Prescription bacterial therapeutics represent a distinct pharmaceutical modality requiring specialized manufacturing, clinical development, and regulatory expertise. Three product architectures have emerged: whole-community products replicating FMT's ecosystem restoration, partial-community products focusing on functional groups, and defined-strain products targeting precise mechanisms. With three regulatory approvals achieved for recurrent C. difficile infection, the modality's viability is established.
The industry now faces its next challenge: demonstrating that rationally designed products can deliver efficacy in complex indications where FMT has shown promise. Success requires solving unique manufacturing challenges, maintaining viability of fastidious anaerobes through fermentation, lyophilization, and storage, while advancing mechanistic understanding to pathway-specific biology. This analysis examines the technical, regulatory, and commercial landscape, emphasizing manufacturing optimization, strain selection criteria, and clinical strategies that align product design with indication characteristics.
Whole-community LBPs leverage ecosystem restoration for microbiome depletion (rCDI); partial-community products balance diversity with manufacturing control; defined-strain consortia and single strains target specific mechanisms. FMT success in complex diseases (melanoma, GvHD, IBD) validates microbiome modulation; challenge is identifying minimal effective strain sets.
Critical derisking: lyophilization parameter optimization (survival varies by strain/growth phase), media reformulation for GMP scale-up (eliminating undefined/animal-derived components), potency assurance through temperature excursions. Co-fermentation offers efficiency but requires holistic purity testing; monoculture-then-blend maintains strain-level control.
Species-level choice insufficient; strain-level phenotype dictates potency (metabolite production, immunomodulation), safety (virulence factors, antibiotic resistance), manufacturability (growth kinetics, lyophilization tolerance), and colonization potential. Early characterization prevents costly pivots.
Defining molecular mechanisms (bile acid conversion, NOD2 pathway activation) strengthens regulatory submissions, IP protection, investor confidence, and trial design. Moving beyond "colonization resistance" to testable biomarkers enables patient stratification and endpoint selection.
FDA/EMA frameworks exist for LBPs; rCDI approvals demonstrate feasibility. Complex indications face standard requirements: dose-ranging, controlled trials, durability. Novel PK/PD considerations (engraftment, strain shedding, microbiome profiling) require thoughtful endpoint design.
While sector-wide venture funding moderated, partnerships remain active: late-stage assets approaching proof-of-concept, biomarker discovery, CDMO infrastructure. Value inflection points: Phase 1b safety/engraftment, Phase 2a mechanistic validation, manufacturing scale-up.
A healthy human body comprises human cells and a structured community of microbial cells, known as "microbiota". Collectively, the genomes of these single celled microorganisms constitute the "microbiome" of a niche, though in practice, the term "microbiome" is also used regularly to refer to the whole microbial community. At microbially colonized sites, including the digestive, respiratory and urinary tracts, skin and vagina, the resident microbiota confers essential health impact and benefits upon its human host. Such benefits may include colonisation resistance against pathogens and favorable immunomodulation. Commensalism and symbiosis are the result of millennia of human-microbe co-evolution and serve as the basis for the nascent modality of microbiome therapeutics.
Microbiome therapies represent a relatively new therapeutic modality and are being developed to treat a broad spectrum of illnesses. The earliest microbiome therapies tended to focus on recurrent Clostridoides difficile infection (rCDI), treatment of which was substantially derisked by the proven success of faecal microbiota transplant (FMT) for this indication1. The scientific rationale behind transferring whole microbial communities to combat disease is restoration of natural resilience and colonization resistance to a depleted ecosystem. Importantly, while FMT from healthy donors, largely regardless of microbiota composition, has proven effective for treating rCDI, FMT for treatment of more complex indications (e.g. irritable bowel syndrome (IBS)2, inflammatory bowel disease3, cancer4) has had lower and often more variable success rates, indicating that the microbiome profiles of the donor(s) and the recipient matter and have a bearing on a successful outcome.
The underlying premise for defined strain LBPs, whether consortia or single-strain, is that the desired therapeutic effect can be achieved without the need to replicate or replenish an entire microbial community. Various strategies have been used to nominate species- and strain- level therapeutic compositions. Some companies have performed discovery in animal models, others have analyzed large clinical datasets, and others still have pursued phenotypic screening to identify strains with a mechanism of action of interest. The genetically modified LBPs have typically been derived from strains of species with a history of use for genetic engineering.
A corpus of scientific literature emphasizes strain uniqueness and cautions against assuming general equivalence of strains of any microbial species. For example, strains of the species Escherichia coli, can be pathogens (i.e. AIEC, ETEC, STEC) or beneficial to health (i.e. probiotic strain E. coli Nissle 1917). The decision of which strains to include in a defined-strain therapeutic should therefore not be arbitrary, but rather a choice consciously made that balances strain-level performance across measures of potency, safety, manufacturability and commercial opportunity.
FMT, also known as "intestinal microbiota transplant" (IMT) involves the transfer of stool and the entire microbial community therein, from a donor (or pool of donors) to the intestines of a recipient with the intention of restoring health or functionality to the system. Stool slurries may be applied via colonoscopy, or where access via the rectum is not possible, (e.g. owing to obstruction or disease), via nasoduodenal tube.
Non-commercial, traditional FMT has not been approved as a medicine in the US or the EU and has often been used "off-label" by clinicians to create a treatment option for patients in need. For several years, until the approval of the first commercial product in the US (REBYOTA, see below), the US FDA used enforcement discretion, where an investigational new drug (IND) submission would otherwise be needed, to allow traditional FMT to be administered to patients with rCDI who had exhausted standard treatment options5. In the EU, centralized regulations for FMT are lacking, which means that regulations vary across the bloc6.
FMT has proven efficacious in clinical trials for treatment of rCDI1, ulcerative colitis (UC)7 and immune checkpoint inhibitor refractory melanoma4. CMC challenges for FMT as a mainstream treatment modality include difficulties in scale-up, given the requirement for donors to repeatedly pass rigorous health screening to remain active donors, and the biological limitations on the amount of material that a single donor can produce daily. As stool is an inherently variable material, with day-to-day variation within individuals and substantial compositional differences between individuals, a high-level view of batch-to-batch uniformity is taken. Composition is therefore typically assessed at a broad taxonomic rank, such as phylum or family, representation of which tends to be reasonably consistent across individuals, rather than at a highly resolved taxonomic rank, like species, which varies substantially.
Material pooled from several stool donors has often been used to ensure the resulting blend represents a rich microbial community, though it may be more diverse than a typical gut microbiota. Blending also attempts to compensate for the absence of key species in any contributing donor. Stool banks that centralised donor recruitment and health screening and which also provided screened, infusion-ready material to medical professionals have been key to the acceptance and delivery of safe FMT.
OpenBiome, a US stool bank, supplied sixty thousand investigational FMT treatments between November 2013 and September 20218, though BIOMICTRA, an FMT treatment produced by Biomebank (Australia) became the world's first regulator-approved microbiome therapy in early November 20229.
In the UK, NICE continues to recommend FMT for the cost-effective treatment of rCDI10. However, the availability of regulator-approved microbiome therapies is expected to reduce future reliance on traditional FMT for this indication. Together, advances in the modes of delivery available for whole stool products and progress in the field of microbiome therapies generally, will likely relegate traditional FMT from a pioneering treatment to a therapeutic option used in extraordinary or exceptional circumstances. Indeed, patients have consistently expressed some aversion towards having faecal material from other people administered to them via FMT11, 12, suggesting that commercial microbiome therapies would be preferred alternatives.
Whole community LBPs offer a technological advancement over traditional FMT. While the scientific rationale is the same – complete replacement of an individual's gut microbiota with the microbiota from a healthy person to restore health – the logistics of delivering the treatment are typically simplified.
The primary mode of action of whole community LBPs is restoration of health through ecosystem replacement, (essentially the same principle as FMT). For infectious diseases, such as rCDI, where natural colonisation resistance has been lost due to antibiotic treatment, such approaches are logical and effective. The modes and mechanisms of action by which whole community products improve clinical presentation and symptomology in the context of complex diseases is less well understood. The treatment itself may also not be suitable for all patients, especially those with anaphylactic food allergies, given the nature of the raw material.
MaaT Pharma (France) is developing enriched whole community LBPs focused on improving survival outcomes in cancer patients. MaaT Pharma refers to its products as "Microbiome Ecosystem Therapies (MET)". These are whole stool, multi-donor derived products, enhanced by the purposeful enrichment of a group of anti-inflammatory, short-chain fatty acid producing bacterial species – MaaT Pharma's so called "Butycore". Ecosystem restoration is the primary driving force behind the development of these high diversity, high richness treatments.
In June 2025, MaaT Pharma submitted a marketing authorisation application to the European Medicines Agency (EMA)15 and announced a commercial agreement with Clinigen16 for its most advanced drug candidate, MaaT013. MaaT013, (Xervyteg), is a third-line treatment delivered by enema for acute graft versus host disease (aGvHD) including gastrointestinal involvement. This is a promising treatment option that addresses a life-threatening complication of allogenic stem cell transplantation for which there are no approved therapies.
VOWST by Seres Therapeutics, (now fully licensed to Nestlé Health Sciences), is an FDA-approved microbiome treatment for rCDI17. It is a partial community product comprising the spore-forming microbial fraction of donated whole stool. The spore fraction is separated from the whole stool material by solvent treatment and purification, which greatly reduces the non-spore matter in the final product18. Each lot is derived from the faecal material of a single donor that has passed an extensive health screening19. The drug is delivered orally via capsules shortly after the patient has completed a course of antibiotics and immediately after a bowel preparation19.
The use of bacterial spores as active ingredients is inherently advantageous. As heat, stomach acid, oxygen and desiccation resistant structures, they can be handled and stored at room-temperature without reducing their viability potential. Spores are hardy structures that germinate in response to bile-acid cues which also means that they are well suited to transiting and colonising the harsh GI tract environment. VOWST is proposed to work by modulating bile acid conversion in the gut which inhibits C. difficile germination and growth20. Effective gut colonisation by the introduced spore-forming species means that the treatment can have a durable effect.
Defined strain-consortium LBPs, unlike the whole or partial community LBPs, consist of purified, identifiable, and well-characterised microbial, often bacterial, strains. These therapeutic strains are archived as pure cultures in master and working cell banks, access to which is carefully controlled and managed. The therapeutic strains may be derived originally from human tissues, such as stool, skin or vaginal swabs, or alternatively, from an environmental source, such as food, water or soil. Defined strain consortium LBPs do not contain any microorganisms other than the strains purportedly required for therapeutic efficacy, and purity is assured by rigorous testing during manufacture. Moreover, the composition of defined strain consortium LBPs is known and considered at the highly resolved taxonomic ranks of species and strain.
The batch-to-batch variability for defined strain bacterial therapeutics is much lower than for whole community products because the starting material is standardized. Biological variability is low, and the manufacturing processes are optimised for the specific therapeutic strains. Defined strain consortium products contain more than one, but usually fewer than twenty strains, though very large defined strain consortia containing more than 100 strains were previously in development, (i.e. FB-001 from Federation Bio (US) included 148 strains). Defined strain consortia are a popular option within the spectrum of LBPs, with several companies adopting this approach:
Like for the defined strain consortium products, single-strain prescription LBPs include a purified, well-characterised medical-grade therapeutic strain. Osel Inc (US) has repurposed a well-known bacterial strain, Clostridium butyricum MIYAIRI 588, that had been used in Japan as a non-prescription probiotic for decades, to augment ICI effectiveness during the treatment of metastatic renal carcinoma (Phase 1 studies NCT03829111, NCT05122546)25, 26. Despite not meeting primary endpoints, (change in Bifidobacterium composition of stool from baseline to week 12), Phase 1 studies yielded promising read-outs for progression-free survival and clinical outcomes. Phase 2 clinical development is therefore expected to commence in 2026, (NCT07037004).
Exeliome's (France) lead product, EXL01, comprises a single strain of Faecalibacterium prausnitzii formulated as a capsule for oral administration. The immunomodulatory properties of this strain, which are reportedly achieved through hyperactivation of NOD2 pathway signalling, effecting metabolic rewiring in monocytes and macrophages27, 28, are thought to benefit ICI effectiveness, which is being investigated in three Phase 2 clinical trials for gastric cancer (NCT06253611), non-small cell lung cancer (NCT06448572), and hepatocellular carcinoma (NCT06551272). In 2023, Exeliome initiated clinical development of EXL01 in Crohn's disease, (NCT05542355), where it was intended to prolong the maintenance of corticosteroid-induced remission in adults with mild-to-moderate disease. Though the study was terminated early owing to poor recruitment, a Phase 2 study, (NCT06925061), evaluating the suitability of EXL01 for the prevention of endoscopic recurrence of Crohn's disease after surgery started recruiting in early 2025. The drug candidate is also being investigated for prevention of rCDI in high-risk patients, (NCT06306014).
Companies developing genetically modified LBPs (GM LBPs) typically adopt a single-strain approach. The chassis is usually a strain from a well characterised bacterial species for which genetic tools are available. This is important, because manipulation of non-model species is often challenging, requiring considerable protocol optimisation and de novo establishment of a genetic manipulation toolbox.
Azitra (US) is developing engineered strains of Staphylococcus epidermidis as investigational new drugs for treatment of dermatological conditions. ATR12-351, (Phase 1b, NCT06137157) targets Netherton syndrome, which is a hereditary, rare and chronic disorder diagnosed in infancy characterised by inflamed skin, poor hair quality and a predisposition to allergies29. The syndrome arises due to an attenuation or loss of function of the LEKTI protein, a SPINK5-encoded serine-protease inhibitor, that plays a role in regulating skin shedding, (desquamation). Azitra's drug candidate is topically applied to the skin and delivers a functional LEKTI protein to the stratum corneum30.
Prokarium (UK) has initiated Phase 1 clinical development of ZH-9, its intravesical treatment for the prevention of recurrences in non-muscle invasive bladder cancer (NCT06181266). The treatment consists of a live attenuated strain of Salmonella enterica Typhi, ZH9, that has inherent anti-tumour properties31. The attenuation was achieved by genetic modification32, 33.
Several other companies developing GM LBPs are in preclinical development, including Endure Biotherapeutics (US), NeoBe Therapeutics (UK), Pulmonobiotics (Spain) and Sanarentero (US). Previously, Synlogic had advanced its GM LBP drug candidate, labafenogene marselecobac, to Phase 3 trials for treatment of phenylketonuria, however, an internal review of interim Phase 3 data indicated that the candidate drug was unlikely to meet its primary endpoint. The trial was therefore discontinued and the company ceased operations in 202434.
Prebiotics are "substrate(s) that (are) selectively utilized by host microorganisms conferring a health benefit"35. Prebiotics are useful because they can selectively boost the population size of the microorganisms that consume them, which should yield a net health benefit for the host. A substrate that enriches the pathogenic microbial population, for example, would not be considered a prebiotic. When prebiotics are matched with non-prescription beneficial strains, the resulting product is commonly known as a "synbiotic". Marketed over-the-counter (OTC)/consumer synbiotics include Bimuno GOS + Probi Defendum and Bimuo GOS + Probi Digestis from Clasado Biosciences, (UK) and Probi AB (Sweden)36. Prescription synbiotics represent an as yet unproven combination drug development option. We are aware of one biotech company, ProbioTech USA (US)37, with IND-ready oral and suppository dual-chamber synbiotic programs for cancer supportive care and GI indications, and a preclinical program for women's health.
Manufacturing prescription grade LBPs for clinical trials and commercial use is not yet routine and can be technically challenging. For the whole and partial community LBPs, the raw material is donated by volunteers who pass a health-assessment. As such, manufacturing involves additional screening to ensure the absence of known pathogens and contaminants, and formulation of the material for administration to patients. Providers of whole community LBPs typically build custom manufacturing suites for manipulating the raw material (e.g. EnteroBiotix38, Maat Pharma39).
Production of defined-strain LBPs follows a different path, the main steps of which typically include large-scale fermentation, biomass harvest, lyophilisation, blending, milling and incorporation into a dosage form, (e.g. a capsule, cream). Purity and viability testing are part of release testing protocols. Methodologies and expertise for handling stool, strict anaerobes, spore-formers and fastidious microorganisms are not commonplace in traditional manufacturing suites. Consequently, some microbiome drug developers, (e.g. Vedanta Biosciences40, Servatus Biopharmaceuticals (Australia)41 and prior to its closure, 4D Pharma (UK)42) brought production "in house".
Processes for large-scale production of microorganisms for foods, vaccines and recombinant protein products have existed for decades. Of note, most of the industrial-scale defined-strain bacterial manufacturing today is conducted at food-grade standards to supply ingredients for dietary supplements and health foods and drinks. The producers of these bacterial strains are wary of adding the additional controls demanded by FDA to classify their bacteria as "prescription-grade". Reimagination of standard processes is needed to support prescription LBP manufacture. Crucially, the viability of the therapeutic strains must be maintained throughout all steps of the production process. This means that exposure to environmental factors and processing steps that may reduce viability, (and therefore potential potency), must be carefully managed, monitored and controlled, and would ideally be avoided. Depending on the characteristics of the strain of interest, it may need to be shielded from exposures to oxygen, heat, cold, moisture, agitation, certain chemicals or buffers, temperature fluctuations or freeze-thaw cycles.
Access to fermentation equipment appropriate for small-scale process development prior to scale-up can also be limiting and risks extending timelines at large-scale if further process optimisation is needed. For example, the sterilisation and harvest conditions that work for bench-top volumes, (i.e. < 50 ml), may not apply to LBP cultures grown in bioreactors with thousands or tens of thousands of litres capacity.
Media composition must also be considered and optimised prior to scale-up. This is judicious to manage costs, assure yield and maintain timelines during production in large bioreactors. While culture failures at small scale may be considered a nuisance, failure or underperformance in large volumes would be extremely costly in terms of time and money. It is therefore imperative that the risks of sub-optimal or failed production runs are minimised. Beyond ensuring technical competency and adequate facilities are in place, derisking steps include careful media optimisation and development of scalable processes as early as possible in the development process. Altering media composition risks altering microbial phenotype and hence strain performance, so it is crucial that the impact of changes to media composition on the target strain's biology is considered before key preclinical experiments are undertaken.
Inclusion of undefined components in bacteriological media risks inconsistent performance owing to lot-to-lot variability in the source material, (e.g. peptone). Use of media recipes that include animal products, such as blood or clarified rumen fluid, may be routine during discovery or preclinical development, but are impractical and undesirable at large-scale. Identifying and sourcing alternative media components that deliver the same growth performance and which meet the requirements for GMP manufacture is not trivial, and media reformulation, if needed, would be expected to greatly extend process development timelines. Media recipes that require an expensive subcomponent may be unaffordable to produce at large-scale and if the subcomponent is highly refined or otherwise purified, it may also prove difficult to procure it in sufficient quantities. Preferentially, small-scale process development would identify the simplest possible media composition that supports microbial growth to high biomass while also being compatible with industrial processes, volumes and standards.
It is advisable to build small-scale lyophilisation survival testing and optimisation into the early stages of process development for LBPs. Prioritising lyophilisation compatibility even earlier could also prove valuable in the long run. Biological factors that influence lyophilisation survivability for bacteria include cell wall structure, (i.e. Gram-positive or Gram-negative), and the metabolic and physiological state of the cells at the time of their harvest (e.g. late-exponential, stationary). Survivability is also influenced by decisions taken at various processing steps. For example, lyoprotectant choice, freezing rate and the temperatures applied during the drying steps all influence survivability. For optimal preservation and recovery, lyophilisation parameters should be defined for each strain individually. This poses an additional challenge for manufacture of co-fermented products, where a one-size-fits-all approach to lyophilisation is necessary, but probably not optimal for any of the component microbes.
As an alternative to lyophilisation, Servatus Biopharmaceuticals has developed a proprietary method for very low temperature spray drying that can be applied to produce high quality LBPs41. This could deliver attractive production efficiencies.
One approach to manufacturing defined strain consortium products is to generate a drug substance from a monoculture of each therapeutic strain before blending them during incorporation into drug product. This allows the sponsor to straightforwardly monitor drug substance purity and culture performance on a single strain basis. Depending on the number of strains included in a consortium, it may not be possible to generate the various drug substances in parallel, owing to the capacity limitations of the production facility.
Alternatively, therapeutic strains may be co-fermented. This yields a drug substance that includes all strains from the outset. Time- and cost-saving efficiencies during production are the main pragmatic drivers for co-fermentation. Savings are achieved because only one production run is required to yield the drug substance. Associated paperwork is also minimized.
Embracing co-culture for production requires a wholehearted commitment to the approach. Community composition is often dynamic for a period before achieving stability and it is unlikely that the microbial profile of the resulting drug substance will be balanced. Tracking the growth performance and final load of individual strains in a mixture is more challenging than with monocultures and purity testing of a co-cultured product requires a holistic approach.
Monocultures and co-ferments may differ biologically, which could be relevant to drug potency. For example, co-fermentation requires the member microbes to form a co-compatible community. Cross-feeding may boost the growth performance of strains that struggle or underperform in monoculture. Cross-feeding and microbe-microbe interactions may also stimulate the production of metabolites that are not produced in monocultures. Microbes may also respond physiologically to co-existence which may yield a hardier cellular structure or architecture (e.g. exopolysaccharide production, stress conditioned cells) and may prime the microbes to better compete and colonise when introduced to a complex ecosystem (e.g. production of pilli, fimbrae, porins, antimicrobials).
Drug formulation post lyophilisation also requires thoughtful decision making. Considering the drug's therapeutic activity influences decision-making on its intended site of therapeutic activity, which in turn, informs which dosage form, (e.g. capsule, tablet, powder sachet, enema, inhaler, topical cream, etc.), is used. For orally dosed, lyophilised LBPs, it is often desirable to deliver the active ingredients to the intestines for rehydration. Dosage forms that protect their contents from gastric acid (e.g. enteric coated capsules) are preferred for this application.
The amount of active bacterial ingredient that can be included in a dosing unit (e.g. tablet, capsule, fingertip unit, actuation), depends on the density of live microorganisms in the lyophilised material. This is best explained with an example, (Table 1):
Size 00 capsules are among the largest that can be comfortably swallowed by adults. These capsules can be filled with ~0.95 ml liquid or 950 mg powder, (assuming a density of 1 g/ml)43. If, for therapeutic efficacy, a dose must contain 8x109 CFU of a therapeutic strain, and the potency of the lyophilised powder is 1x1010 CFU/g, then 800 mg of powder will provide the target dose. This fits inside a size 00 capsule. However, if the potency of the lyophilised material is lower, at 1x109 CFU/g, 8 g of powder would be needed per dose. This would require a patient to swallow nine large capsules. Such a dosing regimen would be burdensome and could lead to low patient compliance or incomplete dosing.
Table 1: Impact of drug substance potency on dosing volume
| Target Dose (CFU/dose) | Potency (CFU/g) | Amt. powder/dose (mg) | No. capsules/dose |
|---|---|---|---|
| 8x109 | 1x1010 | 800 | 1 |
| 8x109 | 1x109 | 8000 | 9 |
For consortium products in which the individual therapeutic strains are prepared as different drug substances, a choice exists about how to combine them during preparation of the drug product. For example, it may be desirable to blend the materials according to their potency, (i.e. a target CFU/strain/dose) or by powder weight (i.e. a target mg/dosing unit). The ratios of strains in the drug product may be approximately equal or deliberately unequal and there can be sound reasoning behind the decision to adopt either approach.
Dosing materials that lack active ingredients may be prepared for use as placebo in placebo-controlled trials involving LBPs. Care should be taken to ensure that the excipients included in these dosing materials are unlikely to have an unintended prebiotic effect on the microbiota, especially if microbiome profiling is a trial endpoint.
Like many other medicines, those containing LBPs are often taken by patients in their homes. Diligent patient compliance with the dosing regimen is therefore key to successfully delivering a course of treatment. Any storage temperature requirements to maintain shelf-life of the drug product would ideally be readily achievable in a standard home (e.g. refrigerated or frozen storage), and the viability of the therapeutic strains should be assured through temperature fluctuations that might reasonably be experienced during product handling at home. Packaging requirements may include innovations to maintain the drug product's shelf-life, (e.g. oxygen scavengers, moisture absorbers) but must also allow for straightforward patient access to the medicine44. Given the sensitivity of many LBP products to temperature, accelerated stability studies may not be acceptable to regulators45, the consequences of which should be considered early in the manufacturing timeline.
Generally, LBP use is incompatible with concurrent antibiotic use, though antibiotic priming (to open a niche for the strains in the drug), often features in LBP treatment protocols19.
Like cell therapies, LBPs are living medicines, which means that the dose delivered may increase rather than decrease after administration, especially for LBPs comprising wild-type strains that are designed to colonise their new environment. Traditional approaches to pharmacokinetics (PK) and pharmacodynamics (PD) therefore need to be revised to sensibly apply to LBPs.
For LBPs, colonisation kinetics form the basis of PK. Studying the PK of therapeutic strains involves monitoring where in the body they are detected and their absolute numbers at these locales, which may change over time as the strains integrate into their new ecosystem and according to the frequency of repeat dosing. To avoid sepsis, most LBPs would not be expected to be found alive in the bloodstream. For orally dosed LBPs, levels of the therapeutic strains in stool are taken as a proxy for colonisation of the intestines, given the difficulty of obtaining samples from the lower, and especially the upper, intestines. If available, quantitative biomarkers indicative of therapeutic strain activity (e.g. a strain-specific metabolite in urine) could be monitored as part of the PK package. The duration of colonisation by the therapeutic strains, whether transient or long-term, is also relevant. Many LBPs are designed to deliver strains that "engraft", that is, permanently colonise their new niche to deliver a sustained therapeutic effect, even after dosing is complete. This could also mean that the therapeutic strains are shed from the recipient over an extended time-period.
PD takes a holistic approach and involves monitoring microbiome and other relevant omics- profiles, (e.g. metabolome, proteome) in an appropriate tissue, (e.g. stool, skin swab, lung aspirate) to detect and measure the impact of the LBP.
Understanding the relationship between the dose administered and therapeutic effect is key to better understanding the pharmacology of these drugs. The absolute numbers of a strain needed for therapeutic effect may vary between individuals, depending on their prevailing microbiome composition, physiology and lifestyle habits. The size of the therapeutic strain community in its target niche is likely to be kept in check by cooperative and competitive interactions with other resident microbes. This means that the relationship between the dose administered, and the eventual size of the therapeutic strain community observed, is not likely to be standard across all strains, or knowable from the outset. For example, strains well adapted for colonisation of their intended niche may outperform other strains that are less well adapted, even if the well-adapted strain is supplied in equal or lower numbers than the less well-adapted strain. Prebiotics, if used, would also be expected to alter the colonisation dynamics of the strains introduced. Eventually, with enough data-points, these community dynamics could be modelled and predicted in silico.
Relating dose administered and the resulting in situ therapeutic strain community metrics to observations of clinical efficacy in large numbers of human subjects will allow the range of effective therapeutic levels to be determined and the relationship between the dose delivered and its therapeutic effectiveness to be ascertained.
The widely accepted definition of probiotics as "live microorganisms that, when administered in adequate amounts, confer a health benefit on the host"46, 47 is broad and sufficiently generic to overlap with the defining expectations of LBPs48. LBPs contain live organisms, e.g. bacteria that are "applicable to the prevention, treatment, or cure of a disease or condition of human beings" and are not vaccines48. Filterable viruses, oncolytic bacteria, gene therapy agents and injectables are not LBPs48. By definition, both probiotics and LBPs require delivery of live microorganisms with health-positive intentions. What therefore is the difference between them? Fundamentally, it is the nature of the intended health-claim that determines whether a product must be considered a drug or a supplement49. Regardless of semantic preferences for "LBP" or any synonym of "probiotic" (e.g. next generation probiotic, precision probiotic, beneficial bacteria), if a product is associated with a therapeutic health claim, it must be considered a drug.
Many people will be familiar with the sight of various health products containing microorganisms available for purchase in pharmacies and supermarkets without prescription. These products are often capsules or gummies or functional foods that include viable bacteria or fungi with proclaimed health benefits. These products are not registered medicines. To date, three microbiome therapies, (BIOMICTRA, REBYOTA, VOWST), have been approved by regulatory authorities worldwide. These products have been rigorously assessed for safety and efficacy through clinical trials, and are drugs.
The US FDA's definition of a drug is not limited to products with therapeutic effect against disease49. It also encompasses products that intend to alter the structure or function of the body without reference to disease, e.g. medications to prevent pregnancy. In contrast, a food supplement that is intended to enhance the flavour or texture of food and that does not claim a therapeutic purpose or any impact on the normal structure or functioning of the human body, is not considered a drug, e.g. sourdough starter cultures. Nor is a dietary supplement that is intended to alter the normal structure or function of the body without therapeutic intention, e.g. "supports digestive function", considered a drug. Where there is an intention to diagnose, cure, mitigate, treat, or prevent disease or to alter the normal structure or function of the human body with therapeutic intent, the product is considered a drug.
Table 1: Determining whether beneficial microbes are drugs or supplements
| Criterion | LBP | Non-prescription live microbial supplement |
|---|---|---|
| Contains whole microbes? | Yes | Yes |
| Microbes are viable? | Yes | Yes |
| Includes a health claim? | Yes | Yes |
| Health claim is therapeutic? | Yes | No |
As a relatively new therapeutic modality, comprehensive regulatory frameworks for microbiome treatments have been forged in real-time to respond to the needs of the sector. To date, globally, approvals have only been issued for complex, donor-derived products intended to treat rCDI. These approvals by the American and Australian regulatory authorities have established precedent. No microbiome drugs have yet been approved as prescription medicines in Europe, though MaaT Pharma has pursued clinical development in Europe and is requesting approval in the EU for its complex donor-derived product, Xervyteg15.
Within the EU, country-specific regulation adds to the regulatory complexity. In recognition of this, the Substances of Human Origin (SoHO) regulations were published to offer greater, harmonised protections to donors and recipients of SoHO-relevant materials across member states, and with the intention of streamlining regulatory approvals within the EU50.
Microbiome products that are not intended to cure or treat disease are not regulated as medicines. Rather, these non-prescription, direct to consumer products, often generically referred to as "probiotics", are frequently treated as dietary supplements if orally dosed but not included in food. In this context, the US FDA does not require companies to demonstrate efficacy, though safety must be assured, and products must be produced to food grade GMP standards and appropriately labelled51, 52. If a dietary supplement intended for the US market includes a structure/function health claim, it must bear the FDA disclaimer, "This statement has not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat cure or prevent any disease"52.
In contrast, the EU's European Food Safety Authority (EFSA) sets a high bar for approval of dietary supplements containing microbes. It prefers applications to include precise health benefit claims that are supported by evidence from human clinical trials. This places a significant burden of proof on companies that wish to bring dietary supplements containing beneficial microbes to market in the EU. EFSA's overarching position is that the use of the term "probiotic" is itself an unauthorised health-claim53, 54 and no health claims apart from a long-standing claim for live cultures that improve lactose digestion of yoghurt or fermented milk in lactose intolerant individuals55 have been approved in the EU. Despite this, several EU member states have independently permitted use of the term "probiotic" within their jurisdictions. The stringent regulatory requirements and inconsistent local regulations within the EU are likely to dissuade companies from targeting this market.
Asia-Pacific has a large and growing probiotics market, and regulations for bringing probiotic foods and supplements to market vary across the region56. Probiotics have a well-established history of use in Japan, where two systems, "foods for specialised health use" (FOSHU) and "foods with function claims" (FFC), exist for the management of health claims liked to orally consumed probiotics. FOSHU was established in 1991. It focuses on physiological health claims (like structure/function health claims in the US), that appear on the labels of foods, pills and capsules. FOSHU approval is a stringent standard that focuses on safety and efficacy and can only be granted following favourable government review. FFC has existed since 2015 and is more flexible than FOSHU. It does not involve government review, and places the responsibility for establishing appropriate, scientifically-supported label claims with the applicant. Evidence from clinical trials is not required for FFC approval.
In Australia, microbial products with health-benefit claims are regulated either as foods by the Food Standards Australia New Zealand (FSANZ) or as therapeutic goods by the Therapeutic Goods Agency (TGA). Products falling within the TGA framework are assessed according to a risk-based classification that determines the level of regulatory scrutiny they will be subjected to. In early 2025, the TGA issued guidance to assure the quality of "probiotic medicines"45.
While microbiome research dates back decades, commercialisation of the microbiome for therapeutics really began in earnest circa 2011, (Figure 1). The sector raised ~$400M USD between 2011 – 2013, and this increased every three-year period to 2020 – 2022 when the sector raised ~$4.5B USD, (Figure 1). Since then, financing within the microbiome sector has decreased, with a substantial drop-off noted in the latest period, 2023 – H1 2025.
Figure 1: Overview of the value of microbiome deals, 2011 – H1 2025. Cumulative value of all financing achieved in the microbiome sector across three-year intervals. For 2025, only deals up to and including 30th June 2025 were considered. Financing achieved through grants, capital raising deals, mergers, acquisitions and strategic alliances were included. Source data retrieved from GlobalData, July 2025. "Microbiome" search term used to identify relevant deals. "Announced" and "Completed" deals were considered.
Grant funding has always been a primary source of non-dilutive capital finance for the microbiome sector, (Figure 2a) and in 2020 – 2022 it amounted to ~ $2.2B USD (Figure 2c). Like grants, venture funding has also been important over time, and at its most recent peak in 2020 – 2022, was worth ~$953M (Figure 2b, c).
Figure 2: Trends in capital financing in the microbiome sector over time. (a, b): Capital raising deals attributable to a given financial category expressed as a percentage of all microbiome deals over three-year time periods. (c): Trends in the value of various sources of capital finance to the microbiome sector across three-year intervals. (a, b, c): For 2025, only deals up to and including 30th June 2025 were considered. Source data retrieved from GlobalData, July 2025. "Microbiome" search term used to identify relevant deals. "Announced" and "Completed" deals were considered.
Venture funding (Series A – D) is most active in North America, particularly in the US (Figure 3a). Activity in Europe and Asia-Pacific is comparable, though the Asia-Pacific market appears more mature, given (i) the generally larger investment values reported and (ii) the greater proportion of Series B investments in Asia-Pacific. Activity in Israel drove all deals in the Middle East, (Figure 3b).
Figure 3: Microbiome focused venture financing deals across geographies. (a): Heat-map of the world illustrating locations of microbiome-focused venture financing activity. The lighter the yellow shading, the fewer deals recorded. No deals were registered in countries coloured grey. (b) Pie-charts indicating the proportion of venture financing deals in each geographical region attributable to various rounds of venture financing, Series A-D. Source data retrieved from GlobalData, July 2025. "Microbiome" search term limited to deals "completed" and "announced" across Series A – E+ between 1st Jan 2010 and 30th June 2025 inclusive were used to identify relevant deals.
Generally, Series A investment started to become more active from 2016, reaching a peak in 2023 and has declined since, reaching low points in 2024 and the first half of 2025, (Figure 4a). Half of 2023's early-stage investments classified as "Series A" represented refinancing of an existing company. This likely reflects the willingness of investors to reinvest to support their existing investees rather than to add new microbiome companies to their portfolios. Overall, Series A investments tend to be worth less than $25M, with deals worth less than $5M most frequently reported, (Figure 4b). Series B investment activity peaked in 2020 and has since waned (Figure 4a). Series B financing values vary, though investments of $10M - $50M are most common, (Figure 4b). Series C investments are comparable in value to Series B investments, though they occur less often. The earliest Series D financing was in 2020, and no Series D rounds have completed since 2021 (Figure 4a). These investments have been worth $50M - $100M (Figure 4b). Overall, venture financing investments of $100M or more are rare in the microbiome sector, (Figure 4b).
Figure 4: Trends in type and value of venture financing. Heat-maps illustrating (a) trends in the number of Series A – D deals over time and (b) trends in values of Series A – D deals 2011 – H1 2025. Source data retrieved from GlobalData, July 2025. "Microbiome" search term limited to deals "completed" and "announced" across Series A – E+ between 1st Jan 2010 and 30th June 2025 inclusive were used to identify relevant deals.
Funding through equity offerings has been effectively stable since 2017-2019 (Figure 2b). IPOs were registered for six companies between 2019 and 2022 (Figure 5a), with amounts, typically less than $100M raised (Figure 5b). When used, private investment in public entity (PIPE) deals have usually been worth less than $5M or between $10M and $25M, (Figure 5b).
Figure 5: Trends in type and value of equity offerings. Heat-maps illustrating (a) trends in the number of equity offering deals over time and (b) trends in values of equity offering deals 2011 – H1 2025. Source data retrieved from GlobalData, July 2025. "Microbiome" search term limited to deals "completed" and "announced" between 1st January 2010 and 30th June 2025 inclusive were used to identify relevant deals.
Private equity and debt offerings have been a minor source of financing for the microbiome sector over time, (Figure 2b, c). Direct investment has been used to raise sums ranging from ~$6.5M to $100M. Debt offerings, which have been the least frequently used method of raising capital in this sector, have featured since 2017-2019, (Figure 2b).
The relative proportions of deals involving mergers and acquisitions have been broadly stable over time (Figure 6). Several mergers completed between 2016 and 2020, though none have completed since (Figure 7a).
Acquisitions have increased since 2020, with greatest activity in 2022 (Figure 7a). Notable acquisitions in 2022 included:
Figure 6: Trends in mergers, acquisitions and strategic partnerships in the microbiome sector over time. Deals attributable to a given category expressed as a percentage of all biotech deals in each three-year time-period. For 2025, only deals up to and including 30th June 2025 were considered. Source data retrieved from GlobalData, July 2025. "Microbiome" search term used to identify relevant deals. "Announced" and "Completed" deals were considered.
Figure 7: Activity in mergers, acquisitions and strategic alliances deals over time. Source data retrieved from GlobalData, July 2025. "Microbiome" search term used to identify relevant deals. Only completed deals were considered.
Asset transactions have been stable over time (Figure 6, Figure 7a). Examples of prominent recent asset transactions include:
Partnership activity was high between 2016 and 2020, especially during the years from 2016 to 2018 (Figure 6, Figure 7b). Most partnerships were forged for asset co-development. The highest-value partnership agreements between 2016 and 2020 were:
Beyond drug supply agreements, notable recent partnerships involving pharma companies include:
Licensing agreement activity has slowed in the sector since 2020 (Figure 6, Figure 7b). Deals since 2023 have been few, but have included:
The significant licensing agreement worth $1905M agreed between Seres Therapeutics and Nestlé Health Science in 2016 for Seres's rCDI and IBD assets remains the largest in the microbiome sector to date.
Human microbiomes have long been touted as potential therapeutic targets and as a reservoir of beneficial microbes for use in medicines, supplements and cosmetics. Market approvals for whole-community LBPs help to validate the "bugs as drugs" concept, the potential of which will be fully realised when defined-strain LBPs also achieve clinical success and market approval.
Belief in the potential of microbiome manipulation as a therapeutic modality has allowed the sector to become established and to grow. Clinical validation of defined-strain LBPs is, however, still lacking and is now essential to assuring the future of this potential therapeutic option. The recent clinical failure of Vedanta Biosciences' VE202 in UC illustrates the challenge of accurately predicting what mix of defined strains will lead to a meaningful clinical effect and emphasises a broader need for a more complete understanding of the intricacies and complexities of host-microbiota symbiosis. Without clinical validation, fundraising to support further discovery and development for microbes as drugs will prove extremely challenging. Consequently, the performance of the defined-strain LBPs currently and recently in clinical trials will be watched very carefully by stakeholders inside and outside the industry.
The fates of the microbiome assets presently in clinical development notwithstanding, there exists tacit evidence of confidence for the future. The merit of the defined-strain approach is implicit in the developers of whole- and partial- community LBPs adding defined-strain LBPs to their portfolios (e.g. BiomeBank, BB265; Maat Pharma, MaaT034; Seres Therapeutics, SER-155). Series A investments to support new and existing start-up activity in recent years, especially in 2023, is also indicative of the potential for value realisation still perceived to exist in the sector. However, prudent investor sentiment prevails, and while there is a willingness to support existing portfolio companies to achieve key milestones or read-outs, fundraising remains challenging in the current economic environment. Much innovation in microbiome drug development has been pioneered by small and mid-sized biotech companies that rely on grant and venture funding to advance their assets through development. Pharma companies have gotten involved through partnerships, supply agreements and late-stage deals linked to market authorisation but, with some exceptions (e.g. Ferring Pharmaceuticals), have now backed away from investing heavily in microbiome drug-discovery and development. Most big pharmaceutical companies seem to be waiting for the field to mature before (re)engaging. Nevertheless, there has been historical pharma and large biotech interest in the microbiome as a source of novel biomarkers, particularly as a source of clinical response biomarkers, which if validated, would prove valuable across treatment modalities.
Investment in the microbiome sector has also advanced microbial sciences generally, creating opportunities for discovery and development across specialisms. Investment in CDMO infrastructure and capabilities means that large-scale exploitation of fastidious and anaerobic microorganisms for broad biotechnology applications is technically enabled, though capacity and widespread capability is still limiting. Likewise, co-fermentation at scale is also feasible now. These advances are important because there will likely be greater emphasis on microbial metabolites, by-products and microbial cell fractions as therapeutic agents and commodities into the future.
Apart from delivering whole cells as a source of therapeutic material, microbially derived metabolites and cellular fractions (a type of "postbiotic"57 may also have therapeutic potential and could be exploited as drugs. Natural products of microbial origin have long served as medicines, (notably, a major source of antimicrobials), representing 28% of all non-mammalian natural products approved by the US FDA to the year 201358. Since the turn of the century, drug discovery focused on natural products has fallen out of fashion in favour of screening synthetic compounds, largely driven by the greater efficiencies perceived to exist in exploitation of the latter. Scientific and technological advances of recent times, including massive discovery of new microbial species through a renewed and enabled interest in microbial culturing, widespread microbiome profiling and capacity for multi-"omics" data generation, have yielded rich and untapped microbiological resources which means that natural product discovery from microbes could find itself in vogue again.
Artificial intelligence (AI)-driven data-mining may be considered a boon to drug developers hoping to exploit the therapeutic value of natural products of microbial origin now. For some, focusing on the molecules from microbes rather than on the microbes themselves as the therapeutic agents is a more attractive way to bring drugs to market. This is because the "drugs-from-microbes" approach would take advantage of the small molecule pharmaceutical infrastructure and know-how that already exists for screening compounds, and the path to and through clinical development is well trodden. Growth and activity in this area are worth watching into the future.
Proving precisely how a medicine exerts its effect boosts confidence in the drug candidate, derisks clinical development, helps to secure IP and to win investment. For LBPs, this means moving beyond generic modes of action, (e.g. "restoration of colonisation resistance" and "restoration of barrier function") to defined mechanisms of action that identify molecule-, gene- and pathway- level biology directly relevant to therapeutic efficacy. As microbiome drug active ingredients comprise living organisms rather than a molecular entity, it is also very likely that most LBPs will prove efficacious through several routes, (e.g. immunomodulation and microbial antagonism), and where whole-community or consortium LBPs are used, though complementary, and potentially synergistic mechanisms, (e.g. synchronised immunomodulation via different effector molecules). This is an area in which significant progress needs to be made.
For GM LBPs, the intended mechanism of action is largely premeditated and forms the basis for the strain engineering. Nevertheless, the intrinsic host-microbe interactions of the chassis strain could impact the pharmacology of the intended drug and must also be carefully considered from the outset.
The microbiome drug-development industry has been built on the foundations laid by decades of academic research, pioneering FMT trials and the heritage of the food-microbiology and non-prescription probiotics industries. It is now well established that microbiomes are valid therapeutic targets and that whole microbiomes can serve as drugs. That rationally chosen, named microbes can serve as effective drugs is currently unproven, though pipeline activity asserts scientific confidence in the approach. Favourable clinical read-outs in the near-term will be key to assuring investor confidence in the LBP sector beyond Series A.
Manipulating the gut microbiota to eliminate a gastrointestinal pathogen such as C. difficile, is obvious and logical. How manipulating the gut microbiota positively affects complex and/or systemic disease is less apparent. Even so, favourable FMT outcomes in complex indications, including ICI refractory melanoma, GvHD and IBD emphasise the relevance and utility of microbiota manipulation in these contexts. Understanding exactly how specific gut microbiota manipulation works for systemic disease, and resolving efficacious FMT microbial communities to a few therapeutic strains that recapitulate the whole community effect, are matters of ongoing research.
The full potential of microbiome therapeutics is incompletely realised at present, though significant progress has been made. The next frontier for microbiome therapies is a thorough understanding of the mechanisms of action underlying their efficacy. Microbiome "omics" experiments are well suited to the AI-era of "big data" analysis. Insights gleaned will foster further innovation and research translation as unappreciated host-microbe biology is discovered and understood.
Explore Alacrita's latest insights on biotech development, regulatory strategy, and life science innovation.
Alacrita delivers specialized expertise in product development, regulatory strategy, CMC, and business planning. Contact us to discuss your specific challenges.