Establishing Best Practice for N-Nitrosamine Read-Across and Surrogate Selection

Read-across is, ultimately, achieving a confident decision on the activity of a compound based on an analogue compound with robust toxicity data. The use of read-across assessments for the determination of acceptable intakes for N-nitrosamine compounds is recommended in regulatory guidance for this critical impurity class and is given an equal weight to the tier-0 screening approach, the Carcinogenic Potency Categorisation Approach (CPCA), as well as in vitro and in vivo toxicological studies. However, read-across assessment is a multifaceted, manual process which requires the expert assessment of many factors. The complexities of performing read-across have been acknowledged as a driving force for the introduction of the CPCA. Prior to this there was a significant backlog of compounds needing a read-across assessment for a compound-specific acceptable intake (AI) as control to the class-specific threshold of toxicological concern (TTC) was unachievable. It was however clear from subjective assessment that control to the class-specific limit was unwarranted from a safety perspective. In this article the authors propose factors that should be considered in both the manual read-across and the new generation of semi-automated in silico tools for read-across assessment, in order to ensure that a robust assessment is performed, and a confident conclusion can be both reached and documented. Key case studies are presented for several NDSRIs, showing the applicability of this approach.