eTable. Study Schedule
Ebbert JO, Hatsukami DK, Croghan IT, Schroeder DR, Allen SS, Hays JT, Hurt RD. Combination Varenicline and Bupropion SR for Tobacco-Dependence Treatment in Cigarette SmokersA Randomized Trial. JAMA. 2014;311(2):155–163. doi:10.1001/jama.2013.283185
Combining pharmacotherapies for tobacco-dependence treatment may increase smoking abstinence.
To determine efficacy and safety of varenicline and bupropion sustained-release (SR; combination therapy) compared with varenicline (monotherapy) in cigarette smokers.
Design, Setting, and Participants
Randomized, blinded, placebo-controlled multicenter clinical trial with a 12-week treatment period and follow-up through week 52 conducted between October 2009 and April 2013 at 3 midwestern clinical research sites. Five hundred six adult (≥18 years) cigarette smokers were randomly assigned and 315 (62%) completed the study.
Twelve weeks of varenicline and bupropion SR or varenicline and placebo.
Main Outcomes and Measures
Primary outcome was abstinence rates at week 12, defined as prolonged (no smoking from 2 weeks after the target quit date) abstinence and 7-day point-prevalence (no smoking past 7 days) abstinence. Secondary outcomes were prolonged and point-prevalence smoking abstinence rates at weeks 26 and 52. Outcomes were biochemically confirmed.
At 12 weeks, 53.0% of the combination therapy group achieved prolonged smoking abstinence and 56.2% achieved 7-day point-prevalence smoking abstinence compared with 43.2% and 48.6% in varenicline monotherapy (odds ratio [OR], 1.49; 95% CI, 1.05-2.12; P = .03 and OR, 1.36; 95% CI, 0.95-1.93; P = .09, respectively). At 26 weeks, 36.6% of the combination therapy group achieved prolonged and 38.2% achieved 7-day point-prevalence smoking abstinence compared with 27.6% and 31.9% in varenicline monotherapy (OR, 1.52; 95% CI, 1.04-2.22; P = .03 and OR, 1.32; 95% CI, 0.91-1.91; P = .14, respectively). At 52 weeks, 30.9% of the combination therapy group achieved prolonged and 36.6% achieved 7-day point-prevalence smoking abstinence compared with 24.5% and 29.2% in varenicline monotherapy (OR, 1.39; 95% CI, 0.93-2.07; P = .11 and OR, 1.40; 95% CI, 0.96-2.05; P = .08, respectively). Participants receiving combination therapy reported more anxiety (7.2% vs 3.1%; P = .04) and depressive symptoms (3.6% vs 0.8%; P = .03).
Conclusions and Relevance
Among cigarette smokers, combined use of varenicline and bupropion, compared with varenicline alone, increased prolonged abstinence but not 7-day point prevalence at 12 and 26 weeks. Neither outcome was significantly different at 52 weeks. Further research is required to determine the role of combination therapy in smoking cessation.
clinicaltrials.gov Identifier: http://clinicaltrials.gov/show/NCT00935818
Smoking accounts for 62% of deaths among female smokers and 60% of deaths among male smokers.1 Innovative pharmacotherapeutic approaches to tobacco-dependence treatment need investigation to reduce smoking-related death and disability.
Quiz Ref IDBupropion SR (sustained-release) and varenicline are nonnicotine pharmacotherapies indicated for tobacco-dependence treatment. Bupropion SR may mediate effects through noradrenergic and dopaminergic systems2 with a competitive inhibitory effect on nicotinic acetylcholine receptors.3Quiz Ref IDVarenicline is a partial agonist that binds with high affinity and selectivity at α4β2 neuronal nicotinic acetylcholine receptors.4,5 Opportunities exist for additive or synergistic therapeutic effects from combination therapy with these 2 medications.
Combination pharmacotherapy for treating tobacco dependence may increase smoking abstinence compared with monotherapy. A combination of bupropion SR and the nicotine patch is more effective than nicotine patch therapy alone,6 suggesting that an additive benefit is achieved by combining therapies. In an open-label pilot study evaluating combination therapy with varenicline and bupropion SR, the combination was well tolerated with smoking abstinence rates exceeding those observed in prior trials with either drug as monotherapy.7 If proven to be more effective than single-drug therapy, this therapeutic approach may have important clinical implications for tobacco-dependence treatment. Exploration of combination therapy with existing drugs may provide the best opportunity to advance treatment in the absence of any new pharmacotherapies for tobacco dependence.
To investigate the efficacy of combination pharmacotherapy with varenicline and bupropion SR for smoking cessation, compared with varenicline monotherapy, we conducted a multicenter, randomized, phase 3 clinical trial.
A randomized, blinded, placebo-controlled clinical trial was conducted at Mayo Clinic in Rochester, Minnesota, a Mayo Clinic Health System site in La Crosse, Wisconsin, and the University of Minnesota in Minneapolis between October 2009 and April 2013. The study consisted of a 12-week treatment period with follow-up through week 52. The institutional review boards of Mayo Clinic and the University of Minnesota approved all study procedures. The trial ended when recruitment was achieved and follow-up was completed.
Individuals were eligible to participate if they were at least 18 years of age, smoked at least 10 cigarettes per day for at least 6 months, were motivated to become smoking abstinent, completed written informed consent, and were in good health.
Potentially eligible participants were excluded if they were pregnant, lactating, or likely to become pregnant and not willing to use contraception or had (1) an unstable medical condition; (2) another household member in the study; (3) bupropion or varenicline allergies; (4) current use (previous 30 days) of tobacco-dependence treatment and unable to discontinue use; (5) unstable angina, myocardial infarction, or coronary angioplasty (previous 3 months) or an untreated cardiac dysrhythmia; (6) a history of renal failure or renal dialysis; (7) a history of seizures; (8) current nonspecific suicidal thoughts or lifetime history of a suicidal attempt (ie, “potentially self-injurious act committed with at least some wish to die, as a result of act” [as defined by the Columbia-Suicide Severity Rating Scale8]); (9) a history of closed head trauma with a greater than 30-minute loss of consciousness, amnesia, skull fracture, subdural hematoma, or brain contusion; (10) a history of psychosis, bipolar disorder, bulimia, or anorexia nervosa; (11) current depression (moderate or severe [as determined by a score of ≥20 on the Beck Depression Inventory, Second Edition9]); (12) active substance abuse other than nicotine; (13) current use (previous 14 days) of antipsychotics, monoamine oxidase inhibitors, or drugs with bupropion SR interactions; (14) recent antidepressant dose change (previous 3 months); (15) systolic blood pressure higher than 180 mm Hg or diastolic higher than 100 mm Hg; (16) current treatment (previous 30 days) with another tobacco dependence investigational drug; or (17) current (previous 30 days) bupropion or varenicline use.
The study consisted of a telephone screening call, 11 clinic visits, and 3 follow-up telephone calls (Figure). One follow-up telephone call occurred during the medication phase at the time of the target quit date and 2 calls occurred after the medication phase. Two clinic visits occurred before the medication phase, 6 during the medication phase, and 3 after the medication phase.
For each participant, demographic data, tobacco use history, and self-reported information on race and ethnicity according to National Institutes of Health guidelines and recommendations for federally funded research were collected.10 Smoking dependence was assessed using the Fagerström Test for Nicotine Dependence (score range, 0-10).11
Depressive symptomatology was assessed using the Beck Depression Inventory, second edition.9 The Columbia-Suicide Severity Rating Scale assessed for suicidal ideation or behaviors.8 Both assessments were completed at baseline and weeks 2, 4, 8, 14, 26, and 52.
A central pharmacy randomly assigned study medication in a 1:1 ratio using a computer-generated randomization sequence with variable-sized blocks ranging from 2 to 8 stratified by study site. Study medication was labeled and dispensed according to participant identification, ensuring that treatment assignment remained concealed from the participant, investigators, and all study personnel having participant contact. Following provision of informed consent, participants received randomly assigned medication at the baseline visit.
During clinic visits, participants received brief (≤10 minutes) behavioral counseling,12 and tobacco use status, vitals signs, exhaled-air carbon monoxide (CO) measurements (measured in parts per million [ppm]), and weight were obtained. Participants completed tobacco craving and nicotine withdrawal assessments using a daily diary containing the Minnesota Nicotine Withdrawal Scale, Revised (MNWS-R).13 The MNWS-R consists of items assessing irritability, anxiety, tobacco craving, depressed mood, difficulty concentrating, hunger, impatience, insomnia, and restlessness. Items were rated on a 5-point scale ranging from 0 (not present) to 4 (severe) and reported symptoms for the previous day. Pill counts were conducted at clinic visits and through self reports of missed doses.
Participants were randomly assigned to receive varenicline + bupropion SR (combination therapy); or varenicline + matching bupropion SR placebo (varenicline monotherapy). Medication was started the day after the baseline visit and the target quit date was the eighth day of therapy.
Varenicline, taken orally, was administered in an open-label fashion and dispensed in blister packs. Participants started with a recommended oral dosage of 0.5 mg once daily for 3 days, titrated to 0.5 mg twice daily for days 4 to 7, and then to the maintenance dose of 1 mg twice daily (total, 2 mg/d) for 11 weeks.
Bupropion SR or identical-appearing placebo tablets were dispensed in pill bottles. Bupropion SR was titrated 1 tablet (150 mg) by mouth once per day for days 1 to 3, then 1 tablet by mouth twice per day (total of 300 mg/d) for a total of 12 weeks. Participants receiving placebo escalated dosing in the same fashion.
The primary end point was the biochemically confirmed prolonged and 7-day point-prevalence smoking abstinence rates at week 12. End points were selected using recommended outcomes for tobacco intervention studies.14 A CO level of 8 ppm or less verified self-reported smoking abstinence.15 Point prevalence was defined as CO-confirmed self-reported tobacco cessation in the previous 7 days. Participants who met criteria for CO-confirmed 7-day point-prevalence abstinence at week 12, 26, and 52 visits were defined as meeting criteria for prolonged abstinence if they submitted negative responses to both of the following questions: “Since 14 days after your target quit date, have you used any tobacco on each of 7 consecutive days?” and “Since 14 days after your target quit date, have you used any tobacco on at least 1 day in each of 2 consecutive weeks?” Secondary outcomes were prolonged and 7-day point-prevalence smoking abstinence rates at weeks 26 and 52, tobacco craving and nicotine withdrawal symptoms, and weight changes.
All analyses were performed using the intention-to-treat approach. Smoking abstinence end points at 12 weeks (end of treatment), 26 weeks, and 52 weeks were analyzed separately using logistic regression. For these analyses, smoking abstinence was the dependent variable, treatment group was the independent variable, and study site was a covariate. Participants with missing smoking status information were adjudicated as smoking. A sample size of 250 participants per group was determined to provide statistical power (2-tailed, α = .05) of greater than 80% to detect a difference between treatment groups for the primary end point of prolonged tobacco abstinence at end-of-treatment week 12. Sample size was based on reported smoking abstinence rates in previous trials of varenicline16,17 and a minimum detectible odds ratio (OR) of 1.7 for the comparison of varenicline and bupropion SR vs varenicline and placebo. In addition, we conducted planned exploratory analyses to assess whether treatment effect was moderated by age, sex, baseline smoking rate (<20 cigarettes per day [lighter smokers] vs ≥20 cigarettes per day [heavier smokers]), or level of nicotine dependence (a Fagerström Test for Nicotine Dependence [FTND] score ≤5 indicating a low/moderate level of dependence vs an FTND score of ≥6 indicating a high level of dependence). For each characteristic, logistic regression analyses were performed with treatment, study site, and characteristic included as explanatory variables along with the treatment × characteristic interaction effect. If a significant interaction effect was detected, supplemental analyses were performed to compare treatment outcomes within subgroups defined by the characteristic.
The MNWS-R was completed daily. Tobacco craving was analyzed separately. A baseline score was calculated using MNWS-R data completed prior to starting medication. Scores obtained for the first 16 days after the target quit date were analyzed as change from baseline. Mixed linear models were used with the daily change score as the dependent variable and a lag-1 autoregressive covariance structure used to take into account the clustering of repeated measurements within participants. Models included effects for treatment group, study day, and the treatment × study-day interaction. Analyses were performed using all days for each participant and also using only data collected prior to the first reported tobacco use following the target quit date.
Among participants meeting criteria for prolonged abstinence, weight change from baseline was compared between groups using the 2-sample t test. Frequency of adverse events was compared between groups using the Fisher exact test. In all cases, 2-tailed P values were reported with values less than or equal to .05 considered statistically significant. Adverse events were adjudicated by the investigators. Analyses were conducted using SAS statistical software (version 9.3, SAS Institute Inc).
Of 635 potentially eligible participants who consented, 506 (80%) were randomly assigned to varenicline and bupropion SR (n = 249) or varenicline and placebo (n = 257) (Figure). Overall study completion rates were 62% (315 participants), 63% (158 participants) in the varenicline and bupropion SR group (combination therapy) and 61% (157 participants) in the varenicline and placebo group (varenicline monotherapy). Patients assigned to study groups were similar at baseline (Table 1).
Combination therapy was associated with significantly higher prolonged smoking abstinence rates at 12 and 26 weeks compared with varenicline monotherapy (Table 2). No significant differences were observed in prolonged smoking abstinence rates between the 2 groups at 52 weeks. No significant differences were observed between the 2 groups in the 7-day point-prevalence smoking abstinence rates at any time point.
After more than 16 days following the target quit date, no significant differences in nicotine withdrawal or craving were observed between the 2 groups (mean treatment difference for nicotine withdrawal, 0.04 [95% CI, −0.02 to 0.10; P = .25] and mean treatment difference for craving, 0.05 [95% CI, −0.20 to 0.30; P = .70]). Similar results were obtained when the analysis included only data for days that participants reported abstinence (mean treatment difference for nicotine withdrawal, 0.03 [95% CI, −0.15 to 0.21; P = .74] and mean treatment difference for craving, 0.06 [95% CI, −0.47 to 0.59; P = .81]).
Among participants meeting criteria for prolonged smoking abstinence at the end of treatment (week 12), the mean weight change from baseline to week 12 was significantly less in the combination therapy group compared with the varenicline monotherapy group (1.1 kg [95% CI, 0.5-1.7] vs 2.5 kg [95% CI, 2.0-3.0]; P < .001). At 26 weeks, differences in weight gain were not observed and participants in the combination therapy group gained 3.4 kg (95% CI, 2.5-4.3), and participants in the varenicline monotherapy group gained 3.8 kg (95% CI, 2.9-4.8) (P = .48). At week 52, weight gain from baseline for the combination therapy group was 4.9 kg (95% CI, 3.6-6.2), and for the monotherapy group it was 6.1 kg (95% CI, 4.6-7.6) (P = .23).
Adverse events occurring in at least 2% of one of the study groups are listed in Table 3. Anxiety was reported more commonly with combination therapy than with varenicline monotherapy (7.2% vs 3.1%; P = .04). Depressive symptoms were also reported more commonly with combination therapy than with varenicline monotherapy (3.6% vs 0.8%; P = .03).
During the medication phase or within 7 days of stopping medication, 4 serious adverse events (SAEs) occurred. In the combination therapy group, 1 participant sustained trauma during a motor vehicle collision after receiving medication for 2 months. In varenicline monotherapy, the 3 events included food poisoning, diverticulitis, and breast cancer. No events were adjudicated to be related to study medication.
During follow-up and after medication discontinuation for at least 7 days, 8 SAEs were reported. Five occurred in the combination therapy group and included acute coronary syndrome, deep vein thrombosis complicated by acute coronary syndrome, prostate cancer, a new coronary artery disease diagnosis, and pneumothorax. In the varenicline monotherapy group, 3 SAEs occurred: 1 death due to complications from human immunodeficiency virus 6 months after study drug was discontinued, 1 attempted suicide 9 months after the study medication was completed, and 1 lung cancer. No events were adjudicated to be related to study medication.
Preplanned exploratory analyses were performed to assess potential moderators of the effect of treatment on abstinence. No evidence was observed showing that treatment effects differed according to age or sex (P > .25 for all age-by-treatment and sex-by-treatment interaction effects). However, evidence was observed showing that an effect of treatment on prolonged abstinence at 6 and 12 months was dependent on baseline smoking rate (interaction effect, P = .04 at 6 months and P = .01 at 12 months) and level of nicotine dependence (interaction effect, P = .03 at 6 months and P = .01 at 12 months). From supplemental subgroup analyses, no differences were observed between the 2 groups at any time point for either prolonged or point-prevalence smoking abstinence among lighter smokers (<20 cigarettes per day). However, heavier smokers (≥20 cigarettes per day) receiving combination therapy were more likely to achieve prolonged smoking abstinence at weeks 12, 26, and 52 (Table 4) and 7-day point-prevalence smoking abstinence at weeks 26 and 52. For smokers with low/moderate levels of nicotine dependence (FTND ≤5), no difference in abstinence outcomes were detected at any time point. However, among participants with high levels of nicotine dependence (FTND ≥6), combination therapy was associated with an increased likelihood of prolonged abstinence at weeks 12, 26, and 52, and 7-day point-prevalence abstinence at week 52 (Table 4).
Quiz Ref IDAmong cigarette smokers, the combined use of varenicline and bupropion SR, compared with varenicline alone, resulted in an increase in prolonged smoking abstinence but not 7-day point-prevalence smoking abstinence at 12 and 26 weeks. Neither outcome was significant at 52 weeks. Our observed rates of prolonged smoking abstinence with varenicline monotherapy were consistent with those of previous varenicline studies at all time points.16,17
Quiz Ref IDWe observed a greater attenuation of weight gain at 3 months in participants continuously abstinent from smoking with combination therapy compared with varenicline monotherapy. Meta-analyses have suggested that bupropion SR attenuates postcessation weight gain more than varenicline at the end of treatment.18 In previous trials, mean weight gain with varenicline among smoking-abstinent participants from baseline to 12 weeks was 2.37 kg16 and 2.89 kg,17 and 2.12 kg16 and 1.88 kg17 for bupropion SR. Most weight gain after smoking cessation occurs in the first 3 months,19 and weight gain has been shown in some studies to lead to smoking relapse.20- 23 Combination therapy could provide a clinical option for patients concerned about weight gain and for whom weight gain may undermine smoking cessation in the short term.
Quiz Ref IDAnxiety and depressive symptoms were reported more commonly in combination therapy. In previous smoking cessation studies with varenicline and bupropion SR, no significant increases in anxiety were observed with either varenicline or bupropion SR compared with placebo.16,17 Bupropion SR is known to be associated with anxiety when used in the treatment of tobacco dependence.24 Tobacco withdrawal has also been associated with both anxiety and depressive symptoms.25 All patients being treated with pharmacotherapy for tobacco dependence should be monitored for changes in anxiety and mood, an approach consistent with standard clinical practice.
This study has limited generalizability to the general population of smokers because patients with serious medical and psychiatric illnesses including those with active substance abuse were excluded. For this study, reported abstinence rates and treatment comparisons need to be interpreted with the knowledge that 38% of participants did not complete the study. This may lead to overestimation or underestimation of the true treatment effects. However, drop-out rate was comparable between the 2 groups and comparable with previous trials using varenicline for smoking cessation.16,17 Additional analyses using multiple imputation to accommodate missing data were performed. An underlying assumption of multiple imputation analyses is that missing outcomes are absent at random. Empirical evidence suggests this is not true for smoking cessation studies in which participants who drop out are likely to have relapsed to smoking.26 However, the assumption of this study that all who have dropped out have resumed smoking could underestimate actual abstinence rates. The analyses of the effect of treatment by smoking rate and nicotine dependence level were exploratory and hypothesis generating.
Among cigarette smokers, combined use of varenicline and bupropion, compared with varenicline alone, resulted in an increase in prolonged abstinence but not 7-day point-prevalence at 12 and 26 weeks; neither outcome was significantly different at 52 weeks. Further research is required to determine the role of combination treatment in smoking cessation.
Corresponding Author: Jon O. Ebbert, MD, MSc, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (firstname.lastname@example.org).
Author Contributions: Dr Ebbert had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Croghan, Schroeder, Hurt, Ebbert.
Acquisition of data: Hatsukami, Croghan, Allen, Hays, Hurt.
Analysis and interpretation of data: Hatsukami, Croghan, Schroeder, Hurt, Ebbert.
Drafting of the manuscript: Hatsukami, Croghan, Ebbert.
Critical revision of the manuscript for important intellectual content: Hatsukami, Croghan, Schroeder, Allen, Hays, Hurt.
Statistical analysis: Schroeder.
Obtained funding: Hurt, Ebbert.
Administrative, technical, or material support: Croghan, Hays, Hurt.
Study supervision: Hatsukami, Croghan, Allen, Ebbert.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Ebbert reports serving as an investigator for clinical trials funded by Pfizer, receipt of consultancy fees from GlaxoSmithKline, research support from Pfizer, and research support from Orexigen and JHP Pharmaceuticals outside of the current study. Dr Hatsukami reports receipt of research support from Nabi Biopharmaceuticals outside of the current study. Dr Hays reports serving as an investigator for clinical trials funded by Pfizer. Dr Hurt reports receipt of consulting fees from Pfizer, an unrestricted grant from Pfizer Medical Education Group, and provision of expert testimony in Florida tobacco litigation cases. The other authors report no disclosures.
Funding/Support: The clinical trial was supported by National Institutes of Health (NIH) grant CA 138417 (primary investigator, Dr Ebbert). Medication (varenicline) was provided by Pfizer.
Role of the Sponsors: NIH and Pfizer had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.