What is the added value of liquid chromatography-tandem mass spectrometry for routine urine drug screening in clinical situations?

Abstract

Replacing the common two-step process of urine drug screening by a single-step methodology based on selective and quantitative liquid chromatography-tandem mass spectrometry (LC-MSMS) seems quite attractive. With this approach, it may be more appropriate to use the limit of analytical quantification instead of a cutoff concentration to report positive results. In this study, we want to evaluate the impact of this modification on the results of urine samples received in clinical situations. Results of all urine drug screens and associated confirmations performed between July 2021 and March 2022 for amphetamine, benzodiazepine and opiate groups, cannabis and cocaine metabolites and methadone were retrieved from the laboratory information system of the Medical Biology Department of Laboratoire National de Santé in Luxembourg. The samples came from hospitals, addiction centers, prisons and occupational health clinics. They were analyzed by an automated immunoassays, Cobas (Roche Diagnostics) and by a targeted quantitative screening assay, MassTox Drugs of Abuse testing, (Chromsystems) on a 6500QTRAP LC-MSMS (Sciex). We calculated the number of samples where results were discordant. Cases where no drugs were identified by LC-MSMS but results of the immunoassay were above the cut-off concentration (based on the screen test cut-off concentration of European guidelines) were considered as false positive (FP) results. Whereas cases where drugs were quantified by LC-MSMS but results of the immunoassay were below the cut-off concentration were considered as false negative (FN) results. The FP and FN results were 1 and 5% for cannabis metabolite ( n = 477), 0 and 14% for cocaine metabolite ( n = 189), 2 and 20% for methadone ( n = 211), 1 and 10% for benzodiazepines ( n = 489), 1 and 20% for opiates ( n = 149) and 48 and 2% for Amphetamines ( n = 92). The median concentrations of FN were 18 ng/ml (range from 15 to 31) for THC-COOH, 62 ng/ml (range from 24 to 121) for benzoylecgonin, 221 ng/ml (from 63 to 785) for methadone and 367 ng/ml (from 66 to 3174) for EDDP. In the benzodiazepine FN results (median concentration 69 ng/ml, range from 26-318), oxazepam, lorazepam and OH-alprazolam were the most frequently detected drugs. Among the opiate FN results, morphine associated or not with codeine was the most frequently quantified drug (median 81 ng/ml, range from 27 to 519). Four out of the 29 FN samples contained 6-monoacetylmorphine (from 17 to 136 ng/ml) in the absence of morphine, 2 samples were positive for pholcodine and one sample contained high concentration of meperidine. The two FN samples for amphetamines contained amphetamine (500 and 800 ng/ml). Nine out of the 42 amphetamines FP samples contained ritalinic acid (from 10,000 to 130,000 ng/ml). This targeted method improves the sensitivity but is restricted to a pre-defined list of drugs, although it covers a wider panel than commercial immunoassays, some relevant molecules may be missing. The hydrolysis protocol used before LC-MSMS analysis also explains the increase in sensitivity for benzodiazepines and opiates (allowing the detection of conjugated metabolites). Decreasing the threshold of detection for urine drug screening allowed for the extension of the drug detection window, improving the detection of drugs administered at low dose or those reacting poorly in immunoassay. This single step approach bypasses the main pitfalls of the immunoassays, namely the lack of analytical specificity (mainly for amphetamines) and a screening limited to well-established (parent) drugs, with consequences on patient management still only partially objectified. This targeted quantitative screening LC-MSMS method with its high sensitivity, accurate quantification, and associated internal quality control, achieves the required analytical performances expected in a clinical setting. The implementation of this approach will largely be conditioned by the optimization of the sample throughput to ensure a turn-around-times compatible with various clinical situations and the availability of the equipment and highly trained staff.