Aseptic preparation of an injectable suspension from the live S. suis ST171 bacterial concentrate requires a closed low-shear environment from bulk harvest to final filling. The bulk biomass is washed by tangential flow filtration against 10–50 mM potassium phosphate buffer at pH 7.0–7.4 to remove spent culture medium. The washed concentrate is adjusted by optical density and viable plate count to a target release titre of ≥5.0 log10 CFU per 2.0 mL dose. The suspending vehicle is prepared separately and contains 5–10% w/v sucrose or sorbitol for isotonicity and membrane stabilization, 0.5–1.0% w/v hydrolyzed gelatin as a protective colloid, and residual buffer salts. The concentrate is mixed into the vehicle in a jacketed stainless-steel vessel held at 2–8°C using a bottom-mounted low-shear impeller at 25–50 rpm; high-shear rotor-stator mixers are not used because they generate local adiabatic heating and fluid shear that reduce viable count. The finished bulk suspension is filled into 20 mL Type I glass vials under ISO 14644-1 Class 5 unidirectional airflow within a restricted-access barrier system or isolator. Terminal sterilization is not possible because the live bacterial cells are retained by 0.2 µm membranes; therefore, all liquid components are pre-sterilized by autoclaving or filtration and confirmed before aseptic addition. The final injectable is a slightly turbid suspension stored at 2–8°C. Freezing is prohibited because ice crystal formation disrupts the bacterial cell wall and capsular material.

What Happens When the Lyophilization Shelf Exceeds the Collapse Temperature of the ST171 Cake?

In lyophilized powder manufacture, the dominant process conflict is the mismatch between the heat supplied for ice sublimation and the low maximally freeze-concentrated glass transition temperature of the suspending matrix. A sucrose-based vehicle without a crystalline bulking agent typically exhibits a Tg' near -32°C to -35°C. If the product temperature at the sublimation front rises above this threshold, the freeze concentrate undergoes viscous flow, the cake collapses, and residual moisture becomes chemically bound in a shrunken, occluded structure. For ST171-specific formulations, published Tg' data are limited; each batch of bulk concentrate must be characterized by modulated differential scanning calorimetry after annealing. A production-scale cycle for such a live bacterial suspension commonly includes freezing on a shelf at -45°C for 3 h, an annealing hold at -10°C for 2 h to crystallize any mannitol present, primary drying at a shelf temperature of -25°C and chamber pressure of 0.07–0.13 mbar for 20–30 h, and secondary drying at +20°C until residual moisture is ≤2.0% w/w by Ph. Eur. 2.5.12. Shelf-mapping studies in a 400 L lyophilizer have shown that edge-vial product temperatures can exceed center-vial temperatures by 1.5–3.0°C during primary drying, which is enough to shift a marginal formulation into collapse and produce batch non-uniformity. Vials are therefore fully nested without perimeter blanking, and the thermal fluid inlet temperature is ramped rather than stepped to avoid overshoot. The dried cake is a white to off-white porous plug with residual moisture of 1.0–3.0% w/w. Above 3.0% w/w, lipid and protein components of the bacterial membrane show accelerated oxidative damage during storage; below 1.0% w/w, desiccation injury may be observed for some live bacterial strains, so the moisture specification is treated as a bilateral design space rather than a simple upper limit.

Mass administration through drinking-water systems shifts the process concern from aseptic filling to water chemistry and short-term stability in a non-sterile environment. The lyophilized ST171 powder or concentrated liquid suspension is first diluted into 5–10 L of non-chlorinated water in a plastic or stainless-steel stock container. Free chlorine in the header tank must be reduced to <0.03 mg/L before vaccine addition; sodium thiosulfate is dosed at approximately 16 mg/L for each 1 mg/L free chlorine present, based on the stoichiometric reduction of hypochlorous acid. The water is buffered to pH 7.0–7.4 with 10–20 mM phosphate; citric acid buffers are avoided because they can lower pH below 6.5 and reduce bacterial viability. The vaccine suspension is stirred gently with a food-grade paddle or a low-speed axial-flow impeller at 10–20 rpm and is distributed through a proportional medicator set to deliver the calculated dose over 4–6 h. Pulsating diaphragm pumps and high-velocity recirculation loops are avoided because hydrodynamic stress can reduce viable count. Dissolved copper and iron in galvanized or copper pipes should be <0.1 mg/L and <0.3 mg/L, respectively, because transition-metal ions accelerate membrane peroxidation. After dilution, the suspension is used within 2–4 h at room temperature; if water temperature exceeds 25°C, the holding time is reduced or the chiller is set to ≤10°C. The terminal finished product is a ready-to-use drinking-water solution for pigs; residual lines are flushed after each dosing event to prevent biofilm formation.

Pellet Binder Compatibility and Post-Pelleting Viability Loss in Swine Starter Prestart Diets

Dry oral delivery as a feed premix or granule is used when individual injection is not practical and where drinking-water administration is not possible due to intermittent water access. The primary formulation task is to maintain water activity below 0.20 and to avoid direct contact between the live bacterial cells and mineral acids, aldehyde-based pellet binders, or high-pH inorganic carriers. Lactose monohydrate and dextrose monohydrate are preferred carriers because they have low hygroscopicity and do not generate reactive fines; wheat middlings and limestone are excluded because of abrasive surfaces and alkaline pH. The lyophilized ST171 powder is blended with carrier in a double-ribbon blender at 20–40 rpm for 10–15 min using geometric dilution to a final batch size of 500 kg. If a free-flowing granule is required, dry granulation by roller compaction at roll pressures 5–10 kN/cm is preferred over wet granulation. Fluid-bed wet granulation with aqueous binder solutions at inlet air temperatures above 35°C can reduce viability by several log10 because the organism is fully hydrated and heat-sensitive. Incorporation into finished feed must avoid the conditioning and pelleting sequence; a pellet mill die temperature of 70–85°C for 30–60 s is incompatible with non-sporulated live bacteria and commonly causes 3–5 log10 CFU reduction. Where pelleted feed is required, the premix is sprayed onto cooled pellets at ≤35°C with a rotary atomizer or applied as a coating with a food-grade oil-based suspension. The medicated feed is then sampled immediately after mixing and after 30, 60, 90 days at 25°C/60% RH to confirm viable count. Antimicrobial feed additives including tetracyclines, sulfonamides, and ionophores must be omitted from the ration during administration because they suppress or kill the vaccine organism.

Enteric-Coated Capsule Feasibility for Oral Mucosal Exposure

Tablet and capsule presentations are not the primary registration route for a live streptococcal vaccine because Streptococcus suis colonizes the palatine tonsils and upper respiratory mucosa rather than the distal intestine. Direct compression of lyophilized powder with microcrystalline cellulose and croscarmellose sodium at compression forces of 10–20 kN has been associated with viability loss of 1–3 log10 in non-sporulated bacterial powders due to frictional heat and particle-particle shear; published data for ST171 in a direct compression matrix are limited. Hard gelatin capsules may be filled at ≤25°C and ≤20% RH using a low-speed capsule filler, then banded and enteric-coated with Eudragit L 30 D-55 aqueous dispersion. However, water uptake during aqueous coating can reduce the viability of the lyophilized organism, so a protective subcoat of hydroxypropyl methylcellulose and an internal desiccant are necessary. The rationale for enteric coating is to protect the live organism from gastric acid and deliver it to the oropharyngeal and tonsillar lymphoid tissue via the oral cavity, but current published data for this specific configuration are limited.