Data-driven insights into the characteristics of liquisolid systems based on the machine learning algorithms

Abstract

Liquisolid systems (LS) represent a formulation approach where liquid drug or its dispersion is transformed into a powder with good flowability and compactibility, leading to enhanced drug dissolution and bioavailability. Many research groups have focused on the preparation and investigation of LS, leading to a higher need for comprehensive evaluation of factors impacting LS characteristics. The aim of this work was to investigate the applicability of machine learning algorithms in the LS evaluation, using data mined from published literature, and provide an insight into critical factors governing the liquisolid system performance.

The dataset was prepared using publication search engines and relevant keywords, with a total of 425 formulations included in the database. The database focused on preparation methods, formulation parameters, and liquisolid system characteristics. Subsequently, critical properties of the liquisolid system, i.e. flowability, compact hardness, and drug dissolution, were analyzed using machine learning algorithms, including Gradient Boosting, Adaptive Boosting and Random Forest.

In addition to conventional preparation methods and excipients, novel technologies (fluid bed preparation, extrusion/spheronization) and materials (Neusilin®, Fujicalin®, and Syloid®) enhanced the properties of liquisolid systems. The analysis revealed that formulation factors, such as carrier and coating agent type and content, liquid phase load, model drug type and content, as well as preparation method, significantly influenced liquisolid system characteristics. The models developed exhibited high prediction accuracy when applied on test data (higher than 80%). This indicates that the machine learning models may provide an insight into the critical attributes affecting the LS performance and may be used as a valuable tool in the development and optimization of these samples.

Introduction

Poor aqueous solubility of drugs represents one of the major challenges in formulation development, impacting both bioavailability and therapeutic effect. Approximately 70% of model drugs in the development phase may be classified as “practically insoluble” (<0.1 mg/ml according to the definition provided within the European Pharmacopoeia and USP), whereas around 65–70% of drugs on the market belong to the Biopharmaceutical Classification System (BCS) classes II and IV (Ku and Dulin, 2012). Many formulation approaches have been developed to overcome poor drug solubility, including liquisolid system formulations (Spireas, 2002; Spireas and Bolton, 1998).

Spireas and Bolton (1998) have described liquisolid systems as “acceptably flowing and compressible powdered forms of liquid medications“. Liquisolid systems are prepared by absorption/adsorption of the drug in liquid form, usually a solution or dispersion, on a porous carrier and subsequent addition of a suitable coating material (coating agent). The liquid form of the drug increases its solubility and dissolution rate, resulting in improved bioavailability of the drug.

Spireas (2002) provided the mathematical approach for the development of liquisolid systems, with the relevant formulation parameters:

R-ratio, representing the ratio between the carrier (Q) and coating agent amount (q), and
liquid load factor (Lf), representing the ratio between the amount of liquid drug (W) and carrier (Q).

The carriers incorporated in the liquisolid systems usually have a high specific surface area, high liquid absorption capacity and good flowability (e.g. microcrystalline cellulose, aluminum magnesium silicate) (Vraníková et al., 2021). On the other hand, commonly used coating materials (e.g. colloidal silica) have very fine particles, inferior flow properties and contribute to the absorption of the liquid phase (Vraníková et al., 2021). Although the composition of liquisolid systems is similar to self-emulsifying drug delivery systems (SEDDS) in terms of the solid carrier and the drug in liquid form, fewer excipients are generally added to these formulations (Almeida and Tippavajhala, 2019; Spireas and Bolton, 1998). In addition, liquisolid systems are usually based on a hydrophilic, non-volatile liquid without the addition of oily phase, which allows the use of a less complex preparation technology and generally results in a faster release of the drug from the final dosage form (Almeida and Tippavajhala, 2019; Vraníková et al., 2015b).
Usually, liquisolid systems are intended for obtaining compacts, so flowability and compactibility represent their critical quality attributes (Spireas, 2002). Additionally, these samples may also be filled into capsules (Barmpalexis et al., 2018; Sheta et al., 2020). Recently, liquisolid–based pellets were developed, as a formulation approach combining favorable drug dissolution with high flowability and smooth pellet surface (Lam et al., 2019; Pezzini et al., 2016; Vasiljević et al., 2021). In order to obtain liquisolid systems suitable for further processing, R–ratio and liquid load factor need to be carefully evaluated and optimized.

Liquisolid system technology is relatively simple, based on conventional excipients and methods, and cost-effective, so it overcomes some limitations of other commonly used methods to improve drug solubility (Aleksić et al., 2020). Many research groups have engaged in the preparation and study of liquisolid systems and highlighted their potential (Aleksić et al., 2020; Glišić et al., 2023; Jhaveri et al., 2020; Lam et al., 2019; Molaei et al., 2018; Shah et al., 2023; Vraníková, et al., 2021). Favorable effects on drug bioavailability have also been recently reported in vivo studies, where liquisolid formulations showed faster onset of drug therapeutic effect and higher drug bioavailability compared to conventional formulations after administration to mice, rats, or human subjects (Badawy et al., 2016; Devi et al., 2022; Sheta et al., 2020; Yadav et al., 2024). Due to incorporation of a wide range of excipients and model drugs, the liquisolid systems described in the literature exhibit different physicochemical and biopharmaceutical properties, such as flowability, compressibility and dissolution behavior (Aleksić et al., 2020; Jaydip et al., 2020; Molaei et al., 2018; Shah et al., 2023; Vasiljević et al., 2021; Vraníková, et al., 2021). Therefore, there is an increasing need for a thorough and systematic evaluation of formulation factors and preparation methods that influence the properties of liquisolid systems.

Machine learning (ML) is a branch of artificial intelligence (AI) that focuses on the development of algorithms that enable AI to learn from large and complex data sets. It is a promising approach for identifying hidden patterns underlying the data and provides a higher level of insight compared to basic data analysis and interpretation. ML has already been applied in various fields and recent pioneering trials in pharmaceutical development have also been reported. It has been described as a useful technique for performance evaluation of dosage forms printed by fused deposition modeling (Elbadawi et al., 2020; Obeid et al., 2021), digital light processing (Tagami et al., 2021) and inkjet printing (Carou-Senra et al., 2023), as well as orodispersible films (Turkovic et al., 2024). Due to the high research interest in the preparation and evaluation of liquisolid system, there is a growing interest in a thorough analysis of the relationships between the formulation factors and liquisolid system characteristics. Recently, Glišić et al. (2023) reported that ML algorithms provided a better understanding of the effects of formulation factors and/or tableting process parameters on the flowability and compaction properties of mesoporous silica-based liquisolid systems. This opens up the possibility of ML application in the analysis of a wide range of liquisolid system samples reported in the literature, where the usual data processing and interpretation would be challenging but beneficial for the further liquisolid system development and application. Therefore, the aim of this study was to investigate the applicability and contribution of ML algorithms in the evaluation of liquisolid systems using the data mined from the literature and to provide insight into critical factors governing the liquisolid system characteristics and performance.

Download the article as PDF here: Data-driven insights into the characteristics of liquisolid systems based on the machine learning algorithms

or read it here

Following excipients are mentioned in the study besides others: Neusilin®, Fujicalin®, Syloid®

Ivana Vasiljević, Erna Turković, Jelena Parojčić, Data-driven insights into the characteristics of liquisolid systems based on the machine learning algorithms, European Journal of Pharmaceutical Sciences, 2024, 106927, ISSN 0928-0987, https://doi.org/10.1016/j.ejps.2024.106927.