| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Doctoral Training Program in Forensic Sci., Univ. of Alabama at Birmingham, Birmingham, AL 35294-2060
a author for correspondence: fax 205-934-2067, e-mail rliu{at}sbs.sbs.uab.edu
Current workplace drug-testing programs require reporting positive results only if the gas chromatography/mass spectrometry (GC/MS) results are at or above a cutoff concentration and the mandated initial immunoassay results are also at or above an identical or higher cutoff value. Because immunoassays are generally also responsive to compounds that are structurally related to the analyte, whereas GC/MS analysis is specific for individual compounds, the cutoff adopted for the immunoassay is generally set higher than that adopted for the GC/MS test. The corresponding immunoassay cutoff adopted for the immunoassay may be correlated to that adopted for the GC/MS test if: (a) the drug of abuse is a mixture of structurally related compounds or (b) multiple structurally related metabolites are produced as a result of drug exposure.
We have previously demonstrated that the apparent11-nor-
9-tetrahydrocannabinol-9-carboxylic
acid (9-THC-COOH) concentrations (derived from immunoassays) that
statistically correspond to a specific 9-THC-COOH concentration
(determined by GC/MS) varied with the immunoassay methodologies used
[1] and with the reagent manufacturing dates
[2]. In the current study, we evaluate
empirical data derived from two RIA kits based on the
same technology and available at the
same time period, but supplied by two
different manufacturers.
Analytical procedures for immunoassays and GC/MS were as previously described (2). Two data sets generated in routine testing with Immunalysis (Glendale, CA) and Abuscreen® (Roche, Branchburg, NJ) cannabinoid RIA kits were used. Data derived from truly negative samples (0 µg/L for both RIA and GC/MS tests) and from those samples with apparent 9-THC-COOH concentrations out of the dynamic ranges of the immunoassays were excluded from this correlation study, thereby permitting a comparison with the samples that are most likely to provide information for differentiation. Of a pool of ~13 200 specimens tested with Immunalysis, 24 had apparent 9-THC-COOH concentrations in the 50–130 µg/L range. Similarly, 88 of ~26 400 specimens tested with Abuscreen fell within the same range. Although significantly more data points were available in the Abuscreen data set, only the latest 24 data points, representing a total sample pool of ~10 300 specimens tested, were included in this comparative study.
Apparent analyte concentration data obtained with the two RIAs were
compared with the respective analyte concentrations determined by GC/MS
by using a linear model without forcing the resulting regression line
through the origin. These comparisons were done with
DeltaGraph® Pro 3 for Macintosh (Deltapoint,
Monterey, CA) on a Macintosh II-CX computer (Apple Computer, Cupertino,
CA). The resulting regression parameters are shown in Fig. 1
.
|
Data included in this study came from a diverse population; collectively, the samples contained all compounds resulting from the metabolic process after the intake of Cannabis products and thus reflected various different metabolic patterns and elapsed time after exposure. Because the RIA reagents used in this study respond (to different degrees) to many of these compounds, the apparent 9-THC-COOH concentration derived from the RIA screen test will be higher than that obtained in the GC/MS procedure. Thus, proper selection of the initial test cutoff will have direct consequences on the overall efficiency of the test program. An improperly low initial screen test cutoff may refer for GC/MS analysis many samples with a targeted analyte concentration below the GC/MS cutoff. This would, overall, increase analytical cost (low laboratory efficiency). Conversely, if the cutoff for the initial test is set too high, some samples that contain the targeted analyte concentration at or above the GC/MS cutoff may be reported as negative in the initial test, thus resulting in low program efficiency.
The apparent analyte concentration in the initial test that
statistically corresponds to a specific GC/MS analyte concentration can
be calculated by inserting the selected GC/MS value into the regression
equation derived from the comparison study. E.g., in the present study,
we used the regression equations to estimate apparent RIA
concentrations corresponding to two specific analyte (9-THC-COOH)
concentrations—15 and 30 µg/L. The statistically derived immunoassay
apparent analyte concentration values that best correspond to these two
GC/MS analyte concentrations are also shown in Fig. 1
.
Data for both immunoassays were collected in two periods of ~3 months each with use of several kits of various ages. Precision data (CV) obtained from 10 and 5 replicates of calibrator and controls, respectively, for both assays were comparable, mostly in the range 3–5%. The age of RIA reagents is known to affect the response slope of the assay's calibration curve and thus the assay's sensitivity. Differences in reagent ages, however, should not generate significantly different apparent concentration of analyte in a specific specimen.
Several general statements can be derived from the information provided
in Fig. 1
:
1) Statistically derived RIA apparent concentrations of 9-THC-COOH corresponding to a parent compound concentration of 15 µg/L are similar (69.2 vs 68.6 µg/L) for the two IRA kits tested. However, the respective RIA apparent concentrations corresponding to 30 µg/L 9-THC-COOH are substantially different (92 vs 83 µg/L).
2) The much larger difference of the apparent Immunalysis RIA concentrations corresponding to 30 and 15 µg/L 9-THC-COOH (92 and 69 µg/L vs 83 and 69 µg/L for the Abuscreen results) indicates greater sensitivity of the Immunalysis reagent toward the change of 9-THC-COOH concentration in test specimens.
3) That statistically derived apparent concentrations corresponding to 15 µg/L 9-THC-COOH (69 µg/L) for both RIAs are higher than the commonly used 50 µg/L cutoff indicates both reagent kits are maximized to avoid generating falsely negative results.
We also compared the Abuscreen results with those we previously reported, i.e., obtained with reagents manufactured at different dates (2). The statistically derived Abuscreen apparent analyte concentration corresponding to 30 µg/L 9-THC-COOH obtained here is statistically indistinguishable from that derived when using the reagent dated June 1989/May 1990. Also, the manufacturer reports the same cross-reactivity characteristics for the reagents used in these two studies.
Acknowledgments
We thank Environmental Chemical Corporation, Inc., of Birmingham, AL, from whom we obtained the data used for this study.
References
9-tetrahydrocannabinol-9-carboxylic acid in urine specimens. Forensic Sci Int 1991;49:43-56.
[ISI][Medline]
[Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
