Correlation of Results From Cannabinoid Immunoassay and GCMS Analysis

A number of studies have been conducted to investigate how well results from cannabinoid immunoassays can correlate to GC/MS analysis and/or to select an appropriate cutoff value for each of the initial test methods (99-105). In all cases the general correlations exist, yet the data points could be rather scattered. Generally speaking, the correlation coefficients are more sensitive to the change of sample groups, in which the distributions in the relative concentrations of THC-COOH and other cross-reacting compounds varies.

The relative concentrations of THC metabolites in plasma and urine have been studied to determine if a temporal relationship could be estimated between marijuana use and metabolite excretion (65,69). With the addition of the P-glucuronidase hydrolysis step in the extraction protocol, the presence of significant quantities of THC and 11-OH-THC in urine could be demonstrated (69). The relative concentrations of THC-COOH and 11-OH-THC can be shown in a scatter plot when all data for urinary THC-COOH and 11-OH-THC concentrations published in the article by Manno et al. (69) were used to create the plot shown in Fig. 4. For samples with THC-COOH levels closely surrounding the 15 ng/mL cutoff, the relative cross-reactivities of an immunoassay with 11-OH-THC, THC-COOH, and their relative abundance may contribute to the immunoassay outcome by rendering the results false positive or false negative when compared to a fixed GC/MS value of THC-COOH.

In addition to the interindividual metabolism and metabolite variability, the correlation of immunoassay and GC/MS results can also be influenced by the total performance characteristics of not only the screening but also confirming techniques used (123-127). Because all analytical techniques have an acceptable range of imprecision, it is essential to note that a value generated from immunoassay or GC/MS analysis is

Thc Cooh Half Life
Fig. 4. Relative concentrations of THC-CODH and 11-OH-TCH in cannabinoids containing urine samples. (Adapted from data from ref. 65.)

Table 1

Examples of the AACC/CAP Forensic Urine Drug Testing (Confirmatory)

Survey Results

Table 1

Examples of the AACC/CAP Forensic Urine Drug Testing (Confirmatory)

Survey Results


Coefficient of

Low value

High value


No. labs


variation (%)



UDC-1, 2003
















UDC, 2002
























Data were obtained with permission from American Association for Clinical Chemistry/College of American Pathologists (AACC/CAP) forensic urine drug testing (confirmatory) Survey UDC-A of 2003 and Survey 2002 year-end critique for A9-THC-COOH.

Data were obtained with permission from American Association for Clinical Chemistry/College of American Pathologists (AACC/CAP) forensic urine drug testing (confirmatory) Survey UDC-A of 2003 and Survey 2002 year-end critique for A9-THC-COOH.

not an absolutely fixed number. These analytical techniques all have to be validated and meet a host of quality-control and quality-assurance requirements. Similar to the requirements for proper utilization of immunoassays, knowledge of the advantages and potential pitfalls of different GC/MS systems as well as ionization and detection modes would facilitate proper optimization for the accuracy of compound quantification and identification (124).

Because GC/MS involves multiple steps of extraction, derivatization, and quantitative analysis, the laboratory has to determine the acceptable criteria for replicate analysis. Generally, the repeatability and reproducibility of GC/MS in a certified laboratory are excellent, even though there are interlaboratory variabilities among the certified laboratories. For years, the College of American Pathologists and American Association for Clinical Chemistry have been conducting quarterly surveys and year-end critiques for all certified laboratories. The survey results of THC-COOH analysis for year-end 2002 and the first quarter of 2003 are listed in Table 1. The results are fairly consistent over the years, and the interlaboratory coefficient of variation has been approx 10-15%. Statistically, the variations may not significantly affect the confirmation of presumptive positives, even though the confirmation rate for near-cutoff samples can be more readily affected.

A semi-quantitative immunoassay produces a numerical concentration that approximates the total amount of THC-COOH along with associated metabolites in the specimen, namely, a value for apparent THC-COOH equivalent. The results of unknown clinical samples are calculated by the automatic analyzers based on a calibration curve. The calibration curve is calculated from prevalidated equations for the best-fit curve. The claimed concentrations of calibrators must be established by repeated

GC/MS analysis to ensure that the THC-COOH concentration in the calibrators stays within the acceptable range of GC/MS values for the entire duration of its shelf life.

Table 2 shows a collection of analytical recovery data or imprecision data from various package inserts of commercial immunoassays. The nominal THC-COOH concentration is the amount of THC-COOH compound spiked into urine for running the immunoassays, and the numerical value of apparent THC-COOH concentration is the average of replicate results obtained from the immunoassays.

In general, the results of semi-quantitative immunoassays provide an indication of the levels of THC metabolites to assist in making dilutions for GC/MS analysis. How closely a semi-quantitative immunoassay result can match the nominal value is affected by a number of factors, including the quantitative accuracy of calibrators, the quantitative accuracy of the spiked samples for evaluation, the constituents of the specimens, the assay precision for the lot of reagents used, and the assay dynamic range. The results may no longer be semi-quantitative in that the absorbance changes of the immunoassay flatten out or reach the plateau (128). Commonly used commercial immunoassays offer applications for multiple cutoff choices to meet the requirement of different drug-testing programs. Depending on the drug-testing program goals and preferences, the more frequently used cutoff concentrations for urinary cannab-inoid immunoassays are 20, 25, 50, and 100 ng/mL.

In a study designed to understand the relationship of THC concentrations in oral fluid and plasma after controlled administration of smoked cannabis, Heustis and Cone observed that results from an RIA selective for THC were higher than those obtained from GC/MS. The mean ± standard deviation ratio of RIA to GC/MS concentration was 3.35 ± 2.16, with a range of 1.1-8.8 (23). The higher estimated THC concentrations in oral fluid by the RIA screen method were attributed to cross-reactivities of the THC RIA antibody to other cannabis constituents. In this study, THC RIA concentrations at 0.2 hour were generally 20-fold or more than those measured at 0.27 hour. With a 1.0 ng/mL screening cutoff concentration, the mean detection times by RIA for the 1.75% and 3.55% doses were 5.7± 0.8 and 8.8 ± 8.3 hours, respectively. The authors also compared the excretion rates in three biological specimens from the same subject by GC/MS analysis of THC (for oral fluid and plasma) and THC-COOH (for urine) and reported half-life estimates of 0.8 hour for oral fluid, 0.9 hour for plasma, and 16.9 hours for urinary specimens.

Continue reading here: Stability of Cannabinoids in Biological Matrices

Was this article helpful?

0 0