In Vivo Mechanism of Action of Sodium Caprate for Improving the Intestinal Absorption of a GLP1/GIP Coagonist Peptide.

Sodium caprate (C10) has been widely evaluated as an intestinal permeation enhancer for the oral delivery of macromolecules. However, the effect of C10 on the intestinal absorption of peptides with different physicochemical properties and its permeation-enhancing effect in vivo remains to be understood. Here, we evaluated the effects of C10 on intestinal absorption in rats with a glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GIP-GLP1) dual agonist peptide (LY) and semaglutide with different enzymatic stabilities and self-association behaviors as well as the oral exposure of the LY peptide in minipigs. Furthermore, we investigated the mechanism of action (MoA) of C10 for improving the intestinal absorption of the LY peptide in vivo via live imaging of the rat intestinal epithelium and tissue distribution of the LY peptide in minipigs. The LY peptide showed higher proteolytic stability in pancreatin and was a monomer in solution compared to that in semaglutide. C10 increased in vitro permeability in the minipig intestinal organoid monolayer to a greater extent for the LY peptide than for semaglutide. In the rat jejunal closed-loop model, C10 increased the absorption of LY peptide better than that of semaglutide, which might be attributed to higher in vitro proteolytic stability and permeability of the LY peptide. Using confocal live imaging, we observed that C10 enabled the rapid oral absorption of a model macromolecule (FD4) in the rat intestine. In the duodenum tissues of minipigs, C10 was found to qualitatively reduce the tight junction protein level and allow peptide uptake to the intestinal cells. C10 decreased the transition temperature of the artificial lipid membrane, indicating an increase in membrane fluidity, which is consistent with the above in vivo imaging results. These data indicated that the LY's favorable physicochemical properties combined with the effects of C10 on the intestinal mucosa resulted in an ∼2% relative bioavailability in minipigs.

Vitamin C Improves Dasatinib Concentrations Under Hypochlorhydric Conditions of the Simulated Stomach Duodenum Model.

PURPOSE: pH-dependent drug-drug interactions (DDIs) with poorly soluble, weakly basic drugs may lead to clinical implications. Dasatinib is a tyrosine kinase inhibitor with reduced absorption in patients on acid-reducing agents (ARAs). The objective of this study is to investigate the influence of gastric pH on dasatinib supersaturation and determine if vitamin C (L-ascorbic acid) can improve dasatinib concentrations under simulated hypochlorhydric gastric conditions. METHODS: A dynamic, in vitro, multi-compartment, simulated stomach duodenum (SSD) model mimicking fluid volumes and transfer rates was used to investigate the concentration of BCS class IIb drugs versus time curves. Dasatinib and lamotrigine were explored under normal, fasted, simulated gastric fluids (pH 2) (FaSGF), hypochlorhydric simulated gastric fluids (pH 4.5) (FaSGFhypo) and FaSGFhypo with 1000 mg of vitamin C. RESULTS: Significant supersaturation of dasatinib was observed in the duodenum compartment of the SSD model in FaSGF. A 90% reduction in dasatinib AUC∞ was observed in FaSGFhypo. Upon addition of vitamin C to FaSGFhypo, drug concentrations were restored to those observed in FaSGF. Lamotrigine AUC∞ in the duodenal compartment were similar in both FaSGF and FaSGFhypo. The in vitro trends observed for dasatinib and lamotrigine are reflective of the trends observed in vivo in subjects receiving treatment with ARAs. CONCLUSIONS: The SSD model serves as a good in vitro tool for assessing the effect of pH-dependent DDIs on bioavailability of weakly basic drugs with solubility/ dissolution limited absorption. Vitamin C provides a promising approach for improving bioavailability of poorly soluble, weakly basic drugs in hypochlorhydric patients.

Identification of a Multi-Component Formulation for Intestinal Delivery of a GLP-1/Glucagon Co-agonist Peptide.

PURPOSE: Oral delivery of therapeutic peptides has been challenging due to multiple physiological factors and physicochemical properties of peptides. We report a systematic approach to identify formulation compositions combining a permeation enhancer and a peptidase inhibitor that minimize proteolytic degradation and increase absorption of a peptide across the small intestine. METHODS: An acylated glucagon-like peptide-1/glucagon co-agonist peptide (4.5 kDa) was selected as a model peptide. Proteolytic stability of the peptide was investigated in rat and pig SIF. Effective PEs and multiple component formulations were identified in rats. Relative bioavailability of the peptide was determined in minipigs via intraduodenal administration (ID) of enteric capsules. RESULTS: The peptide degraded rapidly in the rat and pig SIF. Citric acid, SBTI, and SBTCI inhibited the enzymatic degradation. The peptide self-associated into trimers in solution, however, addition of PEs monomerized the peptide. C10 was the most effective PE among tested PEs (DPC, LC, rhamnolipid, C12-maltosides, and SNAC) to improve intestinal absorption of the peptide in the rat IJ-closed loop model. A combination of C10 and SBTI or SBTCI increased the peptide exposure 5-tenfold compared to the exposure with the PE alone in the rat IJ-cannulated model, and achieved 1.06 ± 0.76% bioavailability in minipigs relative to subcutaneous via ID administration using enteric capsules. CONCLUSION: We identified SBTI and C10 as an effective peptidase inhibitor and PE for intestinal absorption of the peptide. The combination of SBTI and C10 addressed the peptide physiochemical properties and provides a formulation strategy to achieve intestinal delivery of this peptide.

Identification and Prevention of Particulates in a Parenteral Solution of a Protein Kinase B Inhibitor.

The aim of this work was to identify the root cause of particulates in a parenteral solution formulation of a protein kinase B inhibitor (AKT inhibitor) and to devise a formulation fix. While standard potency/purity analyses did not reveal degradation, it was determined that the mass of the particulates corresponds to that of a "dimer-like" degradation product. The dimer-like molecule manifested itself as particulates rather than impurities in standard HPLC analysis. Particulate formation was found to be pH dependent. Consequently, an optimum formulation pH at which the compound has adequate solubility and the rate of particulate formation is slow was identified.

Salt Disproportionation in the Solid State: Role of Solubility and Counterion Volatility.

Disproportionation propensity of salts (HCl, HBr, heminapadisylate) and adipic acid cocrystal of corticotropin releasing hormone receptor-1 antagonist was studied using model free kinetics. Using thermogravimetic weight loss profile or heat flow curves from differential scanning calorimetry, an activation energy plot for salts and cocrystal was generated based on model free kinetics. This activation energy of disproportionation provided qualitative information about the solid state salt stability. To ensure the stability throughout the shelf life, "prototype" formulations of salts and cocrystal in tablet form were stored at 40 °C and several water vapor pressures. Disproportionation kinetics were studied in these prototype tablet formulations using two-dimensional X-ray diffractometry. Formulations containing the adipic acid cocrystal or heminapadisylate salt did not show disproportionation of API when stored at 40 °C/75% RH for 300 days. On the other hand, formulations containing HCl or HBr salt disproportionated. Though isostructural, the disproportionation propensity of HBr and HCl salts was quite different. The HCl salt highlighted the important role that volatility of the counterion plays in the physical stability of the formulations. Solution state stability (i.e., in dissolution medium) of salts and cocrystal was also assessed and compared with solid state stability, by determining their solubility at different pH's, and intrinsic dissolution rate.

Absolute bioavailability of evacetrapib in healthy subjects determined by simultaneous administration of oral evacetrapib and intravenous [(13) C8 ]-evacetrapib as a tracer.

This open-label, single-period study in healthy subjects estimated evacetrapib absolute bioavailability following simultaneous administration of a 130-mg evacetrapib oral dose and 4-h intravenous (IV) infusion of 175 µg [(13) C8 ]-evacetrapib as a tracer. Plasma samples collected through 168 h were analyzed for evacetrapib and [(13) C8 ]-evacetrapib using high-performance liquid chromatography/tandem mass spectrometry. Pharmacokinetic parameter estimates following oral and IV doses, including area under the concentration-time curve (AUC) from zero to infinity (AUC[0-∞]) and to the last measureable concentration (AUC[0-tlast ]), were calculated. Bioavailability was calculated as the ratio of least-squares geometric mean of dose-normalized AUC (oral : IV) and corresponding 90% confidence interval (CI). Bioavailability of evacetrapib was 44.8% (90% CI: 42.2-47.6%) for AUC(0-∞) and 44.3% (90% CI: 41.8-46.9%) for AUC(0-tlast ). Evacetrapib was well tolerated with no reports of clinically significant safety assessment findings. This is among the first studies to estimate absolute bioavailability using simultaneous administration of an unlabeled oral dose with a (13) C-labeled IV microdose tracer at about 1/1000(th) the oral dose, with measurement in the pg/mL range. This approach is beneficial for poorly soluble drugs, does not require additional toxicology studies, does not change oral dose pharmacokinetics, and ultimately gives researchers another tool to evaluate absolute bioavailability.

Leveraging BCS in Development: A Case Study.

In this work, we discuss leveraging the Biopharmaceutics Classification System (BCS) in the development of edivoxetine HCl, a selective norepinephrine reuptake inhibitor. First, the biopharmaceutical and in vivo data are presented and discussed. Solubility studies indicate that edivoxetine HCl meets the BCS "highly soluble" criteria. To determine permeability classifications, in vitro intestinal Caco-2 epithelial cell model with and without cyclosporin A (CsA), a common P-glycoprotein (P-gp) inhibitor, were conducted. Pharmacokinetic (PK) data obtained across phase 1 and 2 clinical studies where single and multiple doses range from the lowest to the highest strength are presented. Neither the Caco-2 nor the in vivo data on their own were sufficient to conclusively classify edivoxetine as highly permeable. However, collectively the data were utilized to support high permeability and consequently BCS1 classification of edivoxetine HCl. BCS1 classification was leveraged throughout development to assess the risk associated with not conducting relative bioavailability (RBA) studies and avoiding bioequivalence (BE) studies. Examples are presented where formulation changes were made between phase I (drug in capsule/drug in bottle formulations) and phase II (tablet) trials in addition to phase III (tablet) and commercial (smaller tablet) without having to conduct any in vivo comparability studies. For the first change, BCS was leveraged to avoid conducting a RBA study even before obtaining official BCS classification. For the later change, official BCS1 classification was relied upon to avoid conducting a BE study.

A novel method for determining the solubility of small molecules in aqueous media and polymer solvent systems using solution calorimetry.

PURPOSE: To explore the application of solution calorimetry for measuring drug solubility in experimentally challenging situations while providing additional information on the physical properties of the solute material. METHODS: A semi-adiabatic solution calorimeter was used to measure the heat of dissolution of prednisolone and chlorpropamide in aqueous solvents and of griseofulvin and ritonavir in viscous solutions containing polyvinylpyrrolidone and N-ethylpyrrolidone. RESULTS: Dissolution end point was clearly ascertained when heat generation stopped. The heat of solution was a linear function of dissolved mass for all drugs (<10% RSD, except for chlorpropamide). Heats of solution of 9.8 ± 0.8, 28.8 ± 0.6, 45.7 ± 1.6 and 159.8 ± 20.1 J/g were obtained for griseofulvin, ritonavir, prednisolone and chlorpropamide, respectively. Saturation was identifiable by a plateau in the heat signal and the crossing of the two linear segments corresponds to the solubility limit. The solubilities of prednisolone and chlopropamide in water by the calorimetric method were 0.23 and 0.158 mg/mL, respectively, in agreement with the shake-flask/HPLC-UV determined values of 0.212 ± 0.013 and 0.169 ± 0.015 mg/mL, respectively. For the higher solubility and high viscosity systems of griseofulvin and ritonavir in NEP/PVP mixtures, respectively, solubility values of 65 and 594 mg/g, respectively, were obtained. CONCLUSION: Solution calorimetry offers a reliable method for measuring drug solubility in organic and aqueous solvents. The approach is complementary to the traditional shake-flask method, providing information on the solid properties of the solute. For highly viscous solutions, the calorimetric approach is advantageous.

Development and evaluation of C10 and SNAC erodible tablets for gastric delivery of a GIP/GLP1 peptide in monkeys.

The permeation enhancers (PEs) sodium caprate (C10) and sodium N-[8-(2-hydroxybenzoyl) amino] caprylate (SNAC) have been utilized for the intestinal and gastric delivery of macromolecules, respectively. However, the potential of C10 for the gastric delivery of a peptide and the ability of SNAC to deliver other peptides to the stomach beyond semaglutide have not been investigated. In this study, we have developed and evaluated C10 and SNAC-containing erodible tablets for the gastricdelivery of a glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GIP/GLP1) dual agonist peptide (LY) in cynomolgus monkeys. We also evaluated the impact of release rates on the in vivo performance of C10 and SNAC. Furthermore, we compared the oral exposure of the LY peptide and semaglutide with different proteolytic stabilities using a SNAC erodible tablet. Additionally, we investigated the mechanism of action of SNAC for improving gastric absorption of the LY peptide via tissue distribution in monkey. C10 and SNAC tablets released the peptide and PE by erosion from the tablet surface with 100 % release within 60 min at pH 6.8. Following a single oral administration to monkeys, C10 and SNAC erodible tablets at 300 mg exhibited similar LY mean absolute oral bioavailability of 5.7 % and 4.2 %, respectively. The C10 immediate release capsule (500 mg) with faster dissolution profile (10 min) showed a decrease in the LY oral bioavailability; however, a faster dissolution profile (15 min) with erodible SNAC tablet resulted in a relatively higher LY oral bioavailability compared to the slow-release erodible tablets (60 min). Using SNAC as the PE, the combination of slow-release tablet design and LY peptide with higher pepsin stability resulted in about 4-fold higher mean oral bioavailability in the monkeys than semaglutide (4.2 % vs 1.2 %, respectively). In the monkey gastric tissue, SNAC was found to reduce tight junction protein levels and increase the peptide uptake into the gastric epithelium suggesting its permeation enhancing mechanism via both paracellular and transcellular pathways. Taking these data altogether, the enhanced proteolytic stability of the LY peptide combined with the optimal erodible tablets enabled the gastric delivery of the LY peptide with a higher oral bioavailability than semaglutide.

Use of first derivative of displacement vs. force profiles to determine deformation behavior of compressed powders.

Displacement (D) vs. force (F) profiles obtained during compaction of powders have been reported by several researchers. These profiles are usually used to obtain mechanical energies associated with the compaction of powders. In this work, we obtained displacement-force data associated with the compression of six powders; Avicel PH101, Avicel PH301, pregelatinized corn starch, anhydrous lactose, dicalcium phosphate, and mannitol. The first three powders are known to deform predominantly by plastic behavior while the later ones are known to deform predominantly by brittle fracture. Displacement-force data was utilized to perform in-die Heckel analysis and to calculate the first derivative (dD/dF) of displacement-force plots. First derivative results were then plotted against mean force (F') at each point and against 1/F' at compression forces between 1 and 20 kN. Results of the in-die Heckle analysis are in very good agreement with the known deformation behavior of the compressed materials. First derivative plots show that materials that deform predominantly by plastic behavior have first derivative values (0.0006-0.0016 mm/ N) larger than those of brittle materials (0.0004 mm/N). Moreover, when dD/dF is plotted against 1/F' for each powder, a linear correlation can be obtained (R2=>0.98). The slopes of the dD/dF vs. 1/F' plots for plastically deforming materials are relatively larger than those for materials that deform by brittle behavior. It is concluded that first derivative plots of displacement-force profiles can be used to determine deformation behavior of powders.

Quantitative mineralogical properties (morphology-chemistry-structure) of pharmaceutical grade kaolinites and recommendations to regulatory agencies.

The physical and chemical characteristics of kaolinite (kaolin) may be variable, and minor amounts of other clay minerals, nonclay minerals, and other impurities may affect the properties of kaolinites. Thus specific technical properties of pharmaceutical grade kaolinites become very important because these clays are used in medical applications, e.g., as pharmaceutical excipients, and will be consumed by humans. Seven pharmaceutical grade kaolinite specimens were used in this study: K1004, KA105, 2242-01, K2-500, Acros, Acros-mono, and KX0007-1. In addition, two kaolinites from the Clay Minerals Society Source Clays, KGa-1b and KGa-2, were used for comparison purposes. The Acros-mono and 2242-01 kaolinites contained minor amounts of illite, which was demonstrated both compositionally and structurally by using inductively coupled plasma spectroscopy and powder X-ray diffraction. The KX0007-1 kaolinite powder was found to be heavily contaminated with quartz, cristobalite, and alunite. Crystal structure computations also showed excess Si in its tetrahedral site, and the mineral no longer has the typical kaolinite crystal structure. These widely-used industrial standards should be quantitatively characterized morphologically, compositionally, and structurally. Results of the mineralogical characteristics should be clearly labeled on the pharmaceutical grade kaolinites and reported to the relevant regulatory agencies.

An IQ Consortium Perspective on Connecting Dissolution Methods to In Vivo Performance: Analysis of an Industrial Database and Case Studies to Propose a Workflow.

Assessment of bioperformance to inform formulation selection and development decisions is an important aspect of drug development. There is high demand in the pharmaceutical industry to develop an efficient and streamlined approach for better understanding and predicting drug product performance to support acceleration of clinical timelines. This manuscript presents an effort from the IQ Formulation Bioperformance Prediction Working Group composed of members from 12 pharmaceutical companies under the IQ Consortium to develop a database around the topic of formulation bioperformance prediction and report findings from the database analysis. Six case studies described in the manuscript demonstrate how bioperformance models were used to predict in vivo performance and to provide guidance addressing questions encountered during oral solid dosage form development. The case studies also described findings of a correlation between in vitro dissolution and in vivo performance and how dissolution data can be incorporated into physiologically based biopharmaceutical modeling. Finally, a workflow for how in vitro dissolution data can be utilized to predict clinical bioperformance of oral solid dosage forms is proposed.

Interpreting deformation behavior in pharmaceutical materials using multiple consolidation models and compaction energetics.

Tableting behavior is often characterized using qualitative analyses of compactibility and compressibility measurements. More quantitative methods use consolidation models to estimate parameters indicative of the predominating deformation mechanism exhibited by a material. It will be shown that a concerted approach, using multiple consolidation models and mechanical energy analysis, presents a more reliable way of evaluating the relative plasticity of pharmaceutical materials and identifying complicating behaviors. Force versus displacement data for compact formation, porosity versus pressure and tensile strength data for ejected compacts were collected with a single instrument. The porosity and tensile strength data were analyzed using two relatively new models and the results were compared to three more classical models. Additionally, the mechanical work measurements were used to interpret the consolidation model predictions. Although the individual models are susceptible to a number of errors, complications and invalid assumptions, confidence can be gained when diverse models provide similar predictions. Disagreement between the model predictions can be taken as a sign of atypical behavior that should be further investigated by looking at the material's mechanical energetics. Finally, the use of work energy associated with compression and decompression as an initial measure of plasticity is supported.

Microstructures and pharmaceutical properties of ferulic acid agglomerates prepared by different spherical crystallization methods.

Spherical agglomerates of an active pharmaceutical ingredient, ferulic acid (FA), were prepared using four different spherical crystallization methods, i.e., quasi-emulsion solvent diffusion (QESD), anti-solvent, pH shift, and the direct method. Both the as-received FA and spherical agglomerates were characterized in terms of specific surface area (SSA), primary crystal shape and size, granule morphology and size, powder flowability, tabletability at two distinct speeds, and dissolution (both powder and tablet). Results showed that the microstructure, which is affected by size, shape, and packing of primary crystals, was the key that determined the flowability, tabletability and dissolution. The QESD powder exhibited the best flowability and tabletability. Both powder and tablet dissolution of FA followed the order of as-received > QESD > anti-solvent > pH shift, which was consistent with the order of the surface area exposed to the dissolution medium and not SSA. Moreover, compression reduced differences in the rates of dissolution of FA powders due to the size reduction of agglomerates by fragmentation.

Direct Compression Tablet Containing 99% Active Ingredient-A Tale of Spherical Crystallization.

Direct compression (DC) is the easiest and most cost-effective process for tablet manufacturing, because it only involves blending and compression. However, active pharmaceutical ingredients generally exhibit poor mechanical and micromeritic properties, which necessitate dilution and the use of high percentage of excipients to enable a robust DC manufacturing process. Consequently, drug loading in DC tablets is usually low (typically <30%, w/w). In this study, spherical crystallization by the quasi-emulsion solvent diffusion method was used to engineer a poorly flowing model compound, ferulic acid (FA), to attain superior mechanical properties, particle size distribution, and morphology. The engineered FA particles enabled the successful development of DC tablets containing 99% FA, which is in sharp contrast to the maximum 10% FA loading using as-received FA. The record high active pharmaceutical ingredient loading in this work illustrates the potential for spherical crystallization to enable high drug loading when developing a tablet product using the DC manufacturing process.

A Simulated Stomach Duodenum Model Predicting the Effect of Fluid Volume and Prandial Gastric Flow Patterns on Nifedipine Pharmacokinetics From Cosolvent-Based Capsules.

Nifedipine is a Biopharmaceutics Classification System class II drug displaying large variability in absorption even when administered as immediate-release soft gelatin capsules of a cosolvent formulation. This in vitro study sought to understand the reasons behind variability in nifedipine absorption, how it can be minimized, and if it can be predicted using in vitro models. A dynamic in vitro simulated stomach duodenum model was used to explore drug concentration-time profiles of nifedipine soft gelatin capsules under conditions simulating how patients take their medicines. Specifically, the effect of prandial gastric emptying patterns and fluid volume administration (250 mL vs. 50 mL water) were investigated. Significant supersaturation of nifedipine was observed. While administration of large and small water volumes gave rise to a similar Cmax and area under the curve (AUC∞), the coefficient of variation in AUC was 4.8% and 49%, respectively, which can be attributed to differences in precipitation kinetics. Fasting and fed gastric emptying patterns also gave rise to a similar AUC; however, Cmax was significantly lower in the fed state. These trends are consistent with previously published in vivo results in healthy volunteers. The simulated stomach duodenum provides a good discriminative screening tool for predicting trends in drug concentration profiles of Biopharmaceutics Classification System class II drugs.

Relative Bioavailability Risk Assessment: A Systematic Approach to Assessing In Vivo Risk Associated With CM&C-Related Changes.

Relative bioavailability (RBA) studies are often carried out to bridge changes made between drug products used for clinical studies. In this work, we describe the development of a risk assessment (RA) tool that comprehensively and objectively assesses the risk of noncomparable in vivo performance associated with Chemistry, Manufacturing, and Controls (CM&C)-related changes. The RA tool is based on a risk grid that provides a quantitative context to facilitate discussions to determine the need for an in vivo RBA study. Relevant regulatory guidances and the required in vitro and in silico absorption modeling data, on which the RA is based, are discussed. In addition, an analysis of previously executed RBA studies at Eli Lilly and Company over a period of several years is presented. The risk grid incorporates individual risk factors for a given study and provides a recommendation on the risk associated with bypassing an RBA study. The outcome of an RA results in 1 of 3 possible risk zones; lower tier risk, intermediate tier risk, and upper tier risk. In cases where the outcome from the RA falls into the intermediate tier risk zone, further in depth data analysis is required.

A critical evaluation of fasted state simulating gastric fluid (FaSSGF) that contains sodium lauryl sulfate and proposal of a modified recipe.

The aim of this work is to evaluate one of the most commonly used fasted state simulating gastric fluids (FaSSGFs), which contains sodium lauryl sulfate (SLS) (FaSSGF(SLS)), and propose a more appropriate surfactant concentration. Surface tension studies clearly show that the critical micelle concentration (CMC) of SLS in the relevant media (a media whose pH and sodium chloride concentration are representative of physiological conditions) is significantly lower (p<0.05) than 8.67 mM, which is the SLS concentration in FaSSGF(SLS). The CMC of SLS in the relevant media was determined to be 1.75 mM. Based on this a modified recipe is proposed in which the concentration of SLS is sufficient to achieve a surface tension similar to that in vivo without causing artificial micellar solubilization. Solubility, intrinsic dissolution, and GastroPlus modeling studies are presented to support and give rationale for the modified recipe. In addition, a comparison between the modified recipe and other FaSSGFs reported in the literature is made.

Effect of storage conditions on compaction behavior of two grades of spray-dried lactose.

In this work we examine the effect of storage conditions (moisture exposure) on the compression behavior of 2 grades of spray-dried lactose (Pharmatose DCL 11 and Pharmatose DCL 14) under 2 different circumstances. The first was to expose powder samples to moisture, then compress them. The second was to expose precompressed tablets to moisture. We clearly show that the effect of moisture exposure and amorphous content crystallization in spray-dried lactoses on compaction behavior depends on whether this moisture exposure takes place before or after compression. In addition, the impact of storage conditions depends on the grade of spray-dried lactose.

Effect of particle size on compaction of materials with different deformation mechanisms with and without lubricants.

This work investigates the effect of excipient particle size on compaction properties of brittle, plastic and viscoelastic materials with and without added lubricants. Sieve cuts of microcrystalline cellulose (MCC), starch and dibasic calcium phosphate dihydrate were obtained by sieving, then samples were tested without lubrication or with added lubricant (0.5% Mg stearate mixed for either 5 or 30-min). Compacts were left overnight before testing. It was found that in the absence of lubricant, compact tensile strength (TS) was dependent on particle size only for starch. With Mg stearate, lubricant sensitivity shows a strong dependence on excipient particle size for both starch and MCC, where smaller particles are less affected by lubricant. Dibasic calcium phosphate dihydrate was not sensitive to lubricant even after 30 min mixing. This study highlights that in the absence of lubricant, initial particle size of excipients has no impact on compact strength not only for dibasic calcium phosphate dihydrate (brittle), but also for MCC (plastic). On the other hand, TS is dependent on particle size both with or without added lubricant for starch (viscoelastic).