Eur Arch Otorhinolaryngol (2002) 259 : 279–284DOI 10.1007/s00405-002-0456-z L.Guarda-Nardini · R. Tito · A. Staffieri · A. Beltrame Treatment of patients with arthrosis of the temporomandibular joint by infiltration of sodium hyaluronate: a preliminary study Received: 20 June 2001 / Accepted: 25 January 2002 / Published online: 24 April 2002 © Springer-Verlag 2002
Jat-niedbalaJournal of Analytical Toxicology, Vol. 25, January/February 2001 Immunoassay for Detection of Cocaine/Metabolites in
R. Sam Niedbala1,*, Keith Kardos1, Tiffany Fries1, Alana Cannon1, and Al Davis2
1STC Technologies, Inc., Bethlehem, Pennsylvania 18018-1799 and 2LabOne, Inc., 10101 Renner Boulevard,
Lenexa, Kansas 66219-9752
gual glands, minor mucous glands, gingival crevicular fluid, and cellular debris. In addition, oral fluids may contain bacteria, There is growing interest in the use of alternate biological fluids
nasal secretions, and residues of ingested fluids and substances.
for drug testing. An advantage of oral fluids is that collection can
The primary contributors to this mixture are saliva secretions be made from individuals under direct observation without undue
from the various glands. Hence, many authors have used other embarrassment or invasion of privacy. This study evaluated the
terms for oral fluids such as "saliva" and "mixed saliva" (1,2).
STC Cocaine Metabolite MICRO-PLATE EIA for use in detection
The composition of oral fluids can change rapidly in response of cocaine and metabolites in oral fluids. Intra- and interassay
to stimulation. The parotid gland normally contributes about precision of the EIA was < 10%. The EIA was cross-reactive to
20% of the total volume of unstimulated saliva. At high stim- benzoylecgonine (100%), cocaine (> 12.9%), and cocaethylene
ulated flow rates, the parotid gland becomes the predominant (13.8%), but did not demonstrate detectable cross-reactivity with
secretor, and its contribution can rise to 50% (3).
other commonly encountered medicants. Evaluation of a series of
Cocaine and many other drugs of abuse have been measured potential adulterants of oral fluids indicated that common
in varying concentrations in oral fluids following different household chemicals and foodstuffs did not alter the outcome of
EIA testing for cocaine metabolite. Analysis by EIA and by gas
routes of administration. Several reviews regarding the occur- chromatography–mass spectrometry (GC–MS) of oral fluids and
rence of drugs of abuse in oral fluids, analysis, and the con- urine specimens collected from current drug users in treatment
centration relationships between "saliva" and plasma have been programs and subjects participating in research studies involving
published (4–10). Disposition of cocaine into oral fluids ap- controlled dosing of cocaine provided assessment of the clinical
pears to occur through a combination of mechanisms including sensitivity and specificity of the STC Cocaine Metabolite EIA.
(1) passive diffusion from blood across acinar cell membranes Analysis of the data by means of receiver operating characteristic
into saliva ducts followed by secretion into the mouth and (2) (ROC) plot indicated that the optimal cutoff concentration for the
direct deposition and sequestration in the oral cavity during oral fluids EIA was 10 ng/mL. In comparison to GC–MS (10-ng/mL
drug administration by oral, smoked, and intranasal routes.
combined cutoff concentration for cocaine and benzoylecgonine),
The concentrations of cocaine and metabolites (benzoylecgo- the EIA (10-ng/mL cutoff concentration) demonstrated a sensitivity,
nine and ecgonine methyl ester) deposited into saliva and sub- specificity, and accuracy of 95%, 82%, and 88%, respectively.
The oral fluids EIA was slightly less sensitive than the urine
sequently into oral fluids by passive diffusion are determined by EIA (300-ng/mL cutoff concentration) for the detection of
their solubility, pKa, degree of plasma protein binding, and cocaine metabolite with a sensitivity, specificity, and accuracy
saliva flow rate (11). To enter saliva by passive diffusion, drugs of 73%, 85%, and 88%, respectively. Overall, testing of oral
must be lipid-soluble, non-ionized, and unbound (free fraction fluids for cocaine metabolite with the STC Cocaine Metabolite
of drug in plasma). Because of these factors, cocaine is fre- MICRO-PLATE EIA appears to offer a viable alternative to urine
quently detected in saliva in higher concentrations than found for detection of recent cocaine use.
in simultaneously collected plasma specimens (12). Ben-zoylecgonine and ecgonine methyl ester are less lipid-solublethan cocaine and are generally found in lower concentrationsthan found in simultaneously collected plasma specimens. In addition to passive diffusion from blood, direct deposition of co-caine into oral fluids occurs when cocaine is administered by Oral fluid is a complex mixture of fluids and solids origi- the smoking and intranasal routes. Shortly after ingestion, the nating from the parotid glands, submandibular glands, sublin- amount of cocaine present in oral fluids as a result of direct de-position and sequestration can greatly exceed the amount con- * Author to whom reprint requests should be sent.
tributed by passive diffusion (13). However, cocaine deposition Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
Journal of Analytical Toxicology, Vol. 25, January/February 2001 into oral fluid through this mechanism clears rapidly in 1–2 h plastic specimen vial containing a preservative solution. The after administration. Thereafter, "saliva"/plasma ratios were fiber pad is impregnated with a mixture of salts and gelatin comparable to those obtained after intravenous cocaine ad- which creates a hypertonic environment and an increase in osmotic pressure wherever it contacts oral mucosal cells. The Methods for the detection of cocaine and metabolites in oral pad was placed between the lower gum and cheek during col- fluids have become a topic of considerable interest because of lection of oral fluids. Following a collection period of approxi- the inherent advantages over other biological specimens. The mately 2.0 min, the pad was placed into the vial containing the single, most important advantage in the use of oral fluids is in preservative solution, and the vial was sealed. Specimens were the ease of collection. Specimens can generally be collected in stored at room temperature and shipped to the laboratory for a matter of minutes under direct observation and without em- barrassment to the donor. An additional advantage may also be The clinical sensitivity and specificity of the EIA were deter- that adulteration of oral fluids is likely to be more difficult mined by testing specimens from chronic cocaine users as well than adulteration of urine specimens. Also, detection times for as volunteer subjects receiving single doses of cocaine. In- cocaine and/or benzoylecgonine may be longer than expected formed consent was obtained from all subjects. Specimens based upon estimated half-lives of approximately 1 h and 4.5 h were collected from cocaine-experienced subjects participating for cocaine and benzoylecgonine, respectively. For example, in research at Johns Hopkins School of Medicine, Baltimore, Cone and Weddington (14) reported detection of cocaine in MD (99 subjects); patients in treatment at the Hooper Detoxi- oral fluids by a sensitive immunoassay method for 5–10 days.
fication Center, Portland, OR (50 subjects); and research vol- Benzoylecgonine has also been reported to be detectable in unteers at the National Institute on Drug Abuse, Intramural oral fluids for 24 h (15). Consequently, the use of oral fluids for Research Program, Baltimore, MD (10 subjects). In the first routine detection of recent cocaine use may represent a feasible study (Study 1), a total of 149 matched Intercept oral fluid and alternative to urine testing.
urine specimens were collected randomly from cocaine-expe- The goal of the present study was the validation of a com- rienced subjects (N = 99) and patients when visiting the clinic mercial immunoassay method for detection of cocaine and for treatment (N = 50). Oral fluid specimens were analyzed metabolites in oral fluids. Specimens were collected from drug- using the STC EIA at the Gulf Coast Regional Blood Center, free human subjects and from cocaine users with the Inter- Houston, TX, and all urine specimens were analyzed by GC–MS.
cept™ Oral Specimen Collection Device (Epitope, Inc., In the second study (Study 2), 163 oral fluid specimens and 156 Beaverton, OR). This device consists of a treated, absorbent, matching urine specimens were collected in a controlled-dosing cotton-fiber pad affixed to a nylon stick. The collection pad is study in which 10 subjects were administered single, intra- impregnated with a mixture of salts and gelatin that creates an venous doses of cocaine. Timed oral fluid specimens were col- increased osmotic pressure when placed in contact with oral lected over 131 h by expectoration into plastic vials. Urine mucosal cells. Placement of the collection pad in contact with specimens were collected at each time point unless the subject the gingival mucosa between the lower gum and cheek en- could not void a specimen. Saliva flow was stimulated by use of hances the flow of oral fluids including mucosal transudate sour candy. All specimens were frozen until time of analysis.
onto the absorptive pad. After several minutes, the collection Aliquots of oral fluid specimens were applied to Intercept col- pad was placed into a vial containing a buffer solution. The pad lection devices and analyzed by the STC EIA and GC–MS at STC collects approximately 400 µL of oral fluid, which is diluted into Diagnostics. Urine specimens were screened by EMIT using a 800 µL of buffer in the transport tube. Specimens were tested 300-ng/mL cutoff and confirmed by GC–MS using a 150-ng/mL for cocaine metabolites with a solid-phase, competitive enzyme immunoassay (EIA). Analytical precision, sensitivity, and speci-ficity were determined for the EIA system. Clinical sensitivity and specificity of the EIA were determined by comparison of re- The STC Cocaine Metabolite MICRO-PLATE EIA (STC Diag- sults to analysis by gas chromatography–mass spectrometry nostics) is a competitive immunoassay for the detection of co- caine and cocaine metabolites in oral fluid collected with theIntercept collection device. To each microplate well, 50 mL ofsample eluate (N = 3), control, or calibrator was added alongwith 50 µL of labeled enzyme and allowed to incubate at room Materials and Methods
temperature in the dark for 30 min. After competitive bindingof cocaine metabolite and enzyme-labeled hapten to the anti- body fixed onto the EIA well, the wells were washed six times All buffer and reagent chemicals were obtained from the with deionized water. Substrate reagent (100 µL, 3,3',5,5'- Sigma/Aldrich Chemical Co. (St. Louis, MO). Test kits were tetramethylbenzidine) was added, and the microplate was again supplied by STC Technologies, Inc. (Bethlehem, PA).
incubated at room temperature in the dark for 30 min. The re-action was then stopped by adding 100 µL of stopping reagent Specimen collection and storage
(1M sulfuric acid), and the color was measured at 450 nm with Oral fluids were collected from human subjects with the In- a Bio-Tek EL312 reader. The absorbance measured was in- tercept Oral Specimen Collection Device. This device consists versely proportional to the quantity of cocaine and cocaine of an absorbent, cotton-fiber pad affixed to a nylon stick and a metabolite in the specimen. The calibrators (N = 3) in the assay Journal of Analytical Toxicology, Vol. 25, January/February 2001 kit were included on every plate and consisted of oral fluid Accuracy = (TP+TN)/(TP+TN+FP+FN) control buffered matrix containing benzoylecgonine at 0, 5,10, and 50 ng/mL.
Receiver operating characteristic (ROC) plot
The optimum cutoff concentration for the STC Cocaine Metabolite MICRO-PLATE EIA was determined by use of an Oral fluid eluates were analyzed for cocaine, ecgonine methyl ROC plot (16). EIA results (5-, 10-, 20-, and 50-ng/mL cutoff ester, and benzoylecgonine by GC–MS with selective-ion mon- concentrations) of oral fluid specimens obtained in the con- itoring (a model 5890A GC coupled with a model 5970 mass se- trolled-dosing study (N = 163) were compared to GC–MS re- lective detector, Hewlett-Packard, Palo Alto, CA). A J&W, sults (10-ng/mL cutoff concentration, sum of cocaine and 0.25-mm o.d., 3-m column was used for the GC. One milliliter benzoylecgonine). GC–MS results were considered to be accu- of each specimen was treated with deuterated analogues of co- rate. The EIA ROC plot was constructed by plotting sensitivity caine, ecgonine methyl ester, and benzoylecgonine (25 ng/mL) versus 1 – specificity.
and extracted with solid-phase columns (#ZSDAU020, WorldWide Monitoring). Ecgonine methyl ester and benzoylecgo-nine were derivatized with pentafluoropropionic anhydride inthe presence of pentafluoropropanol. The following ions were monitored for each analyte: cocaine, m/z 182, 272, 303; co-caine-d3, m/z 185, 275, 306; ecgonine methyl ester, m/z 182, EIA limit of detection (LOD)
314, 345; ecgonine methyl ester-d3, m/z 185, 317, 348; ben- The LOD of the STC Cocaine Metabolite MICRO-PLATE EIA zoylecgonine, m/z 300, 316, 421; and benzoyecgoine-d3, m/z was defined as the mean absorbance (A0) at zero-drug concen- 303, 319, 424. The approximate retention times were as follows: tration minus three times the noise (SD) (LOD = A0 – 3 SD) ecgonine methyl ester, 3.4 min; cocaine, 6.2 min; and ben- (17). The detection limit for benzoylecgonine was determined zoylecgonine, 5.7 min. The limit of quantitation (LOQ) was by obtaining the average absorbance value for 12 drug-free In- defined as the lowest concentration at which 100% of the di- tercept devices and then subtracting 3 SDs of the average. The luted standards yielded all qualifier ions within ± 20% and re- absorbance value minus three SDs was then extrapolated from tention times within ± 2% of the target concentration. The the log-log standard curve and represents the sensitivity of the LOQs (N = 6 replicates) for cocaine and benzoylecgonine in oral assay. The minimum detectable concentration for benzoylec- fluids were 4.0 ng/mL and 7.0 ng/mL, respectively. The intra- gonine was 0.95 ng/mL.
assay precision was cocaine (11%), benzoylecgonine (2%), andecgonine methyl ester (10%).
Precision of EIA
The precision of the STC Cocaine Metabolite MICRO-PLATE Sensitivity, specificity, and accuracy
EIA was assessed by the addition of standard concentrations of The number of true positives (TP), true negatives (TN), false benzoylecgonine to the Intercept preservative solution followed positives (FP), and false negatives (FN) was determined at spec- by assaying over a period of 20 days. The intra-assay precision ified cutoff concentrations by comparison of the results from was determined by analyzing each concentration 16 times per analysis of clinical specimens with the following assays: (1) run for 4 runs in 1 day. Interassay precision was determined by STC Cocaine Metabolite MICRO-PLATE EIA for cocaine and analyzing 2 samples at each concentration in 2 separate runs cocaine metabolites in oral fluids (cutoff concentration = 10 per day for 20 days. The results of testing are listed in Table I.
ng/mL benzoylecgonine equivalents); (2) STC Cocaine Metabo-lite MICRO-PLATE EIA for cocaine metabolites in urine (cutoff Specificity of EIA
concentration = 300 ng/mL benzoylecgonine equivalents); (3) Compounds that demonstrated no cross-reactivity at con- GC–MS analysis of cocaine and benzoylecgonine in oral fluids centrations of 10,000 ng/mL with the STC Cocaine Metabolite (cutoff concentration = 10 ng/mL of combined total concen- MICRO-PLATE EIA are listed in Table II. The cross-reactivity of tration of cocaine and benzoylecgonine); and (4) GC–MS anal-ysis of benzoylecgonine in urine (cutoff concentration = 150ng/mL). In comparisons of EIA results with GC–MS, a result Table I. Precision of the STC Cocaine Metabolite EIA
was considered a true positive or a true negative if both resultswere in agreement. A result in which the EIA was positive and the GC–MS result was below the cutoff calibrator was consid- %CV* (N = 64)†
%CV (N = 80, 20 days)‡
ered a false positive. A result in which the EIA was negative and the GC–MS result was equal to or greater than the cutoff cali- brator was considered a false negative. In comparisons of anal- ysis of oral fluids by EIA to analysis of urine specimens by EIA, the result of the urine assay was arbitrarily designated as thecorrect result. Sensitivity, specificity, and accuracy were calcu- * CV, coefficient of variation.
† N = 64 indicates that 16 devices were analyzed on one day at each concentration lated as follows: Sensitivity = TP/(TP+FN) ‡ N = 80, 20 days indicates that 2 devices at each concentration were analyzed × 2 runs/day × 20 days.
Specificity = TN/(TN+FP) Journal of Analytical Toxicology, Vol. 25, January/February 2001 cocaine and related metabolites compared to benzoylecgonine Stability of benzoylecgonine in stored specimens
are listed in Table III.
Oral fluid specimens were collected from 15 self-reported cocaine-negative subjects with the Intercept collection device.
Oral fluid eluates were combined into a total of 12 specimen Table II. Compounds Exhibiting No Cross-Reactivity with
pools. Each pool was confirmed to be free of cocaine metabolite the STC Cocaine Metabolite EIA when Tested at 10,000-
by EIA. Benzoylecgonine was added to achieve final concen-trations of 0, 10, and 50 ng/mL in each of four pools. Specimens Acetylsalicylic acid were stored at –20°C, 4°C, 23°C, or 37°C for 21 days. In addi- tion, a portion of each subject pool was added to specimen pads and stored at –20°C, 4°C, 23°C, or 37°C for 21 days. Elu- ates were tested in the STC Cocaine Metabolite MICRO-PLATE EIA (10-ng/mL benzoylecgonine cutoff concentration) at 0, 7, 14, and 21 days. No qualitative changes in specimen responses were observed during the 21-day storage period for the speci- mens containing 0 and 50 ng/mL benzoylecgonine. The signal strengths (absorbance) of the specimens containing 10 ng/mL benzoylecgonine were subjected to an analysis of variance to de- termine if significant changes (p < 0.05) had occurred over the 21 days of storage. No significant changes in absorbance had Adulteration testing of EIA
Common foodstuffs and household chemicals were added to Intercept specimens in the absence and presence of ben- zoylecgonine (final eluate concentration of 20 ng/mL) and Naphthalene Trimipramine tested with the STC Cocaine Metabolite MICRO-PLATE EIA (10-ng/mL benzoylecgonine cutoff concentration). As shown in Table IV, there were no false positives with any of the productsin the absence of benzoylecgonine. All specimens tested positive Table III. Cross-Reactivity of Cocaine and Metabolites
when benzoylecgonine was present.
with the STC Cocaine Metabolite EIA
Effect of specimen pH on EIA
Oral fluid specimens were collected from self-reported co- caine-negative subjects with the Intercept collection device.
Specimens were pooled and confirmed negative by EIA. All negative specimen pools were combined into a single spec- imen pool that had a pH of 7.03. This pool was divided into 10 aliquots. The pH of nine individual aliquots was adjusted with Ecgonine methyl ester NaOH or HCl to the following pH levels: 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, and 9.0. The pH of one aliquot was Table IV. Effect of Potential Adulterant on Response of the STC Cocaine
not adjusted. Each aliquot was further divided into three equal subaliquots. Benzoylecgoninewas added to achieve final concentrations of 0, Final EIA
10, and 50 ng/mL in each aliquot. The unad- justed pH sample was tested in duplicate, and the remaining samples were tested in triplicate with the STC Cocaine Metabolite MICRO-PLATE EIA (10-ng/mL benzoylecgonine cutoff Domino Pure Cane Sugar concentration). At all pH levels, specimens P&G Crest (original) without benzoylecgonine tested negative, and specimens containing 50 ng/mL of benzoylec- Tropicana 100% Pure gonine tested positive. Specimens containing 10 ng/mL benzoylecgonine tested positive at Nyquil Adult Nighttime all pH levels except 7.5. The pH 7.5 specimen tested negative but was within 0.01 absorbance units of the 10-ng/mL cutoff calibrator sample.
The percent coefficients of variation (%CV) for * Benzoylecgonine (BZE) was added to provide a final eluate concentration of 20 ng/mL.
responses across the entire pH range tested Journal of Analytical Toxicology, Vol. 25, January/February 2001 for specimens containing benzoylecgonine at 0, 10, and 50 benzoylecgonine. GC–MS concentrations of cocaine and ben- ng/mL were 10.2%, 7.3%, and 6.9%, respectively.
zoylecgonine in each oral fluid specimen were summed forcomparisons to other assays. For example, an oral fluid spec- Clinical sensitivity and specificity
imen that contained 6 ng/mL of cocaine and 5 ng/mL of Oral fluid specimens collected from cocaine users (combined benzoylecgonine by GC–MS was considered positive at a cutoff N = 312 specimens) were analyzed with the STC Cocaine concentration of 10 ng/mL. However, it should be noted that Metabolite MICRO-PLATE EIA and by GC–MS for cocaine and for some specimens, the values were below the LOQ of themethod. Therefore, the result is an estimation of concentration.
Table V. Clinical Sensitivity and Specificity of the STC
Matched urine specimens (combined N = 305) were analyzed Cocaine Metabolite EIA for Cocaine and Cocaine
with the STC Cocaine Metabolite MICRO-PLATE EIA and by Metabolites in Oral Fluids and the STC Cocaine
GC–MS for benzoylecgonine. The sensitivity and specificity of Metabolite EIA for Cocaine Metabolites in Urine
the STC Cocaine Metabolite MICRO-PLATE EIA for analysis oforal fluids and urine is tabulated in Table V. Comparisons were Oral fluids, GC–MS
made between the oral fluids EIA and GC–MS, between oral (10 ng/mL total concentration)
fluids EIA and urine EIA, and between the urine EIA and Oral fluids (N = 312) ROC curve analysis
Figure 1 illustrates the ROC curve analysis for 163 oral fluid specimens by the STC Cocaine Metabolite MICRO-PLATE EIAcompared to GC–MS results (sum of cocaine and benzoylecgo- Urine, EIA (300 ng/mL)
nine concentrations). The GC–MS results were considered to be an accurate indicator of the true presence or absence of cocaine Oral fluids (N = 305) and benzoylecgonine. From the ROC curve analysis, the max- imum sensitivity and specificity were obtained at a cutoff con- centration of 10 ng/mL. The false-negative rate at 10 ng/mL was 4/163 (2%), but rose rapidly at higher cutoff concentrations (e.g., 21/163 [13%]) at 20 ng/mL. At 10 ng/mL, the sensitivitywas 96% and specificity was 87%.
Urine, GC–MS (150 ng/mL)
Urine (N = 305) Immunoassays have become a mainstay in forensic testing for drugs of abuse because of their ability to be performed rapidly, accurately, and inexpensively with automatedequipment. They are a requirement of theUnited States Department of Health andHuman Services (DHHS) Mandatory Guide-lines for Federal Workplace Drug Testing Pro-grams, which states that "[the] initial test shalluse an immunoassay which meets the re-quirements of the Food and Drug Administra-tion for commercial distribution" (18). Thechief function of an immunoassay is to accu-rately identify negative specimens that con-tain no drug or drug at a concentration belowan administrative cutoff level. Specimens thatinitially test positive by immunoassay (pre-sumptive positives) then become the focus offurther testing. A natural result of this processis that negative specimens are eliminated frommore time-consuming, expensive, analyticalprocedures. In addition, positive specimens Figure 1. ROC plot for the STC Cocaine Metabolite MICRO-PLATE EIA compared to GC–MS anal-
are subjected to a minimum of two different ysis of 163 oral fluid specimens. The data points for each cutoff value (5, 10, 20, and 50 ng/mL) test procedures that greatly improve the ac- are indicated on the plot.
curacy of the test result.
Journal of Analytical Toxicology, Vol. 25, January/February 2001 Although the federal workplace drug-testing program in the curacy rate of 77% was also obtained in comparison of EIA U.S. is currently designed for testing only urine, considerable analysis of oral fluids with EIA analysis of simultaneously col- interest has evolved over the last decade in the use of alternate lected urine specimens. These results are quite comparable to biological specimens (e.g., oral fluids, sweat, and hair speci- an accuracy rate of 90% obtained by comparison of EIA analysis mens) (19). In particular, oral fluids testing for cocaine and of urine specimens to GC–MS analysis.
cocaine metabolites offers specific advantages over urine testing Although dilution and adulteration of urine specimens are in that the collection procedure should be easier and less em- common problems in urine testing (21–23), the prospect of barrassing and the specimen should be more difficult to tamper adulteration of oral fluids is unclear. Flushing and dilution may with, substitute, or adulterate. Direct observation of urine col- sometimes be effective in the reduction of cocaine metabolites lection is expressly forbidden in the DHHS program unless in urine below detection limits (23), but this approach is not "…there is reason to believe that a particular donor may alter likely to be effective in oral fluids testing. Excessive water in- or substitute the specimen…" (18). In contrast, direct obser- gestion is not likely to affect cocaine analyte concentrations in vation of oral fluids collection by the collector should not be plasma or in oral fluids. Other commonly employed methods of considered an invasion of privacy by the donor. Consequently, adulteration include the addition of a toxic substance directly to the integrity of specimens and chain-of-custody procedures the specimen. For oral fluids collection, some individuals may should be considerably easier to ensure with oral fluids than attempt to adulterate the specimen by holding liquids in their urine collection. Another potential advantage of oral fluid mouths. This opportunity to adulterate an oral fluid specimen testing for cocaine could be realized in testing programs that could be substantially reduced by rinsing the oral cavity with are directed toward "fitness for duty." Because cocaine and co- water prior to collection, instituting a brief waiting period be- caine metabolite concentrations in oral fluids frequently fore collection, and by observing the collection process. Also, demonstrate direct relationships to blood concentrations (12), evaluation of different household products and foodstuffs is positive findings can be interpreted as evidence of recent drug needed to determine if their presence in the oral cavity influ- use. Some inference might even be made regarding residual, ences test outcome. An examination of several common prod- psychoactive drug effects from recent cocaine use on the basis ucts (Table IV) indicated that their presence did not produce of a confirmed oral fluid test result. However, some prior studies either false-positive results or false-negative results in the oral have shown cocaine to be detectable for 5 to 10 days in chronic users (14). Thus, any conclusions regarding recent use fromoral fluids must be studied further.
Development of an EIA for detection of cocaine and metabo- lites in a new biological matrix such as oral fluids requires con- sideration of similarities and differences between the newbiological matrix compared to existing urine-screening assays.
Analysis of oral fluids by the STC Cocaine Metabolite MICRO- Although benzoylecgonine is the target analyte in urine testing, PLATE EIA provided a rapid and sensitive method for the de- oral fluids may contain a combination of cocaine, benzoylec- tection of cocaine and cocaine metabolites. Data from method gonine, and ecgonine methyl ester. Cocaethylene, an active validation showed acceptable intra- and interassay precision metabolite of cocaine and ethanol, could also be present. With and accuracy. The EIA demonstrated cross-reactivity to ben- the exception of ecgonine methyl ester, the presence of these an- zoylecgonine and cocaine sufficient for detection of cocaine alytes in combination or individually could be detected by the use with an accuracy equivalent to that obtained by urinalysis.
STC Cocaine MICRO-PLATE EIA. Consequently, GC–MS con- The assay was highly selective, and no cross-reactivity was firmation procedures for each of these analytes may be neces- demonstrated with other common medications. An evaluation sary. After very recent use (1–2 h), cocaine may be the of potential adulterants showed no interference in EIA test out- predominant analyte present in oral fluids, whereas after 2 h or come. In a study of oral fluid and urine specimens simultane- with multiple dosing, benzoylecgonine may be predominant ously collected from current cocaine users, the EIA was slightly (15,20). Concentrations of cocaine and metabolites in oral fluids less sensitive than urine (73% sensitivity). Overall clinical sen- will generally be lower than concentrations found in urine sitivity and specificity of the EIA compared to GC–MS were making lower EIA cutoff concentrations necessary on an oper- 95% and 82%, respectively. The performance of the oral fluids ational basis. For example, in the present study, the STC Co- EIA in combination with GC–MS confirmation suggest that caine Metabolite MICRO-PLATE EIA performed optimally at 10 oral fluids testing may be an attractive alternative to urine ng/mL for detection of cocaine use as compared to 300 ng/mL testing for cocaine use.
An oral fluids EIA for detection of cocaine use should gener- ally exhibit comparable clinical sensitivity, specificity, and ac-curacy to urine testing methods. In the current study, the STC Cocaine Metabolite MICRO-PLATE EIA demonstrated a clinicalsensitivity of 95% and specificity of 82% in testing of the overall 1. M.H. Glock, P.A. Heller, and D. Malamud. Saliva as a diagnostic fluid. NLM Bibliography 92-5: 1–102 (1982).
population (N = 312). In addition, an overall accuracy rate of 2. G.J. Di-Gregorio, A.J. Piraino, and E. Ruch. Diazepam concentra- 88% for detection of cocaine and cocaine metabolites in oral tions in parotid saliva, mixed saliva, and plasma. Clin. Pharmacol. fluids in comparison to GC–MS was observed (Table V). An ac- Ther. 24: 720–725 (1978).
Journal of Analytical Toxicology, Vol. 25, January/February 2001 3. P.M. Smith. Saliva and Oral Health, W.M. Edgar and D.M. O'Mul- caine in human saliva and urine after chronic use. J. Anal. Toxicol. lane, Eds. British Dental Association, London, U.K., 1980, p 9.
13: 65–68 (1989).
4. E.J. Cone. Saliva testing for drugs of abuse. Ann. N. Y. Acad. Sci. 15. W. Schramm, P.A. Craig, R.H. Smith, and G.E. Berger, Cocaine and 694: 91–127 (1993).
benzoylecgonine in saliva, serum, and urine. Clin. Chem. 39:
5. H.W. Peel, B.J. Perrigo, and N.Z. Mikhael. Detection of drugs in 481–487 (1993).
saliva of impaired drivers. J. Forensic Sci. 29: 185–189 (1984).
16. M.H. Zweig. Assessment of clinical sensitivity and specificity of 6. J.C. Mucklow, M.R. Bending, G.C. Kahn, and C.T. Dollery. Drug laboratory tests. NCCLS Document 7, 1987.
concentration in saliva. Clin. Pharmacol. Ther. 24: 563–570
17. J.K. Taylor. Quality Assurance of Chemical Measurements. Lewis Publishers, Chelsea, MI, 1988.
7. O.R. Idowu and B. Caddy. A review of the use of saliva in the 18. Substance Abuse and Mental Health Services Administration.
forensic detection of drugs and other chemicals. J. Forensic Sci. Mandatory guidelines for Federal workplace drug testing pro- Soc. 22: 123–135 (1982).
grams. Fed. Regist. 59: 29908–29931 (1994).
8. W. Schramm, R.H. Smith, P.A. Craig, and D.A. Kidwell. Drugs of 19. E.J. Cone, New developments in biological measures of drug abuse in saliva: a review. J. Anal. Toxicol. 16: 1–9 (1992).
prevalence. NIDA Res. Monograph 167: 108–129 (1997).
9. B. Caddy. Advances in Analytical Toxicology, R.C. Baselt, Ed.
20. E.J. Cone, J. Oyler, and W.D. Darwin. Cocaine disposition in saliva Biomedical Publications, Foster City, CA, 1984, p 198.
following intravenous, intranasal, and smoked administration. 10. E.J. Cone and A.J. Jenkins. Handbook of Analytical Therapeutic J. Anal. Toxicol. 21: 465–475 (1997).
Drug Monitoring and Toxicology, S.H.Y. Wong and I. Sunshine, 21. J.T. Cody. Specimen adulteration in drug urinalysis. Forensic Sci. Eds. CRC Press, New York, NY, 1997, p 303.
Rev. 2: 63–75 (1990).
11. N. Samyn, A. Verstraete, C. van Haeren, and P. Kintz. Analysis of 22. F. Luceri, F. Godi, and G. Messeri. Reducing false-negative tests in drugs of abuse in saliva. Forensic Sci. Rev. 11: 1–19 (1999).
urinary drugs-of-abuse screening. J. Anal. Toxicol. 21: 244–245
12. E.J. Cone, K. Kumor, L.K. Thompson, and M. Sherer. Correlation of saliva cocaine levels with plasma levels and with pharmacologic 23. E.J. Cone, R. Lange, and W.D. Darwin. In vivo adulteration: Excess effects after intravenous cocaine administration in human sub- fluid ingestion causes false-negative marijuana and cocaine urine jects. J. Anal. Toxicol. 12: 200–206 (1988).
test results. J. Anal. Toxicol. 22: 460–473 (1998).
13. E.J. Cone, B.A. Holicky, T.M. Grant, W.D. Darwin, and B.A. Gold- berger. Pharmacokinetics and pharmacodynamics of intranasal
"snorted" heroin. J. Anal. Toxicol. 17: 327–337 (1993).
Manuscript received February 22, 2000; 14. E.J. Cone and W.W. Weddington, Jr. Prolonged occurrence of co- revision received June 14, 2000.
1225 West Market St., Akron OH 44313 Phone: 330-869-0500, Fax: 330-869-0518 Safety Data Sheet Sovchem® TETD 1. CHEMICAL PRODUCT & COMPANY IDENTIFICATION Manufacturer: Emergency Contact: Sovereign Chemical Company Chemtrec: 1-800-424-9300 (continental USA) 1225 West Market Street