Challenges and solutions in the bioanalysis of BMS-986094 and its metabolites including a highly polar, active nucleoside triphosphate in plasma and tissues using LC-MS/MS

BMS-986094, a nucleotide polymerase inhibitor of the hepatitis C virus, was withdrawn from clinical trials because of a serious safety issue. To investigate a potential association between drug/metabolite exposure and toxicity in evaluations conducted after the termination of the BMS-986094 development program, it was essential to determine the levels of BMS-986094 and its major metabolites INX-08032, INX-08144 and INX-09054 in circulation and the active nucleoside triphosphate INX-09114 in target and non-target tissues. However, there were many challenges in the bioanalysis of these compounds. The chromatography challenge for the extremely polar nucleoside triphosphate was solved by applying mixed-mode chromatography which combined anion exchange and reversed-phase interactions. The LC conditions provided adequate retention and good peak shape of the analyte and showed good robustness. A strategy using simultaneous extraction but separate LC analysis of the prodrug BMS-986094 and its major circulating metabolites was used to overcome a carryover issue of the hydrophobic prodrug while still achieving good chromatography of the polar metabolites. In addition, the nucleotide analytes were not stable in the presence of endogenous enzymes. Low pH and low temperature were required for blood collection and plasma sample processing. However, the use of phosphatase inhibitor and immediate homogenization and extraction were critical for the quantitative analysis of the active triphosphate, INX-09114, in tissue samples. To alleviate the bioanalytical complexity caused by multiple analytes, different matrices, and various species, a fit-for-purpose approach to assay validation was implemented based on the needs of drug safety assessment in non-clinical (GLP or non-GLP) studies. The assay for INX-08032 was fully validated in plasma of toxicology species. The lower limit of quantification was 1.00ng/mL and the linear curve range was 1.00-500.00ng/mL using a weighted (1/x(2)) linear regression model. Intra-assay and inter-assay precision (CV, %) ranged from 2.3% to 5.5% and accuracy within ±2.2% from nominal. INX-08032 was found to be stable in acidified mouse plasma for at least 24h in wet ice bath, 125 days at -70°C and following at least three freeze-thaw cycles. No endogenous components in plasma were found to interfere with the measurement. The extraction recovery was between 90% and 95%. The assays for BMS-986094, INX-08144, INX-09054 and INX-09114 were qualified with wider acceptance criteria for accuracy and precision. Analyte stability was also evaluated to guide sample collection, storage, and processing. These assays were successfully applied to an investigative toxicokinetic and tissue metabolite profiling study described in the article.