Champix Research

ORIGINAL CONTRIBUTION JAMA-EXPRESS

Effect of Maintenance TherapyWith Varenicline on Smoking CessationA Randomized Controlled TrialSerena Tonstad, MD, PhDPhilip Tønnesen, MD, PhDPeter Hajek, PhDKathryn E. Williams, PhDClare B. Billing, MSKaren R. Reeves, MDfor the Varenicline Phase 3Study Group

ONE OF THE MOST STRIKING

features of attempts to stopsmoking is the discrep-ancy between early suc-

cess rates and long-term outcomes. Inmost studies, as well as in large-scaletreatment services for dependent smok-ers, more than 50% of smokers man-age to achieve abstinence for at least afew initial weeks.1 For the majority ofsmokers, this is the most difficult pe-riod of the quit attempt, and at the endof it, they report diminishing with-drawal discomfort and increasing con-fidence in the possibility of remainingsmoke-free.2 Nevertheless, 50% to 60%of initially successful quitters go on torelapse within a year.3-5

Among the numerous interven-tions proposed to prevent relapse is thatof Marlatt and Gordon,4 who recom-mend that patients should learn to iden-tify situations likely to lead to relapseand be provided with cognitive and be-havioral strategies to cope with thesesituations. This approach is by far the

See also pp 47, 56, and 94.

Author Affiliations: Department of Preventive Car-diology, Ulleval University Hospital, and Universityof Oslo, Oslo, Norway (Dr Tonstad); Department ofPulmonary Medicine, University of Copenhagen,and Gentofte Hospital, Copenhagen, Denmark (DrTønnesen); Wolfson Institute of Preventive Medi-cine, Queen Mary, University of London, London,England (Dr Hajek); and Pfizer Global Research and

Development, Groton, Conn (Drs Williams andReeves and Mr Billing).Members of the Varenicline Phase 3 Study Group arelisted at the end of this article.Corresponding Author: Serena Tonstad, MD, PhD, De-partment of Preventive Cardiology, Ulleval Univer-sity Hospital, N-0407 Oslo, Norway ([email protected]).

Context The majority of cigarette smokers who achieve abstinence relapse withinthe first year and require many attempts before achieving permanent abstinence. Evi-dence to support pharmacological treatment for relapse prevention is insufficient.

Objective To determine whether smokers who quit after 12 weeks of treatment withvarenicline, a selective �4�2 nicotinic acetylcholine receptor partial agonist, maintaingreater continuous abstinence rates (defined as not a single “puff” of a cigarette) thanplacebo controls during an additional 12 weeks of treatment and until 52 weeks aftertreatment initiation.

Design, Setting, and Participants Randomized controlled trial conducted at mul-tiple medical clinics in 7 countries with follow-up to 52 weeks after study baseline. Of1927 cigarette smokers recruited between April 2003 and February 2004 and treated for12 weeks with open-label varenicline titrated to 1 mg twice per day, 1236 (64.1%) didnot smoke, use tobacco, or use nicotine replacement therapy during the last week of treat-ment and 62.8% (n=1210) were randomized to additional treatment or placebo.

Intervention Participants were randomly assigned to receive either double-blind vareni-cline, 1 mg twice per day (n=603), or placebo (n=607) for an additional 12 weeks.

Main Outcome Measures Carbon monoxide–confirmed continued abstinence dur-ing weeks 13 to 24 and weeks 13 to 52 of the study.

Results The carbon monoxide–confirmed continuous abstinence rate was signifi-cantly higher for the varenicline group than for the placebo group for weeks 13 to 24(70.5% vs 49.6%; odds ratio [OR], 2.48; 95% confidence interval [CI], 1.95-3.16;P�.001) as well as for weeks 13 to 52 (43.6% vs 36.9%; OR, 1.34; 95% CI, 1.06-1.69; P=.02). Adverse events reported in the open-label period were mostly mild; nodifference in adverse events between varenicline and placebo was observed duringthe double-blind period.

Conclusions Smokers who achieved abstinence for at least 7 days at the end of 12weeks of open-label varenicline treatment and were randomized to receive an addi-tional 12 weeks of varenicline treatment showed significantly greater continuous ab-stinence in weeks 13 to 24 compared with placebo. This advantage was maintainedthrough the nontreatment follow-up to week 52. Varenicline may be an efficacious,safe, and well-tolerated agent for maintaining abstinence from smoking.

Trial Registration clinicaltrials.gov Identifier: NCT00143286JAMA. 2006;296:64-71 www.jama.com

64 JAMA, July 5, 2006—Vol 296, No. 1 (Reprinted) ©2006 American Medical Association. All rights reserved.

at Cardinal Health, on July 5, 2006 www.jama.comDownloaded from

http://www.jama.com

most widely used. Despite the impor-tance of this area of research, few con-trolled studies have examined such in-terventions. A recent comprehensivereview of existing studies concludedthat currently there is no evidence-based relapse prevention interventionavailable.5,6

Varenicline is a highly selective �4�2nicotinic acetylcholine receptor par-tial agonist developed specifically forsmoking cessation. The efficacy ofa 12-week course of varenicline sur-passes that of bupropion and pla-cebo.7,8 The odds of quitting are nearlyquadrupled at the end of treatment com-pared with placebo7,8 and nearly tripledat 1-year follow-up.8 We report hereinthe results of a randomized, double-blind, placebo-controlled trial evaluat-ing the efficacy of an additional 12weeks of varenicline used for relapseprevention in smokers who success-fully achieved abstinence following aninitial 12-week open-label vareniclinetreatment.

METHODSParticipants

Cigarette smokers between the ages of18 and 75 years were recruited be-tween April 13, 2003, and February 17,2004, at 6 US and 18 internationalmedical clinics experienced in smok-ing cessation. Entry criteria includedmen and women who smoked an av-erage of 10 or more cigarettes per dayduring the past year and over the monthprior to the screening visit, with no pe-riod of abstinence longer than 3 monthsin the past year and who were moti-vated to quit. Women with childbear-ing potential were required to be ac-tively practicing effective contraception.Participants were excluded if they hada serious or unstable disease within thepast 6 months; required treatment fordepression within the past 12 months;had a history of or current panic dis-order, psychosis, or bipolar disorder;had severe chronic obstructive pulmo-nary disease, a history of cancer, evi-dence or history of clinically signifi-cant allergic reactions, laboratoryabnormalities, cardiovascular disease

within the past 6 months, uncon-trolled hypertension, or a history ofdrug or alcohol abuse or dependencewithin the past 12 months; used asmoking cessation aid (ie, nicotine re-placement therapy [NRT], bupro-pion, clonidine, or nortriptyline) withinthe previous month; used tobacco prod-ucts other than cigarettes or mari-juana within the past month and did notagree to abstain from use of these prod-ucts during study participation; had abody mass index of less than 15 or morethan 38 (calculated as weight in kilo-grams divided by height in meterssquared); or used any of the followingmedications: NRT, antidepressants,antipsychotics, mood stabilizers/anticonvulsants, naltrexone, steroids,or insulin.

Study Design

This was a 52-week, multicenter trialapproved by the institutional reviewboard at each site, and all participantsprovided written informed consentprior to any procedures. The study wasconducted in compliance with the ethi-cal principles of the Declaration of Hel-sinki and the International Confer-ence on Harmonization Good ClinicalPractices Guidelines. Following 12weeks of open-label treatment withvarenicline, 1 mg twice daily, partici-pants who had successfully abstainedfrom smoking and use of tobacco andNRT for at least the last 7 days of thetreatment period were randomly as-signed to receive either varenicline orplacebo and continued into a 12-week, double-blind treatment phase.Thereafter, participants continued intoa nontreatment follow-up phase for anadditional 28 weeks (for a total of 52weeks in the study).

Following a 3- to 21-day screeningperiod, participants returned for a base-line visit, which was the start of the 12-week open-label phase. Sociodemo-graphic data were obtained; race andethnicity were assessed by self-report.Participants received varenicline, 0.5mg daily for 3 days, 0.5 mg twice dailyfor 4 days, then 1 mg twice daily for 11weeks. At each visit, participants re-

ceived up to 10 minutes of smoking ces-sation counseling in accordance withUS Public Health Service guidelines.9

Participants were advised to take studymedication for 1 week before attempt-ing to quit smoking but could attemptto quit earlier if they wished. Fromtheir target quit date, participants at-tempted to remain abstinent fromsmoking and other nicotine use. Par-ticipants attended clinic visits weeklyfrom weeks 1 through 8 and at weeks10 and 12. At the end of the open-label treatment phase, participants wereeligible for randomization into thedouble-blind treatment phase if they re-ported that they had not smoked orused tobacco or NRT during the last 7days, had an end-expiratory carbonmonoxide (CO) level of 10 ppm or less,were adherent to the study drug, andwere deemed appropriate for study par-ticipation according to the protocol.

The study used a single centralized,computer-generated randomization se-quence (stratified by center with a blocksize of 4) to assign participants to re-ceive varenicline or placebo during the12-week double-blind phase in a ratioof 1:1. Participants were to continue toabstain from smoking or nicotine useand attended clinic visits at weeks 13,14, 16, 20, and 24. During the double-blind treatment phase, participants whodiscontinued study medication orbegan smoking could continue partici-pation in the study provided they at-tended the remaining visits and com-pleted all assessments. At the end of thedouble-blind treatment phase, all par-ticipants continued in the 28-week non-treatment follow-up phase, regardlessof whether they had completed the full12 weeks of double-blind treatment orhad stopped smoking. The participantblinding was maintained during thisphase. Participants attended fol-low-up visits at weeks 25, 28, 36, 44,and 52 and were contacted by tele-phone at weeks 26, 32, 40, and 48.

Assessments

At each contact (clinic visit or tele-phone call), participants were askedquestions (the Nicotine Use Inven-

EFFECT OF VARENICLINE ON SMOKING CESSATION

©2006 American Medical Association. All rights reserved. (Reprinted) JAMA, July 5, 2006—Vol 296, No. 1 65


http://www.jama.com

tory, designed for the current study andother studies) regarding use of ciga-rettes and other nicotine-containingproducts (during the treatment peri-ods) or tobacco products (during post-treatment follow-up) since the last con-tact and during the past 7 days. At eachclinic visit, exhaled CO was mea-sured.

Safety Evaluations

Complete physical examinations wereconducted at the screening or baselinevisit and at weeks 12 and 24. Visits atscreening, baseline, and weeks 2, 12,and 24 included a 12-lead electrocar-diogram and blood laboratory measure-ments.

Efficacy Evaluations

The primary efficacy end point wasthe CO-confirmed continuous absti-nence rate from week 13 throughweek 24 (double-blind treatmentphase). A key secondary efficacy endpoint was the continuous abstinencerate from week 13 through week 52.Continuous abstinence rates were also

summarized for each clinic visit ortelephone contact. The continuousabstinence rate from week 13 througha given time point was defined as theproportion of participants who wereabstainers from smoking and fromnicotine or tobacco-containing prod-ucts from the randomization visitthrough the time point of interest.Participants were considered to beabstainers from smoking since theprior clinic visit or telephone contactif they responded “no” to the questionof smoking any cigarettes (even asingle “puff”) since the last study visitor contact and had not used any othernicotine-containing products (duringthe double-blind treatment phase) ortobacco products (during the non-treatment follow-up phase) since theprevious study visit or contact. Addi-tionally, participants whose CO valuewas more than 10 ppm were classifiedas smokers.

Those missing a CO value but meet-ing other abstinence criteria were con-sidered nonsmokers, with the excep-tion of the week 52 visit, at which time

participants with a missing CO valuewere considered smokers. Partici-pants were considered nonsmokers atmissed visits if, at the next nonmissedvisit, they reported not smoking or us-ing nicotine or tobacco products sincethe last study visit.

Point prevalence of abstinence wasdefined as a self-report of no smokingor other tobacco use (or other nico-tine product use during the treatmentphases) in the previous 7 days, con-firmed by an expired CO value of notmore than 10 ppm at clinic visits. Par-ticipants with a missing smoking as-sessment were considered smokers forthat 7-day period.

Time to first lapse was calculatedfrom the date of first randomizedtherapy to the date of the first ciga-rette smoked (even a puff).

The Minnesota Nicotine With-drawal Scale (MNWS)10,11 was used toassess the experience of craving andwithdrawal after the end of treatmentwith varenicline. The MNWS was self-administered by participants at weeks13 and 25 to assess symptoms duringthe previous week.

Statistical Analyses

The study was powered to show differ-ences between varenicline and placebofor continuous abstinence rate fromweeks 13 to 24. The sample size esti-mate was based on data published for bu-propion vs placebo.12 Assuming treat-mentdifferences similar to those reportedfor bupropion, the number of partici-pants per treatment group required toprovide 80% power to detect a treat-ment difference using a 2-group conti-nuity-corrected �2 test with a .05 2-sidedsignificance level was estimated to be410. Ultimately, a higher abstinence ratethan expected at the end of the open-label period resulted in a larger numberof randomized participants than planned.

The by-treatment efficacy compari-sons are reported for all randomizedparticipants on an intention-to-treat ba-sis. All significance tests were 2-tailedusing a significance level of P=.05.

All measures of abstinence were ana-lyzed as binary data using a logistic re-

Table 1. Baseline Participant Characteristics

Characteristics

Open-LabelVarenicline Phase

(n = 1927)

Double-Blind Phase

Varenicline(n = 603)

Placebo(n = 607)

Age, mean (SD), y 44.2 (10.7) 45.4 (10.4) 45.3 (10.4)

Male, No. (%) 941 (48.8) 303 (50.2) 293 (48.3)

Race/ethnicity, No. (%)White 1853 (96.2) 583 (96.7) 589 (97.0)

Black 35 (1.8) 9 (1.5) 10 (1.6)

Asian 14 (0.7) 3 (0.5) 4 (0.7)

Other 25 (1.3) 8 (1.3) 4 (0.7)

Fagerstrom score, mean (SD)* 5.55 (2.04) 5.43 (1.96) 5.35 (1.98)

No. of years of smoking,mean (SD) [range]

27.2 (10.7) [2-59] 28.2 (10.4) [3-58] 28.1 (10.5) [2-58]

No. of cigarettes per dayin past month,mean (SD) [range]

21.6 (8.3) [3-99] 20.7 (7.3) [8-60] 20.7 (7.5) [10-65]

Previous serious quit attempts,No. (%)

0 341 (17.7) 88 (14.6) 89 (14.7)

�1 586 (82.3) 515 (85.4) 518 (85.3)

Longest period of abstinencein past year, d

Mean (SD) 7.4 (18.4) 8.3 (19.6) 7.6 (18.3)

Median (range) 0 (0-200) 0 (0-90) 0 (0-90)*The possible range is from 1 to 10, with higher scores indicating greater dependence on nicotine. The numbers of

participants completing all questions of the Fagerstrom test were as follows: open-label varenicline, n = 1922; double-blind varenicline, n = 602; double-blind placebo, n = 605.




http://www.jama.com

gression model, including treatmentand center, and testing was performedout using the likelihood ratio �2 test.In the analyses for all of these param-eters, participants who withdrew fromthe study and therefore did not have datafor subsequent visits were assumed to besmokers for the remainder of the study.The confidence intervals (CIs) for thenumbers needed to treat were calcu-lated by taking the reciprocal of the 95%CI bounds for the difference in the con-tinuous abstinence rates using the nor-mal approximation.

Time to first lapse after randomiza-tion was analyzed using the productlimit (Kaplan-Meier) method with a log-rank test for the comparison of vareni-cline and placebo. Participants who dis-continued prematurely from the studywithout having smoked were assigneda date of first cigarette that was 1 dayafter the date on which they met the cri-terion for continuous abstinence. Par-ticipants who remained abstinentthrough the end of the study were cen-sored at the date of the week 52 visit.

The statistical software used for allanalyses was SAS, version 8.2 (SAS In-stitute Inc, Cary, NC).

RESULTSDemographic and smoking character-istics of the participants are shown inTABLE 1. Of 1927 who participated inthe first 12 weeks of open-label treat-ment, 1236 (64.1%) met the pointprevalence criterion for abstinence dur-ing week 12. Twenty-six chose not tocontinue or had other reasons for notbeing eligible to continue, and 1210were randomized and were included inefficacy evaluations. Participant dispo-sition is shown in FIGURE 1.

Continuous Abstinence Rates

The continuous abstinence rate forweeks 13 to 24 was higher for partici-pants randomized to varenicline thanfor participants randomized to pla-cebo (odds ratio, 2.48; 95% CI, 1.95-3.16; P� .001)(TABLE 2). The continu-ous abstinence rate for weeks 13 to 52was also higher for the vareniclinegroup than for the placebo group (odds

ratio, 1.34; 95% CI, 1.06-1.69; P = .02)(Table 2). The number needed to treatfor the continuous abstinence rate ofvarenicline relative to placebo was 5(95% CI, 4-7) for weeks 13 to 24 and14 (95% CI, 8-71) for weeks 13 to 52.

Point Prevalence of Abstinence

The 7-day point prevalence of absti-nence at week 24 (odds ratio, 2.82; 95%CI, 2.18-3.64; P�.001) and at week 52(odds ratio, 1.33; 95% CI, 1.06-1.67;P=.01) was significantly higher for par-ticipants who received varenicline thanfor participants who received placebo(FIGURE 2).

Time to First LapseThe median time to the first lapse (post-randomization to double-blind treat-ment) was significantly longer forparticipants receiving double-blindvarenicline than for participants treatedwith double-blind placebo (198 [95%CI, 159-260] days vs 87 [95% CI,58-143] days, respectively; log-rankP�.001) (FIGURE 3).

Weight Change

For participants who were continu-ously abstinent from smoking duringweeks13to24,meanweightchangefromopen-label varenicline baseline to week

Figure 1. Participant Flow Through the Study

603 Assigned to Receive Double-Blind Varenicline602 Received Treatment as Assigned

607 Assigned to Receive Double-Blind Placebo604 Received Placebo as Assigned

2416 Individuals Screened

1928 Assigned to Receive Open-Label Varenicline1927 Received Treatment

488 Excluded377 Did Not Meet Inclusion Criteria69 Refused to Participate42 Other

717 Discontinued Treatment132 Lost to Follow-up200 Adverse Events29 Lack of Efficacy71 Protocol Deviation1 Pregnancy

150 Refusal to Participate Further134 Other

494 Completed Study

61 Discontinued Follow-up28 Lost to Follow-up2 Adverse Events0 Lack of Efficacy

27 Refusal to Participate Further2 Deaths2 Other

463 Completed Study

47 Discontinued Follow-up24 Lost to Follow-up1 Adverse Events2 Lack of Efficacy

19 Refusal to Participate Further0 Deaths1 Other

555 Completed Treatment Period

47 Discontinued Treatment12 Lost to Follow-up8 Adverse Events4 Lack of Efficacy3 Protocol Deviation


510 Completed Treatment Period

94 Discontinued Treatment31 Lost to Follow-up8 Adverse Events5 Lack of Efficacy2 Protocol Deviation


603 Included in Efficacy Analysis602 Included in Safety Analysis

607 Included in Efficacy Analysis604 Included in Safety Analysis

1210 Randomized




http://www.jama.com

24 was 3.62 kg (SD, 3.29 kg) for partici-pants who received double-blind vareni-cline vs 4.03 kg (SD, 3.30 kg) for thedouble-blind placebo participants. Themean change in body weight betweenweeks 12 and 24 for these abstinent par-ticipants was 0.80 kg (SD, 2.13 kg) forthose treated with double-blind vareni-cline vs 1.51 kg (SD, 2.31 kg) for the pla-cebo group. Among all participants, themean weight change from open-label

baseline to week 24 for double-blindvarenicline-treatedparticipantswas3.41kg (SD, 3.27 kg) vs 3.53 kg (SD, 3.20 kg)for placebo-treated participants. Themean change in body weight for thedouble-blind varenicline-treated partici-pants between weeks 12 and 24 was 0.71kg (SD, 2.18 kg) vs 1.02 kg (SD, 2.35 kg)for placebo-treated participants.

MNWS Scores

In this study, the MNWS was used toassess experience of craving and with-drawal after treatment with vareni-cline was discontinued at the end of theopen-label and double-blind treat-ment periods. At weeks 13 and 25, thewithdrawal symptoms tended on aver-age to be low (between “slight” and “notat all” in the 5-point rating scale) (dataavailable from authors on request).The differences in the mean urge-to-smoke subscale scores between the pla-cebo and varenicline groups are shownin FIGURE 4.

Safety and Tolerability

TABLE 3 shows the most frequent ad-verse events. During the open-labelphase, 229 participants (11.9%) expe-rienced adverse events that led to treat-ment discontinuation. Events that con-tributed most frequently to treatmentdiscontinuation were nausea (3.2% ofparticipants), headache (1.0%), insom-

nia (0.9%), depression (0.8%), and fa-tigue (0.6%). The median onset to nau-sea was 8 days (interquartile range, 3-15days) and the median duration was 20days (interquartile range, 6-54 days).Vomiting was reported in 56 partici-pants (2.9%).

The incidence of adverse events dur-ing the double-blind treatment phasewas similar for the varenicline and pla-cebo groups (46% and 45%, respec-tively). A total of 10 varenicline-treated participants (1.7%) and 8placebo participants (1.3%) discontin-ued because of adverse events. No ad-verse event resulted in discontinua-tion in more than 1 participant in eithertreatment group.

Three participants died in this study.None of these deaths was considered re-lated to the study drug. One man witha history of depression that was not re-vealed at the time of enrollment diedof suicide 27 days after completingdouble-blind treatment. Another mandied of complications related to lungcancer 19 days after completing double-blind treatment. A third man discon-tinued the open-label phase of the studyon day 25 because of an adverse eventof back pain and died of a rectal sar-coma 197 days after last vareniclinedose. Nonfatal serious adverse eventswere reported for 20 participants dur-ing or after treatment with open-labelvarenicline. Ten participants random-ized to double-blind varenicline and 5participants randomized to double-blind placebo experienced nonfatal se-rious adverse events during or aftertreatment.

During the double-blind treatmentphase, changes from baseline in bloodpressure did not differ between thevarenicline and placebo groups. Themean pulse rate remained approxi-mately the same from the double-blind period baseline in participantstreated with varenicline and de-creased by approximately 2/min in pla-cebo participants.

COMMENTIn this randomized clinical trial, smok-ers who successfully quit after taking

Table 2. Carbon Monoxide–ConfirmedContinuous Abstinence Rate at Clinic Visits

No. (%) of Participants

Double-BlindVarenicline(n = 603)

Double-BlindPlacebo(n = 607)

Double-blindtreatmentphase, wk*

13 576 (95.5) 537 (88.5)

14 551 (91.4) 476 (78.4)

16 509 (84.4) 413 (68.0)

20 454 (75.3) 331 (54.5)

24 425 (70.5) 301 (49.6)

Nontreatmentfollow-upphase, wk†

25 408 (67.7) 293 (48.3)

28 361 (59.9) 282 (46.5)

36 306 (50.7) 257 (42.3)

44 280 (46.4) 239 (39.4)

52 263 (43.6) 224 (36.9)*Weeks 13-24: odds ratio, 2.48; 95% confidence interval,

1.95-3.16; P�.001.†Weeks 13-52: odds ratio, 1.34; 95% confidence interval,

1.06-1.69; P = .02.

Figure 2. Seven-Day Point Prevalence of Abstinence

4 8

Open-LabelTreatment Phase

Double-Blind Treatment Phase

DrugTreatment

12

†

∗

WeekWeek12 16 20 24 28 32 36 40 44 48 52

100

30

60

50

40

70

80

90

20

10

0

Res

pond

ers,

%

100

30

60

50

40

70

80

90

20

10

0

Abs

tinen

ce, %

Double-Blind Varenicline (n = 603)Treatment Group

Double-Blind Placebo (n = 607)Telephone Visit

Telephone VisitVarenicline(n = 1927)

*P�.001.†P=.01.




http://www.jama.com

open-label varenicline for an initial 12weeks experienced a significantly re-duced rate of relapse when taking anadditional 12 weeks of varenicline, 1 mgtwice per day, compared with pla-cebo. There was a substantial differ-ence in continuous validated absti-nence rates at the end of treatment anda smaller but significant difference 6months later. In this study, vareni-cline was safe and well tolerated for pro-longed use, and most adverse eventswere characterized as mild or moder-ate in intensity. In the field of relapseprevention—in which there is a no-table lack of positive findings—these re-sults represent an important new de-velopment.

A recent review of relapse preven-tion literature identified a need for stud-ies randomizing abstaining smokers andusing strict outcome criteria.5,6 Thepresent trial used stringent methods, set-ting standards of rigor in this difficultarea. In particular, the use of continu-ous rather than point prevalence absti-nence rates as the primary outcome cri-terion, with self-reports validated byfrequent CO monitoring, contributes tothe robust results. Study participants inthe open-label phase attended 10 clinicvisits. During the double-blind phase,participants attended 5 clinic visits dur-ing the treatment period, as well as an-other 5 visits and 4 telephone contactsduring the 6-month posttreatment fol-low-up. The visits were brief but werestill likely to provide a level of motiva-tional support not available in routinecare. However, the level of supportwould not be expected to contribute todifferences between the active drug andplacebo. Since intensive support is prob-ably of particular benefit to partici-pants taking placebo, the support couldin fact have reduced the effect size.

It is important to note that the 43.6%rate of continuous abstinence achievedat week 52 reflects only participants whowere abstinent for at least 7 days at theend of the first 12 weeks of treatment.We included only participants who hadmade progress in cessation to under-stand whether prolonged treatmentwould help these participants to re-

mainabstinent.Theweek12pointpreva-lence rate of abstinence was unusuallyhigh (64.1%), and it surpassed the ratesof about 50% that were observed in 2smoking cessation studies of 12 weeksof treatment with varenicline vs bupro-pionorplacebo.7,8 Thisdifference isprob-ably due to the open-label design of thefirst phase of the present study com-pared with the double-blind design of the2 smoking cessation studies.

A temporary acceleration of the rateof relapse (as measured by the strict cri-terion of a single “puff”) occurred in thevarenicline group after the withdrawalof medication. Toward week 52, therelapse rates of the 2 study groups did

not show any further signs of converg-ing (Table 2). A statistically and clini-cally significant difference remained inplace through the follow-up period. The7% increment observed in this study issimilar to the increment in 12-monthabstinence rates achieved with a stan-dard regimen of 8 to 12 weeks of treat-ment with nicotine replacement therapycompared with placebo13 and to the totalof 9.5% difference vs placebo found inthe Cochrane meta-analysis of bupro-pion smoking cessation trials,14 both ofwhich are widely accepted to have clini-cal relevance.15

At the end of this trial, as in all ex-isting literature on smoking cessation

Figure 3. Time to First Lapse

280240

Log-Rank P<.001

20016012080

Time From First Double-Blind Dose, d

40

1.00

0.50

0.75

0.25

0

Pro

port

ion

Abs

tinen

t

Double-Blind Varenicline (n = 603)Double-Blind Placebo (n = 607)

Figure 4. Mean Scores on the Urge to Smoke Subscale of the MNWS

Varenicline Placebo

571No. 560Week 13

499 468Week 25

551 505Week 13

418 328Week 25

1.50

0.75

0.50

1.00

1.25

0.25

0

Mea

n S

core

1.50

0.75

0.50

1.00

1.25

0.25

0

Mea

n S

core

All Participants Abstainers∗

No.

Error bars indicate 95% confidence intervals.*Defined by the 7-day point prevalence abstinence prior to the week of Minnesota Nicotine Withdrawal Scale(MNWS) evaluation, weeks 13 and 25.




http://www.jama.com

with 1 year of follow-up, more than 50%of participants in each group returnedto smoking. Thus, an examination oflonger medication periods is war-ranted. Discontinuation of vareniclinein participants randomized to placeboat week 12 was followed by a meanurge-to-smoke value that was small andonly modestly higher than that of par-ticipants continuing to receive double-blind varenicline. Similarly, at week 24,discontinuation of varenicline was notfollowed by a large difference in urgeto smoke than discontinuation of pla-cebo, both for all participants and forabstainer groups.

A recent Cochrane review5 identi-fied 2 studies of bupropion vs placebothat examined continued abstinence insmokers who had initially quit eitherwith bupropion12 or with the nicotinepatch.16 These trials did not detect a sig-nificant effect at 6 months16 or at 1year12 after the end of treatment.

In this study, varenicline was safe andwell tolerated. Some of the adverseevents recorded in the open-label phaseof the trial are known tobacco with-drawal effects (eg, constipation andsleep disturbance). The only notabledrug effect was nausea, which wasmostly mild and tolerable. Although32% of participants reported some nau-sea, only 3% discontinued the medica-tion because of it. During the second12-week course of varenicline, therewere no adverse events reported morefrequently in the varenicline group thanin the placebo group.

The study population included a gen-erally healthy sample because of thenumber of exclusion criteria, as manyclinical trials of new medications do,so there may be limits to its generaliz-ability. During the double-blind treat-ment phase, the Nicotine Use In-ventory did not distinguish amongparticipants who used NRT or to-bacco products other than cigarettes.Thus, in the analysis of the primary endpoint, 2 participants in the placebogroup were counted as smokers be-cause of the use of other nicotine-containing products. During the non-treatment period, NRT use did not

disqualify the participants from beingconsidered abstinent at these visits.

The number of participants classi-fied as lost to follow-up or who re-fused further participation differed be-tween the varenicline and placebogroups. The differences are explain-able by dropouts due to inability to stopsmoking. Although this study was ini-tiated prior to the publication of thestandardized smoking research meth-ods proposed as the Russell Stan-dard,17 it conformed to many of the pro-posed criteria. All participants wereencouraged to remain in the study andparticipate in assessments even if theydiscontinued treatment. Nonetheless,there were some participants who dis-continued the study for adverse eventsor protocol deviations. The numberswere small and similar between treat-ment groups. Although CO only has a4-hour half-life and does not providea complete check on the smoker’s self-report of abstinence from smoking,regular CO monitoring is in accor-dance with established standards.17 Thefollow-up period ended 1 year from theinitial quit date but 40 weeks from thestart of the double-blind phase. Sixmonths is the accepted minimum for

long-term follow-up,18 but future stud-ies may consider extended follow-upperiods.

In conclusion, extended use ofvarenicline helps recent ex-smokers tomaintain their abstinence and preventrelapse. Varenicline is the first smok-ing cessation treatment to demon-strate a significant long-term relapseprevention effect.

Author Contributions: Dr Tonstad had full access toall of the data in the study and takes responsibility forthe integrity of the data and the accuracy of the dataanalysis.Study concept and design: Williams, Billing, Reeves.Acquisition of data: Tonstad, Tønnesen, Hajek, Billing,Reeves.Analysis and interpretation of data: Tonstad, Tønnesen,Hajek, Williams, Billing, Reeves.Drafting of the manuscript: Tonstad, Tønnesen, Hajek,Billing.Critical revision of the manuscript for important in-tellectual content: Tonstad, Tønnesen, Hajek, Williams,Billing, Reeves.Statistical analysis: Tonstad, Billing.Obtained funding: Reeves.Administrative, technical, or material support: Williams.Study supervision: Reeves.FinancialDisclosures:DrTonstad reports receivinghono-raria for lecturing and consultancies for Pfizer and othermanufacturers of smoking cessation medications; DrTønnesen reports receiving honoraria for participationin advisory boards for Pfizer and Sanofi-Aventis and forpresentations relating to smoking cessation from Pfizerand GlaxoSmithKline; Dr Hajek reports consulting formanufacturers of smoking cessation medications, in-cluding Pfizer; Dr Williams reports stock ownership inPfizer; and Dr Reeves reports stock ownership in Pfizer.No other disclosures were reported.

Table 3. Most Frequent Adverse Events*

Adverse Events

No. (%) of Participants Taking�1 Dose of Study Medication

Open-LabelVarenicline Phase

(n = 1927)

Double-BlindTreatment Phase

Varenicline(n = 602)

Placebo(n = 604)

Gastrointestinal disordersNausea 645 (33.5) 7 (1.2) 4 (0.7)

Flatulence 234 (12.1) 2 (0.3) 0

Constipation 168 (8.7) 0 3 (0.5)

Dyspepsia 133 (6.9) 9 (1.5) 6 (1.0)

Psychiatric disordersInsomnia 377 (19.6) 16 (2.7) 17 (2.8)

Abnormal dreams† 276 (14.3) 6 (1.0) 0

Irritability 97 (5.0) 16 (2.7) 27 (4.5)

Headache 304 (15.8) 17 (2.8) 12 (2.0)

Nasopharyngitis 145 (7.5) 29 (4.8) 32 (5.3)

Fatigue 162 (8.4) 9 (1.5) 11 (1.8)*All-cause adverse events that occurred in 5% or more of participants in the open-label treatment phase or in the double-

blind phase in varenicline-treated participants at a frequency greater than that for placebo; adverse events that be-gan or increased in severity during study drug treatment or up to 7 days after the last dose.

†“Abnormal dreams” is the MedDRA (Medical Dictionary for Regulatory Activities) for several descriptions, includingvivid dreams and increased frequency of dreams.




http://www.jama.com

Funding/Support: This study was sponsored by PfizerInc, which provided funding, study drug and pla-cebo, and monitoring.Role of the Sponsor: Pfizer Inc was involved in all el-ements of this study, including but not limited to thestudy design and monitoring. In addition, the data-base containing the findings of the 25 individual in-vestigator sites was maintained by Pfizer Inc, and sta-tistical analyses were performed at Pfizer Inc by MrBilling and Ann Pennington, MS.Independent Statistical Analysis: Ingar Holme,PhD, professor of biostatistics, University of Oslo,and Ulleval University Hospital, had access to all ofthe data used in the study and performed an inde-pendent analysis. Dr Holme performed analyses ofthe primary and key secondary end points usinglogistic regression and cross-tabulations. Resultswere identical to those obtained by the sponsor. DrHolme received compensation for this reanalysisfrom Pfizer Inc.

Varenicline Phase 3 Study Group: Andrew Pipe, MD,University of Ottawa Heart Institute, Ottawa, On-tario; David I. Stewart, MD, Cornwall Medical Centre,Cornwall, Prince Edward Island; Gerald Brosky, MD,MSc, Capital District Health Authority, Centre for Clini-cal Research, Halifax, Nova Scotia; Ginette Girard, MD,Novabyss Inc, Sherbrooke, Quebec; Gordon Ford, MD,FRCPC, Rockyview General Hospital, Calgary, Al-berta; Lew Pliamm, MD, Canadian Phase Onward, Tor-onto, Ontario; Eva Kralikova, MD, PhD, General Uni-versity Hospital, Dependence Treatment Department,Prague, Czech Republic; Martin Doessing, MD, De-partment of Medicine, Frederikssund, Frederiksborg,Denmark; Ronald Dahl, MD, Department of Respira-tory Diseases, Aarhus University Hospital, Aarhus, Den-mark; Erling Aaserud, MD, Kristianborg StudiescenterDA, Bergen, Hordaland, Norway; Jorn Ossum Gron-ert, MD, Flattum Legesenter, Honefoss, Buskerud, Nor-way; Erik Lunell, MD, PhD, Fagerstrom Consulting, Hels-ingborg, Skane, Sweden; Hans Gilljam, MD, CTP/

Samhallsmedicin, Stockholm, Sweden; Johan Herlitz,MD, Division of Cardiology, Sahlgrenska University Hos-pital, Goteborg, Sweden; Michele D. Reynolds, MD, Ra-diant Research, Dallas, Tex; Peter D. Londberg, MD,Summit Research Network LLC, Seattle, Wash; Sam-uel N. Lederman, MD, Radiant Research, West PalmBeach, Fla; John E. Pappas, MD, Central Kentucky Re-search Associates Inc, Lexington, Ky; Martin Edwards,MD, The Jenner Practice, London, England; PhilipHoward, MD, St Helier Hospital, Clinical Trials Unit, Lon-don, England; Melvin J. Tonkon, MD, Apex ResearchInstitute, Santa Ana, Calif; David P. Sachs, MD, Divi-sion of Pulmonary and Critical Care, Stanford Univer-sity School of Medicine, Palo Alto, Calif.Acknowledgment: We acknowledge the contribu-tion of Ingar Holme, PhD, University of Oslo and Ul-leval University Hospital, Oslo, Norway, for assis-tance in the independent analysis of the data, and KellyStein Marcus, PhD, from Cardinal Health, Wayne, NJ,for editorial support.

REFERENCES

1. Judge K, Bauld L, Chesterman J, Ferguson J. TheEnglish smoking treatment services: short-termoutcomes. Addiction. 2005;100(suppl 2):46-58.2. Hughes JR,Gust SW,SkoogK,KeenanRM,FenwickJW.Symptomsof tobaccowithdrawal: a replicationandextension. Arch Gen Psychiatry. 1991;48:52-59.3. Ferguson J, Bauld L, Chesterman J, Judge K. TheEnglish smoking treatment services: one-yearoutcomes. Addiction. 2005;100(suppl 2):59-69.4. Marlatt A, Gordon J. Relapse Prevention: Main-tenance Strategies in the Treatment of AddictiveBehaviors. New York, NY: Guilford Press; 1985.5. Hajek P, Stead LF, West R, Jarvis M. Relapse pre-vention interventions for smoking cessation. Co-chrane Database Syst Rev. 2005;(1):CD003999.6. Lancaster T, Hajek P, Stead L, West R, Jarvis MJ.Prevention of relapse after quitting smoking. Arch In-tern Med. 2006;166:828-835.7. Gonzales D, Rennard SI, Nides M, et al; for the Vareni-cline Phase 3 Study Group. Varenicline, an �4�2 nico-tinic acetylcholine receptor partial agonist, vs sustained-release bupropion and placebo for smoking cessation:a randomized controlled trial. JAMA. 2006;296:47-55.

8. Jorenby DE, Hays JT, Rigotti NA, et al; for the Vareni-cline Phase 3 Study Group. Efficacy of varenicline, an�4�2 nicotinic acetylcholine receptor partial agonist,vs placebo or sustained-release bupropion for smok-ing cessation: a randomized controlled trial. JAMA.2006;296:56-63.9. The Tobacco Use and Dependence Clinical PracticeGuideline Panel, Staff, and Consortium Representatives.A clinical practice guideline for treating tobacco use anddependence. JAMA. 2000;283:3244-3254.10. Cappelleri JC, Bushmakin AG, Baker CL, MerikleE, Olufade AO, Gilbert DG. Revealing the multi-dimensional framework of the Minnesota NicotineWithdrawal Scale. Curr Med Res Opin. 2005;21:749-760.11. Hughes JR, Hatsukami D. Signs and symptoms oftobacco withdrawal. Arch Gen Psychiatry. 1986;43:289-294.12. Hays JT, Hurt RD, Rigotti NA, et al. Sustained-release bupropion for pharmacologic relapse pre-vention after smoking cessation: a randomized,controlled trial. Ann Intern Med. 2001;135:423-433.

13. Silagy C, Lancaster T, Stead L, Mant D, FowlerG. Nicotine replacement therapy for smokingcessation. Cochrane Database Syst Rev. 2004;(3):CD000146.14. Hughes JR, Stead LF, Lancaster T. Antidepres-sants for smoking cessation. Cochrane Database SystRev. 2006;(1):CD000031.15. National Institute for Clinical Excellence. Guid-ance on the Use of Nicotine Replacement Therapy(NRT) and Bupropion for Smoking Cessation. Lon-don, England: National Institute for Clinical Excel-lence; 2004. Reference No. N0082.16. Hurt RD, Krook JE, Croghan IT, et al. Nicotine patchtherapy based on smoking rate followed by bupro-pion for prevention of relapse to smoking. J Clin Oncol.2003;21:914-920.17. West R, Hajek P, Stead L, Stapleton J. Outcomecriteria in smoking cessation trials: proposal for a com-mon standard. Addiction. 2005;100:299-303.18. Hughes JR, Keely JP, Niaura RS, Ossip-Klein DJ,Richmond RL, Swan GE. Measures of abstinence inclinical trials: issues and recommendations. NicotineTob Res. 2003;5:13-25.




http://www.jama.com

Champix Research

Documents

Transcript of Champix Research