Research

Original Investigation

Effect of Varenicline on Smoking Cessation

Through Smoking Reduction
A Randomized Clinical Trial
Jon O. Ebbert, MD, MSc; John R. Hughes, MD; Robert J. West, PhD; Stephen I. Rennard, MD; Cristina Russ, MD;
Thomas D. McRae, MD; Joan Treadow, RN, BSN; Ching-Ray Yu, PhD; Michael P. Dutro, PharmD; Peter W. Park, PhD

IMPORTANCE Some cigarette smokers may not be ready to quit immediately but may be

willing to reduce cigarette consumption with the goal of quitting.

OBJECTIVE To determine the efficacy and safety of varenicline for increasing smoking

abstinence rates through smoking reduction.

DESIGN, SETTING, AND PARTICIPANTS Randomized, double-blind, placebo-controlled,
multinational clinical trial with a 24-week treatment period and 28-week follow-up conducted
between July 2011 and July 2013 at 61 centers in 10 countries. The 1510 participants were
cigarette smokers who were not willing or able to quit smoking within the next month but
willing to reduce smoking and make a quit attempt within the next 3 months. Participants
were recruited through advertising.

JAMA Report Video and

Author Video Interview at
jama.com
Supplemental content at
jama.com
CME Quiz at
jamanetworkcme.com and
CME Questions page 723

INTERVENTIONS Twenty-four weeks of varenicline titrated to 1 mg twice daily or placebo with

a reduction target of 50% or more in number of cigarettes smoked by 4 weeks, 75% or more
by 8 weeks, and a quit attempt by 12 weeks.
MAIN OUTCOMES AND MEASURES Primary efficacy end point was carbon monoxideconfirmed
self-reported abstinence during weeks 15 through 24. Secondary outcomes were carbon
monoxideconfirmed self-reported abstinence for weeks 21 through 24 and weeks
21 through 52.
RESULTS The varenicline group (n = 760) had significantly higher continuous abstinence
rates during weeks 15 through 24 vs the placebo group (n = 750) (32.1% for the varenicline
group vs 6.9% for the placebo group; risk difference (RD), 25.2% [95% CI, 21.4%-29.0%];
relative risk (RR), 4.6 [95% CI, 3.5-6.1]). The varenicline group had significantly higher
continuous abstinence rates vs the placebo group during weeks 21 through 24 (37.8% for the
varenicline group vs 12.5% for the placebo group; RD, 25.2% [95% CI, 21.1%-29.4%]; RR, 3.0
[95% CI, 2.4-3.7]) and weeks 21 through 52 (27.0% for the varenicline group vs 9.9% for the
placebo group; RD, 17.1% [95% CI, 13.3%-20.9%]; RR, 2.7 [95% CI, 2.1-3.5]). Serious adverse
events occurred in 3.7% of the varenicline group and 2.2% of the placebo group (P = .07).
CONCLUSIONS AND RELEVANCE Among cigarette smokers not willing or able to quit within the
next month but willing to reduce cigarette consumption and make a quit attempt at 3
months, use of varenicline for 24 weeks compared with placebo significantly increased
smoking cessation rates at the end of treatment, and also at 1 year. Varenicline offers a
treatment option for smokers whose needs are not addressed by clinical guidelines
recommending abrupt smoking cessation.
TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT01370356

JAMA. 2015;313(7):687-694. doi:10.1001/jama.2015.280

Author Affiliations: Mayo Clinic,

Rochester, Minnesota (Ebbert);
Department of Psychiatry, University
of Vermont, Burlington, Vermont
(Hughes); Health Behaviour Research
Centre, Department of Epidemiology
and Public Health, University College
London, United Kingdom (West);
Division of Pulmonary, Critical Care,
Sleep, and Allergy, University of
Nebraska Medical Center, Omaha
(Rennard); Pfizer Inc, New York,
New York (Russ, McRae, Treadow, Yu,
Dutro, Park).
Corresponding Author: Jon O.
Ebbert, MD, MSc, Mayo Clinic, 200
First St SW, Rochester, MN 55905
([email protected]).

(Reprinted) 687

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

Research Original Investigation

Varenicline for Smoking Cessation Through Smoking Reduction

orty percent of cigarette smokers make an average of 2

quit attempts annually.1 In a telephone survey of 1000
current daily cigarette smokers, 44% reported a preference to quit through reduction in the number of cigarettes
smoked, and 68% would consider using a medication to facilitate smoking reduction.2 However, US clinical practice
guidelines3 recommend that smokers quit abruptly even
though only 8% of smokers report being ready to quit in the
next month.4 Developing effective interventions to achieve tobacco abstinence through gradual reduction could engage more
smokers in quitting.5-7
Among cigarette smokers not ready to quit, tobacco reduction incorporating nicotine replacement therapy and behavioral interventions decreases cigarettes smoked and increases future smoking abstinence.6,7 Population-based studies
suggest that quitting gradually may be less successful than quitting abruptly.8 However, a systematic review comparing both
approaches suggests that reducing cigarettes before the quit
date and quitting abruptly without prior reduction yields comparable quit rates.9
Almost all prior studies of pharmacotherapy-aided reduction have examined nicotine replacement therapies. Varenicline is a partial agonist binding with high affinity and selectivity at 4 2 neuronal nicotinic acetylcholine receptors. 10
Varenicline significantly increases smoking abstinence rates
among smokers seeking treatment and quitting abruptly.11,12
Among smokers not trying to stop, varenicline significantly reduces cigarette consumption13 and may increase quit attempts.14
Varenicline may be an effective intervention for smokers
who are not willing or able to make an immediate quit attempt but who would be willing to reduce their smoking in
preparation for a quit attempt in the future (ie, a reduce-toquit approach). A prior reduce-to-quit study of varenicline was
small, provided only 8 weeks of varenicline, and obtained
equivocal results.14 We conducted a larger, randomized, placebo-controlled clinical trial providing varenicline for 6 months
to evaluate a reduce-to-quit approach.

Methods
Study Design
Written consent forms and study procedures were approved
by the institutional review boards or ethics committees of
participating institutions and each enrolled participant voluntarily signed the consent form. A randomized, doubleblind, placebo-controlled trial was conducted at 61 centers in
10 countries (Australia, Canada, Czech Republic, Egypt,
Germany, Japan, Mexico, Taiwan, United Kingdom, and
United States) between July 2011 and July 2013. Study sites
included clinical trial centers, academic centers, and outpatient clinics. Study site training was provided at an investigator meeting with training materials maintained and accessible through a shared website. The study consisted of a
24-week treatment period followed by a 28-week nontreatment follow-up phase (protocol in Supplement 1). The first
12 weeks of treatment were the reduction phase and the next
12 weeks were the abstinence phase. Participants were
688

recruited through advertising. Recruitment advertisements

included the following language: Want to quit smoking but
prefer to cut down first? and Are you ready to quit but prefer to do it gradually? and Want to quit smoking, but hate
the idea of going cold turkey? Enrollment ended when
recruitment goals were achieved. Participants received financial compensation for time spent for clinic and phone visits
as well as travel time; the amount of remuneration was determined by each clinical trial site.

Screening and Eligibility Criteria

Eligible participants were 18 years or older, smoked an average of 10 or more cigarettes per day with no continuous abstinence period longer than 3 months in the past year, had an exhaled carbon monoxide level higher than 10 ppm, and were
not willing or able to quit smoking within the next month but
were willing to reduce their smoking and make a quit attempt
within the next 3 months.
Exclusion criteria included a history of a suicide attempt
or suicidal behavior in the previous 2 years as assessed by
the Columbia-Suicide Severity Rating Scale (C-SSRS)15 and the
Suicide Behavior Questionnaire-Revised (SBQ-R)16; major depressive or anxiety disorder assessed by a physician as severe
(lifetime or current) or unstable (ie, medication dose change
or exacerbations in the last 6 months); lifetime diagnosis of psychosis, panic disorder, posttraumatic stress disorder, or schizophrenia; alcohol or substance abuse in the last 12 months; a
diagnosis of severe chronic obstructive pulmonary disease;
clinically significant cardiovascular or cerebrovascular disease in the previous 2 months; taking more than a limited number of doses of varenicline previously; and self-reported inability to abstain from noncigarette tobacco products,
marijuana, or smoking cessation aids (including electronic cigarettes). Women were excluded if pregnant, lactating, or likely
to become pregnant and unwilling to use contraception.

Study Procedures
Participants were randomized to receive varenicline or placebo for 24 weeks of treatment in a 1:1 ratio using a computergenerated block randomization schedule within site (Figure 1).
Investigators obtained participant identification numbers and
treatment group assignments through a web-based or telephone call-in drug management system. Participants, investigators, and research personnel were blinded to randomization until after the database was locked.
Race/ethnicity was self-reported. At each clinic visit and telephone contact, information was collected on cigarette or other
nicotine product use. Exhaled carbon monoxide measurements were obtained at all clinic visits. Tobacco dependence was
assessed with the Fagerstrm Test for Nicotine Dependence.17
Study design is shown in the eFigure in Supplement 2.
Adverse events and US Food and Drug Administration defined serious adverse events (adverse events resulting in death,
hospitalization, or other important medical events) were collected during study visits during the treatment phase and up
to 1 month after last treatment dose. A semistructured interview solicited information about psychiatric adverse events.
Suicidal ideation and behavior were assessed using the C-SSRS

JAMA February 17, 2015 Volume 313, Number 7 (Reprinted)

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

jama.com

Varenicline for Smoking Cessation Through Smoking Reduction

Original Investigation Research

Figure 1. Varenicline vs Placebo for Smoking Cessation Through Smoking Reduction Trial Flow Diagram
1747 Individuals signed informed
consent at screening visit
237 Excluded
185 Did not meet entrance criteria
45 Had significant medical condition or abnormal
laboratory test
45 Had substance use or abuse
44 Did not meet minimum smoking requirement
24 Had a psychiatric history
18 Unable to comply with schedule or procedures
9 Previous varenicline study or use
23 No longer willing to participate
9 Lost to follow-up
20 Othera

1510 Participants randomized

760 Randomized to receive varenicline

751 Received treatment as randomized
9 Withdrew

750 Randomized to receive placebo

742 Received treatment as randomized
7 Withdrew
1 Declined study medication but
completed study

145 Discontinued study during

treatment phaseb
1 Died
13 Adverse events
39 No longer willing to participate
56 Lost to follow-up
3 Protocol violation
6 Insufficient clinical responsec
27 Other

181 Discontinued study during

treatment phaseb
0 Died
12 Adverse events
47 No longer willing to participate
62 Lost to follow-up
2 Protocol violation
28 Insufficient clinical responsec
30 Other

606 Completed treatment phase of study

562 Completed treatment phase of studyd

47 Discontinued study during

follow-up phase after treatmente
0 Died
0 Adverse events
15 No longer willing to participate
20 Lost to follow-up
0 Protocol violation
0 Insufficient clinical responsec
12 Other

46 Discontinued study during

follow-up phase after treatmente
0 Died
1 Adverse event
12 No longer willing to participate
19 Lost to follow-up
1 Protocol violation
0 Insufficient clinical responsec
13 Other

559 Completed study

516 Completed studyd

760 Included in the efficacy analyses

750 Included in the efficacy analyses

751 Included in the safety analyses

742 Included in the safety analyses

Participants excluded due to reasons classified by the investigators as other

included not attending randomization visit; unable to commit to attending
study visits; change in work schedule; change in concomitant medications;
change in personal circumstances; and unavailability of urine drug
screening kits.
Treatment phase was weeks 1 through 24. Discontinuations from study during
treatment phase due to reasons classified by the investigators as other
included new job or change in work schedule; moved out of area;
change in personal or family circumstances; and unwilling or unable to
attend visits.

at baseline and all study visits. Participants completed the Patient Health Questionnaire (PHQ)-918 to assess the frequency
and severity of potential depression-related events every other
week during the treatment phase and at clinic visits during the
follow-up phase.
jama.com

Insufficient clinical response was a prepopulated option chosen by the

investigators on the case report forms.

Includes 1 participant in the placebo group who declined study medication but
completed study participation.

The follow-up phase after treatment was weeks 25 through 52.

Discontinuations from study during follow-up phase after treatment due to
reasons classified by the investigators as other included new job or change in
work schedule; moved out of area; change in personal or family
circumstances; and unwilling or unable to attend visits.

Interventions
Participants were asked to reduce their baseline smoking rate
by 50% or more by week 4 with further reduction to 75% or
more from baseline by week 8 with the goal of quitting by week
12. Counseling training was provided at the investigator meet(Reprinted) JAMA February 17, 2015 Volume 313, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

689

Research Original Investigation

Varenicline for Smoking Cessation Through Smoking Reduction

ing. Smoking cessation counseling was provided consistent

with the recommendations of the US Department of Health and
Human Services Treating Tobacco Use and Dependence clinical practice guideline.3 Participants received the Clearing the
Air: Quit Smoking Today booklet from the National Cancer Institute. Advice on reduction techniques was provided such as
systematically increasing the amount of time between cigarettes and rank-ordering and then eliminating cigarettes from
easiest to hardest to give up.14 Counseling was tailored to the
participants needs during the reduction, abstinence, and posttreatment phases. Counsellors were urged to be consistent and
brief, to focus on problem solving (eg, what triggers the urge
to smoke) and skills training (eg, practical actions to avoid
smoking), and to highlight successes not failures. Individual
counseling lasted 10 minutes or less per visit during 18 clinic
and 10 telephone visits. The last cigarette was to be smoked
prior to midnight on the day before the week 12 visit. Participants could reduce their smoking faster and could make a quit
attempt before week 12. Participants who had not reduced or
made a quit attempt by week 12 were encouraged to continue
medications and visits and make quit attempts, and participants who relapsed after week 12 were encouraged to make new
quit attempts. After the 24-week treatment phase (ie, 12 weeks
of reduction and 12 weeks after a quit attempt) participants
were followed up through week 52 (28-week nontreatment
follow-up phase).
Participants started with a recommended varenicline (or
matching placebo) dosage of 0.5 mg once daily for 3 days, increasing to 0.5 mg twice daily for days 4 to 7, and then to the
maintenance dose of 1 mg twice daily.

A logistic regression model included treatment effect as

the explanatory variable and investigative center as a covariate. In addition, an expanded logistic regression model including the treatment-by-center interaction was used to test for the
interaction effect. However, as prespecified in the statistical
analysis plan, the inferences were based solely on the predesignated logistic regression model including only the main effects of treatment and center, regardless of the significance of
the treatment-by-center interaction (with P .05 considered
significant). Relative risks (RRs) and risk differences (RDs) calculated using Proc Freq (SAS Institute) with option RELRISK
(for RRs) and RISKDIFF (for RDs) are reported for efficacy end
points. We conducted a sensitivity analysis for the primary end
point in which participants who were missing a carbon monoxide measurement during weeks 15 through 24 were classified as smokers.
To preserve the type I familywise error rate of .05, a fixedsequence procedure was used. The treatment comparison was
performed first for weeks 15 through 24, then for weeks 21
through 24, and then for weeks 21 through 52. A post hoc analysis was conducted in the same manner for the end point of the
CAR for weeks 15 through 52. Each test used a 2-sided P value
of .05 or less for significance. We used SAS (SAS Institute), version 9.2, for statistical analyses.
Adverse events were coded using the Medical Dictionary
for Regulatory Activities (MedDRA), version 16.1.21 Adverse
events occurring during treatment and up to 30 days after receiving the last dose of study drug are reported. All safety analyses included all randomized participants who received any dose
of study medication. Our study was not powered to detect significant differences in adverse events between groups.

Study End Points

The primary efficacy end point was the carbon monoxide
confirmed continuous abstinence rate (CAR) during the last 10
weeks of treatment (ie, weeks 15-24). A participant was considered abstinent from tobacco if he or she self-reported tobacco
abstinence throughout the period and had an exhaled carbon
monoxide level of 10 ppm or less at each visit. In the case of a
missing visit, participants were considered abstinent if they were
abstinent at the next nonmissed visit and also reported not
smoking during the missed visit. A missing carbon monoxide
measurement did not disqualify a participant from meeting the
end point if they self-reported not smoking. Secondary efficacy end points were the CARs during weeks 21 through 24 and
during weeks 21 through 52. We also calculated the nonprespecified end point of the CAR for weeks 15 through 52.

Results
Enrollment and Follow-up
Of 1747 potentially eligible participants screened, 1510 (86%)
were randomly assigned to receive varenicline (n = 760) or placebo (n = 750). Overall study completion was defined as completion of the week 52 visit and was 73.6% (559 of 760 participants) in the varenicline group and 68.8% (516 of 750
participants) in the placebo group (Figure 1). Participants assigned to study groups were similar in demographic and smoking characteristics at baseline (Table 1). Participants who discontinued treatment were encouraged to remain in the study.

Smoking Abstinence
Statistics
The efficacy analysis was based on the intent-to-treat population (all randomized participants). Participants who dropped
out of the study were treated as smokers. A sample size of 1404
randomized participants in a 1:1 ratio (702 in each group) was
estimated to provide 90% or more power to detect a difference between varenicline and placebo of 10.3% in the primary end point of CAR during weeks 15 through 24, assuming
a CAR of 17.2% for varenicline and 6.9% for placebo using a
2-group, continuitycorrected, 2-sided 2 test. A P value of .05
or less was considered significant.9,19,20
690

The varenicline group (n = 760) had significantly higher continuous abstinence rates during weeks 15 through 24 than the
placebo group (n = 750) (32.1% for the varenicline group vs 6.9%
for the placebo group; RD, 25.2% [95% CI, 21.4%-29.0%]; RR,
4.6 [95% CI, 3.5-6.1]) (Table 2). The varenicline group had significantly higher continuous abstinence rates vs the placebo
group during weeks 21 through 24 (37.8% for the varenicline
group vs 12.5% for the placebo group; RD, 25.2% [95% CI, 21.1%29.4%]; RR, 3.0 [95% CI, 2.4-3.7]) and weeks 21 through 52
(27.0% for the varenicline group vs 9.9% for the placebo group;
RD, 17.1% [95% CI, 13.3%-20.9%]; RR, 2.7 [95% CI, 2.1-3.5]). No

JAMA February 17, 2015 Volume 313, Number 7 (Reprinted)

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

jama.com

Varenicline for Smoking Cessation Through Smoking Reduction

significant treatment-by-center interaction for the primary end

point was observed in the logistic regression model.
Among the 244 participants receiving varenicline who were
counted as abstinent for the primary end point, there were 25
participants with at least 1 missing carbon monoxide measurement (including those with missing visits) during weeks 15
through 24; among the 52 participants receiving placebo who
met the primary end point, there were 4 participants with at
least 1 missing carbon monoxide measurement. We conducted a sensitivity analysis with participants who were missing a carbon monoxide measurement being classified as smokers for the weeks 15 through 24 continuous abstinence rate (219
participants [28.8%] for varenicline vs 48 participants [6.4%]
for placebo). The RR for this analysis was 4.5 (95% CI, 3.46.1). This value approximates the value we observed using the
prespecified imputation (RR, 4.6 [95% CI, 3.5-6.1]).Among participants meeting the primary end point (abstinent during
weeks 15-24), the median time from baseline to the beginning
of the continuous abstinence period was 50 days for varenicline and 85 days for placebo (P < .001). The varenicline group
also had a significantly higher 7-day point prevalence smoking abstinence rate compared with placebo at weeks 12, 24, and
52 (Figure 2). Varenicline significantly increased the 4-week
point prevalence smoking abstinence rate compared with placebo at week 52 (32.8% for varenicline vs 17.3% for placebo; RD,
15.4% [95% CI, 11.1%-19.7%]; RR, 1.9 [95% CI, 1.6-2.3])

Original Investigation Research

Table 1. Participant Characteristics

Varenicline Group
(n = 760)
Demographic Characteristics
Sex, No. (%)
Men

425 (55.9)

Women

335 (44.1)

324 (43.2)

44.7 (11.8) [19-79]

44.4 (12.0) [18-78]

White

476 (62.6)

463 (61.7)

Black

36 (4.7)

47 (6.3)

Asian

175 (23.0)

177 (23.6)

Other

73 (9.6)

63 (8.4)

5.5 (2.1)

5.6 (2.0)

Age started smoking, y

17.3 (4.3)

17.3 (4.4)

No. of cigarettes per day in

past month

20.6 (8.5)

20.8 (8.2)

Age, mean (SD) [range], y

Smoking characteristics, mean

(SD)
Fagerstrm Test for Nicotine
Dependence scorea

Longest Period of Abstinence

Since started smoking, d
Median (IQR)

30 (3-240)

21 (2-180)

Mean

272

206

In past year, d
Median (IQR)

0 (0-1)

0 (0-0)

Mean

Smoking Reduction (Weeks 1 to 12)

At week 4, 47.1% of participants treated with varenicline (358
of 760 participants) reduced the number of cigarettes smoked
per day (ie, average number of cigarettes during days on which
smoking occurred over the last week) compared with baseline by 50% or more or abstained completely compared with
31.1% of participants treated with placebo (233 of 750 participants) (RD, 16.0% [95% CI, 11.2%-20.9%]; RR, 1.5 [95% CI, 1.31.7]). After 8 weeks, 26.3% participants in the varenicline group
(200 of 760 participants) reduced smoking by 75% or more from
baseline or abstained compared with 15.1% participants in the
placebo group (113 of 750 participants) (RD, 11.3% [95% CI, 7.2%15.3%]; RR, 1.8 [95% CI, 1.4-2.2]).

Since started smoking, No. (%)

The percentage of participants with adverse events was higher

in the varenicline group than in the placebo group (618 of 751
participants [82.3%] in the varenicline group vs 538 of 742 participants [72.5%] in the placebo group). Adverse events with
the greatest risk difference between varenicline and placebo
(>2%) were nausea, abnormal dreams, insomnia, constipation, vomiting, and weight gain (Table 3). Adverse event incidence resulting in permanent treatment discontinuation was
not significantly different between the 2 groups (63 of 751 participants [8.4%] in the varenicline group vs 52 of 742 participants [7.0%] in the placebo group; P = .27). Percentages of participants with serious adverse events were not significantly
different between varenicline and placebo (28 of 751 participants [3.7%] in the varenicline group vs 16 of 742 participants
[2.2%] in the placebo group; P = .07). During treatment and up
to 30 days after the last dose, suicidal ideation or behavior was
jama.com

426 (56.8)

Race/ethnicity, No. (%)

No. of Serious Quit Attempts by Any Method

Safety

Placebo Group
(n = 750)

None

130 (17.1)

159 (21.2)

191 (25.1)

188 (25.1)

139 (18.3)

110 (14.7)

300 (39.5)

293 (39.1)

Abbreviation: IQR, interquartile range.

a

Fagerstrm Test for Nicotine Dependence measures the degree of cigarette

dependence with scores ranging from 0 to 10 with higher scores indicating a
higher degree of dependence.17

recorded on the C-SSRS in 6 of 751 participants (0.8%) in the

varenicline group and 10 of 742 participants (1.3%) in the placebo group. Any increases in PHQ-9 depression scores from
baseline to any time point after baseline occurred in 169 of 751
participants (22.5%) treated with varenicline compared with
145 of 742 participants (19.5%) treated with placebo (P = .16).

Discussion
Among cigarette smokers not willing or able to quit smoking
in the next month but willing to reduce with the goal of quitting in the next 3 months, varenicline produced a statistically
and clinically significant increase in the CARs at the end of
treatment and at 28 weeks after treatment. Varenicline produced greater smoking reduction than placebo prior to quitting. Varenicline was not associated with significant increases in treatment discontinuations due to adverse events.
Smokers enrolled in the current study were not ready to quit
in the next month, and overall smoking abstinence rates would
(Reprinted) JAMA February 17, 2015 Volume 313, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

691

Research Original Investigation

Varenicline for Smoking Cessation Through Smoking Reduction

Table 2. Continuous Carbon MonoxideConfirmed Smoking Abstinence Rates for Periods of the Studya
Continuous Abstinence, No. of Participants (%)
Varenicline Group
(n = 760)

Placebo Group
(n = 750)

Risk Difference, % (95% CI)

Relative Risk (95% CI)

Primary End Pointb

Weeks 15-24

244 (32.1)

52 (6.9)

25.2 (21.4-29.0)

4.6 (3.5-6.1)

Weeks 21-24

287 (37.8)

94 (12.5)

25.2 (21.1-29.4)

3.0 (2.4-3.7)

Weeks 21-52

205 (27.0)

74 (9.9)

17.1 (13.3-20.9)

2.7 (2.1-3.5)

182 (24.0)

45 (6.0)

18.0 (14.5-21.4)

4.0 (2.9-5.4)

Secondary End Pointsb

Post Hoc End Pointb

Weeks 15-52
a

All randomized participants were included at all time points. For nonmissed
visits, a carbon monoxide level higher than 10 ppm disqualified participants
from meeting the end point. Participants with missing carbon monoxide
measurements met the end point if they self-reported not smoking. Participants
with missing visits were considered abstinent if they self-reported not smoking
and did not have a carbon monoxide measurement higher than 10 ppm at the
next nonmissed visit and also reported not smoking during the missed visit.
Participants who withdrew from the study were considered smokers after the
time of withdrawal.

Figure 2. Seven-Day Point Prevalence Smoking Abstinence for

Participants Receiving Varenicline vs Placebo
50

7-Day Point Prevalence, %

Varenicline (n = 760)
40

30

20
Placebo (n = 750)

10

0
0

12

16

20

24

28

32

36

40

44

48

52

Week

RD indicates risk difference; RR, relative risk.

Week 12: RD, 24.5% (95% CI, 20.8%-28.3%), RR, 4.7 (95% CI, 3.5-6.2); week
24: RD, 25.7% (95% CI, 21.2%-30.1%), RR, 2.5 (95% CI, 2.1-3.0); week 52: RD,
15.8% (95% CI, 11.5%-20.2%), RR, 1.9 (95% CI, 1.6-2.2).
All randomized participants were included at all time points. For nonmissed visits,
carbon monoxide level higher than 10 ppm disqualified participants from meeting
the end point when abstinence was self-reported. Participants with missing
carbon monoxide measurements met the end point if they self-reported smoking
abstinence. Participants with missing visits were considered smokers for that visit.
Participants who withdrew from the study were considered smokers from the
time of withdrawal. Error bars indicate 95% CIs.

have been expected to be low. Although overall smoking abstinence rates were low in the placebo group, varenicline increased
the rates of achieving abstinence such that the absolute abstinence rates were similar to those observed in studies of varenicline in smokers motivated to quit after 1 week of treatment.11,12
The mechanism of varenicline action as an aid to gradual
cessation could relate to a reduction in cigarette craving or a
blockade of the reinforcing action of nicotine through partial
agonist activity at the nicotinic acetylcholine receptors.10 Ancillary effects from varenicline may exist with respect to con692

The primary end point was carbon monoxideconfirmed self-reported

abstinence during weeks 15 through 24; secondary end points,
carbon monoxideconfirmed self-reported abstinence for
weeks 21 through 24 and weeks 21 through 52; post hoc end point,
carbon monoxideconfirmed self-reported abstinence for weeks 15
through 52.

fidence in ability to quit. However, this should have been controlled through study blinding and any effect via this route is
likely to be small given limited evidence that confidence plays
a causal role in sustaining quitting attempts.22
Adverse events caused by varenicline were similar to previous observations. In the present study, varenicline was associated with an increased rate of constipation and weight gain.
However, both are established effects of smoking cessation,23,24
and it is possible that the greater incidence of abstinence with
varenicline and not the direct effect of varenicline was the
cause. The incidence of bronchitis was lower in those treated
with varenicline, an effect which is possibly mediated by an
increased rate of smoking cessation.25 Varenicline did not increase the risk of suicidal ideation or behavior or other psychiatric adverse events.
Major strengths of this study include the randomized design,
large sample size, and convergent validity of the findings across
multiple outcomes and measures. One limitation of this study
relates to the exclusion of potential participants if they had severe psychiatric, pulmonary, cardiovascular, or cerebrovascular
disease. As a result, the generalizability of this treatment approach
to a broader population of smokers who need to quit smoking
but may want to achieve it through reduction is unknown. In addition, participants in the current study were provided with
significant support with counseling from trained staff occurring
during 18 clinic and 10 telephone visits. Because of this, the observed abstinence rates with varenicline in actual clinical practice might be expected to be less than that observed in the
current trial. We did not test whether varenicline would be more
effective for the reduce-to-quit cessation approach than other
tobacco treatments such as nicotine replacement therapy.
Uncertainty remains as to the prevalence of smokers in the
general population who meet the definition of smokers enrolled in this study. In a cross-sectional study collecting data
via telephone or face-to-face interview with daily smokers responding to the Current Population Survey, the prevalence of
contemplators (ie, interested in quitting smoking in next 6

JAMA February 17, 2015 Volume 313, Number 7 (Reprinted)

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

jama.com

Varenicline for Smoking Cessation Through Smoking Reduction

Original Investigation Research

Table 3. Adverse Events Occurring During Treatment Plus 30 Days in 2% or More of Participants Who Received
1 or More Doses of Study Drug in Either Treatment Groupa
No. (%)
Adverse Events
Nausea

Varenicline Group
(n = 751)

Placebo Group
(n = 742)

Risk Difference, % (95% CI)

209 (27.8)

67 (9.0)

Nasopharyngitis

98 (13.0)

89 (12.0)

18.80 (14.99 to 22.61)

1.05 (2.30 to 4.41)

Abnormal dreams

86 (11.5)

43 (5.8)

5.66 (2.83 to 8.49)

Insomnia

80 (10.7)

51 (6.9)

3.78 (0.92 to 6.64)

Upper respiratory tract infection

63 (8.4)

63 (8.5)

0.10 (2.92 to 2.72)

Headache

62 (8.3)

54 (7.3)

0.98 (1.74 to 3.69)

Anxiety

52 (6.9)

65 (8.8)

1.84 (4.56 to 0.89)

Fatigue

46 (6.1)

34 (4.6)

1.54 (0.74 to 3.82)

Irritability

39 (5.2)

30 (4.0)

1.15 (0.98 to 3.28)

Constipation

38 (5.1)

13 (1.8)

3.31 (1.48 to 5.14)

Increased appetite

37 (4.9)

30 (4.0)

0.88 (1.22 to 2.98)

Sleep disorder

37 (4.9)

29 (3.9)

1.02 (1.06 to 3.10)

Dizziness

32 (4.3)

27 (3.6)

0.62 (1.35 to 2.60)

Vomiting

31 (4.1)

13 (1.8)

2.38 (0.67 to 4.08)

Back pain

28 (3.7)

29 (3.9)

0.18 (2.12 to 1.76)

Weight increased

28 (3.7)

12 (1.6)

2.11 (0.48 to 3.74)

Diarrhea

27 (3.6)

23 (3.1)

0.50 (1.33 to 2.32)

Depressed mood

26 (3.5)

27 (3.6)

0.18 (2.05 to 1.70)

Depression

25 (3.3)

35 (4.7)

1.39 (3.38 to 0.61)

Decreased appetite

20 (2.7)

19 (2.6)

0.10 (1.52 to 1.72)

Agitation

20 (2.7)

14 (1.9)

0.78 (0.74 to 2.29)

Dyspepsia

19 (2.5)

9 (1.2)

1.32 (0.05 to 2.69)

Abdominal pain upper

18 (2.4)

7 (0.9)

1.45 (0.16 to 2.75)

Middle insomnia

17 (2.3)

11 (1.5)

0.78 (0.59 to 2.16)

Flatulence

17 (2.3)

6 (0.8)

1.46 (0.21 to 2.70)

Influenza

16 (2.1)

12 (1.6)

0.51 (0.86 to 1.89)

Dry mouth

16 (2.1)

11 (1.5)

0.65 (0.70 to 2.00)

Abdominal distention

16 (2.1)

8 (1.1)

1.05 (0.22 to 2.32)

Somnolence

15 (2.0)

10 (1.3)

0.65 (0.65 to 1.95)

Initial insomnia

15 (2.0)

9 (1.2)

0.78 (0.49 to 2.06)

Cough

14 (1.9)

23 (3.1)

1.24 (2.81 to 0.34)

Disturbance in attention

13 (1.7)

17 (2.3)

0.56 (1.98 to 0.86)

Sinusitis

12 (1.6)

16 (2.2)

0.56 (1.94 to 0.82)

Bronchitis

10 (1.3)

23 (3.1)

1.77 (3.26 to 0.28)

months but not in the next 30 days) was 33.2%.4 We were not
attempting to fit smokers into a specific stage of readiness for
behavior change.26 Instead, our approach aimed to reduce barriers to engaging in the quitting process by allowing and facilitating smoking reduction in a precessation phase.27 Our
sample most closely resembles the 33% of smokers who want
to quit sometime between 1 and 6 months in the future. The
approach used in this study would be expected to be of interest to 14 million of the 42 million current smokers.
The US Public Health Service3 and other guidelines recommend smokers set a quit date in the near future and quit
abruptly. However, many smokers may be unwilling to commit to a quit date at a clinic visit. Because most clinicians are
likely to see smokers at times when a quit date in the next
month is not planned, the current study indicates that prescription of varenicline with a recommendation to reduce the
jama.com

All participants had been

randomized and had received at
least 1 dose (including partial doses)
of study drug; and includes all
participants who had experienced
an adverse event from the date of
first dose of study drug and up to 30
d after the last dose of study drug.
Adverse events were recorded in
the clinical report form (volunteered
and observed adverse events) and,
in addition, neuropsychiatric
adverse events were solicited in a
semistructured neuropsychiatric
adverse event interview. Adverse
events were coded according to the
Medical Dictionary for Regulatory
Activities (MedDRA).21 Participants
were counted for each type of
adverse event but were counted
only once for multiple occurrences
of the same type of event.

number of cigarettes smoked per day with the eventual goal

of quitting could be a useful therapeutic option for this population of smokers. The approach of reduction with the goal of
quitting increases the options for a clinician caring for a smoker.

Conclusions
Among cigarette smokers not willing or able to quit within the
next month but willing to reduce cigarette consumption and
make a quit attempt at 3 months, use of varenicline for 24 weeks
compared with placebo significantly increased smoking cessation rates at the end of treatment, and also at 1 year. Varenicline offers a treatment option for smokers whose needs are
not addressed by clinical guidelines recommending abrupt
smoking cessation.
(Reprinted) JAMA February 17, 2015 Volume 313, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

693

Research Original Investigation

Varenicline for Smoking Cessation Through Smoking Reduction

ARTICLE INFORMATION
Author Contributions: Drs Ebbert and Yu had full
access to all of the data in the study and take
responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Hughes, West, Rennard,
Russ, Treadow, Yu, Park.
Acquisition, analysis, or interpretation of data:
Ebbert, Hughes, West, Rennard, Russ, McRae,
Treadow, Yu, Dutro.
Drafting of the manuscript: Ebbert, Hughes, West,
Rennard, Russ, McRae, Treadow, Yu, Park.
Critical revision of the manuscript for important
intellectual content: Ebbert, Hughes, Rennard, Russ,
Yu, Dutro, Park.
Statistical analysis: Yu.
Obtained funding: Park.
Administrative, technical, or material support:
Treadow, Dutro, Park.
Study supervision: Ebbert, Russ, McRae, Treadow,
Park.
Conflict of Interest Disclosures: The authors have
completed and submitted the ICMJE Form for
Disclosure of Potential Conflicts of Interest. Dr
Ebbert reports receiving grants from JHP
Pharmaceuticals, Orexigen, Pfizer, and the National
Institutes of Health; and receiving personal fees
from GlaxoSmithKline during the conduct of the
study. Dr Hughes reports receiving personal fees
from Alere/Free and Clear, Cicatelli, DLA Piper,
Dorrffermeyer, Embera, Equinox, GlaxoSmithKline,
Healthwise, Nicoventures, Pfizer, Pro Ed, Publicis,
Selecta, and nonfinancial support from Swedish
Match. Dr West reports receiving grants, personal
fees, and nonfinancial support from Pfizer,
GlaxoSmithKline, and Johnson & Johnson. His
salary is funded by a Centre grant from Cancer
Research UK. Dr Rennard reports being an advisory
board member for A2B Bio, Almirall, Dalichi Sankyo,
Novartis, Nycomed, and Pfizer; consulting for
Almirall, APT Pharma/Britnall, AstraZeneca,
Boehringer Ingelheim, Chiesi, CSL Behring, Decision
Resource, Dunn Group, Easton Associates, Gerson,
GlaxoSmithKline, MedImmune, Novartis, Pearl,
Roche, Takeda, and Theravance; receiving lecture
fees from CME Incite, Forest, Novis, PriMed, and
Takeda; receiving grants from AstraZeneca,
Boehringer Ingelheim, GlaxoSmithKline, Johnson &
Johnson, Novartis, and Otsuka. Drs Russ, McRae,
Yu, Dutro, Park, and Ms Treadow are employees
and stockholders of Pfizer.
Funding/Support: This study was funded by Pfizer.
Role of the Funders/Sponsors: Pfizer was involved
in the design and conduct of the study; collection,
management, analysis, and interpretation of the
data; preparation, review, and approval of the
manuscript; and decision to submit the manuscript
for publication.
Additional Contributions: We thank the
investigators and study site personnel involved in
the study. They were compensated for their
contributions by Pfizer. Editorial support in the
form of developing tables and figures; editing,
proofing, and formatting the text; collating review
comments; and preparing the manuscript for

694

submission was provided by Abegale Templar, PhD

(Engage Scientific, Horsham, United Kingdom), and
funded by Pfizer.
REFERENCES
1. Borland R, Partos TR, Yong HH, Cummings KM,
Hyland A. How much unsuccessful quitting activity
is going on among adult smokers? Addiction. 2012;
107(3):673-682.

placebo on smoking behaviors and urges among

nontreatment seeking smokers. J Psychopharmacol.
2012;26(10):1383-1390.
14. Hughes JR, Rennard SI, Fingar JR, Talbot SK,
Callas PW, Fagerstrom KO. Efficacy of varenicline to
prompt quit attempts in smokers not currently
trying to quit: a randomized placebo-controlled
trial. Nicotine Tob Res. 2011;13(10):955-964.

2. Shiffman S, Hughes JR, Ferguson SG, Pillitteri JL,

Gitchell JG, Burton SL. Smokers interest in using
nicotine replacement to aid smoking reduction.
Nicotine Tob Res. 2007;9(11):1177-1182.

15. Posner K. Suicidality issues in clinical trials:

Columbia suicide adverse event identification in
FDA safety analyses. http://www.fda.gov/ohrms
/dockets/ac/07/slides/2007-4306s1-01-CU-Posner
.ppt. Accessed January 20, 2015.

3. Fiore MC, Jaen CR, Baker TB, et al. Treating

tobacco use and dependence: 2008 update.
http://www.ahrq.gov/professionals/clinicians
-providers/guidelines-recommendations/tobacco
/index.html. Accessed December 31, 2014.

16. Osman A, Bagge CL, Gutierrez PM, Konick LC,

Kopper BA, Barrios FX. The Suicidal Behaviors
Questionnaire-Revised (SBQ-R): validation with
clinical and nonclinical samples. Assessment. 2001;
8(4):443-454.

4. Wewers ME, Stillman FA, Hartman AM,

Shopland DR. Distribution of daily smokers by stage
of change: Current Population Survey results. Prev
Med. 2003;36(6):710-720.

17. Heatherton TF, Kozlowski LT, Frecker RC,

Fagerstrm KO. The Fagerstrm Test for Nicotine
Dependence: a revision of the Fagerstrm
Tolerance Questionnaire. Br J Addict. 1991;86(9):
1119-1127.

5. Asfar T, Ebbert JO, Klesges RC, Relyea GE.

Do smoking reduction interventions promote
cessation in smokers not ready to quit? Addict Behav.
2011;36(7):764-768.
6. Wang D, Connock M, Barton P, Fry-Smith A,
Aveyard P, Moore D. Cut down to quit with
nicotine replacement therapies in smoking
cessation: a systematic review of effectiveness and
economic analysis. Health Technol Assess.
2008;12(2):iii-iv; ix-xi; 1-135.
7. Moore D, Aveyard P, Connock M, Wang D,
Fry-Smith A, Barton P. Effectiveness and safety of
nicotine replacement therapy assisted reduction to
stop smoking: systematic review and meta-analysis.
BMJ. 2009;338:b1024.
8. Cheong Y, Yong HH, Borland R. Does how you
quit affect success? Nicotine Tob Res. 2007;9(8):
801-810.
9. Lindson N, Aveyard P, Hughes JR. Reduction vs
abrupt cessation in smokers who want to quit.
Cochrane Database Syst Rev. 2010;(3):CD008033.
10. Coe JW, Brooks PR, Vetelino MG, et al.
Varenicline: an 42 nicotinic receptor partial
agonist for smoking cessation. J Med Chem. 2005;
48(10):3474-3477.
11. Gonzales D, Rennard SI, Nides M, et al;
Varenicline Phase 3 Study Group. Varenicline, an
42 nicotinic acetylcholine receptor partial
agonist, vs sustained-release bupropion and
placebo for smoking cessation: a randomized
controlled trial. JAMA. 2006;296(1):47-55.
12. Jorenby DE, Hays JT, Rigotti NA, et al;
Varenicline Phase 3 Study Group. Efficacy of
varenicline, an 42 nicotinic acetylcholine receptor
partial agonist, vs placebo or sustained-release
bupropion for smoking cessation: a randomized
controlled trial. JAMA. 2006;296(1):56-63.
13. Ashare RL, Tang KZ, Mesaros AC, Blair IA, Leone
F, Strasser AA. Effects of 21 days of varenicline vs

18. Kroenke K, Spitzer RL, Williams JB. The PHQ-9:

validity of a brief depression severity measure.
J Gen Intern Med. 2001;16(9):606-613.
19. Hatsukami DK, Rennard S, Patel MK, et al.
Effects of sustained-release bupropion among
persons interested in reducing but not quitting
smoking. Am J Med. 2004;116(3):151-157.
20. Cahill K, Stead LF, Lancaster T. Nicotine
receptor partial agonists for smoking cessation.
Cochrane Database Syst Rev. 2008;(3):CD006103.
21. Medical Dictionary for Regulatory Activities
website. http://www.meddra.org/. Accessed
December 31, 2014.
22. Vangeli E, Stapleton J, Smit ES, Borland R, West
R. Predictors of attempts to stop smoking and their
success in adult general population samples:
a systematic review. Addiction. 2011;106(12):21102121.
23. Hajek P, Gillison F, McRobbie H. Stopping
smoking can cause constipation. Addiction. 2003;
98(11):1563-1567.
24. Aubin HJ, Farley A, Lycett D, Lahmek P, Aveyard
P. Weight gain in smokers after quitting cigarettes:
meta-analysis. BMJ. 2012;345:e4439.
25. Willemse BW, Postma DS, Timens W, ten
Hacken NH. The impact of smoking cessation on
respiratory symptoms, lung function, airway
hyperresponsiveness, and inflammation.
Eur Respir J. 2004;23(3):464-476.
26. Prochaska JO, DiClemente CC. Stages and
processes of self-change of smoking: toward an
integrative model of change. J Consult Clin Psychol.
1983;51(3):390-395.
27. Baker TB, Mermelstein R, Collins LM, et al. New
methods for tobacco dependence treatment
research. Ann Behav Med. 2011;41(2):192-207.

JAMA February 17, 2015 Volume 313, Number 7 (Reprinted)

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/18/2015

jama.com