Suez J, Zmora N, Zilberman-Schapira G, et al. Post-antibiotic gut mucosal microbiome reconstitution is impaired by probiotics and improved by autologous FMT. Cell. 2018;174(6):1406-1423.e16.
Murine and human longitudinal intervention cohort research study
To assess the impact of probiotics and autologous fecal transplant (aFMT) on the microbiome profiles of both humans and mice after antibiotic treatment.
Healthy human volunteers (N=21) and male wild-type (WT) C57BL/6 mice (N=50)
Human volunteers met the following criteria: men and women, aged 18-70, who were currently not following any diet regimen or dietitian consultation and were able to provide informed consent.
Exclusion criteria included: 1) pregnancy or fertility treatments; 2) usage of antibiotics or antifungals within 3 months prior to participation; 3) consumption of probiotics in any form within 1 month prior to participation; 4) chronically active inflammatory or neoplastic disease in the 3 years prior to enrollment; 5) chronic gastrointestinal disorder, including inflammatory bowel disease (IBD) and celiac disease; 6) active neuropsychiatric disorder; 7) myocardial infarction or cerebrovascular accident in the 6 months prior to participation; 8) coagulation disorders; 9) chronic immunosuppressive medication usage; 10) pre-diagnosed type I or type II diabetes mellitus or treatment with antidiabetic medication. Fulfillment of inclusion and exclusion criteria was validated by medical doctors.
Study Parameters Assessed
The authors collected microbial profile information from both human and murine participants at baseline, after broad-spectrum antibiotic exposure, and during/after the intervention for 3 study arms: autologous FMT (aFMT), probiotics, and spontaneous recovery.
Primary Outcomes Measured
The study utilized a variety of both RNA and DNA microbial sequencing techniques to provide microbiome profiles that could identify large shifts in microbial diversity and analyze species-specific details. This analysis was done on microbiome samples collected throughout the gastrointestinal (GI) tract at various timepoints within the various arms of the study outlined below.
The drinking water of adult mice was treated with a broad-spectrum antibiotic regimen of ciprofloxacin and metronidazole for 2 weeks. The immediate impact of antibiotic treatment on gut mucosal microbiome configuration was assessed in 1 group of mice sacrificed after the 2-week antibiotic exposure (n=10).
- The remaining animals (n=30) were divided into 3 post-antibiotic intervention groups.
- In the first group (‘‘probiotics’’), antibiotic treatment was followed by 4 weeks of daily administration by oral gavage of a commercially prescribed probiotics product involving 11 strains that was validated for composition and viability by multiple methods: Lactobacillus acidophilus (LAC), L casei (LCA), L casei sbsp paracasei (LPA), L plantarum (LPL), L rhamnosus (LRH), Bifidobacterium longum (BLO), B bifidum (BBI), B breve (BBR), B longum sbsp infantis (BIN), L lactis (LLA), and Streptococcus thermophilus (STH).
- Each mouse of the second group (‘‘aFMT’’) received, on the day following cessation of antibiotics, an oral gavage of its own pre-antibiotics stool microbiome.
- A third group (‘‘spontaneous’’) remained untreated following antibiotic therapy to assess the spontaneous recovery of the indigenous gut microbiome in this setting.
- An additional group of mice (‘‘control’’) did not receive antibiotics or any other treatment and was followed throughout the study’s duration (n=10).
Sample collection was done in mice after they were sacrificed, and the digestive tract was dissected into 8 parts:
- Stomach beginning at the pylorus
- Proximal 4 cm of the small intestine, collected as the duodenum
- Proximal jejunum
- Distal jejunum
- Ileum, harvested as the distal third of the small intestine
- Colon proximal
- Colon distal
For each section, the content within the cavity was extracted and collected for luminal microbiome isolation, and the remaining tissue was rinsed 3 times with sterile phosphate-buffered saline and collected for mucosal microbiome isolation. A total of 710 fecal, 680 luminal, and 680 mucosal samples were analyzed.
Two cohorts of healthy volunteers were recruited, a naive cohort (n=25) and an antibiotics-treated cohort (n=21), subdivided into 3 interventions:
- Probiotics (n=8)
- aFMT (n=6)
- Spontaneous reconstitution (n=7)
For the antibiotics-treated cohort, volunteers consumed oral ciprofloxacin 500 mg (Ciprodex, Dexcel Pharma) twice a day and oral metronidazole 500 mg (Flagyl, Sanofi) 3 times a day for a period of 7 days.
- Stool samples were collected daily from humans during the baseline and antibiotics phases, daily during the first week of intervention, weekly throughout the rest of the intervention, and twice a month and monthly during the follow-up phase.
- Two endoscopic exams were performed on each participant after completion of the week-long antibiotic course, and 3 weeks later (day 21) luminal content was aspirated from the stomach, duodenum, jejunum, terminal ileum, cecum, and descending colon.
- Antibiotic treatment mildly enhanced probiotic gut mucosal colonization in mice.
- Probiotics delay and aFMT improved the post-antibiotic reconstitution of the indigenous murine microbiome.
- Probiotics in antibiotics-perturbed humans were continuously shed in stool and colonized the lower GI mucosa.
- Probiotics delayed, while aFMT improved, the post-antibiotic reconstitution of the indigenous human fecal microbiome.
- Probiotics delayed the post-antibiotic reconstitution of the indigenous human mucosal microbiome.
- Reversion of antibiotics-associated GI transcriptomic landscape was delayed by probiotics.
- Probiotics-secreted molecules inhibited human microbiome in vitro growth.
Although this study analyzes a large collection of robust data, it still creates more questions than answers. The study showed that antibiotics greatly diminish microbial diversity, something that has been well-established in the literature.1 Additionally, it revealed that aFMT causes a quick and comprehensive return to homeostatic microbial configuration after antibiotic exposure. And, interestingly, it revealed that SupHerb Bio-25, an 11-strain probiotic supplement, given twice a day following antibiotic exposure causes a significant delay in the return to homeostatic configuration when compared to both no probiotic treatment or aFMT. More on this later; first let’s discuss FMT.
What is FMT?
Fecal microbiota transplant, or FMT, involves delivering stool collected from a healthy donor to the GI tract of a recipient. This can be done via oral capsule, retention enema, infusion through upper or lower endoscopy, or via nasogastric tube. The donated stool is often processed via blending, straining, and centrifugation (if orally encapsulated), which allows for a small fraction of that stool to be administered.2 We like to explain it as a super-probiotic that includes all the “compost” of the healthy ecosystem from the stool of the donor. As this study highlights, this technique of reconstituting the human microbiome is piquing the interest of the medical community.
This study indicates that aFMT from stool collected before antibiotic use may help mitigate the extinction of potential commensal species and therefore reduce post-antibiotic dysbiosis.
The growing interest in FMT and its use has prompted various clinical trials to test its efficacy in treatment of a multitude of diseases. However, FMT is most commonly seen clinically in the context of treating patients with antibiotic-resistant Clostridium difficile infections (rCDIs).
This is because FMT was designated by the US Food and Drug Administration (FDA) as an Investigational New Drug (IND) in 2013.3 Its approved use in patients with CDI is through a discretionary enforcement, which allows physicians to treat patients who are acutely ill and meet the criteria of “patients with C. difficile infection not responding to standard therapy.” Fecal microbiota transplant has a cure rate for rCDI ranging from 80% to 98% depending on the study and delivery method used with good safety and low adverse events.4-7
The clinical trial data using FMT for other conditions such as liver disease, metabolic syndrome, and IBD look promising.8,9 However, because of the IND designation use of FMT for any indication other than rCDI can only be done in the context of a clinical trial and not in regular clinical practice.
What is autologous FMT and why aren’t we using it?
The FDA’s designation of FMT as an IND also includes what is being referred to as autologous FMT (aFMT)—FMT that is processed from the recipient’s own stool. Under this designation, FMT can only be used in the context of a clinical trial. With aFMT, novel infectious disease risk is eliminated from donor to recipient, but there still is risk associated with the physical processing of the stool.
In this study, aFMT was done by collecting the healthy volunteers’ stool prior to the administration of antibiotics, then delivering it via intrajejunal infusion after the completion of a 7-day course of antibiotics. The aFMT arm of the human trial (n=6) had the quickest and most complete return of the microbiome to its pre-antibiotic, baseline profile. Additionally, several species of microbes that did not recuperate after antibiotics in the spontaneous recovery arm (n=7) were found to be replenished within the aFMT arm. This study indicates that aFMT from stool collected before antibiotic use may help mitigate the extinction of potential commensal species and therefore reduce post-antibiotic dysbiosis.
Interestingly, aFMT has also been used to treat CDI. In a small study (N=24) of rCDI patients previously treated with vancomycin, a cure rate of 62.5% was reported using aFMT.10 This adds further evidence that aFMT can be helpful in returning a patient’s microbiome to its native diversity, potentially helping prevent or cure opportunistic infections after antibiotic use.
In this study, aFMT revealed itself a promising prospect for mitigation of microbiome depletion post-antibiotic exposure. However future studies with larger sample size are needed before any firm conclusions can be made. And, due to FDA regulation, this therapy isn’t readily available for use in treatment. Probiotics, on the other hand, are.
What can I tell my patients about aFMT?
This research paper provides data that is very exciting and consistent with the last few years of data.11 Both FMT and aFMT provide recipients a fast, robust shift to diversity similar to donor profiles or back to their native profile, respectively.12 If probiotics create a “Pandora’s box” of unknown interplay into each recipient’s unique microbial ecosystems, then an aFMT is the closest thing to native ecosystem restoration we have. Unfortunately, further use is limited by FDA regulations. They have clearly stated that aFMT is within the same designation as donor-derived FMT and therefore can only be used within a clinical trial as an IND.
We are hopeful that as techniques evolve, we will find a way to sufficiently minimize the very legitimate fear of infectious disease that is behind the strict restrictions of FMT and aFMT. Until then, supporting patient’s microbial ecosystem diversity through diet and lifestyle, and minimizing the need for antibiotics are the best tools.
Could probiotics cause harm after broad-spectrum antibiotic use?
The short answer: We don’t know enough about the mechanisms that underlie this complex microbial ecosystem to make clinical conclusions with confidence. From this data (present study under review), we do not know if there is an association between use of probiotics and development of post-antibiotic dysbiosis without access to a relative risk ratio. An analysis of relative risk would better allow us to understand the probability of developing post-antibiotic dysbiosis when comparing probiotic supplementation with spontaneous recovery. However, the results of this study do present some very interesting questions about what is happening to the microbiome after antibiotic use and what role probiotics may play in returning human physiology to homeostasis.
The study revealed that probiotics in both humans and mice prolong the period of time it takes for the microbial ecosystem to return to pre-antibiotic profiles. It also illustrated that the microbial profile in the probiotic arm most resembled the depleted microbiome after antibiotic treatment when compared to spontaneous recovery and aFMT.
Before making any clinical conclusions that this particular administration of probiotics is “bad,” let us step back. The literature has shown that correlation exists between adverse outcomes and decreased microbial diversity in humans.13 This study failed to report about participants' development of adverse outcomes associated with each intervention tested or further additional benefit. These sorts of outcome measures would be important for clinical decision-making.
Current research suggests that probiotic supplementation in healthy populations causes a transient increase in supplement-specific microbes resulting in improved immune system response, bowel movements, and even vaginal flora.14
This study did not discuss these types of quality metrics for us to better understand patient experience. If we begin to observe participants who are immune-compromised, how would probiotic supplementation impact their outcomes? The complexity of the microbial ecosystem may reveal that certain at-risk populations greatly benefit from probiotic administration with the potential to mitigate diarrheal disease or other long-term adverse outcomes compared to spontaneous recovery.
There also might be a difference depending on time and type of probiotics given. Concurrent administration of probiotics is common practice around antibiotic use in the naturopathic community; however, effects of this type of dosing regimen were not assessed in this study. A systematic review and meta-analysis performed in 2013 revealed a reduced risk of antibiotic-associated diarrhea and CDI when probiotics were given concurrently with antibiotic prescription.15
The study did detail that continuous shedding of probiotic strains was present 5 months post-probiotic therapy. If the probiotics could limit complete microbiome depletion during antibiotic exposure, how would this affect long-term outcomes? We don’t know. What is known is the current data is inconclusive and further research is required in order to make clinically relevant decisions.
How and when should you give probiotics after broad-spectrum antibiotics?
One of the conclusions implied by this single study’s results is that probiotic administration of this brand in otherwise healthy patient’s post-antibiotic exposure could negatively impact their microbial ecosystem. However, we do not feel there is enough clinical data to apply these conclusions to other types of probiotics or account for the individuality of every person’s microbiome, especially considering the small sample size (N=21) and the single variety of probiotic measured. More studies are needed to conclude that probiotics cause true clinical harm post-antibiotic exposure.
The largest barrier to drawing clinically relevant decisions from basic microbiome research is that we are dealing with an immensely complex system, one we still do not know much about due to the sheer diversity of microbes present in each individual system and their eco-immunologic interplay.16 Converse to the conclusion this study implies, we caution physicians on diminishing the potential benefits of probiotics based on this data.
Although we know loss of diversity is generally bad, we don’t know what it is about each individual’s microbial ecosystem that triggers a post-antibiotic infection (such as CDI) in one patient and not another. We don’t know what the keystone species are, or which metabolites are most important. We do know that keystone species can be different from one individual to the next, so one microbiome’s keystone species may not matter to another person when it is wiped out from antibiotics.17
We are at the tip of the iceberg when it comes to understanding microbial-host and microbial-microbial communication systems. We need to further elucidate more of this information from a network systems approach.18 We know microbial diversity can change rapidly and is dependent on many complex variables—so many, we can’t even begin to design properly controlled studies in humans. What we do know is the human (and animal) microbiome is an ecosystem. And, therefore, the basic principles of ecosystem biology should be applied to microbiome research to better understand this complex relationship.
When considering ecosystems biology, the outcomes from this study cause further pause and generate many questions. What is the quality of the probiotic they used? How is it manufactured? What was the difference between subjects who returned to their baseline microbial diversity with probiotics versus those that did not? What were the participants eating? Should we start probiotics before giving antibiotics? How would this study have changed if patients were on probiotics one week before antibiotics? And so on.
In light of this data, we will consider what the patient’s risk factors are and continue approaching their wellness from a naturopathic lens. How can probiotics be integrated as food-based products instead of supplemented? Would yogurt, kefir, kimchee, and other food-derived sources of Lactobacillus and Bifidobacterium produce different outcomes? Which foods and other habits are needed to best recover microbiome biodiversity after antibiotic use? We know that this is an immensely complex interplay, as mentioned before, but things like stress interacting through a gut-brain axis can impact microbiome ecosystem dynamics.19 We have to take into account the whole person when we consider the various mechanisms involved in patient health.