The live Bacteroides fragilis, Streptococcus faecalis and Bacillus cereus preparation is a veterinary-grade active pharmaceutical ingredient consisting of three viable bacterial strains co-lyophilized in a cryoprotectant matrix. The label name Streptococcus faecalis is a deprecated heterotypic synonym for Enterococcus faecalis; current strain identity should be confirmed by 16S rRNA gene sequencing or MALDI-TOF MS before use in manufacturing. The API is typically supplied as an off-white to tan free-flowing powder with a water activity below 0.20 and a residual moisture content below 5.0% by Karl Fischer titration. Manufacturer-specific model identifiers, rather than a single compendial monograph, designate the strain ratio and target potency; a supplier code such as LBFSFBC-VET10 may be used but has no regulatory harmonization. The product is intended for further processing into tablets, capsules, powders, granules, premix and oral solutions; injectable use is restricted to aseptic manufacture because live bacterial cells cannot be terminally sterilized by steam, dry heat, gamma irradiation or membrane filtration without loss of viable count.

Why Does the Bacteroides fragilis Component Dictate Cold-Chain Handling?

Bacteroides fragilis is an obligately anaerobic, Gram-negative, non-spore-forming rod. Unlike the Bacillus cereus spore fraction, it cannot survive atmospheric oxygen for prolonged periods; lyophilized cells retain viability only if the powder is kept dry and cool. In a production-scale freeze dryer with a shelf area of at least 1 m² and chamber pressure maintained between 0.2 mbar and 0.5 mbar, the product temperature during primary drying must remain below the glass transition temperature of the maximally freeze-concentrated amorphous phase, typically below −30°C for sucrose/trehalose-based formulations. Cake collapse, microcollapse or incomplete sublimation of ice increases residual moisture, lowers the glass transition of the dried matrix and accelerates loss of B. fragilis viability during storage; batch records should record product temperature at the shelf edge and center because shelf-edge radiation can elevate vial or tray temperature by 2°C to 5°C. Storage at 2–8°C in sealed aluminum foil pouches with desiccant is applied when the B. fragilis fraction is present at high count; products containing only Bacillus spores may be stored at ambient temperature, but this triple-strain preparation should not be treated as an ambient-stable spore probiotic.

Freeze-drying cycles are commonly developed using a lyophilizer with stainless steel shelves, an external condenser and a capacitance or Pirani gauge for chamber pressure. Primary drying endpoint is determined by comparative pressure measurement or dew-point measurement; removing the product based on a fixed 48 h or 72 h cycle without verifying sublimation endpoint risks collapse in the center trays. Batch-to-batch variation in B. fragilis viable count is observed when the same freeze-dry program is used for different fill depths or container types; vial fill depth above 1.0 cm increases drying resistance and can leave residual moisture above 3.0% at the bottom of trays. Production records should therefore include fill depth, shelf temperature ramp rate, chamber pressure and condenser temperature, not only cycle time.

Blending with microcrystalline cellulose, lactose monohydrate or mannitol is performed in a low-shear V-blender or bin blender; high-shear granulation without cooling can increase product temperature above 30°C and should be avoided unless viability loss is confirmed below the manufacturer’s acceptance criterion. Compression into tablets requires a rotary tablet press fitted with low-humidity feed frames; relative humidity below 35% is applied when transferring lyophilized powder from sealed containers to dies. Tablet compression force is not a universal limit but must be validated by comparing viable count before and after ejection; a processing loss above 0.5 log₁₀ CFU per tablet is generally considered significant because it shifts the ratio of the three strains and alters the dose. Capsule filling on automatic dosators with dosator nozzles may generate local frictional heat; viability should be monitored at the beginning, middle and end of each production run. Excipients with reducing sugars, high hygroscopicity or antimicrobial preservatives are incompatible with live microbial API and should be excluded unless compatibility studies demonstrate no inhibition.

Dry granulation by roller compaction may be used to improve flow. Roller compaction pressure should be set at the minimum pressure necessary to produce acceptable compact hardness; viability loss after dry granulation is often 0.1–0.5 log₁₀ CFU/g, but published data for this specific configuration is limited. The main failure mode observed at production scale is hygroscopic excipient carryover and moisture pickup during transfer; therefore open handling time should be limited and controlled by loss-on-drying checks at the beginning and end of the batch. During batch release, identification of each strain is performed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry or by species-specific polymerase chain reaction. Because B. fragilis is anaerobic, sample preparation must minimize oxygen exposure; sealed anaerobic jars or an anaerobic workstation with gas composition 85% N₂, 10% H₂ and 5% CO₂ is employed for dilution and plating. Counts are expressed in colony-forming units per gram of dried API, not per tablet or per dose, to avoid confusion during formulation. If a batch fails per-strain viable count, re-blending with a higher-potency lot is generally not accepted without a documented deviation and stability evaluation because aerobic steps during rework reduce the B. fragilis fraction.

When the API Is Compounded into Injections or Sterile Solutions

Injectable presentations containing live bacterial cells are not conventional feed probiotics and fall under sterile veterinary medicinal product requirements. The API cannot be sterilized after formulation because moist heat at 121°C, dry heat at 160°C, gamma irradiation at typical 25 kGy doses and sterilizing-grade filtration all remove or inactivate vegetative cells and spores. Aseptic processing is therefore mandatory; the lyophilized API is reconstituted in a sterile vehicle within an ISO 5 laminar-flow zone, using components that have been pre-sterilized by a method compatible with the formulation. Endotoxin control is critical because B. fragilis is Gram-negative; parenteral-grade material should be assigned a bacterial endotoxin limit, for example less than 0.5 EU/mg when a parenteral route is proposed, and tested by Limulus amebocyte lysate methods such as Ph. Eur. 2.6.14 or USP <85>. Particulate matter in injectable suspensions must meet the relevant veterinary or pharmacopoeial particle limits after reconstitution; the live bacterial cells themselves contribute turbidity and particle counts, so methods must differentiate microbial cells from foreign particulate matter. Sterility testing of live bacterial preparations is inherently challenging because the API contains viable microorganisms; compendial sterility tests such as USP <71> cannot be applied directly. Instead, bioburden testing of raw materials, environmental monitoring during aseptic filling and absence of specified pathogens under ISO 6579-1, ISO 16649-2 and ISO 6888-1 are used as release controls.

Oral solutions and drinking-water premixes are prepared by dispersing the API in non-chlorinated water at 20–25°C; free chlorine above 0.2 mg/L and acidic pH below 4.0 reduce viability within minutes and should be chelated or buffered before reconstitution. For farm-scale premix, the lyophilized powder is first mixed with a carrier such as lactose or dextrose in a ribbon mixer or paddle mixer; mixing time is determined by homogeneity testing of an added tracer, not by fixed time alone. The final solution should be used within 2 h after reconstitution unless stability data support a longer hold time. Published data for this specific three-strain combination are limited; extended hold-time claims require strain-specific viability studies under simulated field water conditions.

Strain-Specific Viability, Selective Enumeration and Matrix Interference

Viable count determination for the three strains requires separate selective cultivation because the organisms differ in oxygen tolerance, Gram reaction and growth temperature. Bacteroides fragilis is enumerated under anaerobic conditions on Bacteroides bile esculin agar at 35–37°C for 48 h; Enterococcus faecalis is enumerated on Enterococcosel agar at 35–37°C for 24–48 h under aerobic conditions; Bacillus cereus spores are heat-shocked at 70°C for 15 min and plated on mannitol egg yolk polymyxin agar at 30°C for 24–48 h. Excipients can interfere with recovery: mannitol, starch and magnesium stearate may create osmotic or hydrophobic barriers. Method suitability must be evaluated by matrix recovery tests using low-level inocula and the acceptance range 50–200% recovery as described in USP <61>, USP <1223> or ISO 7218. Batch-to-batch variability in the B. fragilis fraction tends to be larger than in spore-forming Bacillus preparations because vegetative cells are more susceptible to oxygen ingress through foil seals and temperature excursions during transport; release data should be trended by control chart, not evaluated as pass/fail alone.

This triple-strain preparation differs from single-strain Bacillus subtilis or Bacillus licheniformis products in four operational aspects. First, it contains one Gram-negative anaerobe, one Gram-positive facultative anaerobe and one spore-former; therefore its cold-chain and oxygen barriers are more demanding. Second, Bacillus cereus is a spore-forming species that can carry enterotoxin and cereulide genes; the API strain should be documented as non-toxigenic by absence of nheA, hblD and cesB genes using PCR. Third, Enterococcus faecalis from some sources can harbor acquired antibiotic resistance; because it is grown as a deliberate microbial API, the strain should be screened for vanA, vanB, ermB and other clinically relevant resistance determinants. Fourth, the Gram-negative B. fragilis cell wall contributes endotoxin; this is generally not a concern in oral powders or premixes but becomes a specification matter for injectable forms. Published data for this specific configuration are limited; users should not extrapolate storage, heat tolerance or antimicrobial susceptibility from products that contain only Bacillus spores or yeast.

Documentation for veterinary-grade live microbial API is generally compiled under VICH GL18 and regional requirements such as EU Regulation 2019/6 for veterinary medicinal products. A certificate of analysis should report per-strain viable count, absence of specified pathogens, residual moisture, water activity, endotoxin for injectable grade, and the strain deposit accession numbers. Resistance to antimicrobials should be interpreted according to the European Food Safety Authority guidance on bacterial susceptibility testing; absence of acquired resistance genes is more relevant than absolute minimum inhibitory concentration values because the product is an intentionally added viable microorganism. Users should request stability data generated under the intended storage condition; accelerated data alone are not sufficient to justify an extended shelf life.