OR WAIT 15 SECS
Cynthia A. Challener is a contributing editor to Pharmaceutical Technology.
Limited guidance and numerous challenges create confusion about the scope and timing of stability testing for drugs in development.
The stability of clinical trial materials, regardless of the trial phase, must be understood to ensure patient safety. Stability can be affected by the nature of the API, the production process, the choice of excipients, the container and container closure system, and other factors. Stability testing is therefore essential for demonstrating that the formulations administered to patients in any clinical trial will remain unchanged throughout the duration of the trial. The type and length of stability tests typically depend on the phase of development and the nature of the clinical trial (e.g., use of placebo or comparator drug).
Various guidance documents from FDA and the European Medicines Agency (EMA) have been published regarding stability testing of clinical trials materials (CTMs). The language regarding duration of testing in these documents tends to be vague, however. For Phase I trials, FDA recommends monitoring of the stability and quality of the drug during the clinical trial (1). Similarly, EMA recommends an ongoing stability program be performed with accelerated and long-term storage studies initiated prior to the study (2).
For Phase II and III trials, FDA expects submission of a description of the stability performance and also suggests the development of stability-indicating analytical procedures that will detect significant changes in the quality of the drug product (3). The agency also encourages the completion of stress studies at Phase II, while in Phase III studies these stability studies should be extended to provide marketing application stability data.
While the International Council for Harmonization (ICH) guidelines on stability testing (4) specifically indicate that formal stability protocols do not apply to CTMs and such protocols are not required for clinical stability studies or submission as part of clinical authorization applications, it is highly recommended that companies do conduct stability studies by means of established procedures.
For CTMs, at a minimum, real-time data must be collected for a sufficient period to demonstrate shelf life for the product in use, as well as cover the duration for any intended clinical trial, according to Alyn McNaughton, technical director at Lonza Pharma & Biotech. The essential purpose of these stability studies, adds Geoff Carr, director of analytical development for Patheon Pharma Services by Thermo Fisher Scientific, is to ensure that CTMs will remain satisfactory over the period that they are intended to be administered to subjects enrolled in the clinical study, which is often dependent on the clinical trial phase.
The duration of a clinical phase will vary depending upon the number of subjects required to complete the study, the therapeutic indication and the data read out period, according to Teresa Iley, director of pharmaceutical development and manufacture for Intertek Pharmaceutical Services. It also tends to get longer as the project progresses, which impacts the required length of the stability study, adds McNaughton.
“As stability testing is commonly on the critical path for any product development, due to the need for real-time data, most stability studies are extended beyond the required time period to gain more knowledge about the product’s robustness and potential maximum shelf life. This approach allows an understanding of product supply needs for future clinical evaluation; for example, if additional manufacturing may be required during a clinical study to supplement product supply for batches that will expire during the clinical study, or if the product’s shelf life is not long enough for future trials or even for a viable commercial product,” he says.
Iley notes that shelf-life justification can be supplemented by data generated from technical batches manufactured in support of clinical trial applications, but it is typically recommended that the actual batches used to supply the trial are also assessed concurrently with the trial.
Initially a CTM is put on stability at both the long-term or storage temperature and under accelerated conditions. If the material is stable under both, the shelf life can be extrapolated up to twice the period covered by the long-term data (5), according to Judy Carmody, founder and principal consultant of Carmody Quality Solutions. “For Phase I trials, for instance, there may be one pull point at three months, which would allow an extrapolated shelf life of six months,” she explains.
CTMs for Phase II or III studies are likely to be much closer to the product intended for commercialization, adds Carr. “These trials are usually larger, longer duration, multicenter, and often in multiple countries, so longer-term stability studies (e.g., two years or more) are likely to be more appropriate,” he says. In addition, Carr notes that because the formulations are likely to be closer to the intended commercial products, data from these studies may be used as supportive in new drug applications.
The use of accelerated data to justify shelf life longer than real-time data can be a source of confusion, according to McNaughton. “Accelerated data collection provides a means of predicting the shelf life beyond the actual age of the product. However, interpreting any changes in the product impurity profile, relative to the toxicology-based safety information, can prove challenging in situations where the product does demonstrate some instability. It is also critical to ensure that real-time data, when it does catch up to the age of the product at the end of the study, remain in specification,” he explains.
The vague guidance regarding stability testing for CTMs can have a number of impacts, including on study designs. For instance, because many companies base their studies on ICH guidelines, especially Q1A(R2) (6), and then make modifications to make protocols more suitable for CTMs, there is often considerable variation in how different companies conduct their studies, according to Carr.
In addition, because it is important to ensure that batches in clinical studies always have the data available to demonstrate they remain in specification during the duration of the trial, it is necessary to continuously update the shelf life as soon as data are available and before the previous shelf life expires, McNaughton stresses. That can be challenging during early-phase studies, but simpler for later-phase trials because more time is available for completing stability studies.
On the other hand, if changes in the chemical or physical stability of a drug product are identified in stability studies, clinical programs may need to be suspended until new batches can be supplied, according to Iley. She notes that data provided from technical batches can be a good indicator of how clinical batches will perform and could save precious time when planning for contingencies.
The need to re-supply batches may also arise due to factors such as subject recruitment delays or dropouts, changes to dosing regimens, the introduction of new active, excipient, or packaging materials or revision to manufacturing processes. “Using a risk-based approach, it may be possible to undertake bridging studies to justify these changes, but it may also require performance of new or extended analytical studies. Any of these factors can impact the trial design and supporting stability program,” Iley says.
The design of bridging studies will depend on the changes that created the need to perform them, according to Carmody. They typically involve the use of the established analytical methods, with comparison of results to previous data to confirm agreement.
Any issues identified related to the chemical or physical stability of the drug product could cause a delay to the entire clinical development program. “A well-designed stability study, initiated early in the program, will facilitate detection of these issues and lessen any impact they may otherwise cause if not understood and resolved quickly,” asserts Iley.
The key challenge is the fact that as candidates progress through the development cycle, typically changes are made to the formulation, manufacturing process, analytical methodology, container closure system, or other aspects. Any such change requires generation of new stability data, according to Carmody. “The new material must also be demonstrated to be stable for the duration of the clinical trial. Any additional time for manufacturing the new CTM further adds to the potential for delays due to the need for additional stability studies,” she says.
A development program would, however, take decades if each of these studies were run sequentially to completion, according to McNaughton. “It is necessary to take risks based on limited data from partial studies, which must be done while ensuring that product used in trials remains covered by a stability study that demonstrates it remains within specification,” he says.
This necessary approach results in a balancing act of timing and risk with respect to determination of the best time to conduct these studies, and sometimes, when shelf life is limited or not yet known, ensuring a contingency to remanufacture to keep a clinical trial supplied is required. “Careful extrapolation of data obtained from samples stored at accelerated conditions can help predict when additional batches may be required to support clinical studies and therefore allow planning with some level of confidence when manufacturing slots will be required,” Iley adds.
There are also often increasing types of stability tests that need to be conducted as candidates move through later clinical trials. For instance, Carmody notes that diluents added to a powder to create a solution for infusion must be subjected to stability studies if the diluent manufacturer cannot provide relevant data. The in-use stability of the solution formed with the diluent must also be demonstrated. This information will impact the product handling protocol for the trial. “In-use studies should therefore be executed during early stages because the results directly impact and inform formulation and process development activities,” Carmody says.
API availability is an important factor as well because synthetic route and process development typically proceed in parallel with clinical programs, according to Carr. “Supplies are often limited around Phase I as synthetic route and scale-up activities have not yet progressed. Limiting stability study duration is important for avoiding wastage of valuable API,” he observes. API availability generally increases as candidates move to Phases II and III, so longer-term stability studies can be more readily supported. In addition, Carr comments that long-term data are more useful at this point because the nature of the API and dosage form is likely to be more representative of future commercial batches.
Complications can also arise in early-phase studies if the product is not tested promptly. In some cases, artificial failures can occur due to product being older than it should when tested, according to McNaughton. “As a result, an incorrect short shelf life can be assigned or the product can even fail specifications during the clinical trial, when in reality it is in specification at the shelf life when it should have been tested. Such situations add expense and uncertainty in a project, and in some cases the risk of withdrawal from the trial and program termination,” McNaughton says.
Often during early-phase studies, knowledge of both products and methods are limited, according to McNaughton. “When unknown degradation occurs, or a method does not behave as anticipated due to an unforeseen change, this can lead to pressure on the project. There exists a need to explain the change and justify it as safe for ongoing work or understand quickly that it is a change that does have an impact on product safety,” he notes.
Establishing methods at the start of the process and how the method lifecycle progresses (e.g., identification of new impurities, formulation changes) can require careful assessment and may identify gaps in planned studies or potentially initiate new studies, Iley says.
What is important, adds Carr, is the strategy for method development/validation. Because CTMs are considered GMP at Phase I, at least some validation of analytical procedures is required. Many companies adopt an approach of “phase-appropriate validation” or “qualification” using methods that have only been partially validated. “The objective is to be economic with the resources applied to validation without compromising patient safety,” he states.
Quality-by-design approaches to analytical method development/validation conflict with this strategy, however. Even so, Carr remarks that analytical development can only proceed to a certain extent until the full details of product strength and formulation are known. He suggests that, given the likelihood of limited knowledge regarding potential degradation products and how to set limits for them, the best approach is to use “alert limits” rather than pass/fail limits or “report results” with no defined limits. “Alert limits may be set applying ICH Q3B(R2) principles (5); an advantage of this approach is that if an alert limit is exceeded, an investigation can be conducted but confirmation of the result does not necessarily require a batch to be rejected, as would be the case with an out-of-specification result,” Carr explains. Alert limits are also useful for pharmaceutical performance testing, such as dissolution for solid oral-dosage forms and viscosity for semi solids.
The issue of validation is perhaps one of the most confusing aspects of conducting stability studies, according to Carmody. It is a Catch-22; stability studies and stability-indicating studies (forced degradation of the CTM) should be validated, but validation is typically not completed until Phase III.
“What should be done,” Carmody says, “is to understand the target product profile, mechanism of action, and other important characteristics (e.g., solubility, chromophoric, etc.) to identify the most appropriate analytical methods to choose from.” The best methods can then be chosen for identification, potency, purity, and impurity analyses and the relevant parameters validated according to regulatory authority expectations. For any deviations, good scientific rational must be documented.
“Even for Phase I CTMs, analytical methods can be developed with the intended use and validation parameters in mind. As important, stability-indicating studies and appropriate methods must be developed to ensure that the possible degradation products for the CTM can be detected,” Carmody states.
Carmody also notes it is important that stability programs be run by qualified personnel with the relevant knowledge. “There are different expectations for stability testing compared to traditional quality control testing. If analytical groups are going to be responsible for stability testing, then the people involved should be trained on the regulations and requirements,” she asserts.
While for the most part the focus of stability testing is on the API and formulated drug product, stability testing of placebos and blinded clinical comparators is also important. Stability testing on placebos is not particularly challenging because, in general, it essentially involves appearance testing and tests for other properties that could impact a blinded study.
The same is not true for blinded comparators, according to Carr. These products typically come from other companies. They often have been repackaged in new containers and may even have been manipulated in some way to ensure appropriate blinding for the study. As a result, there is little information available. “Stability testing is likely to be based on comparisons between the unchanged product in the original packaging versus the repackaged blinded product, and it can be very challenging to establish and validate suitable analytical procedures for this type of study,” Carr observes.
A carefully designed study that stratifies the storage conditions, specific tests, and material batches is required to avoid the often-nebulous growth of a stability program while still generating pivotal development data, according to Iley. “Ensuring strong, validated methods are in place and planning for potential trial extensions in the stability program design with the inclusion of optional time points and sufficient spare samples to support them is essential,” she adds.
The more knowledge of the product and methods that is generated in advance, the more a study can be de-risked, agrees McNaughton. “It is important to ensure sufficient windows for testing are left, before the data are critically required, to allow for investigational work to take place at every time point in the early stages of a product. For this investigational work, there also needs to be sufficient product added to the study to facilitate investigations to examine what may or may not be a critical change,” he adds.
It is also crucial to understand what climate zones the product may be exposed to. “There is little point in generating data for a product in the relatively temperate zone if the product also needs to be used in a tropical zone, and vice versa,” says McNaughton. Determining the behavior of the product components is also key, because if they do not behave according to Arrhenius predictions under accelerated conditions, then a false failure could occur that does not reflect the true performance of the product under its real-time storage conditions.
Using batches of drug product, including any secondary packaging, that have been manufactured to established procedures is also important, according to Iley. “Although requirements may change or vary during the course of product development, the closer to the final processes and procedures, the less influence they may have on the data obtained from the stability program,” she explains. She also says it is best to expect the unexpected and plan for mitigation.
The most important considerations are, according to Carr, to do what is needed to establish the shelf life for CTMs to ensure patient safety and reliable clinical data. It is then important to determine whether it is sensible to extend the study to collect additional stability data that could support a future new drug application filing.
“Full stability programs must be implemented to address all of the different aspects of clinical trials, from different doses to diluents, placebos, and comparator drugs. An understanding of other potential impacts on stability must also be considered, from critical raw materials to reference standards. All may need to go on stability,” Carmody asserts.
Equally important, she says, is consideration of appropriate time points for testing and the trending and statistical analysis of data throughout the testing period. “Shelf life is determined when the stability data crosses specified criteria. That needs to be predicted and not discovered after the fact. It is essential to understand the behavior of the CTM and intervene and make necessary changes if the stability data [are] not trending in the right direction,” she concludes.
1. FDA, Guidance for Industry: cGMP for Phase I Investigational Drugs (Rockville, MD, July 2008).
2. EMA, Guideline on the Requirements for the Chemical and Pharmaceutical Quality Documentation Concerning Investigational Medicinal Products in Clinical Trials (London, Sept. 2017)
3. FDA, Guidance for Industry: INDs for Phase II and Phase III Chemistry, Manufacturing and Controls Information (Rockville, MD, May 2003).
4. ICH, Q1A(R2) Stability Testing of New Drug Substances and Products, Step 5 version (ICH, 2003).
5. ICH, Q1E Evaluation for Stability Data, Step 4 version (ICH, 2003).
6. ICH, Q3B(R2) Impurities in New Drug Products, Step 4 version (ICH, 2006).
Vol. 43, No. 10
When referring to this article, please cite it as C. Challener, “Stability Testing for Small-Molecule Clinical Trial Materials," Pharmaceutical Technology 43 (10) 2019.