Scalable generation of droplet-templated amorphous solid dispersions using continuous antisolvent extraction

Abstract

Amorphous solid dispersion (ASD) is a widely used formulation strategy for improving the apparent solubility and oral absorption of poorly water-soluble drugs. However, established commercial ASD manufacturing processes often generate powders with broad particle size distributions, irregular morphology and poor powder flowability, all of which complicate downstream processing. In this work, we introduce a droplet-templated ASD manufacturing process that combines step-emulsification co-processing of a polymer and drug with continuous anti-solvent extraction to produce highly monodisperse ASD particles. We produce intermediate ASD drug powder products with favorable particle attributes, including high monodispersity, improved flowability, and compatibility with downstream tableting. In contrast with prior reports on microfluidic production, this process allows the production of sufficient quantities of powder to enable compression into drug products (tablets). The tablets thus produced meet stringent drug product specifications (tensile strength, extended supersaturation dissolution profile and residual solvent below ICH limits). These results demonstrate that droplet microfluidics, traditionally viewed as a small-scale process, can be leveraged as a promising alternative manufacturing platform for ASDs with industrially relevant throughput for high-value pharmaceutical products.

Introduction

In recent years, there has been a surge of interest in using amorphous forms of active pharmaceutical ingredients to overcome poor aqueous solubility exhibited in their crystalline form (Wilson et al., 2020). The spotlight on amorphous solid dispersions (ASDs) has intensified as 60–70% of newly manufactured drugs fall under the biopharmaceutics classification system (BCS) Class II and IV compounds (Lugtu-Pe et al., 2021). While Class II drugs have limited absorption due to poor solubility, Class IV drugs exhibit low bioavailability due to a combination of low solubility and permeability (Papich and Martinez, 2015). By leveraging on higher free energy and lower density of amorphous substances, ASD formulations of Class II and potentially Class IV drugs can enhance the bioavailability and therapeutic effects compared to their crystalline counterparts (Babu and Nangia, 2011, Surampalli et al., 2013)

Established ASD manufacturing techniques such as freeze drying, spray drying, and jet milling are energy-intensive and often produce polydisperse particles with large size distributions that can negatively affect dissolution and oral bioavailability (Chu et al., 2012). In addition, these processes suffer from recrystallization which has been reported in spray-dried materials processed under different temperatures and feed rates (Beyer et al., 2016). Converting the resulting powders into suitable dosage forms pose yet another challenge to the downstream processing, often requiring additives and secondary excipients such as lubricants to improve their physical properties for tableting (Pandi et al., 2020). Hot-melt extrusion (HME) has emerged as a more modern approach for cost-effective and high-throughput production of amorphous solid dispersions (ASDs) (Agrawal et al., 2015). However, the risk of thermal degradation of heat-sensitive active pharmaceutical ingredients (APIs) and copolymers often limit the applicability of HME.

Co-precipitated amorphous dispersion (cPAD) is another emerging technology that involves mixing co-dissolved APIs and polymers with a fully miscible antisolvent under high shear, followed by thermal annealing to achieve excellent particle properties (Strotman and Schenck, 2021). Despite its advantages, the heat introduced during high-shear mixing and thermal annealing can promote crystallization of amorphous materials (Kumar and Suryanarayanan, 2022) and many processing challenges remain in preventing such recrystallization behavior of ASDs (Myślińska et al., 2023). More recently, droplet microfluidics has been explored as an alternative platform for co-processing APIs and excipients through droplet solidification via extractive crystallization (Shikha et al., 2022, Ng et al., 2022, Fortt et al., 2020). Although the scope of the studies does not extend to producing amorphous compounds, the absence of thermal treatment in the process, as well as the superior powder rheology of co-processed materials produced may suggest a potential avenue for droplet microfluidics in fabricating amorphous compounds with improved powder properties.

Droplet microfluidics has gained broad popularity in recent years due to its remarkable ability to create particles with a narrow and consistent size distribution (Chong et al., 2015). Having exceptional control over particle size and morphology is highly beneficial for the pharmaceutical industry due to its effect on flowability and hence processibility in the downstream manufacturing (Ekdahl et al., 2019). Amongst many microfluidic droplet generators, T-junctions (Schneider et al., 2011) and cross-junctions (Loo et al., 2016) are most commonly used, which leverage the viscous shear force exerted by the continuous phase on the dispersed phase in a single channel format. Hence, many attempts have been made to scale this system by arraying multiple single-channel generators in parallel to achieve the desirable throughput (Yadavali et al., 2018). However, fluctuations in the flow rates can broaden droplet size distributions and result in the formation of satellite droplets (Loo et al., 2016). Planar shear devices can thus be difficult to scale up in practice as droplet size at each junction is sensitive to the local flow rates in the parallelized network (Wu et al., 2021).

Step-emulsification has recently emerged as a particularly attractive route for scalable droplet generation with precise control over emulsion size and monodispersity (Amstad et al., 2016). Unlike shear-driven methods, step-emulsification leverages preferential wetting, in which the strong affinity of the continuous phase for the channel walls causes a ‘pinching-off’ effect on the dispersed phase flow (Lian et al., 2019). Droplet formation is driven by the difference in Laplace pressure when the liquid confined in a shallow channel expands into a spherical droplet upon entering a deeper reservoir through a step (Eggersdorfer et al., 2018). The droplet size can therefore remain largely independent of flow rates over a wide operating window for both dispersed and continuous phases, as long as the dispersed phase is depleted faster from the nozzle than it is refilled (Ofner et al., 2016). This insensitivity to flow rate fluctuations can significantly simplify the device design and operation for industrial scale-up (Stolovicki et al., 2017).

Step emulsification devices are typically fabricated from polydimethylsiloxane (PDMS) due to its rapid and cost-effective prototyping in laboratory settings (Stolovicki et al., 2017). However, the poor compatibility of PDMS with many organic solvents (Lee et al., 2004) limits its use in pharmaceutical applications. To address this issue, Ofner et al. were the first to fabricate a step-emulsification device in glass using photolithography followed by etching and bonding of two aligned glass pieces (Ofner et al., 2016). Their work demonstrated the feasibility of generating highly monodisperse droplets using pharmaceutical solvents at throughputs of 25 mL/h, far exceeding those of single-channel devices. If such high-throughput droplet generation can be coupled with a suitable process to produce pharmaceutical products, such as ASDs, it could enable a new paradigm for precise and compact commercial-scale pharmaceutical manufacturing– this is indeed what we seek to demonstrate in this study.

This paper presents an integrated process for generating monodisperse ASD microparticles by coupling a glass step emulsification device with a subsequent unit operation of continuous antisolvent extraction. The novelty of the present work extends beyond the use of microfluidics for co-processing polymeric materials with APIs, which is a well-established approach (Sharratt et al., 2021, Shi et al., 2023). Rather, the paper demonstrates the feasibility of an industrially relevant and scalable microfluidics-enabled manufacturing technique that entails continuous operation, controlled solvent extraction and improved powder attributes for robust downstream processing. We chose Felodipine (FLP) as a model API and hypromellose acetate succinate (HPMCAS) as a model excipient in our studies. In contrast to prior microfluidic studies, the process allows the production of sufficient quantities of powder to subsequently enable compression into oral solid dosage forms with suitable tensile strength, an extended supersaturation dissolution profile, and residual DMC below the reporting limit. In this work, we can process droplets into powders at a representative dispersed-phase flow rate of 0.9 mL/min. The device could also be operated at dispersed-phase flow rates up to 10 mL/min for droplet generation, corresponding to a projected ASD solids throughput of up to 720 g/day at 5% w/v solids loading. While the higher flow-rate operation was associated with a modest increase in droplet size, as shown in Supplementary Fig. S2, these results indicate the potential scalability of the microfluidic droplet-generation step. Further optimization of the downstream extraction and recovery process would be required to translate this projected droplet-generation throughput to end-to-end powder production.

Continue reading here

Jun Ho Min, Zheng Jie Liew, Leon Yoon Ho, Zhi Kai Tio, Arif Z. Nelson, Natalia Veronica, Lai Wah Chan, Paul Wan Sia Heng, Abigail Gershman, Samir Kulkarni, John Richard Murphy, Jennifer Ann Dolman, Catherine Michelle Ambler, Shawn LaCasse, Kapildev Arora, Saif A. Khan, Patrick S. Doyle, Scalable generation of droplet-templated amorphous solid dispersions using continuous antisolvent extraction, International Journal of Pharmaceutics, 2026, 127193, ISSN 0378-5173, https://doi.org/10.1016/j.ijpharm.2026.127193.

Read more interesting articles on Amorphous Solid Dispersions here:

• Advancing amorphous solid dispersions through empirical and hybrid modeling of drug–polymer solubility and miscibility: A case study using Ibuprofen
• Dissolution-Microdialysis predicts oral bioavailability from phospholipid-based amorphous solid dispersions
• Amorphous Solid Dispersions of Polyphenols: Current State of the Art (Part I)