|
Abstract INTRODUCTION:
Liquid Chromatography coupled to triple quadrupole-based tandem mass spectrometry (LC-MS/MS) is a dominant approach for confirmatory drug testing in clinical toxicology. However, standard LC-MS approaches are highly targeted and susceptible to interferences from closely related compounds, and are not suitable for untargeted general unknown screening (GUS) or drug supply monitoring in clinical settings. In contrast, liquid chromatography high-resolution mass spectrometry (LC-HRMS) on quadrupole time-of-flight (Q-TOF) systems collect full-scan mass spectra with rapid acquisition of MS/MS spectra in an untargeted manner, enabling robust GUS applications. However, LC-HRMS-based GUS approaches suffer from limited dynamic range, owing to inefficiencies in ion utilization, MS/MS acquisition, and fragmentation energies. Together, these factors can lead to misidentifications and false-negative results for lower abundance or poorly ionizing targets.
To address these limitations in routine toxicological testing, we investigated the value of the ZenoTOF 7600 HRMS platform (Sciex), which implements a novel electron-based fragmentation mechanism (Electron Activated Dissociation/EAD) alongside linear ion trap pulsing (i.e., Zeno trap pulsing) to improve the sensitivity of MS/MS acquisition. For GUS applications, we also compared traditional IDA MS/MS acquisition to Sequential Window Acquisition of all Theoretical Spectra (SWATH) acquisition on this platform, which can potentially improve sensitivity by guaranteeing MS/MS acquisition for all compounds within a defined mass range.
METHODS:
The impact of linear ion trap pulsing on GUS screening was evaluated by establishing limits of detection and matrix effects for 150 toxicological compounds found in the San Francisco drug supply. Compounds were divided into three drug cocktails, and spiked into drug-free urine at concentrations ranging from 1-5000 ng/mL. Samples were diluted in triplicate before untargeted LC-HRMS investigation. Full-scan high-resolution mass spectra was acquired for all precursors between 125-650 m/z, and MS/MS were acquired using either Zeno-SWATH or Zeno-IDA modes (Acc. Time: 25 ms, Cycle Time: 1.25 s) with CID fragmentation (CE: 35 ± 15V). Library spectra were generated by analyzing each compound in pure solvent using the MRM-HR acquisition mode.
The resulting data was analyzed in SciexOS, and limits of detection were defined as the lowest concentration at which an analyte was identified in all replicates. Interferences were evaluated by cross comparison with drug cocktails not containing the drug of interest, and matrix effects assessed by spiking compounds into six different blank urine matrices and at high and low concentrations, and comparing peak areas to matrix-free samples. A method comparison was conducted against our previously published and validated LC-HRMS GUS approach using the Sciex 5600 Q-TOF system using 101 remnant urine samples representing 110 of the 150 compounds in our library.
To evaluate EAD fragmentation, EAD library spectra were generated for all 150 previously characterized toxicological agents using the manufacturer’s recommended settings for small molecule analysis (CE: 12 V, KE: 10 keV, Acc. Time 95 ms). EAD and CID spectra from each molecule were directly compared to identify novel diagnostic fragment ions, and candidate fragments were further compared across isomers and drug classes to evaluate selectivity and potential interferences. Finally, an EAD-based “dilute-and-shoot” test for urine buprenorphine and its metabolites was developed and validated according to ANSI/ASB standard guidelines to establish the value of EAD fragmentation in quantitative toxicology applications.
RESULTS: Zeno-SWATH acquisition enabled the identification of more toxicological agents than Zeno-IDA at every concentration evaluated. In particular, Zeno-SWATH acquisition enabled the detection of 42 toxicological compounds at 1 ng/mL, while Zeno-IDA did not detect any compounds at this threshold. Furthermore, over half of the analyzed compounds exhibited LODs at least 5-fold lower when using Zeno-SWATH than when using Zeno-IDA (p < 0.01). Interferences for both methods consisted exclusively of closely eluting, isobaric compounds with highly similar MS/MS spectra; Zeno-IDA misidentified 8 compounds, while Zeno-SWATH acquisition misidentified 6 compounds. Matrix effects were significantly reduced with Zeno-SWATH acquisition (median ME: -4% vs -8%; p.adj < 0.001). The method comparison reported ~10% (Zeno-IDA) and ~35% (Zeno-SWATH) increases to the number of identified compounds compared to IDA-based GUS on the Sciex 5600 platform.
EAD fragmentation improved the generation of diagnostic fragment ions for 54 analyzed compounds. The most significant qualitative improvements were observed for amine stimulants (e.g. amphetamine, methamphetamine, phentermine), atypical opiates (e.g. buprenorphine, nitazenes), and benzodiazepines (e.g., adinazolam, bromazolam), among others. In particular, we observed that the isomeric amine stimulants methamphetamine and phentermine, which are challenging to separate on column or distinguish by CID fragmentation, productively generated unique fragment ion signatures.
A quantitative EAD-based urine buprenorphine test was validated, and provided analytical measuring ranges of 0.2-800 ng/mL for buprenorphine/norbuprenorphine and 1-4000 ng/mL for the glucuronide metabolites. Precision and bias were well controlled under 20% at all QC levels and at the lower limit of quantitation.
CONCLUSION: Linear ion trap pulsing and EAD fragmentation improve sensitivity and selectivity for toxicological agents in both GUS and targeted applications. Quantitation by EAD is also robust over a wide dynamic range. These factors make the Sciex 7600 ZenoTOF system a versatile platform for clinical toxicology, public health/drug supply monitoring, and research on novel and emerging toxicological agents.
|
= Discovery stage. (53.14%, 2025)
= Translation stage. (22.33%, 2025)
= Clinically available. (24.53%, 2025)
