Nitrosamines without published AI, and possibly negative for carcinogenicity

I have been wondering what the best approach is for nitrosamines with negative carcinogenicity data. For example 4-N-Nitroso-N-methylaminopyridine (CAS#16219-99-1), according to the Lhasa carcinogenicity database there is a study available with negative carc. results (Preussmann, 1979). Would such nitrosamines then be considered Class 5 impurities? Based on a single study?
What is your experience?

1 Like

If an adequate read-across surrogate is identified, the TD50 should be derived from a sufficiently robust carcinogenicity study. Parameters to consider include adequate description of the study design and appropriate histopathological analysis, number of dose groups (i.e., single-dose studies are not considered appropriate), number of animals per dose group, duration of exposure, route of administration, observed dose-response relationship.

Thank you, in this case the data is for the molecule itself, so no Read Across. There are a limited number of animals per group (but 3 dose groups), and only females included. There seems always something to be found in a study. But it was a long term examination.

Basically, the reason for my doubt is mainly the massive difference between the 18 ng/day when there is no (high quality) carcinogenicity data available, and the Q3A/B limits, when such a negative carcinogenicity study would be acceptable (despite some limitations) and the impurity can be considered as ICH M7 class 5.

Hi @Carolien …I think the study you cited is rated as not of good quality and reliable (KL 4)…From publication it seems only one dose each were tested for each three of nitrosamine isomers. In addition, only female were used and the study design is not clear…As per ICH M7 the carcinogenicity studies with following criteria are considered of lesser quality and less realiable

• < 50 animals per dose per sex;
• < 3 dose levels;
• Lack of concurrent controls;
• Intermittent dosing (< 5 days per week);
• Dosing for less than lifetime.

Therefore, as @hgajjar pointed out, it is advisable to go for read across approach with robust reliable carcinogenicity studies data. It will help to support the approach taken scientifically as well.

Thanks. Indeed Read-Across is also an option, maybe for additional information together with this less than optimal study. Some kind of weight of evidence approach may be taken.

I did notice that this molecule is also mentioned as negative in the paper by Dobo et al. 2022 10.1021/acs.chemrestox.1c00369 in Structural Group 3.

1 Like

Its good that you found it in Dolo 2022 article. Dobo 2022 also reported that study has only one dose and ony female was studied. I am sure you will find a suitable approach using collective weigh of evidence.

Hi Carolien, if there is data for a molecule, I would rather use it than doing a read-across to another molecule. Yes, many of these old studies have limitations. However, I would try it and use the Preussmann study with both arguments that 4-NMPY was Ames negative (TA100) and non-carcinogenic in a lifetime bioassay. In the study described by Preussmann, 3 derivatives were tested, only one of them (2-NMPY) was Ames positive and carcinogenic. I would also mention that in this study, more than 120 animals were used in total with a group size of 32. In addition, the doses were selected based on MTD. As an additional argument that the limitation of using female animals only does not necessarily disqualify the study, it could be explained that it has been shown for NDEA, NDMA and NNK that there is no sex difference in metabolic activation using rat nasal mucosa microsomes (Hong, 1991) and that Cyp2e1 mRNA and protein expression levels are comparable or even higher in female rat tissues compared to male rats (Gerges, 2023).
Best, Stephanie


I also agree that the chemical-specific data should be given preference. In addition to the strong points made by @SSimon regarding the Preussman et al. bioassay (and supported by other considerations), note that the ICH M7 guidance states, “The more robust studies were generally used to derive limits. However, studies that did not fulfill all of the above criteria were, in some cases, considered adequate for derivation of an AI…” Whether a bioassay is adequate/fit-for-purpose should be made on a case-by-case basis taking everything into consideration rather than following a check-box approach.


Coming back to the 4-N-Nitroso-N-methylaminopyridine (CAS#16219-99-1), I was using the CPCA, and ended up at the Potency category 2. (Alpha hydrogen score 2, no deactivating features, no activating features).

However I am still unsure as to how now to integrate the literature data.

  • Negative carc data (small groups, females only) (Preussmann R, Habs M, Pool BL. Carcinogenicity and mutagenicity testing of three isomeric N-nitroso-N-methylaminopyridines in rats. J Natl Cancer Inst. 1979 Jan;62(1):153-6. PMID: 364151)
  • Negative AMES (not a nitrosamine enhanced AMES)

@AndyTeasdale @conudel any thoughts here?!