Understanding the impact of excipient variability on oral solid dosage form performance: A possible role for multivariate data analysis, in the form of principal component analysis

Abstract

The development of a robust oral solid dosage form requires knowledge of all the sources of variation that could impact the dosage form’s performance and stability. One of those sources of variability is in the excipients that can make up the bulk of the dosage form. Raw material and manufacturing variability can make a difference to the physical properties and performance of the excipient. It is, however, also important to note that sources of variation in data include analytical and sampling variation, which affect the reported data without directly impacting the performance of the dosage form. Working closely with an excipient supplier, drug product manufacturers can substantially improve their understanding of the sources of variability in the excipients that they use. This helps to strengthen the regulatory design space for their asset and will helps stabilise the product during its commercial life cycle. The use of multivariate analysis to examine reported data, along with targeted experimentation, is a demonstrated way to work towards this goal. In this paper we give examples of how this can be achieved.

Introduction

Quality by design and product robustness

In Solid Oral Dosage forms variability of excipients has been shown to be a potentially consequential issue affecting the performance of medicines, as captured in Table 1. In taking into account the variability of excipients, it has been posited that, in some cases, excipients used at lower than recommended levels are more likely to manifest changes in dosage form performance due to excipient variability.1

Table 1: Some reported examples where excipient variability can affect the performance of oral solid dosage forms.

Issue	Excipient (s) implicated	Reference
Effect on dissolution of tablets	Crospovidone	2
Drug release from matrix tablets	Xanthan gum	3
Dissolution changes in oral solid dosage forms	Magnesium stearate	4, 5
Impact of superdisintegrant variability on dissolution	Sodium starch glycolate, Croscarmellose sodium	6
Performance of amorphous solid dispersions	Povidone K30	7
Performance of matrix tablets	HPMC K4M	8

Traditional, well-established pharmaceutical batch manufacturing processes rely heavily on end-product testing to ensure final product quality.9 If product critical quality attributes (CQAs) are not met, a problem-solving exercise is initiated and, as a result, corrective actions are implemented to ensure quality in future batches. Ensuring quality and safety in the production of medicines is critical to protect patients from potential health risks, rigorous regulatory guidelines – including those from ICH, FDA and EDQM – require continuous oversight, optimisation and swift response to any quality issues or defects that arise.10,11 Pharmaceutical Quality by Design (QbD) represents a paradigm shift in the approach to pharmaceutical development and manufacturing, focusing on building quality into the product from the initial stages onwards.12,13 This pro-active approach integrates quality considerations into the earliest stages of product development, thereby reducing the likelihood of quality issues arising later. A key principle of this paradigm, product and process robustness, is achieved when a formulation and its manufacturing process can consistently deliver the desired quality, despite variations in APIs, excipients, equipment, and environmental conditions. The importance of looking into excipient variability in a QbD approach, not only between materials but also between manufacturers of the same excipient, has been highlighted previously.14,15,16,17,18 By anticipating and accommodating variations, resilient and reliable processes and products can be developed.

Excipients in oral solid dosage forms

Pharmaceutical excipients are substances other than the pharmacologically active drug or pro-drug, which are included in the manufacturing process or in a finished pharmaceutical product dosage form.14 Typically, excipients play a crucial role in the structure, delivery, and stability of a formulation. The selection of excipients for a formulation depends on numerous factors, including the function, dosage form, manufacturing process, compatibility with other ingredients, regulatory status, stability and the ability to create a robust final drug product.19 Excipients can be added to a formulation to serve different functions. They can for example be added to the core of a tablet as antioxidant, anti-adherent, bulking agent, diluent, binder, disintegrant, lubricant, glidant, preservative, or stabilizer20

Drug product manufacturers must evaluate the safety of excipients before they can be used in their formulations.21 Excipients should be inert and non-toxic, they should be approved and listed for use by the regulatory agencies (such as US FDA or EMA), either by prior use in a dosage form or by inclusion of a package of information in the regulatory dossier for a new registration. Additionally, excipients must be compatible with the API and each other. It is preferable but not absolutely necessary they should be produced IPEC-GMP following the IPEC-Excipient GMP guidelines to ensure quality.22 Controls, such as packaging and manufacturing, along with details of storage conditions, also need to be addressed.

In some cases, materials used as excipients have purposes as items of commerce beyond use in pharmaceuticals, and pharmaceutical excipients may be a small part of the overall volume of the material. For instance, it has been suggested that only 1 % of the volume of titanium dioxide used worldwide is in form of E171 for food and pharmaceuticals, where it is used as a stability and dosage form identification aid.23-25 However, this small part of the volume (if not the value, as pharmaceutical excipients command a higher price than other higher volume uses) of the material does not negate the need to ensure that the necessary manufacturing protocols are followed, and the requisite processes are in place to ensure that the material meets the required standard. Controls and/or guidance on how to package and store the excipients, in both unopened and open conditions, should be available to the customer.

Additionally, excipients must be compatible with the API and the other excipients. The acceptable variation in material properties of an excipient typically depends on the robustness of the formulation and the manufacturing process in which the excipient will be used. Overall, the selection of excipients is a multifaceted process that requires thorough understanding of the formulation, manufacturing process, regulatory requirements, and patient needs.

Introduction to multivariate analysis (MVA) and use in pharmaceutical data analysis

Pharmaceutical excipients are routinely characterised by suppliers and pharmaceutical companies to ensure compliance with the registered specifications and to ensure the quality and safety of the finished drug products. Typically, several parameters are routinely tested, with the exact number and type depending on the chemistry and function of the material. Many excipients are manufactured in large volumes by continuous or semi-continuous processes, with large numbers of batches being produced, efficient data analysis methods are needed to handle the complex data sets generated from testing multiple lots of excipients across various attributes. These data-rich environments are well-suited to analysis using “big data” techniques, many of which are already being used extensively in the pharmaceutical industry.²⁶^,²⁷^, ²⁸^,²⁹^,³⁰^,³¹

One of these techniques, multivariate analysis (MVA), encompasses a range of statistical methods which are widely utilised in the pharmaceutical industry for its capacity to extract insights from large and complex data sets.³² Within these, principal component analysis (PCA), a multivariate projection method that compresses the variation of large data sets into a smaller number of latent variables, is well established as a tool to enable extraction of information from pharmaceutical material characterisation data.²⁸^,³⁵^,³⁶ PCA simplifies data representation by focusing on key sources of variability in a data set, while ignoring less relevant noise and correlated measurements. PCA reduces dimensionality by expressing data with fewer variables that are linear combinations of the original variables, which are called principal components (PCs). A detailed description and geometrical interpretation of PCA approaches, methods and algorithms can be found in Esbensen and Geladi³³ or in Geladi and Grahn.³⁴ PCA improves data interpretability and is commonly used to interrogate large data sets and to reveal the underlying relationships in the data (between samples and between variables). This includes the identification of patterns, trends, groups, outliers and significant variables.

Sources of variation in data in reported data

Material variability

All production processes, including those used to manufacture APIs and excipients, have some inevitable degree of variation³⁷ introduced by wear and tear of equipment, fluctuations in environmental conditions, staff changes, and variability in raw materials. This variation (termed common-cause variation) does not indicate that a process is out of control, rather that not every batch will be identical, and will not, of itself, lead to issues in pharmaceutical manufacture, but it is crucial for pharmaceutical manufacturers to understand its impact on the final dosage form. For instance, variations in raw material attributes, such as particle size distribution or polymorphic forms, can significantly impact the physical and chemical properties of the final product.³⁸ Additionally, equipment-related variations, including fluctuations in temperature, pressure, or mixing dynamics, can introduce process variability, potentially affecting the uniformity and quality of pharmaceutical products.

A neat approach to capturing and delineating the sources and consequences of raw material variability has recently been published.³⁹ In the context of formulation development, the hierarchy of variability starts with the most fundamental level, where the presence, amount, and type of an excipient is determined. As development progresses, the focus shifts to different grades and suppliers. Finally, batch-to-batch and within-batch variability are considered to ensure consistent quality in routine manufacturing. This rational hierarchy approach ensures that robustness demonstrated at higher levels extends to subsequent levels, making the development process more efficient and comprehensive.

Sources of raw material variability

It is important to acknowledge that even seemingly simple excipients may exhibit inherent complexity.⁴⁰ The raw materials used to produce excipients can significantly influence the final product properties and their performance. Commercially available starches, for example, can have similar specifications but may be derived from different botanical sources, like corn, potato, or tapioca. The botanical source of starch has been reported to affect amongst others the solubility, moisture content, and gelatinization properties.⁴¹

But even if the botanical source is the same, variability may be intrinsic in natural excipients. Imagine, for example, excipients with complex heterogeneous compositions, for which the polymerisation process has taken place in nature. The chemical composition for these kind of materials can vary as the result of natural variation based on geographic location, climate, or season.⁴²

Raw material variability could however also originate from non-natural sources. Different chemical agents used as raw material in a production process, could result in different impurity levels or different performance attributes (as with drug manufacturing). An example for this kind of variation is the type of cross-liking agent that is used to create the superdisintegrant sodium starch glycolate (type A). The type of cross-link has been shown to affect the resistance to hydrolysis and therefore the stability of performance substantially in wet granulation under alkaline conditions.⁴³ It is however not possible to obtain information on the cross-linking agent from the specifications or the Certificate of Analysis (CoA).

In a previous publication,¹⁷ it was demonstrated that MVA could be used to identify variations in the properties of an incoming raw material (ethylcellulose) that could help explain small differences in performance of an excipient for controlling release.

Download the full article as PDF here Understanding the impact of excipient variability on oral solid dosage form performance

or read more here

Understanding the impact of excipient variability on oral solid dosage form performance: A possible role for multivariate data analysis, in the form of principal component analysis, Janssen, Pauline H.M. et al., Journal of Pharmaceutical Sciences, Volume 114, Issue 12, 104021