For direct compression of canagliflozin hemihydrate at the 100 mg and 300 mg dose levels, the powder properties of the active pharmaceutical ingredient determine whether the blend can be compressed on a high-speed rotary press without unacceptable weight variability. The micronised API is often supplied with a particle-size distribution measured by laser diffraction under USP <429>; because canagliflozin exhibits poor aqueous solubility, the particle-size specification is not merely a handling parameter but a dissolution control point. A production-scale batch record for a direct-compression trial typically begins with a pre-blend of the API and a portion of microcrystalline cellulose in a 600 L bin blender at 12 rpm for 10 min, followed by addition of sieved anhydrous lactose, croscarmellose sodium, and the remainder of the microcrystalline cellulose. The full blend is mixed for 20 min, sampled at 10 locations with a sample thief, and tested for blend uniformity; if the relative standard deviation of content exceeds 5.0%, the batch is not advanced to lubrication because magnesium stearate will not correct a segregation defect. Lubrication is performed by passing magnesium stearate through a 0.5 mm screen and mixing for 3 min at 12 rpm; over-lubrication beyond 5 min produces a hydrophobic surface film that lowers tensile strength and increases friability to the point of edge failure during film coating. Compression on a 16-station or 45-station rotary press fitted with 9.5 mm round or 19.0 mm × 9.5 mm oval tooling is controlled by setting the main compression force to produce tablet hardness in the range of 80–120 N; the force-versus-hardness relationship varies with API lot and residual moisture, so instrumented compression data must be generated for each incoming API lot. After compression, core tablets are film-coated with an aqueous polyvinyl alcohol-based coating system to a weight gain of 2.5–4.0%, with pan inlet air temperature held below the point at which the thiophene-containing molecule undergoes oxidative discoloration. The finished tablet is tested for disintegration under USP <701> in 900 mL of 0.1 M HCl at 37 °C and for dissolution under USP <711>; because published data for this specific direct-compression configuration are limited, the route should be challenged with paired dissolution and content-uniformity studies before scale-up.

Commercial immediate-release canagliflozin tablets are not necessarily manufactured by direct compression; the decision between direct compression and wet granulation is driven by the API particle-size distribution, excipient compatibility, and available press time. Direct compression is operationally simpler because it eliminates the drying endpoint and reduces the number of unit operations, but it is vulnerable to feed-frame segregation when the API bulk density falls below 0.3 g/mL and the diluent bulk density exceeds 0.5 g/mL. One approach to stabilise the feed frame is to substitute spray-dried lactose with a spherical particle shape for angular anhydrous lactose; the spherical excipient reduces interparticulate friction and keeps the blend moving through the feed shoe. However, anhydrous lactose may be preferred when the formulation is moisture-sensitive because spray-dried lactose can contain higher free moisture. The final choice should be justified by a design-of-experiment study relating filler grade, lubricant time, and press speed to content uniformity and dissolution; no single published range is sufficient for all API lots.

Why Does Wet Granulation Shift the Dissolution Failure Mode for High-Dose Canagliflozin Tablets?

High-dose tablets containing 300 mg of canagliflozin hemihydrate place the formulation near the upper limit of direct-compression feasibility because the API can constitute approximately 40–50 wt% of the core; wet granulation is therefore selected to densify the powder and improve content uniformity. The granulation operation begins with a low-shear planetary mixer charged with canagliflozin, lactose monohydrate, microcrystalline cellulose, and croscarmellose sodium; a binder solution of hydroxypropyl cellulose in purified water at 2–4 wt% solids is added while the main blade runs at 60–100 rpm. The endpoint is determined by both amperage and visual mass consistency, but the more critical parameter is the subsequent drying step. A fluid-bed dryer with inlet air temperature of 50–60 °C reduces the moisture content to 1.5–2.5% loss on drying; overdrying below 1.0% shifts granule friability upward and generates fines that segregate during tableting, while residual moisture above 3.0% can raise the risk of punch filming and slow dissolution. The dried granules are milled through a 1.25 mm screen and blended with extragranular croscarmellose sodium before lubrication with magnesium stearate. Dissolution testing per USP <711> in 900 mL of 0.75% sodium lauryl sulfate in 0.1 M HCl at 75 rpm paddle speed is used to detect the characteristic failure mode of this route: over-wetting causes the binder to form a dense film on the API surface, delaying release beyond 15 min even though the tablet disintegrates within specification. Granule particle-size distribution should be held with D50 between 150 µm and 250 µm; published canagliflozin-specific particle-size dissolution limits are limited, so each granulation batch must be qualified by paired disintegration and dissolution profiles rather than by sieve data alone.

When the granulation is scaled from a 25 kg high-shear bowl to a 600 kg production bowl, the water addition rate and final massing time require re-optimisation because the larger bowl generates more shear heat and accelerates binder hydration. The production bowl should be jacketed with chilled water to hold the granulate below 35 °C during massing; otherwise, localised overwetting at the spray nozzle forms dense agglomerates that survive milling and produce superpotent tablets. After drying, the granules are compressed on a rotary press at 30–60 rpm turret speed with pre-compression force set at 3–6 kN to expel air and prevent capping. The terminal dosage form is a film-coated oval tablet with label strengths of 100 mg and 300 mg, tested for assay by USP <621>, content uniformity by USP <905>, and dissolution by USP <711>. The specification must also control degradation products under ICH Q3B(R2); the hemihydrate form must be verified by X-ray diffraction because exposure to high-humidity granulation can promote form conversion and alter dissolution.

A Low-Shear Trituration Fills Hard Gelatin Capsules Without Excipient Segregation

In hard gelatin capsule manufacture for canagliflozin hemihydrate, a dry trituration in a V-blender or bin blender is acceptable only when the target capsule fill weight remains below 300 mg; above this fill weight, the powder bed becomes difficult to dose uniformly on an automatic capsule machine. The recommended order of addition is to pre-blend the sieved API with an equal mass of lactose monohydrate for 10 min, add the remaining lactose and silicified microcrystalline cellulose, mix for 20 min, then add magnesium stearate through a 0.5 mm screen and lubricate for 2–3 min. This short lubrication interval is critical because the low bulk density of a dry capsule blend permits magnesium stearate to distribute rapidly; longer mixing reduces blend compressibility and can cause insufficient powder retention in the dosator. Automatic capsule filling on a 4000–8000 capsule/h dosator machine requires a powder bed height set to 60–70% of the dosator travel, and the dosing chamber spring pressure is adjusted so that the fill-weight variation remains within ±5% of target. Capsules are checked for lock length, leakage, and dissolution in 900 mL of pH 6.8 phosphate buffer with 0.2% sodium lauryl sulfate using USP <711> basket apparatus at 100 rpm. The terminal dosage form is a size 0 or size 1 capsule with a printed identification band; if the capsule shell contains gelatin, storage below 25 °C and 60% RH is required to prevent cross-linking that alters dissolution. For capsule-specific quality, USP <905> covers weight variation, and ICH Q3D limits elemental impurities when the API is sourced from palladium-catalysed intermediates; no separate harmonised monograph exists for canagliflozin capsules, so the tablet monograph is adapted with justification under ICH Q6A.

Capsule-specific dissolution failures are often traced to powder stratification during machine hopper residence, particularly when the hopper is refilled during long runs. The vibration of the dosing disc can sort the blend by density, causing a gradual increase in API content in the lower hopper segments. To reduce this risk, the hopper level should be maintained above 40% capacity, and the capsule machine should be stopped after 8 h for blend re-verification. Published data for canagliflozin capsule formulation stability are limited, but the general requirement under 21 CFR 211.65 is that equipment surfaces contacting the blend be non-reactive; stainless steel 316L is specified because canagliflozin can coordinate trace metal ions under acidic microenvironments. Terminal capsule product should also be evaluated for microbial limits under USP <61> and USP <62> if the capsule is intended for oral administration without further processing.

When canagliflozin hemihydrate is co-formulated with metformin hydrochloride in a fixed-dose combination, the two actives are not pre-blended into a single granulation because metformin hydrochloride is hygroscopic and acidic, and the resulting local pH can accelerate degradation of the canagliflozin thiophene moiety. The bilayer tablet process separates the canagliflozin layer and the metformin layer until final compaction. The canagliflozin layer is prepared by wet granulation with microcrystalline cellulose, croscarmellose sodium, and hydroxypropyl cellulose, then dried to a loss on drying of 1.5–2.5%. The metformin layer is prepared separately by high-shear granulation with povidone or pregelatinised starch, dried, and milled; because metformin hydrochloride has high aqueous solubility and a dense crystalline habit, its granulation endpoint is judged by amperage rather than by moisture alone. Bilayer compression is carried out on a rotary tablet press equipped with pre-compression and main compression stations; the first layer is pre-compressed at 5–10 kN, and the second layer is added to the die and compressed at 15–25 kN to achieve a core hardness of 120–180 N. Weight control for each layer is monitored by press force and by automatic weight-control systems; deviations beyond ±5% of layer target can shift the ratio of the two actives outside the label claim. The terminal product is a non-scored oval film-coated tablet; dissolution of both actives is measured in 900 mL of pH 6.8 phosphate buffer or 0.1 M HCl using USP <711>. Metformin releases faster than canagliflozin, so the specification includes separate time points for each active; the canagliflozin release profile is the more sensitive indicator of bilayer interface integrity because a fracture at the interface can expose a dense metformin face that slows canagliflozin release below the labelled rate.

Where a patient cannot ingest an intact tablet or capsule, an oral granule intermediate is formulated to disperse in water without the need for a disintegration mechanism inside a compact. Instead of a high-shear binder, the API is blended with mannitol, xylitol, and low-substituted hydroxypropyl cellulose, then granulated with a 5–10 wt% povidone K30 solution in a fluid-bed granulator. The inlet air temperature is maintained at 45–55 °C and the spray rate is set to 10–20 g/min/kg to produce open, porous granules with a tapped density below 0.6 g/mL; denser granules sink in water and form a non-dispersible sediment. The dried granules are screened through a 1.0 mm sieve and dusted with colloidal silicon dioxide at 0.2–0.5 wt% to prevent agglomeration during storage in high-density polyethylene bottles with child-resistant closures. Dispersibility is evaluated by adding a unit dose to 200 mL of water at 20 °C; the suspension must pass through a 0.710 mm sieve with no residual lumps within 2 min. Content uniformity of the dispensed dose follows USP <905>, and microbial limits follow USP <61> and USP <62>. The terminal product is a single-dose or multiple-dose oral granule for solution or suspension, but published data for canagliflozin-specific oral granule formulations are limited, so formulators should generate their own dose-proportionality data under ICH Q8(R2) rather than assuming tablet bioequivalence.

When Terminal Sterilisation Encounters Solubility Limits in an Injectable Canagliflozin Presentation

Injectable presentations of canagliflozin present a distinct formulation problem because the molecule is not freely soluble in water and no harmonised injectable monograph has been published; absence of a compendial reference means that a proposed formulation must be developed under ICH Q8 with design-space verification. The first technical screen is solubility in parenterally acceptable solvents: if the API is dissolved in a co-solvent system of propylene glycol, ethanol, and water for injection, the final organic content must be justified by ICH M3(R2) and the 80 µg/day parenteral organic solvent threshold must be considered. Aqueous solubilisation via sulfobutylether-β-cyclodextrin is an alternative, but phase-solubility diagrams must be generated over pH 4–8 because canagliflozin is a weak acid; the degree of substitution and molar ratio of cyclodextrin to API directly affect the free fraction and the risk of precipitation upon dilution with saline. Terminal sterilisation by moist heat at 121 °C for 15 min is preferred for drug products that withstand it; however, if the formulation contains a labile thiophene ring or heat-sensitive cyclodextrin complex, the sterilisation step may degrade the API and shift impurity profiles beyond the qualification threshold. In that case, aseptic processing through a 0.2 µm sterilising-grade filter in a Grade A/ISO 5 environment is required under 21 CFR 211.110 and EU GMP Annex 1. The filled vials must meet USP <1> for sterility, USP <85> for bacterial endotoxins, and USP <788> for particulate matter; closure integrity must be verified by dye ingress or vacuum decay per USP <1207>. The terminal dosage form is a single-use vial or pre-filled syringe containing a sterile solution or lyophilisate; however, because no commercial injectable canagliflozin product has been approved, published data for this specific configuration are limited, and the formulator should report physico-chemical stability data under ICH Q1A(R2) rather than extrapolating from oral stability.

If the completed formulation is lyophilised, the solution is aseptically filtered into depyrogenated vials and freeze-dried under a validated cycle; the sublimation shelf temperature is set below the collapse temperature of the formulation, and the chamber pressure is held between 50 mTorr and 200 mTorr. The resulting lyophilised cake must reconstitute within 2 min with water for injection to a solution free of visible particles larger than 10 µm; subvisible particle counts are controlled by light obscuration under USP <788>. For an injectable canagliflozin presentation, the critical process limits remain undetermined by public literature because no commercial injectable canagliflozin product is approved in the United States or the European Union; therefore, process parameters must be established under an investigational filing rather than by reference to a compendial monograph. The absence of an approved injectable reference also means that the stoichiometric ratio of any cyclodextrin excipient to API, the final pH adjustment range, and the terminal sterilisation cycle must be justified by degradation-kinetic experiments that identify the activation energy of the thiophene oxidation and the hydrolysis rate of the glucitol ether linkage. Without these data, an injectable dosage form cannot be considered robust under ICH Q9 risk-management principles.