This item is only available as the following downloads:
MATERNAL-FETAL COCAINE METABOLISM, DISTRIBUTION,
DIANE LYNNE PHILLIPS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
I would like to thank all of the members of my supervisory committee, Dr. lan
R. Tebbett, Dr. Hartmut Derendorf, Dr. Roger L. Bertholf, and Dr. Kathleen T.
Shiverick, for their significant contributions towards the completion of the research
for this dissertation. I would especially like to thank my advisor, mentor, and friend
Dr. lan Tebbett for all of his support through the years I have spent on my graduate
education both at the University of Florida and at the University of Illinois at
I would like to thank those individuals who spent their own valuable time
teaching me new techniques or supplying samples used for this work, including
C Dr. Stephen Roberts, Center for Human and Environmental Toxicology,
University of Florida, with whom we collaborated for all of the cocaine and
ethanol interaction studies in mice.
c> Dr. Robert Covert, M.D., and Dr. Samir Wassef, M.D., of the University of
Chicago for supplying samples of meconium for the comprehensive
meconium testing and for statistical analysis.
> Dr. Lucky Jain, M.D., of the University of Illinois at Chicago for supplying samples.
o Dr. Joanna Peris, Department of Pharmacodynamics, University of Florida, for
providing samples for the rat cocaine and ethanol interaction study.
0 Ruth Winecker, graduate student for Dr. Bertholf, for her invaluable assistance
with sample derivitization and GC/MS.
1 Scott Masten, graduate student for Dr. Shiverick, for teaching me everything I
ever wanted to know about cell culture and for helping to critically analyze
the IM-9 cell data as experiments were completed.
0 Dr. Janet Karlix and Becky Frieberger, Department of Pharmacy Practice,
University of Florida, for additional cell culture work for the human cord blood
C Kathie Wobie, Project Care, Department of Pediatrics, University of Florida, for
teaching me my "social worker skills" and for her expertise on subject
enrollment for the cord blood PBMC study.
TABLE OF CONTENTS
ACKNOW LEDGMENTS...................................................................... ii
ABSTRACT.... ........................................................................ .............. vi
1 BACKGROUND AND SIGNIFICANCE............................. ....... 1
Pharmacological and Physiological Effects of Cocaine........... 2
Cocaine Administration and Pharmacokinetics.................... 7
Cocaine Metabolism......................................................... 13
Maternal-Fetal Circulation and Pharmacokinetics................. 18
Drug Distribution to the Fetus.............................................. 21
Effects of Gestational Cocaine Exposure on the Neonate...... 24
Immunotoxicology of Cocaine......................................... 28
Cocaine and Alcohol Interactions.................................... 35
Analytical Methods for Detection of Gestational
Cocaine Exposure............... ............................. 37
Hypotheses for this Study..................................................... 41
Specific Aim s ... ................................................. .............. 41
Biological Samples ............................................... ............ 42
Specific Aim #1....................................... 43
Specific Aim #2.................................................................... 57
Specific Aim #3.................................... 58
Specific Aim #4............ ............................ 62
Specific Aim #1 .................................................................. 69
Specific Aim #2................................................... ............... 93
Specific Aim #3.................................................... ............... 94
Specific Aim #4 ........... ................ ...... ................ 123
4 DISCUSSION........ .......... ................. ...... 135
A CALIBRATION CURVES................................... ..................... 158
B RAW DATA FOR INTERDAY AND INTRADAY
VARIABILITY CALCULATIONS............................ ............. 160
BIOGRAPHICAL SKETCH.................................. ................ 181
Abstract of Dissertation Presented to the Graduate School of the
University of Florida in Partial Fufillment of the
Requirements for the Degree of Doctor of Philosophy
MATERNAL-FETAL COCAINE METABOLISM, DISTRIBUTION,
DIANE LYNNE PHILLIPS
Chairman: Dr. lan R. Tebbett
Major Department: Pharmaceutics
Cocaine use by women during pregnancy has prompted interest in cocaine
distribution to the mother and fetus and in long term effects of cocaine on fetal
development. Conservative estimates indicate that 11-15% of all newborns have
been exposed to cocaine during gestation. Hypotheses: the fetus, given unique
routes of exposure in utero and immature metabolism, is exposed to higher
concentrations of cocaine than the mother during maternal cocaine use in
pregnancy; the ingestion of cocaine with alcohol further increases cocaine
concentration in both mother and fetus; and cocaine and metabolites have
immunotoxicological effects thereby making the neonate more susceptible to
Amniotic fluid, cord blood, and neonatal urine were obtained at delivery from
women who revealed cocaine use during pregnancy. Cocaine and metabolites
were extracted from body fluids using solid-phase extraction and quantified using
high performance liquid chromatography and/or gas chromatography-mass
spectrometry. Results of human and animal studies indicate that the fetus is
exposed to high cocaine concentrations in utero and eliminates the drug more
slowly than the mother. In mice, ethanol cotreatment resulted in a fivefold increase
in peak cocaine serum concentrations and a sixfold increase in peak cocaine
concentrations in the brain. Peak cocaine concentrations were reached faster in
ethanol treated mice and rats. These observations suggest that ethanol can alter
the pharmacokinetics of cocaine, which could partially account for reports of
increased toxicity of the drug when combined with alcohol. Cocaine has multiple
actions on several neurotransmitters many of which have a direct effect on immune
function. The immunotoxicologic effects of cocaine are therefore of interest. No
studies have been published examining the effect of cocaine metabolites on the
immune system. Proliferative effects of cocaine and cocaine metabolites
benzoylecgonine, norcocaine, and cocaethylene on B lymphocytes were examined
using the IM9 human cell line. Results suggest that cocaine has an inhibitory effect
on T-cells while having a stimulatory action on B-cells. This effect is similar to that
produced by the human immunodeficiency virus. Cord blood mononuclear cells
isolated from both cocaine exposed and nonexposed neonates showed similar
responsiveness to mitogenic challenge.
BACKGROUND AND SIGNIFICANCE
Cocaine is one of the most widely abused illicit drugs in the United States.
An estimated 50 million Americans (25% of the population) have used cocaine and
5-6 million use it regularly [1,2]. The most recent population estimates on the use
of cocaine were reported by the National Institute on Drug Abuse and the
Substance Abuse and Mental Health Services Administration in the 1992 "National
Household Survey on Drug Abuse." Results indicated an estimated 5 million
Americans used cocaine in the previous year with the majority of users reporting
"crack" cocaine use . As the price of cocaine has dropped in the last 20 years,
the number of users has risen significantly and the use pattern has changed from
recreational experimentation to abuse. It is estimated that 15-30% of all cocaine
users are women of childbearing age [1,3]. Recently, more attention has been
directed to this population and some hospitals perform routine screening of
newborn's urine for drugs of abuse. From such screening studies, it has been
reported that 11-15% of all pregnant women have used cocaine during pregnancy
[4-7]. However, there appears to be significant geographical variation in
prevalence. One recent study from a Chicago hospital yielded a prevalence rate
of 37% of very-low birthweight babies testing positive for cocaine and/or cocaine
metabolites , while 18% of newborns tested positive in a Boston survey and 2-3%
in a Rhode Island survey . In a survey of 36 hospitals from different
geographical regions across the country, the overall prevalence rate was 11% with
reports ranging from 0.4-27% . The effect of cocaine use by the mother during
pregnancy has been associated with medical complications in the newborn. Of
particular concern at this time is the effect of cocaine on immune function as over
75% of all perinatally acquired human immunodeficiency virus (HIV) infections are
associated with intravenous drug use by the mother or her sexual partner .
Pharmacological and Physiological Effects of Cocaine
Cocaine has multiple actions affecting dopaminergic, cholinergic,
serotonergic, and noradrenergic receptor sites. Cocaine enhances the experience
of pleasure with increased alertness, sexuality, and energy, and decreased anxiety
and social inhibition. However, while the use of low doses of cocaine is reported
to enhance feelings of pleasure, the chronic use of high doses of cocaine may lead
to irritability, aggression, paranoia, anorexia, and psychosis. The effects of cocaine
on the central nervous system (CNS) are well characterized and demonstrate a
biphasic patten of intense stimulation followed by depression. Cortical stimulation
is manifested by euphoria, restlessness and excitability. However, none of the
known neurochemical actions of cocaine singly provide a full explanation of how the
drug produces euphoria. In the CNS, cocaine blocks the reuptake of dopamine at
the dopamine transporter in the synaptic cleft thereby producing an accumulation
of dopamine at postsynaptic receptor sites (Figure 1-1). This, in turn, increases
neurotransmission in the mesolimbic and mesocortical dopaminergic pathways in
the brain. These pathways are termed the "reward center" of the brain and function
in producing the euphoria associated with cocaine use.
ATP CYCLASE CAMP
Figure 1-1: Effects of cocaine at dopaminergic synapses .
It is not clear, however, whether the euphoria is a direct effect of cocaine on
dopaminergic neurons or a combination effect due to simultaneous actions on other
neurotransmitters, like serotonin [10-12]. By blocking the reuptake of dopamine
with chronic cocaine use, the tissue stores of this neurotransmitter are gradually
depleted and this results in compensatory upregulation and supersensitivity of
dopamine receptors [1,11]. The blockade of dopamine reuptake is associated with
the intense reinforcing stimulus of cocaine [13,14]. Cocaine also blocks the
reuptake, turnover, and synthesis of serotonin (5HT) and this mechanism may be
involved in the euphoric action of cocaine as well as the onset of tolerance and
effects on sleep, appetite, and aggression [1,14].
Cocaine also blocks the reuptake of catecholamines, primarily
norepinephrine (NE), at adrenergic nerve endings potentiating sympathetically
mediated vasoconstriction, tachycardia, hyperglycemia, and hypertension [1,4,15].
Initially, in the brain, the levels of NE are elevated; however, cocaine activates the
presynaptic alpha-2 receptor which, in turn, produces feedback inhibition of NE. As
with dopamine receptors, chronic cocaine use results in compensatory upregulation
of adrenergic beta receptors in the CNS.
The toxic effects of cocaine have been the subject of a great deal of research
in recent years and several complications associated with cocaine abuse have been
identified. Pyrexia is a prominent feature of cocaine poisoning. Heat retention
occurs due to vasoconstriction, increased muscle activity, and direct effects on the
heat regulating center of the diencephalon . Death by respiratory collapse has
occurred immediately after IV administration of cocaine and pulmonary edema is
a common clinical finding at autopsy following cocaine overdose. The use of crack
cocaine has been associated with characteristic pulmonary injury known as "crack
lung" which is characterized by inflammatory cell infiltration and alveolar
hemorrhage . Oral and nasal ingestion have resulted in death after a symptom
free period of one hour followed by the onset of generalized seizures .
Cerebrovascular accidents, including seizures, intracerebral hemorrhage,
subarachnoid hemorrhage and cerebrovascular infarction, have also been reported.
The mechanism of hemorrhage is thought to be associated with cocaine-induced
hypertension causing the rupture of a pre-existing lesion [15,18]. Cocaine and
metabolites are also reported to be potent vasoconstrictors of cerebrovascular
arterioles [19,20]. The accumulation of catecholamines predisposes the
myocardium to arrhythmias which may compromise cardiac output, in some cases
resulting in myocardial infarction and ischemia due to the myocardial increased
demand for oxygen. Due to the vasoconstrictive action of cocaine, the myocardium
is not able to utilize compensatory autoregulatory mechanisms involving
vasodilation in response to the increased oxygen demand. Cocaine toxicity has
also been implicated in sudden arrhythmic death, aortic rupture, contraction band
necrosis, cardiomyopathy, and myocarditis [15, 21-30]. Intestinal ischemia may
result from cocaine induced catecholamine stimulation of alpha receptors in
mesenteric vasculature, causing intense vasoconstriction and reduced blood flow
[15,30]. Death from cocaine overdose usually results in 1-5 hours, and the number
of cocaine related emergency room visits have tripled in the last 10 years.
Additionally, cocaine also serves as a local anesthetic and this was one of
the earliest uses of cocaine in the United States. Cocaine has been used as a local
anesthetic in dental, nasal, and ophthalmic surgery. The pharmacologic action of
cocaine (as well as other local anesthetics) is the blockade of sodium channels in
the membranes of neuronal cell bodies, neuronal axons, and cardiac muscle
[31,32]. Sodium channel blockade prevents the propagation of action potentials to
neighboring cells and the nerve conducts fewer impulses. While unionized cocaine
penetrates the membrane faster, the cationic form predominates at physiological
pH and is the active form of the drug at the receptor site. When compared to other
local anesthetics, cocaine has a unique effect in that it also acts as a
vasoconstrictor at the site of application. This prolongs the local anesthetic effect
and reduces bleeding. Other local anesthetics are often administered with a
vasoconstrictor, such as epinephrine, to produce a similar effect. Since cardiac
muscle cells also contain sodium channels, these are also blocked by cocaine. This
disrupts normal cardiac pacemaker activity and conduction, which when combined
with the blockade of norepinephrine reuptake, can produce the arrhythmia and
hypertension discussed previously.
Finally, studies in rodents and non-human primates indicate cocaine also
affects normal endocrine function and alters secretion of anterior pituitary hormones
via its action on central dopamine systems [33-37]. Pituitary prolactin secretion is
under tonic inhibitory regulation by dopamine (dopamine is known as prolactin
inhibitory factor (PIF)). Dopaminergic neurons in the tuberofundibular pathway
(connecting the arcuate nuclei with the hypothalamus and posterior pituitary) are
responsible for prolactin release from the pituitary. In human studies, chronic
cocaine use has been associated with persistent hyperprolactinemia in cocaine
users [38-40]. A controlled study with 8 male human cocaine users demonstrated
that hyperprolactinemia was associated with chronic cocaine use; however, there
was no difference in the pulse frequency of prolactin release. This indicated that
the hyperprolactinemia may have resulted from the effect of cocaine on
dopaminergic inhibition of basal prolactin secretion . A follow-up study by this
group with 42 patients undergoing treatment for cocaine addiction found that while
hyperprolactinemia was not common, its presence was a poor prognostic sign for
abstinence from cocaine use . Research in this area is on-going, especially
evaluation of the role of serotonin on hypothalamic-pituitary-adrenal axis
Cocaine Administration and Pharmacokinetics
Cocaine is a naturally occurring alkaloid found in the leaves of the South
American shrub Erythroxylon coca. The leaves are estimated to contain up to 2%
cocaine by weight and are harvested at 3-4 years old [4,42]. South American
natives have chewed the coca leaf for centuries for its stimulant effects, and it was
introduced to the United States in 1885 for use both as a local anesthetic and in
some over-the-counter products, including wine and Coca-Cola [1,4,13]. In the
early 1900s, as the abuse potential of cocaine became evident, laws were passed
to restrict its use. By World War II, all states had laws prohibiting cocaine use for
other than medicinal use as a local anesthetic. Presently, cocaine is placed in
Schedule 2 of the Controlled Drug Act following guidelines stipulated by the federal
government. Drugs placed in Schedule 2 have an accepted medical use, but also
have high abuse potential. Cocaine powder, or cocaine hydrochloride, is extracted
from the coca leaves using acids and an organic solvent like gasoline. This form
of cocaine is a water-soluble salt which can be administered by nasal insufflation
or by intravenous injection of an aqueous solution. However, since 1985, free-base
cocaine, in the form of "crack" cocaine, has become the most widely used form of
the drug. "Crack" cocaine is an impure free-base manufactured from cocaine
powder by mixing with ammonia, baking soda, and warm water. The free base
precipitates in the alkaline water and the water is then boiled away leaving hard
chunks of the drug [43,44]. "Crack" is not water-soluble but is stable at
temperatures required for pyrolysis and is administered by vaporization via smoking
in a water pipe. Crack cocaine can be obtained illicitly for as little as $10 and the
rapid, intense euphoria produced makes it highly reinforcing .
The pharmacokinetics of cocaine is dependent on the route of administration.
Natives of the Andes and Amazon chewed coca leaves with ash thereby forming
a coca paste. The coca paste was either held in the mouth for long periods of time
and gradually swallowed or it was mixed with tobacco and smoked. The chewing
of paste does not produce euphoria, but does provide mild stimulation and
enhancement of endurance. The chewing of powdered coca leaves which
contained 17-48mg of cocaine was reported to produce peak plasma
concentrations of 10-150pg/L within 0.4-2 hours in 6 volunteers . The chewing
of 50g of coca leaves in a 3 hour period produced plasma cocaine concentrations
of 150-450ng/ml in native Peruvians  and similar cocaine levels were reached
much faster when coca paste was smoked .
Cocaine is absorbed from mucous membranes and the gastrointestinal tract,
although oral ingestion is not common. Gastrointestinal absorption is slow and
bioavailability of cocaine is estimated to be 30% by this route . Gastrointestinal
absorption is slow since cocaine (with a pKa of 8.6) would be ionized in the stomach
(pH 2.0) and would perhaps be more readily absorbed from the more alkaline small
intestine (pH 6.0-8.0). One study using human volunteers measured plasma
cocaine concentrations after oral administration of a dose of 2mg cocaine
hydrochloride per kilogram body weight in a gelatin capsule . For comparison,
the same dose was also administered topically to the nasal mucosa. Results
indicated that after oral administration, cocaine was not detected in plasma for 30
minutes. Peak plasma concentrations of 104-424ng/ml were reached in 50-90
minutes and then declined over the next 4-5 hours and the half-life was 0.9 hours.
Results of intranasal application demonstrated detectable plasma cocaine
concentrations by 15 minutes and peak concentrations of 61-408ng/ml at 60-120
minutes with a half-life of 1.3 hours. Subjective measures of "high" by the subjects
were reached faster in the intranasal applications .
Nasal insufflation of cocaine hydrochloride was once the predominant route
of drug administration by users . After topical application of a cocaine dose of
1.5mg/kg body weight to the nasal mucosa of 13 human patients, the absorption
half-life was about 12 minutes and peak plasma levels of 100-500ng/ml were
achieved in 15-60 minutes. Plasma cocaine concentrations then decreased
gradually over the next 3-5 hours, although cocaine was still detectable on swabs
from the nasal mucosa three hours after the application . In a study designed
to parallel the illicit intranasal use of cocaine, 10 human volunteers were instructed
to inhale 100mg of a mixture of cocaine hydrochloride powder (16-96mg) and
lactose through a 5.0 cm straw . On alternate days, subjects were given 16-
32mg of cocaine hydrochloride intravenously in 1ml of saline. Plasma cocaine
concentrations were measured after dosing. After nasal insufflation, peak plasma
cocaine concentrations of 53-206ng/ml were reached in 30-60 minutes; however,
plasma cocaine concentrations persisted for a longer period of time than subjective
drug effects reported by the subjects. After intravenous administration, peak
plasma cocaine concentrations of 200-300ng/ml were reached in 5 minutes and
physiological and subjective drug effects paralleled declining cocaine
concentrations . Jeffcoat et al  used radiolabeled cocaine to examine
disposition after nasal insufflation by human volunteers. A dose of 106mg of [4-3H]-
cocaine (equivalent to 94.6mg of cocaine) was insufflated through a glass tube.
The average bioavailability was 80%. The absorption half-life was 11.7 minutes and
the elimination half-life was 80 minutes with the mean peak plasma cocaine
concentration of 220ng/ml reached by 45 minutes.
Intravenous injection of cocaine and smoking free-base cocaine produce
similar pharmacokinetic profiles and both produce an intense euphoria within 1
minute of administration. Intravenously injected cocaine has been reported to reach
the brain in 16 seconds; however, due to efficient respiratory exchange, smoked
free-base cocaine reaches the brain faster (in 8 seconds), but has a shorter
duration of action . Jeffcoat et al  also administered radiolabeled cocaine to
human volunteers by intravenous injection and smoking. A dose of 23mg of [4-3H]-
cocaine (equivalent to 20.5mg of cocaine) was dissolved in 3 ml of saline and then
injected into an antecubital vein over a 1 minute period. The mean peak plasma
cocaine concentration of 180 ng/ml was reached by 5 minutes with the distribution
half-life reported to be 11 minutes and the elimination half-life 78 minutes . For
administration via smoking, a dose of 50 mg of [4-3H] cocaine free base was placed
in a glass pipe and the pipe was heated to 250"C in a hot oil bath. Subjects were
instructed to inhale vapors for 10 seconds (holding the inhaled vapor for 15
seconds) at 30 second intervals for 5 minutes. The absorption half-life was 1.1
minutes and the mean peak plasma concentration of 203ng/ml was reached by 6
minutes . The average bioavailability was 57%; however, a previous study
demonstrated that most undecomposed cocaine reaches the circulation and the
lower bioavailability is related to decomposition before inhalation . The
inhalation efficiency of "crack" pyrolysis has been measured as 739% and 6211%
at 170"C and 220"C respectively .
Plasma cocaine concentrations after multiple dosing were also reported
following intravenous and pulmonary administration of cocaine to human subjects
by Isenschmid et al . Subjects were given two separate doses of cocaine
spaced 14 minutes apart by either intravenous injection (16mg or 32mg cocaine)
or by smoking (25mg or 50mg cocaine). Plasma samples were analyzed for
cocaine and metabolite concentrations and were found to be dose dependent.
Peak cocaine concentrations ranged from 210ng/ml for 25mg cocaine via smoking
to 470ng/ml for 32mg cocaine via intravenous injection. Results indicate that the
half-life of cocaine was 38-39 minutes regardless of dose or route of administration.
To simulate the binge cocaine use of a typical user, some subjects were given 5-7
doses of cocaine in 90 minutes using a mixture of intravenous injection and
smoking. The maximum cocaine concentration measured was 1200ng/ml from an
individual who received 16mg of cocaine by IV bolus and then smoked 50mg
cocaine 6 separate times over 90 minutes. In subjects smoking 50mg cocaine 6
separate times over 90 minutes, the average plasma cocaine concentration
achieved was 890ng/ml .
The pharmacokinetics of cocaine have been studied by many investigators
with most reporting the half-life to be 43-60 minutes [52,53,56-58]. The volume of
distribution of cocaine is small and reported values range from 1.2-2.7L/kg
[4,18,53,59]. When plasma levels of cocaine and urinary excretion data were
plotted together, the decline of the excretion rate plot paralleled that of the decline
in plasma indicating cocaine follows a linear pharmacokinetic model . Garrett
and Seyda  found the plasma protein binding of cocaine to be less than 10%
and not a significant pharmacokinetic variable. However, this report is challenged
by Edwards and Bowles  who report that cocaine is highly bound to the serum
proteins albumin and alpha-1-acid glycoprotein. Support that this protein binding
is not a significant factor, however, comes from placental transfer studies which
found that while cocaine was 50-90% bound to serum proteins, diffusion across the
placenta was still rapid and cocaine binding to umbilical cord serum was
significantly less than to maternal serum .
Only 1-5% of a dose of cocaine is excreted unchanged in the urine [4,60].
The remainder is metabolized by several pathways to water soluble metabolites
which are excreted in the urine (Figure 1-2):
(1) Cocaine is hydrolyzed by plasma and liver cholinesterases to ecgonine
methyl ester; in addition, non-enzymatic hydrolysis and hydrolysis by liver
carboxylesterases to benzoylecgonine also occurs. Ecgonine methyl ester is
primarily a urinary metabolite and is usually detected in the blood only in post-
mortem cases . These pathways account for 70-90% of cocaine metabolism in
(2) The second pathway is via N-demethylation by mixed function oxidase
to norcocaine, a physiologically active cocaine metabolite . Human cytochrome
P450 3A is the oxygenase responsible for the oxidative metabolism of cocaine.
(3) When cocaine is used simultaneously with ethanol, a transesterification
reaction occurs in the liver (catalyzed by hepatic carboxylesterase) to form
cocaethylene [66-68]. Cocaethylene is an active metabolite retaining cocaine's
pharmacologic and toxicologic properties and has a half-life of more than twice that
of the parent compound.
(4) Other, minor metabolites have been identified in the blood and urine of
cocaine users including, ecgonine, ecgonidine, norecgonine methyl ester, and
Figure 1-2: Routes of metabolism of cocaine in vivo.
Benzoylecgonine and ecgonine methyl ester are water soluble metabolites
readily excreted in the urine and can be detected for 24-72 hours after a cocaine
dose. Acidification of the urine, in combination with diuretics, can increase renal
excretion of cocaine. Renal excretion of cocaine in the neonate may be slow as the
glomerular filtration rate is only 30-40% that of an adult and is directly related to
gestational age. In addition, the rate of tubular secretion in neonates is only 20-
30% that of an adult and the renal blood flow velocity is significantly lower [31,71].
Only 2-4% of a cocaine dose is excreted in the feces and bile.
As mentioned earlier, plasma and liver cholinesterases metabolize cocaine.
Therefore, any factor affecting esterase activity could have a significant effect on
cocaine metabolism. It has been reported that pregnant women, fetuses, and
people with liver disease have lower cholinesterase activity thus potentially
increasing the risk of cocaine toxicity in these groups [4,15]. In addition, due to
genetic polymorphisms, 1-13% of North Americans are considered to have little or
no cholinesterase activity . The formation of norcocaine by N-demethylation
was higher in subjects with lower cholinesterase activity . Ecobichon and
Stephens  were able to demonstrate that plasma pseudocholinesterase activity
in the neonate was 50-60% of that in the adult. Detectable levels of pseudo-
cholinesterase were observed at 28 weeks of gestation and activity increased
rapidly until 1 year of age when the neonatal levels reached those of adults. Their
observations suggested that premature neonates would have significantly different
rates of hydrolysis of ester-type drugs . The activities of the fetal drug
metabolizing enzymes are only 50-70% that found in adults and the half-lives of
various drugs in the neonate are significantly longer than the half-lives in adults .
Aspects of drug metabolism and monoamine transport by the placenta must
also be considered in this discussion. The placenta essentially serves as a sieve
for the passage of drugs from the matemal circulation to the fetus. Lipophilic drugs
readily cross the placental barrier and only drugs of very high molecular weight or
drugs that are highly ionized are unable to cross the membranes. Cocaine and
several cocaine metabolites readily cross the placenta via passive diffusion [63,74].
However, water soluble metabolites like benzoylecgonine do not diffuse well. When
formed in the fetus, these metabolites will tend to accumulate on the fetal side due
to ionization of the drugs since the pH of fetal blood is slightly lower than the pH of
The placenta also serves as a site of oxidative metabolism for some drugs
(i.e. ethanol and pentobarbital) including hydroxylation, demethylation, and
dealkylation reactions . The placenta has considerable amounts of smooth
endoplasmic reticulum and expresses several families of cytochrome P450. The
cytochrome P450 family CYP3A, including 3A3,4, and 7 is responsible for a portion
of cocaine metabolism. CYP3A4 and 7 are expressed in fetal liver and are also
expressed in placental tissue [75-77]. However, in studies by Schenker et al 
and Krishna et al  with perfused human placenta there was no evidence of
placental metabolism of cocaine. Conversely, Roe et al , using human placental
microsomes, showed a 20% decrease in cocaine concentration over a 130 minute
in vitro incubation period. In addition, using human placental villus tissue
homogenates, Ahmed et al  were able to identify a protein binding site with high
specificity for cocaine. Acetylcholinesterase has also been identified in human
placental villus tissue homogenates indicating the placenta may have the capacity
to metabolize cocaine; however, there is evidence that the placental
acetylcholinesterase is different both genetically and functionally from plasma
Human placenta expresses transporters for both serotonin  and
norepinephrine . These transporters are present on the brush border
membranes which are in contact with the maternal blood. While the purpose of
these placental transporters in fetal development has not been elucidated, in the
CNS, the disruption of monoamine transport by cocaine produces significant
physiological responses. The presence of these transporters may indicate the
placenta may actually serve as a primary target organ for cocaine and, therefore,
as an important mediator of the fetal effects of maternal cocaine use. Work with the
placental serotonin transporter has shown that cocaine is a potent competitive
inhibitor of the transporter . It is speculated that serotonin, a vasoconstrictor,
may be involved in regulation of the uteroplacental circulation; thus, the action of
cocaine at this transporter would result in increased concentrations of serotonin and
vasoconstriction of the placental vasculature .
Materno- Fetal Circulation and Pharmacokinetics
During gestation, the placenta serves as the system for fetal respiratory gas
exchange, nutrient supply, and removal of metabolic wastes. The fetal lungs are
filled with fluid until just prior to birth. The maternal and fetal circulatory systems are
separate; however, there is diffusion between the circulations via the placenta.
Oxygenated blood enters the placenta via the maternal uterine and ovarian arteries
and the maternal spiral arteries, where the oxygen is exchanged for carbon dioxide
with the fetal blood. Drugs used by the mother during pregnancy also diffuse from
the maternal circulation through this process. The placento-fetal circulation consists
of the umbilical cord, which has 2 arteries and 1 vein. Since gas and nutrient
exchange is handled by the placenta during gestation, circulatory adaptations are
present in the fetus which are then adjusted on delivery to the circulation present
throughout postnatal life. These adaptations include the ductus venosus, foramen
ovale, and ductus arteriosus. Oxygenated blood enters the fetus via the umbilical
vein and a substantial amount is shunted through the ductus venosus to the inferior
vena cava and right atrium; however, some blood enters the fetal liver through the
portal vein. Blood is then shunted from the right atrium through 2 pathways: (1)
through the foramen ovale, to the left atrium, and then to the ascending aorta. This
pathway provides the circulation to the fetal heart, brain, and upper extremities; (2)
to the right ventricle and then to the pulmonary circulation (less than 10% due to the
high pulmonary resistance) and through the ductus arteriosus to the descending
aorta and lower body. Blood then returns to the placenta via the umbilical arteries.
After birth, once the umbilical cord has been clamped, the ductus venosus closes,
the respiratory center is activated, and blood flow increases through the pulmonary
veins. The foramen ovale closes as left atrial pressure rises over right atrial
pressure and the ductus arteriosus flow reverses and then constricts due to the high
oxygen tension of the blood. If the ductus arteriosus remains patent after birth, it
must be either induced to close by administration of indomethacin (which promotes
closure by inhibition of prostaglandin synthesis) or, as a last resort, it must be
surgically ligated. A recent study has found that prenatal cocaine exposure is
associated with indomethacin failure and subsequent surgical ligation of the patent
ductus arteriosus (R. F. Covert, University of Chicago, personal communication).
Complete evaluations of human maternal-fetal cocaine pharmacokinetic
relationships have proven difficult to measure due to the complex technical issues
confronted. Factors that must be considered in such pharmacokinetic
determinations would include maternal cocaine absorption, distribution, and
elimination, fetal distribution and elimination, placental transfer (bi-directional),
placental disposition, and amniotic fluid disposition. Several compartmental models
have been proposed to model maternal-fetal drug kinetics. Szeto  proposed a
two-compartment model, one compartment representing the mother and the second
representing the fetus. Drug elimination from both mother and fetus would be
considered. Another possible compartmental model would include a
multicompartment arrangement with compartments for the placenta and amniotic
cavity (Figure 1-3).
Figure 1-3: Proposed maternal-fetal pharmacokinetic model.
(Modified from Devane ).
Physiological models have also been proposed [83,84]. These models
consider drug disposition to the tissues and organs of the body as a function of time
and can model altered physiological states. Luecke et al  have developed a
sophisticated, computer driven, physiologically based model for human pregnancy
that considers the rapid growth that occurs during pregnancy. However, since
available data from human pregnancy are generally limited to single time-point
measurements taken at delivery and since cocaine doses are only obtained from
patient self-report, most cocaine pharmacokinetic determinations have utilized
animal models. Sheep are often the animal model of choice given the relatively
large fetal size and the ability to cannulate fetal vessels and withdraw blood for
measurement of drug levels. This may, however, not be a good model for cocaine
metabolism and pharmacokinetics.
Druo Distribution to the Fetus
Maternal drug use during pregnancy is of great concern due to the rapid
passage of drugs across the placenta discussed earlier in this chapter. Therefore,
it would also be useful to determine to what extent the fetus is exposed to drugs
ingested by the mother. Many studies have observed that the fetal concentration
of a drug is lower than the maternal concentration. Drugs studied have included
methadone, morphine, meperidine, lidocaine, aspirin, and dilantin .
Information on fetal concentrations of cocaine is somewhat inconclusive.
Devane and colleagues  reported that the time of sample collection was an
important consideration when comparing maternal and fetal plasma cocaine
concentrations. Pregnant ewes were administered an IV bolus of cocaine in saline
at doses of 0.5-4.0mg/kg and both maternal and fetal plasma cocaine
concentrations were measured over time. Results showed higher maternal cocaine
concentrations until 4 minutes after the cocaine dose at which time the maternal
and fetal concentrations were of similar magnitude. The area-under-the-curve
(AUC) for the maternal samples was substantially greater than the fetal . A
separate study also utilizing pregnant ewes found that fetal blood concentrations of
cocaine observed 5 minutes after maternal infusion were approximately 12% of the
maternal level .
In pregnant rats, following a single intraperitoneal dose of cocaine (30mg/kg),
cocaine was distributed throughout the fetus with the highest concentrations found
in the placenta, followed by the fetal liver, maternal heart, fetal brain, and maternal
brain and plasma . Spear et al  reported fetal plasma cocaine
concentrations, following subcutaneous dosing in pregnant rats, to be 2 to 1 of
those of the dams. It was suggested that the difference could be explained by lower
plasma protein binding in the fetus. Dow-Edwards  administered a single dose
of cocaine (60mg/kg) by intragastric gavage to pregnant rats on day 22 of gestation
and cocaine concentrations were measured at time points for 90 minutes. Cocaine
plasma levels peaked at 15 minutes in both dams and fetuses and the mean peak
maternal concentration was 1.8 times higher than the fetal; however, maternal
cocaine levels decreased faster than fetal levels, and from 30-90 minutes fetal
levels were higher than maternal levels. Cocaine studies in pregnant mice showed
significantly lower cocaine concentrations in fetal tissues following a single cocaine
dose . However, chronic cocaine dosing in rabbits demonstrated fetal cocaine
concentrations near or above maternal levels . In that study, pregnant New
Zealand White rabbits received 6 doses of cocaine (0.5mg/kg IV) between days 22-
28 of gestation. Mothers were sacrificed on day 29 and the maternal and fetal
tissue cocaine concentrations were measured. Fetal brain mean cocaine
concentration was 20% higher than mean maternal brain concentration, but mean
maternal liver cocaine concentration was higher .
Chronic dosing in the pregnant guinea pig (chosen for its metabolic similarity
to humans) demonstrated a similar effect. In that study, pregnant Dunkin-Harley
guinea pigs were given a daily subcutaneous dose of cocaine (6mg/kg) from days
50-59 of gestation. Starting 1 hour after the last injection, sampling was begun as
follows: amniotic fluid collected from around each fetus, cord blood collected from
the umbilical veins, brains removed, maternal blood collected by cardiac puncture,
and maternal urine collected if possible. Maternal and fetal plasma cocaine
concentrations were not significantly different and the cocaine brain to plasma ratios
were also not significantly different. Higher benzoylecgonine concentrations were
found in maternal plasma, but fetal plasma contained higher concentrations of
benzoylnorecgonine. However, amniotic fluid cocaine concentrations were 3 times
that found in either the maternal or fetal plasma .
Controlled pregnancy studies are difficult to implement with non-human
primates due to their tendency to abort the fetus if it is surgically removed,
cannulated, and returned to the womb. In the only published report found, 4
pregnant baboons were studied at days 155-160 of gestation (term is 184 days).
Catheters were placed in the femoral arteries and veins of both the mothers and the
fetuses and catheters were also placed in the amniotic sac. Mother baboons were
administered a single dose of cocaine (2mg/kg IV) over a 2 minute period and then
blood and amniotic fluid concentration-time profiles were measured for 24 hours.
Serum cocaine concentrations were observed to decline in a biphasic manner in
both mother baboons and their fetuses; however, while the mean AUC for the fetal
samples was lower than the mean maternal AUC, the elimination half-life in fetuses
was 25% longer .
Effects of Gestational Cocaine Exposure on the Neonate
Given the documented physiological and toxicological effects of cocaine use
and the evidence that cocaine readily crosses the placental barrier, the effects of
maternal cocaine use in pregnancy on the fetus and newborn have been examined.
Documented labor and delivery complications associated with maternal cocaine use
have included spontaneous abortion [15,94], stillbirth [94,95], preterm labor and
delivery [96-100], premature rupture of the membranes [101,102], and abruptio
placentae [5,97,103,104]. The pathophysiological mechanism believed to be
responsible for the documented obstetric complications is the accumulation of
catecholamines in the periphery produced by cocaine. Accumulation of
catecholamines, in turn, produces vasoconstriction in the maternal and placental
circulation and increased uterine contractility which leads to placental abruption
[15,30,104]. In addition, intrauterine growth retardation has also been reported to
be significantly higher in cocaine users [102,105]. Placental insufficiency and
decreased blood flow produced by chronic cocaine use may be a major factor on
intrauterine growth regulation. Experimental evidence in pregnant sheep
administered intravenous doses of cocaine have demonstrated maternal and fetal
hypertension, decreased uterine blood flow, increased uterine vascular resistance,
elevated levels of serum catecholamines, and fetal hypoxemia [106-109]. Using
conscious fetal lambs, Covert et al  also reported reduced fetal cerebral blood
flow following cocaine administration (acute IV bolus) to the ewe, but increased
cerebral blood flow following cocaethylene administration. However, in work
involving acute IV continuous infusions of cocaine to either pregnant ewes or fetal
lambs, it has been reported that cocaine increased fetal cerebral blood flow .
In this work, cocaine administered to the ewe decreased fetal oxygen tension and
oxygen content; however, cerebral and myocardial blood flows were increased
which compensated for the effects of the fetal hypoxemia. Infusion directly to the
fetus did not alter cerebral blood flow .
On the surface, it would appear these independent reports yield opposing
results; however, there are very significant differences in the experimental
protocols which could account for these observations: (1) cocaine administration:
acute IV bolus (1mg/kg) versus acute continuous infusions lasting 30 minutes
(0.3mg/kg); (2) time of blood flow measurements: Covert et al  began blood
flow measurement immediately upon dosing while Burchfield et al  controlled
for behavioral state. Cocaine administration to either the pregnant ewe or directly
to fetal lambs has been reported to disrupt REM sleep in the fetal lambs [108,111].
Thus, Burchfield et al  monitored the behavioral state of the fetus and began
blood flow measurements only once the fetus had returned to REM sleep after the
infusion was stopped. Usually, a recovery time of 1.5-2 hours was required for the
fetus to resume REM sleep. REM periods were chosen for blood flow monitoring
because cerebral blood flow is as much as 35% higher during REM and REM sleep
is a metabolically active state .
Fetal hypoxemia, secondary to the placental vasoconstriction induced by
cocaine, may also be a contributing factor to the higher incidence of meconium-
stained amniotic fluid seen in pregnant women who have used cocaine during
pregnancy [113-116]. It has been suggested that meconium-stained amniotic fluid
is indicative of fetal stress. Meconium staining is associated with fetal cardiac
disturbances and aspiration of meconium by the fetus can result in respiratory
distress and infection.
Maternal cocaine use in pregnancy is also associated with adverse affects
in the neonate. These problems have included lower gestational age, lower birth
weight, length, or head circumference, a higher incidence of intrauterine growth
retardation and/or small for gestational age infants, fetal hypertension and
ventricular tachycardia, longer newborn hospital stay, and lower Apgar scores (a
measure of newborn health) with poorer general outcome [6,94,99-101,114,117].
Other medical problems reportedly associated with maternal cocaine use have
included retinopathy and ophthalmic abnormalities  and perinatal seizures and
cerebral infarction [8,119]. The risk of sudden infant death syndrome is reported
to be 3-7 times higher in newborns exposed to cocaine during pregnancy than in
the general population [5,96,120,121]. Case reports also indicate an increased risk
of necrotizing enterocolitis (NEC) in cocaine exposed newborns [122,123]. NEC is
a serious illness, characterized by ischemic necrosis of the intestinal tract
complicated by mixed-organism infection, with a high mortality rate . The
illness is generally seen in small, preterm infants and develops after the onset of
feeding and would be exacerbated by the vasoconstriction produced by cocaine.
The evidence that prenatal cocaine exposure is associated with a higher rate
of congenital malformations is weak. Bingol et al  studied 50 cocaine using
women and 110 polydrug abusers, and noted a higher malformation rate in the
cocaine group. However, the teratogenic information reviewed to date indicates
that prenatal cocaine use is not a significant teratogen and the only consistent
teratogenic findings have been those of the genitourinary tract [102,125,126].
Despite limited reports that infants and toddlers exposed to cocaine in utero
are more jittery, irritable, and inconsolable, and less organized and interactive ,
the long-term behavioral and neurological consequences of prenatal cocaine
exposure are unknown. In rodents and primates, gestational cocaine exposure was
associated with decreases in learning and memory . The National Institute on
Drug Abuse is currently funding longitudinal studies in Seattle, Pittsburgh, Miami,
and rural Florida where cocaine exposed newborns are followed for 3-5 years to
study their neurobehavioral outcomes . Results of these studies will provide
evidence for more definitive statements regarding the effect of prenatal cocaine
exposure on neurobehavioral progress.
Finally, prenatal cocaine exposure is also associated with increased fetal and
newborn deaths [96,127-129]. Fetal deaths in many cases were attributed to
placental abruption, respiratory distress, and prematurity. A majority of the mothers
in fatal cases had received no prenatal care. Two additional cases of newborn
death associated with cocaine using mothers, but not prenatal exposure, also
warrant attention as they are not isolated occurrences . One infant who had
been born cocaine positive later died from dehydration and malnourishment
following maternal neglect and a second newborn died from acute cocaine
intoxication resulting from cocaine laced baby formula. Both cases were classified
as homicides .
Immunotoxicolovy of Cocaine
In addition to the toxic effects of cocaine previously discussed, evidence
exists to suggest cocaine has a marked effect on the immune system. The
pharmacologic effect of cocaine on modulation of monoamine transport would
indicate that cocaine may have a significant effect on the regulation of the
hypothalamic-pituitary-adrenal axis and therefore, a direct effect on the immune
system. If cocaine is toxic to the developing fetal immune system, maternal cocaine
use in pregnancy could potentially increase the risk of the newborn acquiring
The immune system is characterized by mechanisms of innate and acquired
immunity that protect the body from attack by foreign substances. Innate immunity
is characterized by mechanisms always present in the body to act as barriers to
infection: mucous membranes, coughing, pH, fever, interferon, granulocytes, and
macrophages. Acquired immunity is the mechanism by which the body responds
to foreign substances (called antigens) after a previous exposure. Acquired
immunity functions through the activation of lymphocytes to respond to the antigen.
The cell types involved in acquired immunity include T and B lymphocytes and
macrophages. Human lymphocytes possess binding sites for dopamine as well as
adrenergic, cholinergic, and opioergic binding sites . These dopamine binding
sites resembled neuronal uptake sites and dopamine binding at this site was
inhibited by cocaine and other biogenic amine inhibitors . In a separate study,
cocaine and other monoamine uptake inhibitors were found to inhibit in vitro
mitogen stimulated B and T lymphocyte proliferation in rats . These reports
indicate cocaine may have the potential to modulate the immune response directly
through effects on lymphocytes as well as indirectly through hypothalmic-pitutitary-
adrenal axis neuroendocrine pathways .
The purpose of the developing fetal immune system is to produce
functionally mature B and T lymphocytes. T cells are responsible for cellular
immunity and recognize foreign antigens in association with a major
histocompatability (MHC) bearing accessory cell (macrophage) that has processed,
and presents, the antigen to the T cell. For example, resting T cells will recognize
cells that are infected with intracellular organisms like a virus or bacteria. The virus
or bacteria is endocytosed by the accessory cell and then the antigen surface
marker proteins and MHC molecules are expressed on the cell surface. The resting
T cell is then activated by its interaction with the accessory cell and the activated
T cells are then stimulated to produce and release lymphokines. B cells constitute
humoral immunity and are responsible for synthesis and secretion of
immunoglobulins (IgG, IgA, IgM, IgE, IgD). These immunoglobulins are able to bind
to circulating free antigens, including bacteria and toxins. The B cell response is,
however, dependent on cooperative interactions with activated T cells. Activated
T cells release lymphokines which are known as interleukins (IL) because of their
effects on other lymphocytes as well as their action as autocrine growth factors.
These interleukins include IL2, 4, and 5 and IL-4 and 5 activate the B cell response
. Additional evidence demonstrates that B cell differentiation may be
stimulated by direct cell to cell contact between B cells and T cells, specifically
CD4+ T helper cells, not just via lymphokines .
B and T lymphoctye development is initiated during the first trimester and
matures during the second trimester . A mature B or T cell is defined as one
that expresses a surface receptor marker for recognizing specific antigens.
Important T cell markers include CD4 or T helper cells and CD8 or T
suppressor/cytotoxic cells. T cells mature and differentiate in the thymus and the
thymus develops early in gestation. Lymphocytes are recognizable by week 10 of
gestation and by week 17, the distribution of T cell surface markers is essentially
that of the adult. By week 15, fetal T cells will respond to the mitogen
phytohemagglutinin (PHA). T cell maturation occurs only during fetal development
and for a short time after birth and the T cells then populate the secondary lymphoid
organs (spleen and lymph nodes). B cells are first observed in the fetal liver by 9-
10 weeks of gestation. B cells are then observed soon after in the fetal bone
marrow, spleen, and blood. At this time in fetal development, B cells only express
IgM and synthesize no other immunoglobulins. Later in gestation, maternal IgG
crosses the placenta and provides the fetus and neonate with some resistance to
infection until the newborn is able to synthesize immunoglobulins. B cells reach the
numerical adult number in the cord blood by week 15 of gestation and, by the
second trimester, B cell generation shifts to the bone marrow . Therefore, any
substance taken during gestation that alters this maturation process could severely
compromise the immune system of the newborn.
Although several studies have investigated the effects of cocaine on human
and animal immune functions, the relationship remains unclear. Studies evaluating
the effect of cocaine on cellular immune function have indicated that the drug
causes suppression of mitogenic responses in human and mouse T lymphocytes
. These studies, however, involved incubation of the cells with relatively high
concentrations of cocaine (12.5-300pg/ml). A complicating factor of many
published studies with cultured lymphocytes has been that the dose levels far
exceed blood plasma levels found in human cocaine users [51,53]. Cocaine has
also been shown to suppress proliferation of phytohemagglutinin (PHA)-activated
human peripheral blood lymphocytes at lower drug concentrations (0.9-12.OM)
. Previous work with cultured mouse splenocytes demonstrated that cocaine
induced suppression of PHA-activated cells was a biphasic process . More
significantly, cocaine (at concentrations observed in human cocaine abusers) was
demonstrated in vitro to augment human T-lymphocyte proliferation stimulated by
the anti-CD3 antibody . This was designed to more accurately mimic the in
vivo stimulation of T-cells via the T-cell receptor complex than is PHA activation.
Additional work demonstrated that the effect of cocaine stimulated T-lymphocyte
proliferation was modulated by alterations in calcium mobilization and IL-2
production . Recently, it was reported that cocaine impairs early activation of
CD4+ T-cells . The specific sub-population of CD4+ T-cells affected was the
population carrying a surface marker that targets the cells to migrate to the lymph
nodes. The in vivo effect of this impaired activation would potentially lead to greater
risk for the development of opportunistic infections.
In one of the more promising reports to date on the immunotoxicology of
cocaine, Ruiz et al  found alterations in specific T cell subpopulations in
cocaine-intoxicated individuals. Total numbers of natural killer cells, T and B
lymphocytes, and T cell sub-sets including memory CD8 T cells, activated T cells,
T helper cells, and T helperlinducer cells were counted using flow cytometry
combined with immunofluorescence staining targeted to surface marker antigens.
Results demonstrated a reduction in the total number of T cells in cocaine-
intoxicated individuals which was primarily a result of a decreased percentage of
CD4+ T helper cells. In addition, CD8+ "memory" cells were also reduced.
However, there was also a preferential stimulation of activated T cells in these
individuals. These results indicate that in cocaine-intoxicated individuals there may
be a dysfunctional immune response. In response to an immunological challenge,
a larger proportion of T cells would be activated but they would not evolve into
memory cells . The effect on the immunocompetence of human fetal T-cell
sub-populations would be of particular concern due to the chronic exposure of the
fetus to cocaine in utero.
Cocaine has been shown to stimulate both natural killer cell activity and
natural killer cell numbers in human volunteers who received an intravenous
infusion of cocaine  and also in cocaine-intoxicated patients admitted to a
hospital emergency room . However, an in vitro study indicated that cocaine
had no effect on natural killer cell activity and no effect on cytotoxic T lymphocytic
activity . Thus, cocaine may have differential effects in vivo on recruitment
versus activation of immune cell subpopulations.
Studies of humoral immunity have produced less definitive information on the
immunomodulatory effects of cocaine. Klein et al  reported that the
lymphoproliferative response to a B cell mitogen was not affected by cocaine at a
concentration which suppressed T cell proliferation. In another study, rats injected
with cocaine for a 10 day period were shown to have elevated B cell responses to
helper T cell independent and dependent antigens. However, in this same study,
in vitro exposure of lymphocytes to cocaine had no effect on B cell proliferation
. Previous work by Havas et al  in mice found that, even at lethal doses
of cocaine, the B cell response was elevated.
There is limited evidence to support indirect effects of cocaine on human B
cells. Elevated serum levels of IgG have been reported in IV drug users; however,
when human peripheral blood mononuclear cells (PBMC's) were cultured with
pokeweed mitogen (a B cell activator) and cocaine there was no effect of the
cocaine on IgG synthesis when compared to drug free controls . This indicates
cocaine, and other IV drugs, may induce B cell activation through indirect
mechanisms; however, the elevated IgG levels may be an effect of the presence of
more infectious agents seen in the IV drug using population . In a related
report in the rat, cocaine (1mg/kg) was administered IV to the jugular vein 2 times
per day for 12 consecutive days. Serum immunoglobulin levels were measured 24
hours after the last cocaine dose and compared to controls. Results demonstrated
significant increases in IgG levels in the cocaine treatment group and smaller,
although significant, decreases in serum IgA .
Finally, the potential role of cocaine as a co-factor in human
immunodeficiency virus (HIV) infection and in the development and progression of
acquired immunodeficiency syndrome (AIDS) has been raised [150-152]. The role
of cocaine as a co-factor in perinatal HIV transmission has also been raised .
Early indications are that HIV infection associated with cocaine users is a result of
intravenous drug use or sexual contact with an intravenous drug user [150,154].
In addition, female cocaine users also demonstrate increased incidence of sexual
transmitted diseases attributed to a more sexual promiscuous lifestyle [151,155].
However, after adjusting for sexual promiscuity and prostitution, the rates of
infection were still higher in cocaine using women indicating involvement of some
other factor . A recent study in Baltimore of 140 IV drug users found that 43%
of females and 57% of males were HIV positive. When correlated to drug usage,
the group with the highest infection rate was cocaine-preferring females with 58%
HIV positive . Cocaine has also been shown to potentiate the replication of the
HIV virus in human peripheral blood mononuclear cells in in vitro culture [157,158].
Cocaine and Alcohol Interactions
Many addicts report the concomitant use of ethanol and cocaine and report
a better "high" when both drugs are used [68,159]. In fact, Gorelick  reports
that clinical and epidemiological data indicate the co-abuse of these substances is
more common than might be expected and that the 1988 NIDA household survey
found that 80% of the cocaine users also reported ethanol use. An earlier survey
found that 30% of cocaine users drank ethanol every time they used cocaine. One
estimate indicates that as many as twelve million people may use these drugs in
combination in any one year . If the combination of cocaine and ethanol were
to increase the concentration of cocaine in the systemic circulation and the brain,
the toxicity experienced by the fetus could be exacerbated.
Ingestion of cocaine and ethanol leads to the in vivo formation of the active
metabolite cocaethylene (also known as ethylcocaine or ethylbenzoylecgonine).
This metabolite has been detected in the urine of cocaine users also testing positive
for ethanol [159,161]. In vitro studies indicate the formation of CE occurs via a
transesterification reaction in the liver catalyzed by hepatic carboxylesterase
forming a 2-carboxyethyl ester in place of the 2-carboxymethyl ester of cocaine
[66,162]. This enzyme is also reported to catalyze the conversion of cocaine to
benzoylecgonine. Controlled studies in vivo in rats and mice have found that serum
benzoylecgonine concentrations are lower with ethanol pretreatment indicating
ethanol inhibits the formation of benzoylecgonine while simultaneously catalyzing
the formation of CE [162,163]. The enzymatic formation of CE has been
demonstrated utilizing human liver extracts and human, mice, and rat liver
microsomal homogenates [66,163,164]. In vitro incubation of cocaine and ethanol
with homogenates of human kidney, brain, liver, lung, and placenta show that only
liver homogenates produce cocaethylene .
Cocaethylene retains the pharmacologic and toxicologic properties of
cocaine yet has a half-life of 2 hours compared to the 45-60 minute half-life of
cocaine . This metabolite has high affinity for the dopamine transporter in the
CNS and may enhance the euphoria experienced by cocaine users due to the
blockade of dopamine reuptake . However, the effect on modulation of the
serotonin system was significantly less than cocaine indicating CE is targeted to
dopaminergic function . An enhanced and prolonged euphoria was reported
by subjects administered both cocaine and ethanol. The plasma cocaine and
norcocaine concentrations were also significantly higher in the acutely intoxicated
individuals as compared to those receiving only cocaine . In primates, CE was
equipotent to cocaine in maintaining self-administration .
Because of the high rate of the co-abuse of ethanol and cocaine, one major
effect of the combination has been an increase in cocaine related hospital
admissions and sudden death [68,169,170]. It has been reported that the co-abuse
of these drugs increases the risk of cocaine-related sudden death 18-fold .
Significant levels of CE have been measured in postmortem blood and brain
samples from cocaine related deaths . In one study, urine and blood, collected
from 15 patients admitted to the hospital for trauma (motor vehicle accidents,
gunshot or stab wounds), were analyzed for cocaine and cocaine metabolites.
Cocaethylene was detected in the plasma of 13 of the 15 patients and some
samples had only cocaethylene detected, no cocaine . Cocaethylene has also
been detected in the urine of newborn babies [8,169]. This cocaine metabolite is
believed to mediate cocaine related toxicity and lethality [171,172] exhibiting an
LDs in mice of 60.7mg/kg versus 93.0mg/kg for cocaine. Cocaethylene has also
been found to be toxic to liver hepatocytes  and ethanol has been found to
potentiate the hepatotoxicity of cocaine [174-176].
Although the cocaine/ethanol metabolite cocaethylene will be used in
immune function studies, the role of ethanol alone will not be studied as it has been
extensively reviewed [177-179].
Analytical Methods for Detection of Gestational Cocaine Exposure
Due to the potential for adverse effects of gestational cocaine exposure,
efficient and sensitive methods of detection are needed to identify those newborns
exposed. The most common analytical method for detection of drugs of abuse is
a urine screen using an immunoassay technique, followed by solid-phase extraction
of the drug from the biological matrix and structural confirmation with gas
chromatography-mass spectrometry. Immunoassay is a rapid, relatively
inexpensive screening procedure which is designed primarily to eliminate drug free
samples. However, immunoassay is fairly non-specific and any sample which tests
positive for a particular drug must be confirmed by a more specific chromatographic
analysis. Immunoassay is based on the reaction of an antibody with the drug for
which the sample is being tested, ultimately producing a measurable response. In
the case of cocaine, the immunoassay test involves the reaction of an antibody with
benzoylecgonine, the principle hydrolysis product of cocaine in adult urine. This
antibody is much less cross-reactive towards cocaine or other metabolites such as
norcocaine or EME. Immunoassay methods commonly used include enzyme
multiplied immunoassay (EMIT), fluorescence polarization immunoassay (FPIA),
and cloned enzyme donor immunoassay (CEDIA) which are homogeneous assays
requiring no separation step of bound drug from unbound drug.
Radioimmunoassay (RIA) is also used and is a non-homogeneous assay requiring
separation of bound from unbound drug. For urine testing, the legal and
commercially established immunoassay cut-off limit for benzoylecgonine is
Chromatographic techniques used for analysis of cocaine and metabolites
have included both gas chromatography-mass spectrometry (GC-MS) and high
performance liquid chromatography (HPLC). GC-MS provides the highest degree
of specificity and is the accepted "gold standard" method of analysis for cocaine,
benzoylecgonine, and EME [56,69,180-185]. The main disadvantage to GC-MS is
that compound derivatization is routinely required prior to analysis. The use of
HPLC is becoming more common as the advent of diode-array and multiwavelength
detectors has improved the selectivity of the method giving ultraviolet (UV)
absorption profiles and derivative spectral data for each peak in the chromatogram.
Very sensitive HPLC methods have been developed for cocaine detection and
quantitation in a variety of biological matrices [18,92,186-189]. The main
disadvantages of HPLC are that EME is not detectable using UV detection and that
matrix interference is more common than that seen with GC-MS.
Urinalysis of newborn urine has limitations. In most cases it only indicates
recent use (within 48-72 hours) of the drug by the mother. This sample may be
highly inaccurate in newborns due to the elapsed time from presentation of the
mother for labor, delivery, and subsequent urine collection in the nursery. However,
when the urine of 70 newborns born to cocaine using women was sampled every
8 hours for 6 days after delivery, benzoylecgonine was detected in some newborns
up to 120 hours (mean 80 hours) after delivery . In the case of detection of
cocaine exposure in newborns, it has been suggested that meconium, the first feces
of the baby, is a better indicator of cocaine exposure [190,191]. It is hypothesized
that drugs in the maternal-fetal circulation may accumulate in the meconium from
the time it is first produced at about week 18 of gestation. Since the initial reports
by Ostrea et al , other investigators have confirmed that meconium was a
better biological matrix than urine for detection of gestational cocaine exposure
[192-195]. However, a recent report employing structured drug use interviews
found that meconium testing did not improve detection of cocaine use occurring in
early to mid-gestation; in fact, overall results found testing of urine (maternal or
fetal) and meconium produced equivalent results when the same analytical methods
were used. Cocaine use had to have occurred in the three weeks prior to delivery
to elicit a positive cocaine test using meconium .
Amniotic fluid may also be a good biological matrix for detection of prenatal
cocaine exposure. Reports from cocaine studies using guinea pigs, sheep, and
humans indicate amniotic fluid often contains high concentrations of cocaine and/or
cocaine metabolites [92,109,129,197-199]. Since the concentrations in amniotic
fluid are often much higher than in maternal urine or serum, cocaine and
metabolites may accumulate in this medium. Since the fetus is exposed to the
contents of the amniotic fluid, both orally via swallowing and transdermally, during
the gestational period, the determination of the presence and concentration of
compounds in this fluid may be significant as it would prolong fetal exposure to the
toxic effects of cocaine. Using esophageal ligation in fetal lambs, Mahone 
demonstrated cocaine and metabolites were still detected in the fetal plasma when
cocaine was infused into the amniotic fluid. It has been reported that the fetus
swallows between 200-450ml of amniotic fluid per day and urinates as much as
500ml per day into the amniotic cavity . Since the fetus urinates into the
amniotic cavity, the presence of benzoylecgonine in the amniotic fluid is indicative
of cocaine metabolism by the fetus since benzoylecgonine does not cross the
placenta from mother to fetus [74,109].
The above review of the literature indicates the following: that fetal
metabolism and elimination of cocaine differs from that of adults, the concomitant
use of alcohol with cocaine may exacerbate the toxic effects of cocaine, and
cocaine has a modulatory effect on the immune system. This research was
therefore designed to investigate the following hypotheses.
Hvyotheses for this Study
(1) The main hypothesis is that the fetus is exposed to higher concentrations of
cocaine than the mother during maternal cocaine use in pregnancy.
(2) The ingestion of cocaine with concurrent ethanol use further increases
cocaine concentration. In pregnancy, cocaine concentration is increased in
mother and fetus.
(3) Cocaine and its metabolites have an adverse effect on the immune
system (in vitro).
1. To develop and validate analytical methods for the analysis of cocaine and
cocaine metabolites benzoylecgonine (BZE), norcocaine (NC), and
cocaethylene (CE) in biological samples.
2. To characterize maternal/fetal cocaine metabolism and distribution using
both human samples and animal models.
3. To characterize cocaine metabolism and distribution with the concurrent use
of ethanol using animal models.
4. To examine the effect of cocaine, benzoylecgonine, norcocaine, and
cocaethylene on T and B lymphocyte proliferation (in vitro).
Human biological fluid and tissue samples from were provided by subjects
identified by neonatologists at the University of Illinois Hospital (Chicago, Illinois),
the University of Chicago Hospitals (Chicago, Illinois) or at Shands Hospital
(Gainesville, Florida). The use of these samples was approved by the Institutional
Review Boards at all participating universities. Animal studies were conducted at
the University of Florida Gainesville and approved by the Animal Care Committee.
Amniotic fluid (1-5ml) was collected at the time of membrane rupture,
delivery, or amniocentesis. Upon arrival in the nursery, a urine bag was placed on
the infant to collect a sample of the first voided urine (1-5ml) and before 24 hours
of age. Urine samples were then transferred to a plastic sample container. The
newbom's meconium was collected from the diaper and placed in a plastic sample
container. All samples were stored frozen at -20C until analysis. Cord blood (5-
20ml) was obtained after delivery and placed in a heparinized blood collection tube.
Serum and tissue samples from animal studies were stored at -20*C until analysis.
Specific Aim #1
To develop and validate analytical methods for the analysis of cocaine and
cocaine metabolites benzoylecgonine, norcocaine, and cocaethylene.
Methanol, acetonitrile, and potassium phosphate monobasic (all HPLC
grade) were purchased from Fisher (Fair Lawn, NJ). Diethylamine, butylamine,
chloroform, isopropanol, sodium hydrogen carbonate, sodium hydroxide, acetic
acid, hydrochloric acid, ammonium hydroxide, and tris buffer were reagent grade
(Fisher, Fair Lawn, NJ). Lipase was reagent grade (Sigma, St. Louis, MO).
Cocaine hydrochloride, benzoylecgonine, and bupivacaine were obtained from
Sigma (St. Louis, Mo). Cocaethylene and norcocaine were obtained from the
National Institute on Drug Abuse (NIDA, Rockville, MD).
Fluorescence Polarization Immunoassav Screening
Instrumentation, materials, and calibration
An Abbott Laboratories ADT fluorescence polarization immunoassay
instrument (Abbott Park, IL) was used for screening of urine, amniotic fluid, and
meconium extracts. This technology uses specific antibodies in the reagent
systems that react with free drug (benzoylecgonine) in the sample to form an
antibody-drug complex. Therefore, when analyzed by this method, a "positive"
sample indicated the presence of cocaine metabolite. The reagent kit contained all
components needed for the reaction of the antibody with any free drug in the
sample which included: antibody, antigen-tracer, and a pretreatment which
prepared the sample for analysis. Kits for this system have been developed by the
manufacturer for analysis of either urine or serum. Dilution buffer, reagent kits,
calibrators, and controls for benzoylecgonine were purchased from Abbott (Abbott
The ADTM was calibrated using 6 calibration solutions prepared in human
urine with a benzoylecgonine concentration range of 0-5000ng/ml. Low, medium,
and high control solutions of benzoylecgonine were included in all runs for quality
control monitoring. All calibrators, controls, and samples were diluted with an equal
volume of buffer (pH 7.7) during sample analysis.
Cocaine hydrolysis experiment
Since the ADxM screens samples only for the presence of benzoylecgonine,
any sample which contained only cocaine would screen "negative". Previous work
however indicated high concentrations of the parent drug (cocaine) in meconium
and amniotic fluid. This experiment was therefore designed to find the optimal
conditions at which cocaine (spiked into urine) was completely hydrolyzed to
benzoylecgonine allowing its subsequent detection by immunoassay techniques.
Drug free urine was obtained from a volunteer. A 100lg/ml sample of cocaine
hydrochloride was prepared by solubilizing 0.01g cocaine hydrochoride in 100ml of
drug free human urine. Standards were prepared by dilution of the stock solution
with human urine to yield cocaine concentrations of 1000ng/ml and 100ng/ml.
Samples were incubated over a range of pH values of 8.1-10.4 (the pH adjusted
using concentrated NaOH), for time periods from 15-60 minutes, in erlenmeyer
flasks placed in a warm water bath (50C). After the incubation period, the pH of
the sample was immediately adjusted using 6M HCI to a final pH range of 8.1-10.4
for analysis by ADTM. Samples were analyzed by ADTM and reported values of
benzoylecgonine were compared to the known concentrations.
Clinical samples of urine and meconium
Subsequently, a series of actual clinical samples of newborn urine and
meconium were screened by both the direct and the hydrolysis method to
determine if hydrolysis provided improved detection of gestational cocaine use. Ten
samples of urine and 5 samples of meconium were used. One aliquot of the urine
sample was analyzed directly by ADTM. A second aliquot of urine (400pl) was
adjusted to pH>9.5 with concentrated NaOH, incubated for 60 minutes at 50"C, and
then adjusted to pH 8.0 for analysis by ADx. The following procedure was used
for meconium analysis by ADx: (1) The meconium sample was weighed and
placed in a test tube (12X75mm); (2) Either 100% methanol (2ml) was added (for
direct analysis) or 3% ammoniacal methanol (2ml) was added (for hydrolysis); (3)
The mixture was vortexed and centrifuged for 10 minutes at 1000g; (4) Followed by
incubation for 60 minutes at 50C; (5) The supernatant was then transferred to a
clean test tube and evaporated to dryness; (6) The residue was reconstituted in
ADTM buffer (0.5ml) prior to analysis by ATM Benzoylecgonine concentrations
measured using both methods were then compared.
Solid Phase Extraction (SPEI Procedures for Biological Samples
Meconium (0.5-1.0g) was extracted by suspension and vortex mixing with
3ml of methanol. After centrifuging for 10 minutes at 1000g, the remaining
methanolic supernatant was transferred to a stoppered tube and vortexed with
0.025M potassium phosphate buffer (1ml) at pH 3 and 10pl of internal standard
(bupivacaine 100pg/ml). This extract was then applied to a Strong Cation
Exchange column (SCX) (Varian, Harbor City, CA.) with a capacity of iml, which
had been conditioned under vacuum on a Vac Elut manifold (Varian) with methanol
(2ml), water (Iml) and 0.25M phosphate buffer (1ml). After application of the
sample, the column was air dried for approximately 30 seconds and then washed
with phosphate buffer (1ml) and 0.1M acetic acid (0.5ml). The column was again
air dried for 30 seconds before eluting the adsorbed drugs with 3% ammoniacal
methanol (2ml). The final extract was evaporated to dryness under nitrogen.
An aliquot of urine (0.5-1ml) was mixed with an equal volume of 0.025M
potassium phosphate buffer at pH 3 and also with 10pl of internal standard
(bupivacaine 100pg/ml) and then applied directly to an SCX extraction cartridge and
extracted as described above for meconium. Extracts were evaporated to dryness
Whole blood or cord blood
Due to the lipophilic nature of whole blood samples, a more polar extraction
column was utilized for the successful extraction of cocaine and metabolites from
this matrix. Whole blood or umbilical cord blood (0.5-1ml) was extracted using the
following procedure: Bond Elut (Varian, Harbor City, CA.) columns, containing C2
packing material with a capacity of 3ml, were positioned on a Vac Elut vacuum
manifold and conditioned with methanol (3ml), followed by 0.1M sodium hydrogen
carbonate (3ml) at pH 8.5. The blood sample was diluted with Iml of carbonate
buffer (pH 8.5) and 10pl of internal standard (bupivacaine 100pg/ml) and then
applied to the extraction column and allowed to dry for 30 seconds before it was
washed with carbonate buffer (3ml), followed by 1ml of 5% v/v methanol in water.
The absorbed drugs were then eluted from the column with 6 X 0.250 ml of
chloroform : isopropanol (4:1). The extracts were evaporated to dryness under
The viscosity of amniotic fluid required the use of a high flow extraction
cartridge as follows: Amniotic Fluid (1-5 ml) was added to phosphate buffer (1-5 ml,
pH 6) and 10pl of internal standard (bupivacaine 100pg/ml). The mixture was
applied to an X-TracT column (United Chemical Technologies, Horsham, PA), with
a capacity of 15ml, previously conditioned with methanol (5ml), water (3ml) and
phosphate buffer (3ml) at pH 6. The sample was drawn through under vacuum and
the column air dried for 30 seconds. The column was then washed with water
(2ml), 100mM hydrochloric acid (2mL), and methanol (3ml). The column was again
allowed to dry before eluting adsorbed compounds with 10ml of a mixture of
chloroform : isopropanol : ammonium hydroxide (78:20:2). The eluent was
evaporated to dryness under nitrogen.
Two methods were used to extract drugs from brain tissue. Due to its
lipophilic nature, brain tissue must be digested prior to extraction. In method 1, the
brain sample was weighed in a sample vial. Then, 3ml of Tris buffer (0.2M) was
added to the vial along with 1mg of lipase and 10pl of internal standard
(bupivacaine 100pg/ml). The mixture was homogenized using a mechanical
homogenizer and then incubated at 60C for 2.5 hours. The sample was then
extracted using Bond Elut (Varian) columns containing C2 packing material and with
a capacity of 1 ml as described above for whole blood. In method 2, the brain tissue
was weighed in a large, conical tube. Then 3ml of 0.025M potassium phosphate
buffer at pH 3 and 10pl of internal standard (bupivacaine 100pg/ml) were added.
The sample was homogenized using a mechanical homogenizer and centrifuged
for 10 minutes at 1500g. The supernatant was then extracted using SCX columns
with a 1ml capacity as described above for urine. Extraction recovery was
determined using both for comparison.
Serum or plasma
Serum and plasma from humans or animals was extracted following the
method described above for urine.
Placenta and liver
Placental and liver tissue were extracted using method 2 described above
for brain tissue and using C2 extraction columns with a Iml capacity.
High Performance Liquid Chromatoaraphv (HPLC)
The initial mobile phase used was that of Browne et al  which consisted
of 0.025M potassium phosphate buffer: acetonitrile (85:15) containing diethylamine
(25ml/L) with the final pH adjusted to 2.9 with concentrated orthophosphoric acid.
Mobile phase 2 consisted of 0.025M potassium phosphate buffer (with 2%
butylamine): acetonitrile (78:22) with the final pH adjusted to 3.0. Mobile phase 3
consisted of 0.025M potassium phosphate buffer : acetonitrile : butylamine
(78:20:2) with the final pH adjusted to 3.0. Mobile phase 4 consisted of 0.025M
potassium phosphate buffer : acetonitrile : butylamine (81:18:1) with the pH
adjusted to 3.0 with concentrated orthophosphoric acid. Solvents used were of
HPLC grade and were degasssed by bubbling with helium prior to use.
Two analytical columns were utilized throughout this research. Column 1
was a Waters (Marlborough, MA) Novapak C18, 4um, ODS column (15cm x 3.9mm
ID). Mobile phase was delivered to this column at 0.5 ml/min. Column 2 was an
Alltech (Deerfield, IL) Lichrosorb RP18 10um column (25cm X 4.6mm ID). Mobile
phase was delivered to this column at 1.5 ml/min.
Analysis was performed using two separate systems. System 1 used a
Waters (Marlborough, MA) Model 510 pump to deliver mobile phase to the
analytical column. A C18 Novapak Guard Pak precolumn (Waters, Milford, MA)
was used to protect the analytical column. The sample was injected using a Waters
model UK6 universal liquid chromatograph injector equipped with a 50pI sample
loop. The detector was a Spectra Physics Focus multiwavelength detector (Thermo
Separation Products, Winter Park, FL) with an IBM Personal System/2 data system
for acquisition, integration, and processing. The eluent was monitored at 230, 255
and 275nm and full spectra were recorded from 190 to 400nm for each peak.
Quantitative analysis was achieved by comparison of peak areas of unknowns with
extracted standards. All peak areas were reported as relative peak areas by
comparison to the peak area of the internal standard (bupivacaine 1 pg/ml). Each
determination was taken as the mean of duplicate injections in most cases. The
calibration curve was produced over the range 0.05-2pg/ml. Method validation
studies were completed as described below.
HPLC system 2 used two Waters Model 501 pumps to deliver mobile phase
to the analytical column. A Waters C18 Novapak Guard pak precolumn was used
to protect the analytical column. The sample was injected using a Waters WISP
7108 autosampler with a capacity for 48 samples. The sample size injected was
50pl. The detector was a Waters model 486 tunable absorbance detector and the
eluent was monitored at 230nm. Instrument control, data acquisition, processing,
and custom reporting was handled using a NEC Powermate 386 computer and
Waters Millennium 2010 (version 2.0) Chromatography Manager software (Waters,
Marlborough, MA). In addition, this system also included Waters Millennium
System Suitability software which provided for trend plotting, quality control, and
method validation following GMP/GLP regulatory protocols. Therefore, in addition
to the method validation procedures described below and performed with System
1, a GLP method validation program was developed for the analysis of cocaine and
metabolites via HPLC. With this program, the chromatographic system was
validated prior to unknown sample analysis to insure that samples were analyzed
on a validated method. If the method failed the validation test, problems were
corrected before analysis of unknowns.
Method Validation and Quality Control
Standard analytical procedures
Standard curves for quantitative analysis were constructed using bupivacaine
as the internal standard for both HPLC analysis and for GC-MS. In each case the
range of the standard curve, precision and accuracy, specificity and sensitivity were
determined using the following standard procedures.
Extraction recovery of cocaine, benzoylecgonine, norcocaine, and
cocaethylene was determined by comparison of the peak area observed for a
nonextracted standard solution in an appropriate solvent injected directly onto the
HPLC to the peak area observed with a solution prepared at the same
concentration in the appropriate biologic matrix and injected onto the HPLC
following extraction. Extraction recovery was determined from all biological
matrices used with five replicates at each concentration.
Range of standard curve and control samples
The standard curve was constructed so that the maximum calibrator was
>20% of the expected maximum concentration and the minimum calibrator was at
least 10% > than the minimum quantifiable value (10 x baseline noise). A 6-8 point
calibration curve was prepared in the range 100-1750ng/ml for each compound
(cocaine, benzoylecgonine, norcocaine, and cocaethylene) with a constant
bupivacaine concentration. Control samples were included in the analysis of
unknown samples by injection of a control for every 5-10 unknown samples
injected. Stock solutions of 1mg/ml of cocaine, norcocaine, benzoylecgonine,
and bupivacaine and 1.5mg/ml of cocaethylene were prepared in methanol. A
mixed standard stock solution containing 10pg/ml of COC, BZE, NC, and CE was
then prepared in methanol. Calibration standards in the range 100-1750ng/ml of
COC, BZE, NC, and CE and 1000ng/ml of bupivacaine were prepared in methanol
and evaporated to dryness. Standards were reconstituted in the HPLC mobile
phase (for direct injection) or in the biological matrix (in preparation for solid-phase
Linearity and precision of standard curves
The standard curve was determined from the calibrators by a linear least
squares fit to the equation y = mx + b, where x = concentration and y = ratio of drug
peak area to internal standard peak area. The regression line was not forced
through zero. An "r" value between 0.97 to 1.0 was considered to represent
acceptable linearity. Precision was assessed from the slope of replicate standard
curves run on the same day and over a minimum two week period. The y-intercept
must be less than 25% of the value of the minimum standard curve calibrator.
Throughout the study period, changes in curve slope, y-intercept and residuals were
assessed to check for occurrence of proportional or determinate error.
Accuracy and precision
Intraday and interday precision were determined on the standards. Replicate
(n = 7) analysis of control samples was performed on the same day and over a
minimum period of two weeks. Precision is expressed as the relative standard
deviation of the intraday and interday replicate analysis for each control sample.
Accuracy is expressed as the relative difference between the actual value and the
mean measured intraday and interday value for each control. Minimal acceptable
accuracy and precision criteria are defined based on the specific requirements of
each study and for this study the minimal acceptable variation was 15% for interday
The limit of detection of each assay procedure was defined as three times
the signal to noise ratio and the limit of quantitation as five times the signal to noise
ratio determined by repetitive analyses of a drug free specimen. Sensitivity was
documented by replicate analysis of blank (drug free) samples and samples
containing drug at minimal detectable and minimal quantifiable concentration in the
appropriate biological matrix.
Samples of the appropriate biological matrix known to be drug free were
analyzed both with and without a known quantity of drug and/or metabolites added
to the sample to demonstrate lack of interference from endogenous materials.
Resolution of chromatographic peaks of drug, metabolites, internal standard and/or
other eluting compounds was determined.
Gas Chromatography-Mass Spectrometry / HPLC Comparison Experiment
A Finnigan MAT Incos 50 quadruple mass spectrometer (San Jose, CA),
Hewlett-Packard Model 7673A autosampler and Hewlett-Packard Model 5890 gas
chromatograph (Palo Alto, CA) were used for quantitative analysis of CO, BZE, NC,
BUP, and CE. The compounds were separated on a 30 m X 0.32 mm i.d. fused
silica capillary column coated with a 0.25pm film thickness of bonded-phase methyl
silicone (DB-1) (J&W Scientific, Folsom, CA). The injector was operated in the
splitless mode at 2800C, and helium was used as carrier gas at a column head
pressure of approximately 6psi and flow rate of 1ml/sec. Specimens were injected
at 500C, the split valve opened after 1 minute, and the temperature was increased
to 2800C at a rate of 33C/min. The ion source was operated at 1800C, with an
accelerating voltage of 70eV, 750pA filament current, and 1 kV electron multiplier
voltage. Ion current was acquired at the following masses for cocaine and
benzoylecgonine: CO: 82, 182, 303 and BZE: 300, 316, 421. A multiple ion
descriptor was determined for the other metabolites before other studies were done.
The total scan time was 0.439 seconds.
Multiple ion descriptor
A multiple ion descriptor was determined for all compounds, including the
internal standard. Using 100pg of drug diluted in 100pl of dimethylformamide
(DMF), each sample was chromatographed and the full range was scanned. The
multiple ion descriptor was selected based on the fragmentation pattern of the
compounds. A minimum of three ions were chosen. Individual metabolites (except
BZE) were also subjected to a derivitization procedure as described below to
determine if the norcocaine, bupivacaine, or cocaethylene would derivitize.
Each sample for GC/MS was transferred to a silanized Reacti-vialTM(Varian,
Harbor City, CA) following SPE and the eluting buffer was evaporated at 60-700C
under a gentle nitrogen sweep. To this residue, 50pL of pentafluoropropionic
anhydride (Aldrich, Milwaukee, WI) and 25pL of pentafluoropropanol (Aldrich) was
added. The vial was tightly capped, vortex-mixed, and the derivatization reaction
allowed to proceed for 20 minutes at 780C, after which time the reagent was
evaporated under a nitrogen sweep. The residue was reconstituted in 50pL of
dimethylformamide prior to GC/MS analysis.
A six point calibration curve was prepared in urine. A sample of drug-free
urine was obtained from a volunteer. The urine was spiked with cocaine,
benzoylecgonine, norcocaine, and cocaethylene to prepare a stock solution
(10pg/ml). Bupivacaine (0.5pg/ml), the internal standard, was added to individual
standards and samples prior to SPE. Concentrations were extracted in the range
of 0-1500ng/ml. Samples were extracted via the solid-phase extraction method
described above for urine. The SPE eluant was then evaporated to dryness under
nitrogen, reconstituted in 100pl of methanol, and split into 2x50pl samples. These
samples were then evaporated to dryness and one sample was reconstituted in
50pl HPLC mobile phase and subjected to HPLC analysis. The second sample was
derivitized as described above, reconstituted in 50pl DMF, and subjected to GC/MS
A calibration curve for each compound was prepared on three separate days
over a 2 week period using this method. Then, 10 urine samples identified as
known cocaine positive were also extracted and analyzed by the procedure
described for standards. Finally, interday variability was also determined by
replicate analysis of 7 standard samples, split and analyzed according to the above
procedure. The HPLC used for this experiment was System 2 described above.
A statistical comparison of the GC/MS and HPLC data was performed for
comparison of sensitivity, precision, quantitation, and reproducibility using paired T-
tests and a standard error of the estimate determination.
Specific Aim #2
To characterize cocaine metabolism and distribution in the maternal-fetal unit
using both human samples and animal models.
Human Studies Meconium and Urine
Two studies were done. In study 1, samples of infant urine and meconium
were collected from 22 consecutive live births. Urine samples were screened using
the Abbott ADT analyzer described above. Any positive samples were confirmed
by HPLC, following solid-phase extraction, as described above under methods for
specific aim #1. Meconium samples were analyzed by HPLC only (following solid-
phase extraction (SPE)). In the second study, comprehensive meconium testing
was compared to targeted urine immunoassay screening. For this study, all
newborns delivered at a large, urban community hospital over a one month period
and all newborns delivered at a large, suburban community hospital over a one
month period were included. The total number of newborns studied was 312.
Meconium was collected from all newborns and analyzed by HPLC following SPE.
Newborn urine samples were collected only from those deliveries considered as
high risk for gestational cocaine exposure. Risk factors identified were history of
cocaine use, history of other illicit drug use, preterm labor, no prenatal care,
unexplained hypertension, and abruptio placentae. Targeted urines were screened
by immunoassay. The incidence of cocaine exposure was determined by each
method and the test sensitivity was then calculated for each method based on the
total number of newborns subsequently assigned to the "cocaine exposed" group.
Incidence and sensitivity were evaluated for statistically significant differences using
Human Study Amniotic Fluid. Cord Blood. and Newborn's Urine
Samples of amniotic fluid, cord blood, and infant urine were obtained from
14 women who either reported a history of cocaine use at some time during
pregnancy or who were strongly suspected of such use. These individuals were
identified by the neonatologists. Samples were collected as described above at the
beginning of the methods chapter. Cord blood (0.5-1ml) was analyzed by HPLC
only (following SPE). Urine samples were screened using the Abbott ADTM
analyzer and any positive samples were confirmed by HPLC (following SPE). All
amniotic fluid samples were analyzed by HPLC (following SPE); however, these
samples were also screened with the Abbott ADxT analyzer for method
development purposes. All samples were analyzed for detection and quantification
of cocaine and cocaine metabolites.
Specific Aim #3
To characterize cocaine metabolism and distribution with the co-ingestion of
High Dose Cocaine--Mice
Two groups of male ICR mice (Harlan, Indianapolis,IN) with 18 mice per
group were used. Group 1 mice were dosed with cocaine (55mg/kg IP). Group 2
mice were dosed with cocaine (55mg/kg IP) and pretreated with ethanol (3g/kg
gavage) 30 minutes prior to the cocaine dose. Mice were euthanized with CO2
asphyxiation at time points of 15, 30, 60, 90, 120, and 180 minutes with 3 mice
sacrificed at each time point. After sacrifice, blood was collected by cardiac
puncture and the serum (300-500pl) was isolated. The brain was also harvested.
Samples were stored frozen at -20C until SPE and HPLC analysis and quantitation.
Statistical analysis was performed using a one-way ANOVA to compare the mean
drug concentration at each time point between groups 1 and 2.
Low Dose Cocaine--Mice
Two groups of male ICR mice (Harlan) with 18 mice per group were used.
This experiment was designed as under "high dose" above except the cocaine
dosage was lower and designed to more closely approximate human physiological
attainable cocaine concentrations. Group 1 mice were dosed with cocaine
(10mg/kg IP). Group 2 mice were dosed with cocaine (10mg/kg IP) and pretreated
with ethanol (2g/kg gavage) 30 minutes prior to the cocaine dose. Mice were
euthanized with CO2 asphyxiation at time points of 15, 30, 60, 90, 120, and 180
minutes with 3 mice sacrificed at each time point. After sacrifice, blood was
collected by cardiac puncture and the serum (300-500pl) was isolated. The brain
was also harvested. In addition, serum ethanol concentrations were measured in
a 10pl aliquot using a commercially available kit (Sigma, St. Louis, MO). Samples
were stored frozen at -20C until SPE and HPLC analysis and quantitation. Area-
under-the curves for the time versus concentration plots were calculated using the
trapezoidal rule and compared for statistically significant differences using a one-
way ANOVA. The time to peak concentration (Tmax) and the peak concentration
(Cmax) were determined for each group from the plot of time versus concentration.
Two groups of female ICR timed-pregnant mice (Harlan) were used with 15
mice per group. On day 17 of gestation, Group 1 mice were dosed with cocaine
(5mg/kg tail vein IV ). On day 17 of gestation, Group 2 mice were dosed with
cocaine (5mg/kg tail vein IV) and pretreated with ethanol (2g/kg gavage) 30 minutes
prior to the cocaine dose. Mice were euthanized with CO2 asphyxiation at time
points of 15, 30, 60, and 90 minutes with 4 mice sacrificed at each time point. After
sacrifice, maternal blood was collected (300-500pl) by cardiac puncture and the
maternal brain was harvested. The fetal brains were also harvested. All fetal
tissues from the same dam were pooled and then split into 2 separate samples if
sufficient tissue had been collected. Samples were stored frozen at -20C until SPE
and HPLC analysis and quantitation. Statistical analysis was performed between
the groups using a one-way ANOVA to compare maternal and fetal brain cocaine
concentrations and maternal blood cocaine and benzoylecgonine concentrations.
In addition, elimination rates for cocaine were calculated on brain cocaine
concentrations over time using a semi-log plot of time versus cocaine concentration.
Elimination rates were evaluated by independent t-tests for statistically significant
Benzovlecaonine Dosed Mice
To examine BZE transport across the blood brain barrier, male ICR mice
(Harlan) were divided into 2 groups as described above for cocaine dosing. Group
1 mice were dosed with BZE (10mg/kg IP). Group 2 mice were dosed with BZE
(10mg/kg IP) and pretreated with ethanol (2g/kg gavage) 30 minutes prior to the
BZE dose. Mice were euthanized with CO2 asphyxiation at time points of 15, 30,
60, 90, 120, and 180 minutes with 3 mice sacrificed at each time point. After
sacrifice, blood was collected by cardiac puncture and the serum (300-500pl) was
isolated. The brain was also harvested. Samples were stored frozen at -200C until
SPE and HPLC analysis and quantitation. Statistical analysis was performed using
a one-way ANOVA to compare the mean drug concentration at each time point
between groups 1 and 2.
Rat Acute and Chronic Dosing with Cocaine / Ethanol
Four groups of Sprague Dawley male rats were used for this experiment and
with five rats per group. Groups 1 and 2 were "acute" groups. Group 1 received
a single dose of cocaine (10mg/kg IP). Group 2 received a single dose of cocaine
(10mg/kg IP) and of ethanol (1.5g/kg IP). Blood samples were collected at 15, 30,
60, 120, and 180 minutes. The brain was harvested after the last blood collection
and the hypothalamus was sectioned from the rest of the brain. Samples were
stored frozen at -20C until extraction and analysis and quantitation by HPLC.
Groups 3-4 rats were dosed daily for 24 days and were considered "chronic"
groups. Group 3 was dosed with cocaine (10mg/kg IP) and saline (1.5g/kg IP).
Group 4 was dosed with cocaine (10mg/kg IP) and ethanol (1.5g/kg IP). Blood
samples were collected at 15, 30, 60, 120, and 180 minutes. The brain was
harvested after the last blood collection and the hypothalamus was sectioned from
the rest of the brain. Samples were stored frozen at -20C until extraction and
analysis and quantitation by HPLC. Area-under-the curves for the time versus
concentration plots were calculated and compared for statistically significant
differences. In addition, the time to peak concentration (Tmax) and the peak
concentration (Cmax) were determined for each group from the plot of time versus
Specific Aim #4
To examine the effect of cocaine, norcocaine, benzoylecgonine, and
cocaethylene on T and B cell proliferation.
IM-9 B-Lvmphoblastoid Human Cell Line
IM-9 cells (CCL#159, ATCC, Rockville,MD) were used to study the effect of
cocaine and metabolites on B cell function. These cells have receptor sites for
human growth hormone, IGFII, insulin, and calcitonin. The cells were cultured in
suspension in a complete medium consisting of 90% RPMI1640 (supplemented with
100pg/ml streptomycin, 100 units/ml penicillin, and 2mmol L-glutamine) and 10%
fetal bovine serum. The cell suspension was replenished every 48 hours.
Cocaine hydrochloride, benzoylecgonine, and phorbol 12-myristate 13-
acetate (PMA) were obtained from Sigma (St. Louis, MO). Cocaethylene and
norcocaine were obtained from the National Institute on Drug Abuse (NIDA,
Rockville, MD). Drug solutions were sterile filtered through 0.2pm syringe acrodisc
filters (Gelman Sciences, Ann Arbor, MI). RPMI 1640 medium (Gibco, Grand
Island, NY) was supplemented with streptomycin (100pg/ml; Gibco), penicillin
(100units/ml; Gibco), and L-glutamine (2mM; Gibco). IM-9 cells (CCL#159) were
obtained from ATCC (Rockville, MD) and fetal bovine serum (FBS) from Hyclone
Laboratories. [3H]-thymidine (5.0 Ci/mmol) was obtained from Amersham Life
Science, scintillation fluid from Research Products Division (Costa Mesa,CA), and
cell culture plates were from Corning (Marietta, GA).
Cocaine hydrochloride, benzoylecgonine, and cocaethylene were dissolved
in sterile phosphate buffered saline (PBS) at concentrations of 40X final well
concentration. Norcocaine was dissolved in a PBS solution containing 1% methanol
to aid solubility, then diluted to 40X final concentration in PBS. Then, all drug
solutions were sterile filtered through 0.2pm syringe acrodisc filters. Solutions were
then diluted just prior to use in complete RPMI serum-free medium (described
under materials) to 4X final well concentration. Final well concentrations of drug
solutions were in the range 0.01 pg/ml 10.Opg/ml with four concentrations of each
drug. Phorbol 12-myristate 13-acetate (40nM) was used as a negative control in
[3H] Thymidine Incorporation Assays
Proliferation assays were performed when the cells were in early log-phase
growth. Cells were cultured in either RPMI 1640 serum-free medium or in RPMI
1640 medium supplemented with 2% fetal bovine serum (final well concentration)
for 24 and 48 hours in a humidified, 37C atmosphere with 5%CO2 in air.
Cocaine, benzoylecgonine, norcocaine, and cocaethylene (50pl, 4X final
well concentration) in serum-free RPMI 1640 medium were added in triplicate to the
96 well plate for final well concentrations of 0.01-10.0pg/ml. Drug-free controls and
PMA controls were included. IM-9 cells were collected by centrifugation, washed
once with serum-free medium, and resuspended at 2X105 cells/ml (100pl added to
each well). This resulted in 2X104 cells per well. Viability and cell density were
determined by trypan blue exclusion. Cells were cultured in 96 well plates in a final
volume of 200pl. The cells were pulsed with 1 pCi of [3H]-thymidine for the last four
hours of incubation. The cells were harvested onto glass fiber filters using a
Brandel automatic cell harvester. The filters were dried overnight and then placed
into scintillation vials with 4ml of scintillation fluid. The amount of incorporated
tritiated thymidine (in counts per minute (CPM)) was determined for both the
controls and the drug treated groups using standard liquid scintillation counting.
Preliminary experiments were completed to determine the best incubation
conditions for experiments. Using the methods described, IM-9 cells were
incubated in varied conditions. Using, (1) Two cell types: Cells cultured prior to the
experiments in RPMI 1640 serum-free medium or in complete medium; (2) Four
medium types for experiments: In either RPMI 1640 serum-free medium or in RPMI
1640 medium supplemented with fetal bovine serum producing a final serum well
concentration of 1, 2, or 5%; (3) Two time periods: 24 and 48 hours; (4) Three cell
densities: 1X104 cells/well, 2X104 cells/well, or 5X104 cells/well.
In addition, a series of experiments were conducted to find external controls
to be included during the proliferation experiments with cocaine and metabolites.
Compounds investigated as controls were tumor necrosis factor (TNFa),
transforming growth factor (TGFP,), interleukin 6 (considered a B cell stimulating
factor), dexamethasone, human growth hormone, and phorbol 12-myristate 13-
acetate (PMA). Experimental conditions were: 4 concentrations of each compound,
2X104 cells/well, 48 hour incubation with a 4 hour [H]-thymidine pulse, and in either
2%FBS RPMI 1640 medium or in RPMI 1640 serum-free medium. Cell proliferation
responses were evaluated in order to choose a suitable control for cocaine and
Drug stability during the incubation period
Since cocaine and cocaethylene, and norcocaine are subject to hydrolysis
to benzoylecgonine and norbenzoylecgonine respectively, drug stability for all
compounds was monitored over the incubation period to insure that any effects
observed were due to the specific drug studied and not a degradation product.
Drug concentrations were also verified by HPLC prior to use. The HPLC system
used was System 1 described above. Stability of the drugs in medium, under
incubation conditions, was determined by incubation of two concentrations of each
drug in both serum-free and 2%FBS RPMI 1640 medium and both with or without
cells for 48 hours. After 48 hours, the medium was removed from the wells with a
pipet and stored frozen at -20C pending HPLC analysis. Medium from wells
containing cells was centrifuged (1000g, 10 minutes) and the supernatant removed
for analysis by HPLC.
The amount of incorporated tritiated thymidine (in CPM) was determined for
both the controls and the drug treated groups. Treated cells were then expressed
as the percentage of control for the respective treatment condition (incubation time
and media type). Means and standard errors for each drug and each drug
concentration were calculated using the percentage of control values from three
separate experiments. One-way ANOVA was performed at each drug
Cord Blood PBMCs (Peripheral Blood Mononuclear Cells)
Cord blood (20ml) from both cocaine exposed (n=5) and non-exposed (n=5)
newborns was collected in heparinized blood collection tubes immediately after
delivery. Women who used cocaine during pregnancy were identified by chart
review upon their arrival for labor and delivery. Cord blood mononuclear cells were
isolated and stimulated with a mitogen to compare proliferative responses between
the cocaine exposed and non-exposed groups. Phytohemagglutinin was selected
as the mitogen for these experiments because it stimulates the first signal
transduction pathway in T cell activation.
Human albumin, histopaque 1077, phytohemagglutinin (PHA-L), RPMI 1640
medium, and ['H]-thymidine were purchased from Sigma (St. Louis, MO).
Scintillation fluid was from National Diagnostics (Atlanta, GA) and cell culture plates
were from Corning (Marietta, GA).
PBMC culture and p3Hl-thvmidine incorporation assays
Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll
Hypaque density gradient centrifugation. Cord blood (up to 20ml) was diluted to
40ml with RPMI 1640 medium supplemented with 5% human albumin. Histopaque
density gradient (10ml) was then layered under the blood. The sample was
centrifuged for 30 minutes at 1200g to isolate the PBMCs. After centrifugation, the
PBMCs form a layer between the histopaque density gradient and excess RPMI
medium. This layer was removed with a sterile transfer pipet, diluted to 50ml with
RPMI 1640 medium supplemented with 5% human albumin, and centrifuged for 10
minutes at 1200g. The cell pellet was resuspended at 1X106 cells/ml (100pl pipeted
to each well). This resulted in 1X105 cells per well. Viability and cell density were
determined by trypan blue exclusion. Cells were cultured in 96 well plates in a final
volume of 200pl.
The PBMCs (1x105 cells/well) were then cultured with phytohemagglutinin
(5pg/ml) and RPMI 1640 medium supplemented with 5% human albumin for 72
hours in a 37C humidified atmosphere with 5% CO2. PBMCs were also cultured
without PHA to serve as unstimulated controls for the PHA stimulated cells. The
cells were pulsed with 1 pCi of [3H]-thymidine for the last 24 hours of incubation.
The cells were harvested onto glass fiber filters using a cell harvester. The filters
were dried overnight and then placed into scintillation vials with 4ml of scintillation
fluid. The amount of incorporated tritiated thymidine (in CPM) was determined for
both the controls and the drug treated groups using standard liquid scintillation
The differences in cell proliferation, in response to stimulation by PHA,
between the cocaine exposed versus non-exposed cord blood lymphocytes was
evaluated using a one-way ANOVA.
Specific Aim #1
Non-enzymatic Hydrolysis of Cocaine in Human Urine Analysis by AD.
Results (Tables 3-1, 3-2, 3-3) indicate cocaine spiked into human urine can
be adequately hydrolyzed to benzoylecgonine for analysis by immunoassay. Ideal
hydrolysis conditions were: hydrolysis pH 9.5 or greater, incubation time of at least
30 minutes at 50"C, and the sample adjusted to pH 8.0 in preparation for analysis
by immunoassay. The next objective was to examine whether this procedure could
be followed for analysis of actual clinical samples.
Table 3-1: Initial experimental conditions for hydrolysis.
Sample Description Concentration reported by AD, (ng/ml)
Medium Control 1440
Blank human urine 0
Cocaine spiked urine (1000ng/ml) 13
Cocaine spiked urine @ pH 8 and 82
left at room temp. for 10min
Cocaine spiked urine @ pH 5.7 22
and left at room temp. for 10min
Cocaine and BZE spiked urine High (>5000ng/ml)
BZE spiked urine (5pg/ml) High (>5000ng/ml)
Table 3-2: Hydrolysis experiment 2 of cocaine to benzoylecgonine at varying pH
after incubation for 60 minutes at 50"C.
Sample Hydrolysis AD, pH b Concentration
pH" (ng/ml) C
Blank urine d -- Low
Blank urine --- --- Low
COC spike no --- 2
Medium Control -- 1417
Spike 1000ng/ml COC 9.5 7.1 1021
Spike 1000ng/ml COC 9.8 6.2 1071
Spike 1000ng/ml COC 10.2 6.4 932
Spike 1000ng/ml COC 10.2 7.9 962
Spike 1000ng/ml COC 10.5 8.1 765
Spike 1000ng/ml COC 10.5 6.6 679
BZE spike --- High
Hydrolysis pH is the pH to which the urine sample is adjusted with NaOH
prior to hydrolysis incubation.
b AD, pH is the pH to which the urine sample is adjusted with 6M HCI prior
to analysis after the incubation period.
c Concentration reported is not quantitative due to sample dilution produced
by pH adjustments. "Positive" results would indicate the sample would
be submitted for confirmatory analysis and quantitation.
d ---- indicates not applicable to this sample.
Table 3-3: Final experiment to determine ideal cocaine hydrolysis conditions at
Sample Hydrolysis pH Hydrolysis AD, pH Concentration
Time (min) (ng/ml)
Blank urine unhydrolyzed ---- Unchanged 6
Blank urine 10.1 60 8.0 9
1000ng/ml unhydrolyzed ---- Unchanged 65
100ng/ml unhydrolyzed ---- Unchanged 5
Low control unhydrolyzed ---- Unchanged 494
1000ng/ml 8.1 15 8.1 209
1000ng/ml 8.0 15 8.0 223
1000ng/ml 8.9 30 7.8 623
1000ng/ml 8.7 30 7.8 585
1000ng/ml 8.5 60 8.5 893
1000ng/ml 8.7 60 8.0 802
1000ng/ml 9.7 15 9.7 NA
1000ng/ml 9.2 30 9.2 843
1000ng/ml 9.8 15 8.3 623
1000ng/ml 9.4 30 7.9 909
1000ng/ml 10.1 15 7.4 574
1000ng/ml 10.4 30 10.4 862
100ng/ml 8.9 30 8.3 34
100ng/ml 8.5 60 8.5 69
100ng/ml 10.0 15 7.8 61
100ng/ml 9.7 30 7.0 96
100ng/ml 10.2 15 7.9 57
100noml 10.2 30 8.3 93
Clinical Samples of Fetal Urine and Meconium
The presence of unmetabolized cocaine in newborn urine and meconium
samples would limit the use of sensitive, rapid immunoassay screening. The results
of experimental non-enzymatic hydrolysis of infant urine and meconium to improve
detection of gestational cocaine exposure are shown in Table 3-4 and 3-5.
Table 3-4: BZE detection in meconium by immunoassay following hydrolysis.
Sample I.D. BZE Concentration by the BZE concentration by the
"direct" method (ng/g) hydrolysiss" method (ng/g)
M0472 0 74
M0474 Low Low
M0476 1 4
M0478 3 Low
M0479 287 281
Low control 485 --
cocaine spiked 89 952
meconium (1pg)_ ___
Table 3-5: BZE detection in newborn urine by immunoassay following hydrolysis.
Sample I.D. BZE Concentration by the BZE Concentration by the
"direct" method (ng/g) hydrolysiss" method (ng/g)
U0439 185 34
U0440 0 4
U0443 2 3
U0446 2 10
U0448 4935 1830
U0449 1 5
U0450 6 1
U0451 6 5
U0455 22 169
U0456 158 59
High Control 2832 ---
Results of the hydrolysis experiments found that this method did not improve
detection of gestational cocaine exposure in actual clinical samples. The method
followed to complete the hydrolysis experiments was as labor-intensive as the solid-
phase extraction and HPLC analysis described in Chapter 2, yet did not improve
detection. For these reasons, this method was discontinued.
Solid-Phase Extraction Procedures for Biological Samples
Solid-phase extraction was the analytical method used to extract cocaine and
metabolites from the biological matrices. Methods used for SPE are as written in
Chapter 2. The percentage of recovery of drug from each biological matrix is
summarized in Tables 3-6 and 3-7 Extraction recovery was satisfactory with all
methods. Extraction recovery from placenta and liver tissue was variable and
requires the use of an internal standard.
Table 3-6: Extraction recovery of cocaine and metabolites (1pg) from biological
Matrix BZE" COCO NC" CE" BUP"
Meconium 8111 879 887 917 915
Urine 849 924 864 936 894
Amniotic Fluid 6212 868 ---- -- 885
Whole Blood 758 864 903 918 854
Cord Blood 8011 888 --- -- 856
Serum 839 936 927 966 933
a Recovery is reported as % recovery % std. dev. of recovery from n=5
b --- indicates recovery not determined for this compound.
Table 3-7: Extraction recovery of cocaine and metabolites (1 pg) from brain tissue.
Extraction BZE" COC" NC" CE' BUP'
1--Digested 858 896 ----b ---- 904
2--Undigested 816 917 886 905 855
Recovery is reported as % recovery % std. dev. of recovery from n=5
determinations for method 1 and n=3 determinations for method 2.
b ---- indicates recovery not determined for this compound.
Results of extraction of cocaine and metabolites from brain tissue indicated
that digestion was not required. Therefore, solid-phase extraction method 2 was
utilized for all brain extractions for this project. In brief, brain tissue was
homogenized with phosphate buffer (3ml) and internal standard (10 pl), centrifuged
for 10 minutes, and the supernatant extracted using SCX columns. However, fetal
brain tissue was sufficiently fluid such that the uncentrifuged homogenate could be
extracted directly without blocking the SCX column. This method was followed for
extraction of fetal brain tissues under Specific Aim#3.
High Performance Liquid Chromatoaraphv (HPLC)
The HPLC mobile phase providing the best resolution of drugs and
endogenous matrix interference was Mobile phase 4 which consisted of 0.025M
potassium phosphate buffer : acetonitrile : butylamine (81:18:1) with the final pH
adjusted to 3.0 with concentrated orthophosphoric acid. This mobile phase
provided excellent chromatographic performance with both analytical columns used
throughout the project. A typical HPLC chromatogram from HPLC system 1 is
shown in Figure 3-1. In addition, system 1 employs a multiwavelength detector
which allows for acquiring full spectra from 190-400nm (Figure 3-2) and the
generation of the ultra-violet (UV) and derivative spectrum for each peak in the
chromatogram (Figure 3-3 and Figure 3-4).
0.00 8.00 16.00 24.00 32.00
Chromatogram Display: \COCAINE6.BFF
Figure 3-1: HPLC chromatogram of cocaine and metabolites at 230nm (system 1).
0 6.00 12.00 18.00 24.00 30.00
Chromatogram Display: \FOCUS\STDMIX.BFF
Figure 3-2: Full spectrum of standard solution of cocaine and metabolites.
Spectra Analysis: \FOCUS\STD1.BFF
Figure 3-3: UV spectrum of a standard solution of cocaine.
220 240 260 280 300
Spectra Analysis: \FOCUS\STD1.BFF
Figure 3-4: UV derivative spectrum of a standard solution of cocaine.
HPLC system 2 employs a tunable absorbance detector which is operated
at 230nm for detection of cocaine and metabolites. A typical HPLC chromatogram
is shown in Figure 3-5. System 2 also employs computer software to manage data
acquisition and processing as well as quality control monitoring to evaluate
chromatographic performance prior to and during sample analysis. Method
validation information acquired after analysis of a set of standard samples for
method calibration of cocaine and metabolites is shown in Figure 3-6 and Figure 3-
7. This data was evaluated after the calibration to insure that the variations in
column efficiency, resolution, and retention times were less than 10% and therefore
indicating suitable chromatographic performance for the analysis of unknown
5.00 10.00 15.00 20.00
Figure 3-5: HPLC chromatogram of cocaine and metabolites (system 2).
Figure 3-5: HPLC chromatogram of cocaine and metabolites (system 2).
Mille -niu Reults Report: S3 Peak Plot Proc Chan: 486 Page 1
For Sample: 1750 Vial: 1 Inj: 1 Chan: 486 Date Processed 10/07/93 05:03 PM
Channel Descr: WATWS5 486
nnill n uium Sa .pl. Inf oration
Acq Meth Set:
10/03/93 03:28 PM
10/07/93 05:03 PI
Peak: BEE, Fi, sia Width of 0.758
0.02 .. t
0o0oi 0 5 ia 1012.
5.00 0.00 7.00
Figure 3-6: Validation report to monitor the chromatographic system.
Figure 3-7: Method validation bar graph indicating with-in run variability.
The linearity and precision of standard curves of cocaine, benzoylecgonine,
norcocaine, and cocaethylene with bupivacaine as an internal standard were
routinely determined throughout the course of the project. A representative
calibration curve is shown in Figure 3-8 for cocaine with the data shown in Table 3-
8. Representative calibration curves for the metabolites are in Appendix A.
Table 3-8: Representative standard calibration data for cocaine.
Concentration Average Standard n
(ng/g) Response Deviation
100 320 10 2
250 650 10 2
500 1440 10 2
750 1770 20 2
1000 2400 10 2
1250 2910 0 2
1500 3490 20 2
1750 4030 10 2
The data is fit to the equation y=b + mx where m=slope and b=intercept.
Results of this calibration show: slope = 2.223
The correlation coefficient (r) = 0.9984
0 500 1000 1500 2000
Figure 3-8: Representative calibration curve for cocaine.
Throughout the study period, changes in curve slope, y-intercept, and
correlation coefficient were assessed (Table 3-9). Results indicate that the average
deviation of slope for all compounds was less than 25% over a two-year period.
These values were also determined utilizing two different HPLC systems, minor
alterations in mobile phase, and two different types of analytical columns.
Table 3-9: Variation in slope, y-intercept, and correlation coefficient over time.
Date HPLC slope y-int r slope y-int" r
3/30/92 1 3.71 -25 0.9825 2.75 15 0.9891
5/26/93 1 _b 3.88 40 0.9927
6/3/93 1 4.60 -24 0.9901 2.40 -70 0.9954
6/10/93 1 4.77 -31 0.9960 2.99 -16 0.9810
6/24/93 1 5.03 -33 0.9962 3.04 -26 0.9897
7/20/93 1 3.99 10 0.9976 2.01 7 0.9995
7/22/93 1 4.21 114 0.9669 2.83 -1 0.9996
12/1/93 1 4.73 75 0.9830 3.28 82 0.9952
7/13/93 2 3.36 1 0.9991 2.65 -33 0.9996
8/13/93 2 3.56 -2 0.9977 2.68 11 0.9930
10/4/93 2 3.92 33 0.9990 2.90 33 0.9989
2/21/94 2 3.89 48 0.9989 2.23 68 0.9984
5/20/94 2 4.80 -52 0.9896 3.80 -39 0.9974
8/10/94 2 4.82 -22 0.9998 3.51 9 0.9993
12/21/94 2 4.80 -17 0.9993 3.54 2 0.9993
10/26/94 2 5.61 -76 0.9923 4.41 -76 0.9944
y-intercept values are reported in units of concentration.
b indicates determination not made on this date.
Date HPLC slope y-int" r slope y-inta r
3/30/92 1 b --- ----
5/26/93 1 --- --- ---- -- --
6/3/93 1 1.90 -13 0.9959 3.20 -138 0.9898
6/10/93 1 4.22 7 0.9840 3.14 -143 0.9764
6/24/93 1 4.66 -67 0.9889 3.23 -156 0.9897
7/20/93 1 1.73 49 0.9957 2.49 -118 0.9759
7/22/93 1 3.70 -15 0.9995 2.63 -113 0.9913
12/1/93 1 3.42 111 0.9962 2.69 58 0.9930
7/13/93 2 2.93 -24 0.9990 2.38 -33 0.9997
8/13/93 2 2.79 20 0.9951 1.71 -25 0.9962
10/4/93 2 3.26 18 0.9988 2.23 -53 0.9988
2/21/94 2 3.14 26 0.9990 2.01 49 0.9977
5/20/94 2 4.03 -47 0.9972 2.90 -100 0.9959
8/10/94 2 4.36 -21 0.9998 2.82 -113 0.9984
12/21/94 2 4.33 -14 0.9996 2.76 -93 0.9982
10/26/94 2 4.37 -101 0.9919 2.69 -93 0.9944
a y-intercept values are reported in units of concentration.
b --- indicates determination not made on this date.
Intraday and interday quantiation variability were also assessed throughout
the course of the study for both HPLC systems to determine the precision of the
method (Tables 3-10, 3-11, 3-12). Raw data points are tabulated in Appendix B.
Table 3-10: Interday variability for HPLC system 1 at 1000ng/ml.
Benzoylecgonine Cocaine Norcocaine Cocaethylene
mean 1033 1044 1037 1005
Std.Dev. 106 94 114 140
%RSD 10.3 9.0 11.0 13.9
n 9 10 8 8
Table 3-11: Interday variability for HPLC system 2 at 1000ng/ml.
Benzoylecgonine Cocaine Norcocaine Cocaethylene
mean 1012 968 1065 991
Std.Dev. 29 34 89 115
%RSD 2.9 3.5 8.4 11.6
n 13 13 13 13
Table 3-12: Intraday variability for HPLC system 2 at 1000ng/ml.
Benzoylecgonine Cocaine Norcocaine Cocaethylene
mean 993 901 1050 976
Std.Dev. 11 24 33 27
%RSD 1.1 2.7 3.1 2.8
n 11 11 11 11
Method sensitivity and specificity were determined to demonstrate the limit
of detection and the lack of interference with endogenous materials in the biological
matrix. The HPLC methods provided adequate specificity for all compounds of
interest; however, GC/MS demonstrated the best specificity of all analytical
Table 3-13: Limits of detection (LOD) and limits of quantitation (LOQ).
Analytical BZE Cocaine Norcocaine CE
Method LOD LOQ LOD LOQ LOD LOQ LOD LOQ
HPLC System 1 1 5 1 5 1 5 4 20
HPLC System 2 1 5 1 5 1 5 2 10
GC/MS 1 5 2 10 10 50 1 5
Gas Chromatoaraohv-Mass Spectrometrv / HPLC Comparison
Results of the experiment to compare HPLC analysis of cocaine and
metabolites to the "gold standard" GC/MS indicate that HPLC is a valid method of
analysis. However, in situations where structural identity is required for confirmation
purposes (for example, in the analysis of unknown urine samples for drugs of
abuse), GC/MS is the preferred method of analysis. In the research laboratory,
HPLC using photodiode array or multiwavelength detection can provide limited
specificity via UV and derivative spectra and can be used for analysis of samples
from designed experiments (for example, when mice are injected with a known
quantity of cocaine). The multiple ion descriptors for mass spectral analysis are
illustrated in Table 3-14. Each compound, except the internal standard, was
described by three major ions each of which must be present for a positive
identification. Calibration results are compiled in Table 3-15 and include mean
deteminations of slope, y-intercept, and correlation coefficient for the three separate
calibrations by each method. Sensitivity of the analytical methods was compared
by paired T-tests on the mean slope determinations (Table 3-16). Precision was
evaluated by a paired T-test on the interday variabilities of the methods (Table 3-
17). Quantitation of benzoylecgonine from human urine samples, previously
identified as positive by a certified drug testing laboratory, showed excellent
correlation between the methods (Figure 3-9). Results indicate that GC/MS
demonstrated better precision than HPLC, but the methods provide equivalent
Table 3-14: Multiple ion descriptors for GC/MS analysis.
Compound m/z monitored
Benzoylecgonine derivitized 300 316 421
Cocaine 182' 198 303
Cocaethylene 196' 272 317
Norcocaine derivitized 194' 313 435
Norcocaine 1368 168 289
SMass used for regression analysis
Table 3-15: Calibration curve comparisons for GC/MS and HPLC (n=3 calibrations).
Compound Slope Intercept (ng/ml) r
Mean SD Mean SD Mean SD
BZE 6.37 0.22 -12 15 0.9972 0.0019
o Cocaine 3.90 0.15 4 18 0.9972 0.0034
I NOR 5.31 0.21 -41 8 0.9970 0.0074
CE 9.86 0.295 3 4 0.9972 0.0035
W D-BZE 6.63 1.30 27 70 0.9954 0.0111
SCocaine 6.07 0.85 25 62 0.9896 0.0168
D-NOR 1.48 1.59 -34 86 0.9980 0.1009
CE 8.37 0.49 51 76 0.9942 0.0068
Table 3-16: Comparison of method precision.
* p < 0.05 using Paired T-test
Table 3-17: Comparison of method sensitivity.
Benzoylecgonine Cocaine Norcocaine Cocaethylene
Equivalent GC/MS" HPLC Equivalent
** p < 0.05 using Paired T-test
The standard error of the estimate (Table 3-18) was calculated at the 95%
confidence interval for each drug. Three calibrations were done using each method
(HPLC and GC/MS) for each drug. The slopes and intercepts were determined by
linear regression (y=mx+b). Then, using the calculated values of slope and
intercept, the "calculated" concentration was determined. The "calculated" values
were than plotted on the y-axis and the "actual" concentrations (50-1500ng/ml) were
plotted on the x-axis. A linear regression of these plotted data yielded the
calculation of the standard error of the estimate.
Table 3-18: Standard error of the estimate (ng/ml) for the calculated versus
actual calibration concentrations (residuals) by GC/MS and HPLC.
Calculated concentrations determined by solving the y=mx+b equation
following method calibration.
Method and Cocaine BZE Norcocaine Cocaethylene
GC/MS #1 142.46 7.56 -" 43.57
HPLC # 1 62.48 62.91 38.05 33.38
GC/MS #2 50.24 106.78 --- 86.42
HPLC #2 62.05 47.37 17.31 51.98
GC/MS #3 97.16 16.88 57.30 91.78
HPLC #3 28.88 48.05 85.81 68.58
---- indicates value not calculated.
IPLE IRIFIE CONTROL 1589 WVnL
RANGEC 1,129 L9ELI N I, 4.0 AM> A 1, 1.0 J 0 W U 29. 3 9
lee.O 219025 .
RIC 2532. 933.
15 2171 8. 237151 .
5l51 6.13 6,35 6,457 7.H I,41 *u IIM
Figure 3-9: Gas chromatogram of cocaine and metabolites after derivitization.
81/27': 94 8:42:! + 7:34
SaIPLE: COCAETHYLENE, I UG/SI
COINS.: SCA141 50-500
TEMPl 264 DEG. C
51 122 212
JJ, .230. 1. 2tn [ '$1 302
M/Z 58 108 150 200 251 3E0
Figure 3-10: Mass spectrum of cocaethylene extracted from human urine.
0 500 1000 1500 2000 2500 3000
GCIMS Concentration (nglml)
Figure 3-11: Method comparison for BZE quantitation in urine from human cocaine
Specific Aim #2
Human Studies -- Meconium and Urine
Results of analyses of newborn urine and meconium are in agreement with
previously published studies indicating meconium is a better matrix for detection of
prenatal cocaine exposure (Tables 3-19). In addition, the procedure of "targeting"
women or their newborns for urine toxicology testing was evaluated against
comprehensive meconium testing for identifying the incidence of cocaine exposed
newborns (Table 3-20).
Table 3-19: Analysis of 22 paired samples of newborn meconium and urine.
Both Negative Both Positive -ve urine/+ve +ve urine/-ve
I meconium meconium
18 2 2 0
Total positive = 4/22 = 18.2%
Of 312 infants tested by comprehensive meconium testing or targeted urine
immunoassay screening, 62 were identified as cocaine exposed by either maternal
report of drug use or detection by either method of analysis. When the cocaine
exposed group was evaluated for presence of clinical risk factors, it was found that
69% presented no risk factor. Of the 53 positive meconium samples, cocaine was
detected in 40 (75%), benzoylecgonine in 36 (68%), norcocaine in 12 (23%), and
cocaethylene in 1 (2%). Cocaine concentrations ranged from 140-3498ng/g, BZE
from 92-8433ng/g, and norcocaine from trace-1357ng/g.
Table 3-20: Targeted urine screening versus comprehensive meconium testing.
Targeted Urine Comprehensive Meconium pa
Incidence 10/312 (3.2%) 53/312 (17.0%) 0.0001
Sensitivity 10/62 (16.1%) 53/62 (85.5%) 0.0001
Total cocaine exposed (by both methods)= 62/312 = 19.9%
a p values were determined by chi-squared analysis.
Human Studies -- Amniotic Fluid and Cord Blood
Analyses of newborn urine, cord blood, and amniotic fluid from cocaine
exposed newborns and their mothers indicate the highest concentrations of
benzoylecgonine (the major cocaine metabolite) are found in amniotic fluid. The
fetus is continually exposed to amniotic fluid throughout gestation. Results also
used in evaluation of hypothesis (1) are reported under results for specific aim #3.
Table 3-21: Human study using biological fluids from 14 women with a history of
cocaine use during pregnancy
Fluid Type # Confirmed cocaine BZE Concentration
exposed Range (ng/ml)
Amniotic Fluid 10/14a 420 3100 ng/ml
Cord Blood 6/8" 120 770 ng/ml
Infant urine 5/12" 100 -1110 ng/ml
4/14 individuals had all fluids confirmed cocaine unexposed.
3/12 individuals had a negative newborn urine but a positive amniotic fluid.
Unmetabolized cocaine was detected in 4/14 amniotic fluid samples.
Refers to the number exposed/the number of samples available for testing.
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E5BQ6S4CS_T4RN3Q INGEST_TIME 2012-09-24T14:16:17Z PACKAGE AA00011837_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC