Advanced quench correction model for quantitative offline radioactivity counting in in vivo drug metabolism studies

A novel workflow for quantification of a drug and its metabolites in in vivo studies has been developed in the context of a radiolabeled human mass balance study. Samples are analyzed with ultra-high-performance liquid chromatography, and fractions are collected in a 384-well plate, which is subjected to offline counting, providing improved detection limits over online radioactivity detection. We discuss an advanced strategy to account for signal suppression or quenching, which significantly affected results in the offline counting of feces and urine samples in the selected case example, to provide more accurate quantification. The new quench model fits 2 data sets from 384-well plates with the actual matrices present to perform counting efficiency correction. Improved results were obtained over the existing approach, where a generic quench curve is defined by only a limited number of points made from a dilution series of a quenching agent. To account for outliers, a robust quartic model was applied. The new model effectively describes matrix-induced quenching and corrects for this, resulting in correct profiles with improved overall recovery as corroborated by comparison with online radioactivity detection and liquid scintillation counting, and can generically be applied postacquisition. The strategy was applied to all 36 fecal extracts from a human absorption, distribution, metabolism, and excretion study, where half of the samples present less than 20,000 disintegrations per min/mL, increasing the average column recovery (sum of individually quantified peaks relative to the total injected radioactivity) to >85%. SIGNIFICANCE STATEMENT: To improve interpretation in radiolabeled absorption, distribution, metabolism, and excretion studies, a matrix-based quench correction model is developed. It compensates for matrix-induced signal suppression when analyzing in vivo samples via offline radioactivity counting. It greatly improves data quality and enables accurate assessment of the true significance of detected metabolites.