Fecal microbiota transplantation for GI disorders

Introduction

Fecal microbiota transplantation (FMT) is a technique that involves transferring the intestinal microbiome from a healthy donor to a diseased recipient in order to improve the latter’s microbiome and decrease disease severity. Although the technique was actually mentioned in a Chinese textbook of emergency medicine in 320 A.D, it has rarely been used in traditional medicine until the start of this century, when knowledge of the intestinal microbiome and dysbiosis expanded substantially. In humans, gastrointestinal (GI) conditions are by far the most common reason to perform FMT, but multiple studies have been performed using the technique for other indications, including hepatic disorders, metabolic syndrome, treatment of antibiotic-resistant microbes, psychiatric disorders and obesity 1,2. In animals, FMT has been proven to have a beneficial effect in puppies with parvovirus enteritis 3 and also appears to be promising for dogs with chronic diarrhea 4,5, but to date there is only one feline case report available 6. There are currently no evidence-based guidelines or any consensus on donor screening, FMT dosing or the best protocol to follow for animals, but a recently formed group of international experts, the Companion Animal Fecal Bank Consortium, is working on such guidelines, with preliminary results due this year. Despite the lack of consensus, FMT is regarded as a fairly safe treatment for dogs with acute or chronic GI disorders, and has the potential to decrease disease severity in many cases. This paper will review various reports on the use of FMT in dogs with GI disorders, present a description of the procedure, and discuss some clinical cases. 

FMT for GI disorders

As noted above, the beneficial effect of FMT has been shown in various studies. One study looked at parvovirus enteritis in puppies 3, where 66 animals with parvovirus at two veterinary hospitals were treated with either “standard” measures alone or standard treatments plus FMT in a randomized controlled trial. FMT significantly reduced hospitalization periods and the time to recovery (median time 3 days versus 6 days in the control group), and survival was higher in the dogs treated with FMT (26/33, 79%) compared to the other group (21/33, 64%), but the difference was not statistically significant. In another study of 18 dogs with acute diarrhea, a single FMT at presentation improved fecal scores to the same extent as metronidazole treatment, as evaluated at day 7, and at day 28 the dogs treated with FMT had significantly better fecal consistency compared to the metronidazole treated group 7. Furthermore, FMT helped to restore the intestinal microbiome of the first group to healthy levels at day 28, whereas the metronidazole treated dogs had significant dysbiosis at this time point, when compared to both dogs treated with FMT and healthy dogs. However, in a small placebo-controlled pilot study of 8 dogs with acute hemorrhagic diarrhea, no clinical benefit was seen in dogs given FMT compared to sham-treated controls 8.

With respect to dogs with chronic diarrhea and/or chronic enteropathy, one case report and one case series on successful FMT treatment have been published, as well as two scientific abstracts 4,5,9,10. In the case series, 9 dogs with refractory inflammatory bowel disease (IBD) which were unresponsive to food trials, antibiotics, corticosteroids or cyclosporine, were included 4. A significant decrease in the canine inflammatory bowel disease activity index (CIBDAI 11 – Box 1)) post-FMT was seen in all dogs, as well as a significant increase in fecal Fusobacterium spp. 7/9 dogs had lower fecal concentrations of Fusobacterium compared to the donor dogs prior to FMT. Fusobacterium is a major producer of short chain fatty acids (SCFA) and an important component of a healthy canine intestinal microbiome, whilst dysbiosis and decreased levels of SCFA-producing intestinal microbes are very common in chronic canine enteropathies (Box 2) 12. Dysbiosis was also present in a study of 16 dogs with chronic diarrhea which were each given one FMT, with a significant improvement in the fecal dysbiosis index* reported one week after treatment 10. The retrospective study behind the other abstracts 5,9 is discussed in more detail in the next section.

Box 1. The CIBDAI scoring system. Six parameters are scored, each from 0-3, where 0 = normal, 1 = mild changes, 2 = moderate changes, and 3 = severe changes. The scores are added together to give the CIBDAI.

Box 2. What are SCFAs?

Very limited information is available on the use of FMT in cats (Figure 1); there is currently only one case report, which was a cat with non-responsive ulcerative colitis that responded to FMT 6.

FMT for poorly responsive chronic enteropathies

The efficacy of FMT on chronic enteropathy (CE) cases was demonstrated by the following study. This involved a retrospective data review from a cohort of 36 dogs (aged 0.6-13 years, median 6.3) with CE that had FMT as adjunctive therapy in the author’s hospital between 2019-2021 5. All dogs had shown either poor or no response to standard evidence-based treatments, and a follow-up period of at least 3 months post-FMT was required for inclusion. Exclusion criteria were (i) if any current maintenance therapy dose was increased during the period reviewed (ii) intestinal parasites, or (iii) starting a new immunosuppressive treatment or diet in parallel with FMT. FMT was given using a standardized protocol to all dogs, using two different donor dogs, both of which had a dysbiosis index* below -2 (normobiosis) 12. 

All 36 dogs had been treated for CE for between 1-110 months (median 21) at inclusion, with the main complaints being refractory diarrhea (28/36), lethargy (15/36) and various side effects of medication (10/36). 34/36 dogs were treated with corticosteroids at inclusion, and 20/36 received second line immunosuppressive drugs, including mycophenolate, chlorambucil, cyclosporine or azathioprine. 26/36 dogs were fed a hydrolyzed diet, 8/36 a single protein diet, and 2 were fed a highly digestible “intestinal” diet.

34 dogs received between 2 and 5 FMTs, with the majority (26 dogs) receiving 3 treatments. The other 2 dogs, both non-responders, had received one FMT each. Clinical improvement based on CIBDAI was noted in 75% of dogs (27/36) after treatment, with the most common improvements being increased activity level (20/36), improved fecal scores (19/36) and weight gain and/or increased appetite (10/36). This latter group had previously shown a poor appetite and/or subnormal body condition scores. The maintenance corticosteroid dose could be tapered in 6 of the dogs to a level lower than what had been possible prior to FMT. One dog which had previously developed frequent flare-ups of diarrhea that only responded to tylosin did not need antibiotics for 21 months after the third FMT (case 2# in the next section), and another dog that was previously treated with metronidazole and immunomodulatory drugs could stop the metronidazole after FMT.

CIBDAI at inclusion was 2-17 (median 6) and this decreased significantly to 1-9 (median 2) during the first month following the last FMT. Fecal samples for analysis of the dysbiosis index* (reference interval ≤ 0) were available from 23 dogs at inclusion. Dogs unresponsive to FMT had a significantly higher result compared to good responders at inclusion. A high dysbiosis index has been previously shown to correlate with decreased microbial diversity with fewer bacterial taxa present (In people, low microbial diversity prior to FMT is a negative prognostic factor for responding to FMT 13). Side effects were mild and uncommon; 6/36 dogs (3 responders and 3 non-responders) had diarrhea within 48 hours post FMT, with two of these dogs also showing clinical signs of abdominal or rectal pain within 24 hours after FMT. All side effects were, however, self-limiting.

This study has, however, several limitations. It is a retrospective study, the microbiome and metabolome were not followed over time, and there was no control group included. Nevertheless, results suggest that FMT can be used as adjunctive therapy in dogs with poorly responsive CE.

FMT procedure

As previously mentioned, there is currently no consensus or evidence-based guidelines on donor screening or the best protocol for FMT 14. The following recommendations are based on the author´s personal clinical experience and recent studies 5,7. 

Donor screening

A donor animal should be an individual that is clinically healthy, with a normal body condition score and a CIBDAI score of 0-3 (i.e., no clinical signs of a chronic GI disease) 11; essentially the aim is to find a donor with plenty of beneficial microbes and no potential fecal pathogens. In addition, the animal should not be fed a raw food diet, should not be receiving any long-term medication, and must not have had any antibiotics for at least 6 months, preferably longer. For feline donors, indoor-only cats are preferred in order to avoid exposure to parasites from small rodents etc. Intestinal parasites, including Giardia intestinalis, should be excluded from all potential donors. To ensure high levels of beneficial microbes, such as short-chain fatty acid (SCFA)-producing bacteria and Clostridium hiranonis, potential donors should be screened with the canine or feline dysbiosis index* 12. The fecal canine donors at the author’s hospital are also free from Salmonella spp., Campylobacter jejuni, Clostridioides difficile and Clostridium perfringens enterotoxinogen, including Clostridium perfringens netF-toxin. However, such extensive screening of donors may not be necessary – it is likely most important to exclude intestinal parasites and ensure high levels of beneficial microbes, as microbial composition and diversity of the donor transplant is vital for successful treatment of ulcerative colitis in humans 13. Furthermore, this study reported that recipients with a good response to FMT had increased fecal microbial diversity both before and after FMT compared to non-responders, as well as increased fecal levels of SCFAs and secondary bile acids post FMT.

FMT dosing and procedure

The amount of feces used for FMT in dogs can vary considerably 14. The author currently uses 5 g of donor feces per kg bodyweight of the recipient for dogs up to 30 kg and cats; for recipient dogs over 30 kg, 2-3 g of feces per kg bodyweight is used. This is a relatively large amount, but has been associated with a good outcome in the majority of dogs with CE 5. Food should be withheld from the recipient for 6 hours prior to FMT, but water is allowed, and the recipient dog should be walked for 30-40 minutes just prior to the procedure in order to defecate. A low dose of acepromazine (0.1 mg/kg SC) can be given 15 minutes beforehand unless contraindicated; although some clinicians omit this if the recipient is calm, premedication usually makes it easier for the dog to relax and rest after the procedure, allowing a long contact time between the transplant and the colonic mucosa. In the author’s experience, cats need to be fully sedated prior to FMT.

The fecal transplant can be delivered via the upper or lower GI tract. In people, the route of administration does not appear to be outcome-related for GI indications (recurrent Clostridioides difficile infection, ulcerative colitis and Crohn’s disease) 15,16,17, but in published reports on FMT in dogs, the rectal route is by far the more commonly used delivery method, using a retention enema or colonoscopy.

Fresh or frozen feces can be employed; if the latter, it should first be thawed overnight in a fridge. (In people with recurrent or refractory Clostridioides difficile-infection, FMT using frozen feces has been shown to be as effective as fresh material 18). The feces should be blended and mixed with sterile saline (20-120 mL) until a desirable texture is achieved, before being filtered through a sieve. The filtrate is then aspirated into 60 mL sterile syringe(s) and either left at room temperature, or warmed to body temperature in a water bath prior to use, since it is highly unpleasant for the recipient to receive a large volume directly from the fridge. The transplant is administered rectally using a 12-16 FG catheter 7. The catheter should be well lubricated before insertion, with the tip placed approximately at the level of the last rib (Figure 2). FMT can be given with the dog in either a standing position, sternal or lateral recumbency (Figure 3). The owner is then instructed to minimize the dog´s physical exercise for 4-6 hours in order to increase the contact time between the intestinal mucosa and the transplanted feces. Food should also be withheld for the same period, since the presence of food in the stomach stimulates colonic contractions. At the author´s hospital, the standard protocol (Box 3) is for dogs with CE to receive a series of three FMT’s, with 10-20 days interval between each, as experience has shown that one treatment is often ineffective in reducing clinical signs in many dogs, or not lasting long enough. However, if no beneficial effects have been noticed after two treatments, a third FMT is not given 5.

Box 3. The author’s preferred FMT protocol.

Case 1# – “Alma”

Alma (Figure 4) is a spayed female Golden Retriever who developed a steroid-responsive CE at 3 years of age. At the age of 5 she was on a maintenance dose of oral methylprednisolone (0.4 mg/kg EOD) and a hydrolyzed, soy-based diet. This controlled the clinical signs of CE to some extent, but she was still suffering from lethargy, signs of abdominal pain, occasional vomiting, diarrhea and a subnormal body condition score (BCS) of 3.5/9 (15% below ideal BCS), with mild to moderate muscle atrophy. Several attempts were made to decrease the methylprednisolone dose, but each time the clinical signs worsened. Treatment with mycophenolate was started as an add-on immunosuppressive, but this did not allow any reduction in the methylprednisolone dose. The owner agreed to try FMT as adjunctive therapy, and three separate FMTs were given as a rectal retention enema with 10-14 days intervals. Alma showed a very positive and rapid clinical response; she was much more active and alert, played more with other dogs, and gained 2 kg in weight, allowing a gradual tapering of methylprednisolone to 0.2 mg/kg EOD. Fecal analysis showed that Alma had a dysbiosis index* of -1.2 (normobiosis) at baseline, but she had marked alterations of fecal lipid profiles, such as sterols and fatty acids, and the most striking abnormality was a fecal coprostanol concentration 24 times that of a normal dog. Cholesterol in the gut lumen is metabolized to coprostanol by intestinal microbes, and this compound is poorly absorbed from the gut 19, so Alma had an exaggerated conversion of cholesterol to coprostanol. Two weeks after FMT 1, the fecal lipid profile was normalized, which correlated with normalization of her BCS. The positive effects of FMT lasted for 7 months, but then Alma again became lethargic and lost weight; however, a second series of FMT and a temporarily increased dose of methylprednisolone reversed the clinical signs. 

Case 2# – “Moltas”

Moltas is an intact male German Shepherd that has suffered from chronic, partially refractory diarrhea all his life. He also suffers from atopic dermatitis, recurrent pyoderma and chronic otitis. At 1.5 years of age, he was clinically fairly stable on high daily doses of prednisolone, but he had a BCS of 3/9 and tapering of the prednisolone led to worsening of clinical signs. Azathioprine had no effect, and multiple dietary trials, including a highly digestible diet and two different single protein diets, were unsuccessful. During the worst flare ups of diarrhea, Moltas did respond to tylosin or metronidazole, and at this point he was referred to the author. He was started on a hydrolyzed novel protein diet and cyclosporine, which had some effect, allowing for some tapering of the prednisolone. At 2.5 years of age the cyclosporine was replaced with chlorambucil, which led to a clinical improvement and weight gain to a normal BCS. During chlorambucil treatment, prednisolone could be replaced with 3 mg budesonide every other day (EOD), which has fewer side effects. Moltas was also treated with allergen-specific immunotherapy, twice weekly medical baths with chlorhexidine, and 4 mg methylprednisolone EOD as a maintenance dose for his skin condition. During the following 2.5 years Moltas was relatively stable, but had flare-ups of diarrhea every few months. Minor exacerbations could be controlled with a temporary increase in the dose of budesonide (3 mg daily for 3-10 days). More severe flare-ups occurred roughly every six months, and these did not respond to immunosuppression, so Moltas was prescribed tylosin (25 mg/kg q24h for 7 days). At 5 years of age, the GI signs had increased, such that there were monthly flare-ups of diarrhea, regurgitation and lethargy. This increased disease activity had prompted increased polypharmacy, with more frequent use of tylosin alongside budesonide (3 mg EOD), methylprednisolone (4 mg EOD), chlorambucil (3 mg EOD) and cobalamin (1 mg orally once weekly). 

On clinical examination, marked abdominal pain was obvious on palpation. Serum biochemistry revealed a mild hypoalbuminemia (28 g/L; reference interval 30-45 g/L) and mild-moderate decrease in total protein (51 g/L; reference interval 61-75 g/L). These parameters had been within the reference range at the last check-up six months previously. Serum cobalamin concentrations had also dropped significantly to 221 pmol/L (reference range 180-708 pmol/L), despite weekly maintenance therapy. Fecal samples were negative for intestinal parasites.

Moltas was treated with 1 mg of cobalamin EOD and three FMTs via rectal retention enemas at 14-day intervals. After FMT 1, the regurgitation episodes stopped, and after FMT 2 the fecal quality improved and Moltas became more playful and active (Figure 5). After the third FMT, diarrhea had stopped and abdominal palpation did not induce signs of pain. Furthermore, serum albumin and total protein concentrations had increased and were back within the reference range. During the next 21 months, Moltas was much more stable, although there were mild flare-ups of diarrhea every third month, which lasted for 1-2 days and were self-limiting. After 21 months, the fecal quality became progressively worse, and a severe flare-up occurred. Increasing the dose of corticosteroids had only a limited effect, and Moltas was again treated with tylosin for a week, followed by a second series of 3 FMTs, which had the same positive effect as the first treatment.

Case 3# – ”Harold”

Harold is an intact male French Bulldog (Figure 6) who had a persistent Giardia intestinalis infection as a puppy and young dog. The infection finally cleared, but diarrhea, melena and weight loss continued. The referring vet had treated Harold with metronidazole and corticosteroids, which only led to marginal improvement, and full-thickness surgical biopsies from the small intestine and the colon were taken at a year of age. The histopathologic diagnosis was granulomatous colitis and moderate lymphocytic-plasmacytic enteritis with moderate lacteal dilation. Sulfasalazine was added to the treatment without any effect, so Harold was referred to the GI-service at the author’s hospital at the age of 1.5 years. At this point he was slightly lethargic and had a BCS of 3/9. He was started on a 6-week course of enrofloxacin for granulomatous colitis, which quickly led to resolution of clinical signs, including weight gain. At a check-up just after the treatment finished, Harold was asymptomatic and had a BCS of 4/9. However, 3 weeks later, diarrhea (of predominantly colitis type) and vomiting recurred. As no colonic biopsies had been sent for culture and sensitivity testing when the biopsies were collected, it was unknown if Harold was already harboring multidrug-resistant E. coli prior to enrofloxacin treatment. Since resistance to fluoroquinolones develops rapidly during treatment, it was very likely that multidrug-resistant E. coli was now part of his intestinal microbiome 20. In Boxer dogs with granulomatous colitis it has been shown that the presence of fluoroquinolone-resistant E. coli is associated with failure to respond fully to enrofloxacin treatment, as well as concurrent antimicrobial resistance to chloramphenicol, rifampicin, and trimethoprim-sulfa 20, and multidrug-resistance and treatment failure often leads to euthanasia in affected dogs. Carbapenem has been reported as an alternative antibiotic in dogs with granulomatous colitis and fluoroquinolone-resistant E. coli 21, but it is a critically important class of antibiotics in human medicine and is prohibited for veterinary use in many countries. 

At this time point, the owner agreed to try FMT. The first procedure was followed by 2-3 days of flatulence, smelly feces and mild vomiting, and although the fecal quality then improved slightly, diarrhea recurred after 14 days. The second FMT 16 days after the first one was again followed by 2-3 days of similar signs, but this time the subsequent improvement of the fecal quality was more pronounced. Harold was also started on a multi-strain probiotic at this point. After FMT 3, no side effects occurred, the stool was normal and Harold was much more active and alert. He continued on the multi-strain probiotic every other day along with a hydrolyzed protein diet, and at the latest check-up (14 months after FMT 3), he was still in complete remission.

Case 4# – Ina

Ina is an intact female German Shepherd who had shown signs of CE since she was a year old, although these had responded to a hydrolyzed protein diet combined with a multi-strain probiotic. At 2 years of age she developed a urinary tract infection that was treated with (unknown) antibiotics by her local veterinary clinic. After the antibiotics, Ina became markedly flatulent, lethargic and hyporectic, clinical signs similar to those she presented with during the initial work-up for CE. Intestinal dysbiosis following antibiotic treatment was suspected, and analysis of a fecal sample showed a dysbiosis index* of 6.2 (Figure 7), consistent with severe dysbiosis. Ina was still lethargic and hyporectic six weeks after the antibiotics had finished, and a FMT series was scheduled. After FMT 1, Ina improved but relapsed before FMT 2; however, after two further FMT treatments she was again extremely alert with a normal appetite, and the dysbiosis index went from severe to mild classification after FMT 1, followed by normobiosis after FMT 2 (Figure 7). 

Conclusion

Fecal microbiota transplantation (FMT) is a promising treatment in companion animal gastroenterology, with published studies reporting very few unwanted side effects. At present, FMT dosage and protocol will vary somewhat among small animal clinicians, but a consensus on treatment guidelines is pending. FMT can be used in various cases, including puppies with parvovirus infection, and appears to be beneficial for the treatment of many dogs with poorly responsive chronic enteropathies. Treatment with FMT may also allow reduction in the use of antibiotics in selected cases. 

*The Dysbiosis Index is provided by the GI Laboratory at Texas A&M University, USA