An accelerating trend in biotechnology is the use of plastic components in systems used to produce an active pharmaceutical ingredient (API) or a finished drug product (FDP) including:
- Tubing, connectors and devices (such as syringes) used for solution transfer,
- Filters used with starting solutions, process intermediates and/or the final API (or FDP),
- Containers (and associated connecting technology), used for solution storage and transfer,
- Tangential-flow filters used for concentration and diafiltration of process intermediates,
- Chromatography columns used for the concentration/purification of intermediates,
- O-rings used to seal connectors and fittings,
- Final bulk containers,
- Sensors and process-facilitating devices.
If any process stream solution used to generate an API or FDP (such as media, buffers, eluents, diluents) is contacted by a plastic component during API or FDP production, the component and the solution may interact. If the component or solution is modified by the interaction, then the component or solution, and ultimately the API or FDP, could be rendered unsuited for its intended use.
While all potential interactions between production equipment and process stream solutions are important to the suitability of an API or FDP, this commentary focuses on the situation wherein a substance from a production system component (extractable) moves into a process solution, persists in the process stream through the entire production process, and accumulates in the API or FDP as a production-related impurity (PrI), also referred to as a process-related leachable. Clearly, the impurity’s direct and indirect effect on the safety and efficacy of the API or FDP must be established.
It stands to reason that the most direct means of establishing the impact of the production system on the composition of an API or FDP is to screen the API or FDP for PrIs. Practically, however, such a direct approach may be difficult, if not impossible, to accomplish. Furthermore, such an approach is reactive in that potential issues associated with PrIs surface only after the API or FDP has been produced.
A more practical and proactive approach to managing the risk that APIs or FDPs contain unacceptable levels of PrIs is to establish the production system’s extractables profile. Establishing a production system’s extractables profile requires that the system be operated using simulating solvents and representative (and typically accelerated) conditions and that the system’s effluent, more or less reflecting the API or FDP, be screened for extractables.
For simple production systems, a system extractables profile may be readily and directly obtained by system testing. The more complex the production process is, the more difficult it is to directly test the entire system to obtain its extractables profile. In such cases, one could infer the system’s extractables profile from the individual extractable’s profiles of the system’s components. Such an inferred system profile considers (a) the individual component’s propensity to be extracted by the process stream solution(s) it contacts, (b) how the extractables profile evolves as the process stream moves from component to component through the production process, and (c) the ability of extracted substances to persist in the process stream and become entrained in the API or FDP. That is, extractables from production components become PrIs in the API or FDP by being serially extracted from the system’s components by the process solutions and then persisting in the process stream throughout the entire production process. Establishing the component’s propensity to be extracted by process stream solutions involves a laboratory investigation consisting of two steps: (1) extracting the component under relevant experimental conditions and (2) analytically characterizing the extract for extractables. Such an investigation generates the component’s extractables profile, consisting of two vital pieces of information, a list of the extractables and their concentration in the extract.
Organizations that produce and market APIs or FDPs (and who therefore are users of production systems) derive value from component extractables profiles in two ways. On one hand, an extractables profile can provide a basis for selecting production system components. Alternatively, an extractables profile suggests what the manufactured API’s or FDP’s PrI profile might be. Such information would be relevant, for example, in registration dossiers that secure regulatory approval to market the API or FDP.
As one might imagine, there is great enthusiasm on the part of users to have access to comprehensive and relevant extractables profiles that are generated once (and thereafter are universally available) via a standardized methodology which is applicable to every possible process situation encountered by every individual user. To an appropriate extent this enthusiasm is shared by other stakeholders, including component suppliers and regulators. However, this enthusiasm should be tempered somewhat as the diversity of process situations is so great that a standardized methodology applicable to all process situations would contain an inordinate and unsustainable number of method variations, specifically considering the conditions used to generate extracts. Furthermore, requiring that the testing be performed once and thereafter the results be universally available is not without its practical implications, including establishing that entity who generates, secures and maintains the extractables profiles.
The considerable challenge in developing such a universal standard is mitigated somewhat by “knocking the challenge down to size” and “managing it in smaller pieces”. In the following discussion I suggest four means of transforming the objective of converting standardized testing from an impossible dream to a practical and effective reality, noting that:
- Standards for component selection are very different from standards for component qualification,
- A risk-based approach is relevant and appropriate,
- The needs of the many outweigh the needs of the few, and
- Standards do not replace thinking and efficient standards require thinking.
Concept 1
API or FDP producers must establish the quality of their API and FDP to the satisfaction of regulators to secure approval to market the API or FDP within the regulator’s jurisdiction. So doing requires that the production process be qualified as being suitable for use. As noted previously, this necessitates that either the API or FTD be profiled for PrIs or that the production system be profiled for extractables which presumably could become PrIs. As an alternate to full system profiling, individual system components can be profiled and the system profile inferred as an aggregate of the component profiles. In this way component extractables profiles could be used to address PrIs during product registration. Although such a practice is applicable in certain circumstances, its applicability in specific situations must be established as an individual component’s extractables profile establishes how the component behaves as an isolated entity and may not address how an individual component will behave in a system or how all components in a system behave collectively.
While component extractables profiles may be useful in system qualification, it is questionable whether component extractables profiles are required in component selection. While extractables profiles facilitate the selection of components, component selection can be achieved if the component has been well-characterized because its extractables behavior can be inferred from the characterization data. Such a practice has been established for plastic materials used in packaging systems and is equally applicable to plastic components used in production systems. Consider, for example, the case of a plastic whose characterization reveals that it contains Irganox 1010. From this, one can infer the material’s probable extractables, as the extractables associated with Irganox 1010 are well-known and well-documented. As characterization testing may be less extensive than extractables profiling, one can envision different standards for component selection (based on characterization) and system qualification (based on extractables profiling).
Concept 2
It is well established in regulatory practice that the risk associated with a certain circumstance establishes the level of testing necessary to verify that the circumstance is acceptable. For example, Table 1 in the FDA Container Closure Guidance establishes the relative risk that packaging-related impurities (leachables) would accumulate in packaged drug product at levels that would adversely affect the drug product’s quality attributes. By considering the dosage form’s route of administration and the likelihood for interaction, the relative risk associated with a dosage form is established and the required testing to verify acceptability is specified.
It is reasonable to anticipate that a production process can be examined and that the relative risk that (a) a component would extracted by process stream solutions, (b) that the extractables would persist in the final process output, and (c) that the resultant PrIs would adversely affect output quality could be established. As was the case with packaged dosage forms, such a relative risk assessment could drive the level of required testing, with lesser testing being required in low risk situations and greater testing, building on the foundation of the lesser testing, being required in the higher risk situations.
Concept 3
As noted previously, all the circumstances experienced in all the possible API or FDP production systems represent a large universe of situations. A standard that seeks to cover all circumstances will necessarily be a complex standard, requiring an inordinate and unsustainable number of variations. Considering an extractables standard, a universally applicable standard would require numerous extraction processes.
Some variation of the “80-20 rule” explains the issue with this circumstance. If a set of extraction processes were developed to cover the entire universe of production circumstances, it is undoubtedly the case that 20% of the extraction processes would cover 80% of the production circumstances and that the remaining 80% of the extraction processes would be required to cover the remaining 20% of production circumstances. In such a circumstance, one easily understands the consternation of a concerned party who needs to test a component by the universal protocol when in fact the component is used in a way which fits in the 80% of circumstances that require only 20% of the testing.
In such a circumstance it might be advisable to develop a standard targeted at the “easy” 80% and let the remaining “difficult” 20%” augment or adjust that standard as required.
Concept 4
Concept 4 is, to a certain extent, a variation of Concept 3. In some ways, a standard that “does all the thinking” is highly desirous. Such a standard would contain those tests and specifications required to support the approach that “if a component is tested with all the tests and meets all the specifications contained within a standard then it has been established to be suited for its intended use”. Consistent with the observations made in Concept 3, such a standard, applied to production system components to address API or FDP quality, would have a large number of tests and exacting specifications, some of which are applicable to most situations and many of which would be applicable to only certain situations. In such a circumstance, one can easily understand the consternation of a concerned party whose component must be tested by methods in excess of the component’s intended use and must meet specifications that far exceed reasonable quality expectations.
Several organizations have taken up the challenge of developing a standard for the extractables profiling of plastic components used in API and FDP production. I wish those organizations god-speed as the mission is challenging but the rewards are great. In light of the previous discussion, I offer these organizations the following:
- Match the standard to its purpose,
- Leverage a robust risk-based approach,
- Seek out and leverage related successes,
- Aim for the middle, and
- Find a way to collaborate, as the industry will only benefit from a unified result. Multiple standards drive chaos, not stability, and being first does not necessarily mean being right.
Dr. Dennis Jenke is a Baxter Distinguished Scientist at Baxter Healthcare Corporation, where he works to establish the suitability for use of packaging systems, manufacturing systems, and administration devices for pharmaceutical products (for example, leachables/extractables and drug binding). He has published extensively in analytical chemistry, environmental science, and material/solution compatibility and serves as an expert reviewer for numerous pharmaceutical and analytical journals. He is the author of Compatibility of Pharmaceutical Solutions and Contact Materials; Safety Considerations Associated with Extractables and Leachables and is a contributing author to the Leachables and Extractables Handbook. Dr. Jenke is a member of several professional organizations that establish best demonstrated practices in the area of material/solution compatibility and is a frequently invited speaker on that subject.