Flexible Approaches to Accommodate Excipient Variability Using the Principles of QbD

This article is part of PharmTech's supplement "Solid Dosage and Excipients 2010."

Pharmaceutical formulations generally comprise the active pharmaceutical ingredient (API) and excipients, which are then combined using specific processing. Quality by design (QbD) requires understanding the raw-material variabilities and their effect on finished-product quality. Traditionally, excipient-supplier involvement in the formulation-development process has been limited, with the supplier often unaware of the application or functionality, and the user unaware of supplier's process capabilities. This lack of mutual understanding must change in order for QbD to succeed.

Excipients are a potential source of product variability, and they are generally less well characterized than APIs. QbD, therefore, has emphasized the role of excipients—once referred to as "inert" ingredients—and they are now properly recognized as enabling the API to be converted to a medicinal product that can be administered safely and efficaciously to the patient. Excipients are also now subject to increasing scrutiny from users, regulators, pharmacopeias, and other standard-setting and educational bodies such as the International Pharmaceutical Excipients Council (IPEC), the International Society of Pharmaceutical Engineering (ISPE), the American Society for Testing and Materials (ASTM), and the National Institute for Pharmaceutical Technology and Education (NIPTE).

This paper will discuss several topics related to excipients in a QbD context (1, 2), including: excipient functionality; approaches to raw-material variability; excipient manufacture and quality control; functionality versus composition; and the building of flexibility into the development process (2).

Excipient functionality

Traditionally, pharmacopoeias have emphasized safety and purity of raw materials and focused on chemical composition. Excipient monograhs do not address efficacy or functionality, and typically include identification tests, possibly assays, tests related to minor components, and occasionally, a limited series of physical tests intended to further characterize the material. Specification of pharmaceutical excipients has thus generally emphasized consistency of composition rather than consistency of functionality. Although such chemical and physical testing may be insufficient to guarantee functionality, it is useful for grade specification. This difference may be due in part to the fact that the use and performance of an excipient in a given process and formulation may not be an entirely intrinsic raw-material property.

For example, biopolymers such as carrageenan are sold "pure" for pharmaceutical use, whereas for food use, the same materials are diluted with other food-grade materials to provide a consistent or standardized functionality. This functionality is usually defined, along with agreed-upon methods, in the purchase agreement. The pharmaceutical industry has been somewhat unique in the past in that it has promoted consistency of composition at the expense of consistent performance.

Another example of consistency is control of color. Because of the visual impact, even pharmaceutical users will prefer constant color to constant composition when offered the choice. For example, a supplier producing an orange film-coating may formulate the product using not only an orange pigment, but also small amounts of red and yellow pigments. This way, if there is batch-to-batch variability in the orange pigment, adjustment of the proportions of red and yellow can be used to guarantee a consistent orange color. Slightly different quantitative formulations would be necessary for each batch of constant-color excipient.

Pharmacopeial compliance has often been erroneously perceived as a guarantee of performance for excipients, when in fact, compliance only establishes minimum safety and purity standards. Functionality transcends the molecular composition of the excipient or its particle form, and may depend on the application (reason for use), the formulation (effect of other ingredients), and the process (the details of how the API and excipients are combined together in a particular application). There is often a trade-off between competing formulation objectives, and a given excipient may be multifunctional in a particular application.

Because excipient functionality or performance is application-specific, it can only be properly assessed in the particular application. However, manufacturing a test batch of product for every excipient lot received is not an economical option. Thus, surrogate tests are necessary whereby some type of test—often physical, but sometimes chemical—that correlates with performance in the medicinal product can be used to assess the suitability of a particular lot of an excipient or API.

Most in industry agree that surrogate testing may be necessary and appropriate. Opinion is divided, however, as to the best approach to use. Two approaches to such testing have been proposed. The European Pharmacopoeia (Ph. Eur.) has introduced a nonmandatory Functionality Related Characteristics (FRCs) section to some individual monographs. FRCs, however, as advocated by the pharmacopeia may not be relevant to functionalities in all cases, and they run the risk of overspecification beyond what is required for use in a particular application.

For example, lactose is included in hundreds of dry-powder systems for film coating. The dry powder formulation is reconstituted in water to form the film coating suspension that is then sprayed onto the tablets. One FRC for lactose is particle-size distribution. In this film-coating application, the lactose is used to help disperse the other components and promote the hydration of the powder mix. Particle-size distribution is not particularly relevant in this application because the lactose is dissolving. Obviously, lactose particle size may have an influence on the rate of dissolution of the lactose, but in the context of the application, the effect is not relevant because the hydration of the polymer takes considerably longer than even the coarsest grade of pharmaceutical lactose would take to dissolve. Unfortunately, even though the FRC section of the Ph. Eur. clearly states that the test is nonmandatory, customers frequently demand the information. Many customers consider the monograph specification to be mandatory, even if the particular test is included in the FRC section.

The United States Pharmacopeial Convention (USP) is proposing a new general chapter on excipient performance, General Chapter <1059>. A first draft was published as a "Stimuli to Revision" article in 2007 (3). Based on received comments, USP revised and published a second version of the chapter as an in-process revision in 2009 (4). USP's general-chapter approach is considered more useful because it addresses the relevant attributes pertinent to an excipient application rather than trying to arrive at a set of tests for FRCs that will be applicable for all the typical uses of a particular excipient.

For example, microcrystalline cellulose (mcc) is frequently used in solid dose forms as a filler/binder and disintegrant, and also to improve the rheological properties of the wet mass in wet granulation and extrusion-spheronization. Mcc can also be used, however, in aqueous suspensions as a suspension stabilizer (auxiliary suspending agent) to provide a sufficient disperse phase to make the suspension more stable. The FRCs for Microcrystalline Cellulose Ph.Eur. are particle-size distribution and powder flow. Powder flow is irrelevant to the ingredient's use as a rheology modifier or suspending aid. However, because powder flow is included in the monograph, customers will often request it, despite the fact that the fine-particle grade used in aqueous suspensions does not flow like a powder because it is very fine and cohesive.

New approaches to raw-material variability

Under the FDA 21st-century CGMP initiative and ICH Q8(R2), pharmaceutical excipient users are now encouraged to apply QbD principles to develop formulations and processes that are flexible enough to cope with anticipated raw-material variability and to produce products that are robust and provide consistent performance (1, 2). At one extreme, the process could be controlled by input parameters from the raw materials so that incoming variability could be compensated by process control to yield finished product that consistently meets the Quality Target Product Profile (QTPP) (2). At the other extreme, the raw material's critical quality attributes (CQAs) could be specified tightly enough to ensure consistent performance to the QTPP.

Of these two extremes, tighter raw-material control is unlikely to be viable in practice; thus, industry is compelled to consider adjustments during the manufacturing process to compensate for inherent raw-material variability. For example, such compensation is facilitated by moving to end-point control instead of using the traditional fixed-processing times.

One fundamental concept within QbD is the design space. ICH Q8(R2) defines design space as:

"the multidimensional combination and interaction of input variables (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality. Working within the design space is not considered as a change. Movement out of the design space is considered to be a change and would normally initiate a regulatory postapproval change process. Design space is proposed by the applicant and is subject to regulatory assessment and approval" (2).

Scientists in new-product development need to be aware of the expanded scope for adjustments afforded by QbD as they develop their design of experiments to establish their design space. Adjustments could include changes in the physical grade of the excipient (including any potential use of nonpharmaceutical-grade material in the early non-GMP phase of development), blending different grades, quantitative changes in excipient levels, fractionation of excipients (e.g., sieve cuts) as well as process adjustments. Understanding raw-material characteristics and how they influence product CQAs is necessary to derive the controlling algorithms for implementation of such adjustments within a quality system.

This understanding is essential for establishing a design space. Simply mapping the process and raw-material properties used, without understanding, merely defines the experience space, and no adds no benefit beyond the traditional pharmaceutical-development approach. To use the flexibility afforded by the 21st- century CGMP initiative and the ICH guideline, it is necessary to understand the overall context of excipient manufacture and variability, integrate such materials understanding and process engineering, and switch emphasis from consistent composition to consistent performance.

Excipient manufacture and quality control

Unlike the unit operations in pharmaceutical manufacturing, which are primarily batch-oriented, manufacture of excipients often employs continuous processing. An excipient batch or lot number may simply represent a defined period of time, rather than a discrete batch, with production quantities into the hundreds of tons. A Certificate of Analysis result provides a composite or average figure, and the supplier may have a much higher frequency of in-process testing on certain specification parameters. Access to such "raw" data will afford more meaningful correlations between supplier raw-material variability and the variability in CQAs of the user's finished products. Sharing such data will depend on the establishment of good supplier-user communications and trust. In the future, this information-sharing could be part of a supply or quality agreement, or a separate "QbD agreement." If necessary, relevant information could also be supplied under a specific, confidential disclosure agreement.

Requesting excipient samples at the extremes of specification for continuously produced materials is a problem. Driving a continuous process on the edge of specification means that half of the output will be out of specification. Understanding the raw-material variability in terms of the supplier's process capability greatly facilitates integration of materials understanding and process engineering (user and supplier). In a QbD universe, should we be looking to obtain materials at the extremes of specification when it is possible to use other means to establish a relevant design space? As discussed in the previous section, industry can use alternate grades, blends of grades, quantitative changes in excipient level, and fractionation to establish a design space.

Excipient composition

One aspect of excipients that is not well understood is excipient composition. Most excipients work because they are not "pure" materials. By accident or intention, they are, in effect, mixtures, and the other components often play a significant role in the functional performance of an excipient. Unfortunately, industry does not know enough about the composition of any excipient to state with certainty which minor components are necessary for functionality in a particular application. These minor components are not "impurities" as we might understand impurities in the context of an API. Rather, these minor components of an excipient may actually be necessary for optimum functionality.

For example, despite the hydrophobic nature of magnesium stearate, its functionality as a lubricant depends on the water of crystallization. Water situated between oriented fatty acid chains facilitates shear, and the anhydrate is less effective as a lubricant (5, 6).

Dibasic calcium phosphate dihydrate (DCP) provides another example. The coarse grade of DCP is used in the formulation of tablets by direct compression. During the compaction process, DCP deforms via brittle fracture. It is possible to manufacture very pure DCP with a very low level of ionic "impurities" using precipitated calcium sources and so-called "green" phosphoric acid. Unfortunately, DCP prepared in this way does not provide adequate performance in direct compression because of a lower fragmentation propensity. In this case, foreign ions are required for optimum performance. Their incorporation into the crystal lattice disrupts the lattice structure and causes defects that facilitate brittle fracture. The highly purified material does not have sufficient lattice defects to fracture adequately during compaction, and the resulting tablets are much weaker.

Building in flexibility

Standardization of functionality, as practiced by other industries, offers lessons for QbD. Rather than focusing on the extremes of specification for selected raw-material parameters, it may be possible to bracket the specific excipient with complementary grades or materials. For example, the particle-size distribution of mcc is the basis of several grades, and particle size may be of importance in many applications. If substituting some or all of one mcc grade in your formulation (e.g., 100 µm) with other size grades (e.g., 50µm or 200µm) causes no unanticipated problems (other than flow), then the design space will be wider than the particle-size limits for the 100 µm grade. This bracketing example is one approach to constructing a design space during development, but the approach also can be incorporated into the formulation. Incorporation of complementary grades or materials builds in the necessary flexibility for continuous improvement throughout the product life cycle.

Not every aspect of the product and process can be optimized before launch. More than 95% of development formulations are never likely to be commercialized because of clinical or safety failures during development. For this reason, most commercial-product formulations are not developed until Phase II clinical development is nearing completion. Given that commercial formulations are developed relatively late during clinical development, many will not have been fully optimized when they are launched. Limited availability of the API during development and other considerations may afford only a 3- to 4-sigma level of operational excellence at launch. To achieve 6-sigma operational excellence requires the ongoing analysis of manufactured product and raw-material statistics to identify and control subtle raw-material contributions. These contributions may not have been identified at launch, or may have arisen subsequently. Raw-material contributions to finished-product variability should be minimized using proper evaluation of the design space during development, but they will be of greater significance to subsequent operational-excellence programs post-launch because under QbD, the product is continuously improved.

Building formulation flexibility into the development process will pay great dividends by allowing continuous improvement, complementing process flexibility and raw-material specification options. Ultimately, continuous production may be possible. A smooth continuous improvement in quality and scale is likely to become even more important with the trend of pharmaceutical blockbusters becomingmore progressive or incremental in life-cycle growth. Formulation flexibility, in which compensatory quantitative adjustments in excipient levels have been included in the design will protect against both short-term raw-material variability and long-term changes due to material supply or increased operational-excellence requirements.

Brian Carlin, MRPharmS, MRSC CChem, is director of open innovation at FMC BioPolymer in Princeton, NJ, and chairman of the IPEC–Americas QbD Commitee, brian.carlin@fmc.com. R. Christian Moreton is vice-president of pharmaceutical sciences FinnBrit Consulting in Cambridge, MA, and a member of the Pharmaceutical Technology Editorial Advisory Board.

References

1. FDA, Pharmaceutical CGMPs for the 21st Century: A Risk-Based Approach (Rockville, MD, Aug. 2002).

2. ICH, ICH Q8(R2) Pharmaceutical Development, Step 4 Version (Aug. 2009).

3. G.E. Amidon et al., Pharm. Forum 33 (6), 1311–1323 (2007).

4. <1059> Excipient Performance," Pharm. Forum 35 (5), 1228–1250 (2009).

5. Y. Wada and T. Matsubara, Powder Tech. 78 (1994) p. 109.

6. K.D. Ertel and J.D. Carstensen, J. Pharm. Sci.77 (1988) p. 625.