An assessment is required regarding the appropriate nitrosamine control strategy for products falling within the scope of ICH S9. According to the relevant guideline, nitrosamines should be evaluated in accordance with ICH Q3A/B. Under this approach, the applicable acceptance limits become substantially higher than the nitrosamine-specific Acceptable Intake (AI) limits.
In this context, is it still necessary to develop and validate analytical methods and perform sample testing for potential nitrosamines? Based on our current experience, nitrosamines are not expected to be detected at such elevated limits, even for those with the highest theoretical potential for formation.
As an alternative approach, a scientific risk assessment could be established based on the nitrite sources present in the product and the maximum theoretical amount of nitrosamine that could be formed from these nitrite sources. In this way, it could be scientifically demonstrated that the nitrosamine levels would not exceed the applicable ICH Q3A/B limits under any foreseeable circumstances.
Would such a scientifically justified risk assessment be considered an acceptable justification by regulatory authorities in lieu of analytical method development, validation, and routine testing?
1 Like
A risk assessment is always required. Regarding the analytical method, nitrosamine analyses will be included in the ICH Q3A/B impurities methods. It will still require to demonstrate specificity, and thus that the method is stability indicating including with respect the nitrosamine. There were some previous discussions on this site, you may want to check out.
I agree with the fact that risk assessment is needed, but analytical work may not be required. Normally the ICH Q3A / Q3B are too high for nitrosamines. You have two possiblities, if the precursor amine is an impurity, normally will already be controlled at ICH Q3A / B thresholds, there will be no way of the potential nitrosamine to exceed that limit (even more considering that nitrite will be the limiting reagent). If the API is the precursor amine, normally nitrites will again limit the capability of transformation into the nitrosamine.
I personally would always go first for worst-case calculation of transformation of nitrite into nitrosamine rather than moving into analytical efforts
1 Like
Risk assessment always comes first and informs the extent of analytical effort required. If there is sufficient evidence indicating that there is no risk of contamination, analysis is not required.
I have seen however many drug substances falling within the scope of ICH S9 already loaded with nitrosamines detectable at ICH Q3A levels and sometimes declared by the suppliers. Nitrosamines do not form exclusively within the drug product; therefore, risk cannot be mitigated solely on the basis of worst-case calculations in the drug product.
3 Likes
In a recent case I have considered an API with a specification limit of NMT 0.20% for a secondary amine impurity. As the API falls within the scope of the ICH S9 guideline, a limit for the corresponding NDSRI would be NMT 0.10% according to ICH Q3A. In the nitrosamine evaluation I have declared that the possibility of conversion of 0.20% of the impurity to 0.10% of the corresponding nitrosamine is negligible (a conversion of 50% of the impurity to the corresponding nitrosamine would be needed, which is extremely unlikely), and that therefore no confirmatory test is required, and therefore that no development of a specific analytical method would be required. I should also add that the formation of the nitrosamine of that particular impurity up to the specification limit of 0.10% during the manufacturing process is unlikely, due to the absence of any nitrite source in the process, with the exception of traces of nitrites in the water used in the process. In our experience the only case in which nitrosamines are formed in a manufacturing process at levels at or over 0.10% is when sodium nitrite is used as a reagent in the presence of DMF and/or triethylamine, and/or a secondary amine. And this is not the case for the API which I have been evaluating.
I am convinced that this approach is correct for this particular API, but I am also aware that this might not be accepted by some of the regulatory authorities.
3 Likes
i would agree definitely that in most cases only the risk assessment is necessary but our experience has shown that some assessors want also analytical work to be done, at least the specificity as Mircea has said above.
My recommendation is to try initially the theoretical approach and be prepared for the experimental part.
Christos
3 Likes