Content uploaded by Larry W Hawk
Author content
All content in this area was uploaded by Larry W Hawk
Content may be subject to copyright.
Effects of Dopamine Transporter and Receptor Polymorphisms
on Smoking Cessation in a Bupropion Clinical Trial
Caryn Lerman
University of Pennsylvania Peter G. Shields
Georgetown University
E. Paul Wileyto and Janet Audrain
University of Pennsylvania Larry H. Hawk Jr.
State University of New York at Buffalo
Angela Pinto and Susan Kucharski
University of Pennsylvania Shiva Krishnan
Georgetown University
Ray Niaura
Brown University Leonard H. Epstein
State University of New York at Buffalo
This study examined the role of dopaminergic genes in prospective smoking cessation and response to
bupropion treatment in a placebo-controlled clinical trial. Smokers of European ancestry (N⫽418)
provided blood samples for genetic analysis and received either bupropion or placebo (10 weeks) plus
counseling. Assessments included the dopamine D2 receptor (DRD2) genotype, dopamine transporter
(SLC6A3) genotype, demographic factors, and nicotine dependence. Smoking status was verified at the
end of treatment (EOT) and at 6-month follow-up. The results provided evidence for a significant
DRD2 ⫻SLC6A3 interaction effect on prolonged smoking abstinence and time to relapse at EOT,
independent of treatment condition. Such effects were no longer significant at 6-month follow-up,
however. These results provide the first evidence from a prospective clinical trial that genes that alter
dopamine function may influence smoking cessation and relapse during the treatment phase.
Key words: tobacco, smoking, treatment, genetics, bupropion, dopamine
To improve treatment of nicotine dependence, researchers are
increasing attention to the genetic basis of smoking behavior.
Although the task of identifying the specific genetic variants that
predispose individuals to smoking behavior has proved daunting,
initial studies are yielding some promising results. Much of this
research has focused on the neurotransmitter dopamine and genetic
variants relevant to dopamine synthesis, reuptake (i.e., removal of
excess dopamine from the neuronal synapse), and receptors. Do-
pamine is considered an important substrate for the reinforcing
effects of nicotine (Henningfield, 1995; Pontieri, Tanda, Orzi, &
Di Chiara, 1996) as well as for the behavioral conditioning pro-
cess in general addiction (Garris et al., 1999; Schultz, 1998).
These processes have obvious implications for nicotine self-
administration, the reinforcing properties of smoking, and smoking
cessation treatment.
In considering the biological evidence relating dopamine to
nicotine reinforcement, one should note that genes affecting do-
pamine function may contribute to individual differences in nico-
tine dependence. To be specific, individuals with lower activity
genetic variants (i.e., those associated with low-endogenous dopa-
mine levels) may be more likely to self-administer nicotine
through smoking to boost dopamine levels. Preliminary support for
the effects of genes in the dopamine pathway is found in two
studies of a polymorphism (variation) in the dopamine transporter
(SLC6A3) gene. In a case control study of smokers and nonsmok-
ers, persons with at least one copy of the 9-repeat allele (variant)
of SLC6A3 were significantly less likely to be smokers than those
with no copies of the 9-repeat allele (mostly SLC6A3-10 geno-
types; Lerman et al., 1999). Although the proportion of variance
accounted for by SLC6A3 alone was modest, there was evidence
for an interaction that was statistically and clinically significant.
Caryn Lerman, E. Paul Wileyto, Janet Audrain, Angela Pinto, and Susan
Kucharski, Department of Psychiatry, Abramson Cancer Center, Univer-
sity of Pennsylvania; Peter G. Shields and Shiva Krishnan, Department of
Oncology, Georgetown University; Larry H. Hawk Jr., Department of
Psychology, State University of New York at Buffalo; Ray Niaura, Brown
Medical School, Brown University; Leonard H. Epstein, Department of
Pediatrics, State University of New York at Buffalo.
This research was supported by National Cancer Institute and National
Institute on Drug Abuse Grants CA/DA P50 84718 and CA63562 to Caryn
Lerman. We acknowledge Jodie Jaroni and Vyga Kaufmann for their
assistance with project coordination and Maryanne Foster for her assis-
tance with manuscript preparation. We thank Colin Feyerabend and Martin
Jarvis for assistance with saliva cotinine determination.
Correspondence concerning this article should be addressed to Caryn
Lerman, Tobacco Use Research Center, University of Pennsylvania, 3535
Market Street, Suite 4100, Philadelphia, Pennsylvania 19104. E-mail:
clerman@mail.med.upenn.edu
Health Psychology Copyright 2003 by the American Psychological Association, Inc.
2003, Vol. 22, No. 5, 541–548 0278-6133/03/$12.00 DOI: 10.1037/0278-6133.22.5.541
541
Among persons with dopamine D2 receptor (DRD2)-A2 geno-
types, 62% of nonsmokers had SLC6A3-9 genotypes compared
with 46% of smokers. The association of SLC6A3 with smoking
cessation history was also significant. Sabol et al. (1999) docu-
mented the effect of SLC6A3 on smoking cessation, but two other
studies examining SLC6A3 found no association with smoking
status (Jorm et al., 2000; Vandenbergh et al., 2002). We should
note, however, that these two studies did not determine DRD2
genotype and, therefore, could not test for the DRD2–SLC6A3
interaction observed in the initial report (Lerman et al., 1999).
Several additional genetic factors in the dopamine pathway have
been examined for associations with smoking. Previous studies
found a higher prevalence of the more rare A1 or B1 allele of the
DRD2 gene among smokers than among control subjects (Com-
ings et al., 1996; Noble et al., 1994; Spitz et al., 1998). However,
a family-based analysis (Bierut et al., 2000) did not provide
evidence for significant linkage of smoking to the DRD2 gene
locus. Genetic variation in dopamine beta hydroxylase, the enzyme
that catalyzes the conversion of dopamine to norepinephrine
(thereby reducing dopamine levels), has been linked with cigarette
smoking rates (McKinney et al., 2000). Studies relating smoking
behavior to other candidate genes in the dopamine pathway (do-
pamine D5 receptor, tyrosine hydroxylase) have reported negative
findings (Lerman et al., 1997; Sullivan et al., 2001).
Emerging research on genetic contributions to smoking behav-
ior has the potential to advance the science of smoking cessation
treatment by generating new knowledge about genetic factors that
influence clinical treatment outcome. This approach, termed phar-
macogenetics, is based on the premise that individual differences
in genotype influence drug metabolism and drug targets, thereby
influencing treatment outcome (Evans & Relling, 1999). The ul-
timate objective is to develop empirically based methods of tai-
loring the type and dose of treatment to an individual’s genotype
to improve efficacy and minimize side effects (Roses, 2000). The
potential for this approach to enhance the treatment of drug de-
pendence is supported by evidence that response to bromocryptine
treatment for alcohol dependence is predicted by DRD2 genotype
(Lawford et al., 1995).
Bupropion is an antidepressant medication with proven efficacy
for smoking cessation (Holm & Spencer, 2000). Six-month absti-
nence rates for smokers treated with bupropion are reported to be
25%–35%, as compared with 15%–20% for placebo (Hurt et al.,
1997; Jorenby et al., 1999). Whereas male gender and lower
smoking rate have been shown to predict abstinence in bupropion
clinical trials (independent of treatment group), studies have yet to
identify pretreatment factors that predict response to bupropion
treatment (Dale et al., 2001; Hayford et al., 1999).
Research has begun to elucidate bupropion’s pharmacokinetic
and pharmacodynamic effects. With regard to the former, bupro-
pion is metabolized to its primary metabolite, hydroxybupropion,
by the cytochrome P450 2B6 (CYP2B6) enzyme (Faucette et al.,
2000). We recently reported that the CYP2B6 genotype is a sig-
nificant predictor of smoking relapse and that bupropion attenu-
ated this effect in women (Lerman, Shields, et al., 2002). CYP2B6
is also involved in nicotine metabolism, and the genotype effect in
this study appeared to be mediated by craving, rather than by
bupropion metabolism (Lerman, Shields, et al., 2002). With regard
to pharmacodynamic effects, available evidence suggests that bu-
propion is a weak inhibitor of dopamine and norepinephrine re-
uptake (Ascher et al., 1995; Sanchez & Hyttel, 1999). This would
result in greater availability of these neurotransmitters and positive
psychobiological effects. These proposed neurochemical effects of
bupropion are consistent with documented beneficial effects on
postcessation withdrawal and mood symptoms in a placebo-
controlled bupropion trial (Lerman, Roth, et al., 2002).
The present pharmacogenetic study investigated whether the
SLC6A3 and DRD2 genes predicted smoking abstinence and re-
sponse to treatment in a placebo-controlled trial of bupropion for
smoking cessation. On the basis of previous findings relating
smoking to genes involved in dopamine regulation (Lerman et al.,
1999; Sabol et al., 1999) and on evidence suggesting that bupro-
pion inhibits dopamine reuptake (Ascher et al., 1995), we made the
following two predictions. First, smokers with the SLC6A3-9 and
DRD2-A2 genotypes would have higher abstinence rates than other
groups. Second, the effects of bupropion on smoking outcomes
would be most pronounced for smokers with the SLC6A3-10 and
DRD2-A1 genotypes, because these individuals would derive the
greatest benefit from increases in dopaminergic activity.
Method
Participants
Participants included 193 men and 225 women of European Caucasian
ancestry who reported smoking 10 or more cigarettes a day. They were
recruited using flyers and newspaper advertisements for a free smoking
cessation research program from May 1999 to September 2001. Potential
participants were first screened by telephone for eligibility and again
during an in-person medical screening session. Exclusion criteria included
pregnancy, a history of psychiatric disorder (Diagnostic and Statistical
Manual of Mental Disorders, 4th ed.; DSM-IV; 1994), seizure disorder,
and current use of psychotropic medications. Analyses were limited to
Caucasians of European ancestry to reduce bias due to racial admixture
(N⫽418). Data from participants in this clinical trial were included in a
previous report of the psychological mechanisms of bupropion treatment
(Lerman, Roth, et al., 2002) and the effects of CYP2B6 genotype on
treatment outcome (Lerman, Shields, et al., 2002).
The sample included 418 participants of European Caucasian decent;
227 participants received bupropion and 191 received placebo. Of the
participants, 54% were women and 46% of the total group completed a
college education. The average age of participants was 44 ⫾11 years, and
the average smoking rate was 25 ⫾12 cigarettes per day.
Procedure
Participants were enrolled at Georgetown University in Washington,
DC, and the State University of New York at Buffalo in Buffalo, NY.
Participants at both sites were ascertained in the same manner and received
identical assessments and treatments, and all participants provided written
consent for genotyping and treatment. At an initial visit to the smoking
clinic, participants provided a 40-ml blood sample and completed a set of
standardized self-report questionnaires (see Measures section). All partic-
ipants received 10 weeks of either placebo (n⫽191) or bupropion (n⫽
227), as determined by randomization. Bupropion treatment was delivered
according to the standard therapeutic dose (150 mg/day for the first 3 days,
followed by 300 mg/day). Participants also received standardized behav-
ioral counseling, focusing on self-monitoring and behavioral-modification
approaches. Trained health educators delivered this counseling during
seven sessions over a 10-week period in a group format. Weekly supervi-
sion and videotaping of sessions were performed to maintain the integrity
of the structured intervention protocol. All participants were instructed to
542 LERMAN ET AL.
quit smoking on a target quit date (TQD) 2 weeks after initiating medica-
tion, which coincided with Session 3 of behavioral counseling.
Self-reported data on smoking status were recorded weekly, at the end of
treatment (EOT; 8 weeks post-TQD) and at a 6-month follow-up using a
validated timeline follow-back method (Brown, Burgess, Sales, & White-
ley, 1998). To biochemically verify self-report data, we performed saliva
cotinine testing for participants who reported abstinence at EOT and at
6-month follow-up using a gas–liquid chromatography method (Feyera-
bend & Russell, 1990). Weekly pill counts were also recorded using pill
bottles returned on a weekly basis by participants. Monetary incentives
were provided for study participation but not for successful quitting.
Measures
Smoking outcomes. Consistent with the recommendations of the Soci-
ety for Research on Nicotine and Tobacco (Hughes et al., 2003), we
considered prolonged abstinence to be the primary outcome and 7-day
point prevalence as the secondary outcome. Prolonged abstinence is de-
fined as sustained abstinence after a 2-week grace period, with 7 consec-
utive days of smoking constituting a treatment failure (Hughes et al., 2003).
Seven-day point prevalence reflects smoking status for the prior 7 days and
was biochemically verified on the basis of cotinine levels. Only those
participants with cotinine levels less than or equal to 15 ng/ml were
classified as abstinent for the point-prevalence outcome. Participants lost to
follow-up (5% at EOT, 7% at 6-month follow-up) were classified as
smokers for all outcome analyses.
Genotype. Consistent with previous reports (Lerman et al., 1999; Sabol
et al., 1999), the SLC6A3 genotype was classified as the presence or
absence of the 9-repeat allele. Because the most common alleles are the
9-repeat and 10-repeat alleles and other variants are very rare (⬍3%), we
refer to these groups as 9/9 or 9/10 versus 10/10. The DRD2 genotype was
classified as the presence or absence of the A1 allele (A1/A1 or A1/A2 vs.
A2/A2), consistent with previous reports (Comings et al., 1996; Lerman et
al., 1999). The assays were validated by confirming a polymorphic inher-
itance pattern in seven human family lines that encompassed three gener-
ations (National Institute of General Medical Studies, n.d.). Quality control
procedures included positive and negative controls with each assay and
independent repeat genotyping for 20% of the results.
Demographic factors. Gender, education, marital status, age, and eth-
nic ancestry of grandparents were assessed by self-report during the pre-
treatment assessment visit.
Nicotine dependence. The Fagerstrom Test for Nicotine Dependence
(FTND) is a 6-item self-report measure of nicotine dependence derived
from the Fagerstrom Tolerance Questionnaire (Heatherton, Kozlowski,
Frecker, & Fagerstrom, 1991). The FTND was administered during the
pretreatment assessment visit.
Statistical Considerations
An association (case-control) study design was used to evaluate geno-
type effects on treatment outcome. This approach was chosen over a
family-based analysis (Spielman & Ewens, 1996) both because of the
limited feasibility of recruiting family members of adults in a clinical trial,
and because the case-control design is more powerful. Intent-to-treat meth-
ods were used to evaluate the trial outcomes. Chi-square tests of associa-
tion and ttests were used to identify factors associated with the smoking
outcomes. All statistical tests were two-sided. Variables with significant
(p⬍.10) associations with smoking status were entered into logistic
regression models to identify predictors of abstinence at EOT and at
6-month follow-up. Main effects of genotype and covariates were tested in
a first step, and interaction terms were included in a second step. All
models controlled for study site. Cox regression was used to examine the
main and interacting effects of genotype on time to relapse (the first day of
a prolonged abstinence failure). The time to event for participants lost to
follow-up was the last date of contact.
Results
Descriptive Data
Participants were classified according to variation in the
SLC6A3 and DRD2 genes. The frequencies for the different
SLC6A3 genotype groups are as follows: 210 participants (50.2%)
had the 10/10 genotype, 162 (38.8%) had the 9/10 genotype, 34
(8.1%) had the 9/9 genotype, and 12 (2.9%) had genotypes con-
taining other allelic combinations (10/11, 10/6, 10/8, 9/11, 9/6, or
9/3). Thus, 200 (47.9%) participants were classified as SLC6A3-9
(9/9, 9/10 or 9/*, in which * refers to alleles other than 9 or 10) and
218 (52.1%) as SLC6A3-10 (10/10 or 10/*). For the DRD2 gene,
genotype frequencies were as follows: 28 participants (6.7%) had
A1/A1 genotypes, 152 (36.4%) had A1/A2 genotypes, and 238
(56.9%) had A2/A2 genotypes. Thus, 180 (43.1%) participants
were classified as DRD2-A1 (i.e., A1/A1 or A1/A2), and 238
(56.9%) were classified as DRD2-A2 (i.e., A2/A2). Genotypes for
both genes were in Hardy–Weinberg equilibrium, and there were
no differences by study site. There were no significant differences
between the bupropion and placebo groups by genotype or the
controlling variables, or between participants retained and those
lost to follow-up (ps⬎.10). Pill counts indicated that, on average,
participants consumed four pills less than the total recommended
number of pills for the entire course of treatment. Pill counts did
not differ significantly by treatment or genotype (ps⬎.10).
Bivariate Analysis of the Effects of Genotype and
Treatment on Smoking Abstinence
Predictors of prolonged abstinence are shown in Table 1. There
was a significant effect of treatment condition at EOT,
2
(1, N⫽
418) ⫽9.69, p⬍.01, and at 6-month follow-up,
2
(1, N⫽
418) ⫽4.40, p⫽.04. The main effects of SLC6A3 and DRD2
genotypes were not significant at either timepoint (ps⬎.10).
Additional factors associated significantly with prolonged absti-
nence included nicotine dependence and gender. The results of
univariate analyses of point-prevalence abstinence were similar;
treatment condition was significant at the end of treatment (p⬍
.01) and at 6-month follow-up (p⫽.03), but the individual
genotype variables were not.
Figure 1 shows the combined effects of the two genotypes on
prolonged abstinence rates. A significant effect of SLC6A3 on
abstinence at EOT was found among participants with A2 geno-
types (odds ratio [OR] ⫽1.74; confidence interval [CI]
⫽1.03, 2.93; p⫽.03), but not among those with A1 genotypes
(OR ⫽0.67, CI ⫽0.37, 1.21; p⫽.18). The Mantel–Haenszel test
for homogeneity of ORs indicated that these values are signifi-
cantly different,
2
(1, N⫽418) ⫽5.74, p⫽.02. A similar trend
for an effect of SLC6A3 on abstinence in the A2 genotype group
was observed for point-prevalence abstinence at EOT, but the
results were not significant ( p⫽.11). No significant effects for the
two genotypes combined were found for either abstinence measure
at 6-month follow-up.
Figure 2 shows the effect of treatment on prolonged abstinence
by SLC6A3 and DRD2 genotypes. These descriptive data suggest
that the largest effect of treatment occurs in the DRD2-A2/
SLC6A3-10 group (OR ⫽2.80; CI ⫽1.30, 6.00; p⬍.01), because
of a lower abstinence rate on placebo. By contrast, the A1/10
543
SMOKING CESSATION IN A BUPROPION CLINICAL TRIAL
group has the smallest treatment effect (OR ⫽0.87;
CI ⫽0.37, 2.00; p⫽.74), because of a high abstinence rate on
placebo. The treatment effects in the A1/9 (OR ⫽2.23;
CI ⫽0.92, 5.41; p⫽.07) and A2/9 (OR ⫽1.91; CI ⫽0.88, 4.14;
p⫽.10) groups were intermediate. The Mantel–Haenszel statistic
indicated that these values are not significantly different,
2
(3,
N⫽418) ⫽4.53, p⫽.21.
Multivariate Models of Abstinence
Variables with significant (p⬍.10) associations with absti-
nence were entered into logistic regression models to test the main
and interacting effects of genotypes and treatment. The best fitting
model for prolonged abstinence at the end of treatment is shown in
Table 2. The final model included significant effects for treatment
condition, gender, nicotine dependence level, DRD2 genotype, and
the DRD2 ⫻SLC6A3 interaction. Participants treated with bupro-
pion, women, and those with lower baseline levels of nicotine
dependence were more likely to be abstinent at the end of treat-
ment. The gene–gene interaction (see Figure 1) indicates that,
among participants with DRD2-A2 genotypes, SLC6A3-9 repeat
genotype is associated with a higher abstinence rate (p⫽.03).
Among participants with DRD2-A1 genotypes, the effect of
SLC6A3 is not significant (p⫽.18). The two-way and three-way
interactions for Treatment ⫻Genotype were not significant in the
logistic regression model.
Similar modeling approaches were used to examine prolonged
abstinence at 6-month follow-up. In this model, only treatment
(p⫽.05) and gender (p⬍.01) were significant predictors. To
Table 1
Predictors of Prolonged Abstinence
Variable and level
End of treatment 6-month follow-up
% abstinent p% abstinent p
Treatment
Placebo 39.8 22.5
Bupropion 55.1 ⬍.01 31.7 .04
SLC6A3
9/9, 9/10 50.0 29.0
10/10 46.3 .45 26.1 .51
DRD2
A1 51.7 26.1
A2 45.9 .20 28.6 .58
Gender
Female 41.8 22.2
Male 55.4 .01 33.7 ⬍.01
Site
Washington, DC 49.4 24.7
Buffalo, NY 46.1 .51 31.7 .11
Variable and
abstinence status
End of treatment 6-month follow-up
MSD p MSDp
Nicotine dependence
Abstinent 4.8 2.1 4.8 2.2
Nonabstinent 5.4 2.1 ⬍.01 5.2 2.1 .06
Age
Abstinent 44.1 10.9 44.5 10.6
Nonabstinent 44.6 11.6 .61 44.3 11.5 .90
Figure 1. Interacting effects of SLC6A3 and DRD2 genotypes on pro-
longed abstinence at the end of treatment.
Figure 2. Effects of bupropion treatment on prolonged abstinence at the
end of treatment, by SLC6A3 and DRD2 genotypes.
Table 2
Logistic Regression Model of Prolonged Abstinence
at the End of Treatment
Variable OR CI p
Drug 1.87 1.24, 2.82 ⬍.01
Site 0.67 0.43, 1.03 .07
Gender 0.51 0.34, 0.78 ⬍.01
Nicotine dependence 0.83 0.75, 0.92 .00
SLC6A3 0.67 0.36, 1.20 .20
DRD2 0.54 0.30, 0.96 .04
SLCA3 ⫻DRD2 2.40 1.07, 5.48 .03
Note. OR ⫽odds ratio; CI ⫽confidence interval.
544 LERMAN ET AL.
determine whether the effects of genotype were significantly dif-
ferent at EOT and 6-month follow-up, we performed logistic
regression using general estimating equations. There was a signif-
icant effect of time point (p⬍.01) and a significant Time ⫻
DRD2 ⫻SLC6A3 interaction, indicating that the genotype effect
was significantly different at the two time points.
We also examined the effects of treatment and genotype on our
secondary outcome, point-prevalence abstinence. At EOT, only
treatment (p⫽.01), gender (p⫽.03), and nicotine dependence
(p⫽.03) were significant predictors of abstinence. At 6-month
follow-up, the only significant predictor was treatment (p⫽.03).
Survival Analysis
Cox regression was used to model the time to prolonged absti-
nence failure by EOT as a function of individual genotype effects
and Genotype ⫻Treatment interactions, controlling for study site,
gender, and nicotine dependence. Significant effects on days to
relapse were found for the following variables: treatment (hazard
ratio [HR] ⫽0.64; CI ⫽0.46, 0.80; p⬍.01); site (HR ⫽1.36;
CI ⫽1.01, 1.80; p⫽.04), gender (HR ⫽1.46; CI ⫽1.10, 1.92;
p⬍.01), nicotine dependence (HR ⫽1.14; CI ⫽1.07, 1.22; p⬍
.01), DRD2 (HR ⫽1.50; CI ⫽1.02, 2.23; p⫽.04), and the
DRD2 ⫻SLC6A3 interaction (HR ⫽0.50; CI ⫽0.29, 0.87; p⫽
.01). Among participants with DRD2-A2 genotypes, the average
number of days to relapse during the treatment phase for the
SLC6A3-9 group was 28.2, compared with 21.2 for the SLC6A3-10
group, t(416) ⫽2.56, p⫽.01. The pattern of results was similar
for days to relapse by 6-month follow-up; however, the Cox
regression results were not statistically significant.
Examination of Effects of Ethnic Ancestry
We directly explored the potential impact of ethnic admixture in
our European Caucasian sample on bias in the ORs for SLC6A3
and DRD2, by adjusting for self-reported ethnic heritage (beyond
White, non-Hispanic). Heritage was clustered into categories for
Northern European (e.g., English, German), Southern European
(e.g., Spanish, Italian), and Slavic or Eastern European. Partici-
pants were excluded if no one category was predominant (i.e., at
least 3 of 4 grandparents). In this reduced sample (n⫽283), the
Mantel–Haenszel statistic showed that there was no evidence of
heterogeneity: for SLC6A3,
2
(2, N⫽283) ⫽0.32, p⫽.85; and
for DRD2,
2
(2, N⫽283) ⫽2.00, p⫽.36. Thus, it appears that
in a European Caucasian sample, any bias due to ethnic ancestry
would be minimal.
Discussion
We investigated the effects of two dopaminergic genes, SLC6A3
and DRD2, on smoking abstinence and response to treatment in a
placebo-controlled bupropion clinical trial. The results revealed a
significant gene–gene interaction effect on prolonged abstinence,
independent of treatment condition. Among smokers with
DRD2-A2 genotypes, those with SLC6A3-9 genotypes (compared
with SLC6A3-10 genotypes) had significantly higher abstinence
rates at the end of treatment (53% vs. 39%) and a longer latency
to relapse at EOT (28 vs. 21 days) and at 6-month follow-up (83
vs. 65 days). These effect sizes are comparable with those
achieved in placebo-controlled pharmacotherapy trials for smok-
ing cessation (Silagy, Mant, Fowler, & Lodge, 1994). By contrast,
among smokers with DRD2-A1 genotypes, the effect of SLC6A3
on abstinence rates and time to relapse was not significant.
These results provide new evidence for dopaminergic genetic
effects on prospective smoking cessation and extend previous
reports of associations of smoking status with SLC6A3 and DRD2
(Comings et al., 1996; Lerman et al., 1999; Sabol et al., 1999;
Spitz et al., 1998). The presence of a gene–gene interaction,
mirroring results from a previous study of smoking status (Lerman
et al., 1999), underscores the importance of not limiting genetic
investigations of smoking behavior to single-gene effects (Lerman
& Swan, 2002). In fact, the lack of attention to multigenic effects
on a complex trait such as smoking may explain, in part, the failure
of some studies to replicate genetic effects on smoking behavior
(Bierut et al., 2000; Jorm et al., 2000; Vandenbergh et al., 2002).
On the basis of existing biological and epidemiological data on
DRD2 and SLC6A3, one can postulate a biobehavioral mechanism
for the observed findings. Available neurobiological correlative
data suggest that individuals with DRD2-A1 genotypes exhibit
altered receptor density and binding characteristics and therefore
may have lower levels of neuronal dopaminergic activity com-
pared with individuals with DRD2-A2 genotypes (Noble, Blum,
Ritchie, Montgomery, & Sheridan, 1991; Ritchie & Noble, 1996;
Thompson et al., 1997). This is consistent with epidemiological
evidence linking DRD2-A1 with altered neurocognitive function
(Anokhin, Todorov, Madden, Grant, & Heath, 1999; Blum,
Braverman, Dinardo, Wood, & Sheridan, 1994) as well as with a
variety of addictive behaviors (Noble, 2000). The 9-repeat
SLC6A3 genotype has been associated with lower levels of dopa-
mine transporter protein expression (Fuke et al., 2001) and lower
brain protein levels (Heinz et al., 2000). This would be expected to
result in less neuronal dopamine reuptake and higher levels of
dopamine, possibly resulting in less reinforcement from dopamine-
stimulating drugs such as nicotine (Heinz et al., 2000). This
assertion is consistent with evidence relating SLC6A3-9 genotypes
with cocaine-induced paranoia, a state attributed to excess dopa-
mine levels (Gelernter, Kranzler, Satel, & Rao, 1994), and with
smoking cessation history in a cross-sectional study (Sabol et al.,
1999). Thus, one could speculate that, in the presence of normal
receptor function (i.e., DRD2-A2), lower dopamine transporter
levels and higher levels of dopamine (i.e., SLC6A3-9) would
minimize the phasic effects of nicotine on dopamine release,
thereby reducing positive reinforcement from smoking. This, in
turn, would make it easier for smokers with DRD2-A2/SLC6A3-9
genotypes to maintain smoking abstinence. By contrast, low do-
pamine levels (i.e., SLC6A3-10) and normal receptor density (i.e.,
DRD2-A2) might produce the greatest need for and reinforcement
from nicotine. To provide support for this hypothesis, we con-
ducted a post hoc analysis of levels of nicotine dependence by the
DRD2 and SLC6A3 genotypes. Consistent with the findings for
smoking abstinence, there was no effect of SLC6A3 in the
DRD2-A1 group and a trend for an effect in the DRD2-A2 group
(FTND scores of 5.0 ⫾2.1 vs. 5.3 ⫾2.2 for the SLC6A3-9 vs.
SLC6A3-10 groups, respectively; p⫽.12).
This interpretation, although speculative, may also explain why
the DRD2 ⫻SLC6A3 interaction was significant for the prolonged
abstinence measure but weaker for the point-prevalence measure at
the end of treatment. The prolonged abstinence measure allows for
545
SMOKING CESSATION IN A BUPROPION CLINICAL TRIAL
slips, provided the participant does not engage in 7 days of
consecutive smoking (Hughes et al., 2003). By contrast, the point-
prevalence measure requires that the participant does not smoke
even a puff of a single cigarette during the 7 days prior to
assessment. Smokers who carry the putative protective genotypes
may be able to smoke intermittently without experiencing a full-
blown relapse. By contrast, other smokers who experience more
reinforcement following a slip may be more likely to return to
smoking. If this is true, then the DRD2 ⫻SLC6A3 effect could be
stronger for the outcome that allows for intermittent slips than for
the outcome that requires continuous abstinence. The weaker ef-
fect for the point-prevalence outcome alternatively may relate to
the requirement for biochemical verification. A recent analysis
from a large number of smoking trials indicated that it is not
uncommon for abstinence rates to differ when biochemical veri-
fication is used (Society for Research on Nicotine & Tobacco
Subcommittee on Biochemical Verification, 2002). In the present
study, the rate of misreporting of abstinence was 20%, consistent
with previous reports (Gariti, Alterman, Ehrman, Mulvaney, &
O’Brien, 2002). However, because the rates of misreporting in this
study did not differ by genotype or treatment group, inaccurate
self-reporting is unlikely to bias the study results (Society for
Research on Nicotine & Tobacco Subcommittee on Biochemical
Verification, 2002).
Although the present findings supported our hypothesis about
genetic effects on smoking abstinence at the end of treatment, the
results were negative for genetic effects at 6-month follow-up. In
fact, at 6-month follow-up, abstinence rates were substantially
lower for all groups, and the treatment effect was also substantially
weakened. Previous research has documented the key role of
environmental and psychological factors in smoking relapse, par-
ticularly after active treatment has ended (Brandon, Zelman, &
Baker, 1987; Hall, Munoz, Reus, & Sees, 1993; Mermelstein,
Cohen, Lichtenstein, Baer, & Kamarck, 1986). Thus, although
genotype may play an important role in maintaining abstinence
while smokers are receiving behavioral counseling or pharmaco-
therapy, environmental factors may be relatively more important
after active treatment has ended. The role of psychological and
environmental factors in promoting relapse may also depend, in
part, on genetic predisposition. Thus, future smoking outcome
studies should incorporate data on these influences, making it
possible to test hypotheses about gene–environment interactions in
smoking cessation.
Contrary to predictions, we did not find evidence that DRD2 and
SLC6A3 significantly modify response to bupropion treatment.
Inspection of the descriptive data (see Figure 2) shows a relatively
equivalent response to bupropion in all four genotype groups, but
a differential response to placebo. To be specific, the DRD2-A2/
SLC6A3-10 group had the lowest quit rate on placebo (27%),
resulting in a highly significant 24% difference in abstinence rates
for the treatment groups. This is consistent with our hypothesis that
less endogenous dopamine (SLC6A3-10) in the presence of normal
receptor density (DRD2-A2) would produce the greatest need for
reinforcement from nicotine. Although bupropion appears to com-
pensate for a possible genetic predisposition to relapse in this
group, our sample size of 418 may not have been sufficient to
detect a Treatment ⫻Gene ⫻Gene interaction effect.
Dopaminergic processes are but one mechanism postulated for
the effect of bupropion on smoking cessation (Sanchez & Hyttel,
1999). Additional data suggest that bupropion influences serotonin
and norepinephrine activity (Cooper et al., 1994; Dong & Blier,
2001) and that it acts as a neuronal nicotinic receptor antagonist
(i.e., reduces nicotine binding in the brain; Slemmer, Martin, &
Damaj, 2000). Given the lack of specificity of the neuropharma-
cologic effects of this medication, any individual genetic effect on
treatment outcome would be expected to be quite small. Thus, the
greatest promise of pharmacogenetic research on smoking cessa-
tion may be realized in larger collaborative studies that examine
multiple genetic polymorphisms across different neurotransmitter
and receptor pathways.
As an investigation in the emerging area of behavioral genetics
and smoking, the present study has both strengths and weaknesses.
The strengths include the collection of DNA samples from all
study participants, thereby limiting self-selection bias as well as
the use of more refined longitudinal smoking phenotypes. One
limitation of the study is that we did not collect posttreatment
blood samples to determine levels of bupropion and to biochem-
ically verify medication compliance; however, genotype and treat-
ment effects on pill count were not found, suggesting that medi-
cation compliance is unlikely to exert a substantial bias. An
additional issue to consider is the use of an intent-to-treat ap-
proach, in which participants lost to follow-up were assumed to be
smokers. We selected this approach to intervention evaluation
because of concerns about reduced power and, more important, the
potential for biasing the study outcomes by excluding noncompli-
ers (Hall et al., 2001). Because only 5% of participants were lost
to follow-up, any bias due to our choice of analytic design is likely
to be minimal. In fact, when the final model was rerun excluding
participants lost to follow-up, the results were the same, and the
DRD2 ⫻SLC6A3 interaction remained significant.
Another potential limitation of this study is that the results may
be biased by ethnic admixture in the study population. In other
words, ethnic heterogeneity in the European Caucasian population
may be associated with both allele frequencies and smoking prac-
tices. Although population stratification can be eliminated in
family-based genetic analyses (Spielman & Ewens, 1996), it is not
very feasible to accrue family members of adults participating in a
randomized trial. Moreover, the significance of population strati-
fication remains controversial. Wacholder, Rothman, and Capo-
raso (2000) used the confounding risk ratio (the ratio of crude and
adjusted relative risks) as a measure of bias in the estimation of
relative risk and determined by simulation that ignoring genetic
admixtures of non-Hispanic European descent would produce a
bias of less than 10%. Millikan (2001) found a similar result using
confounding ORs in a study of genetic markers for breast cancer
and extended the results to include African Americans in the
admixture. We have addressed the potential impact of admixture in
our sample by comparing the crude ORs for genotype effects with
that adjusted for ethnic ancestry. These estimates differed by less
than 2%, indicating that accounting for ethnic ancestry within our
European Caucasian sample is not important.
Despite these potential limitations, the present study provides a
first step toward identifying genetic determinants of smoking
relapse in the context of smoking treatment. Although our inter-
pretations of the data remain speculative until the biobehavioral
mechanisms of genetic influences can be empirically tested, these
results provide directions for future studies. Such studies could
include controlled laboratory investigations to elucidate genetic
546 LERMAN ET AL.
influences on the pharmacodynamic effects of bupropion as well
as larger scale collaborative clinical trials that are amply powered
to detect complex multigenetic and treatment interactions and to
adjust for ethnic ancestry. Although the collection of genetic data
in intervention trials does present a number of unique challenges
(e.g., increased cost, expanded discussion of confidentiality, and
potential risks in the informed consent process), this line of re-
search could enhance our knowledge of nicotine addiction and lead
to interventions that are tailored to individual smokers’needs.
References
American Psychiatric Association. (1994). Diagnostic and statistical man-
ual of mental disorders (4th ed.). Washington, DC: Author.
Anokhin, A. P., Todorov, A. A., Madden, P. A., Grant, J. D., & Heath,
A. C. (1999). Brain event-related potentials, dopamine D2 receptor gene
polymorphism, and smoking. Genetic Epidemiology, 17(Suppl. 1), S37–
S42.
Ascher, J. A., Cole, J. O., Colin, J. N., Feighner, J. P., Ferris, R. M.,
Fibiger, H. C., et al. (1995). Bupropion: A review of its mechanism of
antidepressant activity. Journal of Clinical Psychiatry, 56, 395–401.
Bierut, L., Rice, J., Edenberg, H., Goate, A., Foroud, T., & Cloninger, C.
(2000). Family-based study of the association of the dopamine D2
receptor gene (DRD2) with habitual smoking. American Journal of
Medical Genetics, 90, 299–302.
Blum, K., Braverman, E. R., Dinardo, M. J., Wood, R. C., & Sheridan, P. J.
(1994). Prolonged P300 latency in a neuropsychiatric population with
the D2 dopamine receptor A1 allele. Pharmacogenetics, 4, 313–322.
Brandon, T. H., Zelman, D. C., & Baker, T. B. (1987). Effects of main-
tenance sessions on smoking relapse: Delaying the inevitable? Journal
of Consulting and Clinical Psychology, 55, 780–782.
Brown, R., Burgess, E., Sales, S., & Whiteley, J. (1998). Reliability and
validity of a smoking timeline follow-back interview. Psychology of
Addictive Behaviors, 12, 101–112.
Comings, D., Ferry, L., Bradshaw-Robinson, S., Burchette, R., Chiu, C., &
Muhleman, D. (1996). The dopamine D2 receptor (DRD2) gene: A
genetic risk factor in smoking. Pharmacogenetics, 6, 73–79.
Cooper, B. R., Wang, C. M., Cox, R. F., Norton, R., Shea, V., & Ferris,
R. M. (1994). Evidence that the acute behavioral and electrophysiolog-
ical effects of bupropion (Wellbutrin) are mediated by a noradrenergic
mechanism. Neuropsychopharmacology, 11, 133–141.
Dale, L. C., Glover, E. D., Sachs, D. P., Schroeder, D. R., Offord, K. P.,
Croghan, I. T., & Hurt, R. D. (2001). Bupropion for smoking cessation:
Predictors of successful outcome. Chest, 119, 1357–1364.
Dong, J., & Blier, P. (2001, April 1). Modification of norepinephrine and
serotonin, but not dopamine, neuron firing by sustained bupropion
treatment. Psychopharmacology (Berl), 155, 52–57.
Evans, W. E., & Relling, M. V. (1999, October 15). Pharmacogenomics:
Translating functional genomics into rational therapeutics. Science, 286,
487–491.
Faucette, S. R., Hawke, R. L., Lecluyse, E. L., Shord, S. S., Yan, B.,
Laethem, R. M., et al. (2000). Validation of bupropion hydroxylation as
a selective marker of human cytochrome P450 2B6 catalytic activity.
Drug Metabolism and Disposition, 28, 1222–1230.
Feyerabend, C., & Russell, M. A. (1990). A rapid gas-liquid chromato-
graphic method for the determination of cotinine and nicotine in bio-
logical fluids. Journal of Pharmacy and Pharmacology, 42, 450–452.
Fuke, S., Suo, S., Takahashi, N., Koike, H., Sasagawa, N., & Ishiura, S.
(2001). The VNTR polymorphism of the human dopamine transporter
(DAT1) gene affects gene expression. Pharmacogenomics Journal, 1,
152–156.
Gariti, P., Alterman, A. I., Ehrman, R., Mulvaney, F. D., & O’Brien, C. P.
(2002). Detecting smoking following smoking cessation treatment. Drug
and Alcohol Dependence, 65, 191–196.
Garris, P. A., Kilpatrick, M., Bunin, M. A., Michael, D., Walker, Q. D., &
Wightman, R. M. (1999, March 4). Dissociation of dopamine release in
the nucleus accumbens from intracranial self-stimulation. Nature, 398,
67–69.
Gelernter, J., Kranzler, H. R., Satel, S. L., & Rao, P. A. (1994). Genetic
association between dopamine transporter protein alleles and cocaine-
induced paranoia. Neuropsychopharmacology, 11, 195–200.
Hall, S. M., Delucchi, K., Velicer, W., Kahler, C., Ranger-Moore, J.,
Hedeker, D., et al. (2001). Statistical analysis of randomized trials in
tobacco treatment: Longitudinal designs with dichotomous outcome.
Nicotine & Tobacco Research, 3, 193–202.
Hall, S. M., Munoz, R. F., Reus, V. L., & Sees, K. L. (1993). Nicotine,
negative affect, and depression. Journal of Consulting and Clinical
Psychology, 61, 761–767.
Hayford, K. E., Patten, C. A., Rummans, T. A., Schroeder, D. R., Offord,
K. P., Croghan, I. T., et al. (1999). Efficacy of bupropion for smoking
cessation in smokers with a former history of major depression or
alcoholism. British Journal of Clinical Pharmacology, 174, 173–178.
Heatherton, T., Kozlowski, L., Frecker, R., & Fagerstrom, K. (1991). The
Fagerstrom Test for Nicotine Dependence: A revision of the Fagerstrom
Tolerance Questionnaire. British Journal of Addiction, 86, 1119–1127.
Heinz, A., Goldman, D., Jones, D. W., Palmour, R., Hommer, D., Gorey,
J. G., et al. (2000). Genotype influences in vivo dopamine transporter
availability in human striatum. Neuropsychopharmacology, 22, 133–
139.
Henningfield, J. (1995). Nicotine medications for smoking cessation. New
England Journal of Medicine, 333, 1196–1203.
Holm, K. J., & Spencer, C. M. (2000). Bupropion: A review of its use in
the management of smoking cessation. Drugs, 59, 1007–1024.
Hughes, J. R., Keely, J. P., Niaura, R. S., Ossip-Klein, D. J., Richmond,
R. L., & Swan, G. E. (2003). Measures of abstinence in clinical trials:
Issues and recommendations. Nicotine & Tobacco Research, 5, 13–25.
Hurt, R., Sachs, D., Glover, E., Offord, K., Johnston, J., Dale, L., et al.
(1997). A comparison of sustained-release bupropion and placebo for
smoking cessation. New England Journal of Medicine, 337, 1195–1202.
Jorenby, D. E., Leischow, S. J., Nides, M. A., Rennard, S. I., Johnston,
J. A., Hughes, A. R., et al. (1999). A controlled trial of sustained-release
bupropion, a nicotine patch, or both for smoking cessation. New England
Journal of Medicine, 340, 685–691.
Jorm, A., Henderson, A., Jacomb, P., Christensen, H., Korten, A., Rodgers,
B., et al. (2000). Association of smoking and personality with a poly-
morphism of the dopamine transporter gene: Results from a community
survey. American Journal of Medical Genetics, 96, 331–334.
Lawford, B. R., Young, R. M., Rowell, J. A., Qualichefski, J., Fletcher,
B. H., Syndulko, K., et al. (1995). Bromocriptine in the treatment of
alcoholics with the D2 dopamine receptor A1 allele. Nature Medicine, 1,
337–341.
Lerman, C., Audrain, J., Main, D., Boyd, N., Caporaso, N., Bowman, E.,
et al. (1999). Evidence suggesting the role of specific genetic factors in
cigarette smoking. Health Psychology, 18, 14–20.
Lerman, C., Roth, D., Kaufmann, V., Audrain, J., Hawk, L., Liu, A., et al.
(2002). Mediating mechanisms for the impact of bupropion in smoking
cessation treatment. Drug and Alcohol Dependence, 67, 219–223.
Lerman, C., Shields, P. G., Main, D., Audrain, J., Roth, J., Boyd, N. R., et
al. (1997). Lack of association of tyrosine hydroxylase genetic polymor-
phism with cigarette smoking. Pharmacogenetics, 7, 521–524.
Lerman, C., Shields, P. G., Wileyto, E., Audrain, J., Pinto, A., Hawk, L.,
et al. (2002). Pharmacogenetic investigation of smoking cessation treat-
ment. Pharmacogenetics, 12, 627–634.
Lerman, C., & Swan, G. (2002). Non-replication of genetic association
studies: Is DAT all folks? Nicotine & Tobacco Research, 4, 247–249.
McKinney, E., Walton, R., Yudkin, P., Fuller, A., Haldar, N., Mant, D., et
al. (2000). Association between polymorphisms in dopamine metabolic
547
SMOKING CESSATION IN A BUPROPION CLINICAL TRIAL
enzymes and tobacco consumption in smokers. Pharmacogenetics, 10,
483–491.
Mermelstein, R., Cohen, S., Lichtenstein, E., Baer, J. S., & Kamarck, T.
(1986). Social support and smoking cessation and maintenance. Journal
of Consulting and Clinical Psychology, 54, 447–453.
Millikan, R. C. (2001). Re: Population stratification in epidemiologic
studies of common genetic variants and cancer: Quantification of bias.
Journal of the National Cancer Institute, 93, 156–158.
National Institute of General Medical Sciences. (n.d.). Human genetic
mutant cell repository. Camden, NJ: Coriell Institute. Retrieved from
locus.umdnj.edu/nigms/overview.html
Noble, E. P. (2000). The DRD2 gene in psychiatric and neurological
disorders and its phenotypes. Pharmacogenomics, 1, 309–333.
Noble, E. P., Blum, K., Ritchie, T., Montgomery, A., & Sheridan, P. J.
(1991). Allelic association of the D2 dopamine receptor gene with
receptor-binding characteristics in alcoholism. Archives of General Psy-
chiatry, 48, 648–654.
Noble, E. P., St. Jeor, S. T., Ritchie, T., Syndulko, K., St. Jeor, S. C., Fitch,
R., et al. (1994). D2 dopamine receptor gene and cigarette smoking: A
reward gene? Medical Hypotheses, 42, 257–260.
Pontieri, F., Tanda, G., Orzi, F., & Di Chiara, G. (1996, July 18). Effects
of nicotine on the nucleus accumbens and similarity to those of addictive
drugs. Nature, 382, 255–257.
Ritchie, T., & Noble, E. P. (1996, July 18). [
3
H]naloxone binding in the
human brain: Alcoholism and the TaqI A D2 dopamine receptor poly-
morphism. Brain Research, 718, 193–197.
Roses, A. D. (2000, June 15). Pharmacogenetics and the practice of
medicine. Nature, 405, 857–865.
Sabol, S. Z., Nelson, M. L., Fisher, C., Gunzerath, L., Brody, C. L., Hu, S.,
et al. (1999). A genetic association for cigarette smoking behavior.
Health Psychology, 18, 7–13.
Sanchez, C., & Hyttel, J. (1999). Comparison of the effects of antidepres-
sants and their metabolites on reuptake of biogenic amines and on
receptor binding. Cell and Molecular Neurobiology, 19, 467–489.
Schultz, W. (1998). Predictive reward signal of dopamine neurons. Journal
of Neurophysiology, 80, 1–27.
Silagy, C., Mant, D., Fowler, G., & Lodge, M. (1994). Meta-analysis on
efficacy of nicotine replacement therapies in smoking cessation. Lancet,
343, 139–142.
Slemmer, J. E., Martin, B. R., & Damaj, M. I. (2000). Bupropion is a
nicotinic antagonist. Journal of Pharmacology and Experimental Ther-
apeutics, 295, 321–327.
Society for Research on Nicotine & Tobacco Subcommittee on Biochem-
ical Verification. (2002). Biochemical verification of tobacco use and
cessation. Nicotine & Tobacco Research, 4, 149–159.
Spielman, R., & Ewens, W. (1996). The TDT and other family-based tests
for linkage disequilibrium and association. American Journal of Human
Genetics, 59, 983–989.
Spitz, M., Shi, H., Yang, F., Hudmon, K., Jiang, H., & Chanberlain, R.
(1998). Case-control study of the D2 dopamine receptor gene and
smoking status in lung cancer patients. Journal of the National Cancer
Institute, 90, 358–363.
Sullivan, P. F., Neale, M. C., Silverman, M. A., Harris-Kerr, C., Myaki-
shev, M. V., Wormley, B., et al. (2001). An association study of DRD5
with smoking initiation and progression to nicotine dependence. Amer-
ican Journal of Medical Genetics, 105, 259–265.
Thompson, J., Thomas, N., Singleton, A., Piggott, M., Lloyd, S., Perry,
E. K., et al. (1997). D2 dopamine receptor gene (DRD2) Taq1 A
polymorphism: Reduced dopamine D2 receptor binding in the human
striatum associated with the A1 allele. Pharmacogenetics, 7, 479–484.
Vandenbergh, D., Bennett, C., Grant, M., Strasser, A., O’Connor, R.,
Stauffer, R., et al. (2002). Smoking status and the human dopamine
transporter variable number of tandem repeats (VNTR) polymorphism:
Failure to replicate and finding that never-smokers may be different.
Nicotine & Tobacco Research, 4, 333–340.
Wacholder, S., Rothman, N., & Caporaso, N. (2000). Population stratifi-
cation in epidemiologic studies of common genetic variants and cancer:
Quantification of bias. Journal of the National Cancer Institute, 92,
1151–1158.
548 LERMAN ET AL.
A preview of this full-text is provided by American Psychological Association.
Content available from Health Psychology
This content is subject to copyright. Terms and conditions apply.
