Introduction

Top Abstract Introduction Materials and Methods Results Discussion and Conclusions References

Infants exposed to cocaine in utero may be subject to a variety of physical and behavioral abnormalities (Chasnoff et al., 1989). These risks would be expected to be dose-dependent and are supported by the relation between an increased impairment of fetal growth associated with increased exposure (Mirochnick et al., 1995). However, in most cases, the extent of fetal exposure is often difficult to assess, although levels of cocaine and its metabolites can be quantitated from amniotic fluid (Winecker et al., 1997), infant urine (Ostrea et al., 1989; Lewis et al., 1995), meconium (Ostrea et al., 1989), and hair (Graham et al., 1989).

Recently, meta-hydroxybenzoylecgonine (m-OH BE)³ was identified by gas chromatography-mass spectroscopy (GC-MS) during quantitative analysis for cocaine (Steele et al., 1993). Lewis et al. (1994) showed that identification of m-OH BE by GC-MS confirmatory assays may increase detection of cocaine-exposed infants and decrease false negative results. However, it is not known whether m-OH BE is directly derived from benzoylecgonine or from hydroxylated cocaine, or whether it is produced in the fetus or transferred across the placenta from the maternal circulation. Figure 1 illustrates the known metabolic pathways for cocaine and the possible derivation of m-OH BE (Turner et al., 1988).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Metabolic scheme for cocaine. Benzoylecgonine is formed by enzymatic and nonenzymatic hydrolysis of cocaine. The formation of m-OH BE is unknown. Adapted from Turner et al. (1988).

Thus this study was designed to test the hypotheses that m-OH BE is a metabolite derived from benzoylecgonine and that detection of m-OH BE would increase the potential identification of fetuses exposed to cocaine. We quantitated levels of cocaine, benzoylecgonine, and m-OH BE in amniotic fluid, fetal meconium, fetal intestine, and maternal urine after single dose administration of either cocaine or benzoylecgonine to pregnant time-bred guinea pigs. We found that although m-OH BE was detectable from meconium after both maternal cocaine and benzoylecgonine injection, the greatest yield of m-OH BE from meconium occurred later than that observed for cocaine or benzoylecgonine. Minimal concentrations of m-OH BE were found in amniotic fluid. In addition, detection of m-OH BE in meconium increased the identification of in utero exposed guinea pigs.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion and Conclusions References

This study was approved by the Animal Care and Use Committee, Tripler Army Medical Center. Procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication 85-23, revised 1985) and the Animal Welfare Act and Amendments. Tripler Army Medical Center is accredited by the American Association for Accreditation of Laboratory Animal Care.

Animal Preparation. Sixty-three-days timed-pregnant (term 68 days) Duncan-Hartley guinea pigs (Charles River Breeding Laboratories, Wilmington, MA) were given 10 mg/kg cocaine HCl (10 mg/ml) or benzoylecgonine hydrate (10 mg/ml) i.p. (Sigma, St. Louis, MO). At 24, 48, and 72 h postcocaine (n = 8, 11, and 8, respectively) and postbenzoylecgonine injection (n = 11, 7, and 11, respectively), pregnant guinea pigs were sedated with 25 mg/kg ketamine (Vetalar; Parke-Davis, Morris Plains, NJ) i.m.; and euthanized with 50 mg/kg intracardiac sodium pentobarbital (Wyeth, Philadelphia, PA).

Specimen Collection. Maternal urine was collected by direct bladder aspiration, then hysterotomies were performed. Amniotic fluid was aspirated under direct visualization from each sac before delivery of each fetus. Fetuses were delivered and euthanized with sodium pentobarbital (250 mg/kg i.p.). Meconium was removed from the colon from each fetus. Amniotic fluid and meconium from each fetus were analyzed separately for cocaine, benzoylecgonine, and free m-OH BE, and results were reported for each animal. A 96-h postcocaine injection period (n = 8) was included because the recovery of m-OH BE from meconium was still increasing at 72 h.

Extraction. Cocaine and metabolites were extracted from meconium and intestine. Isolute HCX mixed mode solid-phase extraction columns (200 mg/10 ml) were obtained from Jones Chromatography (Lakewood, CO). All reagents were of American Chemical Society grade or better and all solvents of high performance liquid chromatography grade. Deuterated and unlabeled analytical standards were obtained from Radian Corporation (Austin, TX).

Derivatization. Extracts were reconstituted in butyronitrile (40 µl), transferred to autosampler vials, and capped. N-methyl-N-(tert-butyldimethylsilyl) trifluoroacetamide (30 µl) was added and the extracts were heated at 80°C for 20 min.

Analysis for Cocaine, Benzoylecgonine, and Free m-OH BE. Analysis of the extracts was carried out using GC-MS in electron impact, selected ion monitoring mode. The instrument used was a Hewlett-Packard 5890 gas chromatograph coupled to a 5971A mass selective detector. The column was a DB-5 MS (25 m length × 0.2 mm i.d. × 0.33 µm film thickness). The injector temperature was set at 270°C, the detector at 310°C, and the oven program was as follows: Initial temperature 100°C for 1 min; ramped at 30°C/min to 230°C, ramped at 3°C to 249°C, ramped at 30°C/min to 310°C; held for 6.8 min. The injection mode was splitless and the ions monitored were D3-cocaine: 306, 185; cocaine: 303, 182, 198; D3-cocaethylene: 320, 199; cocaethylene: 317, 196, 212; D3-benzoylecgonine: 285, 406; benzoylecgonine: 282, 403, 346; and m-OH BE: 533, 476, 282. The limit of quantitation of the GC/MS assay is 5 ng/g of sample and the limit of detection of the assay is 2 ng/g.

Data Analysis. Quantitative recovery of cocaine and metabolites at different time periods after injection were compared using one-way ANOVA with Neuman-Keuls multiple comparison procedure. Statistical analysis was performed with the software package, SigmaStat (Jandel Sci., San Rafael, CA). Normality of data distribution was tested by the Kolmogorov-Smirnov test. P values <.05 were considered significant. Values are expressed as mean ± S.E.

	Results

Top Abstract Introduction Materials and Methods Results Discussion and Conclusions References

Amniotic Fluid. After maternal cocaine injection, cocaine was detected in amniotic fluid for up to 48 h (Fig. 2). Benzoylecgonine was found through 96 h (n = 8), but by 72 h (n = 8), the mean level was only 2 ± 0.8 ng/ml. m-OH BE was not consistently detected.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Concentrations of cocaine, benzoylecgonine, and m-OH BE from amniotic fluid at 24, 48, 72, and 96 h (n = 8, 11, 8, and 8, respectively) after maternal guinea pig cocaine injection. No cocaine was detected by 72 h after injection. Benzoylecgonine levels at 72 and 96 h were significantly less than those at 24 and 48 h. m-OH BE was not consistently detected.

View larger version (47K): [in this window] [in a new window]   Fig. 3.   Concentrations of benzoylecgonine from amniotic fluid at 24, 48, and 72 h (n = 11, 7, and 11, respectively) after maternal guinea pig benzoylecgonine injection. By 72 h, levels of benzoylecgonine were significantly less than at 24 h. m-OH BE was not detected.

Meconium. Cocaine was found in meconium 96 h after single dose maternal cocaine injection (6 ± 2 ng/g) (Fig. 4). m-OH BE was also present through 96 h (105 ± 10 ng/g), although benzoylecgonine was found only up to 72 h post injection (12 ± 8 ng/g). At 48 and 72 h, m-OH BE levels were significantly higher than those at 24 h (P < .05), whereas at 96 h post maternal cocaine injection, m-OH BE levels were significantly lower than at 48 h (P < .05). In addition, by 72 h, levels of cocaine and benzoylecgonine were significantly lower than at 24 h (P < .05).

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Concentrations of cocaine, benzoylecgonine, and m-OH BE from fetal guinea pig meconium at 24, 48, 72, and 96 h (n = 8, 11, 8, and 8, respectively) after maternal cocaine injection. m-OH BE levels at 48 and 72 h were significantly greater than at 24 h, but were less at 96 h than at 48 h. Benzoylecgonine levels at 72 h were significantly less than at 24 h.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Concentrations of benzoylecgonine and m-OH BE from fetal guinea pig meconium at 24, 48, and 72 h (n = 11, 7, and 11, respectively) after maternal benzoylecgonine injection. Levels of m-OH BE were significantly increased at 48 h, whereas levels of benzoylecgonine were significantly decreased at 72 h compared with 24 h.

Maternal Urine. After maternal cocaine injection, cocaine was not detected in any of the maternal urine sample collections. Benzoylecgonine was detected from 24 through 96 h (average 14 ng/ml of two samples at 96 h), whereas m-OH BE was detected through 72 h (average 5 ng/ml of two samples at 72 h).

Fetal Intestine. There was minimal recovery of cocaine and benzoylecgonine from fetal intestine after either maternal cocaine or benzoylecgonine administration (data not shown). Furthermore, m-OH BE was not detected from any of the intestine samples, suggesting that its presence in meconium was not the result of intestinal metabolism and secretion.

Percent Detection. The GC-MS assay for cocaine and benzoylecgonine in amniotic fluid 24 h after single dose maternal cocaine injection identified 100% of the exposed fetal guinea pigs (Table 1). However, by 96 h, only 12% of the guinea pigs had cocaine and 50% had benzoylecgonine in their amniotic fluid. Detection of m-OH BE in amniotic fluid did not increase identification of exposed guinea pigs.

	Discussion and Conclusions

Top Abstract Introduction Materials and Methods Results Discussion and Conclusions References

The plasma half-life of cocaine is very short, whereas its metabolites persist and may be biologically active. Although cocaine enters the fetal circulation rapidly and fetal exposure is prolonged (Sandberg and Olsen, 1992; Mahone et al., 1994), several lines of evidence suggest that many of the adverse effects of maternal cocaine use on the developing fetus may be due to the metabolite benzoylecgonine rather than to cocaine. Madden and Powers (1990) showed that benzoylecgonine is more potent than cocaine, causing a concentration-dependent vasoconstriction in isolated cannulated adult cat cerebral arteries. Schreiber and coworkers (1994) confirmed these findings in isolated fetal sheep cerebral arteries. Furthermore, Konkol et al. (1992) showed that intraventricular benzoylecgonine resulted in multiple and prolonged seizures in rats as compared with intraventricular cocaine. Thus, many of the adverse effects of cocaine on the developing fetus may be due to cocaine metabolites as well as to differences in kinetics and tissue distribution (DeVane et al., 1989; Spear et al., 1989; Konkol et al., 1994). Taken together, quantitative analysis for cocaine metabolites may conceivably improve our understanding of the dose-response effects of fetal cocaine exposure.

m-Hydroxy benzoylecgonine is a cocaine metabolite newly detected by GC-MS. It was first described by Steele and coworkers (1993) as an immunoreactive compound that cross-reacts with benzoylecgonine in screening immunoassays. Lewis et al. (1994) found that 23% of meconium samples collected for assay by GC-MS were positive only for m-OH BE, suggesting that detection of this compound may increase identification of cocaine-exposed infants. Although screening immunoassays do not differentiate m-OH BE from benzoylecgonine, positive samples are confirmed by GC-MS. Thus, detection of m-OH BE by the confirmatory assay would verify samples from cocaine-exposed infants (Steele et al., 1993). Moreover, this may be particularly relevant in improving the sensitivity of detection in situations where amount of exposure is low or when exposure has occurred earlier in gestation.

The actual formation and sites of metabolism of m-OH BE are presently unknown. As cocaine is metabolized through several different pathways, m-OH BE may be a breakdown product of benzoylecgonine or of hydroxylated cocaine (Fig. 1). In this study, we found that m-OH BE was present in fetal meconium after both maternal cocaine and benzoylecgonine injection, suggesting that m-OH BE is a metabolite of benzoylecgonine.

The formation of m-OH BE may be maternal with subsequent transfer to the fetus or it may be formed by the fetus after precursors are transferred across the placenta. In this study we showed that maternally administered cocaine and benzoylecgonine cross the placenta to the fetus, with subsequent metabolism to m-OH BE. Although cocaine is lipid-soluble and diffuses easily across the trophoblastic cell surface (Simone et al., 1994), the transfer of hydrophilic, lipid-insoluble molecules, such as benzoylecgonine and m-OH BE, via extracellular membrane channels is limited (Olsen et al., 1989; Simone et al., 1994). Nevertheless, the hemochorial placenta of humans and guinea pigs allows for permeability of hydrophilic molecules under molecular weight 5000 (Willis et al., 1986).

Although elimination of m-OH BE may occur via glucuronide conjugation in the liver and excretion into the bile (Gregus and Klaassen, 1992), the presence of glucuronide-conjugated m-OH BE is variable in the fetus and newborn. Steele et al. (1993) reported that the unconjugated fraction is 59 to 94% of total m-OH BE. Various factors may affect the levels of glucuronide conjugates in the fetus, for example, immature mechanisms of hepatobiliary uptake and biotransformation interfere with glucuronidation (Gregus and Klaassen, 1992). In contrast, the lack of intestinal bacteria limits -glucuronidase activity and impedes unconjugation and reabsorption. In fetal and neonatal guinea pigs, however, there is increased intestinal -glucuronidase activity (Lucier et al., 1977). It is important to note that only the free, unconjugated form of m-OH BE was measured in this study. Thus conclusions related to glucuronidation are beyond the scope of this study.

Meconium, commonly used to detect in utero exposure to cocaine and other substances subject to abuse, was believed to be a static repository for drugs (Ostrea et al., 1989). In this study, levels of cocaine, benzoylecgonine, and m-OH BE in fetal guinea pig meconium were shown to decrease over time after late gestation single dose maternal cocaine administration. In a previous study in which a similar model was used, methamphetamine levels in meconium were also found to decrease after maternal injection (Nakamura et al., 1992).

Garcia et al. (1996) detected high levels of cocaine and benzoylecgonine in gastric fluid, representing swallowed amniotic fluid, from infants soon after delivery. These infants were presumed to have been exposed to multiple doses of cocaine. Although cocaine has a relatively long half-life in amniotic fluid (Sandberg and Olsen, 1991; Mahone et al., 1994), in our study cocaine and benzoylecgonine levels were minimal in amniotic fluid by 48 h after single dose maternal cocaine. m-OH BE could not be consistently detected in amniotic fluid. Thus, compared with levels detected in samples from other sites (vide supra), amniotic fluid does not appear to provide a good source for analysis when exposure is minimal.

In summary, m-OH BE was measured from meconium after both maternal cocaine and benzoylecgonine injection. Detection of m-OH BE from amniotic fluid was minimal. The greatest yield of m-OH BE was from meconium, but peak concentrations were obtained later than that observed for cocaine or benzoylecgonine. In addition, detection of m-OH BE in meconium increased the identification of in utero exposed guinea pigs.