We’ve just updated the data available in Vitic (to Vitic members) and in the freely-available Lhasa carcinogenicity database. The carcinogenicity data in the two is now identical, though presented in different ways - I personally use both tools, for different use cases…
Here are the news article and LinkedIn post
This contains several key papers (including Druckrey et al!) that weren’t in the original CPDB, and thus TD50s for ~20 more nitrosamines than previously available - and updated data (and thus summary TD50s) for many more compounds. To do this, especially for Druckrey, we’ve had to make the most conservative assumption for studies where no concurrent positive control was used, that a tumour incidence of zero was observed in a “control group” equal in size to the smallest dosed group.
For a good example of the new data, there are now entries for NEIPA and a Lhasa TD50 is now provided for NDPhA
We also include an automated reliability score for the data (e.g. were there multiple dose groups, was the experiment time and dosing regimen acceptable…), and chi2 p-values for the fit of the TD50 model to the data. A poor fit may either mean that there were threshold effects or saturation; it’s not inherently a reason to assume a different dose-response and discard the TD50 (for example, there are compounds with extremely reliable data, but p > 0.05 because saturation at higher doses, possibly combined with toxicity, results in the poor fit) but does mean that you should double-check. Likewise, the reliability score information is an automated check, so the conclusions should again be reviewed before being incorporated into your decision.