Synthetic cannabinoids are a fast-growing group of designer drugs that tend to have higher activity at the CB1 cannabinoid receptor than traditional recreational or medical cannabis products, which can lead to serious adverse health effects or even death. The current gold standard method for detecting synthetic cannabinoids is liquid chromatography-high resolution mass spectrometry (LC-HRMS), however, a faster and less expensive method for screening emergency department patient samples could help improve treatment for patients presenting with acute toxicity. A team of researchers led by Christophe P. Stove of Ghent University has now developed a new method based on a novel machine learning model that speeds up screening for synthetic cannabinoids while maintaining high sensitivity.
The new test is an activity-based bioassay in which cells designed to express the CB1 cannabinoid receptor on their surfaces are exposed to patient blood samples. Synthetic cannabinoids in the blood cause the CB1 receptors to be activated and the cells to emit fluorescent light. The machine learning model assesses the level of fluorescence over time in order to determine which samples are positive for synthetic cannabinoids. The assay is designed to be easy to use and can be used for non-targeted screening of new synthetic cannabinoids that may be missed through targeted screening methods.
The researchers tested their new method by using it to screen 968 blood samples obtained from adult patients with acute recreational drug toxicity. The samples were screened using the new assay with results interpreted both manually by an expert and using the machine learning model, as well as via LC-HRMS so the results could be compared. The activity-based tests when reviewed manually by an expert detected synthetic cannabinoids in 141 of the 149 samples confirmed positive by LC-HRMS, and the machine learning model closely matched the manual review while allowing positives and negatives to be determined in a fast, automatic manner. Overall, the activity based screening showed a sensitivity of 94.6% and specificity of 98.5%, and the machine learning model delivered both a sensitivity and specificity of 94%. This study was published in Clinical Chemistry.
“In conclusion, the bioassay continued to demonstrate outstanding performance, confirming its potential as an ideal untargeted screening assay, capable of sensitively and universally detecting new circulating [synthetic cannabinoids],” said Stove. “Although the bioassay itself is already simple and quick, adopting a machine learning approach could potentially speed up sample scoring substantially, reducing the workload, which is ideal for a first-line screening approach complementing conventional analytical methods.”
As many as 10-30 new synthetic cannabinoids are emerging each year, making these substances one of the fastest growing groups of designer drugs and posing a major health threat. A faster, less costly universal screening method can both improve treatment for emergency department patients affected by these substances and assist public health and government officials in understanding and managing the issue.