Introduction

Live biotherapeutic products (LBPs) have progressed from experimental fecal microbiota transplantation to FDA-approved biological products, while development efforts increasingly include defined microbial consortia and genetically modified strains. The approvals of Rebyota in 2022 and Vowst in 2023 showed that living microbial products can meet U.S. standards for safety, purity, potency, and effectiveness. However, these approvals did not create a single development template. Sponsors must still define the active product, control biological variability, reduce pathogen and antimicrobial-resistance risks, establish meaningful potency methods, and justify nonclinical models for microorganisms that may persist, shed, or interact with the host microbiome.

As development expands toward defined consortia and engineered strains, early alignment across CMC, nonclinical, clinical, and regulatory functions will be critical to converting microbiome science into reproducible and approvable therapies.

What Is a Live Biotherapeutic Product?

FDA’s guidance generally excludes filterable viruses, oncolytic bacteria, and products intended as gene-therapy agents. Recombinant LBPs may fall within the guidance, but genetically modified organisms are likely to raise additional considerations.

LBPs may include a single microbial strain, a defined cultured consortium, a complex donor-derived community, a preparation enriched for selected organisms or spores, or a genetically modified microorganism designed to perform a therapeutic function.

Not every microbiome-related product is an LBP. Bacteriophages, non-living microorganisms, microbial metabolites, purified proteins, and other postbiotic products may follow different pathways depending on composition and intended use. A microorganism may also appear in a food or dietary supplement, but a product intended to diagnose, prevent, mitigate, treat, or cure disease may be regulated as a drug and biological product. Clinical investigation generally requires an effective IND, while commercial distribution requires an approved BLA unless an exemption applies.

FDA’s June 2016 guidance, Early Clinical Trials With Live Biotherapeutic Products: Chemistry, Manufacturing, and Control Information, is a key reference for early development. It includes recombinant LBPs within its scope, with additional considerations, but excludes products intended as gene-therapy agents, oncolytic bacteria, and filterable viruses. Sponsors should therefore avoid applying it wholesale to bacteriophages or microbial gene-therapy vectors.

Evolution of the U.S. Regulatory Landscape

From FMT to Licensed LBPs

Early clinical development centered on fecal microbiota transplantation, or FMT, for recurrent Clostridioides difficile infection. Antibiotic exposure can disrupt the intestinal microbial community and permit C. difficile to proliferate. Transferring microorganisms from a qualified donor offered a way to restore microbial functions and reduce recurrence.

FDA considers fecal microbiota products intended to diagnose, prevent, treat, or cure disease to be biological products under Center for Biologics Evaluation and Research oversight. FDA also issued safety communications concerning transmissible pathogens and established a limited enforcement policy for specific FMT use.

The field reached a turning point with the approvals of Rebyota in November 2022 and Vowst in April 2023. Rebyota is administered rectally, while Vowst is an orally administered fecal microbiota spores, live product. Both are licensed biological products indicated to prevent recurrent CDI in adults following antibacterial treatment. These approvals demonstrated that complex donor-derived products can satisfy the BLA standard.

Enforcement Discretion Remains Limited

FDA’s November 2022 enforcement-policy guidance describes limited circumstances in which the agency intends to exercise enforcement discretion concerning IND requirements for FMT used to treat CDI that has not responded to standard therapies. The policy does not apply to FMT obtained from a stool bank and does not extend broadly to other diseases, commercial distribution, or direct-to-consumer use.

FDA reinforced these limits in a March 2025 warning letter to MG Infusions. The agency determined that the company’s FMT products were unapproved new drugs and unlicensed biological products. Promotion beyond nonresponsive CDI, direct shipment for self-administration, and inadequate donor screening placed the products outside the enforcement policy.

Trends Shaping the Next Generation

Moving Toward Defined Products

Donor-derived products established proof of concept, particularly in recurrent CDI. Development is now moving toward selected strains and controlled microbial consortia. Defined LBPs may offer tighter control over identity, strain ratios, starting materials, and manufacturing consistency, and may support a clearer mechanism of action.

Greater definition does not eliminate complexity. Sponsors must understand strain interactions during manufacture, storage, and administration. For engineered LBPs, additional questions may involve genetic stability, containment, persistence, shedding, and transfer of inserted material. FDA encourages sponsors of recombinant LBPs to seek product-specific advice before submitting an IND.

Moving from Composition to Function

Sequencing can show which organisms or genes are present, but composition alone may not establish therapeutic activity. Similar taxonomic profiles may perform differently because of viability, strain balance, metabolic state, processing, or storage. Development is therefore shifting toward functional measures such as pathogen inhibition, enzyme activity, metabolite production, bile-acid transformation, or other mechanism-linked effects.

Expanding Beyond Recurrent CDI

LBPs are being investigated in inflammatory, metabolic, immune-mediated, neurological, and oncological conditions. These indications may be harder because microbiome differences associated with disease do not establish causation. Sponsors must demonstrate that the selected LBP modifies a relevant pathway and produces a clinically meaningful effect. Engraftment, diversity, or biomarker changes may support the rationale but do not independently establish clinical benefit.

CMC: Controlling a Living Biological Product

For an LBP, CMC development must translate a living and potentially variable microorganism, or combination of microorganisms, into a consistently manufactured biological product. FDA’s LBP CMC guidance applies a phase-appropriate approach: each IND must contain enough information to assure identity, quality, purity, and strength, but the required depth depends on the development phase, dosage form, study duration, and available product knowledge. Controls directly related to participant safety should be established from the outset.

Define and Characterize the DS

For cultured LBPs, the drug substance generally consists of unformulated live cells derived from a controlled cell-bank system. FDA recommends documenting the organism’s biological name and strain designation, source, culture and passage history, phenotype, genotype, relevant functional attributes, and any intentional genetic modification. Identity should be established at the species and strain levels using complementary methods. Characterization should also address antimicrobial susceptibility, transferable resistance, toxin or virulence attributes, genetic stability, and translocation risk, where applicable.

Master and working cell banks should be characterized for identity, purity, adventitious agents, functional markers, passage history, and storage stability. For donor-derived LBPs, equivalent emphasis should be placed on donor eligibility, pathogen testing, collection controls, quarantine, requalification, and end-to-end traceability.

Control the Manufacturing Process and DP

The IND should include a manufacturing flow diagram identifying raw materials, equipment, processing areas, in-process tests, holding steps, and time and temperature limits. Variables such as media composition, growth conditions, harvest timing, oxygen exposure, cryopreservation, formulation, and thawing may affect viability, phenotype, purity, strain balance, or biological activity. Contamination and cross-contamination controls should distinguish intended organisms from extraneous microorganisms, particularly in multiproduct facilities. FDA’s Phase 1 CGMP guidance supports proportionate early controls while maintaining quality oversight necessary to protect participants.

Drug-product specifications should be scientifically justified and progressively refined using manufacturing, stability, nonclinical, and clinical data. Depending on the product, release testing may include identity, viable count, strain ratio, microbial purity, specified pathogens, potency, physical attributes, residual materials, and container-closure performance. For quantitative assays, actual numerical results should be reported rather than only pass-or-fail conclusions.

Establish Potency, Stability, and Comparability

Viable count may measure strength but does not necessarily demonstrate potency. Sequencing may describe composition without confirming function. A potency strategy may therefore combine viable count with a mechanism-linked assay, such as pathogen inhibition, enzyme activity, metabolite production, or another relevant biological response. Potency development should begin early enough to support release, stability, comparability, pivotal studies, and process validation.

Stability studies should evaluate viability, potency, purity, strain balance, genetic or phenotypic stability, shipping conditions, temperature excursions, preparation, and in-use handling. Analytical methods may be phase-appropriate initially but should mature toward validation following FDA’s analytical-method guidance.

Comparability should be planned before changes to cell banks, donors, media, scale, facilities, formulation, analytical methods, or storage. The assessment should consider identity, composition, potency, purity, antimicrobial-resistance profiles, and stability. FDA’s process-validation guidance supports a lifecycle approach that progresses from process design to qualification and continued verification.

Nonclinical Development: A Product- and Risk-based Strategy

Nonclinical development for an LBP should establish that the proposed clinical study can proceed with an acceptable level of risk. FDA’s LBP guidance requires adequate pharmacology and toxicology information from animal or in vitro studies to support the proposed trial. Studies using the intended clinical formulation or its individual components are generally more informative than data from similar or incompletely characterized products.

Unlike conventional drugs, LBPs may persist, replicate, interact with resident microorganisms, or cross a compromised mucosal barrier. A risk-based nonclinical program should therefore reflect the organism, product composition, route, dose, treatment duration, patient population, manufacturing process, and prior human experience. FDA recommends addressing general toxicity, target organs or systems of toxicity, dose and duration relationships, pharmacological activity, and, where relevant, reproductive or genetic toxicity associated with product components.

Characterize Risks Specific to Living Organisms

Product characterization should guide the nonclinical strategy. FDA’s guidance specifically recommends determining antimicrobial susceptibility and assessing whether resistance may transfer from the product strain to relevant microbial flora. Where mucosal translocation is a potential safety concern, a reproducible assay should be considered, preferably using an appropriate animal model. For attenuated or genetically modified organisms, additional product-specific considerations may include the stability of attenuation or genetic modification, persistence, shedding, and the potential consequences of reversion or genetic transfer.

Select Models That Answer Defined Questions

Model selection should be scientifically justified. Differences in microbiome composition, anatomy, diet, immune status, and antibiotic exposure can limit the relevance of conventional animal species. Specialized models, including germ-free, gnotobiotic, humanized-microbiota, disease-specific, or immunocompromised models, may be useful when they address a defined question. Model selection should be based on scientific relevance rather than routine availability.

Where an animal model has limited predictive value, the package may integrate in vitro assays, genomic and phenotypic characterization, virulence testing, antimicrobial-resistance analysis, published literature, and relevant prior human experience. The tested material should be representative of the proposed clinical product. Differences in strain, formulation, viable count, manufacturing process, purity, or route should be identified, and their effect on the applicability of the findings assessed.

Integrate Nonclinical Findings with CMC and Clinical Controls

Nonclinical development should not operate independently of CMC. A resistance gene, toxin-associated sequence, impurity, or unexpected phenotype identified during characterization may trigger targeted safety studies. Conversely, nonclinical findings may support new release specifications, manufacturing controls, donor exclusions, dose limits, eligibility criteria, safety monitoring, or stopping rules.

Because LBP risks are product-specific, early FDA interaction is valuable. Pre-IND discussions can align the proposed studies, model relevance, tested material, and use of published or prior human data with the intended clinical population and trial design. FDA’s Biologics Guidance collection provides additional cross-cutting resources that may apply according to product type and development stage.

Rebyota and Vowst: Two Approval Pathways, Shared Regulatory Lessons

The approvals of Rebyota and Vowst established that complex, donor-derived fecal microbiota products within the broader LBP field can satisfy the U.S. BLA standard. Both products are indicated to prevent recurrence of Clostridioides difficile infection in adults following antibacterial treatment for recurrent CDI. Neither product is approved to treat active CDI. Their differences in composition, processing, dosage form, and administration demonstrate why FDA applies a product-specific development and control strategy.

Regulatory Lessons for LBP Developers

Together, the approvals provide four important lessons:

• • Use a product-specific strategy. Differences in composition, processing, dosage form, and administration require tailored CMC, clinical, and risk-control programs.
  • Keep the indication precise. Rebyota and Vowst are approved to prevent recurrent CDI after antibacterial treatment, not to treat active CDI or support broader microbiome-restoration claims.
  • Link processing to product control. A rectal microbial suspension and an oral spore preparation require different specifications, stability programs, and administration controls.
  • Manage residual donor-related risk. Screening and testing reduce but do not eliminate the risks of transmissible agents and food allergens, requiring traceability, labeling, and pharmacovigilance.

Preparing for the Next Generation

1. Confirm product classification and seek early FDA feedback.
2. Define the product and link critical quality attributes to safety and activity.
3. Develop potency and comparability strategies before pivotal development.
4. Integrate pathogen and antimicrobial-resistance controls across CMC and nonclinical programs.
5. Maintain effective oversight of manufacturers, laboratories, donor organizations, and other critical vendors.