Prediction of N-Nitrosamine Partition Coefficients for Derisking Drug Substance Manufacturing Processes

The risk assessment and control of N-nitrosamine impurities in medicines is a topic of great concern for patients, regulators, and the pharmaceutical industry. This paper discusses a modeling approach to predict the solvent–water partition coefficient of 28 N-nitrosamines in 134 water immiscible solvents, including those solvents typically used in pharmaceutical manufacturing. The Conductor-like Screening Model for Real Solvents model has been used because its implementation from first principles does not require regressing specific interaction parameters for the N-nitrosamine functional group. It is proposed that predicted partition coefficients combined with partition modeling should be used to support the industry and regulators in making scientifically justified predictions of the control of N-nitrosamine impurities potentially formed during the manufacture of pharmaceutical products.