | 14 (43.8%) | |
| GI and CDI History | |||
| GI disease, no. (%) | 72 (30.6%) | 8 (25%) | 0.317 |
| IBD, no. (%) | 19 (9.36%) | 0 (0%) | 0.084 |
| Functional GI disease, no. (%) | 23 (11.3%) | 4 (12.5%) | 0.712 |
| Prior CDI antibiotics courses, median (IQR) | 4 (3-5) | 4 (3-4) | 0.264 |
| Medical History | |||
| Immune compromised, no. (%) | 56 (27.6%) | 7 (21.9%) | 0.668 |
| Diabetes, no. (%) | 26 (12.8%) | 6 (18.8%) | 0.404 |
| Cardiovascular disease, no. (%) | 72 (37.4%) | 11 (34.4%) | 1.000 |
| Neurologic disease, no. (%) | 26 (12.8%) | 4 (12.5%) | 1.000 |
| Mood disorder, no. (%) | 74 (36.5%) | 12 (37.5%) | 1.000 |
| Baseline narcotics use, no. (%) | 64 (31.5%) | 11 (34.4%) | 0.839 |
Lack of symptoms or positive stool C. difficile PCR test within 60 days following single FMT.
Recurrent symptoms or positive stool C. difficile PCR test within 60 days following FMT.
Table 4:
Factors associated with late CDI recurrence after successful initial FMT (n = 203)
| Clinical Variables | CDI Recurrence1 (N = 35) | No Recurrence2 (N = 168) | P-value |
|---|---|---|---|
| Age (median, IQR) | 65 (47-78) | 65 (52-79) | 0.545 |
| Body mass index, median (IQR) | 23.03 (20.41-29.05) | 23.79 (20.37-27.46) | 0.802 |
| Sex | |||
| Male, no. (%) | 7 (20.0%) | 47 (28.0%) | 0.404 |
| Female, no. (%) | 28 (80.0%) | 121 (72.0%) | |
| FMT Procedure | |||
| Colonoscopic delivery, no. (%) | 31 (88.6%) | 161 (95.8%) | 0.098 |
| Delivery to terminal ileum, no. (%) | 17 (48.6%) | 71 (42.3%) | 0.338 |
| Loperamide used, no. (%) | 22 (62.9%) | 126 (75.0%) | 0.148 |
| Stool volume infused, median mL (IQR) | 300 (250-450) | 250 (250-300) | 0.164 |
| Inpatient procedure, no. (%) | 5 (14.3%) | 10 (5.95%) | 0.145 |
| Most distal location of infusion | 0.087 | ||
| Oral (pill), no. (%) | 0 (0%) | 1 (0.60%) | |
| Upper GI, no. (%) | 3 (8.57%) | 3 (1.79%) | |
| Terminal ileum, no. (%) | 1 (2.86%) | 15 (8.93%) | |
| Right colon, no. (%) | 21 (60.0%) | 110 (65.5%) | |
| Transverse colon, no. (%) | 7 (20.0%) | 23 (13.7%) | |
| Left colon, no. (%) | 0 (0%) | 12 (7.14%) | |
| Bowel prep | 0.050 | ||
| Poor/inadequate, no. (%) | 4 (11.4%) | 7 (4.17%) | |
| Adequate, no. (%) | 4 (11.4%) | 11 (6.55%) | |
| Fair, no. (%) | 5 (14.3%) | 21 (12.5%) | |
| Good, no. (%) | 14 (40.0%) | 110 (65.5%) | |
| GI and CDI History | |||
| GI disease, no. (%) | 14 (40.0%) | 58 (34.5%) | 0.564 |
| IBD, no. (%) | 6 (17.1%) | 13 (7.74%) | 0.106 |
| Functional GI disease, no. (%) | 4 (11.4%) | 19 (11.3%) | 0.548 |
| Prior CDI antibiotics courses, median (IQR) | 4 (3-5) | 4 (3-5) | 0.650 |
| Medical History | |||
| Immune compromised, no. (%) | 6 (17.1%) | 50 (29.8%) | 0.149 |
| Diabetes, no. (%) | 4 (11.4%) | 22 (13.1%) | 1.000 |
| Cardiovascular disease, no. (%) | 11 (31.4%) | 61 (36.3%) | 0.699 |
| Neurologic disease, no. (%) | 7 (20.0%) | 19 (11.3%) | 0.170 |
| Mood disorder, no. (%) | 11 (31.4%) | 63 (37.5%) | 0.566 |
| Baseline narcotics use, no. (%) | 15 (42.9%) | 49 (29.2%) | 0.160 |
Recurrent symptoms or positive stool C. difficile PCR test more than 60 days following successful FMT.
Lack of symptoms or positive stool C. difficile PCR test more than 60 days following successful FMT.
Discussion
Our study identified multiple patient- and procedure-related factors associated with long-term success following FMT. Most interestingly, we found that delivery of stool to the terminal ileum during colonoscopic FMT was strongly associated with positive outcomes from the procedure, as was patient history of neurologic disease. In contrast, poor bowel prep, BMI, prior CDI antibiotic exposure, underlying GI disease, and immune compromise were associated with lower likelihood of long-term FMT success. When we examined early (<60 days) and late (>60 days) outcomes, bowel prep remained significantly associated with FMT failure and CDI recurrence.
Many of these findings are consistent with prior literature on variables associated with failure of FMT. Two recent metanalyses also found that poor bowel prep was significantly associated with failure of the procedure across multiple studies, along with IBD, inpatient FMT, and severe CDI7,9. Most of the included studies defined failure as recurrence of symptoms or positive C. difficile stool studies within 1–3 months of FMT and did not examine recurrences after this time point, in contrast to our study. One study did examine early (within six months) versus late (after six months) failures of the procedure and found that IBD and use of non-CDI antibiotics after FMT were associated with both early and late failures10. In contrast to our findings, this study found that bowel prep quality was only associated with early FMT failure, although this is complicated by the difference in definition of early and late failure in our study.
In addition to examining early compared to late failures of the procedure, our study also examined variables associated with the procedure itself which may be modifiable by the endoscopist. These included location of stool delivery, whether or not the procedure was performed via colonoscopy, use of loperamide, and the volume of stool infused during the procedure.
Our major finding differs from that of a prior report which also investigated the impact of location of stool delivery on FMT success11. These authors found no difference in outcomes following TI delivery compared to delivery in the colon. However, this study only considered failures within 12 weeks following the procedure. In contrast, our findings indicate that stool delivery to the TI is most beneficial in preventing later recurrences of CDI. Our study cohort was also much larger, with 235 FMT procedures compared to 109, and represents multicenter findings. Two other reports have also examined the use of anti-secretory therapy including loperamide prior to FMT and did not find a difference in procedure failure5,10. This is consistent with our finding that loperamide use was not associated with either early or late FMT failure.
While prior studies have included patient historical variables such as sex, age, immune compromised status, and history of GI disease, very few have investigated other patient baseline factors. One study noted the proportion of patients with co-morbidities including hypertension, diabetes, cancer, and chronic kidney disease and evaluated whether a high Charlson co-morbidity index was associated with FMT failure12. However, none of these appeared to be significant. In contrast to prior work, our study included an expanded set of patient variables and indeed found that lower BMI and neurologic disease were associated with long-term success following FMT.
Potentially, our findings may be related to increased transplanted stool retention. For example, delivery of stool into the TI, above the ileocecal valve, might result in extended release of stool into the colon and allow for improved bacterial colonization. This is supported by findings that colonoscopic delivery of stool is associated with a higher rate of FMT success compared to enema13. Similarly, patients with neurologic disorders frequently have slowed colonic motility14, which might increase contact time of transplanted stool with the colonic mucosa and improve colonization. However, use of loperamide prior to the procedure as well as baseline narcotics use, which both slow intestinal motility15, were not associated with improved success following FMT. One possibility to reconcile this incongruity is that a single dose of loperamide may not substantially impact GI motility compared to a chronic GI-slowing condition. Alternatively, these neurological diseases are associated with other factors that could contribute to the efficacy of FMT, such as baseline microbiome composition or host bile acid composition16,17. Beyond simply increasing GI transit time of donor stool, effects on ileal microbiome and/or small intestinal immunological responses could also contribute to the long-term benefit of TI stool delivery that we found in our study. For example, similar to the appendix, it is possible that the ileum could act as a microbial “reservoir” that is protective against future recurrences of CDI18. In addition, there is some emerging evidence that microbial and immunological changes to the small intestine may impact colonic inflammation19, and it is possible similar feedback may occur in FMT. This feedback might limit the ability of C. difficile to re-germinate or re-colonize in the colon or could limit the severity of inflammatory response against the pathogen. Possible de facto delivery of stool material to the appendiceal orifice may have been beneficial to patients in our study for these same reasons, although such specific delivery is difficult to assess given the retrospective nature of our work.
A higher number of prior antibiotics courses for CDI was significantly associated with FMT failure in our study. Current guidelines suggest that FMT should be used when patients have experienced at least their second recurrence of CDI3 and our findings strengthen the suggestion to treat with FMT at this point without delaying until after further recurrences. Of note, however, available data suggest that the use of FMT as treatment for the initial episode of CDI is not superior to antibiotics20,21. Whether FMT should be offered as a treatment for primary CDI or a first recurrence of CDI is under investigation22,23.
For most patients in our study and prior published work, it appears that a single FMT is sufficient for prevention of further CDI24. Of those who fail to recover following their initial FMT, many appear to respond to a repeat procedure5. This suggests that multiple FMT procedures are not necessary for the majority of patients; however, identifying who may be at risk for recurrent CDI following FMT may reveal a population who could benefit from a second FMT procedure a priori. One such group may well be patients with severe CDI5, although these patients may also be at higher risk of failing a second FMT25. Of note, no patients in our study met criteria for severe CDI per published guidelines3.
Our findings may have significant impact on how FMT is delivered colonoscopically. Unlike many prior factors that have been identified in successful FMT, delivery of stool into the TI is typically modifiable by the endoscopist performing the procedure. This represents a small change in practice with minimal increased risk to the patient which could result in long-term improvement in FMT outcomes.
In addition to these promising findings, our study has several strengths. We evaluated patients from multiple sites representing both academic and community settings, which allowed us to evaluate outcomes across a variety of real-world settings. As above, we also evaluated factors associated with FMT outcomes which have not previously been examined. There are nonetheless limitations to our study. This is a retrospective, nonrandomized study and therefore cannot evaluate causation. Although we have defined early FMT failure versus late recurrence of CDI here to be 60 days, across prior literature this time point has varied from one to six months (most commonly one to three months), making comparison to prior findings somewhat challenging. We have overcome this limitation to some extent by focusing on long-term success of the procedure. However, this also increases the potential for additional confounders we have not accounted for.
Overall, our work has uncovered several new factors contributing to long-term success of FMT. Most importantly, delivery of stool to the TI is associated with FMT success, and given the relative ease and safety of this action could easily be incorporated into the standard of practice for colonoscopic FMT. Prospective evaluation of TI delivery of stool will be necessary to clarify its role in successful FMT.
Supplementary Material
Figure 1.
Flowchart of procedures included in study and outcomes.
External Grant Support:
NIDDK T32DK007056-46 (ARW)
Footnotes
Disclosures: ARW holds the following patent which relates to the content of this manuscript: Compositions and methods for transplantation of colon microbiota, US Patent 2014/0147417 A1, filed 9 March 2012, accepted 29 May 2014.
References
- 1.Ma GK, Brensinger CM, Wu Q, Lewis JD. Increasing Incidence of Multiply Recurrent Clostridium difficile Infection in the United States: A Cohort Study. Ann Intern Med. 2017;167(3):152. doi: 10.7326/M16-2733 [DOI] [PubMed] [Google Scholar]
- 2.Balsells E, Shi T, Leese C, et al. Global burden of Clostridium difficile infections: a systematic review and meta-analysis. J Glob Health. 2019;9(1):010407. doi: 10.7189/jogh.09.010407 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kelly CR, Fischer M, Allegretti JR, et al. ACG Clinical Guidelines: Prevention, Diagnosis, and Treatment of Clostridioides difficile Infections. Am J Gastroenterol. 2021;116(6):1124–1147. doi: 10.14309/ajg.0000000000001278 [DOI] [PubMed] [Google Scholar]
- 4.McDonald LC, Gerding DN, Johnson S, et al. Clinical Practice Guidelines for Clostridium difficile Infection in Adults and Children: 2017 Update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis. 2018;66(7):987–994. doi: 10.1093/cid/ciy149 [DOI] [PubMed] [Google Scholar]
- 5.Fischer M, Kao D, Mehta SR, et al. Predictors of Early Failure After Fecal Microbiota Transplantation for the Therapy of Clostridium Difficile Infection: A Multicenter Study. Am J Gastroenterol. 2016;111(7):1024–1031. doi: 10.1038/ajg.2016.180 [DOI] [PubMed] [Google Scholar]
- 6.Ianiro G, Valerio L, Masucci L, et al. Predictors of failure after single faecal microbiota transplantation in patients with recurrent Clostridium difficile infection: results from a 3-year, single-centre cohort study. Clin Microbiol Infect. 2017;23(5):337.e1–337.e3. doi: 10.1016/j.cmi.2016.12.025 [DOI] [PubMed] [Google Scholar]
- 7.Tariq R, Hayat M, Pardi D, Khanna S. Predictors of failure after fecal microbiota transplantation for recurrent Clostridioides difficile infection: a systematic review and meta-analysis. Eur J Clin Microbiol Infect Dis. 2021;40(7):1383–1392. doi: 10.1007/s10096-021-04163-z [DOI] [PubMed] [Google Scholar]
- 8.Feuerstadt P, Aroniadis OC, Svedlund FL, et al. Heterogeneity of Randomized Controlled Trials of Fecal Microbiota Transplantation in Recurrent Clostridioides difficile Infection. Dig Dis Sci. Published online July 18, 2021. doi: 10.1007/s10620-021-07141-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Beran A, Sharma S, Ghazaleh S, et al. Predictors of Fecal Microbiota Transplant Failure in Clostridioides difficile Infection: An Updated Meta-analysis. J Clin Gastroenterol. 2022;Publish Ahead of Print. doi: 10.1097/MCG.0000000000001667 [DOI] [PubMed] [Google Scholar]
- 10.Tariq R, Saha S, Solanky D, Pardi DS, Khanna S. Predictors and Management of Failed Fecal Microbiota Transplantation for Recurrent Clostridioides difficile Infection. J Clin Gastroenterol. 2021;55(6):542–547. doi: 10.1097/MCG.0000000000001398 [DOI] [PubMed] [Google Scholar]
- 11.Patron RL, Hartmann CA, Allen S, et al. Vancomycin Taper and Risk of Failure of Fecal Microbiota Transplantation in Patients With Recurrent Clostridium difficile Infection. Clin Infect Dis. 2017;65(7):1214–1217. doi: 10.1093/cid/cix511 [DOI] [PubMed] [Google Scholar]
- 12.Meighani A, Hart BR, Mittal C, Miller N, John A, Ramesh M. Predictors of fecal transplant failure: Eur J Gastroenterol Hepatol. 2016;28(7):826–830. doi: 10.1097/MEG.0000000000000614 [DOI] [PubMed] [Google Scholar]
- 13.Tariq R, Pardi DS, Bartlett MG, Khanna S. Low Cure Rates in Controlled Trials of Fecal Microbiota Transplantation for Recurrent Clostridium difficile Infection: A Systematic Review and Meta-analysis. Clin Infect Dis. 2019;68(8):1351–1358. doi: 10.1093/cid/ciy721 [DOI] [PubMed] [Google Scholar]
- 14.Camilleri M Gastrointestinal motility disorders in neurologic disease. J Clin Invest. 2021;131(4):e143771. doi: 10.1172/JCI143771 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Benyamin R Opioid Complications and Side Effects. Pain Physician. 2008;2s;11(3;2s):S105–S120. doi: 10.36076/ppj.2008/11/S105 [DOI] [PubMed] [Google Scholar]
- 16.Sampson TR, Debelius JW, Thron T, et al. Gut Microbiota Regulate Motor Deficits and Neuroinflammation in a Model of Parkinson’s Disease. Cell. 2016;167(6):1469–1480.e12. doi: 10.1016/j.cell.2016.11.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li P, Killinger BA, Ensink E, et al. Gut Microbiota Dysbiosis Is Associated with Elevated Bile Acids in Parkinson’s Disease. Metabolites. 2021;11(1):29. doi: 10.3390/metabo11010029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sanders NL. Appendectomy and Clostridium difficile colitis: Relationships revealed by clinical observations and immunology. World J Gastroenterol. 2013;19(34):5607. doi: 10.3748/wjg.v19.i34.5607 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bousbaine D, Fisch LI, London M, et al. A conserved Bacteroidetes antigen induces anti-inflammatory intestinal T lymphocytes. Science. 2022;377(6606):660–666. doi: 10.1126/science.abg5645 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Juul FE, Garborg K, Bretthauer M, et al. Fecal Microbiota Transplantation for Primary Clostridium difficile Infection. N Engl J Med. 2018;378(26):2535–2536. doi: 10.1056/NEJMc1803103 [DOI] [PubMed] [Google Scholar]
- 21.Singh T, Bedi P, Bumrah K, et al. Fecal Microbiota Transplantation and Medical Therapy for Clostridium difficile Infection: Meta-analysis of Randomized Controlled Trials. J Clin Gastroenterol. 2021;Publish Ahead of Print. doi: 10.1097/MCG.0000000000001610 [DOI] [PubMed] [Google Scholar]
- 22.Grant J Fecal Microbiota Transplantation Versus Standard Medical Therapy for Initial Treatment of Recurrent Clostridium Difficile Infection. clinicaltrials.gov; 2019. Accessed September 22, 2022. https://clinicaltrials.gov/ct2/show/NCT02255305
- 23.Garborg K COmparative Effectiveness of IntestinaL MicrobiOta Versus VaNcomycin for Primary c. Difficile Infection - RandomiZEd Trials. clinicaltrials.gov; 2022. Accessed September 22, 2022. https://clinicaltrials.gov/ct2/show/NCT03796650
- 24.van Nood E, Vrieze A, Nieuwdorp M, et al. Duodenal Infusion of Donor Feces for Recurrent Clostridium difficile. N Engl J Med. 2013;368(5):407–415. doi: 10.1056/NEJMoa1205037 [DOI] [PubMed] [Google Scholar]
- 25.Allegretti JR, Mehta SR, Kassam Z, et al. Risk Factors that Predict the Failure of Multiple Fecal Microbiota Transplantations for Clostridioides difficile Infection. Dig Dis Sci. 2021;66(1):213–217. doi: 10.1007/s10620-020-06198-2 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.