| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Key Words: indexing terms: abused drugs • screening • gas chromatography–mass spectrometry • urine
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Methanol, methylene chloride, isopropyl alcohol, isooctane, dimethyl sulfoxide, ethyl acetate, hexane, glacial acetic acid, and methyl iodide were purchased from Fisher Chemical Co., Pittsburgh, PA, all ACS grade. KOH pellets and KH2PO4 were purchased from Mallinckrodt, Paris, KY. 1,1,1,3,3,3-Hexafluoro-2-propanol was purchased from Sigma Chemical Co., St. Louis, MO; pentafluoropropionic anhydride from Supelco, Bellefonte, PA; and tetramethylammonium hydroxide (2 mol/L) from Aldrich Chemical Co., Milwaukee, WI.
Bond-Elut Certify-LRC columns were purchased from Varian, Harbor City, CA. Phosphate buffers and KOH and HCl solutions were prepared as described in the Bond-Elut instruction manual.
For initial testing, Abuscreen OnLine Carboxy-THC 50 calibrator (chiral; carboxy-THC, 50 µg/L) and Abuscreen OnLine calibrator Level 3 (BZE, 300 µg/L) were purchased from Roche Diagnostic Systems, Branchburg, NJ. For GC-MS testing, BZE Calibrator (150 µg/L) and Carboxy-THC Calibrator (15 µg/L) were purchased from ElSohly Labs., Oxford, MS. Appropriate dilutions of the calibrators were made with blank urine.
immunoassay
Screening tests for COC and THC were done with Abuscreen OnLine
reagents; the subsequent light scatter was measured with an Olympus
AU5131 (Olympus Corp., Lake Success, NY) to follow the agglutination of
microparticles in solution. The degree of inhibition of agglutination
is proportional to the amount of analyte in the sample. The COC assay
detects BZE, the primary metabolite of COC in human urine; the THC
assay detects 11-nor-delta-9-THC-9-carboxylic acid (carboxy-THC).
Reagents were prepared according to the manufacturer's instructions
for using COC and THC on the AU5131. The instrument was calibrated in
the AB mode with a single-point calibrator, and the settings used were
those recommended by Roche. For the COC study, specimens that produced
readings 
100 vs the 300 µg/L calibrator were tested further by
GC-MS. Subsequent experiments determined that the 100 reading is equal
to a BZE concentration of ~60 µg/L.
Specimens with results 20 vs the OnLine Carboxy-THC 50 calibrator
were rescreened with the Cannabinoid 20 Assay from Behring (formerly
Syva), San Jose, CA. Reagents and calibrators were reconstituted
according to the manufacturer's instructions for use on the B10 COBAS
BIO (Roche Diagnostic Systems); instrument settings were those
recommended by Behring. Specimens that produced readings equal to or
greater than the 20 µg/L calibrator were considered to be positive
and were tested further by GC-MS.
extraction and derivatization
BZE was extracted from 5.0 mL of urine as described in the
Bond-Elut instruction manual but with 150 µg of
BZE-D3 internal standard added to each sample. This
was followed by derivatization with
hexafluoro-2-propanol/pentafluoropropionic anhydride and reconstitution
with toluene for analysis by GC-MS.
Carboxy-THC was extracted from 5.0 mL of urine as described in the Bond-Elut manual but with 25 ng of internal standard added to each sample. The extract was then methylated with tetramethylammonium hydroxide/dimethyl sulfoxide and methyl iodide and then reconstituted with isooctane for analysis by GC-MS.
gc-ms analysis
All GC-MS analyses were performed on a Hewlett-Packard (HP; Palo
Alto, CA) Model 5890 gas chromatograph interfaced with a Model 5970A
quadrupole mass spectrometer. Samples were injected with an HP7673
Automated Liquid Sampler. The gas chromatograph was equipped with a
12 m x 0.020 mm (i.d.) HP-1 capillary column (0.33 µm film
thickness). The mass spectrometer was operated in selected ion mode
(SIM). For BZE, GC column conditions were set to give retention times
between 4.5 and 5.5 min. SIM ions used to monitor BZE were
m/z 318, 272, and 439. SIM ions used to monitor
BZE-D3 were m/z 321 and 275. For carboxy-THC,
the GC column variables were set to give retention times between 6.0
and 8.0 min. The SIM ions used to monitor carboxy-THC were
m/z 313, 357, and 372; for carboxy-THC-D3, they
were m/z 316 and 375.
The limit of detection (mean + 3SD) was calculated by analyzing 20 blank urines; for BZE, the limit of detection was ~1 µg/L.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
). (No donor samples from the SAMHSA-regulated program were used
in this study.) Use of a 100 µg/L screening cutoff and a confirmation
cutoff of 50 µg/L designated 105 specimens as positive (2.1%). If
the confirmation test cutoff was lowered further to 30 µg/L, the
number of positives increased to 119 (2.4%), and using the method's
limit of detection (1 µg/L) as the confirmation test cutoff increased
the number of positives to 139 (2.8%). Soldin et al. (3)
previously had reported similar results in a smaller study done with
samples from an adolescent population.
|
In an additional study, done to determine whether the precision of the COC confirmation method was satisfactory at the 30 µg/L cutoff, several minor modifications were made to the method described in Materials and Methods. The 150 µg/L calibrator was diluted to 30 µg/L with blank urine, and the concentration of internal standard added was 30 µg/L. SAMHSA has set 40% of the cutoff concentration as a target for the lower limit of quantification. To see if this target could be obtained, 11 replicate samples (target concentration = 12 µg/L) were analyzed. The resulting mean and CV were 12.5 µg/L and 5.6%, respectively. Moreover, the chromatography was acceptable, showing no interfering peaks and adequate abundances for each peak.
The increase in the number of positives for COC obtained by first
lowering the initial testing and then lowering the confirmation cutoffs
was very important. Using the screening cutoff of 300 µg/L instead of
100 µg/L would have missed 80 COC-positive samples, all of which were
subsequently confirmed by GC-MS. In addition, 21 samples with screening
readings of 100 µg/L were negative by GC-MS—primarily because the
routine COC method was calibrated with a 300 µg/L calibrator.
Clearly, the discrimination of positives and negatives would be
considerably improved if the OnLine reagents and instrument settings
were optimized for a cutoff considerably less than the 300 µg/L
currently in use. This could be achieved by minor modifications in the
reagent or instrument settings, e.g., increasing the sample volume
(personal communication, Alan McNally, Roche
Diagnostic Systems).
Most importantly, the increases in the positive rate shown in Table 1
were accomplished with only minor changes in the confirmation method.
Calibrating the current GC-MS method at 30 µg/L presented no problem.
The GC-MS was run with the electron multiplier set at 1600 eV and an
injection volume of 1.5 µL. Under these conditions, the control at
40% of cutoff (12 µg/L) produced acceptable chromatography.
No special instrument maintenance was done for this study. The samples were injected in the same manner as the routine batches. Most likely, other forensic toxicology laboratories also are capable of achieving similar results with little or no alterations in their current methodologies. This approach would appear to be a less cumbersome way to increase the positive rate for COC than a recent method that analyzes for ecgonine (4).
Lowering the BZE cutoff increases the concern that positive results might be caused by passive inhalation or occupational exposure. A recent study by Cone et al. (5) obtained a urine concentration of BZE as high as 123 µg/L in one subject exposed to vapor of free base cocaine for 1 h. The maximum urine BZE concentration for aides assisting in that study and who were exposed to sidestream smoke for 4 h was 6 µg/L. More intense exposure to smoke from crack cocaine might lead to higher BZE concentrations in the urine. Whether one could unwillingly expose oneself to this degree is doubtful. Le et al. (6) previously reported BZE concentrations >1000 µg/L in the urine of criminalists who were handling or analyzing large amounts of powder cocaine while conducting criminal investigations.
marijuana
In all, 6427 urines were screened for THC at a 50 µg/L cutoff
with the OnLine reagent. Specimens giving readings 20 µg/L were
rescreened with the Cannabinoid 20 Emit Assay. Specimens that still
gave readings equal to or greater than the 20 µg/L calibrator were
tested further by GC-MS. The number of positives reported for the 50
µg/L screening cutoff and the 10 µg/L confirmation cutoff was 179
(2.8%) (Table 2
). Using a screening cutoff of 20 µg/L (and keeping the
confirmation cutoff at 10 µg/L) increased the number of positives to
262 (4.1%). If, in addition, the confirmation test cutoff was lowered
to 5 µg/L, the number of positives increased slightly, to 281
(4.4%). As described in Materials and Methods, calibration
at 5 µg/L was accomplished by making only minor changes in the
routine method. The chromatography was acceptable, with no interfering
peaks, and the peak areas were adequate for quantification. Additional
studies done by diluting the low-concentration control demonstrated
accurate quantification at 1 µg/L (electron multiplier set at 2000
eV, injection volume = 2 µL).
|
Lowering the initial test cutoff to 20 µg/L greatly increased the number of positives. The Emit Cannabinoid 20 assay used in this study demonstrated excellent discernment between negatives and positives. Only 1 of 104 specimens that had positive initial test results failed to be confirmed positive at 5 µg/L. Recently, an application of the OnLine reagent at the 20 µg/L cutoff has been published (7). However, the problem of differentiating between marijuana use and passive inhalation becomes of greater concern when the lower cutoffs are used. Several clinical studies, now ~10 years old, have attempted to evaluate the effects of passive inhalation. Cone et al. (8)(9), Mulé et al. (10), and Perez-Reyes et al. (11) all demonstrated that detectable amounts of cannabinoids can be produced in urines of volunteers purposely exposed to substantial amounts of marijuana smoke. Whether an individual could knowingly and unwillingly expose him- or herself to such amounts is highly questionable.
An earlier study of Huestis et al. (12) found that lowering the initial screening cutoff to 50 µg/L from 100 µg/L resulted in a large increase in the number of THC positives. Their study was done on urines collected from six volunteers and the percentage increase varied from 23% to 53%, depending on which of the eight immunoassay reagents was used. In the present large-scale study, looking at results from 6427 urines tested in a production laboratory, lowering the initial test cutoff even further to 20 µg/L clearly demonstrates substantial additional increases in the number of positives confirmed by GC-MS.
concluding comments
Currently, numerous urine specimens containing low concentrations
of COC metabolites or THC metabolites are being reported as negative.
This supports the findings of Huestis et al. (13), who
showed that detection times with the 50 µg/L THC cutoff provide only
a 1–2 day window of detection for acute users. Lowering the cutoffs
for both drugs should increase the window of detection and permit a
greater percentage of specimens to be confirmed as true positives at
concentrations as low as 50 µg/L for BZE and 10 µg/L for
carboxy-THC.
Clearly, each laboratory individually must address the issue of increased cost of the additional confirmation required because of using the lower screening cutoff values. Ultimately, these costs will be passed on to the consumers—clients who would benefit from avoiding costs generated by problem employees. This justification for doing drug of abuse testing has been made many times in the past decade. Perhaps it is time to reevaluate the cutoff values used to report these results in light of the technology now available.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 Nonstandard abbreviations: SAMHSA, Substance Abuse and Mental Health Services Administration; COC, cocaine; THC, tetrahydrocannabinol (marijuana); D, deuterium; BZE, benzoylecgonine; HP, Hewlett-Packard; and SIM, selected ion monitoring. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  V. I. Luzzi, A. N. Saunders, J. W. Koenig, J. Turk, S. F. Lo, U. C. Garg, and D. J. Dietzen Analytic Performance of Immunoassays for Drugs of Abuse Below Established Cutoff Values Clin. Chem., April 1, 2004; 50(4): 717 - 722. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
