A practical computational tool to predict formulation and process variables during the development of spray-dried amorphous solid dispersions

Íris Duarte1,2, José Luís Santos1, João F. Pinto2, Márcio Temtem1*

1 Hovione Farmaciência SA, Sete Casas, 2674-506 Loures, Portugal; *mtemtem@hovione.com or +351 219 847 569

2 iMed, Research Institute for Medicines and Pharmaceutical Sciences. Faculdade de Farmácia da Universidade de isboa, Av. Prof. Gama Pinto, 1649-003, Lisboa, Portugal

Today, 70 to 90% of the new chemical entities (NCEs) in the pharmaceutical pipeline belong to BCS Class II or IV, thus presenting at least solubility constraints [1]. Promising therapeutic molecules with limited aqueous solubility is a major concern especially when considering the development of oral- dosage forms. To overcome this issue a number of enabling platforms have been used, such as particle-size reduction, use of surfactants, salt formation, complexation with cyclodextrines, production of self-emulsifying drug delivery systems or amorphous solid dispersions (ASDs) [2].

The development of new amorphous solid dispersions has increased exponentially since its appearance in the late 90â??s [3]. ASDs have become popular among the pharmaceutical industry and their efficacy is supported by the significant number of marketed products (e.g. Incivek® (2011), Zelboraf® (2011), Kalydeco® (2012)). The requirements for a successful development of a solid dispersion are related with performance and chemical/physical stability aspects. In what regards the former, an amorphous formulation should present an acceptable dissolution profile and should be able to sustain supersaturation in solution sufficient time to allow for absorption. In what concerns physical

stability, this concept is related with the inherent perceived risk of drug recrystallization from the amorphous state, followed by consequent loss of the solubility advantage.

The development of ASDs with the desirable long-term stability and performance is a challenging process, due to the wide number of formulation and process variables that influence both physical and chemical properties of the product (e.g. several existing polymeric stabilizers, surfactants, different drug-stabilizer ratios, solvents, preparation methods, temperatures, etc). In most cases the selection of the best formulations is based on trial and error experiments, combined with the experience of the researchers themselves. The development of rationale screening programs is of utmost importance for narrowing the scope of formulations and to rapidly identify suitable systems with synergistic interactions for subsequent clinical evaluation.
In the scope of this work the applicability of a computational tool to guide the early identification of potential physically stable systems was evaluated, being based on molecular miscibility predictions between the drug and the polymer. The final goal is to include this tool in a screening methodology that is under development for the production of stable ASDs. This screening methodology aims a reduction in costs, time and risk during product development.
The computational tool is based on diffuse interface theories and presents advantages over commonly applied strategies to guide rationale polymer selection and drug load [4]. Examples of current
approaches are the analysis of the Hildebrand and Hansen Solubility Parameters (SPs) to infer about drug-polymer compatibility [5] or, more recently, the implementation of lattice-based solution theories, such as the Flory-Huggins (F-H) theory to study the thermodynamics of drug-polymer mixing [6]. The most significant differences of the computational tool presented herein over other approaches is the potential to evaluate a ternary system made of drug, polymer and solvent (the traditional approach considers two-components) and the consideration of time-dependent phenomena, such as components mass diffusion and solvent evaporation. This model includes the effect of Thermodynamics, Kinetics and Evaporation (i.e. process variables) on the phase behavior of drug- polymer amorphous systems, and is hereafter named Ternarius. It can be regarded as a pre- formulation tool in the development of amorphous dispersions using spray drying (Figure 1).

Figure 1: Input parameters of the model and two limiting situations given by the tool (outputs).

To assess the applicability of this tool and obtain experimental evidence of the theoretical miscibility estimates, the development of solid dispersions of two poorly-water soluble drugs was considered. The first molecule tested was itraconazole, a BCS Class II drug, commonly used through literature as a model drug. Itraconazole was combined with structurally different polymers, namely Kollidon VA64, Eudragit EPO and HPMCAS. The second molecule (hereinafter defined as Drug A) is poorly permeable and has a Tm/Tg ratio of about 1. Drug A was combined with Kollidon VA64, Eudragit L100, HPMCAS and HPMC. Solid dispersions were produced using different solvent-based methods, namely solvent casting and spray drying. The manufacturing conditions defined for each system were simulated using the Ternarius model. The miscibility of the casted films and spray dried powders was
evaluated using modulated differential scanning calorimetry (mDSC), considering the number of Tgâ??s detected and the presence of endothermic peaks (potential crystallization).
As an example of the predictive capacity of this tool, Figure 2 compiles the predictions given by Ternarius and the experimental results for the different ITZ-polymer systems studied at increasing drug loading. The uncertain region bars correspond to the concentration region that includes the drug loading, from which phase-separation is observed.

Figure 2. Summary of the results obtained: early miscibility and phase behavior predictions (Ternarius) and physical stability assessment by mDSC (SC and SD).

According to Figure 2 and comparing the predicted and experimental results, two important conclusions can be drawn. First, Ternarius can be successfully used in the ranking of the most stable amorphous formulations and with the highest API/polymer ratio, and secondly, it can lead to a reduction in the number of experiments needed to the area of uncertainty, reducing time and resources.
The methods and correlations used to estimate the F-H interaction parameters (??!" ) (Figure 1) are very important to Ternarius accuracy. In this first case study, the ??!" values were obtained from melting
point depression experiments and from the Hansen solubility parameters [7]. These methods, despite being fast and simple to implement, suffer from inherent limitations, which may introduce bias into the system.
In the second case-study with Drug A, the solubility parameters and ??!" values were also obtained
from an inverse gas chromatography experimental methodology using the second generation of
surface energy analyzers (SEA, from SMS). Results obtained using the two methodologies were compared with the experimental miscibility results obtained from solvent casting and spray drying. This allowed further validation of the model and a detailed understanding regarding the importance of the accurate determination of the F-H interaction parameters, which is the thermodynamic component of the model.
On the basis of the computational predictions and experimental results obtained, it can be concluded that Ternarius is useful as a screening tool in the sense that supports a more detailed understanding of the synergistic effects of formulation parameters affecting physical stabilization in ASDs. At this point of the work, by using the model it was possible to set a polymer ranking which was consistent with the experimental data. By ascending order of miscibility capacity and physical stability for the ITZ systems tested the following ranking was obtained: Eudragit® EPO<< PVP/VA 64 < HPMCAS-MG. Moreover, and although there were some differences between the predicted and the experimental
maximum drug load attainable, it is still possible to use the information given by the tool to create
guidelines to define a narrow drug load range to be tested in the following stages of process development, thus saving time and resources.

References:

[1] Smithey, D. et al. Amorphous Solid Dispersions: An Enabling Formulation Technology for Oral Delivery of

Poorly Water Soluble Drugs. AAPS Newsmagazine, 16 (1), 10-14 (2013);

[2] Perrie Y., Rades T. Pharmaceutics - Drug Delivery and Targeting: Pharmaceutical Press; 2010;

[3] Sekiguchi K., Obi N., Studies on absorption of eutectic mixture I: A comparison of the behavior of eutectic mixture of sulfathiazole and that of ordinary sulfathiazole in man. Chem Pharm Bull, 9, 866-872 (1961);

[4] Saylor, D. et al., Predicting microstructure development during casting of drug-eluting coatings. Acta

Biomaterialia, 7, 604-613 (2011);

[5] Van Krevlen, D. W. and Nijenhuis, K. te. Properties of Polymers. 4th Ed., Elsevier (2009); [6] Flory P., Thermodynamics of High Polymer Solutions, J Chem Phys, 9, 660 (1941);

[7] Tian Y. et al., Construction of Drug?Polymer Thermodynamic Phase Diagrams Using Flory?Huggins Interaction Theory: Identifying the Relevance of Temperature and Drug Weight Fraction to Phase Separation within Solid Dispersions. Mol. Pharmaceutics, 10, 236-248 (2013).