Predictions of biorelevant solubility change during dispersion and digestion of lipid-based formulations

Computational approaches are increasingly explored in development of drug products, including the development of lipid-based formulations (LBFs), to assess their feasibility for achieving adequate oral absorption at an early stage. This study investigated the use of computational pharmaceutics approaches to predict solubility changes of poorly soluble drugs during dispersion and digestion in biorelevant media. Concentrations of 30 poorly water-soluble drugs were determined pre- and post-digestion with in-line UV probes using the MicroDISS Profiler^TM. Generally, cationic drugs displayed higher drug concentrations post-digestion, whereas for non-ionized drugs there was no discernible trend between drug concentration in dispersed and digested phase. In the case of anionic drugs there tended to be a decrease or no change in the drug concentration post-digestion. Partial least squares modelling was used to identify the molecular descriptors and drug properties which predict changes in solubility ratio in long-chain LBF pre-digestion (R² of calibration = 0.80, Q² of validation = 0.64) and post-digestion (R² of calibration = 0.76, Q² of validation = 0.72). Furthermore, multiple linear regression equations were developed to facilitate prediction of the solubility ratio pre- and post-digestion. Applying three molecular descriptors (melting point, LogD, and number of aromatic rings) these equations showed good predictivity (pre-digestion R² = 0.70, and post-digestion R² = 0.68). The model developed will support a computationally guided lipid-based formulation strategy for emerging poorly water-soluble drugs by predicting biorelevant solubility changes during dispersion and digestion. This facilitates a more data-informed developability decision making and subsequently facilitates a more efficient use of formulation screening resources.

Introduction

In the discovery phase, computational tools are well established to identify new lead candidates with optimal receptor binding affinity and physicochemical profiles with mechanistic models or data-driven approaches, like quantitative structure activity relationship and quantitative structure property relationship (C. Bergström et al., 2016). The drug candidates are often evaluated based on their “drug-ability”, which is generally defined as the likelihood of the drug being able to functionally interact with the biological target. However, a consequence of discovering drugs with more potent binding often includes the selection of more lipophilic drug candidates, which are often not readily amenable for oral delivery (C. Bergström et al., 2016). Bio-enabling formulation strategies have been developed to increase the bioavailability of these drug compounds. Examples of these bio-enabling formulation strategies are solid dispersions (Singh & Van den Mooter, 2016; Van Den Mooter, 2012), nano- and micro-sizing (Merisko-Liversidge et al., 2003), and lipid-based formulation (LBF) (Feeney et al., 2016). There is no formulation that works for all candidate drugs, therefore, the development of a lead formulation for pivotal clinical trials often includes screening of a range of excipients and formulations until a formulation with acceptable bioavailability is identified. Alternatively, multiple formulations with the drug compound are developed in parallel to increase the chance of success. Each of these approaches has high development costs, long development timelines, and includes a risk of failure to develop the new medicine (Kuentz et al., 2021; Lennernäs et al., 2014). This is particularly concerning because contemporary drug product development continues to depend on trial-and-error and/or in-house knowledge within individual pharmaceutical companies (Ditzinger et al., 2019). However, computational approaches to guide drug product development have substantial potential to make drug product development faster, better, and cheaper (C. Bergström & Larsson, 2018).

LBF is a bio-enabling formulation strategy where the drug compound is presented pre-dissolved in a lipid matrix before administration. LBFs hold for a very promising technology for solubility-rate-limited drugs and for drugs that exhibit significant positive food effects (Feeney et al., 2016). However, the development of LBFs has declined over the last decade (Bennett-Lenane et al., 2020). This can be an indication of challenges among pharmaceutical companies to adopt LBF strategies and a need for guidance on development (Holm, 2019). Drug absorption from LBFs is a dynamic process where the solubility of the drug compound dramatically changes upon both dispersion and digestion also depending on the lipid excipients in the formulation (Boyd & Clulow, 2021; Gautschi et al., 2016; Khan et al., 2016). The dispersion of LBFs in biorelevant media can lead to increased drug solubilisation, transient supersaturation, and delayed precipitation, meaning that the drug can be present in the biorelevant media in a concentration higher than the equilibrium concentration. The pH-stat method is the most widely used in vitro lipolysis method for testing LBFs. With this approach, the pH of the digestion media is maintained throughout the experiment by adding sodium hydroxide to counteract the formation of free fatty acids because of the digestion of the LBFs. This method requires relatively high quantities of media, formulation, and drug compound. Moreover, it is relatively time consuming as it is a low throughput method that requires withdrawing of samples for off-line analysis (Williams et al., 2012). Optimized in vitro lipolysis methods have been suggested to comply with the need for higher throughput, smaller quantities, and real-time analytics (Devraj et al., 2014; Khan et al., 2022; Mosgaard et al., 2015, 2017; Tanaka et al., 2022). Recently, our research group has shown proof of concept of a small-scale lipolysis method using the MicroDISS ProfilerTM which facilitates in-line higher throughput data generation of drug concentrations during dispersion and digestion (Ejskjær et al., 2023).

A number of studies have demonstrated the use of data-driven prediction models in the development of LBFs to estimate lipid solubility to act as a guide for maximal dose loading (L. Alskär et al., 2016; Persson et al., 2013; Sacchetti & Nejati, 2012). Even though it is useful to guide the initial understanding of the maximal dose loading in the LBFs this approach does not represent the sole criterion for LBF suitability. As previously mentioned the drug absorption from LBF is a dynamic process where the solubility can change dramatically upon both dispersion and digestion. Therefore, these aspects are also important to consider when developing a LBF. Furthermore, the solution stability, content uniformity, capsule filling and precipitation are also important aspect to consider later in the development phase (Feeney et al., 2016). A study by Bennett-Lenane et al., (2021) showed the utility of using statistical modelling to predict the biorelevant solubility change of drugs in two types of LBFs. This study provided important and valuable information about solubility upon dispersion in biorelevant media and the use of predictive tools in drug product development. However, a key limitation of the study was that the effect of digestion of the LBFs and the effect on solubility thereof was not considered. The digestion is essential to assess as most excipients used in LBFs are naturally digestible in the intestine which can significantly affect solubility (Koehl et al., 2020; Zupančič et al., 2023).

Accordingly, the objective of this study was two-fold; firstly, to demonstrate a broader utilization of the higher throughput in vitro lipolysis setup by assessing the influence of both dispersion and digestion for a range of drugs within an LBF. Secondly, to enhance the use of computational approaches in LBF strategies by determining whether the solubilised drug concentrations change pre- and post-digestion can be accurately captured in a statistical prediction model, and development of two new interpretable equations which will allow rapid assessment of the viability of the LBF approach.

Table 2. Composition of the LBF investigated.

Formulation	Excipients
Medium-chain LBF	40% Miglyol 812
	40% Tween 85
	20% Cremophor RH 40
Long-chain LBF	40% Olive oil
	40% Tween 85
	20% Cremophor RH 40

 

Download the full article as PDF here Predictions of biorelevant solubility change during dispersion and digestion of lipid-based formulations

or read it here

Lotte Ejskjær, René Holm, Martin Kuentz, Karl J. Box, Brendan T. Griffin, Patrick J. O’Dwyer, Predictions of biorelevant solubility change during dispersion and digestion of lipid-based formulations, European Journal of Pharmaceutical Sciences, 2024,106833,  ISSN 0928-0987, https://doi.org/10.1016/j.ejps.2024.106833.