Abstract
The safety aspects of elemental impurities in finished drug products are a topic of considerable importance in the pharmaceutical community, and guidelines such as ICH Q3D and USP <232> and <233> have been published to provide directions on how to assess finished drug products with respect to such impurities. Although a drug product's packaging system has been identified as a potential source of elemental impurities, comparable guidelines have not been established for assessing packaging systems and their materials and components of construction with respect to their potential to contribute leached elements to packaged drug products. In this commentary, the author considers the critical questions associated with selecting materials and components of construction and qualifying components and packaging with respect to their potential to add elemental impurities to packaged products and suggests means for accomplishing this objective.
LAY ABSTRACT: Elemental impurities in drug products can adversely affect the drug product's quality attributes. Regulatory guidelines that establish how to assess drug products for elemental impurities have been published. Although the drug product's packaging system has been identified as a potential source of elemental impurities, no guidelines have been published to specifically address packaging. In this commentary, the author considers the key issues associated with elemental impurities derived from packaging and suggests means for selecting and qualifying packaging on the basis of extractable elements.
Introduction
ICH Q3D (1) and the aligned USP <232> (2) monograph establish elemental impurities requirements for finished drug products, including a list of elements that, if present in drug products at levels above established permissible daily exposures (PDEs), could represent a patient safety issue. Furthermore, both documents clearly establish container closure systems as sources of elemental impurities, noting that “elemental impurities (that) have the potential to be leached out into the drug substance and drug product from the container closure system.” Moreover, ICH Q3D notes that:
The identification of potential elemental impurities that may be introduced from the container closure system should be based on a scientific understanding of the likely interactions between a particular drug product type and its packaging. When a review of the materials of construction demonstrates that the container closure system does not contain elemental impurities, no additional risk assessment needs to be performed … Studies to understand potential leachables from the container closure system … should be performed. The source of elemental impurities will typically be addressed during evaluation of the container closure system for the drug product.
The Mandate for Testing Container Closure Systems and Their Materials of Construction To Establish Their Contribution of Elemental Impurities To Packaged Drug Products
Although there may be questions related to the nature, extent, and form of the elemental testing required for packaging systems and their materials of construction, there can be no doubt that ICH Q3D sets the stage for packaging system and materials of construction testing. The above text is clear on this subject, and the intent is unmistakable. Moreover, ICH Q3D definitively promotes the concept of risk assessment and requires its use to address elemental impurities. Because the proper foundation of a valid and rigorous risk assessment includes data and information, risk assessment performed on a container closure system must be based on data secured for that system and its materials of construction.
However, even if ICH Q3D were equivocal on this subject, other factors necessitate the testing of packaging and materials of construction as potential sources of elemental impurities. For example, consider quality by design (QbD), which, simply stated, is the concept that quality is best served by building it into the product as opposed to establishing that it has been achieved by testing the product. Surely QbD supports the concept that selecting materials of construction based on their ability to contribute elemental impurities is an important step in ensuring that the drug product will not contain excessive levels of packaging-related elemental impurities. Moreover, it is logical that the selection of materials should be based on a consideration of relevant information and data concerning the materials' ability to contribute elemental impurities.
Lastly, elemental impurity data for container closure systems and their materials of construction may be the basis of an efficient and effective change control process for container closure systems. It would be most unfortunate and inefficient indeed if change control of container closure systems could only be exercised by elemental impurity profiling of packaged drug products.
Testing Container Closure Systems To Establish Their Ability To Contribute Elemental Impurities To Packaged Drug Products
Once the mandate for testing container closure systems to determine their potential to contribute elemental impurities to packaged drug products has been established, the practical question becomes how to accomplish this task. Keys aspects to consider include:
What articles should be tested and for what purpose?
How should the articles be tested?
What elements should be targeted in the testing?
How should the results of the testing be reported and interpreted?
Each of these questions will be addressed as follows.
What articles should be tested and for what purpose?
Figure 1 illustrates the high-level process for developing a packaged drug product from the perspective of its container closure system. In its simplified and generalized form, the flow diagram captures the essence, if not detail, of actual processes used by packaging or drug product development organizations.
The process starts with the selection of the container closure system's materials of construction. In fact, the actual first step is the generation of a requirements definition listing the requirements that the packaging system and its materials of construction must meet. For the sake of this discussion, assume that a requirements definition has been generated and contains a requirement that “the container closure system does not cause the packaged drug product to contain elemental impurities at levels that adversely affect the drug product's suitability for its intended use” (note that in making this statement, we have already moved beyond ICH Q3D, which addresses only one aspect of suitability for intended use [patient safety]).
As noted in Figure 1, actions performed at the materials of construction stage differ from actions performed at the other stages. While selection and qualification are performed for both components and container closure systems, only selection is specified for materials of construction. To understand this difference, one must first understand the difference between selection and qualification. Selection is a decision to use (or not use) a specific article in a particular application, presumably based on the article's ability to meet defined requirements. Selection is a not a guarantee that the specific material will produce an acceptable item. Rather, selection is a risk-based decision that the material is likely to produce an acceptable item. In contrast, qualification is used to confirm a material's ability to meet defined requirements (acceptance criteria), where such requirements ensure that the produced item will be acceptable. Thus, qualification can be considered as a “guarantee” that the produced item is acceptable.
While one could envision the desire to qualify materials of construction (and the value in this process), the propriety of doing so is questioned based on the simple observation that materials are selected so early in the product development project. This is important for two reasons. First, materials are typically highly processed by the time the final container closure system is constructed. Thus, it is questionable whether tests performed on unprocessed materials could be definitively linked to performance requirements for fully processed, packaged drug products. Moreover, the means by which materials of construction are processed to produce packaged drug products vary so widely, drug product to drug product, that it would be challenging to devise test methods for unprocessed materials that would be applicable in a majority of packaging applications.
The second reason addresses the question “Whose responsibility is it to perform materials of construction testing?” Note that this is a different question than “Whose responsibility is it to possess materials of construction data?” The answer to the latter question surely is, “users of the materials of construction.” Because “Who should generate the data?” is a different question than “Who should possess the data?” it should not be a surprise that the answer is different. To provide the answer, consider the simple mathematics of economics. For example, consider a vendor of a material for which there are 100 potential users. Is it not more economical for the vendor to do the testing once and provide the test data to all 100 users than for the users to independently perform the same testing 100 times?
The conclusion that vendor testing of materials is appropriate is another reason why materials are tested for selection and not for qualification. In the example of the one vendor and 100 users, if each potential use of the material is profoundly and fundamentally different and if the vendor accepts the responsibility for qualification, then the vendor would have to perform 100 individual tests to match each user's conditions of use. While it may be a more reasonable request that a vendor implement a well-defined, efficient, and effective test protocol once, it is surely a less reasonable demand that the vendor perform 100 customized assessments.
Most certainly, there are other dimensions of the “who should do the test” question that need to be considered. However, the point is not to make a definitive statement as to who should do the test but rather to note that because there is a question of who should do the test, testing at the materials stage is geared toward selection and not qualification.
Considering components, one can envision container closure development and/or manufacturing processes where the term component has little meaning, as materials of construction are directly converted to packaging systems. Even when components are a relevant consideration, selection may be irrelevant (as the optimal component may already have been specified) or qualification may not be appropriate (as organizations may decide to perform qualification solely at the system level). Nevertheless, component selection may be a critical part of product development and could rely on extracted elements' data as a selection criterion.
Considering the container closure system stage, the required action is most likely qualification, as it is a global regulatory requirement that a container closure system be qualified as suited for its intended use. Less frequently, container closure systems are tested for the purpose of selection, but one cannot completely discount this possibility.
This is a very long discussion indeed to establish that the answer to the question, “What articles should be tested?” is, “materials of construction, components, and container closure systems.” However, because selection and qualification serve two different purposes, selection and qualification are best accomplished by two different testing and reporting strategies. Furthermore, because materials are typically selected and systems are typically qualified, material testing and reporting and system testing and reporting will differ. Lastly, because components may (or may not) fall in the middle, components can either be selected or qualified (or both) based on methods that are suited for these individual purposes.
In summary:
Materials of construction are tested for the purpose of selection.
Components are tested for selection and/or qualification.
Systems are tested for qualification.
Testing for selection and testing for qualification are different by necessity and purpose.
While users are responsible for possessing test data, vendors may be the more economic choice for generating test data, specifically if the data is used to support selection.
How should the articles be tested?
Generally, testing articles to establish their ability to contribute elemental impurities to packaged drug products consists of two independent processes, preparing the article for testing and testing the prepared article. Each is considered separately as follows.
Preparing the Article for Testing: This question of how to prepare a container closure system to test its ability to contribute elemental impurities to packaged drug products is easily answered, as the answer has already been provided in regulatory guidelines and is well understood in the technical community. Specifically, the ability of the container closure system to contribute leachables to a packaged drug product is best addressed by profiling the packaged drug product for leachables over shelf life, as leachables information is used to support the qualification of the container closure system. While this is a straightforward answer for organic leachables, there is a catch when applied to elemental impurities. Specifically, when a packaged drug product is screened for elemental impurities, the results obtained reflect all contributions to elemental impurities, including not only the container closure system's contribution but also the combined contributions of the drug substance, excipients, water, and manufacturing system. While the total levels of elemental impurities in a packaged drug product are necessary and relevant information, it does not delineate the contribution of the container closure system.
As measuring elemental impurities in a packaged drug product does not establish the container closure's contribution to the elemental impurities, this information must be obtained by another means, such as a controlled extraction study of the container closure system performed using extraction conditions that reflect, mirror, or simulate the filled container's clinical conditions of use. Because the conditions for performing such a controlled extraction (simulation) study closely match the clinical conditions of use of the packaged drug product, it is anticipated that the resulting extractable elements profile will be similar to the leached elements profile, both in terms of which elements are measurable and their measured levels.
Because drug products differ so greatly in terms of their inherent “leaching power” for elemental impurities, the clinical conditions of use of drug products can vary widely, and the scope of the study can vary (e.g., performing a study that considers one packaged drug versus many packaged drugs), the extraction conditions used in these controlled extraction (simulation) studies vary from circumstance to circumstance and thus are customized as opposed to standardized.
The question of how to prepare a material of construction for testing is less clear, as there are two potential options, digestion and extraction. In digestion, the material is completely solubilized, typically by the action of strong acids and elevated temperatures. Thus, the prepared sample contains all potential elemental impurities that were present in the material at a level reflecting the total amount of the potential elemental impurities present in the material. As such, a digestion establishes what is present in the material, reflecting its greatest potential to contribute elemental impurities (i.e., the situation where “it all comes out”) and providing the absolute worst-case situation.
Alternatively, the material can be prepared for testing by extraction, as was the case for the container closure system. An extraction, as opposed to a digestion, establishes “what can come out,” which more directly address the issue at hand (i.e., substances in a drug product have to extract out of the test article before they can be leachables). It is for this reason that extractions are the appropriate means of preparing a material of construction for testing. This is a compelling reason, as it is well established that generally only small portions of an item's total pool of elements are leached into drug products under conditions of use (3).
Although both materials and systems are prepared for testing by extraction, the nature of the extraction is likely to be different, material versus system, as the purpose of the material testing is selection, as opposed to qualification. If the selection is made between multiple candidate materials, there is considerable value if the same conditions of extraction are employed for all materials, as this ensures that materials can be compared on an “apples-to-apples” basis. This is exactly the opposite case with qualification, which is a process that is customized to match each situation and which places a high premium on testing items with the conditions under which they are used.
If the goal for material testing is standardization, then it is relevant to consider the extent to which extraction methods can be standardized across material classes. In general, materials used in pharmaceutical packaging fall into the following four broad classes: plastics, elastomers, glasses, and metallics. Each of these materials is sufficiently different that it is unlikely that a single extraction solvent and process can be established that is equally relevant to and appropriate for each class. Thus, different extraction solvents and different extraction processes may be needed for these different classes of materials.
For any material, aggressive extraction using an appropriate extraction medium (e.g., an inorganic acid) may be justifiable as being a worst-case exaggeration of pharmaceutical contact conditions. Use of an aggressive extraction will likely lead to two possible conclusions:
If I judge the test article is suitable for use under aggressive experimental conditions, it is likely to be suitable for use under actual-use conditions.
If I judge a test article to be unsuitable for use under aggressive experimental conditions, then additional testing will be needed to establish whether the test article might still be suitable for use under actual-use conditions.
Similar discussion and conclusion are relevant for components. For the purpose of selection, components should be tested via an approach that is similar to that of materials. To the extent that it is practical, components should be tested for selection in a manner that is consistent with a way in which the component contacts the drug product during the product's storage in its packaging. Furthermore, components should be tested for selection in their fully processed form. When tested for qualification, the word “should” in the last two sentences must be replaced with the term “must.”
Testing of the Prepared Article (Extract): In theory, elemental impurities could encompass the entire periodic table, although practically speaking, many elements are unlikely to be encountered in pharmaceutical packaging systems (3). Thus, the method for testing the extracts must be comprehensive, addressing a large population of elements. As the adverse effect of some elements may be exerted at low levels, the test method should be highly sensitive. Alternatively, as the levels of extracted elements could vary substantially, the test method should have a wide dynamic range. Given the chemical complexity of test articles and extraction solvents, the test method should be highly selective and robust.
Considering the current state of the art in elemental testing, atomic spectroscopic methods, such as inductively coupled plasma atomic emission spectroscopy (ICPAES) and inductively coupled plasma mass spectrometry (ICP-MS), are recommended for elemental impurity testing. For example, see USP <233> (4).
It is difficult, if not impossible, to talk about analytical testing without addressing the aspect of validation. In the pharmaceutical world, it is clear that test methods that directly address and establish a test article's suitability for use (qualification) must be validated using appropriate procedures and specifications. USP <233> may be an appropriate reference to consider when designing validation studies and setting acceptance criteria. While it is also clear that test methods for selection must be suited for this intended use, it is less clear that they need to be rigorously validated, as selection at an earlier stage is always followed by qualification at a later stage. For this and other reasons, methods for selection are typically qualified for use via a process that is less rigorous than formal validation.
In either case, selection or qualification, methods must be established as effective at the time of use via appropriate system suitability assessment.
In summary:
Materials are extracted under aggressive, standardized conditions, which may be optimized for different material types, regardless of the material's possible use.
Components are extracted using conditions that simulate their intended use.
Systems are extracted using conditions that simulate their actual use.
Extracts are tested for targeted elements using atomic spectroscopic methods as indicated in USP <233>.
Test methods for selection must be qualified to establish their suitability for use.
Test methods for qualification must be more formally and rigorously validated to establish their suitability for use.
Test methods for either selection or qualification must include appropriate system suitability assessment at time of use.
What elements should be targeted in the testing?
On one hand, the answer to this question is simple enough; elements that are present in the test article and which present a potential hazard should be targeted. Alternatively, the individual aspects of “what elements are present” and “what elements are potential hazards” are sufficiently complex that this answer is difficult to execute.
From a practical perspective, the analytical methods referenced previously, ICPAES and ICP-MS, have the capability to provide data for many more elements than could be present in materials used in pharmaceutical packaging and/or which are pharmaceutical hazards. Thus, one possible answer is “target them all.” If it were as simple as that, then I might agree with that approach. However, for one thing, data tables that have a long list of targeted analytes and concentration entries such as “ND = not detected” are inefficient. More importantly, as the test methods must be either qualified or validated, each unnecessary element increases the possibility of a test failure, raising the specter of endless analytical runs being rejected because of unnecessary elements. A similar situation could also arise during system suitability testing.
Thus, testing should not include superfluous or nonessential elements.
Now, someone will point out that given advances in drug product development and material science, one never knows what element will be the next “bad actor.” Although I acknowledge the truth in that statement, I point out that the concept here is risk management and not risk avoidance. Should such “bad actors” surface, then of course the list can be adjusted to include them.
I would also expect that the reverse would be true; that is, once sufficient evidence has established that certain elements are not associated with certain materials, then those elements can be removed from the list for those materials.
Considering the potential impact of elemental impurities on patient safety, ICH Q3D provides a list of 24 “elements to be considered in the risk assessment if intentionally added (all routes).” As such, it is the foundation of our generic answer to the question, as generally speaking, it is difficult to always know (a) whether a particular element is intentionally added or not and (b) what dosage form a particular item will be used with. However, it is only a foundation, as it is understood that the ICH Q3D Table 5.1 list is not a comprehensive list of potential hazards. If we consider hazard in the strict context of safety, ICH Q3D itself acknowledges that elements that have “differences in regional regulations are not addressed in this guidance.” This is why aluminum, manganese, and zinc, for example, are not assigned a PDE in ICH Q3D, even though their presence in certain drug products poses a potential patient safety risk. If we consider hazard in the broader sense of being a potential threat to a product's suitability for use, ICH Q3D also notes that it does not address product quality issues, pointing to tungsten's documented ability to adversely affect therapeutic proteins as an example of a potential element of interest. Thus, in addition to providing Table 5.1, ICH Q3D notes that “some of the elements [to be] considered include aluminum, boron, calcium, iron, potassium, magnesium, manganese, sodium, tungsten, and zinc.”
Thus, the answer to, “What elements should be targeted in testing?” is, “The minimum number of targeted elements includes the 24 elements listed in ICH Q3D Table 5.1 as well as the 10 elements specified in the text as ‘considered elements.’”
In summary:
To address safety, the 24 elements in Table 5.1 of ICH Q3D are targeted.
To address safety and/or quality, 10 additional elements specified in ICH Q3D are targeted.
These targeted elements establish a baseline that can be augmented with additional elements as appropriate for a specific test article.
With experience, elements may be deleted from the list if it is established that they are not relevant to a particular test article.
How should the results of the testing be reported and interpreted?
The question can be restated as, “Should there be acceptance criteria for extracted elements?” as the answer to this question will dictate the answer to the title question. To support the answer of, “No, there should not be specifications,” consider the following. Materials and components can be significantly altered as they are used to construct packaging systems. Thus, it is questionable whether extractable elements results for materials or components can easily and rigorously be linked to elemental impurity limits in drug products. Therefore, elemental impurity specifications for finished drug products cannot be easily associated with extractable element specifications for materials. This is not only a question of whether the link can be made mathematically but whether the link can be justified scientifically. If specifications for materials cannot (or should not) be linked to elemental impurity specifications for drug products, then it is not clear what else would serve as the appropriate basis for material specifications. If there is no appropriate basis for material specifications, then there should be no material specifications.
Moreover, selection processes may vary across organizations and situations. Given the great diversity in organizations and situations, it is unlikely that useful, relevant, and generally applicable specifications can be established to aid material selection. It is the responsibility of a given organization to justify their material and component choices, and it is unreasonable to expect the publication of universally applicable specifications by an organization with standards-setting authority.
Thus, extracted elements' results for materials are not correlated with general specifications, unless specifications have been developed by the material's vendor and/or user.
On the other hand, complete container closure systems are tested under conditions that simulate their intended use, and it may be acceptable and relevant under such circumstances to interpret extractable element results in the context of ICH Q3D PDEs, provided that the clinical use of the packaged drug product is well understood and can be linked to the extraction conditions. However, it is noted that PDEs as acceptance criteria do not address product risks other than safety.
It is clear that for extractable elements to be interpreted in either manner, the results of extracted elements testing must be reported in terms of concentration units and not in general terms, such as “pass” or “fail.”
Once it has been established that extractable elements test results are reported as absolute concentrations, a reasonable follow-up question is, “At what concentration level should they be reported?” Somehow, the obvious answer that they should be reported at levels at which they are actionable and relevant seems imprecise. Considering selection, extractable elements information is actionable and relevant at levels consistent with the capabilities of appropriately state-of-the-art methods, expertly applied. Thus, for example, if it is the consensus of qualified analytical experts that elements extracted into a certain extraction solvent can be reliably and routinely measured above a level of X ng/mL, then X ng/mL becomes the reporting threshold, where values above the threshold are reported and values below the threshold are not.
Considering qualification, the action taken is the assessment of whether the extracted element will adversely affect the quality of the drug product into which it is leached. Thus, the extracted elements should be reported at levels well below the level at which quality is adversely affected. Considering patient safety as a critical product quality attribute, it is noted that ICH Q3D has established 30% of an element's PDE as the control threshold above which control measures should be implemented to assure that the level of the element in the drug product does not exceed the PDE. Because this is a control threshold for elements in the drug product, it is not directly applicable to extracted metals, because it does not consider other contributors of elements to the drug product. In this context, then, extractable elements are reported at values somewhat less than 30% of published PDEs (e.g., 5% of the PDE if such levels are analytically achievable), so that approach to such control thresholds can be anticipated.
In summary:
Results to inform selection should be reported as concentrations above a reporting threshold based on analytical capability. It is the prerogative of users to interpret the results in terms of their own selection process and selection criteria.
Results to support qualification should be reported as concentrations to enable interpretation with respect to safety by considering drug product PDEs and the clinical use of the packaged drug product. Results should be interpreted in the context of other product quality aspects by means consistent with the circumstances and the individual organization's risk management processes.
Other Considerations
The Issue of the Chicken and the Egg
When the process of drug product development is started from the beginning, then a staged testing strategy of material selection, component selection/qualification, and system qualification makes sense from a QbD perspective. However, if one already has a fully developed packaged drug product that is compliant with ICH Q3D, it makes little sense to go back and perform the testing required to qualify the container closure system, select or qualify the components, and select the materials, as obviously the system, components, and materials are suited for their intended use if the packaged drug product is suited for its intended use. Similarly, if one has already qualified a container closure system, it makes little sense to go back and perform the testing required to select or qualify the components and select the materials. Thus, it is always the case that qualification at a later stage of development satisfies the need, if such a need exists, to test, for qualification or selection, at an earlier stage of development.
To Align or Not To Align, That Is the Question
The reality of the global marketplace is that globally marketed drug products must be tested via local procedures and meet local requirements. It is unfortunate that globalization of tests and specifications has not kept pace with the globalization of the marketplace, and the frustrating and inefficient reality is that the same material, component, and system tested and used in one market region may need to be tested by varying and differing means to be used in a different market region.
When regional differences are driven by science or overwhelming practical circumstances, then they may be justifiable, if inconvenient. When regional differences arise because of concepts such as “not invented here” or “that is the way we have always done it (even if we cannot remember why),” then they are unjustifiable and unacceptable. Surely, it is desirable to drive toward the global alignment of extractable elements testing of materials, components, and system, even if the mechanism for doing so is unclear.
What Is That in Your Other Hand?
The role of the vendor of a material or component in this process has been previously considered, specifically addressing who could perform testing. However, vendors can supply relevant information without actually performing testing. Thus, for example, it is typical for vendors to supply potential users of their articles with documents that address elemental impurities by specifying what elements the vendor did not intentionally add to the article. For example, if a vendor certified that “none of the ICH Q3D elements are intentionally added,” this could be valuable information for the potential users.
However, one must understand the limitations of such certifications, as they do not specify:
What elements the vendor has intentionally added.
What elements might have been added by others involved further downstream in the manufacturing process.
What elements might have been unintentionally added.
Given these limitations, it must be recognized that while “not intentionally added” certificates have value in terms of assessing an item's risk of contributing elemental impurities, at best, their contribution provides only an incomplete assessment of risk.
In other circumstances, vendors may supply extractable elements data from controlled extraction studies they have performed or commissioned. As was the case with certificates, this information is valuable and can be the basis for thoughtful material or component selection. However, vendor-to-vendor comparisons can be difficult if the conditions of extraction are not consistent across vendors. Of course, consistency among vendors would be facilitated by a standardized test procedure.
In summary:
While sequentially testing materials, components, and systems is logical and appropriate during the packaging system or drug product development process, retroactively testing materials or components in a system that has been tested and found to be satisfactory provides little additional value and should not be required.
When scientifically feasible and practically appropriate, test methods for extracted elements should be aligned across geographic and market regions.
While vendor information related to extractable elements is always useful, particularly in regard to selection, the decision to use vendor data should be based on a careful consideration of the conditions under which, and the means by which, the data was generated.
Conflict of Interest Declaration
The author declares that he has no competing or conflicting interests.
- © PDA, Inc. 2019
PDA members receive access to all articles published in the current year and previous volume year. Institutional subscribers received access to all content. Log in below to receive access to this article if you are either of these.
If you are neither or you are a PDA member trying to access an article outside of your membership license, then you must purchase access to this article (below). If you do not have a username or password for JPST, you will be required to create an account prior to purchasing.
Full issue PDFs are for PDA members only.
Note to pda.org users
The PDA and PDA bookstore websites (www.pda.org and www.pda.org/bookstore) are separate websites from the PDA JPST website. When you first join PDA, your initial UserID and Password are sent to HighWirePress to create your PDA JPST account. Subsequent UserrID and Password changes required at the PDA websites will not pass on to PDA JPST and vice versa. If you forget your PDA JPST UserID and/or Password, you can request help to retrieve UserID and reset Password below.