Fecal Microbiota Transplantation: Redefining Surgical Management of Refractory Clostridium difficile Infection

Abstract

Fecal microbiota transplantation (FMT) is the process of transplanting stool from a healthy donor into the gut of a diseased individual for therapeutic purposes. It has a clearly defined role in the treatment of recurrent Clostridium difficile (reclassified as “ Clostridioides difficile ”) infection (CDI), with cure rates over 90% and decreased rates of subsequent recurrence compared with anti-CDI antibiotics. There is emerging evidence that FMT is also effective in the treatment of severe and fulminant CDI, with associated decreases in mortality and colectomy rates compared with standard antibiotic therapy. FMT shows promise as salvage therapy for critically-ill CDI patients refractory to maximum medical therapy and not deemed to be surgical candidates. FMT should be considered early in the course of severe CDI and should be delivered immediately in patients with signs of refractory CDI. Expansion of FMT's use along the spectrum of CDI severity has potential to decrease associated rates of mortality and colectomy.

Fecal microbiota transplantation (FMT) describes the process of instilling homogenized stool from a healthy donor into the gut of a diseased recipient with a therapeutic intent. The origins of FMT dates back to the fourth century in China, when human fecal suspension was administered by mouth of patients who had food poisoning or severe diarrhea yielding positive results. ¹ In the modern era, interest in FMT for therapeutic purposes has taken off at a feverish pace due to the emergence of research attempting to understand the complex interactions of host and microbiome. FMT is now an accepted therapeutic modality for Clostridium difficile infection (CDI), with interest in treatment of other diseases related to alterations in gut microbiota on the horizon as well, such as with ulcerative colitis, metabolic syndrome, hepatic encephalopathy, and irritable bowel syndrome.

Mechanism of FMT

C. difficile (recently reclassified as Clostridioides difficile ² ) is one of the most common superbugs of our time. Its outbreak has reached epidemic proportions due to widespread use of antibiotics and resultant disruption of the innate gut flora, ³ making hosts susceptible to invasion, colonization, and ultimately infection by C. difficile . Paradoxically, first-line therapy for CDI involves antibiotics that, while effectively clear the infection, further dysbiosis at the same time. It is to no surprise that the likelihood of CDI recurrence is amplified after each successive CDI infection despite treatment, ⁴ a phenomenon attributable to progressively decreasing levels of colonic microbial diversity proportional to the number of previous antibiotic treatments. ⁵ FMT quickly restores microbial diversity and the dominance of favorable organisms, suppressing C. difficile activity via multiple mechanisms of action. First, healthy microbiota competes for niche and nutritional resources. ⁶ Second, the microbiota actively produce bacteriostatic or bacteriocidal peptides, the so-called bacteriocins. ³ Nisin and thuricin, produced by Bacillus thuringiensis and Lactococcus lactis , respectively, are highly effective against C. difficile . ^{7 8} Third, metabolism of primary bile acids into secondary bile acids with ratios inhibitory to C. difficile's spore germination and vegetative growth is restored by transplanted microbes. ⁹ Lastly, indigenous microbial organisms in the gut achieve homeostasis with the host which optimizes gut barrier function and promotes a quiescent mucosal immune system. ¹⁰

U.S. Guidelines for FMT

FMT has been recommended for the third or subsequent CDI recurrence following a vancomycin taper by the American College of Gastroenterology (ACG) since 2013. ¹¹ FMT is now also supported by the Infectious Disease Society of American (IDSA) in the latest iteration of their guidelines released in 2018 for the treatment of the second or subsequent CDI recurrences. ¹² These recommendations are based on a large number of cohort studies and multiple randomized controlled trials (RCTs) on FMT for recurrent CDI (RCDI) with documented success rates approaching 90%. ¹³ Furthermore, rates of CDI recurrence are lower after FMT (5–15%) ^{14 15} compared with antibiotics vancomycin (35–65%) ¹⁶ and fidaxomicin (25%). ¹⁷

FMT Delivery

The method of FMT administration varies widely among institutions, with differences in stool donor type, route of delivery, and stool preparation. The process of FMT begins with rigorous evaluation of potential stool donors for any potentially transmittable diseases. Specific recommendations for comprehensive donor screening can be found in a consensus statement published by the European FMT Working Group. ¹⁸

Multiple studies have compared rates of RCDI cure between donor type, stool processing method, and delivery routes for FMT. There are no significant differences in efficacy when anonymous (universal) versus patient-directed donors is used, ¹³ thus banked universal donor stool is preferred in urgent cases as some donor screening tests may take up to 2 weeks to complete. With regard to stool type, frozen/thawed stool was found to be noninferior to fresh stool in terms of clinical efficacy in a randomized clinical trial. ¹⁹ Studies have produced mixed results upon comparison of freeze-dried (lyophilized) versus standard fresh stools. In one double-blind randomized trial where FMT was delivered via colonoscopy, success was significantly impacted; 78% for lyophilized and 100% for fresh stool. ²⁰ However, with further optimization of the lyophilization process, Khoruts and colleagues achieved a cure rate of 88% while administering only 2 to 3 capsules (encasing 2.1–2.5 × 10 ¹¹ bacteria). ²¹ They showed that even without bowel preparation, the microbiota engrafted well and the host's gut microbiome normalized by 1 month. There were only a few minor adverse events and patients reported no aesthetic concerns associated with lyophilized capsules.

Stool banks have revolutionized the FMT landscape. The availability of rigorously screened and processed donor stool immensely simplified the logistics of FMT for clinicians, thus breaking down the main barrier in access to therapy. In 2017, 87.5% of U.S. population lived within 1-hour driving distance from an FMT provider, while 99.9% could reach a health care facility offering FMT within a 4-hour drive. ²²

There have been numerous randomized clinical trials comparing rates of CDI resolution with different routes of delivery. Upper gastrointestinal (GI) delivery is generally performed via nasoduodenal/nasojejunal tube, esophagoduodenoscopy, push enteroscopy, or capsules, while lower GI routes include delivery by enema or sigmoidoscopy/colonoscopy. A collaborative analysis of 14 studies demonstrated higher rates of failure when upper GI FMT was performed compared with lower GI delivery, with a hazard ratio of 3.4. ²³ At 30 and 90 days, rates of clinical failure were 5.6 and 17.9%, respectively, in the upper versus 4.9 and 8.5%, respectively, in the lower GI delivery group.

Overall, the treatment of RCDI in a nonurgent, outpatient setting may be most ideal with oral application of lyophilized FMT capsules. Capsules containing freeze-dried microbiota are easy to handle, remained stable at a wide range of temperature at least up to 4 days, and associated with fewer aesthetic concerns. However, we recommend FMT via colonoscopy in cases of severe or fulminant CDI, previous FMT failure for RCDI, toxic megacolon, ileus, and cases where direct visualization of the colonic mucosa may be necessary to guide further therapy, rule out alternative etiologies, and to ensure proper delivery of FMT material.

Severe and Fulminant CDI

Surgical Management

Historically, approximately 30% of patients with severe CDI undergo surgical intervention ²⁴ for disease that is refractory to medical therapy or results in generalized peritonitis or colonic perforation. ²⁵ The 30-day postoperative mortality rate is high, reported between 32.7% in a large National Surgical Quality Improvement Program (NSQIP) database study ²⁶ and 41.3% in a meta-analysis. ²⁴ These high rates of mortality are likely due to emergent surgeries or delayed surgical interventions after medical failure. Unfortunately, it is challenging to perfect the timing of surgery because it is difficult to foresee the course of disease and to identify patients who will ultimately fail medical management. Furthermore, patients who develop severe/fulminant CDI are often the same ones who have risk factors for poor postcolectomy outcomes and/or have multiple comorbidities that make them poor surgical candidates or not surgical candidates at all.

Several risk models have been developed in an attempt to identify patients who are likely to progress to severe or fulminant CDI ^{27 28 29} however, subsequent external validation using a cohort with low rates of fulminant CDI resulted in only moderate performance. This study demonstrated low rates of sensitivity for a complicated CDI course, ranging from 26 to 61% using the aforementioned predictive models. ³⁰ A preoperative model to risk stratify patients with medically refractory CDI being considered for total abdominal colectomy (TAC) is a step in the right direction to avoid futile surgery; however, external validation will need to be performed before this risk score can be widely adopted. ²⁶

Alternative surgical approaches to CDI management have also been proposed. Neal and colleagues introduced the diverting loop ileostomy with intraoperative colonic lavage followed by anterograde vancomycin flushes (DLI) approach in 2011. When compared with matched historical controls, treatment with TAC demonstrated a decrease in mortality from 50 to 19%. ³¹ This technique (DLI) also preserved the colon in 93% of patients. Unfortunately, issues have been raised regarding the study design and its results have not been reproduced. A subsequent retrospective study demonstrated no significant difference between TAC and DLI; 30-day mortality was 30% in DIL versus 23% in TAC ( p  = 1.0) with similar mortality rates at 1 year (40 vs. 46%). ³²

FMT for Severe and Fulminant CDI

The success of FMT in RCDI therapy has naturally led to exploration its potential in severe and fulminant cases. ³³ A limited number of studies have broached this topic but initial data appears promising. Early on, success rates as high as 79% for severe CDI ³⁴ and 66% for fulminant CDI were reported. ³⁵ However, following a remarkable but transient improvement in clinical status and objective laboratory parameters, many patients relapsed and some ultimately succumbed to the infection despite delayed reinitiation of anti-CDI therapy or repeat FMT. ^{15 36} This observation led to the development of an inpatient sequential FMT protocol with selected use of oral vancomycin. ³⁷ In this protocol, patients with severe or fulminant CDI undergo FMT via colonoscopy. If pseudomembranes are visualized, vancomycin is restarted within 48 hours post-FMT. If diarrheal symptoms further persist for 3 to 5 days of post-FMT, subsequent FMT(s) are delivered until resolution of pseudomembranes. When pseudomembranes are no longer appreciated, vancomycin is not restarted. This protocol has demonstrated superior cure rates in 57 patients, with 100% cure in severe CDI, and 87% cure in fulminant CDI. ³⁸ A recent RCT compared single FMT infusion followed by 14 days of vancomycin therapy to pseudomembrane-guided multiple FMT infusions (every 3 days until resolution of pseudomembranes) combined with oral vancomycin for a total of 14 days. ³⁹ Cure rates were 75% for the single FMT arm and 100% for the multiple FMT arm ( p  = 0.01). The patients in the multiple-FMT arm received on average four FMTs (range: 3–6). No serious adverse events were associated with either treatment protocol.

Retrospective studies have also demonstrated that FMT produces more favorable outcomes compared with standard medical therapy in hospitalized patients with severe and fulminant CDI. In a French hospital, where infectious disease specialists recommended early FMT for hospitalized patients with CDI, among 111 patients, the 3-month mortality was 12.1% in the FMT group and 42.2% in the standard antibiotic group, with no patients undergoing surgery. ⁴⁰ The odds of dying at 3 months significantly decreased in severe CDI patients who received an FMT (OR = 0.08, p  = 0.001). The study also found that just two patients with severe CDI would need treatment with FMT to save one life. A separate study by an Italian group reported an abrupt drop in CDI-related colectomy cases after initiation of an FMT program. ⁴¹ In another retrospective study, integration of an inpatient sequential FMT program was also associated with decreased CDI-related mortality and colectomy when compared with traditional medical therapy. ⁴² Among hospitalized patients with severe or fulminant CDI, mortality decreased from 10.2 to 4.5% ( p  = 0.021) and colectomy decreased from 6.8 to 2.7% ( p  = 0.042).

Perhaps the most impactful role for FMT is in its potential to cure CDI refractory to maximum medical therapy. This is a category of CDI where clinicians are often presented with a dilemma. Early surgical consultation is recommended in patients with underlying IBD, vasopressor requirements, and impending sepsis or multiorgan failure. ²⁵ While early colectomy has been associated with improved rates of survival in fulminant and refractory CDI, ⁴³ the substantial morbidity and mortality associated with colectomy is difficult to rationalize unless no other options are available. Patients tend to continue with medical management, and unfortunately a significant portion of these patients further deteriorate, become too unstable for surgery, and eventually succumb to their disease. The introduction of FMT to this subset of patients not only constitutes a new treatment modality but may also stabilize partial responders to FMT such that they are better candidates for surgery.

Comparing a 4-year time period before and after availability of inpatient FMT, one study demonstrated significant improvement in mortality (43.2 vs. 12.1%, p  < 0.001) and decreased rates in colectomy (31.8 vs. 7.6%, p  = 0.001) among patients with severe or fulminant CDI that were refractory to medical therapy (defined as continued worsening and/or nonresponse to at least 5 days of oral vancomycin). ⁴² Notably, this study included 16.4% of patients with toxic megacolon and 29.1% with ileus. Within this study, 32 patients with refractory severe or fulminant CDI were also determined to be too ill to be considered for colectomy by a surgical team. Seventeen of these patients continued medical therapy while 15 entered the sequential FMT protocol. In comparing these two groups, patients who received FMT had significantly decreased CDI-related mortality (20.0 vs. 82.4%, p  = 0.001).

While data from these retrospective studies are compelling, RCTs are lacking. Without them, some experts wonder if it is too early to recommend FMT for severe and fulminant CDI. ⁴⁴ Still, multiple authors have advocated for FMT to be formally recommended as a therapeutic option in severe, fulminant, and medically refractory cases of CDI, and even as first line therapy, despite the lack of RCTs. ^{40 45} Bypassing RCTs may seem sacrilegious, but is not an unprecedented phenomenon when evaluating novel treatment options for diseases with high rates of mortality and a lack of alternative therapies. Byar and colleagues proposed five criteria that should be used to identify cases where nonrandomized studies instead of RCTs could be justified. ⁴⁶ These five criteria appear to be met in severe and fulminant cases of CDI, as first asserted by Hocquart and colleagues. ⁴⁰ First, overall rates of death in fulminant CDI are comparable to that of the recent Ebola outbreak, as well as the 1990s HIV epidemic. ^{46 47} Limitation of FMT in patients with severe and fulminant CDI result in a uniformly poor prognosis. Second, FMT has shown an excellent safety profile in the critically-ill population with no serious adverse events relating to the FMT material having been reported. ⁴⁸ Third, the previously mentioned studies have demonstrated a significant and “unambiguous” benefit in those receiving FMT versus medical therapy. Fourth, significant decreases in CDI recurrence, mortality, and colectomy, as well as recent breakthroughs in mechanistic, insights make a convincing argument for FMT and why it is effective. Lastly, designing an RCT with a suitable comparator to FMT has been elusive for researchers for multiple reasons. Comparison to a surgical arm would be difficult given the fact that a portion of patients would not be surgical candidates. Moreover, it may be impossible to standardize the timing of surgery. Late intervention unnecessarily would place patients at risk of progression toward medical nonresponsiveness, hemodynamic instability, perforation, and ischemia. On the other end of the spectrum, surgery prior to determining if patients respond to medical therapy inordinately exposes patients to the high rates of morbidity associated with colectomy.

FMT for CDI in Special Populations

Inflammatory Bowel Disease

Patients with underlying inflammatory bowel disease (IBD) have higher rates of CDI and its associated complications of colectomy and mortality are substantially higher relative to those with CDI alone or IBD alone. ^{49 50 51} FMT is highly effective in curing CDI in patients with underlying IBD ⁵² however, there is concern that introduction of foreign microbiota may incur a disproportionate immune response resulting in worsening IBD activity. ⁵³ In a meta-analysis of 29 studies where FMT was provided to IBD patients for either IBD management or CDI treatment, the pooled rate of worsening IBD activity after FMT was 14.9%. ⁵⁴ However, the rate of IBD worsening was only 4.6% when restricting the analysis to randomized controlled trials. These findings suggest that FMT should still be utilized for CDI treatment in patients with underlying IBD; however, their post-FMT course should be monitored closely for signs of IBD activity flare or worsening.

Solid Organ Transplant Recipients

Patients with a history of solid organ transplantation (SOT) have a higher incidence of CDI, ⁵⁵ and a larger percentage of patients have a recurrence or progression to severe/fulminant CDI. ^{56 57} Unfortunately, the American Society of Transplantation in 2013 discouraged the use of FMT in SOT due to a lack of data regarding FMT in immunocompromised patients at the time of publication, as well as theoretical concerns, for bacteremia. ⁵⁸ Since that time, studies have found no increased incidence in infectious complications or adverse events related to FMT in SOT patients compared with their immunocompetent counterparts. ^{59 60} However, there appears to be a lower rate of cure after a single FMT, 74% for recurrent CDI, and <30% for severe and fulminant cases. ⁶⁰ Subsequent antibiotics and/or FMT(s) given in a multicenter study resulted in higher overall rates of cure, reaching 94, 85.7, and 71.4% in patients with recurrent, severe, and fulminant CDI, respectively. SOT patients should be monitored closely post-FMT for signs of early FMT failure in which case subsequent administration of empiric antibiotics and/or FMT(s) should be considered.

Conclusion

FMT has a well-defined role in the treatment of RCDI, with an emerging body of evidence demonstrating that it is also effective in severe, fulminant, and refractory CDI. The widespread availability of FMT, low risk of adverse events, as well as high rates of efficacy (particularly with multiple infusions), make it an attractive therapeutic option. Colectomy for refractory CDI will always be necessary particularly in cases of concurrent ischemia or perforation. However, where clinicians have traditionally faced the dilemma of choosing between continuing ineffective antibiotic therapy versus subjecting patients to surgery, we now have a promising treatment option in FMT.