|
Abstract INTRODUCTION:
Liquid chromatography–mass spectrometry (LC-MS) is a cornerstone of clinical toxicology, enabling detection of drugs contributing to complex clinical presentations and population-level drug use trends. Untargeted high-resolution mass spectrometry (HRMS) using information-dependent acquisition (IDA) allows broad detection and retrospective analysis of data. However, standardized guidance for validation of qualitative HRMS-based comprehensive drug tests is limited. Here, we describe the validation of a comprehensive drug test following transition from a SCIEX TripleTOF®5600 to an X500R QTOF platform. The assay also incorporates an expanded spectral library reflecting drug use trends in our patient population.
METHODS:
Urine samples were prepared by dilute-and-shoot; serum, plasma, and whole blood samples underwent protein precipitation. Chromatographic separation was performed on a Kinetex C18 column using a 10-minute gradient. Data were acquired on a SCIEX X500R QTOF in positive-ion mode using full-scan TOF MS with IDA (up to 20 dependent MS/MS scans). Data was analyzed on Sciex OS.
Compound identification utilized a weighted combined score based on accurate mass, retention time, isotope pattern, and MS/MS library matching, with manual review of all results, as described (Colby et al.) An in-house spectral library (>200 compounds) was developed, including both illicit and prescription drugs as well as many of their metabolites. Classes of compounds represented in the library include but are not limited to opioids, stimulants, benzodiazepines, hallucinogens, anti-psychotics, anti-depressants, sulfonylureas, and sedatives. Compounds newly added to the method while transitioning to the new instrument include fentanyl analogs and designer benzodiazepines.
Validation included accuracy (88 urine, 38 blood patient samples) compared across three platforms (5600, X500R, and 7600 ZenoTOF), with consensus positivity defined as ≥2 methods. Limits of detection (LOD) were determined across 4-5 matrices per specimen type using spiking studies and defined as the lowest concentration at which the compound was reportable in 3 out of 4 (urine) or 4 out of 5 (serum) matrices. Precision of MS1 peak areas and retention times was assessed using inter-day (n=25 across 11 days) and intra-day (n=10) QC replicates. Carryover was evaluated using high-positive samples followed by blanks, with percent carryover calculated from MS1 peak areas. Urine matrix effects were assessed across 10 matrices. Serum process efficiencies (combined matrix effects and recovery efficiencies) were determined across 4-5 matrices. Stability was evaluated over 72 hours in processed samples store at 10°C in the autosampler and original specimens stored at 4°C by comparing MS1 peak areas. Matrix equivalency across serum, plasma, and whole blood was assessed using patient comparison samples.
Positive and negative determinations used to calculate clinical sensitivity were determined for drug “groups” formed of compounds and their metabolites, synthetic precursors, and/or byproducts, where applicable.
RESULTS:
The X500R assay demonstrated 94% (urine) and 92% (serum) analytical sensitivity, 94% clinical sensitivity (urine and serum), and >99% specificity. Sensitivity estimates are likely conservative due to sample degradation and comparison to the higher-sensitivity 7600 ZenoTOF platform. LODs ranged from 5–5000 ng/mL in urine and 1–1000 ng/mL in serum. Precision studies showed low variability in retention times and acceptable variability in MS1 peak areas for a qualitative drug test. Carryover was negligible in most cases; rare instances of detectable signal in blanks were <1% of the preceding sample, and review criteria were implemented to mitigate false positives. Matrix effects in urine and process efficiencies in serum were compound-dependent, with median values of 86% (interquartile range (IQR): 76-98%) and 82% (IQR: 58-108%), respectively. Urine samples were stable for up to 72 hours in both processed and unprocessed conditions. Serum samples showed minor degradation (15% on average) after 72h in both processed and unprocessed conditions. Compounds that degraded by more than 50% in processed serum samples stored in the autosampler at 10°C for 72h were: bupropion (85%), cocaine (96%), ecgonine methyl ester (68%), cocaethylene (80%), norcocaine (55%), and psilocin (55%). Plasma, whole blood, and serum demonstrated equivalent sensitivity.
The expanded library enabled detection of newly included compounds in approximately one-third of urine patient samples, most commonly 4-ANPP (fentanyl precursor) and beta-hydroxyfentanyl (fentanyl analog and metabolite).
CONCLUSION:
The SCIEX X500R platform demonstrates robust performance across key validation parameters, including sensitivity, specificity, and carryover; the comprehensive drug test was implemented clinically. Expansion of the spectral library to reflect evolving drug use enhances clinical utility and supports toxicology surveillance efforts.
REFERENCES:
1. Colby, J.M., Thoren, K.L., Lynch, K.L. Optimization and Validation of High-Resolution Mass Spectrometry Data Analysis Parameters. J. Anal. Toxicol. 2017. 41(1): p. 01-05. |
= Discovery stage. (57.21%, 2026)
= Translation stage. (23.38%, 2026)
= Clinically available. (19.40%, 2026)
