analyzed by immunoassay matched samples of urine, blood, and salivary ultrafiltratefrom 69 patients who had used cocaine within 24 h of sample collection.
CLIN.CHEM.39/3. 481-487 (1993)
Cocaine and Benzoylecgonine in Saliva, Serum, and Urine Wififried
Schramm,’
Paul A. Craig,’ Richard H. Smith,’ and Gregory
We studied the feasibility of using saliva to detect cocaine
and benzoylecgonine. Saliva was collected as an ultrafiltrate directly in the mouth with an osmotic device. We analyzed by immunoassay matched samples of urine, blood, and salivary ultrafiltrate from 69 patients who had used cocaine within 24 h of sample collection. Cocaine concentrations were 4.9 times higher in saliva than in serum; benzoylecgonine concentrations were 2.5 times higher in serum. Seven urine and two serum samples had undetectable concentrations of cocaine, but all 69 saliva samples were positive for the drug. For benzoylecgonine detection, all urine samples were positive and three serum and one saliva sample were n3gative. We also analyzed 43 samples of saliva by gas chromatography/ mass spectrometry: all were positivefor benzoylecgonine, whereas 40 were positive for cocaine. We conclude that simultaneous measurement of cocaine and benzoylecgonine in saliva is suitable in screening for recent cocaine use. Indexing Terms: screening
abused drugs
saliva ultra fil-
trate
For many diagnostic evaluations, the ability to obtain body fluids noninvasively presents new opportunities, e.g., screening for drugs of abuse. Although a urine sample is preferable to a blood sample, collecting urine is still controversial in light of the potential violation of privacy. Another disadvantage of urine as a diagnostic medium is that urine is not always available on demand, whereas saliva can be collected from a subject at any time, especially with the method used in this study. Saliva as an alternative medium has been investigated and reviewed for its utility for the measurement of therapeutic drugs and biological markers (1-8). Investigation of the measurement of cocaine in saliva is still in its infancy. Since an early study on the distribution of radioactively labeled cocaine in serum and saliva in 1978, almost 10 years elapsed before similar research was reported. We have found no publication on the direct measurement of benzoylecgonine (BEC), the major metabolite of cocaine, in saliva.3 The 1 BioQuant Inc., 1919 Green Rd.,Ann Arbor, Ml 48105. 2Sacred Heart Rehabilitation Center, 400 Stoddard Road, Memphis, MI 48041, and Wayne State University School of Medicine, Detroit, MI 48201. 3Nonstandard abbreviations: BEC, benzoylecgonine; GC/MS, gas chromatography/mass spectrometry; PBS, phosphate-buffered saline; and gel-PBS, phosphate-buffered saline containing gelatin. Received April 8, 1992; accepted September 11, 1992.
E. Berger2
half-life of BEC in blood is about five times longer than that for the parent substance (9-11), and this metabolite ismeasured in urine samples to determine previous use of cocaine. Cocaine is in equilibrium between blood and saliva in the body (12, 13); i.e., it is exchanged between the two fluids. The ratio of its concentration in saliva to that in serum ranged between 3 and 0.5 in one subject over time (13). After cessation of intake, the concentrations in urine, blood, and saliva decreased in parallel (14). In saliva, cocaine can be detected by immunoassay as long as 10 days after the last intake (14); it is detectable only for -24 h after intake by the less sensitive method of gas chromatography/mass spectrometry (GC/MS), which was developed for urine samples. The objective of this investigation was to determine the concentration of BEC in saliva of recent cocaine users (samples collected within 24 h of drug use) and to compare the concentrations of cocaine and BEC in saliva, serum, and urine. Saliva collection was simplified by using an osmotic device that permits the collection of an ultrafiltrate directly in the mouth. This device has been previously evaluated for saliva collection for the determination of cotinine, a metabolite of nicotine (15); steroid hormones (16, 17); and anticonvulsant drugs (18). This being the first report on the correlation of cocaine and BEC in saliva, serum, and urine for a relatively large number of subjects, we will present the results in detail. MaterIals and Methods Reagents From the Sigma Chemical Co. (St. LOUIS, MO) we purchased BEC, [N-methyl-2H3JBEC, cocaine hydrochloride, [N-methyl-2H3jcocaine, formic acid, gelatin, potassium phosphate, sodium acetate, sodium chloride, sodium fluoride, sodium phosphate, sodium sulfate (anhydrous), sucrose, and thimerosal. We obtained acetonitrue, methanol, methylene chloride, and chloroform from Burdick and Jackson (Baxter Healthcare Corp., Muskegon, MI); 2-propanol from EM Science (Cherry Hill, NJ); tetrabutylammonium hydroxide from Fisher Scientific (Fair Lawn, NJ); ammonium hydroxide, 1-butanol, ethyl acetate, and pentafluoropropanol from Aldrich Chemical Co. (Milwaukee, WI); and pentafluoropropiomc acid anhydride from Pierce (Rockford, IL). Phosphate-buffered saline (PBS) consisted of sodium phosphate, 10 mmol/L, pH 7.4, plus 9 g of sodium chloride and 0.2 g of thimerosal per liter. CLINICALCHEMISTRY, Vol.36, No.3,1993 481
We prepared BEC from cocaine by neutral hydrolysis of our preparation was similar to a reference standard (DAC no. 5315-1022-1; Research Technology Branch, National Institute on Drug Abuse, Rockville, MD), as determined by HPLC. We prepared cocaine standard solutions by dissolving cocaine in PBS containing 1.0 g of gelatin per liter (gel-PBS), and stored the aliquots at -20 #{176}C. Gelatin was added to provide a defined protein to the buffer to reduce nonspecific binding of the analyte to glass and plasticware during processing. We prepared BEC standard solutions for radioimmunoassay (RIA) also in gelPBS and stored them at 4#{176}C. RIA kits for the determination of cocaine metabolite (BEC) were purchased from Diagnostic Products Corp. (Los Angeles, CA) and Roche (19). The purity
Diagnostic
Systems (Nutley,
NJ).
Subjects and Specimens
sis (Cuprophan; Enka AG, Wuppertal, Germany). Two membrane discs (35 mm diameter, 20 thick, molecular cutoff 12000 Da) were bonded with polyurethane (Tycel; Lord Corp., Erie, PA) at the periphery, the adhesive forming a 5-mm-wide ring. Granular sucrose (0.75 g) was deposited inside the membrane. The performance of the collector has been described elsewhere (17, 18).
We measured the uptake of an 1251-labeled BEC derivative by the membranes of the osmotic device as an indication of nonspecificbinding of nonlabeled BEC and cocaine to the device. To measure uptake, we incubated several collectors with 125 mL of the radiolabeled BEC derivative (-17 finol/mL, -50000 cpmimL) dissolved in gel-PBS at 37#{176}C. After removing the osmotic devices at various times, we opened and rinsed them for 2-3 s in a beaker containing 100 mL of de-ionized water. We determined the amount of BEC derivative bound to each device by counting the radioactive decay of ‘I. We studied the transfer of BEC and cocaine into the osmotic device in an in vitro model that simulated conditions encountered in the mouth during collection (18). Solutions of BEC and cocaine, 140 mL each, were placed in separate plastic bags and the osmotic collectors were added. We gently massaged the bags with the collectors by hand, keeping the bags in a water bath at 37#{176}C for various periods, and determined by RIA the concentrations of BEC and cocaine in the osmotic pouch and in the bulk solution.
Subjects. We collected specimens from 69 volunteers (52 men and 17 women, ages 16 to 50 years) attending substance abuse programs who admitted using cocaine within the previous 24 h. Upon admission, staff members administered questionnaires to the subjects to determine the time of last use (self-reported times) and the route of administration. According to their reports, 67 patients ingested cocaine by smoking and 2 took cocaine by nasal inhalation. Specimens were also collected from 12 volunteers who had not taken cocaine for 6 months. The procedures followed were in accordance with the ethical standards defined in the Helsinki Declaration of 1975, as revised in 1983. Specimen collection. Urine, serum, and saliva speciProcedures mens were collected from subjects upon admission; blood Immunoassays. We determined the concentrations of and saliva were taken simultaneously and urine was BEC and cocaine with commercial RIA kits: Abuscreen collected within 2 h of this time. Blood was collected in (Roche Diagnostics) according to the manufacturers glass tubes (Corvac#{174}; Sherwood Medical, Charlotte, NC) guidelines, and Coat-A-Count Cocaine Metabolite Asand allowed to clot; the serum was transferred to polysay (Diagnostic Products Corp.) modified on the basis of propylene tubes and frozen. To sample an ultrafiltrate of cocaine standards as described elsewhere (14). Serum saliva, we immersed two osmotic collection devices (see and saliva ultrafiltrate specimens were usually assayed below) in tap water for a few secondsand then placed at 100 L per tube; for urine assays we usually used them in the subject’s mouth. We instructed subjects to serum, although individual values overlapped between sample types. The difference in average concentration was prevalent in all samples collected at different (self-reported) times after ingestion. Cocaine concentrations in saliva could reach values well above 1 mg/L, even in samples collected >16 h after ingestion. The correlation between the concentrations in all body fluids was poor except for urine vs serum at 8-16 h and 16-24 h after drug intake. BEC. The average concentrations of BEC in urine are -1000-fold greater than in saliva ultrafiltrates (Table 2). Serum contains a higher concentration than saliva but substantially less than urine. Individual values again vary over a wide range, and the concentrations overlap in the different body fluids. There is a trend toward decreasing concentrations of metabolite with increasing elapsed time between sample collection and time of ingestion. The concentrations of BEC in urine, serum, and saliva ultrafiltrate correlate much better than do those of cocaine (Tables 1 and 2). Drug detection in matched samples. Based on the criteria mentioned above for cocaine measurement, 7 of 69 urine samples from cocaine users were classified as negative.
All of the saliva samples were positive for
cocaine but results for two blood samples were below the cutoff value (
