What limits content uniformity in direct compression of a 0.5–5.0 wt% peptide API?

Low-dose oral tablet manufacture from small peptide APIs exposes a direct conflict between particle size reduction and powder flow. Micronised peptide fractions with d90 ≤ 20 µm and d50 ≤ 10 µm are typically required to satisfy USP <905> Uniformity of Dosage Units acceptance value ≤ 15.0 at a 0.5–5.0 wt% drug load, but these fractions exhibit bulk density below 0.45 g/cm³ and Carr index above 28, which promotes segregation during hopper discharge. On a 10-station rotary tablet press operated at 20–40 rpm with main compression force 6–12 kN and precompression force 3–5 kN, direct compression blends containing spray-dried mannitol, microcrystalline cellulose PH102, croscarmellose sodium, and 0.5–1.0 wt% sodium stearyl fumarate show acceptable tablet hardness of 50–90 N only if the peptide is geometrically diluted in three to five steps in a 600 L bin blender at 12 rpm for 20 min per step. Batch-to-batch variance is driven by peptide particle size span; spans above 2.0 have produced content uniformity failures at 2.5 wt% load in a 150 mg tablet core. Pre-drying of non-peptide excipients to loss on drying ≤ 2.0% w/w is required before blending because residual moisture above 2.0% accelerates deamidation of Asn-containing peptides during storage at 25°C/60% RH.

Dissolution testing with USP <711> Apparatus 2 at 50 rpm in 900 mL phosphate buffer pH 6.8 commonly requires Q = 75% at 45 min for immediate-release tablet specifications; peptide tablets with slow disintegration often fail because croscarmellose sodium at 2.0–3.0 wt% loses effectiveness when exposed to compression forces above 12 kN. A direct compression process window of 6–10 kN is therefore maintained for formulations with hardness 50–70 N to avoid over-compression-induced delayed disintegration. Process analytical technology using near-infrared spectroscopy in a 25 mm static cup can monitor blend uniformity with root mean square error of prediction below 1.5% w/w, but the calibration set must include production-scale samples from both hopper-top and hopper-bottom positions. Tablet press dust extraction above 5 m/s at the compression zone can selectively remove low-density peptide particles, shifting the in-process content uniformity by more than 2.0% relative to mean blend assay; baffled entablature covers and reduced extraction velocity of 1.5 m/s mitigate this loss. A 15-min blend lubrication with sodium stearyl fumarate is preferred over magnesium stearate because the latter can raise the contact angle of the tablet matrix and delay peptide release in 0.1 N HCl. Lactose monohydrate is excluded from this matrix because reducing sugars form Maillard adducts with primary amino groups at 40°C/75% RH during accelerated stability evaluation.

Capsule-Filling Dosator Pin Compression and Low-Fill-Weight Peptide Blends

Hard gelatin and HPMC capsules of size 3 or size 4 with fill weights of 100–150 mg are used for oral peptide APIs when tablet compression is not suitable because of high excipient sensitivity or low bulk density. On a dosator-type capsule filling machine operating at 70,000 capsules/h with dosing discs equipped with 3.5 mm diameter pins, the fill weight relative standard deviation must remain below 2.0% at a 100 mg target to pass Ph. Eur. 2.9.40 Uniformity of Dosage Units. Powder bridging inside the dosator chamber is a recurring failure mode when the peptide blend contains more than 15% w/w of particles below 75 µm; the remedy is not additional glidant but controlled agglomeration of the peptide fraction to a d50 of 100–150 µm via roller compaction or slugging at 1–3 kN. Colloidal silicon dioxide at 0.2–0.5% w/w is added only after geometric dilution because higher concentrations can reduce powder flow by increasing interparticle cohesion.

In-line capsule weight checkweighers with electrostatic capacitance sensors reject capsules deviating ± 3 mg from target at full speed, but low-dose peptide capsules with fill weight 100 mg and API content 1.0 mg require destructive content uniformity sampling per USP <905> every 15 min. A standard sampling plan of 20 capsules at start, middle, and end of the batch has revealed positional bias between dosator stations exceeding 4.0% assay when the powder bed height in the hopper drops below 20% of charge volume; therefore hopper level is maintained between 30% and 70% of capacity. HPMC capsules require storage below 35% RH during filling because shell moisture below 5.0% can cause brittleness and splitting at machine speeds above 90,000 capsules/h. If the peptide contains methionine, the capsule fill is protected by nitrogen overlay in the hopper and finished capsules are packaged with oxygen scavenger sachets because oxidative degradation at 40°C/75% RH can increase methionine sulfoxide content above 0.5% of label claim within 6 months.

When Spray-Dried Peptide Granules Are Fed to a High-Shear Rotary Press

When a peptide API is spray-dried with mannitol and trehalose at a 1:1:1 mass ratio, the resulting amorphous dispersion can be compacted on a high-shear rotary press only if the residual moisture is kept below 1.8% w/w and the inlet air dew point of the spray dryer is maintained at or below 4°C. The spray dryer is configured with a rotary atomizer at 12,000 rpm, a feed solids content of 15–25% w/w, an inlet temperature of 120–140°C, and an outlet temperature of 60–70°C; these conditions produce a d50 of 60–90 µm and a span of 1.5–2.0. Lactose monohydrate is not included in this matrix because reducing sugars form Maillard adducts with primary amino groups at the powder surface when exposed to 40°C/75% RH for 90 days. The spray-dried powder is then tableted on a 16-station high-shear rotary press with force feeder speed 30–50 rpm, main compression force 8–15 kN, and turret speed 35–50 rpm. A recurring defect on this line is edge chipping of the tablet when the powder electrostatic charge exceeds 2.0 kV/cm; the charge is suppressed by conditioning the powder in a 45% RH room for 2 h before loading.

Amorphous spray-dried peptide-mannitol-trehalose powder has a low glass transition temperature and can plasticize at tablet press temperatures above 35°C; if tablet press bearings raise die table temperature to 38°C, sticking and picking occur on the punch faces. Countermeasures include chilling the die table to 18–22°C and using external lubrication with magnesium stearate applied to the punch tips at 0.1–0.2 mg/punch. Open storage at 60% RH for 30 min can increase powder moisture by 1.5% w/w, shifting flow and compaction; containment with dry nitrogen purge below 20% RH is required between spray dryer discharge and tablet press loading. In-process particle size is monitored by laser diffraction per USP <429>, and batches are rejected when the fraction below 45 µm exceeds 20% w/w because fines accelerate granule overwetting in the force feeder and create density gradients across the die cavity.

The aseptic manufacture of a parenteral small peptide solution from a pharma-grade API begins with compounding in a Grade C cleanroom under ISO 14644-1 Class 7 conditions, followed by bioburden reduction through a 0.22 µm polyethersulfone filter installed in a Grade A laminar-flow zone complying with EU GMP Annex 1 (2022) Section 8.8. The API is dissolved at 1.0–20.0 mg/mL in a buffer selected to maintain pH 4.0–7.5; acetate buffer is used for peptides with an acidic isoelectric point, histidine buffer for neutral pH stability, and phosphate is avoided in lyophilised formulations because disodium phosphate can crystallize during freezing and shift the pH of the freeze-concentrate by up to 3 units. Filtration pressure differential across a 0.22 µm PES capsule filter with an effective filtration area of 0.5 m² must not exceed 1.0 bar; higher pressure can shear peptide aggregates into smaller particles that pass through the filter and later reform as subvisible particles under storage. Pre-filtration bioburden is controlled at ≤ 10 CFU/100 mL by a 0.45 µm prefilter or by holding the solution at 2–8°C for no more than 12 h. Residual trifluoroacetic acid from peptide synthesis is controlled to ≤ 0.1% w/w by ion chromatography because the counterion shifts the final formulation pH and can increase local acid-catalyzed degradation during lyophilisation.

Lyophilisation is performed in a stainless-steel lyophilizer with shelf temperature uniformity ± 1.0°C across a 5.0 m² heat-transfer surface. The solution is filled into 2R or 6R Type I borosilicate vials at a fill volume of 1.0–3.0 mL, placed on cooled shelves at −40°C for 4–6 h to ensure complete solidification. Primary drying shelf temperature is ramped to −20°C at 0.5°C/min under a chamber pressure of 80–100 mTorr. Product temperature during primary drying must remain at least 2.0°C below the collapse temperature determined by freeze-drying microscopy; peptide-mannitol cakes with Tg' around −32°C require shelf settings no higher than −25°C during the first 48 h of primary drying. Secondary drying at 25°C for 6–12 h reduces residual moisture to ≤ 1.0% w/w, measured by Karl Fischer titration per USP <921>. The finished lyophilized cake is inspected for visible particulates per USP <790>, and reconstitution time is specified at ≤ 3 min with 2.0 mL of water for injection.

Aggregation of small peptide APIs in liquid injections remains the dominant failure mode in long-term storage at 5°C or 25°C. Subvisible particle counts can increase from fewer than 100 particles ≥ 10 µm per container at release to more than 3000 within 6 months if the formulation pH is within 0.5 units of the peptide isoelectric point or if polysorbate 80 is present at 0.02% w/v with high peroxide content. A polysorbate-free formulation with 5% w/v mannitol or trehalose as a lyoprotectant is preferred for lyophilized products. For liquid injections stored at 2–8°C, the pH is maintained at least 2.0 units away from the isoelectric point, and an antioxidant such as L-methionine at 2–5 mg/mL may be added if oxidative degradation is confirmed by forced oxidation with 0.3% hydrogen peroxide. Any addition of antioxidant must be justified by photostability testing per ICH Q1B; L-methionine can generate methionine sulfoxide and increase absorbance at 280 nm, complicating peptide purity assays by size-exclusion chromatography.

Oral peptide granules intended for sachet or stick-pack delivery are manufactured by dry granulation or alcohol-assisted wet granulation to retain the peptide in a matrix that can be dispersed in 50 mL of water before administration. Stick-pack lines running at 60–100 packs/min with 10 mm seal width require free-flowing granules with d10 ≥ 150 µm, d50 250–425 µm, and d90 ≤ 850 µm; fines below 75 µm above 15% w/w cause dust accumulation on the heat-sealing jaws and increase seal failure rates above 0.5%. Enteric protection of peptide granules is achieved by bottom-spray Wurster coating using Eudragit L100-55 at a coating weight gain of 8–12% w/w, with triethyl citrate at 20% w/w of polymer and talc at 50% w/w of polymer as anti-tack. Inlet air temperature is held at 45°C and product temperature at 30–35°C; exceeding 35°C during coating softens the enteric polymer at the Wurster partition and creates agglomerates with a d90 above 1.0 mm. Two-stage dissolution per USP <711> delayed-release method A—0.1 N HCl for 2 h followed by pH 6.8 phosphate buffer—is used to confirm no more than 10% peptide release in acid and no less than 75% release in 45 min at pH 6.8.

For oral peptide granules, absolute bioavailability often remains below 1–2% without permeation enhancement; sodium caprate at 10–15% w/w in the granule matrix can increase Cmax but can loosen tight junctions in a dose-dependent manner, so exposure margins must be established under ICH M3(R2). Published data for the specific oral bioavailability of individual small peptides in this granule configuration is limited; the formulation ratio of sodium caprate to peptide must be fixed during phase 1 clinical manufacturing and maintained within ± 2.0% w/w during scale-up to commercial batches. Granule moisture is controlled to 1.0–2.0% w/w before stick-pack filling because higher moisture plasticizes the enteric coating and reduces the acid protection threshold below pH 5.5, while lower moisture causes brittle fracture of coated granules during auger dosing at speeds above 80 packs/min.