The International Council on Harmonization, ICH M7 guideline, “Assessment and control of DNA reactive (mutagenic) impurities in pharmaceuticals to limit potential carcinogenic risk” classifies the mutagenic impurities into one of five classes, numbered 1–5.
The first step is to determine whether there are any relevant historical experimental data, including bacterial mutagenicity data as well as carcinogenicity data. After assessing any available experimental study results, it may be possible to assign the impurity to one of three classes: Class 1 (where it is a mutagenic carcinogen), Class 2 (where it is a DNA-reactive mutagen with no carcinogenicity data) Class 5 (where it is not mutagenic). In the absence of any relevant or adequate data, it is possible to perform a computational assessment to evaluate the mutagenic potential of each impurity. An overall assessment of the results, including a potential expert review, is then used to determine whether the impurity is predicted to be DNA-reactive mutagenic (Class 3) or predicted to be non-mutagenic (Class 5). In situations where there is a mutagenic prediction with a clearly identified structural basis or structural alert, and this alert is also shared with an experimentally determined non-mutagenic chemical, it may be assigned to Class 4.
In the absence of adequate experimental data, the ICH M7 guideline recommends the use of two complementary computational toxicology methodologies that predict the results of the bacterial mutagenicity test:
These two methodologies are often referred to as (Q) SAR approaches with “Q” (standing for quantitative) in parenthesis to include the non-quantitative expert rule-based methodology. As such, any models used should be transparent and interpretable to support such a review. Quantitative structure-activity relationship (QSAR) modelling is a computational method used to predict the biological activity of chemicals, in this case, mutagenicity based on their molecular structure. The models identify patterns and relationships between the structural features of the compounds and their mutagenic activity, allowing them to predict the mutagenicity of new or untested compounds based on their structural similarity to the training set.
Software that has been used for many years in the regulatory environment are
Computational toxicology is a safe, cost-effective (because they often avoid the need to perform a reverse bacterial mutation assay) and high-throughput approach to assess the mutagenic potential of impurities. If the model's predictions are accurate, they can be used to predict the mutagenicity of new compounds with a high degree of confidence. This can be expensive in cases when the impurity needs to be synthesized in sufficient quantities and purity to enable experimental testing.
Here’s your opportunity to join a unique, global company with an incredible, life-changing mission
APPLY NOW