What does it take to develop and manufacture a Microbiome / Live Biotherapeutic drug?
Microbiome and live biotherapeutic products (LBPs) break the usual drug-development mold because the active ingredient is alive. Whether your candidate is a single defined strain, a rationally designed consortium, or a donor-derived community, the program runs on bacteriology and anaerobic handling rather than the synthesis and purification logic that governs small molecules or antibodies. That difference reaches into every stage: how you characterize the drug substance, how you prove it is what you say it is, and how you keep it viable from a frozen master cell bank to a patient dose.
On the discovery and preclinical side, the work that matters is strain isolation and banking, whole-genome sequencing and strain identity, metagenomic and 16S community profiling, and functional characterization (which metabolites the strain produces, how it colonizes, how it behaves in gnotobiotic or humanized-microbiome animal models). Antibiotic-resistance and virulence-factor screening is not optional for a live organism going into people, and the FDA expects a clear safety rationale for each component. GLP toxicology still applies, but the study design has to account for a colonizing, replicating product rather than a fixed dose that clears.
The manufacturing side is where specialist CDMOs separate themselves from generalists most sharply. Many LBP strains are strict anaerobes, so fermentation, harvest, formulation, and fill-finish have to happen under controlled low-oxygen conditions that a conventional biologics suite simply does not have. Then there is lyophilization or other stabilization to keep cells alive on the shelf, enteric or delayed-release encapsulation so the dose survives stomach acid and reaches the gut, and cold-chain logistics throughout. Release testing is its own discipline: viable cell count (CFU or flow-based viability), potency tied to a relevant biological function, strain identity confirmation, and a sterility concept adapted to a product that is deliberately full of live bacteria. A CDMO that does excellent monoclonal antibody work can be the wrong partner here, because none of that anaerobic, live-cell tooling transfers.
How do you choose a CRO or CDMO for Microbiome / Live Biotherapeutic?
The single best filter is whether the supplier has actually made and tested live bacterial products before, ideally ones that reached the clinic. Microbiome manufacturing is unforgiving, and a track record on real LBP programs tells you more than any capability brochure. Walk through the checklist below before you shortlist, and ask for specifics rather than reassurances on each point.
- Relevant platform and track record: prior work on defined strains, consortia, or donor-derived products, and named LBP programs they have taken through IND or into clinical supply.
- Anaerobic and live-cell capability: controlled low-oxygen fermentation, harvest, and fill-finish, plus stabilization (lyophilization or equivalent) and the encapsulation needed for gut delivery.
- Analytical and GxP depth for this modality: validated viability and CFU methods, potency assays tied to function, strain identity by sequencing, residual-DNA and bioburden controls, and a sterility concept that fits a live product, all under GLP and GMP.
- Capacity and scale: ability to move from research banks to GMP master and working cell banks, and to scale fermentation without losing viability or shifting the strain ratio in a consortium.
- Regulatory experience: familiarity with the FDA's LBP guidance and CMC expectations, EMA equivalents if you are filing in Europe, and a clean inspection history on live-biologic suites.
- IP and data ownership: clear terms on who owns strains, banks, process know-how, and the characterization data, especially when the supplier contributes a manufacturing platform.
- Cold chain and supply continuity: stability data behind the storage claim, and a credible answer for backup capacity if a single anaerobic suite goes down.