Process development to overcome drug substance agglomeration and restore dissolution performance

Abstract

Dissolution testing is performed to understand the rate and extent of drug release for assurance of in vivo bioavailability. In this case study, an immediate-release tablet of an investigational BCS Class II compound exhibited slow and incomplete dissolution at pH 4.5 and 6.8, indicating a risk of reduced exposure in patients with elevated gastric pH. To diagnose and remediate the issue, we evaluated three factors: (1) drug substance particle size (milled versus unmilled), (2) formulation wettability via incorporation of a surfactant alongside water-soluble versus insoluble fillers, and (3) shear mixing of the powder blend. Dissolution profiles for tablets containing milled and unmilled drug were comparable, and the addition of a surfactant did not improve performance. Subsequent investigation identified presence of drug agglomerates as the root cause of slow dissolution. Introducing a simple shear deagglomeration step as a pre-blend step markedly enhanced dissolution, allowing the clinical formulation composition to be retained and avoiding re-formulation and associated timeline impacts.

Introduction

Oral solid dosage forms must first disintegrate in the gastrointestinal tract to enable drug dissolution, a prerequisite for absorption. The extent and rate of oral drug absorption are governed by a combination of biopharmaceutical and physiological determinants. Key physicochemical drivers include solubility, intrinsic dissolution rate, ionisation (pKa), and effective surface area, while gastrointestinal pH, bile salt composition, and membrane permeability are among the principal physiological factors modulating absorption.1, 2 Formulation excipients influence performance beyond enabling manufacturability — they also have an impact on content uniformity and disintegration, drug substance stability and bioavailability.

A significant proportion of clinical candidates are poorly soluble, which complicates preclinical characterisation, clinical pharmacology, and formulation development. During early clinical studies, where drug and formulation physiochemical properties are typically held constant, variability in exposure thus predominantly reflects inter-individual physiological differences.3 In later phases, with larger and more diverse populations, variability tends to increase due to broader distributions in gastrointestinal conditions and other factors.4, 5, 6 Patients taking concomitant medications risk altered plasma drug concentrations due to pharmacokinetic and/or pharmacodynamic interactions, potentially leading to therapeutic failure (under-exposure) or adverse effects (over-exposure).7, 8, 9

One prevalent source of risk is concomitant use of acid-reducing agents (ARAs), including proton pump inhibitors (PPIs), H2-receptor antagonists, and antacids, which elevate gastric pH and thereby reduce the dissolution of basic drugs exhibiting pH-dependent solubility.10,11 Given the high prevalence of ARAs therapies, formulation design for such candidates should anticipate elevated gastric pH. Formulations that achieve robust performance across a wide pH range can reduce inter-individual pharmacokinetic variability, support accurate PK/PD characterisation, and avoid dosing restrictions that, if unaddressed, may ultimately be reflected in product labelling.

In vitro dissolution testing across biorelevant pH range (typically pH 1.2–6.8) is used to evaluate drug release and de-risk absorption-related variability. Where rapid and complete dissolution is demonstrated, in vitro approaches may support biopharmaceutics risk mitigation and, in some cases, underpin in vitro bridging in lieu of certain in vivo studies.12, 13, 14, 15 Conversely, incomplete dissolution within this pH range signals potential variability in exposure and a need for formulation optimisation.

Micronized drug substances, while offering increased surface area for enhanced dissolution, are inherently prone to agglomeration due to elevated van der Waals forces, electrostatic interactions, and capillary forces. This phenomenon is particularly pronounced for cohesive, poorly soluble compounds where particle size reduction intensifies interparticulate attractions. Drug agglomeration during powder mixing significantly impacts dissolution performance, with the degree of agglomeration dependent on drug properties, excipient selection, and processing conditions. Das and Stewart demonstrated that cohesive powders form matrices of non-homogeneous powder strength, with deagglomeration following a non-linear relationship with applied energy. Critically, they showed that specific surface area may increase only marginally despite substantial particle size reduction after milling, indicating that porous agglomerate structures persist and mask the true available surface area.16 Kale et al. showed that mixing intensity and duration play critical roles in determining whether agglomerates are disrupted or reinforced during blend preparation, with incomplete deagglomeration observed even after extended mixing periods.17 The tensile strength of drug agglomerates, governed by packing fraction and particle size, determines their resistance to dispersion during both manufacturing and dissolution. Compact agglomerates may persist through tableting and remain intact during dissolution, dramatically reducing effective surface area and resulting in incomplete drug release even under sink conditions.

Various strategies have been developed to address the dissolution challenges. Thommes et al. demonstrated that solid crystalline suspensions prepared via hot melt extrusion with mannitol as a matrix can maintain fine particle dispersions; however, this approach requires hot-melt extrusion — additional unit operations that increase process complexity and manufacturing costs.18 Similarly, high-energy forms such as amorphous solid dispersions, involve substantial reformulation efforts, specialized equipment, and rigorous stability testing. While effective, these approaches are particularly challenging to implement in late-stage development where formulation changes necessitate costly in vivo bridging studies and cause significant timeline delays.

AZDX (a proprietary AstraZeneca investigational compound) is a BCS Class II drug substance (DS) with low solubility at higher pH (<50 µg/ml at pH 6.5). Immediate-release tablets of AZDX showed slow and incomplete dissolution at pH 4.5, despite sink conditions, indicating a heightened risk of sub-therapeutic exposure in patients with elevated gastric pH due to PPI co-administration. This imposes constraints on clinical trial inclusion and threatens to limit patient access post-approval via restrictive labelling. Mid-study reformulation to address this risk can be time-consuming and costly, particularly if in vivo bridging becomes necessary, thereby hindering timely progression to late-stage development.

The present work addresses a critical gap by demonstrating that agglomeration-driven dissolution failure can be systematically diagnosed and remediated using conventional unit operations without reformulation. The novelty lies in three key aspects: (1) systematic integration of complementary analytical techniques (dissolution testing, SEM, X-ray micro-CT) to definitively identify agglomeration as the root cause despite inconclusive particle size distribution data; (2) demonstration that conventional formulation levers (particle size reduction, surfactant addition, water-soluble fillers) were individually insufficient to overcome agglomeration in this system; and (3) development of a simple, scalable process solution—targeted shear pre-mixing or co-milling micro-mixing—that disrupts agglomerates within existing manufacturing workflows (blending or dry granulation) and restores rapid dissolution across the biorelevant pH range. This process-focused approach offers significant practical value for pharmaceutical development, particularly for late-stage programs where maintaining formulation composition and avoiding additional unit operations can save months of development time and millions in costs while de-risking clinical studies.

Continue reading here

Materials

Model drug AZDX of the unmilled and milled grade was sourced from AstraZeneca, UK. Microcrystalline cellulose (MCC, Avicel® PH102, FMC Biopolymer), dibasic calcium phosphate anhydrous (DCPA, Emcompress® anhydrous, JRS Pharma) and mannitol (Pearlitol® 100SD, Roquette) were utilized as fillers. MCC and DCPA are water-insoluble functional fillers. MCC has attractive dry binding (compressibility) properties, enabling the manufacture of tablets by direct compression.

Mayank Singhal, Kalyan Nidadavole, Dean S. Murphy, Dhaval Kalaria, Process development to overcome drug substance agglomeration and restore dissolution performance, Journal of Pharmaceutical Sciences, 2026, 104368, ISSN 0022-3549, https://doi.org/10.1016/j.xphs.2026.104368.

Read also our introduction article on Mannitol here: