The use of High Resolution Mass Spectrometry (HRMS) for comprehensive (non-targeted) screening approaches has increased over the past decade in forensic toxicology. To fully implement this technique in a routine laboratory with increasing demands concerning throughput and quality (e.g. ISO 17025), the method should be easy to apply and data interpretation should be straightforward. Few guidelines of the forensic field specifically addressed specific issues concerning identification criteria, adequate validation experiments and quality assurance (QA) for HRMS approaches. The aim of this study is to describe a practical HRMS approach and to propose suggestions for validation and QA.

Sample preparation consisted of a protein precipitation in ice-cold acetonitrile before evaporation and reconstitution in ammonium formate buffer. ß-theophylline was added as internal standard (IS). Three μL of the extract were then injected onto a 150mm×2.1mm HSST3 C18 UPLC column. The applied gradient started with 87% water/ammonium formate 5mM pH3 (A) and 13% acetonitrile 0,1% formic acid (B). At 10minutes, 50% A and B was linearly obtained at a constant flow rate of 0,4mL/min. At 16min., 1% A and 99% B was programmed to be reached exponentially (curve 7) and at 18,5min. starting conditions were applied again up to the 21th minute. Data were acquired on a Xevo G2 XS-QTOF (Waters, Milford, US) and processed using the UNIFI toxicology screening solution, combined with the web-based HighResNPS database.

Our solution is a 21minutes run based on the Waters UNIFI Forensic Toxicology Screening Application with improvements to detect New Psychoactive Substances (NPS) such as synthetic cannabinoids (SCRAs) which originally eluted in the washing phase. About 15 samples in 5hours could be analysed with simultaneous data processing and reporting. The automated workflow filters data on mass accuracy (< 5ppm), fragment ion (min 1) and a response (>10,000). Additional information for visual inspection concern isotope match, retention time (observed and expected) and fragments (found and count) is given. QA is set-up via an initial suitability test looking into sensitivity, reproducibility, mass accuracy and resolution (full system information including LC, MS and contamination level control), quality control samples (mix, blanks) and proficiency tests (ACQ Science EEQ, GTFCh QSA). The method was validated for blood with regards to selectivity, sensitivity (LOI), matrix effects, stability and carry-over for 25 compounds eluting throughout the analytical run (3-Fluorophenmetrazine, 5F-ADB, Acryloyl fentanyl, Acryloyl fentanyl, Bromazepam, Deschloroetizolam, Diazepam, Diclazepam, Diphenidine, EG-018, Fentanyl, Flunitrazolam, Fluoxetine, FUB-144, Furanyfentanyl, Gabapentin, Ketamine, Mephedrone, MMB-022, NM-2AI, NSI-189, Oxazepam, Risperidone, Tramadol, Trazodone, Valeryl fentanyl). Sensitivity throughout the batch was normalized via the intensity of the IS to establish the robustness and potential impact on data interpretation. Selectivity was evaluated via blanks versus other samples in the same batch to evaluated ‘false positive’ hits in the database. In addition, authentic samples were applied to compare the HRMS TOF data with established LC-MS/MS methods.

Discussion As more laboratories consider HRMS for comprehensive drugs screening, a consistent approach to data analysis and compound identification will be crucial. Carefully selecting, validating and controlling the positivity criteria used in HRMS data analysis is a step in this direction and necessary for ISO17025 accredited laboratories. The choices made in the proposed workflow are evaluated via authentic cases. Identification and validation criteria will be checked against current international guidelines.

The developed targeted HRMS workflow improves the simplicity of the procedure and limits the time-constraints and the need of highly experienced personnel for interpretation.