Quantification of THC, metabolites, and related compounds in whole blood using LC-MS/MS

Abstract

The objective of this study was to develop and validate a method for cannabinoids in whole blood. Cannabis is a frequently used recreational drug in Sweden. Even though smoking is the preferred route of administration an increase in the use of edible products have increased during recent years. Edibles can be products mainly containing cannabidiol (CBD) and only minor amounts of tetrahydrocannabinol (THC) but also products with THC as the main active compound. Studies have shown that the metabolism of THC is different when consumed by the oral route compared to smoking. To enable better interpretation, we aimed to develop an automated supported liquid extraction routine method including the analytes THC, carboxy-THC, 11-hydroxy-THC, CBD, tetrahydrocannabinolic acid A (THCA-A), and cannabidiolic acid (CBDA). Sample preparation was performed on a Hamilton Star and a Biotage Extrahera robot. A 0.2 g whole blood aliquot was transferred to a 96-well plate and diluted with 0.1% formic acid and internal standard (1:1) before extraction with 3 × 600 μL of tertbutylmethylether. After evaporation samples were reconstituted in 100 μL 0.2% formic acid in water/methanol 30/70. All analytes were detected using Waters Xevo TQ-XS triple quadrupole MS with a UniSpray ionization source operated in both positive and negative ion mode. Chromatographic separation was achieved using a 10 μL injection and an 8.0 min gradient elution on a Waters Acquity Premier HSS T3 column with a Waters Acquity I-Class UHPLC system. Mobile phase A was 0.2% formic acid in water and mobile phase B was methanol. Method validation included selectivity, qualitative matrix effects, extraction recovery, calibration function, limit of quantitation, precision and accuracy, carry over, dilution integrity, stability in matrix as well as in extracted samples. The method was applied to authentic cases as a proof of concept. All analytes were baseline separated. No interferences were seen at the retention times of the analytes. Qualitative matrix effects were less than 10% except carboxy-THC (27%) and 11-hydroxy-THC (66%). Extraction recoveries were 29% for THC, 68% for carboxy-THC, 49% for 11-hydroxy-THC, 29% for CBD, 63% for THCA-A and 67% for CBDA. All analytes were best described by quadratic calibration curves using 1/X weighting with observed mean deviations from the nominal value of -8% to 8%. Calibration ranges were established from 0.0002–0.03 μg/g for all analytes except carboxy-THC (0.002–0.3 μg/g). Within-day precision was less than 5% for all analytes with accuracies between 93 and 101%. Between-day precision was less than 10% for THC, carboxy-THC, 11-hydroxy-THC, and CBD but 14% and 25% for THCA-A and CBDA, respectively. All accuracies within 95–108%. A maximum carry over of 0.8% was seen but did not result in false positive samples. Up to 4 times dilution gave results within 20% of the true value. All analytes except THCA-A were stable during two days at room temperature, three weeks at 4 °C as well as in extracts stored at 10 C. THCA-A showed significant degradation over time at room temperature as well as in extracts. Results from 254 samples run during the first weeks of routine showed that 99.9% of concentrations were within the calibration ranges. The mean concentrations (μg/g) were 0.0053, 0.0023, 0.042, 0.00073, and 0.00094 for THC, 11-hydroxy-THC, carboxy-THC, CBD, and THCA-A, respectively. 97.7% of 11-hydroxy-THC/THC ratios were below 1.0 pointing towards smoking as the route of administration but showed detectable THCA-A in 23.1% suggesting that combustion was not complete. Only one case had a CBD concentration higher than THC suggesting intake of CBD products. We conclude that the method showed acceptable validation results and was suitable for routine analyses of cannabinoids in whole blood samples.