A practice of expert review by read-across using QSAR Toolbox

Area of Science:

• Pharmaceutical Chemistry
• Toxicology
• Computational Chemistry

Background:

• The International Council for Harmonisation (ICH) M7 guideline allows quantitative structure-activity relationship (QSAR) systems as alternatives to Ames testing for mutagenicity assessment of pharmaceutical impurities.
• The guideline lacks specific protocols for expert reviews when QSAR predictions conflict or are inconclusive.
• A conservative strategy is needed to identify potentially mutagenic chemical substances.

Purpose of the Study:

• To evaluate the efficacy of the read-across method as a conservative expert review approach for assessing mutagenicity of pharmaceutical impurities.
• To determine the mutagenicity of 36 chemical substances where initial QSAR analyses yielded conflicting or negative results.

Main Methods:

• Utilized the QSAR Toolbox to apply the read-across method for expert review.
• Selected 36 chemical substances from Japan's Industrial Safety and Health Act list, known for strong mutagenicity, where QSAR analyses were inconclusive or negative.
• Assessed mutagenicity by comparing model substances to structurally similar chemicals with known endpoint information.

Main Results:

• The read-across expert review concluded 23 out of 36 (63.8%) model chemical substances as mutagenic.
• Nine of eleven substances predicted as negative by both QSAR systems showed positive mutagenicity through their analogues.
• The read-across method demonstrated utility in a conservative approach to identify mutagenic substances.

Conclusions:

• The read-across method is a valuable tool for expert review in pharmaceutical impurity assessment, particularly when aiming for a conservative identification of mutagenic substances.
• This approach supports the ICH M7 guideline's intent by providing a practical method for handling ambiguous QSAR results.
• Expert review using read-across enhances the safety assessment of pharmaceuticals by uncovering potential mutagenic risks missed by automated QSAR analyses.