Detecting Marijuana In Saliva

Sun, 10/17/2010 - 7:29pm
Eric ChiJason Cole

Using an MRM method on a GC-triple quadrupole MS to confirm and quantitate THC in oral fluid is an effective alternative to blood and urine samples.

Analyzing saliva for the presence of marijuana is an effective alternative to blood and urine samples.Marijuana is the most widely used illegal drug in the United States. Each year U.S. law enforcement agencies apprehend over two million pounds of marijuana in various forms. Initial evidence is first submitted to forensic laboratories and screened for marijuana by microscopic inspection and simple chemical tests. The identification of cannabinoids must then be confirmed as showing positive. Oral fluid analysis is growing in popularity as an alternative to blood and urine analysis for delta-9-tetrahydrocannabinol (THC) detection. Oral fluid can be analyzed without the embarrassment of urine sample analysis or the pain incurred by blood analysis. Similarly, oral fluid analysis necessitates much less training in comparison with the extensive training required for blood analysis. Due to the low concentration of THC normally found in oral fluid and the low volume of oral fluid readily available for this type of analysis, oral fluid analysis is not, however, without its challenges.

This article discusses the regulatory guidelines in the U.S. for the analysis of THC, examines alternative methods used for this type of analysis, and demonstrates how a forensics crime laboratory uses a triple stage quadruople GCMS/ MS system for the confirmation and quantitation of THC in an oral fluid matrix.

The Drug
Marijuana is the world’s most commonly used illicit drug. Although marijuana contains a minimum of 400 different chemicals, the main active chemical is THC. The membranes of certain nerve cells in the brain contain protein receptors that bind to THC.Once securely in place, THC sparks a series of cellular reactions that ultimately lead to the euphoric high that users experience when they smoke marijuana. Marijuana has also been proven to produce a sedative effect.1

Recent studies have shown that the usage of marijuana directly increases the users’ susceptibility to certain brain, heart, and lung diseases. THC has also been proven to impair the immune system’s ability to fight disease. Data released by the National Survey on Drug use and Health in 2007 noted that 14.4 million Americans aged 12 or older used marijuana at least once in the month prior to being surveyed.

Short-term effects of marijuana include problems with memory and learning, distorted perception, difficulty in thinking and problem solving, loss of coordination, and increased heart rate. Research for long-term marijuana usage indicates changes in the brain similar to those seen after long-term abuse of other major drugs. For example, cannabinoid (THC or synthetic forms of THC) withdrawal in chronically exposed animals leads to an increase in the activation of the stress-response system and changes in the activity of nerve cells containing dopamine. Dopamine neurons are vital to the regulation of motivation and reward and are directly or indirectly affected by all drugs of abuse.2

Public health concerns over marijuana have sparked the strict legislation regulating the production, distribution, and use of marijuana in the U.S.

Regulatory Overview
The U.S. Drug Enforcement Administration (DEA) regulates marijuana under the Controlled Substances Act (CSA), which was introduced to prevent the abuse of drugs and other substances.

The CSA act consolidates multiple existing federal laws regulating the manufacture and distribution of narcotics, stimulants, depressants, hallucinogens, anabolic steroids, and chemicals used in the illicit production of controlled substances.

The CSA categorizes all substances into one of five schedules according to specific criteria. Within this CSA framework, marijuana is placed into Schedule 1. Schedule 1 drugs are “classified as having a high potential for abuse, no currently accepted medical use in treatment in the United States, and a lack of accepted safety for use of the drug or other substance under medical supervision.” The quantity of Schedule 1 substances that may be produced in the U.S. in a calendar year is strictly limited by the DEA.

While alternative methods for detecting THC in saliva include the radio immunoassay (RIA)method, gas chromatography with electron capture detection (GC-ECD), and liquid chromatography with electrochemical detection, gas chromatography-mass spectrometry (GC-MS) is still the preferred method for confirmatory analysis or for screening and confirmation in one step.3 Gas chromatography tandem mass spectrometry (GC-MS/MS) offers high analysis speeds and is extremely selective and sensitive, enabling routine analysis of THC in oral fluid at the low levels required by regulatory bodies.

Application Example
A triple stage quadrupole was implemented in a working forensic laboratory to monitor THC, cannabidiol, and cannabinol in oral fluids. The triple stage quadupole was preferred above a single quadrupole due to its ability to detect analytes at very low quantitation limits in complex biological matrices. This is achieved without the requirement of a multi-dimensional GC approach which requires significant time and effort to set up and maintain and often makes the progress of additional assays on the instrument more difficult. Software was used to provide automated sample analysis and quantitation for method validation, more specifically the evaluation of precision and linearity.

This method utilized a productive procedure for high-throughput GC- MS/MS confirmation and quantitation of A9-THC in oral fluids. In this instance, the method of solid phase extraction (SPE) was used due to its ease-of-use and the untainted nature of the consequent extracts.

Sample Preparation
Negative calibrator oral fluid was used for sample preparation. A sample size of 200μL was decided upon. Calibrators, quality controls, and linearity samples were spiked with the necessary volume of THC. Three-point calibration at 0.2, 2, 20ng/mL was used for the measurement of all quantitative amounts. Prior to extraction, samples were brought to an approximate pH of 6 by adding 2ml of pH6 phosphate buffer.

Once extracted, the samples were evaporated to dryness at 40°C under nitrogen, with care taken not to excessively dry the extracts. Dried samples were then derivatized with 30μL of BSTFA at 80°C for 20 minutes, following which the access BSTFA was evaporated to dryness at 40°C under nitrogen. For analysis, 50μL of toluene was added to the derivatized extracts, with the consequent samples being loaded onto an auto sampler for GC/MS analysis (Table 1).

Summary of the sample preparation steps.

Instrumental Analysis
A gas chromatograph was equipped with a standard split/splitless injector. A 5mm ID deactivated glass liner was used in the injector with a glass wool plug and the split/splitless injector temperature was set to 250°C.A2μL injection volume was programmed on the auto sampler and a splitless injection was used. The analytical column was a 15m x 0.25mmID x 0.25μm film (p/n 260F130P), which was installed 64mminto the injection port.

Carrier gas flow was set to a constant rate of 1.2mL/min of helium. The initial temperature on the gas chromatograph was set to 60°C. Upon injection of the sample, the oven temperature was immediately ramped at 35°C/min to a final temperature of 320°C with no final hold. This took place for a total run time of 7.43 minutes and a THC retention time of 5.77 minutes. Following this, the temperature was set to 200°C and the mass spectrometer was tuned using default parameters. These tune settings were used for acquisition, with the default detector gain for the multiple reaction monitoring (MRM) mode left at 2 x 106.

For initial mass spectrometer method development, high concentrations of derivatized THC and THC-D3 were injected and analyzed in electron ionization (EI) full scan to determine precursor masses for EIMRM. Methods were then created to measure each precursor ion’s product ion scan at various collision energies and collision pressures. From these product ion scans, the most intense ions were selected for each MRM transition at the optimum collision energy and collision cell pressure.

The transition from mass 386 to mass 303 was used as the quantitative transition for THC, with the transition from mass 371 to mass 289 as the confirming transition. For THC-D3, the quantitative and confirming transitions were mass 389 to 306 and mass 374 to 292 respectively.

Sample Processing and Result Derivation
For sample acquisition, peak detection, and quantitation,GC/MS reporting software was utilized. By incorporating all of the vital components of analysis into a unified workflow oriented application, the software provided an integrated solution for THCGC-MS/MS confirmation. To make use of the software for method validation, an instrument method was first created for the mass spectrometer, auto sampler, and GC instrument.

Following this, a master method was created within the software, including processing parameters for component identification and quantitation and quality control (QC) criteria specific to the method. Batch creation was performed through the software, which greatly simplified and streamlined sample entry, particularly for the longer validation batches. This highlights the applicability of this software to routine analysis of toxicological samples.

Concentration calculations were based on a three-point calibration at 0.2, 2, and 20ng/mL, using THC-D3 as the internal standard. All validation batches had to conform to QC criteria, including quantitative and qualitative bounds checking. Quantitative criteria for the batch included acceptable quantitation ranges for all samples in each batch. All calculated amounts for QC and calibration samples had to fall within ±20% of the expected concentration in order to accept the sample. Failure of a QC sample within a batch would mean the entire batch would need to be repeated. In addition to this quantitative window, negative controls were evaluated based on two additional criteria. One means of assessing a negative control is a quantitative value for THC less than the method limit of detection (LOD), which in this case was 0.2ng/mL. An alternate criterion for negative controls is that the calculated amount must be less than a pre-determined percentage of the method cutoff.

For this method, a level of 5% of the cutoff (0.1 ng/mL) was used as a second criterion and all negative controls were evaluated for compliance to both criteria. Qualitative criteria included ion ratio and retention time target ranges based on an average of the calibrators, along with peak shape considerations. These criteria were applied to all sample types.

Ratios were calculated for THC-D3 (292:306) and THC (289:303) and for each ratio an acceptable range of ±20% was established. Similarly, the target retention time for THC and THC-D3 was set using a ±2% retention time window based on an average of the calibrators’ retention times.

The analysis of THC in oral fluids using the triple quadrupole system was verified through the evaluation of linear range, carry over, and precision. The preparation and analysis of three separate batches was carried out, one for linearity/carry over and two for precision. Batch acceptability was determined by applying the previously described QC criteria. Carry over was assessed throughout the linearity study and precision analyses were performed on two separate days.

The assessment of assay linearity was performed at a varied range of concentrations. The linearity batch included an extracted standard, a negative control, the 0.2, 2, and 20ng/ml calibrator, a 40% control sample (8ng/ml), and 125% control sample (2.5ng/ml). In order to assess method linearity, samples at 0.2, 0.4, 0.8, 2, 4, 8, 20, and 40ng/mL were prepared alongside the calibrator and controls. Samples were then individually injected four times, with the consequent 32 data points quantified on the basis of the three point calibration. Without exception, all quantitative values were within +20% of their intended concentrations. A least square fit analysis comparing the average quantitative value for each level to the predicted value was revealed to have a correlation coefficient of 0.997 (Figure 1).

The correlation coefficient of the three calibrations used for all quantitations was confirmed as 1.0000. Chromatography for the quantitation ions and all qualifiers was excellent, as shown in Figure 2.

The results indicate that the analyzer, operated in MRM mode, is simultaneously sensitive and selective enough to accurately trace THC measurements in oral fluid. The GC-MS/MS system achieved superior quantitative precision, with coefficients of variation (CV) of 6% or less at 0.8 and 2.5ng/mL and exceptional linearity and accuracy from 0.2 to 20ng/mL with no significant carryover.

Assay linearity demonstrated between 0.2 and 20ng/ml THC in oral fluid.

Example of extracted ion profiles of THC quantitative and qualifier ions monitored at 0.2ng/ml.

The triple quadrupole operated in selected reaction monitoring mode proved to be both selective and sensitive enough to routinely measure THC in oral fluid at a 2ng/mL cutoff level. This was exemplified by the excellent accuracy at the 0.2ng/mL sample level analyzed during the linearity study, where all four injections at the level quantitated within 5% of the actual amount. The linearity study also demonstrated ample linear range for the assay, determined to be between 0.2 to 20ng/mL. Across this range, all samples also gave ion ratios which were within 20% of the ion ratios of the calibrator. Furthermore, the intra- and inter-day precision studies showed that the coefficients of variation for the assay at 0.8 and 2.5ng/mL were well under the 10% value required by many regulatory bodies. As instrument method development and validation were performed in an extracted oral fluid calibrating solution, the results demonstrate how the system for method validation would perform within a working laboratory.

The triple quadrupole method also offered ease of use and excellent speed of analysis relative to alternative approaches. Setup and daily use of this method was as easy as for a typical single quadrupole confirmation method, without requiring a complex multi-dimensional GC approach. In addition, the ease of developing an MRM confirmation method on the triple quadrupole system allows the user more flexibility in expanding to other confirmation assays that prove difficult to analyze on a GC single quadrupole instrument.At a retention time of less than six minutes, the methodology described offers a productive means for high throughput laboratories to confirm and quantitate the use of THC through oral fluid sampling.


  1. Baselt, Randell C. Disposition of Toxic Drugs and Chemicals in Man. Eighth Edition. Biomedical Publications. Foster City, California, 2008. pp 1513-1518
  2. Marijuana (Cannabis),, (2010, July)
  3. Moeller, Manfred R. and Kraemer, Thomas, “Drugs of Abuse Monitoring in Blood for Control of Driving under the Influence of Drugs,” 2002/04000/Drugs_of_Abuse_Monitoring_in_Blood_ for_Control_of.3.aspx


Eric Chi is Senior GC/GC-MS scientist for Thermo Fisher Scientific with experience in a variety of industries focusing on application methods development.

Jason Cole is a Product Manager for GC triple quadrupole technologies at Thermo Fisher Scientific. Jason has held previous roles with the company focusing on methods development as a GC and GC-MS Applications Scientist.


Share this Story

You may login with either your assigned username or your e-mail address.
The password field is case sensitive.