BIOMEDICAL AND ENVIRONMENTAL MASS SPECTROMETRY, VOL. zyxwv 16, 367-372 (1988) zyxw Optimization of the Quantitative Analysis of the Major Cannabis Metabolite (1 1-nor-9-COOH- A9-tetrahydrocannabinol) in Urine by Gas Chromatography/Mass Spectrometry M. Congost, R. de la Torre and J. Segura-f Institut Municipal d'Investigaci6 Mtdica, Passeig Maritim 25-29,08003 Barcelona, Spain A gas chromatographic/mass spectrometric electron impact method is presented for the detection and quantifica- tion of 1 zyxwvutsr l-nor-9-carboxy-A9-tetrahydrocannabinol (THC-COOH) in urine, for use in the confirmation of presump tive results obtained by other techniques. Four extraction procedures, two solid-liquid and two liquid-liquid, have been compared. A comparison of two trimethylsilylating methods demonstrates that the best results are obtained by the use of a mixture containing N-methyl-N-trimethylsilyl-trifluoroacetamide, trimethyliodosilane and dithioerithritol (100:0.2:1) v/v/w. The use of ketoprofen as a new internal standard for the quantification of THC-COOH has proved to be very effective. Both spiked samples and samples from cannabis users have been successfully analysed. It has also been demonstrated that the presence of other drugs of abuse in urine samples do not interfere with cannabis quantification by the method reported here. INTRODUCTION The abuse of zyxwvutsrqp Cannabis sativa L. products, either as marihuana or hashish, is a world-wide problem. The detection of cannabis intake is becoming increasingly important in different situations with clinical, legal, occupational and epidemiological implications. The analysis is generally performed in urine and directed towards the detection of 1 1-nor-9-carboxy-A9-tetra- hydrocannabinol (THC-COOH), the major metabolite of A9-tetrahydrocannabinol (THC) in this biological fluid, and is present mainly as an ester conjugate of glucuronic acid.' The method most commonly used in laboratories is the enzyme immunoassay EMIT,' which detects THC-COOH, but with high cross-reactivity towards other urinary metabolites including 1 1-OH-A9- THC, 8-B-OH-A9-THC, 8-B-1 1-OH-A9-THC.3*4 It is always recommended that positive results obtained by immunological techniques be confirmed because of the lack of specificity of the antib~dy.~.~ Negative results should also be confirmed if it is suspected that adulter- ants which could interfere with the assay (e.g. soap preparations, vinegar, salt) have been added to the sample. zyxwvutsrq ' Several techniques have been reported for the confir- mation of presumptive cannabis detection by immuno- logical methods including thin-layer chromatography (TLC),'-'' high-performance liquid chromatography (HPLC),' '*l' HPLC combined with radioimmuno- assay,I3 gas chromatography with electron capture detection (GC/ECD)6 and gas chromatography/mass spectrometry (GC/MS).'"I6 A highly sensitive and spe- t Author to whom correspondence should be addressed. 08874134/88/240367-06 $05.00 zyxwvuts 0 1988 by John Wiley & Sons, Ltd. cific method for the quantitative detection of THC- COOH in urine is presented in this report. Different procedures for extraction and derivatization have been compared. EXPERIMENTAL Extraction of urine samples Four different extraction procedures, two solid/liquid (S/L) and two liquid/liquid (L/L), were studied. An automated centrifugal sample processor (Prep I, Du Pont, USA) was used for the S/L methods. Urine samples (3 ml) were hydrolysed (0.3 ml 10 N KOH) for 15 min at 60°C after addition of internal standard (ketoprofen, 500 ng ml-'). Procedure 1 used an octadecilsilane-bonded silica resin previously washed with methanol and water. The hydrolysed sample is adjusted to pH 34 (0.45 ml acetic acid) and passed through the column. Two consecutive washing steps are carried out (2 ml water and 2 ml acetonitri1e:water (40: 60)) before eluting with methanol (2 ml). Procedure 2 is based on ion exchange (NH4+C1- silica resin), washing with methanol (2 ml) and eluting with ethyl acetate : methanol : acetic acid (90 : 10: 4) v/v/v (2 ml). When using the L/L procedures, internal standard is added to the urine sample (5 ml) and it is hydrolysed as before. Samples are immediately acidified in procedure 3 (0.5 ml 10 N HCl plus 1 ml 1 N H3PO4) and extracted over 30 min in a rotatory shaker (40 cycles min ~ ') with hexane :ethyl acetate (7 : 1). The organic phase is then extracted with 0.5 N KOH (5 ml). This aqueous phase is acidified (2 N HCl, 2 ml) and finally extracted in