Addiction Biology (June 2003) 8, 123 – 139

INVITED REVIEW

Biological markers of cocaine addiction:

implications for medications development

AHMED ELKASHEF & FRANK VOCCI

Division of Treatment, Research and Development (DTR&D), National Institute on Drug

Abuse (NIDA), National Institutes of Health (NIH), Bethesda, MD, USA

Abstract
The search for effective medications for cocaine addiction has been elusive. The failure to find such medications
so far could be due to poor understanding of the underlying biology both in the premorbid condition and following
the disease state of chronic cocaine use. Population heterogeneity could be a major factor in response to
medications. In an attempt to highlight the issue of biomarkers we reviewed physiological, neuroendocrine and
neuroimaging studies to identify specific biological changes/markers that could be used to characterize subgroups
among chronic cocaine users. Merging the biology within medications studies of cocaine abusers could prove
useful for targeting specific pharmacological agents to subgroups of patients, prediction of response to medication
and relapse to use.

Introduction trials1 has shown that most of these agents are

So far there has not been a single medication that
is FDA-approved for the treatment of cocaine
dependence. Psychotherapy is still the mainstay
of treatment with some proven efficacy; however,
the rising numbers of cocaine abusers, especially
in the adolescent age group, and the high relapse
rate among treatment seekers make the search for
effective treatment highly pressing. The pub-
lished medications studies for cocaine depen-
dence in the 1970s and early 1980s have mostly
utilized dopamine agonists, e.g. amantadine and
bromocripitine, as well as antidepressants, mainly
desipramine. The results from these studies
varied greatly because of different patient selec-
tion criteria, different methodology and outcome
measures. A recent meta-analysis of some of these
ineffective. It is also clear from the data that even
though results were largely negative there were
some subgroups of patients who responded to
these medications; however, the studies do not
provide an explanation as to why. We believe that
population heterogeneity is a major factor in
explaining the inconsistent findings of these
studies. It is generally agreed upon that cocaine-
dependent individuals differ in their response to
stress, ability to experience pleasure, self-regard
and risk-seeking behavior. These differences,
coupled with high co-morbid mental disorders,
e.g. depression, anxiety, attention deficit, bipolar
disorders and personality disorders/traits, make
this population far from being a homogeneous
group.

Correspondence to: Ahmed Elkashef MD, Division of Treatment, Research and Development (DTR&D),
National Institute on Drug Abuse (NIDA), National Institutes of Health (NIH), 6001 Executive Boulevard,
Room 4123, MSC 9551, Bethesda, MD, 20892-9551, USA. Tel: 301 443 5055; Fax: 301 443 2599;
E-mail: ae8a@nih.gov
Received for publication 14th May 2002. Accepted 5th August 2002.

ISSN 1355–6215 print/ISSN 1369-1600 online/03/020123-17

# Society for the Study of Addiction to Alcohol and Other Drugs Taylor & Francis Ltd
DOI: 10.1080/1355621031000117356
124 Ahmed Elkashef & Frank Vocci
The goal of this review is to identify specific
biological changes/markers that could be used to
characterize subgroups among chronic cocaine
users. This could be helpful for:
(1) Targeting specific pharmacological agents
for subgroups of cocaine abusers, which
can maximize the positive outcome of
these trials.
(2) Determining the length of treatment.
(3) Preventing relapse by initiating treatment
at the earliest sign of an increase or re-
emergence of a certain marker.
(4) Understanding the ‘‘biological/genetic’’
makeup that predisposes certain indivi-
duals to become addicted to cocaine.
Methods
We reviewed the literature on biological markers
in cocaine abusers using Medline Grateful med.
Studies for the last 15 years on physiological,
biochemical/neuroendocrine, cognitive and neu-
roimaging indices were collected. We used the
term ‘‘marker’’ loosely in this review to indicate a
change from normality that occurs at certain
phase of the illness, as none of these has been
established as markers for cocaine dependence.
Based on this review, new approaches are
formulated.
Results
It was evident from the search that most of the
biological data investigated the dopamine system.
Fewer more recent studies have also addressed
other systems, e.g. serotonin. Because of the
volume of data and in an effort to keep this review
focused, and because the goal is to present a
working model that can be applied to any
neurotransmitter system, we decided to focus
upon dopamine where most of the data are
published. However, this by no means represents
our bias and should not bias the reader towards
dopamine only.
Biochemical/neuroendocrine
Neuroendocrine studies have investigated both
the acute and chronic effects of cocaine use on
serum levels of prolactin (PRL) and growth
hormone (GH) as indirect measures of change
in dopamine concentrations. Prolactin (PRL)
release is suppressed tonically by dopamine
release into the tuberoinfundibular tract. On the
other hand, growth hormone (GH) secretion is
stimulated by dopamine. Both actions seem to be
mediated through D2 receptors. It is expected
that the acute effects of cocaine will be to
decrease PRL and increase GH release. Less
frequently cortisol, ACTH, LH and testosterone
levels were also measured (Table 1).
Biochemical studies measured serum and CSF
levels of dopamine and its metabolites, mainly
HVA, as well as the norepinephrine metabolite
MHPG.
Four studies measured the acute effects of
cocaine or methylphenidate on prolactin levels.
Three studies2 – 4 reported decreased levels as
expected, whereas in one study5 no change was
detected. These differences were attributed to
possible tolerance to cocaine effects after chronic
use, or a burn-out effect where after prolonged
use of cocaine the dopaminergic neurons in some
subjects become non-responsive to cocaine or
dopamine-depleted. This finding highlights the
difference in dopamine system plasticity follow-
ing chronic cocaine use.
Six of eight studies that measured PRL levels
showed elevation in serum PRL above their
normal laboratory values in chronic cocaine
abusers who abstained from cocaine use for a
few days to weeks. This is again consistent with
decreased dopaminergic tone in the tuberoinfun-
dibular tract upon cessation of cocaine use. When
individual data were presented it appeared that
the elevations were in 40 – 70% of subjects, while
30 – 60% of subjects showed no change or even a
decrease in levels. The other two studies reported
no change or even a decrease in PRL.6,7
Two studies7,8 included subjects off cocaine
for up to 43 – 60 weeks, and reported no
differences in PRL or HVA levels. Unfortunately,
these studies did not include prolactin levels at
baseline to compare with long-term levels, which
might suggest normalization of the proposed
hypodopaminergic state with time and the useful-
ness of PRL levels in the follow-up of the patient’s
response to treatment.
Weiss et al.9 followed 42 cocaine-dependent
patients for 6 months and reported that patients
with hyperprolactinemia at baseline were signifi-
cantly more vulnerable to relapse to cocaine use
than those with normal prolactin at baseline.
Similarly, Kranzler & Wallington10 reported that
patients with elevated prolactin levels on admis-
sion had a greater likelihood of non-compliance
 Biological markers of cocaine addiction 125

Table 1. Studies of neuroendocrine and neurochemical changes in cocaine-dependent patients

Study Subjects Tests Results Comments

(6) 21 cocaine abusers, Serum PRL, GH Decrease PRL in 15

withdrawn 4 – 10 days men, no change in 6
women. Increased GH
in both sexes
(37) 18 cocaine abusers, 20 Serum PRL and TRH Elevated PRL in
NC on admission and 2 cocaine, slight decrease
weeks later in 2 weeks. Blunted
TSH response to TRH,
no change in 2 weeks
(38) 10 cocaine abusers Serum PRL Elevated in 7, decreased 70% of subjects had
after 2 weeks of elevated PRL, 30%
abstinence but not showed no change
normalized
(39) 16 cocaine abusers (use Serum PRL, LH, Patients had elevated 9 patients (about 60%)
1 – 10 years, off 4 cortisol, testosterone on PRL persisted. Other had high levels, 3
weeks) admission and 4 weeks hormones WNL borderline, 2 normal
later
(40) 9 cocaine abusers (use Plasma MHPG Cocaine patient’s Some cocaine patients
44 months) off (6 – 27 MHPG was 53% of on desipramine
days), 18 drug normal
abstinent, and 9 normal
controls
(41) 6 cocaine abusers (off Plasma HVA, MHPG Increased HVA 1 – 2
21 days weeks after last use,
correlated with
increased craving
(7) 23 cocaine abusers (off Serum PRL No significant elevation Very low levels 5 5 ng/
2 – 43 weeks) ml in 6 patients
(42) 8 cocaine, 8 Prolactin, GH (basal 3/8 cocaine patients had 40% had elevated PRL,
cocaine + alcohol, and and after apomorphine elevated baseline 60% no change
8 normal controls challenge) Prolactin than normal.
No differences with
challenge
(2) 18 cocaine- and Prolactin, ACTH, LH Sign. increase in ACTH Acute effects
opoid-dependent (baseline and after and LH after cocaine
30mg i.v. cocaine infusion. Sign. decrease
infusion in prolactin from base-
line after both cocaine
and placebo
(10) 33 cocaine (off 48 hrs) serum PRL 13 patients had elevated Patients who had
PRL, no correlation to elevated PRL (39%)
cocaine use or craving also had premature
discharge from hospital
(8) 9 cocaine (use 6 years, CSF HVA No difference. HVA
off 12 – 60 weeks), 9 corr. with craving
normal controls
(43) 22 cocaine (off 3 Plasma Prolactin, GH, Increase in PRL from
weeks), normal controls HVA at baseline & at 3 baseline, no diff in GH
weeks or HVA
(11) 15 cocaine and 15 Plasma DA, adr., Dopamine sulfate was
normal controls noradr. and their sign. elevated in
sulfated metabolites cocaine, and sig.
correlated with severity
of use. No diff. in the
other metabolites

(continued overleaf )
126 Ahmed Elkashef & Frank Vocci
Table 1. (continued )

Study Subjects Tests Results Comments

(3) 25 cocaine dep. Cocaine 40 mg and Sign. decrease in PRL Acute effects
saline infusions, serum and sign. increase in
PRL and cortisol cortisol with cocaine
(5) 8 cocaine, heroin and Cocaine 40 mg and Sign. increase in Acute effects
alcohol abusers saline i.v. infusions. cortisol but no change
Plasma cortisol and in prolactin
prolactin
(4) 10 cocaine (use 3 – 6 Methylphenidate Patients had higher Acute effects
years, off 1 – 2 days) and challenge acutely and PRL acutely than
9 controls after 1 week Plasma controls. No change in
PRL and GH 1 week in patients.
MPD caused decreased
PRL and GH in
patients

with treatment and early discharge against med-
ical advice.
Fewer studies measured GH, cortisol, ACTH,
LH, HVA and MHPG levels, with inconsistent
findings. Two studies of the acute effects of
cocaine report elevation of cortisol levels in
response to cocaine infusions. Dopamine sulfated
metabolites were reported to be elevated signifi-
cantly in two studies by the same group.11 The
data on other hormones and biochemical mea-
sures are too few and noisy to make any
conclusions.
In summary:
(1) Of all the peripheral parameters measured
in these studies elevated PRL seems to be
the most replicable.
(2) Long-term follow-up suggests that it is a
possible state marker.
(3) The noise in the data could be addressed
as follows:
. patient’s heterogeneity as far as the
dopamine system plasticity;
. different periods of withdrawal from
cocaine, i.e. different recovery time-
lines;
. mixed drug use with opposing ef-
fects on prolactin (serotonin vs.
dopamine); and
. different methods of collecting
blood for PRL levels. Prolactin
levels are very sensitive to stress,
exercise, food and caffeine intake.
Blood samples for prolactin should
be drawn in a relaxed, resting con-
dition, preferably also fasting condi-
tion. Deviations from these
requirements could result in erro-
neous results.
Physiological
Electroencephalogram (EEG) and evoked related
potential (ERP)
Nine of the reviewed EEG studies16 – 20 provide
consistent evidence for frontal lobe changes in the
form of excess relative alpha power, deficit of
absolute and relative delta and theta power and
increased beta activity in chronic cocaine abusers.
One study19 reported decrease in the beta 2 band
mainly in the temporal regions. Except for one
study,18 most subjects included were off cocaine
for only a few days. The same studies also seem to
show correlation between amount of recent
cocaine use and specific EEG changes, suggesting
an acute effect. This makes it difficult to draw
conclusions regarding the extent of these
changes. However, Prichep et al.18 and Alper et
al.’s12 studies seem to suggest that some alpha
and delta wave changes persisted in the subset of
subjects that were followed-up 1 month and 6
months later while abstinent. A similar finding
was reported by Bauer et al.21 of patients off
cocaine for 1 – 5 months, suggesting that these
changes could last longer after the patient
stopped using and is not related directly to acute
withdrawal. The reported EEG changes, espe-
cially the delta deficit in the frontal cortex, are
thought to be related to the mesocortical dopa-

mine system and sensitization. Similar findings
were reported in patients with attention deficit
disorder (ADD) and schizophrenia, where it may
be responsible for the abnormal gating to external
stimuli.
Studies of evoked potential21,23 – 25 reported
reduced frontal P30021,22 with possible correla-
tion to relapse. Other studies using different
stimuli reported reduced P5023 reduced P124 and
increased P100 latency.25 One interesting finding
in the study by King et al.13 of gender-specific
differences suggests a possible protective role for
estrogen from cocaine effects. Gender differences
were also reported in one blood flow SPECT
study26 in cocaine patients, where females were
found to have normal scans compared with males
who showed perfusion defects. However, similar
EEG differences were not reported in the gender
analysis in Perichep et al.’s QEEG study.27
The same study27 detected three distinctive
subtypes of patients based on specific mathema-
tical analysis of their baseline QEEG changes.
These changes correlated highly with retention in
treatment, but not with duration or severity of
cocaine use. This finding is similar to the finding
of the Bauer study21 (Table 2).
In summary:
(1) QEEG data offer fairly consistent find-
ings, mostly frontal.
(2) The noise in the data can be addressed by
several factors very similar to the bio-
chemical issues addressed above, in addi-
tion to differences in the EEG
methodology employed in these studies.
(3) Pilot data suggest there may be gender
differences, and different subtypes among
the clinically homogeneous cocaine de-
pendent population.
(4) Possible use for predicting relapse and or
retention in treatment.
(5) Long-term studies ( 4 6 months) may be
useful in investigating how long-lasting
these changes are and whether at some
point they become undetectable or not
different from a normal control compar-
ison group.
Eye tracking
Eye tracking is the testing of eye movement while
tracking a target moving either sinusoidally or at a
constant speed. The components of eye move-
ment involve smooth pursuit (SPEM) and
Biological markers of cocaine addiction 127
saccades (SEM). They both work interactively
to keep the eye positioned on the target; once the
fovea is placed on a target by the saccadic system
it is kept in place by the smooth pursuit system
which stabilizes the image of the target on the
fovea by matching the eye speed to the target
speed.
A measure of precision of smooth pursuit is the
ratio of eye velocity to target velocity. A perfect
score is 1; abnormalities include a low gain (slow
5 1) or a high gain (fast 4 1). Two types of
saccades may occur during smooth pursuit,
compensatory and intrusive. Compensatory sac-
cades are of two kinds, catch-up and back-up;
they serve to correct a faulty eye position.
Intrusive saccades are of several types and do
not correct for errors in eye position. The most
common ones are anticipatory saccades.
Very few studies investigated the effect of
chronic cocaine use on eye tracking. The
rationales for the studies are weak, and the results
are inconsistent simply because each addressed a
different question and used a different paradigm.
It is difficult to draw conclusions based upon
these few studies. However, the study of eye
tracking in cocaine abusers could be of value, as
there is evidence from cognitive and EEG studies
of impaired frontal lobe function in cocaine
abusers. The frontal eye field is an integral part
in the eye-tracking circuitry. There could also be
a link between the study of eye movement and
dopamine. One study28 found similar eye-track-
ing abnormalities in patients with schizophre-
nia—an illness attributed to dopamine imbalance
to cocaine abusers.
The study of eye tracking in a large group of
chronic cocaine abusers and their relatives as well
as the study of the acute effects of cocaine on eye
tracking could be helpful in understanding
possible genetic predisposition to cocaine addic-
tion and population subtypes (Table 3).
Electroretinogram (ERG)
The electroretinogram is the recording of elec-
trical potentials from the retina in response to
flashes of light. The human retina contains 4 – 6
million cones. Cones are responsible primarily for
pattern detection and color vision. Color vision is
mediated by three different types of cones, long,
medium and short, based on their spectrum
absorption. Short cones respond to blue light.
There are two types of dopamine cells (amacrine
128 Ahmed Elkashef & Frank Vocci

Table 2. Studies of EEG and ERP in cocaine-dependent patients

Study Subjects Test Results Comments

(12) 17 cocaine QEEG at baseline, and Deficit of absolute and

dependents after 1 and 6 months of relative power delta,
abstinence and increased relative
alpha. No significant
change overtime
(13) 40 cocaine abusers, EEG resting Sign. elevated B- Gender-specific
(20 males and 20 activity and reduced differences suggest
females) and 32 alpha in male cocaine that females are more
normal controls (12 users compared to protected from the
females and 20 males) females and controls effects of cocaine
(14) 33 cocaine dependents EEG resting Greater B-activity in Cocaine group have
(use 119 months, off the cocaine group than comorbid dx of
10 days), 10 non-drug, the other two Frontal phobia/panic, and
20 drug but not and central areas cor- PTSD
cocaine dep. (use 98 related with frequency
months) of cocaine use
(21) 49 cocaine dependents P300 with a visual Reduced frontal P300
(use 10 – 12 years, off attention task in the cocaine patients
1 – 5 months), 20 with ASDP dx with
normal non-drug sign. corr. with
controls childhood conduct
disorder hx. P300 also
identified 70.6% of
patients who later
relapsed and 53.3%
who did not
(15) 37 cocaine (use 68 EEG resting. ERP Increased B-activity in
months, off 9 days) (auditory, visual CPT the cocaine and the
31 cocaine and other and Sternberg memory substance abusers
substances and 17 task) groups than normal
normal controls controls which sign.
corr. with amount of
cocaine used the week
prior to admission.
EEG alpha corr. with
duration of cocaine
use. The cocaine-
dependent have
reduced P3 amp in
all tasks
(22) Cocaine (off 90 days), ERP (auditory). Visual Sign. reduced P300b Cocaine patients had a
alcohol and normal and auditory divided in cocaine dep greatest 4 – 6 Hz resting
controls attention at frontal, no diff in tremor, and slow
attention reaction time
(16) 67 cocaine dep.(use QEEG at baseline, 39 Deficit absolute and Study replicates 1995
6.9 years, withdrawn patients retested at 1 relative delta and theta finding
ave.3.2 days) month and 17 patients power, excess relative
at 6 months alpha power. Abn.
were greater in
anterior than post.
regions. Changes
persisted after 1 & 6
months

(continued overleaf )
 Biological markers of cocaine addiction 129

Table 2. (continued )

Study Subjects Test Results Comments

(23) 10 cocaine dep.(use Auditory P50 Cocaine dep.

11.2 years, off 2 markedly depressed
weeks), 15 active P50 amp. and decrease
alcoholics and 10 suppression compared
normal with alcohol and
normal
(24) 19 cocaine (use 12 Olfactory evoked Diminished P1 in Possibly due to lesion
years, off 1 – 5 potential (OEP) cocaine and alcohol in peripheral/central
months), 7 alcohol, 10 olfactory system
nicotine, 12 normal
(25) 11 cocaine (use 10 Pattern shift visual Increased P100 Possibly secondary
years, off 1 – 5 evoked potential latency in cocaine to vascular or
months), 11 alcohol, (VEP) by checker- dep. only neurological effect of
11 both (use 10 years), board reversal cocaine
10 normal controls
(17) 52 cocaine dep. (use QEEG at baseline and Deficit absolute and Changes secondary to
9.9 years, off 5 – 10 f/up study of 39 relative delta and theta cocaine or trait related
days) subjects at 1 month power, excess relative to vulnerability to
alpha power. Abn. cocaine addiction
were greater in
anterior than post.
regions. Changes
persisted after 1
month.
(18) 6 children exposed in QEEG Deficit absolute and Results similar to
utero to cocaine relative delta and theta adults. Four of the six
power, excess relative children have dx of
alpha power. Abn. ADHD
were greater in ante-
rior than post. regions
(19) 36 cocaine dep. (use QEEG at day 2 and 6. Sign. decrease in the Changes were greater
4.7 years, off 2 – 6 9 had f/up study at 1 beta2 band more pro- in i.v abusers vs.
days) month nounced on day 6 smokers
mainly centrally, with
further decrease at 1
month mainly in tem-
poral regions
(20) 14 polydrug with EEG + cocaine (20 Cocaine sign. Increased beta could
cocaine abuse and 40 mg) vs. increased beta in be secondary to
placebo frontal and central decreased cortical
regions and alpha in activity and metabolic
frontal and temporal demand

and interplexiform) in the human retina. They
receive presynaptic input from cone bipolar cells.
Dopamine cells modulate the input from cones to
other common neurons and directly modify the
photoreceptors signals. Light stimulates L-aro-
matic amino acid decarboxylase activity leading
to increased synthesis of dopamine from l-DOPA
and subsequent release of dopamine which binds
to D2 and D4 receptor types that have been
identified on cones.
Patients with decreased dopaminergic tone,
e.g. Parkinson’s disease patients or schizophrenic
patients on neuroleptics, have abnormal ERGs,
particularly reduced blue cone amplitude. The
following studies have investigated whether co-
caine dependence is associated with abnormal
ERG.
Similar to the eye-tracking studies, the ERG
studies are few and generated mainly by the same
group; however, the findings are more consistent.
The above studies show that about 60% of
cocaine-withdrawn patients have reduced blue
cone b-wave amplitude and that this reduction
correlates with craving, and digit span: a test for
130 Ahmed Elkashef & Frank Vocci

Table 3. Studies of eye tracking in cocaine-dependent patients

Study Subjects Tests Results Comments

(40) 9 cocaine abusers Electro-oculography No change in Some cocaine

(use 44 months, off Visual-vestibular SPEM; however, patients on
6 – 27days), 18 ocular reflex (VVO) cocaine patients had desipramine
polydrug abstinent sign. reduction in
and 9 normal con- the VVO reflex gain
trols (a crude measure of
saccades)
(44) 32 cocaine Saccadic No differences Negative, possible
dependents (use 7.2 distractibility found; however, link between
years, off 16.3 days) when the cocaine cocaine use and
and 15 normal patients were split OCD
controls for cocaine-induced
compulsive
foraging, those with
(n = 26) had higher
sacc. distractibility
(45) 19 recently Visual scanning Intensity of cocaine
abstinent cocaine (preattentive and craving positively
users (7.6 years) attentive fixations correlated with the
and saccades upon number of attentive
viewing a cocaine fixations, and
cue) – Questionnaire inversely correlated
of cocaine urges with the number of
(QCU) preatt. fixations and
saccades
(47) 11 cocaine, 4 SPEM Cocaine sign. better Result is possibly
alcoholic and 16 SPEM than the secondary to
controls other two stimulatory effect
(28) 14 cocaine- SPEM Both cocaine and
dependent (off 15.6 schizophrenic
days), 15 patients had sign.
schizophrenics and less SPEM gain and
22 normal controls more large intrusion
saccades

perceptual motor speed. Studies of larger samples
of cocaine patients will be useful to categorize
those patients clinically as well as studies of the
acute effects of cocaine. Similar to prolactin,
ERG may be a useful screening test for dopami-
nergic tone; however, logistically ERGs require a
special set-up and a trained ophthalmologist,
which may limit its application in large studies
(Table 4).
Cognitive/neuropsychological
Studies have investigated the question of
cognitive deficits secondary to cocaine use.
Two mechanisms have been postulated: one
involves a possible direct neurotoxic effect of
cocaine and the second involves a dementia-like
model, possibly secondary to cerebral micro-
infarcts resulting from the vasoconstrictive
effects of cocaine. Measures of attention,
memory/learning, abstract, executive, spatial
and motor functions were employed in chronic
cocaine users.
Most studies suggest that chronic cocaine
abusers suffer some deterioration in cognitive
functions, specifically attention, reaction time,
verbal memory and visuospatial construction.
Three studies report improved performance on
verbal memory and executive function tests,
which was attributed to stimulant effect of
cocaine. However, this explanation seems im-
plausible because some subjects were tested 3 – 8
weeks off cocaine.
Other limitations include lack of drug screen-
ing in some studies, lack of data on premorbid
functions (e.g. IQ, SAT scores), and mixed
 Biological markers of cocaine addiction 131

Table 4. Studies of ERG in cocaine-dependent patients

Study Subjects Tests Results Comments

(47) 20 co. dep.(use 7.6 ERG Patients sign. Evidence for

years, off 3 – 62, reduced blue cone hypodopaminergic
mean 26 days) and response (11/20 state, results
20 normal controls patients have b-wave replicate 1996 and
5 0.5 mV compare 1995 studies
to 4/20 controls and
5/20 absent blue
cone ERG
compared to o/20
controls)
(48) 14 cocaine (use 11.9 ERG Cocaine 6 had low Indirect evidence for
years,off 1 – 180, Craving ( 5 0.5 mV) b-wave relationship between
mean 27 days) Questionnaire and 8 4 0.5 mv. reduced b-wave on
(CCQ) Low had sign. the ERG and
higher cocaine craving
craving scores than
high. Neg. corr. with
factor 3 on the CCQ
(49) 6 cocaine withdrawn -ERG Sig. corr. with digit
with blunted b-wave -Groove Peg Board symbol a test for
(from Roy et al. (40) -WAIS-R perceptual motor
study) (arithmetic, digit speed
symbol, digit span)

drug use or dependence, especially alcohol. No
consistent relationship between the quantity and
duration of cocaine use and cognitive deteriora-
tion was found or reported. Only one study
included subjects 6 months off cocaine29 with
report of impairment in attention, concentra-
tion, visual and verbal memory. Further work is
needed to investigate the extent of cognitive
deterioration, and to establish direct correlation
between cocaine use and cognitive impairment/
improvement. A subset of patients could
possibly be identified with marked cognitive
deterioration, particularly if it involves executive
functions and decision-making abilities, for
which therapy has to be tailored to accommo-
date for the deficit and for the treatment to be
efficacious (Table 5).
Neuroimaging
PET studies
PET studies of cocaine subjects investigated two
main questions: the role of dopamine in the acute
effects of cocaine and the dopaminergic tone of
chronic cocaine users.
These studies provide evidence for a key role of
dopamine in the acute action of cocaine and the
subjective high. Similarly, there is evidence for a
decrease in D2 receptors, dopamine uptake sites,
dopamine release and glucose metabolism in
chronic cocaine users.
The question of how long the PET changes
persist following cessation of cocaine use remains
largely unanswered; however, one study suggests
that these changes could persist for up to 3
months30 after cessation of cocaine use, while two
studies31,32 suggest that these changes are short-
lived and return to normal in 2 – 4 weeks. These
differences shed light on the differences in
dopamine system plasticity among a seemingly
homogeneous group of patients. In a recent
presentation at CPDD 2001, Dr Volkow pre-
sented FDG PET data in chronic methampheta-
mine users showing that brain metabolism
returns to normal levels 9 months following
cessation of use. These data are helpful, as they
provide some guidance for medications studies
regarding duration of treatment or follow-up
specifically in trials where relapse prevention is
an endpoint.
PET imaging, although expensive, can be a
useful tool in characterizing the underlying
biology and holds promise for medication match-
ing (Table 6 near here).
 Table 5. Studies of cognitive functions in cocaine-dependent patients
132

Study Subjects Tests Results Comments

(50) 33 cocaine abusers withdrawn for 59
days and 21 normal controls
(51) 20 cocaine patients (use 49 months,
off 23.6 days) and two matched
normal controls groups 20 subjects
each
(29) 8 cocaine (off 6 months) to set of
normal data. patients also had
Miller – Selfridge Lexical decision
motor synchrony signal detection
MMSE, maze-tracing test, oral lan-
guage
WAIS-R, Booklet Category, finger
oscillation, neuropsychological
screening battery, Rey – Osterith
figure design, Benton multilingual
aphasia exam, numerical attention
test, BDI
Neuropsych SPECT MRI
Cocaine group worse on the delayed
recall on MMS and signal detection.
Cocaine better in the maze-tracing
test (visual memory)
Cocaine group worse on arithmetic
Symbol digit and verbal memory.
Cocaine group better on oral fluency
Cocaine group impaired in atten-
tion, concentration, visual and ver-
Improvement attributed to stimulant
effect
Correlation with cocaine use and
duration. No correlation between
depression and neuropsych. data
No comparison group
Ahmed Elkashef & Frank Vocci

SPECT and MRI bal memory. Hypoperfusion in the

(52) 16 cocaine patients and 21 normal
controls at 72 hours and at 2 weeks
of last cocaine use
(53) 12 cocaine (off 12 weeks), 5 alcohol,
16 normal controls
(54) 23 cocaine (use 7.1 years, off 21 – 40
days), 24 alcoholics and 22 normal
controls
(55) 61 cocaine (use 2 years, off 2
months), 59 polydrug users and 63
normal
(56) 12 cocaine, 14 cannabis, 7 multiple
drugs and 21 no drug history
Stroop, trails A– Osterith, oral word
fluency, digit symbol, block design,
Rey auditory verbal learning, new
adult reading test (for premorbid
intellectual functions)
Visual & auditory stimulus discri-
mination task at 1 week, 3 weeks and
3 months
WMS, Face – name learning test,
WCS, SILS abstraction, arithmetic
subtest of the WAIS, Trails A&B,
BDI and ADD
WAIS, WMS, BNT, Trails A&B,
Verbal fluency (semantic and pho-
nologic, Rey – Osterrieth complex
figure, WCS
WAIS, Trails A&B, Porteus Maze
test
frontal and the periventricular areas
At 72 hours cocaine group impaired Depression unrelated to finding
on visuospatial construction, verbal
memory and concentration At 2
weeks remain impaired as 72 hours
with more impairment on visual
memory and trails B
Cocaine group slow reaction time, Cocaine effect on motor function
persistent reflex arc (basal ganglia)
Cocaine and alcohol groups sign. Deficits unrelated to pre/comorbid
worse on all tests, except sustained conditions e.g. depression or
attention (no diff.) ADHD. Difference in premorbid IQ
between cocaine and controls
Cocaine and polydrug groups per-
form worse on short-term memory,
attention and concept formation
tests
Cocaine group had worse verbal IQ,
and information, vocabulary, com-
prehension, picture completion, and
trails B. The other groups no diff.
from the non-users

(continued overleaf )
 Biological markers of cocaine addiction 133

and verbal memory were attributed

SPECT

The improved executive functions

Of six SPECT studies, five used 99mTc-HMPAO
to investigate blood flow and one used [123I]B-

to possible stimulant effect

CIT investigated dopamine uptake sites. The
blood flow studies provide a strong case for the
vasoconstrictive effect of cocaine evidenced by
perfusion defects. One intriguing finding26 is the
difference between males and females, where
Comments

female patients lacked these perfusion defects

suggesting a possible ‘‘protective’’ effect for
estrogen. The [123I] B-CIT study showed a linear
relationship between cocaine doses and DAT
availability, as evidenced by displacement of
executive functioning, and psycho-

[123I]B-CIT. SPECT offers a more economical

Dose-related effects in attention,

strongly related to impairment if

Trail B Cocaine better on WCS

cocaine use 9 g/week) was more

motor performance. Intensity of
memory, confrontation naming,
Cocaine group worse on spatial

method of investigating questions such as blood

neurobehavioral performance

flow and receptor occupancy, with less radiation

and oral word association

exposure and repeated examinations than PET.

There are very few fNMR and NMRS stu-
dies33,34 investigating proton and phosphorus
spectra in cocaine patients from which to draw
any conclusions; however; emerging techniques
that assess neurotransmitters concentrations, e.g.
Results

GABA, hold promise and may in the near future

Table 5. (continued )

offer wider applications of NMR than PET or

SPECT.
functions, reaction time. Tests were

admission. Relationship to cocaine

use of living scale) , verbal fluency
auditory-visual memory, executive

In summary, neuroimaging tools PET and

Extensive neuropsychological bat-
Booklet category test, WCS, oral

(verbal memory) (WMS-R), Rey

tery (IQ, verbal fluency, verbal-

conducted on day 28 and 29 of

SPECT are useful for studying the underlying

word association, trails A&B

auditory verbal learning test

dopaminergic tone but they are expensive, there-

fore impractical for large phase II studies, with
more applications in proof of concept studies in
early drug development (Table 7).
(RAVLT)

Discussion
The idea of characterizing our patients biologi-
Tests

cally to possibly match medications to patients is

not new and is being pursued across different
30 cocaine dependents with positive

then remained inpatient for 30 days

medical disciplines from cancer to mental illness.

urines on admission to the study,
38 patients (use 7.3 years, off 24

The main goal of this review is to bring the issue

to the surface and to address two specific
questions:
and 21 healthy controls
days) and 54 normal

(1) Is there evidence from the literature that

chronic cocaine users are different biolo-
gically?
(2) What biological parameters can be incor-
Subjects

porated into medication trials that are

meaningful and practical that will be
helpful in characterizing patient’s sub-
types?
Study

It was apparent that the literature was a mess and

(57)

(58)

that there has not been a systematic approach to

134 Ahmed Elkashef & Frank Vocci

Table 6. PET studies in cocaine-dependent patients

Study Subjects Tests Results Comments

(59) 11 polydrug abusers 11C-raclopride, with Sign. decrease in D2 All subjects reported
(10 years cocaine use, i.v. cocaine 48 mg occupancy with euphoria Direct
10 use opiates, all cocaine evidence of dopamine
marj.) involvement in
cocaine-related
pleasure
(60) 17 cocaine dep. (off 11C-cocaine with i.v. 60 – 77% of the DA
5 – 7 days, 10 years cocaine 0.3 – 0.6 mg/ transporter was
use) kg blocked. Self-
reported high
correlated with degree
of occupancy. At least
47% occup. is required
for a high
(61) 20 cocaine dep. (off 11C-raclopride, with Cocaine group showed
3 – 6 weeks) and 23 i.v. methylphenidate decrease DA release in
normal controls 0.5 mg/kg striatum and decrease
euphoria than
controls, and increase
DA release in the
thalamus which
correlated with craving
(30) 12 cocaine dep. (off 11C-cocaine (4 Patient had sign.
3 – 6 weeks) and 20 patients and 8 NC decreased uptake of
normal controls rescanned after 3 11C-cocaine persisted
mos). 10 patients and after 3 months. Patient
9 NC also had also had decreased D2
18F-NMS scan receptors. No corr.
with the DA
transporter
(62) 21 cocaine addicts (off FDG, 7 cocaine Decreased frontal
1 – 6 weeks) and 18 subjects retested after metabolic activity in
normal controls 3 months cocaine subjects,
persisted after 3
months
(31) 15 cocaine addicts (off FDG Cocaine subjects with- 12 subjects had
1 – 4 weeks) and 17 drawn 5 1 week had depression symp. at
normal controls increase metabolism of scan time, but no
orbitofrontal cortex relationship with
and BG compared depression and meta-
with cocaine subj. Off bolic changes. Strong
2 – 4 weeks and NC. corr. between frontal
no diff. between the metab. activity, and
latter two groups craving
(32) 10 cocaine dep.(7 off 18F-N-methyl Decreased D2 recep- Decrease D2 receptors
2 – 7 days, 3 off 4 – 5 spiperone tors occupancy in the 7 initially but then
weeks), 10 normal patients with short recover with prolonged
controls withdrawal and not in withdrawal
the other 3
(63) 8 polysubstance FDG, with 40 mg Cocaine caused 14% Negative corr.
abusers i.v. cocaine decrease in global between amygdala and
metabolism, and euphoria
regional decrease in
26/29 areas (mainly
neocortex, hippocam-
pus, BG, thalamus and
midbrain

(continued overleaf )

Table 6. (continued )
Study Subjects Tests
(64) 1. Cocaine abusers 1. FDG.
(off 2 – 5 days), two 2. F-DOPA
gps. of unipolar and
bipolar depressed. 2.
Cocaine abusers,
Parkinson’s and NC
attempt to answer these questions. The reasons
are:
. different durations and severity of cocaine
use as well as different periods of with-
drawal from cocaine, i.e. different recovery
time-lines;
. mixed drug use which may have opposing
or additive effects on specific biological
parameters; and
. different methods of measurements.
However, from the apparent chaos of the data,
there are some possible ‘‘leads’’ that we believe
are worth pursuing. Of the data reviewed, serum
PRL, EEG, PET and possibly ERG provide some
consistent findings in chronic cocaine users.
Based on the reported data on patient variability,
it appears that about 40 – 60% of chronic cocaine
users show evidence for low dopaminergic tone as
evidenced by increased prolactin, decreased b-
wave on the ERG and decreased D2 occupancy
and dopamine release on PET scans. These
measures could be useful tools to screen subjects
for their dopaminergic tone, keeping in mind the
limitations of each technique. Of the three types
of measures, serum prolactin may be the easiest
and most practical to obtain, although a specific
protocol has to be followed to assure data
accuracy as mentioned in the biochemistry
section. Although we focused on the dopamine
system, other biological measures could also be
included that address the serotonergic system as
well as the HPA stress axis (Table 8).
For future studies that may attempt to answer
these questions we would like to propose a
Biological markers of cocaine addiction 135
Results Comments
1. No difference in left
lat. prefrontal cortex
metab. between
cocaine b. and NC.
Cocaine sub. had
sign. elevated metab.
comp. with the
depressed groups.
2. Cocaine subj. had
sig. decreased f-dopa
uptake similar to
Parkinson’s patients
systematic approach that will probably help the
field.
(1) Ascertain the reliability and validity of
these measures alone and in combination
by assessing more than one and at least
two measures that would be expected to
change.
(2) The distinction between a state or a trait
marker has to be established. State mar-
kers would be extremely helpful if they
change consistently in relation to the drug
dependency state distinguishing acute vs.
chronic effects, or in the abstinence state
distinguishing between acute withdrawal
and the more prolonged craving state, and
lastly relapse.
(3) Trait markers would be major break-
throughs in this field. However, almost
impossible to detect without conducting
high-risk long-term studies, the type
would be most appropriate for genetic
studies. There is no consensus on the
definition of a ‘‘high-risk subject’’; how-
ever, family history of drug abuse, specific
personality traits, e.g. aggressiveness and
impulsivity and co-morbid mental illness
could certainly be incorporated into a
profile of the definition of a high-risk
patient.
(4) Studies have to adequately powered to
account for clinical issues, e.g. severity of
cocaine use, gender, co-morbid psychia-
tric conditions, e.g. depression, and
ADD. A stratification approach may be
useful.
136 Ahmed Elkashef & Frank Vocci

Table 7. SPECT studies in cocaine-dependent patients

Study Subjects Tests Results Comments

(65) 4 cocaine addicts 99m Tc-HMPAO Cocaine caused Vasoconstriction

with 40 mg i.v. 30% reduction in due to cocaine
cocaine absolute whole brain
blood flow
(66) Six cocaine addicts 123[I] B-CIT with 50% displacement
20 and 40 mg i.v. of CIT at 2.8 mg/kg
cocaine dose of cocaine
(26) 13 female (use 15.3 99m Tc-HMPAO Only male group Perfusion abn.
years) and 13 males (all 13 females used showed perfusion involved the frontal
(use 8.2 years) cocaine the day of Abn. females had and temporal cortex
cocaine addicts and the study (4 also normal scans except
26 NC used heroin), 9 men for the 4 that also
used cocaine used heroin
(67) 14 cocaine dep. 99m Tc-HMPAO Markedly decreased
(off 48 hours) with blood flow, 12 had
neurological abn. multiple lesions
(seizures, TIA) (43% in the parieto-
occipital region,
29% in the frontal)
(68) 10 polydrug abusers 99m Tc-HMPAO, Perfusion defects
after 2 – 3, 7 – 8, more common in
17 – 29 days of ab- patients dep. on
stinence. Bupronor- cocaine + alco. and
phine added on day heroin than cocaine
10 to end alone. Partial
improvement on
Bup
(69) 18 polydrug ab. 99m Tc-HMPAO 16/18 had perfusion
(7.7 years of defects mainly in the
cocaine) off 1 – 16 inferoparietal, tem-
days and 15 NC poral, anterofrontal,
and BG

Table 8. Proposed biological profile for two subgroups of in the dopaminergic system plasticity, in response
cocaine-dependent patients based on data reviewed
to chronic cocaine use (i.e. state). However, they
could also be present premorbidly (trait), but
;DA tone :DA tone
there are no data to support that. One recent PET
study35 in normal subjects reported that subjects
PRL : ;
ERG ;b-wave :b-wave with low D2 receptor density experienced a
EEG :b-Activity ;b-Activity pleasurable effect from methylphenidate com-
Cognitive Impaired Normal/im- pared with subjects with normal D2 density. It is
proved also possible to predict that response to medica-
PET ;D2 and DA No change
release tions could vary between subjects based on their
DA agonist Rx Good response Poor/no underlying biology. For example, subjects with
(prediction) response low dopaminergic tone may respond favorably to
dopamine agonist medications vs. subjects with
normal or high dopaminergic tone, which may
respond better to modulators of other neuro-
Based on this data review one could propose that transmitters or to an anxiety-reducing medication
chronic cocaine-dependent subjects have differ- (Fig. 1 near here).
ent underlying dopaminergic tone: low, high or Inclusion of these biomarkers in medications
normal (Fig. 1). These dopaminergic differences trials could prove useful in medication matching,
are probably a reflection of individual differences and in predicting response to medication. A

Figure 1. Graph representing proposed percentage of
subgroups of chronic cocaine users based on reviewed data.
recent publication36 addressed precisely this
issue. The authors investigated the relationship
between response to the SSRI citalopram and
serotonergic tone in a group of chronic alcoholics
using prolactin levels. Patients with high prolactin
levels had a significant therapeutic response to
citalopram, as manifested by reduction of alcohol
intake compared to those with low prolactin levels
who did not respond to citalopram.
Based on the success rate of the medications
studies for cocaine abuse so far it appears that we
will have multiple weak signals of therapeutic
efficacy from multiple medications than one
single medication that will be efficacious for all
patients. This makes the task of identifying
subgroups of patients with specific clinical and
biological characteristics that respond to a spe-
cific medication more crucial. The earlier we
include biomarkers in our medication studies, the
closer we will be to answering this question.
References
1. Lima MS, Reisser AAP, Soares BGO, Farrell M.
Antidepressants for cocaine dependence (Cochrane
Review). The Cochrane Library, Issue 1,2002.
Oxford: Update Software Ltd.
2. Mendelson JH, Teoh SK, Mello NK, Ellingboe J,
Rhoades E. Acute effects of cocaine on plasma
adenocorticotropic hormone, luteinizing hormone
and prolactin levels in cocaine-dependent men. J
Pharmacol Exp Ther 1992;263:505 – 9.
3. Rothman RB, Gorelick DA, Baumann MH, et al.
Lack of evidence for context-dependent cocaine-
induced sensitization in humans: preliminary stu-
dies. Pharmacol Biochem Behav 1994;49:583 – 8.
Biological markers of cocaine addiction 137
4. Elangovan N, Barbato M, Cooper T, Winsberg B.
Neurohormonal and behavioral response to methyl-
phenidate in cocaine abstinence. Psychiatry Res
1996;65:65 – 71.
5. Baumann MH, Gendron TM, Becketts KM,
Henningfield JE, Gorelick DA, Rothman RB.
Effects of intravenous cocaine on plasma cortisol
and prolactin in human cocaine abusers. Biol
Psychiatry 1995;38:751 – 5.
6. Gawin FH, Kleber HD. Neuroendocrine findings
in a chronic cocaine abusers: a preliminary report.
Br J Psychiatry 1985;147:569 – 73.
7. Swartz CM, Breen K, Leone F. Serum prolactin
levels during extended cocaine abstinence. Am J
Psychiatry 1990;147:777 – 9.
8. Knoblich G, Curtis D, Faustman WO, et al.
Increased CSF HVA with craving in long-term
abstinent cocaine abusers. Biol Psychiatry
1992;32:96 – 100.
9. Weiss RD, Hufford C, Mendelson JH. Serum
prolactin levels and treatment outcome in cocaine
dependence. Sc Biol Psychiatry 1994;35:573 – 4.
10. Kranzler HR, Wallington DJ. Serum prolactin
level, craving, and early discharge from treatment
in cocaine-dependent patients. Am J Drug Alcohol
Abuse 1992;18:187 – 95.
11. Faraj BA, Camp VM, Davis DC, Kutner M,
Cotsonis GA, Holloway T. Elevated concentrations
of dopamine sulfate in plasma of cocaine abusers.
Biochem Pharmacol 1993;46:1453 – 7.
12. Alper KR, Prichep LS, Kowalik S, Rosenthal MS,
Roy John E. Persistent QEEG abnormality in crack
cocaine users at 6 months of drug abstinence.
Neuropsychopharmacology 1998;19:1 – 9.
13. King DE, Herning RI, Gorelick DA, Cadet JL.
Gender differences in the EEG of abstinent cocaine
abusers. Neuropsychobiology 2000;42:93 – 8.
14. Herning RI, Guo X, Better WE, et al. Neurophy-
siological signs of cocaine dependence: increased
electroencephalogram beta during withdrawal. Biol
Psychiatry 1997;4:1087 – 94.
15. Herning RI, King DE. EEG and evoked potentials
alterations in cocaine-dependent individuals.
NIDA Res Monogr 1996;163:203 – 23.
16. Prichep LS, Alper K, Kowalik SC, Rosenthal M.
Neurometric QEEG studies of crack cocaine
dependence and treatment outcome. J Addict Dis
1996;15:39 – 53.
17. Prichep LS, Alper KR, Kowalik HM, MeeLee T,
John ER, Rosenthal MS. Quantitative electroence-
phalographic characteristics of crack cocaine de-
pendence. Biol Psychiatry 1996;40:986 – 93.
18. Prichep LS, Kowalik SC, Alper K, de Jesus C.
Quantitative EEG characteristics of children ex-
posed in utero to cocaine. Clin Electroencephalogr
1995;26:166 – 72.
19. Noldy NE, Santos CV, Politzer N, Blair RD,
Carlen PL. Quantitative EEG changes in cocaine
withdrawal: evidence for long-term CNS effects.
Neuropsychobiology 1994;30:189 – 96.
138 Ahmed Elkashef & Frank Vocci
20. Herning RI, Glover BJ, Koeppl B, Phillips RL,
London ED. Cocaine-induced increases in EEG
alpha and beta activity: evidence for reduced
cortical processing. Neuropsychopharmacology
1994;11:1 – 9.
21. Bauer LO. Frontal P300 decrements, childhood
conduct disorder, family history, and the prediction
of relapse among abstinent cocaine abusers. Drug
Alcohol Depend 1997;44:1 – 10.
22. Bauer LO. Psychomotor and electroencephalo-
graphic sequelae of cocaine dependence. NIDA
Res Monogr 1996;163:66 – 93.
23. Fein G, Biggins C, MacKay S. Cocaine abusers
have reduced auditory P50 amplitude and suppres-
sion compared to both normal controls and
alcoholics. Biol Psychiatry 1996;39:955 – 65.
24. Bauer LO, Mott AE. Differential effects of cocaine,
alcohol, and nicotine dependence on olfactory
evoked potentials. Drug Alcohol Depend
1996;42:21 – 6.
25. Bauer LO, Easton C. Pattern shift visual evoked
potentials in abstinent cocaine-dependent, alcohol-
dependent, and cross-dependent patients. Drug
Alcohol Depend 1996;40:203 – 9.
26. Levin JM, Holman BL, Mendelson JH, et al.
Gender differences in cerebral perfusion in cocaine
abuse: technetium-99m-HMPAO SPECT study of
drug-abusing women. J Nucl Med 1994;35:1902 –
9.
27. Prichep LS, Alper AR, Sverdlov L, et al. Outcome
related electrophysiological subtypes of cocaine
dependence. Clinical Electroencephalogr
2002;33:8 – 20.
28. Rosse RB, Risher-Flowers D, Peace T, Deutsch SI.
Evidence of impaired smooth pursuit eye move-
ment performance in crack cocaine users. Biol
Psychiatry 1992;31:1238 – 40.
29. Strickland TL, Mena I, Villanueva-Meyer J, et al.
Cerebral perfusion and neuropsychological conse-
quences of chronic cocaine use. J Neuropsychiatry
Clin Neurosci 1993;5:419 – 27.
30. Volkow ND, Wang GJ, Fowler JS, et al. Cocaine
uptake is decreased in the brain of detoxified
cocaine abusers. Neuropsychopharmacology
1996;14:159 – 68.
31. Volkow ND, Fowler JS, Wolf AP, et al. Changes in
brain glucose metabolism in cocaine dependence
and withdrawal. Am J Psychiatry 1991;148:621 – 6.
32. Volkow ND, Fowler JS, Wolf AP, et al. Effects of
chronic cocaine abuse on postsynaptic dopamine
receptors. Am J Psychiatry 1990;147:719 – 24.
33. Christensen JD, Kaufman MJ, Levin JM, et al.
Abnormal cerebral metabolism in polydrug abusers
during early withdrawal: a 31P MR spectroscopy
study. Magn Reson Med 1996;35:658 – 63.
34. MacKay S, Meyerhoff DJ, Dillon WP, Weiner
MW, Fein G. Alteration of brain phospholipid
metabolites in cocaine-dependent polysubstance
abusers. Biol Psychiatry 1993;15:261 – 4.
35. Volkow ND, Wang GJ, Fowler JS, et al. Prediction
of reinforcing responses to psychostimulants in
humans by brain dopamine D2 receptor levels. Am
J Psychiatry 1999;156:9.
36. Berggren U, Eriksson M, Fahlke C, Balldin J.
relationship between central serotonergic neuro-
tranmisson and reduction in alcohol intake by
citalopram. Drug Alcohol Depend 2001;63:263 – 7.
37. Dackis CA, Gold MS. New concepts in cocaine
addiction: the dopamine depletion hypothesis.
Neurosci Biobehav Rev 1985;9:469 – 77.
38. Cocores JA, Dackis CA, Gold MS. Sexual dysfunc-
tion secondary to cocaine abuse in two patients. J
Clin Psychiatry 1986;47:7.
39. Mendelson JH, Teoh SK, Lange U, et al. Anterior
pituitary, adrenal, and gonadal hormones during
cocaine withdrawal. Am J Psychiatry
1988;145:1094 – 8.
40. Demer JL, Volkow ND, Ulrich I, Krajewski K,
Davis CM, Porter FI. Eye movements in cocaine
abusers. Psychiatry Res 1989;29:123 – 36.
41. Martin SD, Yeragani VK, Lodhi R, Galloway MP.
Clinical ratings and plasma HVA during cocaine
abstinence. Biol Psychiatry 1989;26:356 – 62.
42. Lee MA, Bowers MM, Nash JF, Meltzer HY.
Neuroendocrine measures of dopaminergic func-
tion in chronic cocaine users. Psychiatry Res
1990;33:151 – 9.
43. Satel SL, Price LH, Palumbo JM, et al. Clinical
phenomenology and neurobiology of cocaine ab-
stinence: a prospective in patient study. Am J
Psychiatry 1991;148:1712 – 16.
44. Rosse RB, McCarthy MF, Alim TN, Deutsch SI.
Saccadic distractibility in cocaine dependent pa-
tients: a preliminary laboratory exploration of the
cocaine – OCD hypothesis. Drug Alcohol Depend
1994;35:25 – 30.
45. Rosse RB, Johri S, Kendrick K, et al. Preattentive
and attentive eye movements during visual scan-
ning of a cocaine cue: correlation with intensity of
cocaine cravings. J Neuropsychiatry Clin Neurosci
1997;9:91 – 3.
46. Bauer LO. Eye movements in recovering substance
abusers: a prospective study. Addict Behav
1993;18:465 – 72.
47. Roy M, Roy A, Smelson D, Brown S, Weinberger
L. Reduced blue cone electroretinogram in with-
drawn cocaine dependent patients: a replication.
Biol Psychiatry 1997;42:631 – 3.
48. Roy M, Smelson DA, Roy A. Abnormal electro-
retinogram in cocaine-dependent patients. Rela-
tionship to craving. Br J Psychiatry 1996;168:507 –
11.
49. Smelson DA, Roy A, Roy M. The electroretino-
gram and neuropsychological functioning in co-
caine addicts [letter]. Can J Psychiatry 1996;6:415.
50. Manschreck TC, Schneyer ML, Weisstein CC, et
al. Freebase cocaine and memory. Compr Psychia-
try 1990;31:369 – 75.
51. O’Malley S, Adamse M, Heaton RK, Gawin FH.
Neuropsychological impairment in chronic cocaine
abusers. Am J Drug Alcohol Abuse 1992;18:131 –
44.
52. Berry J, Van Gorp WG, Herzberg DS, et al.
Neuropsychological deficits in abstinent cocaine
abusers: preliminary findings after two weeks of
abstinence. Drug Alcohol Depend 1993;32:231 – 7.

53. Roberts LA, Bauer LO. Reaction time during
cocaine versus alcohol withdrawal: longitudinal
measures of visual and auditory suppression.
Psychiatry Res 1993;46:229 – 37.
54. Beatty WW, Katzung VM, Moreland VJ, Nixon SJ.
Neuropsychological performance of recently absti-
nent alcoholics and cocaine abusers. Drug Alcohol
Depend 1995;37:247 – 53.
55. Rosselli M, Ardila A. Cognitive effects of cocaine
and polydrug abuse. J Clin Exp Neuropsychol
1996;18:122 – 35.
56. Easton C, Bauer LO. Neuropsychological corre-
lates of urine toxicology results. Prog Neuropsy-
chopharmacol Biol Psychiatry 1996;20:969 – 82.
57. Hoff AL, Riordan H, Morris L, et al. Effects of
crack cocaine on neurocognitive function. Psychia-
try Res 1996;60:167 – 76.
58. Bolla KI, Rothman R, Cadet JL. Dose-related
neurobehavioral effects of chronic cocaine use. J
Neuropsychiatry Clin Neurosci 1999;11:361 – 9.
59. Schlaepfer TE, Pearlson GD, Wong DF, Marenco
S, Dannals RF. PET study of competition between
intravenous cocaine and [11C] raclopride at
dopamine receptors in human subjects. Am J
Psychiatry 1997;154:1209 – 13.
60. Volkow ND, Wang GJ, Fischman MW, et al.
Relationship between subjective effects of cocaine
and dopamine transporter occupancy. Nature
1997;4:827 – 30.
61. Volkow ND, Wang GJ, Fowler JS, et al. Decreased
striatal dopaminergic responsiveness in detoxified
cocaine-dependent subjects. Nature 1997;24:830 –
3.
Biological markers of cocaine addiction 139
62. Volkow ND, Hitzemann R, Wang GJ, et al. Long-
term frontal brain metabolic changes in cocaine
abusers. Synapse 1992;11:184 – 90.
63. London ED, Cascella NG, Wong DF, et al.
Cocaine-induced reduction of glucose utilization
in human brain. A study using positron emission
tomography and [fluorine 18]-fluorodeoxyglucose.
Arch Gen Psychiatry 1990;47:567 – 74.
64. Baxter LR Jr, Schwartz JM, Phelps ME, et al.
Localization of neurochemical effects of cocaine
and other stimulants in the human brain. J Clin
Psychiatry 1988;(Suppl 49):23 – 6.
65. Wallace EA, Wisniewski G, Zubal G, et al. Acute
cocaine effects on absolute cerebral blood flow.
Psychopharmacology (Berl) 1996;128:17 – 20.
66. Malison RT, Best SE, Wallace EA, et al. Euphori-
genic doses of cocaine reduce [123I]beta-CIT
SPECT measures of dopamine transporter avail-
ability in human cocaine addicts. Psychopharma-
cology (Berl) 1995;122:358 – 62.
67. Mena I, Giombetti RJ, Miller BL, et al. Cerebral
blood flow changes with acute cocaine intoxication:
clinical correlations with SPECT, CT, and MRI.
NIDA Res Monogr 1994;138:161 – 73.
68. Holman BL, Mendelson J, Garada B, et al.
Regional cerebral blood flow improves with treat-
ment in chronic cocaine polydrug users. J Nucl
Med 1993;34:723 – 7.
69. Holman BL, Carvalho PA, Mendelson J, et al.
Brain perfusion is abnormal in cocaine-dependent
polydrug users: a study using technetium-99m-
HMPAO and ASPECT. J Nucl Med
1991;32:1206 – 10.