For& Science Znternut&mul, 56 (1992) 1’7’7 - 181 Elsevier Scientific Publishers Ireland Ltd.

DETERMINATION OF COCAINE AND BENZOYLECGONINE IN HUMAN AMNIOTIC FLUID USING HIGH FLOW SOLID-PHASE EXTRACTION COLUMNS AND HPLC

CHRISTINE MOOREB, SUSAN BROWNEb, L4N TEBBE’IV, ADAM NEGRUSZa, WILLIAM MEYERb and LUCKY JAINC

aDepartment of Phurmacodynumics, College of Pharmacy, Univereity of Zlliti at Chiwgo, Chicago, ZL 60680, bDepartment of Obstetrice and Gynecology and CDepartm.ent of Pediatrkx, Unive-rsity of Zlliwis Hoepital, Wood Street, Chicago, ZL 60612 (USA)

(Received April 30th 1992) (Revision received June 24th, 1992) (Accepted July lst, 1992)

Summary

A new solid-phase extraction procedure for the determination of cocaine and benzoylecgonine in amniotic fluid, using high flow co-polymerie sorbents is reported. The recoveries of cocaine and benzoylecgonine within the range O.l- 1 mg/l were 95.7% and 50.3%, respectively. The use of high- flow sorbents allowed the easy extraction of amniotic fluid regardless of sample viscosity or physical nature. The use of these solid-phase cohnnns provided many advantages over the more commonly used solvent extraction, including an increase in extraction speed and efficiency, reduced operator time, reduced solvent use and disposal volumes and exceptional extract quality. Further, the de- termination of amniotic fluid obtained from pregnant cocaine users may provide important informa- tion about handling of cocaine by the fetus at various gestational ages. The procedure was successfully applied to amniotic fluid from suspected cocaine abusers.

Key words: Cocaine; Amniotic fluid; High flow columns; HPLC

Introduction

The determination of cocaine abuse amongst pregnant women has become a topical issue for both legal and emotional reasons. The analysis of urine from both the mother and the baby using immunoassay procedures is the most com- mon method of determining the presence of cocaine and its metabolites [ll. Unless exposure to cocaine has been recent, false negative results may be obtain- ed therefore timing of sample collection from the neonates is crucial [2]. More recently, meconium, baby stools collected one, two or three days postnatal which is a dark-green mass of cells, mucus and bile accumulated in the bowel during fetal life and discharged shortly after birth, has been used to determine cocaine

Correspondaee to: Christine Moore, 443 Elgin, Coach House, 2R, Forest Park, Illinois 60130, USA.

178

abuse. It has been suggested that meconium is accumulated throughout the gestational period and therefore may be a more accurate sample for determining drug abuse during pregnancy than the more commonly used urine [3,4]. Since concern exists about the teratogenic effects of cocaine and the current methods of post-natal drug testing are inadequate in assessing timing and extent of fetal exposure, it was thought that amniotic fluid may provide more information about the magnitude of fetal exposure. Amniotic fluid is present throughout the preg- nancy, however, its use as a suitable sample for abuse determination has not been investigated. This could be due to the physical inconsistency of the fluid depending upon the time into the gestational period the sample is taken. Amnio- tic fluid tends to decrease in viscosity as the pregnancy progresses due to urina- tion of the fetus. The determination of amniotic fluid may provide important information about the handling of cocaine by the fetus at various gestational ages. Swallowing of amniotic fluid by the fetus is likely to result in recirculation of cocaine metabolites excreted in fetal urine. Hence, the testing of fluid may give more positive results owing to the prolonged presence of drug metabolites. The solid-phase extraction of amniotic fluid for the determination of methylcitric acid has been reported [5] and so has a microassay for the determination of co- caine, norcocaine, benzoylecgonine and norbenzoylecgonine in the amniotic fluid of guinea pigs [6].

We report a new solid-phase extraction procedure for cocaine, norcocaine, cocaethylene and benzoylecgonine which is independent of the viscosity of the fluid and sample sizes of up to 10 ml can be used (if available). The use of high- flow co-polymerie columns allows the easy passage of the fluid through the sor- bent. The extract was determined using HPLC with full wavelength UV detection.

Experiment81

Sample colkction Amniotic fluid was collected from eight pregnant women via amniocentesis

(n = 5) or transvaginally (n = 3) who were known, or thought, to have abused cocaine during pregnancy. The women verbally consented to the collection of this sample. The fluid was stored at - 20°C prior to extraction and analysis. Amniotic fluid from drug free pregnant women was obtained and spiked with cocaine and benzoylecgonine at three different concentrations (0.1, 0.5 and 1 mgll. Norco- caine and cocaethylene were also included in this study. Bupivacaine was used as an internal standard (20 mgll).

Drug standards and chemicals Cocaine, benzoylecgonine and bupivacaine were obtained from Sigma

Chemical Co. (St. Louis, MO). Norcocaine and cocaethylene were obtained from NIDA (Research Technology Branch, Division of Research). Al1 chemicals used were analytical grade or better and al1 solvents HPLC grade (Fisher, Itasca, IL).

Extraction procedure Solid-phase extraction was performed on X-TrackT@ high flow columns

179

(Worldwide Monitoring, Horsham, PA), containing 500 mg of a co-polymer with dual activity (cation exchange and non-polar interactions, 15 ml capacity).

Amniotic fluid (up to 10 ml, spiked or authentic sample) was added to 100 mM phosphate buffer (2 ml; pH 6) and bupivacaine (0.1 mg/ml; 20 uL) as an internal standard. The mixture was applied to an X-TrackTe column which had been previously conditioned with methanol (5 ml), water (5 ml) and buffer (3 ml; pH 6). The sample was drawn through under vacuum and the column was air-dried. The sorbent was washed with water (3 ml), 100 mM hydrochloric acid (3 ml) and methanol (3 ml). The column was allowed to dry for 5 min under vacuum. Finally, the isolates were eluted with chloroform-isopropanol ammonium hydroxide (78:20:2, v/v) (10 ml). The eluent was evaporated to dryness without heating. The residue was reconstituted in methanol (0.1 ml) prior to analysis.

Chromatographic equipment and conditions A Perkin Elmer Series 2 pump was used to deliver solvent at a rate of 1.5

mllminute onto a Cl8 ODS mu Bondapak (30 cm x 3.9 mm i.d.) column (Waters, Milford, MA). A Cl8 guard column (Guard-Pak, Waters) and a Rheodyne injec- tion system with a fixed loop (20 ~1) were also incorporated into the system. A Spectra Physics Focus multiwavelength detector connected to an IBM Personal Data System 2 data system was used to monitor the eluent from the column at 230,255 and 275 nm and full spectra were recorded over the wavelength range 190 -400 nm. The mobile phase consisted of 0.025 m potassium dihydrogen phosphate-acetonitrile-butylamine (500:125:12.5, v/v). The final pH was adjusted to 2.9 with orthophosphoric acid.

Results and Discussion

Optimisation of elution solvent required The extraction procedure was very rapid and the resulting extracts were suffi-

ciently clean for direct injection onto the HPLC system without further clean up stages. Initially, various elution volumes were tried (Table 1). As the volume ap- proached 10 ml, the recovery of the drugs improved. The recovery of benzoyl- ecgonine, though acceptable, was stil1 markedly lower than the other metabolites (Table 2).

TABLE 1

PERCENTAGE OF DRUG REMOVED FROM SORBENT USING VARIOUS ELUTION VOLUMES

Amount of drug added was 0.5 mg/I, BZE is benzoylecgonine.

Drug 3 ml 6 ml 10 ml

BZE 19.3 39.0 58.8 Cocaine 66.7 95.1 96.4 Norcocaine 61.8 78.8 92.0 Bupivacaine 67.1 86.5 97.3 Cocaethylene 69.9 90.4 90.4

180

TABLE 2

ABSOLUTE RECOVERY OF DRUGS USING 10 ML OF ELUTlON SOLVENT

Amount add_ed (mgll) 1 (n = 3) 0.5 (n = 5) 0.1 (n = 6)

BZE 50.3 f 8.6% 55.8 f 7.3% 58.5 f 6.1% Cocaine 95.7 f 9.3% 96.8 * 4.8% 96.4 f 0.2% Norcocaine 80.6 * 5.7% 90.2 f 5.4% 84.4 f 4.7% Bupivacaine 86.1 f 7.6% 97.3 f 5.4% 100 f 3.2% Cocaethylene 89.1 f 10.9% 90.4 f 8.0% 86.4 f 0.1%

High perfownance liquid chromatography (HPLC) The HPLC system was linear for spiked amniotic fluid extra& over the con-

centration range 0.1 - 10 rg/ml. The minimum quantitation leve1 of cocaine by HPLC was 100 ng/ml and the minimum detection leve1 30 nglml. The retention times of benzoylecgonine, cocaine and bupivacaine were 6.2, 7.9 and 10.1 min, respectively (Fig.1).

Authentic samples Al1 the samples taken from suspected cocaine abusers were found to contain

benzoylecgonine andlor cocaine (Table 3). The average gestational age of the fetuses was 35.2 weeks (range 33-38 weeks) and the mean birth weight was 2312 g (range 1924- 3169 g). These values are significantly higher than the average gestational ages and birth weights of neonates in other studies con- ducted to establish or detect cocaine abuse, suggesting that not al1 full-term, substantial birth weight babies are free from drug abuse. In one case (Case 3) the amount of benzoylecgonine was particularly high (3.06 mg/l) compared to the other values. The presence of the parent drug in the same sample may suggest recent use of cocaine. The passage of water soluble benzoylecgonine across the

0.0200

0.0160

0.0120

0.0080

0.0040

0.0000

Fig. 1. HPLC chromatogram of standard containing 10 pglml cocaine (C) benzoylecgonine (B) and 20 pglml bupivacaine (BU) as internal standard.

181

TABLE 3

LEVELS OF COCAINE AND BZE FOUND IN AMNIOTIC FLUID

BZE, benzoylecgonine; NQ, not quantified; ND, not detected; mean gestational age, 35.2 weeks; mean birthweight, 2312 grams.

Case Number BZE (mgll) Cocaine (mgll)

0.23 +,NQ 0.04 +, NQ 3.06 0.25 1.98 ND 1.65 ND 0.75 ND 0.39 0.014 0.29 0.07

placenta is limited and recovery from amniotic fluid is indicative of fetal metabolism of cocaine. The absente of cocaine in cases 4,5,6 suggests total con- version of the parent drug to the benzoylecgonine.

We suggest that amniotic fluid be seriously considered as an alternative sam- ple for the determination of cocaine abuse among pregnant women, since neonatal urine and meconium are often difficult to collect and the timing of that collection is crucial to accurate determination of drug use. The recirculation of cocaine andlor its metabolites through fetal urination and amniotic fluid swallow- ing may lead to prolonged cocaine exposure in utero.

The extraction procedure described is efficient, rapid and independant of the physical nature of the sample which may vary according to the gestational time period.

References

1 E. Com and J. Mitchell, Validity testing of commercial urine cocaine metabolism assays: 11 Sen- sitivity, specificity, accuracy and confirmation by gas chromatography-mass spectrometry. J. Forens. Sci., 34 (1989) 32-35.

2 A.C. Halstead, W. Godolphin, G. Lockitch and S. SegaI, Timing of specimen collection is crucial in urine screening of drug dependent mothers. CZin. Bioch., 21 (1988) 59- 61.

3 E.M. Ostrea, P.M. Parks and M.J. Brady, Rapid isolation and detection of drugs in meconium of infants of drug dependent mothers. Clin. Chem., 34 (1988) 2372-2373.

4 E.M. Ostrea, M.J. Brady, P.M. Parks, D.C. Asensio and A. Nalus, Drug screening of meconium in infants of drug-dependent mothers: An alternative to urine testing. J. Pediatr., 115 (1989) 474-477.

5 R.E. Rretschmer and C. Bachmann, Methylcitric acid determination in amniotic fluid by electron-impact mass fragmentography. J. CZ&. Cti. Cl&. Biochem., 26(5) (1988) 345 - 348.

6 J.A. Sandberg and G.D. Olsen, Microassay for the simultaneous determination of cocaine, nor- cocaine, benzoylecgonine and norbenzoylecgonine by high-performance liquid chromatography. J. Chromatogr., 525(l) (1990) 113- 121.