Stuart R. Gallant, MD, PhD
Pharmaceutical shelf-life determination affects patient safety by assuring that the strength, quality, purity, and potency of an investigational or commercial drug are preserved until the moment that the drug is delivered to the patient. At the same time, storage condition affects the clinical plan by dictating how the drug will be shipped and handled prior to administration. Further, storage condition and shelf life factor into the marketing plan by dictating how the drug will be maintained and inventoried in the supply chain.
Today’s PharmaTopo post addresses: 1) planning and execution of pre-clinical and clinical stability studies to support shelf-life determination and 2) tracking and trending data to provide ongoing oversight of clinical supplies and ensure adequate inventory to support clinical commitments.
Overview and Terminology
Stability studies monitor the stability of drug substance and drug product. Drug substance is commonly stored in:
- Powders: Contact layer of low-density polyethylene (LDPE); outer bag of heat sealable polyethylene and aluminum foil.
- Liquids: Stainless steel or a multilayer bag with an ethyl vinyl acetate (EVA) contact layer.
Because drug substance bulks are significantly larger than the small samples used in stability testing, thought should be devoted to surface area affects to ensure that accurate modeling is achieved. For drug product, individual examples of the drug product units are held as stability samples. Important terminology includes:
- Stability indicated assays: Assays, often HPLC methods, which monitor the assay, purity, and impurity levels of the drug substance and drug product.
- Product related impurities: Impurities generated during the manufacturing process which do not change in level during storage.
- Degradants: Impurities which increase during storage—the levels of degradants are tracked during stability studies.
- Stress studies: Studies in which samples are exposed to higher levels of moisture, oxygen, pH, temperature, and light are important to documenting the stability indicating nature of stability methods.
The ICH stability pull points are:
The ICH storage conditions are:
| Drug Substance or Product Storage Condition | Studies (Condition, Minimum Duration) |
| Room Temperature | Long Term (25°C/60% RH, 12 months) Intermediate (30°C/65% RH, 6 months) Accelerated (40°C/75% RH, 6 months) |
| Refrigerated | Long Term (5°C/Ambient RH, 12 months) Accelerated (25°C/60% RH, 6 months) |
| Frozen | Long Term (-20°C/Ambient RH, 12 months) |
Significant changes for accelerated conditions include: 5% potency change, failure to meet specification for any degradant, failure to meet specification for appearance or physical properties, pH out of specification, or dissolution failing to meet specification.
Stability Data
Stability data is the end result of numerous processes, each which add some degree of variability to the data:
- Raw Materials: Formulation agents and packaging components each have their own lot to lot variability; there may be more substantial variability if multiple vendors with separate manufacturing processes are used for individual raw material components.
- Active Ingredient and Drug Product: The drug substance can vary in quality. Properties such as particle size distribution and polymorph can contribute to drug product stability. Drug product manufacturing can add significant variability—for instance, gradients in lyophilizer temperature may expose freeze dried drug products to different freezing profiles.
- Drug Product Conditions and Container Integrity: Stability chambers are designed to provide uniform storage conditions, but individual drug product packages may possess greater or lesser seal integrity. Desiccants lots may vary in efficacy.
- Testing: In some cases, the window for testing a given pull point can be quite wide, yet for example, the “6 mo” point will be labeled as such, even if it was pulled early or late. Sample preparation and hold prior to testing can vary, as can operator technique, not to mention variability of the testing methods themselves.
Over product lifecycle, the lot-to-lot variability of drug substance and drug product should decrease, but particularly early on, the variability can be large enough to prevent use of all the data together. In shelf-life determination, application of statistical testing to the individual lots allows determination if the batch data can be pooled. The steps are:
- Fit data with a linear (slope, intercept) models: The example data used below is from Chow and Shao [1]. Each batch of data is fit separately; Chow and Shao provide potency data for 24 batches, so there are 24 slopes and 24 intercepts that result from fitting.
- Apply statistical criteria to fitted slope and intercept values: The goal is to see if the data can be pooled. Typically, for clinical lots, the data will be analyzed individually, rather than being pooled because the number of lots is few and the sources of variability are larger, compared to commercial lots. Using analysis of covariance (ANCOVA), the slope and intercept values are compared; a p-value of 0.25 must be met or exceeded for both the slope and the intercept for the data to be pooled; otherwise, each batch is treated separately in determination of shelf life. The shortest shelf life becomes the shelf life of the product [2].
An example of this type of analysis using the data from Chow and Shao is provided below. 24 lots of data were fit for slope and intercept, and the statistical criteria were not met, so the lots had to be treated individually. The individual shelf lives obtained for the 24 batches had the following summary statistics:
A histogram of the 24 individual shelf lives reveals a diverse set of product stabilities for the individual lots:
Lot 21 had the lowest shelf life (12 months); the Lot 21 data (dots), the linear fit to the data (solid black line), the lower 95% confidence interval for the mean regression line (dashed line), and shelf-life marker line (solid red line) are shown in the graph below:
Tips
Here are a few tips that are worth considering as the project team plans stability studies:
- Later points are “cheap”: In the world of clinical development, funding is uncertain. Using 3-year-old drug substance to manufacture drug product may seem unlikely when a project starts, but if raising money becomes difficult later in a program, the team may want to use 5-year-old drug substance because the capital is not available to run another manufacturing campaign. Always hold enough stability retains to take a stability study out to at least 5 years, if not longer. Only terminate the study when the samples no longer meet the acceptance criteria of the study.
- Pilot studies reduce risk: As soon as development team believes that representative drug substance and drug product can be manufactured, initiate non-GMP stability studies of the drug substance and drug product. These studies will front run the GMP studies allowing the team to anticipate problems in the GMP studies before they appear.
- Backups are a great thing: A backup formulation helps you reduce risk from the primary formulation if the primary formulation has problems. Similarly, a backup storage condition (for example, refrigerated when room temperature is the primary storage condition) helps control risk. Keep taking data on the backup formulation and backup storage condition as long as practically possible. It is even possible to start a clinical study with a conservative storage condition (for example refrigerated) and relax it later in the study when stability data indicates the more permissive condition is acceptable.
- Expiry dating: Labeling drug product with expiry is required in the EU, Canada, Australia, and pretty much everywhere except for the US. The US accepts that the sponsor with track stability, so no expiry date is required for US clinical studies. You will need to remind your clinical repository which drug product packages are destined for which regions, so that the repository is clear on the labeling requirements as they label product for distribution.
Conclusions
Shelf-life determination, based on the principles outlined by the ICH, brings together methodical laboratory experiments and rigorous statistical analysis. Once a shelf life has been established, the project team can have confidence that the most advantageous presentation of the product (in terms of self-life and storage temperature) has been established. The team can go forward planning manufacturing campaigns to feed the distribution network with assurance of product integrity from manufacturing floor to the moment the patient receives a dose.
[1] Chow, S.C. and Shao J. “Estimating drug shelf-life with random batches,” Biometrics, 47, 1071-1079 (1991).
[2] ICH Harmonized Tripartite Guideline, Evaluation for Stability Data, Q1E (2003).
Disclaimer: This post is reprinted from PharmaTopoTM. PharmaTopoTM provides commentary on topics related to drugs. The content on this website does not constitute technical, medical, legal, or financial advice. Consult an appropriately skilled professional, such as an engineer, doctor, lawyer, or investment counselor, prior to undertaking any action related to the topics discussed on PharmaTopo.com.