SCHLAEPFER,

Am J Psychiatry
PEARLSON, 11
154:9, September
WONG,1997
ET AL.

INTRAVENOUS COCAINE AND [ C]RACLOPRIDE

Regular Articles

PET Study of Competition Between Intravenous Cocaine

and [11C]Raclopride at Dopamine Receptors in Human Subjects

Thomas E. Schlaepfer, M.D., Godfrey D. Pearlson, M.D., Dean F. Wong, M.D.,

Stefano Marenco, M.D., and Robert F. Dannals, Ph.D.

Objective: Animal data suggest that the strong euphoriant effects of cocaine are related to
the drug’s enhancement of available dopamine at the synaptic cleft. The authors’ goal was to
determine whether this mechanism is the same in humans because the development of putative
pharmacological agents for treatment of cocaine dependence depends on this knowledge.
Method: Positron emission tomography with [11C]raclopride was used to examine the effects
of the intravenous administration of 48 mg of cocaine (a typical “street” dose) on the occu-
pancy of dopamine 2 receptors in the putamen of 11 self-identified intravenous drug abusers.
Results: All 11 subjects reported subjective stimulation and euphoria in response to cocaine
administration. Radioligand occupancy at dopamine receptors was decreased significantly
after cocaine administration, suggesting that higher dopamine concentrations were competing
at the receptor site. Conclusions: These results support the concept of dopamine system in-
volvement in human cocaine abuse.
(Am J Psychiatry 1997; 154:1209–1213)

C ocaine is a commonly abused (1) stimulant and

euphoriant (2) with strong reinforcing effects (3–
5). The mode of action of cocaine has been the subject
of scientific interest for more than a century (2). The
more recent awareness of the existence of a “cocaine
Presented in part at the 148th Annual Meeting of the American
Psychiatric Association, May 20–25, 1995, Miami. Received March
28, 1996; revised Dec. 2, 1996, and Feb. 13, 1997; accepted Feb. 25,
1997. From the Division of Psychiatric Neuroimaging, Department
of Psychiatry, and the Division of Nuclear Medicine, Department of
Radiology, The Johns Hopkins Medical Institutions, Baltimore, and
the Psychiatric Neuroimaging Group, Department of Psychiatry, Uni-
versity of Bern, Switzerland. Address reprint requests to Dr. Schlaep-
fer, Psychiatric Neuroimaging Group, Department of Psychiatry, Uni-
versity of Bern, Murtenstrasse 21, 3010 Bern, Switzerland; schlaepf@
welchlink.welch.jhu.edu (e-mail).
Supported by fellowship grant 81BE-33483 from the Swiss Na-
tional Science Foundation, the ROCHE Research Foundation, and
the CIBA Research Foundation (Dr. Schlaepfer); the Johns Hopkins
Hospital Inpatient Clinical Research Center through NIH grant RR-
00035 from the Department of Research Resources (Dr. Pearlson and
Dr. Schlaepfer); NIDA grant DA-05317 (Dr. Pearlson); and the Na-
tional Alliance for Research on Schizophrenia and Depression and
NIMH grant MH-42821 (Dr. Wong).
The authors thank David Clough for technical assistance with the
positron emission tomography studies and Roger Noss for reviewing
the manuscript.
abuse epidemic” has added a sociopolitical dimension
to research on this drug (6) and has resulted in a more
vigorous effort in the search for a specific pharma-
cological approach to cocaine dependency. To establish
a putative treatment algorithm, cocaine’s euphorigenic
effects on the human brain must be explained.
It has been demonstrated in animals that cocaine ex-
erts its effects by interacting with dopaminergic syn-
apses (7). Cocaine likely interacts with the dopa-
minergic system by binding to the dopamine
transporter, thus inhibiting dopamine reuptake from
the synaptic cleft into the presynaptic terminal (8).
This mechanism leads to an increase of available dopa-
mine. The neural basis of cocaine reinforcement iden-
tified in animal models seems to involve an enhance-
ment of dopaminergic neurotransmission in dopamine
cells situated in the ventral tegmental area (9, 10). Al-
though broadly supported by animal studies, the “do-
pamine hypothesis” of cocaine’s actions has not been
supported by studies in humans (11).
It is difficult to compare directly the euphoriant and
reinforcing effects of cocaine exhibited in animal models
with these effects in humans. Therefore, we set out to
correlate possible changes in dopamine concentration
with subjective self-ratings of euphoria after an acute

Am J Psychiatry 154:9, September 1997 1209

INTRAVENOUS COCAINE AND [11C]RACLOPRIDE
FIGURE 1. Six Regions of Interest on PET Slice on Which Quantita-
tive Analyses Were Performed for 11 Intravenous Drug Abusersa
aThe two slices on which the basal ganglia were best visible were se-
lected from a total of 15 axial slices and averaged. The resulting
image was filtered by using a Hanning filter, and three regions of
interest were placed interactively on each left and right putamen by
a rater who was blind to the experimental condition. The activities
of the six regions of interest were averaged to obtain activity in the
putamen. The figure represents a 4-minute sampling period 18 min-
utes after injection of [11C]raclopride in one subject.
challenge with intravenous cocaine. Our design used
positron emission tomography (PET) with a radiola-
beled, reversibly binding dopamine antagonist that is
easily displaced by competition of endogenous dopamine.
[11C]Raclopride is a positron-emitting selective dopa-
mine 2 (D2) and dopamine 3 (D3) receptor antagonist.
This tracer has previously been used to demonstrate dif-
ferential D2 receptor occupancies in the basal ganglia of
patients with schizophrenia treated with several differ-
ent competitive antagonists (12). [11C]Raclopride has a
relatively lower affinity for D2 and D3 receptors than
the frequently used tracers [18F]N-methylspiroperidol
or [11C]N-methylspiroperidol; therefore, [11C]raclo-
pride can be more easily displaced by an increase of
endogenous dopamine. This displacement has been
demonstrated in animal (rodent) studies (13) as well as
in human autopsy work (14). Volkow et al. (15) re-
ported displacement of [11C]raclopride in human sub-
jects by increased extracellular dopamine levels result-
ing from the administration of the dopamine uptake
blocker methylphenidate. Using [11C]raclopride in hu-
man subjects, Volkow et al. (15) demonstrated changes
in dopamine concentration induced by methylpheni-
1210
date. Therefore, we hypothesized that acute intrave-
nous administration of cocaine in human subjects
would raise the intrasynaptic concentration of endo-
genous dopamine sufficiently to displace [11C]raclo-
pride from its competitive binding with the D2 receptor.
METHOD
Subjects
We studied 11 men (mean age=36 years, SD=5.2, range=28–44)
who had been abusing intravenous drugs for more than 10 years. All
11 men had the DSM-III-R diagnosis of cocaine abuse, and all regu-
larly abused cocaine and marijuana. Ten men also abused opiates
intermittently. Ten men were African American and one was Cauca-
sian. Subjects were recruited by word of mouth from the East Balti-
more community. All of the men were healthy as assessed by a thor-
ough physical examination and basic medical laboratory results,
including chest X-ray and performance on a stress ECG. Plasma
pseudocholinesterase activity was measured in all of the men to en-
sure adequate metabolism of cocaine-like agents as an additional
safety measure. Written informed consent was obtained according to
the guidelines of the Joint Committee on Clinical Investigation of
Johns Hopkins Hospital.
Subjects refrained from taking any drugs for at least 48 hours be-
fore hospitalization. Twenty-four hours before the beginning of the
PET study, the men were admitted to the inpatient Clinical Research
Center of Johns Hopkins Hospital, where an observed urine toxico-
logical screen of commonly abused substances was performed to
check for compliance with the instructions on drug abstinence. None
of the subjects screened positive for any drug of abuse.
PET Studies
On the day of the study, subjects were scanned with a GE4096-plus
whole-body PET camera (resolution was 5–6 mm in plane, 6–7 mm
z axis). The men were given a mean of 20.6 mCi (SD=10.5) of intra-
venous [11C]raclopride at baseline and a mean of 19.7 mCi (SD=20.7)
of intravenous [11C]raclopride during the drug scan; injected doses
were not significantly different (t=1.45). The specific activities be-
tween baseline and the drug scans (mean=3347 mCi/mmol, SD=
2963, and mean=2096 mCi/mmol, SD=1832) were also not different
(t=1.40). After 0–10 minutes, subjects were injected over a period of
approximately 40 seconds with either saline in the placebo scan or 48
mg of cocaine hydrochloride in the drug scan. The cocaine scan al-
ways followed the placebo scan by 4–6 hours. Since the time to reach
equilibrium is variable among subjects, it was difficult a priori to
choose an ideal time for the cocaine injection. Therefore, we system-
atically varied the time of drug injection: six men received placebo or
cocaine 10 minutes after the [11C]raclopride injection, two subjects
received placebo or cocaine 5 minutes after, one subject received pla-
cebo or cocaine 0 minutes after, and two subjects received placebo or
cocaine in a split-dose regimen of 32 mg after 4 minutes and 16 mg
after 10 minutes. Throughout the study, the subjects’ ECG, blood
pressure, and heart rate were monitored. Blood pressure and heart
rate were also recorded.
Starting 2 minutes before the injection of cocaine or placebo and
every 2 minutes thereafter, a simple visual analog scale questionnaire
(16) was administered for the first 20 minutes of the study. Subjects
used the fingers of the dominant hand to indicate, on a scale of 0–5,
the subjectively perceived strength of the drug effect. This particular
procedure was chosen because verbal participation of subjects would
result in head movement and lead to difficulties in image analysis.
Subjects wore individualized thermoplastic face masks to stand-
ardize positioning in the scanner with the aid of a laser beam. A CT
scan determined which angle of alignment was parallel to the orbi-
tomeatal line. A radial arterial line was placed to obtain blood sam-
ples for input sampling. Scans were acquired at 50 time points over a
period of 90 minutes. Acquisition duration varied from 15 to 240
Am J Psychiatry 154:9, September 1997

FIGURE 2. Individual Time-Activity Curves Showing Displacement of [11C]Raclopride for 11 Intravenous Drug Abusers During Administration
of Saline (Placebo) or Cocaine Hydrochloride
seconds; the first eight scans were acquired over 15 seconds, the next
16 over 30 seconds, the next eight over 60 seconds, and the remaining
18 over 240 seconds.
Image Analysis
Time-activity curves for putamen and cerebellum were derived in
both the placebo and the drug scans. Fifteen axial slices of 6.5-mm
full width at half maximum were acquired. Images were recon-
structed in a 128×128 matrix with a pixel size of 2 mm. The two scan
slices where the basal ganglia were best visible were averaged, and the
resulting image was low-pass filtered by using a 6-mm cutoff Hanning
filter to reduce statistical noise. The same procedure was applied to
two slices where the cerebellum was best visible. To assess a time-ac-
tivity curve of total binding to receptors in the basal ganglia, three
4×4-pixel-wide regions of interest were subsequently placed by a
rater, blind to the experimental condition, for each left and right pu-
tamen on every one of the 50 averaged slices (figure 1) of the dynamic
acquisition. A 6×8-pixel-wide region of interest was placed in each
left and right cerebellar slice to assess nonspecific binding. Specific
binding was calculated by subtracting the cerebellar activity from the
activity in the putamen, both corrected for the amount of injected
radioactivity.
Statistical Analysis
We used the Wilcoxon signed rank test to compare the scores of
the subjective measures for cocaine effects because the distribution of
values was not normal. We used paired t tests to compare blood pres-
sure and heart rate as well as activity measures 7.5 minutes after in-
jection of cocaine or placebo, respectively. For both conditions we
Am J Psychiatry 154:9, September 1997
SCHLAEPFER, PEARLSON, WONG, ET AL.
used a paired t statistic to calculate and compare the total area under
the time-activity curve (corrected for injected activity) as assessed
with the Riemann method.
RESULTS
All subjects tolerated the experimental procedures
well. According to their ratings on the visual analog
scale (16), all of the men reported various degrees of
greater subjective euphoriant effects after cocaine than
after placeo administration. Peak values for self-re-
ported “rush” and “high” were reached 2 minutes after
cocaine injection. All subjects reported increases over
baseline after cocaine administration: the mean increase
in ratings of “rush” was 3.4 (SD=1.5) (Wilcoxon Z=
2.8, N=11, p<0.005); the mean increase in ratings of
“high” was 2.8 (SD=1.7) (Wilcoxon Z=2.8, N=22,
p<0.005).
Blood pressure and heart rate also differed between
baseline conditions and 7.5 minutes after cocaine injec-
tion: the mean change in systolic pressure was 14.5 mm
Hg (SD=9.54) (one-tailed t=5.1, df=10, p<0.001); the
mean change in diastolic pressure was 8.2 mm Hg
(SD=8.1) (one-tailed t=3.4, df=10, p<0.01); the mean
change in heart rate was 29.6 bpm (SD=24.3) (one-
tailed t=4.0, df=10, p<0.01). There was a significant
1211
INTRAVENOUS COCAINE AND [11C]RACLOPRIDE
correlation between the self-reported values for “rush”
and “high” 2 minutes after the first cocaine injection
and the degree of change in systolic (r=0.64 for “rush
and r=0.68 for “high”) and diastolic (r=0.69 for “rush”
and r=0.66 for “high”) blood pressure from 1 minute
before cocaine injection to 7 minutes after.
As shown in figure 2, eight of the 11 men showed
lower activity in the putamen during the cocaine condi-
tion than in the placebo condition. Previous data (17)
suggested that maximal physiological cocaine effects
are reached 8 minutes after injection. Therefore, we cal-
culated a paired means comparison for the difference in
specific binding 7.5 minutes after injection of cocaine.
This analysis revealed significant differences (t=4.0, df=
10, p<0.01). The total mean area under the time-activ-
ity curve (corrected for the amount of injected dose of
radioactivity) as assessed with the Riemann method
was 251.0 (SD=45.5) in the placebo condition and
224.1 (SD=39.1) in the cocaine condition. A paired t
test revealed a significant difference between the two
conditions (t=6.7, df=10, p<0.0001).
DISCUSSION
Converging evidence from preclinical research gen-
erated over the last 25 years has convincingly estab-
lished the pivotal role of dopamine in stimulant reward
(18–22). The neuroanatomical correlates of these
stimulant euphorigenic mechanisms in the dopa-
minergic reward regions of the human brain is very
important because it might provide new avenues
for potential treatments (23). Despite this, we are
aware of only one study that tried to correlate cocaine-
induced euphoria with changes in dopamine concen-
tration in human subjects (24).
Our study of 11 men demonstrated significant dis-
placement of a radiolabeled D2 receptor antagonist in
response to intravenous administration of 48 mg of co-
caine hydrochloride. This dose, similar to typical
“street” doses, exerted clear subjective pharmacologi-
cal effects that are in agreement with those we reported
previously (16). In view of the available in vitro and in
vivo animal data, this effect is most likely associated
with an increase in intrasynaptic dopamine levels, pro-
duced by a blockade of dopamine reuptake due to the
binding of cocaine to the dopamine transporter. Our
results also demonstrate that increases in dopamine in
the putamen, at least as measured by [11C]raclopride
binding, are not necessarily correlated with cocaine
“rush” and “high.” Since it is dopamine release in the
nucleus accumbens (25) and possibly the ventral pal-
lidum (26), but not the striatum, that seems to be re-
sponsible for the euphoric effects of cocaine and am-
phetamine, the assumption must be made that the
changes in dopamine levels in the putamen observed in
the current study do not reflect similar changes in the
accumbens.
A weakness of this study is the use of varying time
intervals between [11C]raclopride and cocaine admini-
1212
stration, which mars the comparability of the results.
This procedure was chosen because we were not able to
identify an ideal time for the cocaine injection a priori.
Cocaine acts as a potent vasoconstrictor and is asso-
ciated clinically with both cardiovascular problems (27)
and stroke (28). One might argue that the effects of co-
caine demonstrated in this study are attributable to de-
creased cerebral blood flow and resulting slower deliv-
ery of [11C]raclopride to D2 basal ganglia receptors. We
tried to defend against this potential criticism by ad-
ministering cocaine to most subjects 10 minutes after
[11C]raclopride administration. The similar time-activ-
ity curves in the cerebellum in both placebo and drug
conditions additionally suggest that blood flow differ-
ences did not contribute to our findings.
Continued use of cocaine reportedly can cause both
tolerance and sensitization, suggesting adaptive changes
in brain neurochemistry (29). Our observation of a
rather large standard deviation in the maximal activi-
ties in the putamen is consistent with such effects. Fur-
thermore, the degree of subjective euphoriant drug ef-
fects, as assessed with a visual analog scale, correlated
negatively with age, analogous to a report by Volkow
et al. (15). The subject who had the lowest activity (the
second graph from the left in the top row of figure 2)
reported the longest history of cocaine abuse (22 years).
This observation is consistent with previous findings
(30) suggesting that chronic cocaine abuse may affect
dopaminergic receptors.
In summary, we believe that our data offer support
for the application of the dopamine hypothesis of co-
caine’s actions (developed with animal data) to human
subjects. Furthermore, our data confirm that it is pos-
sible, by using functional radioligand imaging tech-
niques, to demonstrate acute central actions of pharma-
cological agents of abuse on human dopaminergic
neurotransmitter systems in vivo.
REFERENCES
1. Anthony JC: Epidemiological research on cocaine use in the
USA. Ciba Found Symp 1992; 166:20–33
2. Freud S: Über Coca. Zentralblatt für die Gesamte Therapie 1885;
2:289–314
3. Kornetsky C, Esposito RU, McLean S, Jacobson JO: Intracranial
self-stimulation thresholds: a model for the hedonic effects of
drugs of abuse. Arch Gen Psychiatry 1979; 36:289–292
4. Spealman RD, Madras BK, Bergman J: Effects of cocaine and
related drugs in nonhuman primates, II: stimulant effects on
schedule-controlled behavior. J Pharmacol Exp Ther 1989; 251:
142–149
5. Griffiths RR, Brady JV, Bigelow GE: Predicting the dependence
liability of stimulant drugs. NIDA Res Monogr 1981; 37:182–
196
6. Nicholi AJ: The nontherapeutic use of psychoactive drugs: a
modern epidemic. N Engl J Med 1983; 308:925–933
7. Ritz MC, Kuhar MJ: Psychostimulant drugs and a dopamine hy-
pothesis regarding addiction: update on recent research. Bio-
chem Soc Symp 1993; 59:51–64
8. McMillan BA: CNS stimulants: two distinct mechanisms of ac-
tion for amphetamine-like drugs. Trends Pharmacol Sci 1983;
4:429–432
9. Bozarth MA: New perspectives on cocaine addiction: recent find-
Am J Psychiatry 154:9, September 1997

ings from animal research. Can J Physiol Pharmacol 1989; 67:
1158–1167
10. Kuhar MJ, Ritz MC, Boja JW: The dopamine hypothesis of the
reinforcing properties of cocaine. Trends Neurosci 1991; 14:
299–302
11. Kuhar MJ: Molecular pharmacology of cocaine: a dopamine hy-
pothesis and its implications. Ciba Found Symp 1992; 166:81–
89
12. Farde L, Nordstrom AL, Wiesel FA, Pauli S, Halldin C, Sedvall
G: Positron emission tomographic analysis of central D1 and D2
dopamine receptor occupancy in patients treated with classical
neuroleptics and clozapine: relation to extrapyramidal side ef-
fects. Arch Gen Psychiatry 1992; 49:538–544
13. Young LT, Wong DF, Goldman S, Minkin E, Chen C, Matsu-
mura K, Scheffel U, Wagner HN: Effects of endogenous dopa-
mine on kinetics of [3H]N-methylspiperone and [3H]raclopride
binding in the rat brain. Synapse 1991; 9:188–194
14. Seeman P, Niznik HB, Guan H-C: Elevation of dopamine D2
receptors in schizophrenia is underestimated by radioactive rac-
lopride. Arch Gen Psychiatry 1990; 47:1170–1172
15. Volkow ND, Wang GJ, Fowler JS, Logan J, Schlyer D, Hitze-
mann R, Lieberman J, Angrist B, Pappas N, Macgregor R, Burr
G, Cooper T, Wolf AP: Imaging endogenous dopamine compe-
tition with (11C)raclopride in the human brain. Synapse 1994;
16:255–262
16. Pearlson GD, Jeffery PJ, Harris GJ, Ross CA, Fischman MW,
Camargo EE: Correlation of acute cocaine-induced changes in
local cerebral blood flow with subjective effects. Am J Psychiatry
1993; 150:495–497
17. Fischman MW, Schuster CR, Resnekov L, Shick JF, Krasnegor
NA, Fennell W, Freedman DX: Cardiovascular and subjective
effects of intravenous cocaine administration in humans. Arch
Gen Psychiatry 1976; 33:983–989
18. Fibiger HC, Phillips AG, Brown EE: The neurobiology of co-
Am J Psychiatry 154:9, September 1997
SCHLAEPFER, PEARLSON, WONG, ET AL.
caine-induced reinforcement. Ciba Found Symp 1992; 166:96–
111
19. Pulvirenti L, Koob GF: Dopamine receptor agonists, partial
agonists and psychostimulant addiction. Trends Pharmacol Sci
1994; 15:374–379
20. Koob GF: Neurobiological mechanisms in cocaine and opiate de-
pendence. Res Publ Assoc Res Nerv Ment Dis 1992; 70:79–92
21. Holman RB: Biological effects of central nervous system stimu-
lants. Addiction 1994; 89:1435–1441
22. Kiyatkin EA: Dopamine mechanisms of cocaine addiction. Int J
Neurosci 1994; 78:75–101
23. Gawin FH, Ellinwood EH Jr: Cocaine and other stimulants: ac-
tions, abuse, and treatment. N Engl J Med 1988; 318:1173–1182
24. Malison RT, Best SE, Wallace EA, McCance E, Laruelle M, Zogh-
bi SS, Baldwin RM, Seibyl JS, Hoffer PB, Price LH, Kosten TR,
Innis RB: Euphorigenic doses of cocaine reduce [123I]beta-CIT
SPECT measures of dopamine transporter availability in human
cocaine addicts. Psychopharmacology (Berl) 1995; 122:358–362
25. Koob G, Caine B, Markou A, Pulvirenti L, Weiss F: Role for the
mesocortical dopamine system in the motivating effects of co-
caine. NIDA Res Monogr 1994; 145:1–18
26. Gong W, Neill D, Justice J: Conditioned place preference and
locomotor activation produced by injection of psychostimulants
into ventral pallidum. Brain Res 1996; 707:64–74
27. Goldfrank LR, Hoffman RS: The cardiovascular effects of co-
caine. Ann Emerg Med 1991; 20:165–175
28. Klonoff DC, Andrews BT, Obana WG: Stroke associated with
cocaine use. Arch Neurol 1989; 46:989–993
29. Woolverton WL, Johnson KM: Neurobiology of cocaine abuse.
Trends Pharmacol Sci 1992; 13:193–200
30. Volkow ND, Fowler JS, Wolf AP, Hitzemann R, Dewey S, Ben-
driem B, Alpert R, Hoff A: Changes in brain glucose metabolism
in cocaine dependence and withdrawal. Am J Psychiatry 1991;
148:621–626
1213