OR WAIT 15 SECS
Cynthia A. Challener is a contributing editor to Pharmaceutical Technology.
Knowing the source and understanding the impact on CQAs is crucial to optimum drug formulation and processing.
Excipients are essential components in drug products. Although sometimes defined as “functional” or “inert,” each has a specific role within the formulation that is necessary for the overall function of the medicine. Variability in the properties and performance of excipients may, depending on the nature of the variability and the role and quantity of the excipient, have a significant impact on the safety and/or efficacy of the final drug product. Managing-minimizing-excipient variability is, therefore, an essential part of drug product development and manufacture.
Excipient variability depends on many factors. One important determiner relates to whether the excipients are prepared from natural materials or via chemical synthesis, according to Liam Cullen, an engineering specialist with Servier. For example, he points to cellulose-based materials, which are derived from wood pulp that is then chemically treated. “Natural materials, particularly those that are grown, can be subject to environmental conditions such as soil type, weather, and the use of fertilizers. Secondary processing steps can also provide further material variability,” he explains.
Materials that are manufactured by chemical synthesis alone tend to have less variability, but changes in the manufacturing environment, including the raw materials used, processes, and manufacturing locations can occur. For instance, Cullen notes that suppliers may elect to move the manufacturing of an excipient from one plant to another for cost reasons, and then replace some or all of the raw materials with locally sourced alternatives. “While the overall material specification ranges do not change, the once-typical values may be impacted,” he says.
Variability is most commonly encountered at ACD/Labs in the impurity profiles of excipients, according to Andrew Anderson, the company’s vice-president of innovation and informatics strategy. Generally, excipient attributes such as rheology, granulometry, viscosity, and impurity profiles have been better indicators of change compared to more specific material characteristics such as purity, appearance, or loss on drying, according to Cullen.
Because both changes in the raw materials and processes used to manufacture excipients can impact final drug product performance, Mike Tobyn, research fellow in the materials science and engineering group at Bristol-Myers Squibb (BMS) stresses that it is important to understand the sources of variability for any excipients used in drug formulations.
In addition, he observes that the analytical methods used to generate results for certificates of analyses (CoAs) and by pharmaceutical companies upon receipt of excipients are often an unappreciated source of variability. “When debating whether a material is ‘different’ based on these results, it is necessary to consider whether the variability is in fact due to the different capabilities of the methods being used. In some cases, a change in the analytical results may be due to analytical ‘noise’ and not because the attributes of the material have changed,” states Tobyn.
Understanding the impact of variability in excipient properties or specifications on the critical quality attributes (CQAs) and performance of the final drug product is key to developing a robust strategy for selecting the right formulation and process that will result in a robust drug product, according to Anil Kane, executive director and global head of technical and scientific affairs at Thermo Fisher Scientific. “Challenges (and failures) start in development. Hence it is key to evaluate variability in the excipients, the API, and the process parameters during the development and pilot stages before pivotal clinical trials,” he states.
Excipient variability can impact product consistency, stability, and dissolution profiles, among other properties. Most often these impacts occur with excipients that have functional roles, such as in drug release. Variability in inert excipients (diluents, fillers, etc.) can have an impact on product performance, but the impacts may be more difficult to detect at the development stage, according to Cullen.
At both the development and commercial stage, excipient variability can result in a batch failing to meet specifications, which equates to a loss of time and money, according to Tobyn. “Failure could be obvious, such as sticking or tablet breakage for oral solid dosage products. Or it could be more subtle, such as interference with an established analytical method. In the latter case, the process understanding and control provided by the method may no longer be in place, which slow down release or threaten the batch,” Tobyn says.
While all levels of development and commercial production can be affected by changes in the performance of excipients, some stages may not display any negative effects. “It is important to take a holistic approach that considers all of the information gathered from various sources and includes comparison with historical data as well as computer modeling of the data gathered,” Cullen asserts.
He notes as an example issues that can develop with particle segregation that typically aren’t detected at development scale but become noticeable in the plant. In development, the distance from the feed intermediate bulk container to the tablet press feeder is approximately one meter; in commercial production, the feed IBC can be on the floor above the tablet press, and the acceleration involved when the material is released from the IBC to the press can result in a de-mixing phenomenon called segregation that affects smaller particle sizes in the feed mixture. “This issue is not immediately obvious in production but is normally identified during validation or standard product testing,” Cullen adds. As a result, he suggests that design-of-experiment (DoE) and quality-by-design (QbD) models currently used in smaller-scale development should be expanded to commercial-scale activities.
Lack of variability at the development stage can also be a problem when processes are scaled up, according to Tobyn. “It is important during development to establish potential excipient variability issues that might occur and how they might affect product quality and performance. Tales abound of products developed using unusual or atypical batches throughout the development stage with no issues observed that fail in the plant due to an unexpected change. The formulator must work to avoid this situation,” he comments.
In continuous manufacturing, material characterization of each and every excipient is extremely crucial because continuous processes rely completely on uniform continuous blending and feeding to tableting or encapsulation machines, according to Kane. Physicochemical properties such as particle size distribution, densities, flow properties, surface morphology, and other characteristics play a critical role in developing an efficient continuous process.
Unlike for conventional processes where the first step is usually blending, Tobyn adds that in a continuous process there is a period where the excipient is not blended but is transported via a feeder to the blender. “The flow of this powder is a consideration that does not exist for batch processes. Although loss-in-weight feeders can cope with changes in flow and feed rates, it is still beneficial to keep these changes to a minimum,” he notes.
Here again, employing QbD for continuous manufacturing is an effective approach to risk mitigation, according to Anderson. He also notes that the use of process analytical technology is imperative for QbD implementation.
Drug makers face many challenges when approaching the management of excipient variability. One basic issue is the locations of excipient production, according to Cullen. Excipients are often not manufactured nearby or even in the same country or on the same continent, particularly for specialized materials. “Issues of distance, language, and time zones make it challenging to maintain concurrent or even regular contact,” he notes. Online help desks often provide limited support due to insufficient staffing or unwillingness of suppliers to share process or material sensitive information, even where dedicated secure client accounts have been created, adds Cullen.
There are also different types of suppliers, according to Cullen. Some are dedicated to the pharmaceutical industry and understand the extensive requirements of the industry. Others serve multiple sectors with the pharmaceutical industry comprising a significant part of their businesses. These suppliers generally recognize the special requirements of their pharma customers, but may need some additional support. There are also suppliers for whom the pharmaceutical industry is a small percentage of their business. These suppliers may not fully understand the specific needs of pharma companies and communication may be difficult.
As one specific example, Cullen notes that some suppliers will indicate on CoAs that a specific attribute “complies” with the specification rather than providing an actual value, which makes it difficult to identify variations and to use data for trending analyses to determine patterns of variability that can assist in understanding the subtle changes in the CQAs of an excipient.
A deeper understanding of how variability in excipients can affect drug product performance and proposed control strategies is essential for improving drug product development, according to Kane. Gaining that understanding can be difficult, however, because pharma companies don’t have the internal capability to manipulate excipients and their manufacturing processes. “Variability in excipients is a function of the control strategy used by the suppliers of these materials, and it can be challenging for drug makers to obtain ideal sets of samples for adequate investigations,” he observes.
Furthermore, the number of excipient material properties combined with the number of excipients in a drug product formulation present cost and logistics challenges for executing manageable experimental designs. “Risk-based approaches may be helpful for identifying the excipient material properties (typically functional) that have the greatest impact on product performance,” says Kane.
In addition to finding sufficient time to run a systematic DoE to evaluate the influence of excipients, Thermo Fisher is often also challenged by the need to work with minimal quantities of API at the clinical and pilot scale, according to Kane. For ACD/Labs’ customers, data analysis and reviews are a challenge, and the company is working with a variety of customers to address this issue, according to Anderson.
Tobyn agrees that there are business and quality barriers to completing the processes necessary to ensure that excipient variability is minimized from the start. “The technical processes to establish which materials have low variability exist. We currently need the business processes to exploit that,” he adds.
One of the most common tactics for pharma companies is to, where possible, choose excipients that have worked well in the past and to purchase them from reliable suppliers with a track record of providing excipients with minimal variability.
For instance, BMS uses platform formulations that allow the company to work closely with and gain a greater understanding of a limited number of materials. “Because the platform is known, and how those materials interact has been demonstrated, the only material ‘variable’ is the API,” Tobyn remarks. He adds that the platform utilizes high-volume, widely available excipients that are expected to demonstrate lower variability.
BMS has also used a range of techniques to identify vendors with proven track records in producing low-variability materials. “Use of these vendors, along with ongoing monitoring of the variability of their products, provides a stable base for the platform,” Tobyn explains. Other vendors can be introduced, along with more variability, at key stages if the business requires.
Beyond the use of well-known excipients, the most basic step at Servier is to monitor and analyze the data from both CoAs and on-site testing results, according to Cullen. Functional excipients are monitored more closely than less functional materials, but all can have an impact. The data are trended along with the performance data of the finished product, and over time it is possible to identify which materials have the greatest effect on finished product. Servier also conducts regular supplier audits.
Once the most impactful excipient material properties are identified, this understanding can be combined with knowledge of the API properties and the process parameters used to manufacture the drug product to develop an appropriate control strategy that ensures consistent supply of safe and efficacious drug product, according to Kane. He also stresses that the experience of formulation scientists is essential for developing robust optimized formulations.
In addition, employing a QbD approach, and specifically accounting for each material’s CQAs, can help with developing appropriate procedures and mitigating the risk of commercial production failures when any excipient has attribute variability, according to Anderson.
Of course, suppliers play a crucial role in reducing excipient variability. In general, BMS has found that the more experience a vendor has with a material the less variable it is likely to be, according to Tobyn. “A major excipient that has been produced for 50 years on a tonne-per-day basis and is in thousands of products is less likely (although not totally free from risk) to have variability that will influence final drug products,” he says. On the other hand, suppliers of niche, specialist, novel materials may need some time to develop process and product consistency.
Cullen would like excipient suppliers to openly discuss planned and unplanned process changes with their suppliers, provide specific values of CoAs, not just “Complies”, and prepare annual summaries of CQAs for their processes and finished materials that also include timelines for any changes. He would also like to see more direct communication and opportunities for site visits to suppliers, such as workshops and plant tours, so pharma companies can gain a greater understanding of excipient manufacturing processes.
Excipient suppliers should, in fact, be open to sharing their data, with appropriate controls, on the variability of their products, according to Tobyn. He notes that some vendors, mainly the ones who are aware that their products are low variability, do so, but others are reluctant. “Data sharing would allow customers to make informed decisions about which materials are low and high variability and potentially provide pressure for vendors with high variability to improve that variability,” he observes.
BMS would also like to monitor excipient properties in real time with vendors in order to be able to select batches during the purchasing process, either in a regular process or opportunistically when suitable batches became available. “This type of purchasing, however, will require new processes for our company and the vendors,” Tobyn notes. In addition, he would like to see more focus on analytical methods and reduction of analytical “noise” so that “real” variability can be readily probed, identified, and minimized.
Both Kane and Anderson would like to see more excipient suppliers adopt and apply QbD principles in the manufacture of their excipients.
Essential to minimization of excipient variability is the building of relationships with suppliers, according to Cullen. “Waiting to commence a dialogue with suppliers until an issue arises or until it is time for an audit does not provide good results. It is important to establish and maintain contacts within supplier organizations so that issues can be discussed freely and directly as they occur or ideally in advance of their occurrence where issues during the manufacture of an excipient have arisen. It is an area for constant development. Drug manufacturers must work closely with their excipient suppliers at all times,” he asserts.
Vol. 43, No. 5
When referring to this article, please cite it as C. Challener, “Avoiding Excipient Variability,”Pharmaceutical Technology 43 (5) 2019.