Quality by Design in Excipients

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology-03-01-2013, Volume 2013 Supplement, Issue 2

Factors for assessing excipient variability, the associated challenges developers need to address to design and manufacture solid oral drug products, and solutions for such challenges are examined.

Managing excipient variability is an essential element in designing and manufacturing robust solid oral drug products and is an integral task in when applying quality-by-design (QbD) principles. The article examines factors in assessing excipient variability, the associated challenges developers need to address to design and manufacture solid oral drug products, and solutions for such challenges.

Excipients play a crucial role in a formulation, and as part of the formulation, influence the quality of the final drug product. Managing excipient variability is an essential element in designing and manufacturing robust solid oral drug products and is an integral task in applying quality-by-design (QbD) principles. Pharmaceutical Technology examined excipient performance in QbD in a webcast (www.pharmtech.com/excipient) in 2012. Participating in that webcast were Catherine Sheehan, director of excipients at the US Pharmacopeia, Ian Robertson, global QbD manager at Colorcon, Chris Moreton, vice-president of pharmaceutical sciences at FinnBrit Consulting, and Greg Amidon, research professor of pharmaceutical sciences in the College of Pharmacy at the University of Michigan and chair of the USP Excipient Performance Expert Panel. In this article, Robertson offers insight on managing excipient variability in a QbD paradigm.

Creative Crop/Getty Images

Assessing excipient variability

PharmTech: What are key considerations in assessing excipient variability and what are the associated challenges developers need to address to design and manufacture solid oral drug products?

Robertson (Colorcon): Understanding the effects of normal excipient variability is important to ensure that potential risks to drug-product quality are effectively managed. With an increasing focus on QbD, both innovator and generic-drug developers will be required to demonstrate excipient risk management in relation to the drug products' critical quality attributes. So, while it is the developer's responsibility to demonstrate robust formulation design and manufacture, it is important that all stakeholders collaborate early in the product lifecycle to ensure mutual success. Often, knowledge and tools to proactively assess the effects of excipient variability upon drug-product quality are available. Of course, this not only provides sufficient development but, importantly, supports implementation of an effective control strategy for subsequent commercial manufacture.

So one may ask, why are excipients variable, and what are the sources of variability? Well, excipients used within the pharmaceutical drug products are diverse in terms of their chemistry, origin, manufacture, and also control. Broadly speaking, we could consider two classes of excipient products. Those described as natural or naturally derived, and those described as synthetic or semisynthetic. The natural or naturally derived excipient products and factors, including the crop source along with environmental variables, such as the growing and processing conditions, can all affect the physicochemical properties of the resultant excipient product. Even for synthetic or semisynthetic excipient, differences in the feedstock, such as the organic and inorganic components from crude oil or gas, along with variability in chemical processing, will all contribute to the overall variability of the excipient product itself.

For an individual manufacturer of an excipient product, there may be lot-to-lot variability of the physicochemical properties. Often, a manufacturer may make related products, such as different physical or chemical grades, or even supply different industries. Therefore, different manufacturing campaigns may be adopted, which can lead to variability associated with switchover, start-up, or shutdown upon a single production line.


Indeed, between the multiple manufacturers of an excipient, it is likely there will be differences in the raw materials and the production process used, the scale of manufacture, and location, all which can contribute to the variability of an excipient product's physicochemical properties.

I think it's also important to recognize the variability in the controls used. Although excipients are required to meet the appropriate pharmacopeia standards, there may be differences in the excipient specifications from vendor to vendor. This can include the ranges specified within standard tests as well as the inclusion of noncompendial physicochemical property tests.

Along with the specified parameters, it is important to also consider both unspecified parameters as well as nonroutine parameters. The variability of unspecified or unreported parameters, could impact drug-product performance depending upon its design and intended use. Although unreported, these properties may be measured by the manufacturer as part of its control strategy. In addition, excipients can vary in what is termed nonroutine parameters (e.g., residual reactants and residual catalysts). Variation of these parameters may not be commonly measured by the manufacturer.

So in the majority of cases, robust drug products are successfully developed and manufactured to manage such excipient variability. Failure, however, to understand and control excipient variability has indeed led to some serious quality issues.

FDA, for example, has spoken publicly about excipient variability and its relation to drug-product recalls from the market. It is clear that some existing development and manufacturing practices can fail to ensure that medicines of the appropriate quality standards that are routinely produced. Furthermore, demonstrating quality by testing does not prevent these substandard drug products from entering the market. Obviously, quality failures, including those related to excipient variability, can have serious consequences to the manufacturer in terms of financial and reputational damage. Ultimately, the end user, the patient, is put at risk.

Substandard drug products reaching the patient can impact safety and efficacy. And even those that are captured and rejected by quality control or those recalled from the market can lead to drug shortages. This, again, impacts the patient. So, with increasing public focus upon the pharmaceutical industry and regulators alike, it is clear that managing excipient variability will be an area of increasing industry importance.

It is critical, therefore, for excipient manufacturers and vendors to closely collaborate with the developer to better reduce the risks of quality failures. This is important from early on in the drug product's development, whereby knowledge and experience of the excipient can proactively support robust formulation design and efficient manufacturing control. Failure to account for normal excipient variability can lead to the design and development of nonrobust formulations. Unfortunately, this can necessitate tighter excipient-specification test limits, batch selection, or even custom manufacture to control the performance of the finished drug product.

It must be recognized that, in some instances, the ability of the excipient manufacturer to meet such requirements may not be easily achieved, if achievable at all. This, therefore, presents additional risk and cost to the overall product.

It also is important to consider the interplay of all variables including those of the API and the manufacturing process. The use of risk assessment is helpful to identify those primary variables that may affect the drug product's critical quality attributes.

So, the ability of the developer to determine the effects and interactions of excipient material attributes, along with API and processing parameters, using approaches such as design of experiments, can provide valuable insight into their true criticality upon drug-product quality.

Using QbD, therefore, presents an opportunity to ensure quality is built in by design. Early collaboration by all stakeholders can better ensure that the potential risks of normal excipient variability can be efficiently designed out by our prior knowledge and available tools.


PharmTech: Can you elaborate on that knowledge and related tools and solutions.

Robertson (Colorcon): As I mentioned previously, formulation design is critical to ensuring robust product performance. And to illustrate this point, I would really like to show an example of how normal excipient variability of a rate-controlling polymer, hypromellose (METHOCEL K15M CR) was determined for drug release from a hydrophilic extended-release matrix. The study examined the importance of METHOCEL concentration in regard to managing variation of particle size, viscosity, and percent hydroxylpropoxyl (HP) substitution or percent HP material attributes. And the ranges studied were at the extremes and center point of the cell-specification limits.

So using propranolol hydrochloride as the model drug, it could be seen that drug release was slower when one increased the METHOCEL concentration from 15% to 30% (w/w), and this, of course, may be expected. The drug release was generally consistent across the extremes of viscosity and percent hydroxylpropoxyl irrespective of drug concentration. Therefore, the formulation was robust to the normal variation of these two material attributes.

Interestingly, however, for particle size it was clearly shown that lower METHOCEL concentration, 15% (w/w) resulted not only in a difference of release rate, as seen previously, but also increased tablet variability. Here, faster and more variable drug release was shown for the formulation containing the core subparticles of METHOCEL K15M CR polymer. At 30% METHOCEL concentration, however, this variation was not observed. So robust dissolution performance was shown irrespective of the particle size.

These examples show that it is prudent to determine if such material attributes are a potential risk to drug-product quality attributes during early formulation development. Clearly, in this stage, it is generally easier to adjust the formulation and/or the process to manage such effects rather than implementation of changes during the latter stages of drug product development.

The impact of normal excipient variability, however, can also be managed in other ways. Certainly, QbD supports approaches to develop formulation flexibility or formulation design space. Adjustment of the quantitative composition of the drug-product batch can be justified to manage the variability of the ingoing excipient material attribute. As discussed in ICH guidelines, enhanced knowledge of material attributes is needed to develop and justify such an approach. Nevertheless, it does provide an opportunity for both the developer and the excipient vendor to further collaborate and ensure that the necessary physicochemical properties of the excipient can be evaluated to develop such approaches.

Furthermore, it also is important not to view the excipient sales specification as the endpoint for control of excipient variation. Excipient sales specifications and test limits can, of course, be variable. Therefore, these may not be representative of the excipient manufacturer's control capability. And I think this is an important point, particularly for risk assessment. Enhanced knowledge here can reduce the perceived risk of excipient property variability upon the drug-product quality attributes. And, indeed, excipient manufacturers may be able to provide prior knowledge of their capability, for example, demonstration of process capability, representative material-attribute trend data, or even descriptive statistics illustrating the control for the individual manufacturing approach.

In the propranolol matrix case study, another tool available to the developer is that of the excipient material attribute sample. These may be available to assist the developer in proactively testing the effect of normal excipient variability upon its particular formulation. It may not always, however, be useful to study all of the excipient material attributes, particularly if the manufacturer's control is tight and if variability lot-to-lot is insignificant. Such approaches, however, may be valuable to support experimental design and subsequent development of design space.