P61: Implementation of a systematic toxicological screening method for post-mortem analysis using a UPLC-QToF MS system

Abstract

Introduction General unknown screening (GUS) is the procedure used by toxicologists in order to unambiguously identify the xenobiotics involved in an intoxication case. Due to its high specificity, GC-MS was long considered the gold standard for GUS in clinical and forensic toxicology. Still, GC-MS is poorly suited for the detection of several xenobiotics and their more hydrophilic metabolites. Here we describe the optimization and implementation of a comprehensive high-resolution drug screen in biological matrices using a Waters UPLC XEVO G2 QToF MS system. We also examined the transferability of the high-resolution spectral lists established on mirror systems by comparing our data against two other instruments of the same model and manufacturer using the same chromatographic separation (Humbert et al., 2012; Rosano et al., 2013). Methods Drugs and metabolites were extracted from spiked matrices using SPE cartridges or liquid-liquid extraction. Extracts were sequentially injected into a UPLC coupled to a hybrid quadrupole- time of flight system (MSe) in positive and negative ESI mode. Data was collected using MSe acquisition which performs two sets of acquisition simultaneously and compared to a list containing the exact mass, the retention time and the exact mass fragments as previously described (Humbert et al., 2012; Rosano et al., 2013). In order to optimize the extraction procedure, we modeled different approaches (reverse phase SPE, mixed-mode cation exchange SPE and liquid-liquid extraction) using 18 different xenobiotics frequently encountered in toxicological investigations. Results The most efficient extraction process was chosen using the following criteria: ion suppression, recovery and the limit of detection. Based on our data, mixed-mode cation exchange SPE demonstrated the best overall performance and was selected to be compared to our reference GC-MS screening method. We then evaluated the recoveries, ion suppression and the limits of detections for 225 xenobiotics which are part of our GC-MS screen. Overall our transferability study showed few discrepancies in mass or fragments generated. Retention time reproducibility was 99% using a retention time tolerance of 0.4min and 96% using a 0.3 window. Finally, we performed a comparison between the test LC-QToF method and our reference GC-MS method on real samples to evaluate its performance; several cases will be presented. In summary, these cases demonstrated that the LC-QTOF method was able to identify more xenobiotics in the following categories: benzodiazepine and stimulant metabolites (7-amino-clonazepam, 4-methylamphetamine and desmethyl-diphenhydramine), anti hypertensive drugs metabolite (OH-propranolol), antiretroviral drugs (such as ritonavir and darunavir) and hydrophilic anticonvulsant pregabalin. Conclusion Overall, the LC-QToF method demonstrates increased sensitivity compared to GC-MS and should be more specific. Our evaluation of the transferability of lists between labs indicates that retention variability could be an issue for a small fraction of compounds. Also, in order to detect some families of xenobiotics, such as barbiturates and NSAIDS, the addition of a second injection (using negative ionization) was necessary.