Finding the proverbial needle: Non‐targeted screening of synthetic opioids in equine plasma

Abstract

Synthetic opioids are a class of compounds that are of particular concern due to their high potency and potential health impacts. With the relentless emergence of new synthetic opioid derivatives, non-targeted screening strategies are required that do not rely on the use of library spectra or reference materials. In this study, product ion searching, and Kendrick mass defect analysis were investigated for non-targeted screening of synthetic opioids. The estimated screening cut-offs for these techniques ranged between 0.05 and 0.1 ng/mL. These techniques were designed to not be reliant on a particular vendor's software, meaning that they can be applied to existing drug screening protocols, without requiring the development and validation of new analytical procedures. The efficacy of the developed techniques was tested through blind trials, with spiked samples inserted amongst authentic plasma samples, which demonstrated the usefulness of these methods for high-throughput screening. The use of a non-targeted screening workflow that contains complementary techniques can increase the likelihood of detecting compounds of interest within a sample, as well as the confidence in detections that are made.

Towards compound identification of synthetic opioids in nontargeted screening using machine learning techniques

“…Previous work has shown that different subclasses of synthetic opioids display class-specific diagnostic ions that can be used for nontargeted screening methods. 1,2,4 By exploiting this same phenomenon, class prediction modelling can be attempted using MS 2 data and the presence of abundant product ions within the resultant spectra. The generic structures of the three opioid subclasses included in the class prediction models, namely, fentanyl analogues, AH series and U series opioids, can be found in Figure 1 below.…”

“…Synthetic opioids are of significant concern to society due to their large public health threat. The potential of opioids to be used as performance‐enhancing drugs in horses also raises concerns in the equine antidoping community 1,2 …”

“…Recently, there has been an increasing focus on non‐targeted analysis methods to detect illicit drugs in high‐throughput screening applications, without needing to rely on certified reference materials (CRMs) or library databases 3 . Previous studies have investigated the use of product ion searching to detect novel analogues of known drug classes, exploiting class‐specific cleavages giving rise to common product ions 1,2,4,5 . Other studies have looked at more advanced data analysis methods, incorporating mass filtering techniques such as Kendrick Mass Defect analysis to detect structurally related compounds within a complex matrix 1,6 .…”

“…15,16 This study aimed to develop a machine learning approach to assist with the identification of unknown compounds indicated from nontargeted screening workflows such as those presented by our group. 1 The development of a classification model to predict the subclass of synthetic opioids based on MS 2 dissociation data is presented, alongside the development of a regression model to predict the retention time of suspected compounds from calculated molecular features. CRMs of novel drug analogues can be difficult and costly to procure.…”

Non‐targeted screening of drugs

“…Klingberg et al demonstrates a non-targeted LC-QToF-MS workflow for the screening of synthetic opioids in equine plasma using product ion searching and Kendrick mass defect analysis. 8 The detection of synthetic opioids in equine plasma continues with machine learning approaches to produce a classification model capable of predicting opioid subclasses and a regression model to predict experimental retention times. 9 The potential of broader non-targeted screening workflows using LC-QToF-MS for equine anti-doping is then explored in the short communication by Keen et al 10 Continuing the application for equine anti-doping, the Racing Laboratory of the Hong Kong Jockey Club demonstrated the required diversity for non-targeted screening workflows.…”