What FMT is, and how thin the canine evidence base still is.
Faecal microbiota transplantation moves a screened donor's faecal community into a recipient. In dogs it is most often given as an enema, which can be done in a consulting room without general anaesthesia, though it has also been delivered by duodenoscope, nasogastric tube and oral capsule.
The most useful review of the canine evidence, by Chaitman and Gaschen in 2020, is blunt about how much there is. The published record consists of results from five clinical trials, a few case series and several case reports, drawn from peer-reviewed journals, conference abstracts and a university thesis. A footnote to their summary table states plainly that none of the comparative studies was blinded.
That is the honest frame. What follows is a small, unblinded, mostly uncontrolled literature with some genuinely interesting results in it.
Where it works best: acute diarrhoea.
The strongest canine result compares a single FMT by enema against a seven-day course of oral metronidazole at 15 mg/kg twice daily, in adult dogs with non-infectious acute diarrhoea of less than 14 days. Eleven dogs received FMT, seven received metronidazole, and 14 healthy dogs served as controls.
Faecal consistency improved equally in both groups by day 7. By day 28 the FMT dogs had firmer faeces than the antibiotic group. The difference underneath that was larger: the dysbiosis index normalised at day 7 and remained within the reference range at day 28 in most FMT dogs, and did not normalise in most dogs given metronidazole. Clostridium hiranonis and Faecalibacterium rose and E. coli fell only in the FMT group, and the raised proportion of primary bile acids seen at presentation normalised after FMT only.
The authors' reading is worth quoting for what it implies about antibiotic stewardship rather than about FMT: a single FMT should be considered instead of antibiotic therapy in order to prevent the effects of antibiotics on the intestinal microbiome and to decrease overall antibiotic use. The study was not randomised and not blinded, and 11 against 7 is a small comparison.
Parvovirus: faster, not safer.
In puppies with parvovirus, 33 received standard treatment and 33 received standard treatment plus FMT. Survival did not significantly improve. What did improve was speed. Diarrhoea resolved within 48 hours in 61.5 per cent of FMT puppies against 4.5 per cent on standard treatment alone, and median hospitalisation fell from 6 days, range 2 to 15, to 3 days, range 1 to 6.
One confounder is stated by the authors and matters: the FMT puppies were significantly older than the controls, four months against three, and better survival might be expected in older puppies anyway.
Where it does not work, which is most of the rest.
Two randomised canine trials in this literature found nothing.
Daily oral FMT given to puppies for five days before weaning, using their own dam's faeces, did not prevent post-weaning diarrhoea. Eleven puppies received FMT and 12 received sham treatment, with no difference in faecal consistency. In acute haemorrhagic diarrhoea syndrome, four dogs received a single FMT and four received saline, with no difference in clinical score.
Chronic enteropathy is the indication owners ask about most and the one with the weakest support. The review states directly that the available evidence is weaker in dogs with chronic enteropathies. A case series of 16 dogs with refractory inflammatory bowel disease of more than a year's duration reported apparent improvement in clinical activity scores, but the authors conclude that the confounding effects of the different clinical presentations and the other treatments given make interpretation of the impact of FMT impossible.
The practical pattern the reviewers describe from their own caseload is the important one: dogs with chronic diarrhoea may improve for a few days to a week after FMT, but generally relapse thereafter, so multiple transplants are needed in most situations. That is a maintenance therapy, not a cure, and it is described as unpublished clinical observation rather than trial data.
The donor screen, which is the part that carries the risk.
In human medicine, two patients have died after receiving transplants containing a multidrug-resistant organism, following bacteraemia with drug-resistant E. coli traced to the transplant. The reviewers draw the obvious conclusion, which is that diligent donor screening is essential.
Canine screening protocols vary between published studies. The criteria Chaitman and Gaschen propose are the most complete published set, and they are worth reproducing because an owner offered FMT should be able to ask which of these were done.
History and physical examination
- Preferably between 1 and 10 years of age
- Preferably no travel history outside the local area
- No health issues in the last 6 or 12 months
- No history of chronic gastrointestinal disease, allergies or immune-mediated disease
- No antibiotics in the last 12 months
- Regularly vaccinated according to existing guidelines
- Fed a balanced diet
- Not overweight or underweight, body condition score 4 to 6 of 9
- Normal faecal consistency, and healthy on physical examination
Laboratory screening
- Normal complete blood count and serum biochemistry
- Consider basal cortisol and thyroxine
- Negative for parasite ova on faecal flotation, with empirical broad-spectrum deworming to consider
- Negative for Giardia on faecal flotation and ELISA
- Consider testing for faecal pathogens such as Salmonella spp. and Campylobacter spp.
Faecal microbiome evaluation
- Dysbiosis index below 0
Note the last one. A dysbiosis index is not being used here to diagnose the donor. It is being used to confirm that the material being transplanted is worth transplanting, which is one of the few uses of that number with a clear rationale behind it.
For comparison, the review quotes a commercial canine faecal bank operating a less stringent standard: dogs that go outdoors regularly, have had no antibiotic treatment in the prior six months rather than twelve, are not overweight and have diverse microbiomes, with faeces screened for C. difficile toxin B, Cryptosporidium, Salmonella, Giardia and canine parvovirus 2. Six months against twelve is a real difference in an era of antimicrobial resistance.
One gap is worth naming. Screening donors for multidrug-resistant bacteria is recommended in the human protocol quoted in this review and does not appear in the canine criteria. Given that MDR transmission is what killed the two human patients, that is a question worth asking any provider.
Practical points from the protocols.
Doses in the two authors' own protocols differ by more than twofold, at 2.5 to 5 g of donor faeces per kg of recipient bodyweight in one and 1 to 2 g/kg in the other, both given rectally by red rubber catheter. Recipients are not fed and are kept quiet for four to six hours afterwards to reduce the chance of defecating the transplant back out.
Storage matters more than people expect. Bacterial viability in homogenised canine faeces held up for a week refrigerated at 4 degrees. At three months, adequate viability survived only in frozen samples with 10 per cent glycerol added, and at six months only at minus 80 degrees with glycerol.
FMT is not recommended for patients receiving antibiotics at the same time, since the antibiotic is likely to undo it.
What is still missing.
No adverse event in a dog is reported anywhere in this review, which is reassuring and also a limitation: absence of reported harm in a literature this small is not the same as demonstrated safety. The review makes no statement about long-term safety in dogs either way, and none about regulatory status in any jurisdiction.
There are no canine studies comparing routes of administration, and the reviewers say so directly. The one canine head-to-head, an abstract comparing enema against oral dosing in healthy dogs given tylosin, found no difference in faecal score between either form and controls.
The authors' own conclusion is that more data are needed to define which canine patients could be helped by FMT. That remains the state of it.
Where BAARK fits.
BAARK does not perform, supply or arrange faecal microbiota transplantation.
FMT is a veterinary procedure and the decision to use it, on which animal and from which donor, belongs entirely with a treating veterinarian. Where a vet wants a picture of the recipient's community before and after, a faecal profile can provide one, and the published canine work uses exactly that approach to show what changed.
BAARK is wellness profiling, not a diagnostic test. It does not detect, diagnose, stage or predict the progression of any disease, and nothing here is a recommendation to seek or avoid FMT for any individual animal.
The bottom line.
On the published canine evidence, a single FMT beat a week of metronidazole for acute diarrhoea at 28 days and left the microbiota in better shape, and it shortened hospital stays in parvovirus without improving survival. It failed to prevent post-weaning diarrhoea, failed in acute haemorrhagic diarrhoea syndrome, and in chronic enteropathy tends to buy days to a week before relapse. None of the comparative studies was blinded, and the donor screen is the part of the procedure with a documented human death behind it. Ask which screening criteria were used, and ask about the antibiotic washout.