Representative drug-substance batches.
A representative batch of DS, or DP, is one that is expected to have similar stability to the batches used in clinical production,
based on a science- and risk-based assessment of attributes expected to influence stability and knowledge obtained during
development. Often non-GMP or GLP (good laboratory practice) DS batches are manufactured first and placed on stability to
support a variety of product development activities. These batches often are representative of GMP batches from a stability
perspective and can be used to establish an initial retest period for the DS and support a clinical submission. Stability
data included in the CMC section are limited at this stage of development (5). The DS batch used for Phase 1 clinical supply
manufacture may be placed on stability concurrent to the clinical trials, if additional study is warranted.
In early development, it is common for the manufacturing process to be improved as a normal part of development. As the DS
process evolves, an evaluation is needed to determine whether the initial batch placed on stability is still representative
of the improved process. The authors advocate a science- and risk-based approach for deciding whether stability studies on
new process batches are warranted. The basic principles are discussed below and further details of such an approach can be
found in Ref. 10 and its citations (10).
The first step is to determine which DS attributes have an effect on stability. This step can be completed through paper-based
risk assessments, prior knowledge, or through a head-to-head short-term stability challenge (as discussed in the next section)
comparing the new batch with the earlier batch. If the revised process impacts one or more of these stability-related quality
attributes, the new batch should be placed on stability. If the revised process does not result in a change to a stability-related
quality attribute, the new batch does not need to be placed on stability. There are certain changes that almost always require
a new study. Different polymorphic form, counter-ion, or solvate forms of a compound cannot be assumed to have the same rates
of degradation or the same degradation pathways. Therefore, these changes will almost always require new stability data (11).
Typical changes encountered in early development include changes in synthetic pathway (including a change in the order of
the bond forming steps), batch scale, manufacturing equipment or site, reagents, source materials, solvents used, and crystallization
steps. In most cases, these changes will not result in changes in DS stability. A common misconception is that changes to
the impurity profile will adversely affect stability. Most DS impurities, particularly organic impurities, are essentially
inert and have no effect on DS stability. Some impurities, such as catalytic metals, acidic or basic inorganic impurities,
or significant amounts of residual water or solvents, may affect stability, and if the new process changes the level of these
impurities, additional stability may be warranted. Packaging changes of the bulk material to a less protective package may
require stability data to support the change. Likewise, changing to a more protective package in order to extend the retest
period, beyond what the original representative batch study could support, may require additional data in the new package.
Physical attributes that may impact stability include particle size/surface area, the degree of crystallinity/amorphicity,
polymorphic form, hydrate form, and moisture content. If a new process changes any of these attributes, a risk assessment
is recommended to determine whether additional stability is warranted. In some cases, short-term accelerated or stress stability
data can help decide if these attributes are in fact stability related quality attributes. The concepts detailed in the above
two paragraphs attempt to differentiate major and minor changes and may result in different stability strategies to support
A risk assessment may help to determine whether the current batch is sufficiently representative. Alternatively, additional
data may be needed when the batch is not representative or when there is substantial doubt after the assessment.
Drug-substance stability data collection.
Although the stability testing practices and the amount of collected data vary from company to company, three commonly used
approaches are discussed here. One is that an early, representative DS batch is placed under real-time and accelerated conditions
(e.g., 25 °C/60% RH and 40 °C/75% RH) and stability results for a few time points (e.g., 1–6 months) are generated to support
an initial retest period (e.g., 12 months or more). If needed, the stability study can be continued to extend the retest period
or the GMP material can simply be retested, as needed, as part of a retesting program commonly used for manufacturing components.
The selected analytical tests for stability studies should cover the quality attributes related to stability. Typically, these
tests would include impurities and description and may include assay. Chiral purity, polymorphic form (e.g., x-ray), and water
content should be tested, as appropriate. It may not be necessary to perform all tests at every time point depending on the
purpose of the test.
A second approach is to use high stress conditions with a short time such as a high temperature and high humidity (HT/HH)
model. For example, the DS is stored at 70 °C/75%RH in open and closed containers, while monitoring the chemical and physical
stability up to three weeks. The chemical stability results are used to extrapolate the initial DS retest period up to 18
months, more or less if justified. The extrapolation is based on the principle of the classic Arrhenius relationship between
temperature and the reaction (degradation) rate, assuming a pseudo-zero order reaction. It is a simple and fast prediction
of DS stability (12). This approach is most applicable for DS that is considered to have good physical and chemical stability
as demonstrated during drug discovery. Acceptance of this approach has been gained in a number of filings.
A third approach is the use of stress studies at several conditions coupled with modeling, such as the Accelerated Stability
Assessment Program (ASAP) (13). ASAP combines an accelerated aging protocol with a humidity-corrected Arrhenius equation to
provide an early prediction for DS and DP use period. It involves statistical analysis for more complex modeling of non-single
order kinetics of degradation in DS and DP, and it produces faster and reliable predictions of chemical stability. For DS
or DP with potential chemical stability challenges, this approach can be very useful to define appropriate storage condition
and use period within a relatively short time to enable progressing early development programs. More importantly, ASAP has
potential for a broader scope of stability assessment for developing control strategy of DP stability, such as long-term stability
prediction and packaging requirements in implementing QbD for commercialization. ASAP has been used by companies as part of
regulatory filings in multiple countries and has provided the basis for formulation, packaging, and moisture content justifications.
There is an ongoing dialog between industry and regulatory agencies about these newer ways for industry to meet its obligations
with respect to assuring the acceptability of products at the end of their shelf lives.
The knowledge gained from these studies can then be used to: support storage conditions and initial retest period assignments;
design future stability studies; assess the impact of temperature excursions; and select packaging. The results can also be
used to establish the initial use period for DS-based products (e.g., PIB or PIC). The accelerated or stress approaches may
save time in arriving at a first estimate of a material's stability.
The retest period derived from these types of accelerated or stress studies can be later verified by placing the first clinical
batch into real-time stability studies under ICH accelerated and long-term conditions. Future extensions of the retest/use
period can be based on real-time data.
Assignment of retest date.
Drug substance stability data, commensurate with the stage of development, can be generated to support use of the DS to manufacture
clinical supplies through a system using retest dates. When a longer retest period is needed, the extensions are most commonly
handled either through a stability program or through a batch-specific retest program. Wording may be provided in filings
to outline how the retest period will be extended in the future. If using stability data from the second (HT/HH) or the third
approach (ASAP), a retest period is extrapolated from the Arrhenius equation or moisture corrected Arrhenius equation (12).
Stability testing also is used to support use period dating of simple DS-based products.