Clinical Practice and Infrastructure Review of Fecal Microbiota Transplantation for Clostridium difficile Infection

Abstract

A substantial proportion of Clostridium difficile infection (CDI) cases recur after completion of antibiotic therapy, and antibiotic cure rates diminish with each recurrence of CDI. Fecal microbiota transplantation (FMT) is an effective therapy for recurrent FMT, which otherwise requires prolonged or indefinite antibiotic treatment. FMT is performed by introducing the fecal microbial community obtained from a healthy donor or pool of donors into the stomach, small intestine, or colon of a patient with CDI. Multiple clinical trials support the usefulness of FMT in treating recurrent CDI, and CDI treatment guidelines now include consideration of FMT at the third CDI recurrence. However, there remain challenges to incorporating FMT into clinical practice. First, methods of fecal bacterial community processing vary, as do methods of FMT administration. Second, the optimal dosing strategy and expected benefit of FMT for refractory CDI, particularly for severe and severe complicated cases, are uncertain. Third, the US Food and Drug Administration (FDA) considers FMT an investigational treatment. Fourth, insurance reimbursement for FMT usually falls short of FMT administration costs. In the setting of rising C difficile incidence and growing evidence for FMT efficacy, the demand for FMT has increased. However, uncertainty surrounding optimal FMT preparation and administration methods, FDA oversight, and insurance reimbursement presently limits the clinical practice of FMT.

Clostridium difficile infection (CDI) is a clinical syndrome typically characterized by loose, frequent bowel movements and abdominal pain, which occur as a result of toxin production by colonic C difficile.^{1, 2, 3, 4} CDI occurs in patients colonized with toxigenic C difficile and in patients who have newly acquired C difficile. Estimates of the relative contributions of persistent colonization and new acquisition to incident CDI vary.^{5, 6} In both cases, the pathogenesis of CDI involves depletion of non-C difficile colonic microbiota, altered bile acid metabolism, germination of (resident or recently ingested) C difficile spores, expansion of a population of vegetative C difficile, toxin production, and colonic inflammation.^{7, 8, 9, 10, 11, 12}

The primary event in CDI pathogenesis appears to be depletion of non-C difficile colonic microbiota with a loss of colonization resistance and a concomitant increase in C difficile spore germination.^{8, 13, 14, 15, 16} The risk factors known to correlate with risk for C difficile colonization and CDI—antibiotic exposure, gastric acid suppression (with proton pump inhibitors, in particular), and age—each plausibly contributes to this process.^{17, 18, 19, 20} Consistent with current understanding of CDI pathogenesis, delivery of colonic microbiota from a healthy donor, or fecal microbiota transplantation (FMT), long has been known to eliminate the symptoms of CDI.^{21, 22, 23, 24, 25, 26} Such treatment appears to succeed in relieving the symptoms of CDI in proportion to its success in reconstituting the bacterial community of the donor, though complete engraftment is not necessary, and microbes other than bacteria may play a role as well.^{27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37}

The first-line treatment for CDI remains anti-C difficile antibiotics (metronidazole, oral vancomycin, or fidaxomicin).^{2, 3} Treatment of CDI with anti-C difficile antibiotics is typically successful in relieving CDI symptoms, but anti-C difficile antibiotics may suppress other colonic microbiota, and C difficile may persist for months as a spore. As many as 20% of CDI cases treated with anti-C difficile antibiotics recur, and the risk of recurrent CDI (R-CDI) increases with each subsequent CDI episode.^{38, 39} Enthusiasm for FMT has increased as CDI incidence has increased, and clinical trials consistently have demonstrated the efficacy of FMT in resolving R-CDI.^{27, 31, 40, 41, 42, 43, 44, 45} However, the clinical use of FMT remains difficult because of uncertainty surrounding optimal methods of fecal bacterial community processing and FMT administration, as well as requirements related to US Food and Drug Administration (FDA) oversight and insurance reimbursement. Reviews provide valuable insight into CDI pathogenesis,^{7, 15, 46} setting up a frozen stool bank,^{47, 48} FMT administration methods,^{49, 50} and the details of FDA oversight of FMT.^{51, 52} Here we review current FMT clinical practice, paying particular attention to issues relevant to FMT administration for patients who are hospitalized and critically ill.

CDI Diagnosis and Eligibility for FMT

Accurate identification of CDI is difficult but essential to the successful use of FMT. Accurate identification of CDI demands (1) evaluation of risk factors for C difficile colonization and infection, as well as the consistency of the manifesting syndrome with the common clinical features of CDI; (2) consideration of alternative diagnoses; and (3) assessment of stool C difficile test results. The first and second steps are crucial given ongoing debate regarding the diagnostic usefulness of the nucleic acid amplification test (NAAT) for toxigenic C difficile. NAAT is a first-line test for CDI according to current US guidelines.^{2, 3} However, NAAT has come under scrutiny because some studies suggest that NAAT may be too sensitive and fail to discriminate CDI from C difficile colonization. Patients with positive NAAT but negative enzyme immunoassay (EIA) results for C difficile toxin have less C difficile bacterial load, fecal inflammation, and diarrhea and fewer CDI-related complications.^{53, 54, 55} Given this evidence, published guidelines recommend that positive NAATs be validated with a more specific toxin A and B EIA (a two-step protocol).⁵⁶

The pretest probability of CDI is higher if one or more risk factors for C difficile colonization and CDI are present. Key risk factors include exposure to antibiotics, exposure to gastric acid suppressants (proton pump inhibitors, in particular), and age.^{17, 38, 39, 57, 58, 59, 60, 61, 62, 63, 64, 65} Health-care exposure is another risk factor for CDI, but health-care exposure must be understood to extend beyond hospitalization. A 2015 survey of 15,461 CDI cases in the United States demonstrated that 65.8% of cases were associated with health-care exposure but that only 24.2% started during hospitalization.³⁸

The typical clinical manifestation of CDI includes abdominal pain and cramping accompanied by frequent, watery stools.^{1, 66} Melena and hematochezia are rare.^{2, 67} Signs of a systemic inflammatory response to colitis, including fever and leukocytosis, may be present, and the presence of such symptoms defines severe complicated CDI (SC-CDI). Symptoms typically improve within a week of starting appropriate anti-CDI antibiotics, though dose escalation may be required to induce a response in some cases.³ Postinfectious irritable bowel syndrome (IBS), which can occur in the post-CDI period even after successful CDI treatment and may include prominent lactose intolerance, sometimes interferes with the full resolution of GI symptoms.⁶⁸

The diagnosis of CDI typically is based on (1) NAAT, (2) a two-step algorithm that combines EIA testing for glutamate dehydrogenase (GDH) and C difficile toxins A and B, or (3) a three-step algorithm that uses GDH and toxin EIA testing up front and NAAT if GDH and toxin results are discordant.^{2, 3, 4, 69, 70} As discussed, the diagnostic validity of positive NAAT results in the absence of positive toxin EIA results is under dispute.^{53, 54} In our experience, it is essential to consider alternative diagnoses if NAAT results are persistently positive but response to high doses of anti-CDI antibiotics is equivocal. Inflammatory bowel disease (IBD) is a risk factor for both C difficile colonization and CDI. Particularly when hematochezia is a prominent clinical feature, response to appropriate anti-CDI therapy is minimal, and diagnosis of CDI rests on NAAT results alone, IBD should be considered.⁶⁸

R-CDI is defined as CDI that recurs after initially successful therapy. Ten percent to 20% of appropriately treated CDI cases recur within 8 weeks of treatment, and the recurrence rate increases with subsequent recurrences.^{27, 71} Recurrences may indicate relapse of the same C difficile strain responsible for the primary infection or reinfection in the setting of persistent perturbation of the colonic microbiota.^{28, 72, 73} R-CDI is the best established indication for FMT, with efficacy > 80% now supported by multiple randomized controlled trials. Current CDI treatment guidelines recommend consideration of FMT after the third recurrence.^{3, 27, 31, 40, 41, 42, 43, 44, 45, 74}

FMT also may be considered for cases of refractory CDI or SC-CDI. Refractory CDI, or CDI unresponsive to high-dose oral vancomycin or fidaxomicin, often fits the criteria for SC-CDI (ie, associated with hypotension, fever, ileus, abdominal distention, mental status change, leukocytosis or leukopenia, elevated serum lactate levels, end-organ dysfunction, or requiring intensive care unit admission).³ Randomized clinical trial data are lacking to guide FMT treatment for SC-CDI. Observational studies suggest a possible benefit from FMT, but there is debate as to the optimal FMT dosing regimen in this population; some reports suggest that patients with SC-CDI may benefit from multiple serial FMT procedures (as opposed to a single FMT procedure, the standard for R-CDI) or from continuation of anti-CDI antibiotics concurrent with FMT (as opposed to prior cessation of antibiotics, the standard for R-CDI).^{75, 76, 77, 78, 79} The decision to proceed with FMT in SC-CDI must take into account uncertainty regarding the optimal FMT strategy this population, as well as uncertainty regarding whether the expected benefit matches that of FMT for R-CDI.

In our experience, the alternative diagnosis of ischemic colitis should be considered in patients suspected of having SC-CDI, particularly if C difficile test results are equivocal. Ischemic colitis may manifest with symptoms similar to those of SC-CDI, and interaction between C difficile toxin and colonic ischemia may contribute to a more severe manifestation of CDI.^{80, 81, 82} Colonoscopic examination and biopsy results may not help discriminate well between the two entities.^{83, 84} Although adverse events attributed to FMT have been rare, such events include perforated viscus, sepsis, and death in a patient with documented vascular disease.⁸⁵

Early trials for FMT excluded patients with underlying bowel disease or impaired immune function, but case series suggest that FMT may be administered safely in patients with IBD; HIV; solid organ transplantation; leukemia or lymphoma; and recent exposure to antineoplastic or immunosuppressive agents, including monoclonal antibodies to B and T cells, antitumor necrosis factor, glucocorticoid, antimetabolite, and calcineurin-inhibitor medications.⁸⁶ Neutropenia remains a strict contraindication to FMT in most trials and observational studies and in our clinical practice. Patients with diseases or medications that may increase the theoretical risk of FMT-related infection should be counseled accordingly.

The FMT Procedure: Preparation, Administration, Expectations, and Alternatives

FMT requires healthy stool donors, infrastructure for the processing and storage of donated stool, and a mechanism for introducing stool from the healthy donor into the infected patient’s small intestine or colon. Donor screening, stool processing, storage, and administration practices vary, but some standards have emerged, and ongoing trials will no doubt refine practice further.

FMT stool donors typically are screened for chronic medical conditions according to history and active infections according to serologic and stool testing. FMT screening guidelines published by leadership groups and centers with clinical experience generally concur that potential donors should provide a detailed history and be excluded if they report risk of transmitting an infectious agent (known HIV or viral hepatitis, illicit drug use, unprotected sex, sex with a high risk of HIV or hepatitis C virus acquisition, and so on), GI disease (IBD, IBS, GI malignancy, abdominal surgery, and so on), or other medical conditions or treatments that might impact the gut microbial community composition (obesity, metabolic syndrome, atopic disease, recent antibiotic use, and so on).

FDA guidance recommends its Full-Length Donor History Questionnaire (DHQ version 2.0) as a framework for clinical screening.⁸⁷ The extent of recommended serologic and stool screening is more variable: Some authors favor abbreviated screening if the medical history is reassuring and if the donor is a family member or intimate of the recipient; others favor extensive screening, including standard blood-donor serologic screening and additional stool tests, for all donors.^{22, 47, 88} A report of weight gain after FMT prompted recommendations for more rigorous biometric and serologic screening, including “normal BMI, testing of inflammatory markers, lipids, blood counts, electrolytes, and liver and renal function,” as well as recommendations that healthy, volunteer donors be preferred over familial or intimate donors.^{89, 90} Table 1 summarizes a subset of the screening recommendations published by centers with clinical experience and by advisory groups.^{22, 47, 50, 88, 91, 92, 93, 94, 95}

Table 1.

Recommendations for FMT Donor Screening From Authors With Clinical Experience and From Advisory Panels

Donor Exclusion Criteria	Bakken et al/22 2011	Hamilton et al/47 2012	Health Canada/92 2015	Costello et al/50 2016	OpenBiome/93 2016
Clinical history
Physician concern prompted by health history and lifestyle screen	+	+	+	+	+
Age < 18 or > 50 y	…	…	…	…	+
Age < 18 or > 65 y	…	…	…	+	+
Known HIV	+	+	+	+	+
Known HBV, hepatitis C virus	+	+	+	+	+
Exposure to HIV, viral hepatitis (within 12 mo)	+	+	+	+	+
High-risk sexual behavior	+	+	+	+	+
Use of illicit drugs	+	+	+	+	+
Tattoo or body piercing (within 6 mo)	+	+	+	+	+
Incarceration or history of incarceration	+	+	+	+	+
Known current infection (eg, URI)	+	+	+	+	+
Risk factors for vCJD	+	…	+	+	+
Travel to areas with endemic diarrheal diseases (within 6 mo)	+	+	+	+	+
History of IBD	+	+	+	+	+
History of IBS, idiopathic chronic constipation, or chronic diarrhea	+	+	+	+	+
History of GI malignancy or known polyposis	+	+	+	+	+
Antibiotics within 3 mo	+	+	+	+	+
Antibiotics within 6 mo	…	…	…	…	+
Major immunosuppressive medications	+	+	+	+	+
Systemic antineoplastic agents	+	+	+	+	+
Recent ingestion of a potential allergen (recipient with allergy)	+	…	+	+	+
History of major GI surgery (eg, gastric bypass)	+	+	…	+	+
Metabolic syndrome	+	+	…	+	+
Systemic autoimmunity (MS, SLE, and so on)	+	+	…	+	+
Atopic diseases (asthma and eczema, eosinophilic GI disorders)	+	+	…	+	+
Chronic pain syndromes (fibromyalgia, chronic fatigue)	+	…	…	+	+
Family history of colorectal carcinoma	…	…	…	+	+
Malnutrition (BMI < 18)	…	…	…	+	+
Risk for MDROs, including clinical work	…	…	…	…	+
Stool testing
Clostridium difficile toxin at PCR or EIA	+	+	+	+	+
Routine stool culture enteropathogens	+	+	+	+	+
Shiga toxin EIA	…	…	+	…	+
Fecal Giardia antigen	+	+	a	+	+
Fecal Cryptosporidium antigen	+	+	a	+	+
Acid-fast for Cyclospora, Isospora	+	…	a	+	+
Ova and parasites at microscopy	+	+	+	+	+
Helicobacter pylori fecal antigen	+	…	+	+	+
Rotavirus EIA or PCR	…	…	+	+	+
Norovirus PCR	…	…	+	+	+
Adenovirus EIA or PCR	…	…	+	+	+
Entamoeba histolytica PCR	…	…	+	+	…
Vancomycin-resistant Enterococcus screen	…	…	+	+	+
Extended-spectrum beta-lactamase screen	…	…	…	…	+
Carbapenemase-producing enterobacteriaceae screen	…	…	…	…	+
16S rRNA sequencing	…	…	…	…	+
Serologic and other testing
HIV 1/2 antibody	+	+	+	+	+
HIV p24 antigen	…	…	+	+	+
HAV IgM	+	…	+	+	+
HBV surface antigen	+	+	+	+	+
HBV core antibody (IgG and IgM)	+	+	+	+	+
HBV surface antibody	+	+	+	+	+
HCV antibody	+	+	+	+	+
RPR	+	…	+	+	+
Treponemal antibody	+	…	+	…	+
HTLV 1/2 antibody	…	…	+	+	+
EBV IgG and IgM	…	…	…	+	…
CMV IgG and IgM	…	…	…	+	…
Strongyloides antibody	…	…	+	+	…
E histolytica antibody	…	…	+	+	…
H pylori antibody	…	…	+	+	…
Complete blood cell count	…	…	…	+	+
C-reactive protein	…	…	…	+	…
Erythrocyte sedimentation rate	…	…	…	+	…
Basic metabolic panel	…	…	…	+	…
Liver function tests	…	…	…	+	+
Fasting lipids	…	…	…	+	…
MRSA nasal culture	…	…	+	…	+
Malaria and babesiosis screen	…	…	+	…	…
Gonorrhea and chlamydia screen	…	…	+	…	…
Other guidance
Discretion to limit donor screening for intimates	Yes	No	Yes	NA	NA
Single donor or pooled stool product	Single	Single	Single	Single	Single

The plus sign (+) indicates specifically recommended screening or exclusion criteria. CMV = Cytomegalovirus; EBV = Epstein-Barr virus; EIA = enzyme immunoassay; FMT = fecal microbiota transplantation; HAV = hepatitis A virus; HBV = hepatitis B virus; HCV = hepatitis C virus; HTLV = human T-cell lymphotropic virus; IBD = inflammatory bowel disease; IBS = irritable bowel syndrome; MDRO = multidrug-resistant organism; MRSA = methicillin-resistant Staphylococcus aureus; MS = multiple sclerosis; NA = not applicable; PCR = polymerase chain reaction; rRNA = ribosomal RNA; RPR = rapid plasma reagin; SLE = systemic lupus erythematosus; URI = upper respiratory infection; vCJD = variant Creutzfeldt-Jakob disease

^aHealth Canada guidelines call for “protozoa” screening, but do not specify individual tests.

As standards for donor screening evolve, so do standards for FMT donation processing and storage. The first, to our knowledge, large randomized controlled trial of FMT used freshly donated stool.²⁷ Results from subsequent studies have supported the equivalent efficacy of frozen stool preparation.^{41, 42, 44} Given the imperative for rigorous donor screening and the difficulty of timing donation, most institutions performing regular FMT have developed local banks of frozen stool preparations or work with outside stool banks such as OpenBiome to provide the same product.⁴⁸ Processing donated stool typically involves three steps: (1) homogenization and filtration to remove large particles, (2) concentration of bacteria via centrifugation, and (3) the addition of glycerol to permit frozen storage without bacterial cell lysis, which may impact the viability of bacterial cells and the efficacy of the FMT product.^{47, 96, 97} The production of capsulized FMT for oral delivery also may involve the addition of a lipid emulsion so that capsules remain stable for ingestion.⁹⁸ A single FMT dose typically uses 50 g of fecal material, but a wide range of masses have been used, and one randomized controlled trial used 100 g of fecal material per dose.^{24, 27, 41, 42, 44, 47, 99} It is expected that each gram of stool contains 10¹⁰ to 10¹² bacterial cells.^{100, 101}

The FMT dose can be administered via the upper or lower GI tract. Upper GI delivery methods include oral capsules, nasogastric tubes, nasoduodenal tubes, and esophagogastroduodenoscopy; lower GI delivery methods include colonoscopy, sigmoidoscopy, and enema. Studies directly comparing delivery methods are scarce. Several authors have reported a 5% to 10% greater cure rate with lower GI delivery compared with the rate for upper GI delivery.^{41, 76, 99, 102, 103, 104} However, there are reports of cases refractory to lower GI delivery that responded to FMT via oral capsules,¹⁰⁵ and to date, to our knowledge, only a single, small trial has compared upper vs lower GI delivery directly.⁴¹ Each approach has potential adverse effects: FMT delivered in the upper GI tract may not arrive promptly in the colon if CDI is accompanied by ileus, as is often true in SC-CDI. Delivery of FMT by means of lower GI endoscopy or enema may risk perforation if colonic inflammation is advanced, which also may be the case in SC-CDI.

Just as there is little consensus on the optimal mode of FMT delivery, there is little consensus on how best to prepare the recipient prior to FMT. Bowel lavage is known to reduce the luminal bacterial load significantly,¹⁰⁶ but whether it should be performed prior to FMT is uncertain. Large, randomized controlled trials that support the efficacy of FMT for R-CDI have been performed both with bowel lavage prior to FMT²³ and without.⁴⁴ It is typical to stop anti-CDI antibiotics 24 to 48 hours prior to the FMT procedure, but the impact of this timing on FMT success has not been evaluated rigorously, nor is there robust evidence to support which anti-CDI antibiotic should be used prior to FMT or at which dose. Some practitioners advocate the use of gastric acid suppression (typically a proton pump inhibitor) for 24 hours prior to upper GI FMT,²² though this procedure has been used variably in randomized controlled trials.^{27, 41, 42} Clinical evidence to determine whether bowel lavage and antacid treatment should be given is lacking, and the optimal pre-FMT bowel lavage, antibiotic, and antacid regimen is unknown.

In a broad array of observational studies and trials, as well as in our own experience, FMT cures > 80% of R-CDI cases with a single dose. Repeat dosing typically achieves cure in > 90% of patients with R-CDI.^{3, 27, 31, 40, 41, 42, 43, 44, 74, 107} In several reports and our own experience, the probability of successful FMT is reduced in inpatients, patients with SC-CDI, and patients with concurrent IBD, especially if diarrheal symptoms may be attributable to IBD itself and CDI was diagnosed by means of NAAT results alone.^{108, 109} Although case series results suggest that FMT has a role in the treatment of SC-CDI and CDI with concurrent IBD,^{76, 77, 110} we presently lack adequate trial evidence to predict the probability of cure in patients with SC-CDI or CDI complicating IBD. A recent trial of FMT for secondary prevention of CDI recurrence failed to demonstrate an advantage over prolonged use of oral vancomycin.¹¹¹

It is essential to inform patients seeking FMT for R-CDI not only of the probability of treatment success but also of the symptoms to be expected after the procedure.^{112, 113, 114} In the first 24 hours after FMT, bloating, flatulence, and loose stools are common. These symptoms typically resolve quickly, though some patients complain of bloating and flatulence for weeks. IBS following CDI and after FMT is well documented and may be characterized by lactose intolerance. Diarrhea may recur on isolated days, but patients whose FMTs are successful typically progress to formed stools over 1 to 2 weeks. Early FMT failure is far more common than late FMT failure: A recent cohort study found the rate of early (within 1 month) failure to be 18.6% and of late (1-3 months after FMT) failure to be 2.7%.¹⁰⁹ Failure typically manifests as recurrent diarrhea that progresses over several consecutive days. Such symptoms should prompt repeat stool testing for C difficile. Some practitioners advocate routine post-FMT stool testing, but we have avoided this practice given the uncertain significance of persistent positive NAAT results, as discussed.

The primary alternative to FMT for R-CDI is a prolonged antibiotic taper with pulsed dosing, which commonly is paired with a high-dose probiotic such as kefir.^{71, 115} The monoclonal antibody bezlotoxumab has been shown to reduce the risk of CDI recurrence when administered with standard antibiotic therapy for CDI, but its optimal use in secondary prevention of R-CDI has not yet been established.^{116, 117, 118} The primary alternative to FMT for SC-CDI is surgery, either colectomy or diverting loop ileostomy with colonic lavage.¹¹⁹ The details of these treatments are beyond the scope of this review.

Impact of FMT Beyond CDI

FMT is an effective means for treating CDI, particularly R-CDI. Its use for the treatment of other diseases is an area of active exploration. When advising FMT for the treatment of CDI in patients with other underlying bowel disease, it is essential to consider the potential for off-target FMT effects. In particular, FMT may impact IBD activity.

Whether FMT aids IBD control or provokes IBD exacerbation is an area of active debate, and at present no conclusive statement can be made on the balance between FMT’s benefits and risks in this population. Several studies have been performed to evaluate the efficacy of FMT as a treatment for IBD, and they have produced conflicting results. Reports of FMT benefitting patients with IBD include remission rates as high as 27%,^{120, 121, 122} but reports suggesting benefit have been matched by reports of adverse events that may be attributable to FMT, including IBD exacerbation and microscopic colitis.^{108, 123, 124} The possibility of IBD exacerbation induced by FMT should highlight the importance of discriminating CDI from IBD.⁶⁸

FMT from donors without recent antibiotic exposure also has been observed to reduce colonization by multidrug-resistant organisms (MDROs), and mechanisms by which FMT contributes to MDRO colonization resistance have been proposed.^{125, 126, 127, 128, 129, 130, 131} This off-target effect appears purely beneficial, but its impact on subsequent risk for MDRO infection has not been studied thoroughly. Until the impact of FMT on the risk for subsequent MDRO infection is better characterized, this potential benefit should not sway the decision on whether to proceed with FMT.

Infrastructure for Clinical FMT: Oversight, Billing, and Anticipated Changes

The greatest impediments to the broad dissemination of FMT for the treatment of R-CDI are uncertainty surrounding regulation by the FDA and the difficulty of obtaining reimbursement. Present regulation is not unduly onerous for FMT practitioners, but the FDA is reviewing its FMT guidance. The FDA considers FMT an investigational drug but declared in 2013 that it would exercise “enforcement discretion” and not require an Investigational New Drug (IND) application from physicians who perform FMT for the treatment of CDI that does not respond to anti-CDI antibiotics.⁵² The consideration of FMT as a drug places it under the “Guidance for Industry: Enforcement Policy Regarding Investigational New Drugs”¹³² rather than the regulations that the FDA uses for the oversight of blood transfusions (including donor screening, blood testing, donor deferral lists, quarantine, and regulations to address deficiencies), which might be more appropriate.

An IND application is required if FMT is to be performed for any indication other than CDI, and such treatment should be performed only in the setting of a clinical trial.^{39, 40} In March 2016, the FDA announced an intention to revise its FMT oversight, with specific consideration given to IND application requirements for large stool banks.¹³² The revised policy has not yet been finalized.

Cost-effectiveness analyses have demonstrated substantial cost savings and increased efficacy of FMT for R-CDI, relative to those for oral vancomycin therapy.^{133, 134, 135} Despite these data, in our experience, insurance reimbursement for FMT is not reliable, and reimbursement rates rarely cover the substantial cost of donor screening or purchase of the FMT dose from a stool bank. If FMT donation processing and dose administration are performed by the FMT practitioner, two Current Procedural Terminology codes can be billed: 44705 for the preparation of fecal microbiota for instillation (alternative for Medicare: G0455) and 44799 for the FMT instillation via naso- or oroenteric tube or enema. Esophagogastroduodenoscopy and colonoscopy should be billed as they would be normally. If the FMT material is obtained from a stool bank, it is recommended that code J3590 (“unclassified biologics”) be used in lieu of 44705 and 44799, but no reimbursement is to be expected from this code.¹³⁶

Conclusions

We have reviewed the clinical application of FMT for the treatment of CDI, paying particular attention to the care that must be taken in determining eligibility for FMT, the (evolving) present standards for FMT preparation and administration, the uncertainty of FMT regulation and difficulty of FMT billing, as well as emerging data on FMT effects beyond CDI. FMT has revolutionized the treatment of R-CDI, but much work remains to be done in defining the optimal FMT dose, as well as optimal protocols for donor screening, recipient preparation, and post-FMT monitoring. We remain in the early stages of establishing FMT’s role in the treatment of SC-CDI, early CDI, and other bowel diseases. Given the benefits that FMT has demonstrated already, it is imperative that work to improve FMT efficacy and extend its benefits proceed rapidly.