Treatment Options in Smoking Cessation: What Place for Bupropion Sustained-Release?

Abstract

Bupropion SR is approved for the treatment of tobacco dependence in adult smokers. Bupropion SR is an atypical antidepressant that has been shown to double the likelihood of quitting smoking (to roughly 19%-24% six months into a quit attempt), perhaps by acting on dopaminergic and noradrenergic systems and by acting as an antagonist of nicotine acetylcholine receptors. Head-to-head comparisons of bupropion SR and other stop-smoking treatments suggest that bupropion SR is as or more efficacious than nicotine replacement therapies, equally efficacious as nortriptyline, and less efficacious than varenicline. The evidence available regarding the effectiveness of bupropion SR in real-world settings suggests that abstinence rates are similar to those seen in controlled clinical trials. Bupropion SR appears to be safe and efficacious for both men and women and for people with comorbid medical or mental health conditions. Evidence collected to date supports the use of bupropion SR as a safe, tolerable pharmacotherapy for smoking cessation among adult smokers without a predisposition to seizures, but also suggests that benefits in terms of abstinence last only as long as treatment continues. This review focuses on recent evidence regarding bupropion SR effects and highlights important questions regarding the duration of effects, relative efficacy, effectiveness in clinical use, mechanisms of action, and utilization of bupropion SR that remain unanswered.

Introduction to Bupropion Sustained-Release

Smoking remains the leading preventable cause of death in the United States ¹ and tobacco dependence continues to be a global epidemic that may take one billion lives in the 21st century. ² Both prevention and treatment are essential tools in efforts to address this public health crisis. Bupropion sustained-release (SR) is an atypical antidepressant medication that is U.S. Food and Drug Administration-approved as a smoking cessation aid and recommended as a first-line pharmacotherapy for smoking cessation. ³ Meta-analyses of randomized, placebo-controlled clinical trials suggest that bupropion SR roughly doubles the odds of maintaining or achieving abstinence six or 12 months post-quit, to approximately 19%-24% when compared with placebo.^3,4 The mechanisms of action of bupropion SR are still not fully understood, and the cognitive and behavioral mechanisms of change that account for bupropion SR effects on abstinence have not been explored thoroughly. In the current review, we will briefly summarize what is known about bupropion SR in terms of its pharmacology, efficacy, effectiveness, safety, acceptance, and uptake by smokers, while highlighting some of the unanswered questions regarding this smoking cessation treatment. We will focus, in particular, on recent trials and will re-examine the place for bupropion SR in smoking cessation treatment, given what is currently known about its short- and longer-term efficacy relative to other treatments including varenicline, its effectiveness in clinical use, and its safety and acceptability. We will close with a brief discussion of directions for future research.

Review of Pharmacology, Mode of Action, Pharmacokinetics of Bupropion SR

Bupropion SR is an extended release formulation of bupropion hydrochloride (amfebutamone) that was initially developed as an antidepressant. Its mechanism of action appears to be inhibited reuptake of dopamine and norepinephrine.^5,6 Bupropion may affect dopamine transport in the nucleus accumbens and striatum^7,8 and may cause reduced firing of neurons in dopaminergic mesolimbic reward circuitry, ⁹ although the clinical significance and loci of bupropion SR effects on dopamine transmission are not fully known. ¹⁰ Reuptake of norepinephrine in the locus coeruleus may result in reducing firing of noradrenergic neurons ¹¹ whose activation is associated with exacerbated withdrawal symptoms.^10,12 In addition, some evidence suggests that bupropion binds to neuronal acetylcholine receptors and acts as a noncompetitive nicotine antagonist.^13,14 Each of these identified mechanisms of action of bupropion SR appears relatively weak and does not fully account for the clinical benefits observed for smoking cessation.^10,15 Doses used in vitro or animal models in vivo sometimes exceed those used for smoking cessation and differences in metabolism of bupropion SR across species complicate interpretation.^10,15 In addition, some research suggests that bupropion may have effects on serotonergic neuron firing mediated by increased noradrenergic activity in the locus coeruleus, and no effect on dopaminergic neuronal firing,^16,17 despite the lack of apparent direct effects of bupropion on serotonin transmission.^5,18 As such, the mechanisms of action responsible for enhanced smoking cessation rates are still not fully understood.

The pharmacokinetics of bupropion SR are better understood, although questions remain about the clinical relevance of the effects of the metabolites of bupropion SR. Metabolism of bupropion and its active metabolites is extensive and takes place in the liver. ¹⁹ Three of the metabolites of bupropion (hydroxybupropion, threohydrobupropion, and erythrohydrobupropion) are pharmacologically active, with hydroxybupropion and threohydrobupropion being the most active.^19,20 Peak values in plasma concentrations of bupropion SR are achieved within three hours of taking the medication, which is little affected by administration with food. ¹⁹ Bupropion and hydroxybupropion have elimination half-lives of 20-21 hours, whereas the other two active metabolites have elimination half-lives of 33-37 hours. Bupropion and hydroxybupropion also appear to have similar potency. Steady state levels of bupropion are achieved within 5 days after starting the drug (8 days for metabolites).^19,21 Johnston et al ¹⁹ provide a thorough review of the pharmacokinetics of bupropion SR in healthy volunteers and those with liver disease. Studies show little in the way of differences in pharmacokinetics as a function of gender, age (except for slightly higher peak plasma concentrations in the elderly), smoking status, or mild- to moderate-hepatic impairment. ¹⁹ A recent study suggests that adolescents metabolize bupropion SR more rapidly than do adults, leading the authors to recommend that this population begin with divided doses until more is known about the effects of bupropion metabolites. ²² Because bupropion SR is metabolized by CYP2B6 and inhibits CYP2D6 activity, ²³ caution should be used when taking bupropion SR with drugs that affect CYP2B6 activity or rely on CYP2D6 activity, including over-the-counter medications such as diphenhydramine. ¹⁹ Based on their review of population pharmacokinetic results, Johnston et al ¹⁹ recommend prescribing 150 mg 1-2 weeks prior to a cessation attempt and then increasing to 300 mg (typically 3-7 days later), as this dose appears to best predict successful cessation and therapeutic plasma concentrations.

Efficacy Studies

Recent meta-analyses have summarized the results of approximately two dozen diverse randomized, controlled trials of bupropion SR efficacy.^3,4 Trials included in meta-analyses were diverse in terms of target smoker population, length of treatment (with a minimum treatment period of seven weeks), abstinence outcome measures, funding source, and length of follow-up. These meta-analyses suggest that bupropion SR doubles the odds of quitting successfully, relative to placebo (Odds Ratio [OR] = 2.1, 95% Confidence Interval [CI] = 1.8-2.4 in Hughes et al; ⁴ OR = 2.0, 95% CI = 1.8-2.2 in Fiore et al), ³ such that roughly 24.2% (95% CI = 22.2%-26.4%) of those receiving active bupropion SR achieve abstinence six months post-quit. ³ Another recent meta-analysis of 16 randomized controlled clinical trials supports these estimates and suggests that bupropion SR treatment increases the odds of achieving abstinence for at least six months by an odds ratio of 2.1 (95% Credible Interval in Bayesian meta-analysis = 1.7-2.6). ²⁴ Wu and colleagues ²⁵ estimated the effect of bupropion SR on both short- and longer-term abstinence rates in a meta-analysis of 12 studies and found that bupropion SR increased the odds of achieving abstinence by an odds ratio of 2.1 (95% CI = 1.7-2.6) three months post-quit and by an odds ratio of 1.56 (95% CI = 1.1-2.2) at 12 months post-quit. Note that these meta-analyses differ slightly in the selection criteria used and the meta-analytic techniques applied, yet still converge on highly similar estimates of bupropion SR efficacy relative to placebo. These meta-analyses and additional critical and systematic review articles that evaluate the role of bupropion SR as a smoking cessation pharmacotherapy have already been published.^17,25,26 As such, the present article will provide a brief summary of evidence regarding the efficacy of bupropion SR in samples of healthy smokers using intent-to-treat analyses, with particular emphasis on recently published randomized clinical trials, and will then briefly discuss evidence regarding bupropion SR efficacy in specific populations, effectiveness, and mediators of effects on abstinence.

Trials of efficacy in healthy adult daily smokers

The initial industry-sponsored report on bupropion SR described the efficacy of bupropion SR, compared to placebo, among adult daily smokers without current depression who were in good health and had no prior use of bupropion or contraindications to bupropion use. Results indicated that both 150 mg and 300 mg daily doses of bupropion SR significantly increased biochemically verified 7-day point-prevalence abstinence through one year post-quit in intent to treat analyses. ²⁷ Subjects meeting similar criteria and passing screening for contraindications for transdermal nicotine patch use were randomized to nicotine patch, bupropion SR, bupropion SR plus patch, or placebo pill and patch in Jorenby et al. ²⁸ Results of this trial indicated that bupropion SR treatment significantly increased point-prevalence and continuous abstinence rates through one-year post quit and that both bupropion conditions were superior to the patch only condition through the first year. Adding the patch to bupropion SR treatment did not significantly improve abstinence rates. Gonzales et al ²⁹ later showed that bupropion SR improved abstinence rates relative to placebo for six months among people who had previously used bupropion SR without abstinence success. Recently, Piper and colleagues ³⁰ reported on an NIH-sponsored randomized clinical trial that compared conditions in which smokers received active bupropion SR and active nicotine gum, active bupropion SR and placebo gum, or both placebo pills and placebo gum and found that bupropion SR improved point-prevalence abstinence rates, relative to the double placebo condition up to six months post-quit, but not at one-year. Gum did not significantly improve abstinence rates in those taking bupropion SR. ³⁰

Recent trials testing bupropion SR and varenicline separately against placebo and one another have supported the efficacy of bupropion SR in promoting abstinence, relative to placebo. Whereas Gonzales et al ³¹ reported that both point-prevalence and continuous abstinence rates were significantly higher among those receiving bupropion SR than placebo at both six months and one year post-quit, Jorenby et al ³² found that bupropion SR was superior to placebo in terms of point prevalence abstinence at all time points, but only significantly improved continuous abstinence rates through six months post-quit. The benefits of bupropion SR in promoting continuous abstinence was even more short-lived in Nides et al ³³ who reported that bupropion SR was superior to placebo at 12-weeks post-quit but not at 24 or 52 weeks post-quit. Bupropion SR also failed to improve abstinence rates relative to placebo in a sample of mostly male smokers given a two-month course of open-label nicotine patch treatment and a three-month course of cognitive behavioral therapy. ³⁴

In all of the trials cited above, all medications and placebo were offered in the context of standard, brief smoking cessation counseling or more intensive psychotherapy. ³⁴ Hall and colleagues ³⁵ tested bupropion SR and nortriptyline (a tricyclic antidepressant) against placebo in the context of minimal medical management or a fairly intensive group psychotherapy condition to determine whether bupropion SR efficacy was dependent on concurrent support. Results of this randomized controlled trial indicated that bupropion SR and nortriptyline have significant benefits relative to placebo in terms of point-prevalence abstinence, but not continuous abstinence, and indicated that psychological intervention did not augment abstinence rates in the active medication conditions, relative to the medical management control group. As such, bupropion SR efficacy was not moderated by smoking cessation counseling condition. A recent randomized clinical trial using a fully crossed factorial design of bupropion SR and eight 10-minute sessions of individual counseling similarly found that bupropion SR effects were independent of counseling. ³⁶ Specifically, results indicated that bupropion SR increased abstinence rates during the eight weeks of treatment, but not at six or 12 months post-quit, whether offered with or without counseling. Counseling did not significantly improve abstinence rates in the context of active bupropion SR or placebo medication treatment. ³⁶

All of the studies described above took place in the US with samples that were mostly Caucasian. Analyses regarding bupropion SR are rarely stratified by race or ethnicity in trials that include samples of diverse races and ethnicities. ³⁷ One trial in African-American smokers, however, indicated that bupropion SR treatment significantly improved abstinence rates through six months post-quit when compared to placebo. ³⁸ Randomized controlled trials in other countries (or in multiple countries) have supported the efficacy of bupropion SR in promoting abstinence, relative to placebo, up to six months or one year post-quit.^39–41 Another multi-country trial reported that bupropion SR only resulted in statistically significant increases in abstinence rates relative to placebo while treatment was ongoing and not beyond, however. ⁴²

In summary, randomized, placebo-controlled clinical trials of generally healthy smokers strongly support the efficacy of bupropion SR in promoting abstinence from smoking while treatment is ongoing (usually 7-12 weeks). Evidence regarding lasting benefits of bupropion SR treatment is mixed. Recent trials published since meta-analyses were conducted suggest that bupropion SR benefits may dissipate by six or 12 months post-quit (depending on the length of treatment).

Relapse prevention efficacy

To address the drop-off in abstinence that seems to follow the end of treatment, several trials have examined whether extended treatment with bupropion SR prevents relapse among those who initially achieve abstinence. In 2001, Hays and colleagues provided open-label bupropion SR to a sample of daily smokers motivated to quit for seven weeks and then randomly assigned these subjects to an additional 45 weeks of bupropion SR treatment or placebo. All subjects received smoking cessation counseling and were followed for a year following the end of treatment. Extended treatment with bupropion SR increased biochemically verified point-prevalence abstinence rates during treatment (to 55.1% relative to 42.3% in the placebo group) and six months later (41.1% in the bupropion SR condition, 37.7% in the placebo condition), but not a full year after the end of extended treatment (41.6% in the bupropion SR condition, 40.0% in the placebo condition), when abstinence rates were exceptionally high in both conditions. Extended bupropion SR treatment did not significantly improve continuous abstinence rates over those achieved with placebo beyond six months post-quit in this trial. ⁴³ A re-analysis of these data that examined bupropion SR effects on prolonged abstinence (which allows “slips” but not returns to smoking seven days between clinic visits or nine days out of 365) found a similar pattern; bupropion was associated with significantly higher prolonged abstinence rates through nine-months post-quit, but only marginal effects thereafter. ⁴⁴ Others have found that bupropion SR seems to prevent progression from a smoking lapse to a full-blown relapse, defined as smoking at least seven days in a row during the standard brief treatment period.^45,46 Another trial of a six-month course of bupropion SR vs. placebo following an eight-week open-label course of nicotine patch treatment failed to show any benefit of bupropion SR for relapse prevention among those who quit successfully with the patch. ⁴⁷ A standard eight-week course of bupropion SR also failed to improve cessation rates among those who failed to maintain abstinence using the patch. ⁴⁷ Similarly, Killen and colleagues ⁴⁸ recently reported that an additional 14 weeks bupropion SR treatment conferred no benefit compared to placebo following open-label treatment with bupropion SR for 11 weeks, nicotine patch for 10 weeks, and 180 minutes of one-on-one relapse prevention training. Likewise, extending bupropion SR treatment for an additional nine months in people who quit successfully with a three-month active bupropion SR course did not improve abstinence rates at one-year post-quit or beyond, relative to extended placebo treatment. ⁴⁹ Another recent trial of maintenance therapy with bupropion between weeks eight and 24 post-quit demonstrated that those who received active bupropion SR and placebo gum (but not those who received active bupropion SR and active gum) were significantly less likely to relapse at the end of treatment, but were only marginally significantly less likely to relapse at 48 weeks post-quit. ⁵⁰

In summary, data on prolonged treatment with bupropion SR seem to suggest that extending treatment can forestall relapse while treatment is ongoing, but treatment may not result in lasting benefit when treatment is discontinued. In this way, results of the relapse prevention trials are generally consistent with the 7-12 week treatment trials reviewed above.

Comparative efficacy

Randomized controlled clinical trials have compared bupropion SR to transdermal nicotine patch, nicotine gum, nicotine inhaler, nortriptyline, and varenicline. Jorenby et al ²⁸ showed that bupropion SR combined with placebo patches resulted in superior point-prevalence abstinence rates at six months and one year (34.8% and 30.3%, respectively) relative to placebo pills and active nicotine patch treatment (21.3% and 16.4%). Simon et al, ³⁴ in contrast, found that equally high proportions of smokers assigned to bupropion SR (40%) or placebo pill (42%) achieved point-prevalence abstinence six months post-quit in a trial in which all subjects received nicotine patches and individual smoking cessation counseling. Results at 12 months post-quit similarly showed no benefit of bupropion SR in terms of point-prevalence abstinence (22%) versus nicotine patch and placebo pill treatment (28%) in intent-to-treat analyses. The recent update of the U.S. Public Health Service clinical practice guideline ³ compared abstinence rates in 32 nicotine patch treatment arms and those in 26 bupropion SR treatment arms, using conservative alpha levels, and found that bupropion was not significantly better than patch (OR = 1.0, 95% CI = 0.9-1.2). The Fiore et al ³ results must be interpreted cautiously as this analysis collapses over diverse studies that typically did not include both patch and bupropion arms.

One comparison of maintenance gum and/or bupropion SR treatment ⁵⁰ indicated that abstinence rates were similar in subjects who received bupropion SR and placebo gum (55% at the end of treatment 24 weeks post-quit and 32% at 48 weeks post-quit) and those who received placebo pills and active nicotine gum (46% at 24 weeks post-quit and 26% at 48 weeks). Thus, bupropion SR has not been shown to be superior to nicotine gum treatment.

A comparison of three months of bupropion SR and nicotine inhaler treatment ⁴⁹ indicated that bupropion SR resulted in significantly higher point-prevalence abstinence at the end of treatment (26%) than did nicotine inhaler (14%), although continuous abstinence rates were very similar and low (8% vs. 7%). The only trial comparing bupropion SR and the inhaler provided some support for the superiority of bupropion SR in achieving abstinence when early slips were allowed.

Two randomized clinical trials compared bupropion SR with the atypical antidepressant nortriptyline.^35,41 Neither found any significant difference in abstinence rates as a function of medication assignment. Comparisons of bupropion SR with varenicline, in contrast, indicate that a regimen of 2 mg of varenicline daily is associated with a significant increase in the likelihood of abstinence relative to 300 mg of bupropion SR.^31–33 A recent meta-analysis ²⁴ found that varenicline increased the odds of quitting by an odds ratio of 2.18 (95% Credible Interval = 1.09-4.09) over bupropion SR. In summary, bupropion SR appears to have effects similar to nortriptyline and inferior to 2 mg of varenicline. The varenicline results should be interpreted with caution, as bupropion SR tended to underperform in the varenicline trials (OR = 1.7 relative to placebo), relative to its usual effect size (OR = 2.1), as Hughes et al ⁴ pointed out in a recently updated Cochrane review. The superiority of varenicline in these pre-FDA approval trials may thus reflect underperformance of bupropion SR rather than or in addition to truly superior performance of varenicline. Additional, independent replications in diverse populations of smokers would help to clarify the robustness and magnitude of varenicline vs. bupropion SR differences.

Efficacy in specific populations

Randomized trials have examined bupropion SR efficacy in African-American adults; nonpregnant women; adolescents; and people with chronic obstructive pulmonary disease (COPD), cardiovascular disease, major depression, alcohol dependence, schizophrenia, and specific genotypes. As noted above, bupropion SR has been shown to boost abstinence rates through six months post-quit in a sample of African-American smokers. ³⁷

A meta-analysis of 12 placebo-controlled bupropion SR treatment arms suggests that gender does not interact with bupropion SR and that men and women respond equally well to this medication, at least through the end of treatment. ⁵¹ Gonzales et al ⁵² also reported no gender difference in bupropion SR efficacy during a relapse prevention phase (lasting from 7 to 52 weeks post-quit). Some have argued that bupropion SR equalizes abstinence rates for women and men, as women tend to have lower abstinence rates than do men in other treatment conditions. ⁵³ Randomized, controlled trials of bupropion SR have excluded pregnant women. One study, however compared smokers who had been exposed to bupropion SR during the first trimester of pregnancy with pregnant women who were not taking the medication. Subjects in the two groups were matched for cigarettes smoked per day and nicotine dependence level. Results indicated that those taking bupropion SR were significantly more likely to quit than those who did not. ⁵⁴

A recent randomized, placebo-controlled trial ⁵⁵ of 150 mg or 300 mg of bupropion among smokers aged 14 to 17 years indicated that point-prevalence abstinence rates were significantly higher in the 300 mg condition relative to placebo while treatment was ongoing but only slightly higher 26 weeks post-quit (13.9% vs. 10.3%), consistent with results from trials with adult smokers. Another study of adolescent smokers indicated that adding bupropion SR to patch therapy does not significantly improve abstinence rates at 10 or 26 weeks post-quit. ⁵⁶

Smokers with comorbid medical or mental health conditions have also participated in bupropion SR randomized trials. Successfully treating smokers with depression, alcohol dependence, and schizophrenia continues to be a challenge, but some evidence suggests that these populations respond well to bupropion SR treatment. A sample of smokers with a variety of cardiovascular conditions (and some with comorbid COPD or diabetes) showed significantly higher abstinence rates at six and 12-months post-quit when treated with bupropion SR versus placebo. ⁵⁷ Evidence from another study suggests that bupropion SR significantly improves continuous abstinence between 4 and 26 weeks post-quit among those with COPD, but not those at risk for COPD (not yet diagnosed), and among those with depressive symptom scores suggestive of moderate to severe depression. ⁵⁸

A trial of bupropion SR in another sample of COPD patients similarly showed a benefit of treatment relative to placebo. ⁵⁹ Secondary analyses of general clinical trials have shown that those who enter these trials with a history of depression (but not current depression) are equally as likely to respond to bupropion SR treatment as those without a history of depression.^53,60 A trial of bupropion SR and cognitive behavioral therapy for depression (relative to placebo and standard cessation counseling conditions) indicated a similar benefit for bupropion SR among those with a history of major depression or elevated depressive symptoms (both of which were rare in this sample) and those without such risk factors. ⁶¹ There was no significant interaction between depression status and medication condition in this trial. Likewise, history of alcoholism appeared not to interact with bupropion SR in the Hurt et al ²⁷ trial, although there were few cases of alcoholism identified in this sample. ⁶⁰ A study of 58 (mostly male) smokers seeking various levels of care for alcohol dependence failed to detect a benefit of bupropion SR relative to placebo, however. ⁶² Among individuals with schizophrenia, randomized clinical trials of bupropion SR added to nicotine replacement therapy and smoking cessation counseling suggest that bupropion SR may improve abstinence rates while treatment is ongoing,^63,64 but perhaps not beyond the end of treatment.

In summary, randomized clinical trials comparing bupropion SR to placebo in specific populations of smokers suggest that the populations studied to date respond in a similar manner as the general population of smokers studied initially. That is, there is evidence to suggest that bupropion SR increases abstinence rates during treatment, but may not have lasting benefits beyond the end of treatment for all the groups studied to date.

Some recent studies have suggested that genotype may moderate bupropion SR efficacy. For example, recent trials have shown that people who rapidly metabolize nicotine (which is related to CYP2A6 genotype) benefit from bupropion SR whereas slower metabolizers do not ⁶⁵ and that people who metabolize bupropion more slowly (which is related to CYP2B6 genotype) benefit from bupropion SR whereas rapid metabolizers of bupropion SR do not. ⁶⁶ Other studies have examined the role dopaminergic and other genotypes play in moderating responses to bupropion SR. Although some trials have found evidence of interactions between bupropion SR condition and genes related to catechol-O-methyltransferase gene variants, ⁶⁷ dopamine receptors and dopamine transporters,^68,69 or genes related to dopamine receptors and bupropion metabolism, ⁷⁰ these results are still considered preliminary pending replication. ⁶⁹

Effectiveness

In addition to the clinical trials reviewed above, studies of bupropion SR in health care and employment settings have also been conducted. Holmes and colleagues ⁷¹ published a review of several effectiveness studies documenting (often self-reported) abstinence rates among those using bupropion SR to stop smoking and found that six- and 12-month abstinence rates were similar to those observed in the active medication conditions in randomized clinical trials. For example, Swan and colleagues ⁷² found that over 25% of smokers who were prescribed 300 mg of bupropion SR achieved self-reported point-prevalence abstinence 12 months post-quit. Another prospective study of abstinence among those who presented new prescriptions for bupropion SR for smoking cessation at community pharmacies in British Columbia found a similar 12-month point-prevalence abstinence rate of 21% (verified by CO testing) in this self-selected sample of bupropion SR users. ⁷³ A similar study of 227 general practice patients seeking smoking cessation services reported CO-confirmed continuous abstinence rate of 22% from 8 to 52 weeks post-quit. ⁷⁴ A recent randomized placebo-controlled trial of bupropion SR with brief counseling among Italian primary care patients found that over 25% of those receiving active bupropion SR achieved CO-verified continuous abstinence one year post-quit, compared to 14% of those receiving placebo. ⁷⁵ As such, the data on bupropion SR effectiveness in real-world use are consistent with the abstinence rates reported in more tightly controlled clinical trials with carefully screened participants, although only one randomized effectiveness clinical trial has been published to date.

Mechanisms of change

Relatively little attention has been paid to the question of how bupropion SR improves abstinence on a psychological or behavioral level. Basic animal and human laboratory research has suggested withdrawal, craving, and positive affect as mechanisms of change affected by bupropion SR. Animal research suggests that bupropion both alleviates withdrawal and reduces the inflation of reward thresholds (anhedonia) that occurs in withdrawal. ⁷⁶ In a laboratory study of smokers who abstained from smoking for three-day stays on a closed research ward, without intending to quit long-term, Shiffman et al ⁷⁷ reported that bupropion SR reduced negative moods, difficulty concentrating, and anhedonia, but not craving. In contrast, Teneggi and colleagues ⁷⁸ found that bupropion SR improved craving, but not other symptoms of withdrawal.

Although early clinical trials presented secondary analyses of withdrawal, craving, and weight gain outcomes among those treated with bupropion SR and placebo, formal mediation analyses that preserve the temporal ordering of the putative mediators and smoking outcomes have been rare. Several clinical trials have shown that bupropion SR attenuates increases in withdrawal symptoms, including negative affect and craving to smoke, during the first several weeks of a quit attempt, relative to placebo^{6,26,27,31,32,39,79,80} although some studies have failed to detect bupropion SR effects on withdrawal symptoms.^{33,38,41,48,81} To date, however, only three formal mediation analyses have examined whether changes in affect, craving, or composite withdrawal symptoms mediate the effects of bupropion SR on abstinence.

In the first mediation study, Lerman and colleagues ⁸² reported that change in negative affect upon quitting partially mediated the effect of bupropion SR on abstinence eight weeks post-quit. Lerman et al ⁸² did not find evidence to suggest that bupropion SR improved positive affect or that withdrawal (which was improved by bupropion SR) predicted abstinence. The second mediation analysis ⁸³ reported that bupropion SR effects on abstinence at the end of treatment were partially mediated by decreases in a composite of withdrawal symptoms and in ratings of cravings to smoke during the first week of the quit attempt. A mediation analysis of a third randomized clinical trial of bupropion SR ⁸⁴ found that active medication effects on early abstinence (one month post-quit) were partially mediated by improvements in positive affect on the quit date, momentary ratings of cravings to smoke during the first week post-quit, and average levels of motivation to quit and self-efficacy related to quitting during the first week post-quit. Tests of estimated mediated effects in Piper et al ⁸³ and McCarthy et al ⁸⁴ were statistically significant using state of the art measures of mediation. ⁸⁵

Another candidate mediator of bupropion SR efficacy is reduced weight gain. Evidence regarding bupropion SR on cessation-related weight gain suggests that bupropion SR at best delays weight gain until the end of treatment but does not significantly reduce weight gain.^{27,28,38,39,42} Temporary delays in weight gain may be enough to account for bupropion SR efficacy, as medication effects tend to occur early in the quit attempt, while treatment is ongoing and not beyond (see above), but this hypothesis has not been tested directly.

In summary, although candidate mediators of bupropion SR have been identified, hypotheses regarding the cognitive, affective, and behavioral mechanisms of change that account for bupropion SR effects on abstinence have not been thoroughly tested. Results of the tests conducted to date are mixed, but provide some, albeit inconsistent, support for mediation through negative affect, craving, positive affect, motivation, and self-efficacy.

Safety and Tolerability

Extensive data have been collected regarding the safety and tolerability of bupropion, in both clinical trials and post-marketing studies since the drug was approved in the U.S. for treatment of depression in 1989 and for smoking cessation in 1997. Data suggest that bupropion SR is safe and well tolerated, even in people with cardiovascular disease or pulmonary disease. ⁸⁶ Randomized placebo-controlled clinical trials identified the following adverse events that occurred more frequently in the active medication condition than in the placebo condition: insomnia,^{28,31–33,39,40} dry mouth,^{28,31–33,38,39} and diarrhea. ⁴¹ Dose decreases may reduce these side effects. ⁸⁷ Other common side effects, including headache and nausea, occur in similar rates in the placebo and active medication conditions in the trials reviewed above.

Although adverse events were common in both active and placebo conditions, drug discontinuation was generally rare in clinical trials (ranging from 8% to 15%; e.g.^31,48 Early trials of the immediate release formulation of bupropion hydrochloride detected small increases in the risk of seizures. ⁸⁸ Rare seizures (0.1% incidence) have been linked to bupropion SR use for smoking cessation in people without a history of seizure or major risk factors for seizure such as a history of head injury or eating disorder. ³¹ Some have argued that this seizure risk is comparable to that conferred by use of other antidepressants. ¹⁷ Other rare, serious adverse events that were deemed to be bupropion SR-related in smoking cessation trials included cholecystitis, urticaria, and angiodema.^30,39

Post-marketing research supports the safety and tolerability of bupropion SR for the treatment of depression and tobacco dependence, but also suggests that hypersensitivity reactions (serum-sickness-like or anaphylactic-like reactions) occur in approximately 0.1% of people taking bupropion SR. ⁸⁶ Post-marketing research suggests that hypertension risk increases when bupropion SR is offered in conjunction with nicotine patch therapy. Blood pressure monitoring is recommended for combination pharmacotherapy. ⁸⁶ In addition, post-marketing surveillance documents cases of ischemia, psychiatric changes (such as suicidal ideation), and death that have occurred in those using bupropion SR, but the role of bupropion SR in these events is not fully known. ⁸⁷ The U.S. Food and Drug Administration has warned users about potential suicidal thought and behavior risks possibly associated with antidepressant use, although the basis for this recommendation is not clear. Post-marketing surveillance in England suggested that constipation, anxiety, dyspepsia, rash, and itching, all expectable reactions, were more likely during use of bupropion SR than following use among those monitored by a general practitioner. ⁸⁹ No differences in all-cause or condition-specific mortality were observed in this surveillance study. ⁸⁹

Bupropion SR is not recommended for use during pregnancy due to concerns about potential drug effects on the fetus. One prospective study compared women exposed to bupropion SR during the first trimester of pregnancy with matched samples of women exposed to other antidepressant medication and women not exposed to teratogens. ⁹⁰ Results indicated that the rate of spontaneous abortions was significantly higher in women taking bupropion SR, for smoking cessation or depression, than in women in the non-teratogen group, even after matching the samples for smoking status, alcohol use, maternal age, and gestational age at recruitment (when available). Rates of major malformations, stillbirths, and neonatal death were similar across the bupropion SR and non-teratogen groups. Birth weight and gestational age at birth also did not differ.

Patient Focused Perspectives

Bupropion SR is prescribed fairly frequently, but the demand for it as a smoking cessation aid, patient satisfaction with it, and adherence to recommendations regarding use are still largely unknown. Smoking cessation aids are used by a minority of smokers, generally, ⁹¹ and this is true for bupropion SR as well. Bupropion SR was prescribed under the trade name Zyban (NC, US) 383,870 times in the U.S. in 2005, which equates to roughly three prescriptions for every 10,000 ambulatory care visits. ⁹² Bupropion SR was prescribed an additional 1,090,773 times as Wellbutrin SR in 2005 and Wellbutrin XL, the once-daily extended release formulation of bupropion was prescribed 3,444,972 times in 2005. ⁹² Past surveillance reviews found that roughly 28% of bupropion SR prescriptions between 1997 and 2001 were for smoking cessation rather than depression. ⁸⁷ As such, some of the 2005 drug mentions for Wellbutrin SR or XL may have been prescribed for smoking cessation.

A study of prescribing practices in General Medical Services in Ireland found that the rate of prescribing bupropion SR for smoking cessation fluctuated between 3 and 6 per 1,000 patients between September 2000 and March 2001, but then declined precipitously to one per 1,000 in April 2001, when media reports of deaths among people using bupropion SR were circulated in Ireland, concerns about seizure risk and drug interactions were rising, and nicotine replacement therapies were made available to patients at no cost. ⁹³ Prescription rates remained low throughout 2001. The average length of these prescriptions was one-month, below the minimum seven weeks recommended. ⁹³ These data suggest that bupropion SR may be less acceptable to smokers and prescribers than are nicotine replacement therapies, and that bupropion SR may be underused when prescribed. Results from surveillance in Australia suggest a similar decline in bupropion SR prescribing and use. ⁹⁴ In 2001, when bupropion SR was made available at affordable prices through the Pharmaceutical Benefits Scheme, over 350,000 prescriptions were processed, but this declined sharply to below 100,000 in 2002 and stayed at or below 100,000 through 2005. The authors estimate that this decline corresponds to a decline in the proportion of smokers filling prescriptions for bupropion SR from a high of 11% in 2001 to roughly 3.6% in 2005, whereas nicotine replacement therapy prescriptions increased during the same period. ⁹⁴

In contrast, data from patients visiting general practitioners in England indicated that 134 (56%) of 239 heavy daily smokers who had not used bupropion SR before and were medically eligible accepted an invitation to make a quit attempt with the assistance of bupropion SR and support from a nurse. ⁹⁵ Of these, 54 (40%) completed all the steps necessary to make a quit attempt. As such, 54 of 239 (22%) eligible smokers accepted bupropion SR and made a quit attempt in this sample of primary care patients in England. In the U.S. Veterans Administration Services system, in contrast, health services data suggest that roughly 25% of the 278,129 smokers who received prescription smoking cessation aides between 1999 and 2002 were prescribed bupropion SR, which represents only 2% of the estimated 4,088,807 smokers served during that period. ⁹⁶ Researchers in the VA system have shown, however, that making an on-call smoking cessation counselor available can boost bupropion SR prescription rates to 17% of smokers. ⁹⁷ As such, uptake of bupropion SR seems to differ across time and setting in important ways and acceptance rates tend to be low overall.

Adherence to recommended bupropion SR dosing regimens has not been fully assessed or reported in many of the clinical trials that established the efficacy of bupropion SR. Although most clinical trials report some form of drug discontinuation rates, and these tend to be low, as reviewed above, pill counts and medication event monitoring reports are less common. Self-reported adherence to the twice daily dosing regimen used in McCarthy et al ³⁶ was excellent. All subjects in this trial were reminded by an electronic diary to take their pills in the morning upon waking and in the evening before bedtime and were asked to confirm when they had done so. Subjects reported that they took over 90% of both morning and evening doses. This may be an overestimation, however, as a study using medication event monitoring systems (MEMS) to monitor pill bottle openings with a microchip indicated that only 75% of doses were taken by women who received feedback and counseling about adherence and only 54% by women receiving usual-care counseling. Rates of taking bupropion SR at the proper times were even lower (54% in the enhanced feedback condition and 32% in usual care; ⁹⁸ see Schmitz et al ⁹⁹ for a similar study with a high attrition rate). As such, levels of adherence to bupropion SR use recommendations are not well documented in clinical trials or routine clinical use.

Conclusions

Evidence suggests that bupropion SR has an important place in the treatment of tobacco dependence in adult smokers as a treatment that is equivalent or superior to nicotine replacement therapies, perhaps especially for women and those with a history of depression. ⁵³ Although varenicline showed superior efficacy in the trials conducted to date, bupropion SR may be a good alternative to offer smokers who have contraindications to varenicline use (e.g. severe renal impairment, history of nausea or other side effects with past use) or who decline to use it. Given the evidence that bupropion SR benefits may end when treatment ends, however, it may be important to offer effective relapse prevention treatments toward the end of the course of bupropion SR treatment. The safety and tolerability of bupropion SR are well-established, although post-marketing surveillance is needed to continue to monitor adverse events that may be rare or may be more likely to occur in broad populations of smokers who may differ from volunteers selected for clinical trials in important ways (e.g. medical comorbidities).

Meta-analyses of well-designed and carefully conducted randomized, controlled clinical trials show that bupropion SR doubles placebo abstinence rates to result in approximately 19%-24% abstinence rates six-months post-quit. Such double-blinded studies and meta-analyses are considered the gold standard of evidence regarding efficacy. It is important to acknowledge, however, that the methodology is not immune to problems. Blinding, for example, may be compromised by side effects, withdrawal distress, or success in quitting. ¹⁰⁰ A recent analysis of blind integrity in a randomized controlled trial of bupropion SR found that people in the active medication condition were twice as likely to guess that they received bupropion SR than were those taking placebo (who were twice as likely to say they received placebo than were those in the active condition) and accuracy of condition identification was related to success in quitting. ¹⁰⁰ Compromised blinding for either subjects or researchers could have a substantial impact on study results and interpretation. Likewise, problems in assessing treatment adherence and smoking cessation outcomes complicate interpretation of data from randomized clinical trials. Even the most sophisticated measures of medication adherence (e.g. pill counts, MEMS devices, plasma drug concentrations) are still somewhat error prone. As such, being randomly assigned to treatment is not necessarily equivalent to receiving treatment.

The choice of abstinence outcome and verification of claims of abstinence may also influence outcomes in important ways. As reflected in the review above, bupropion SR did not always have uniform effects across different measures of abstinence (i.e. point prevalence, prolonged, continuous) and means of obtaining abstinence information (i.e. self-report, collateral verification, CO or cotinine testing). Meta-analyses can help to identify the variability associated with methodological factors, if there is sufficient variability in the measures administered and results reported. To date, results of meta-analyses of bupropion SR as a smoking cessation aid have been remarkably consistent, however.

Methodological issues become even more complex as researchers strive to answer additional questions beyond that of basic efficacy and conduct sub-group, mediation, or effectiveness analyses. The complexity inherent in designing and conducting appropriate studies to address such questions may help account for the scarcity of information currently available regarding the efficacy of bupropion SR in subpopulations of smokers identified by demographic factors such as race or ethnicity, by tobacco dependence levels, or physical or mental health status. In order to address critical questions about how well, for whom, by what mechanisms, under what circumstances, and at what risk bupropion SR works in treating tobacco dependence, however, such research must be conducted.

Directions for future research

Research in the following areas may better establish the place of bupropion SR in the changing landscape of smoking cessation treatment options. First, the relative efficacy and relative effectiveness of bupropion SR should continue to be investigated in large-scale trials that will provide adequate statistical power to detect clinically meaningful differences among active treatments (e.g. whether the combination of bupropion SR and NRT is better than either alone) and to detect treatment by individual difference interactions (e.g. with gender, depression history/status, race/ethnicity, dependence level, genotype). Such work is currently underway^101,102 but needs to be extended and replicated in order to establish an evidence base for making specific treatment and treatment matching recommendations for smokers interested in quitting. A specific priority will be for independent research teams to replicate the superiority of varenicline and the value of adding bupropion to NRT, as results that do not support the relative efficacy of bupropion SR may suggest that it should be downgraded to a second-line treatment in the future, given its more stringent contraindications and slightly higher risks, relative to NRT in particular.

Future clinical trials may also generate novel and important knowledge by improving and diversifying their methodology. More detailed assessment and reporting of adherence may facilitate interpretation of results, for example. In addition, examining bupropion SR effects on survival and the probabilities of key transition events (i.e. from abstinence to lapse, from lapse to relapse) may generate a more fine-grained understanding of what bupropion SR does and how this could be best used clinically. To date, results suggest that bupropion SR works while treatment is ongoing and for a short period thereafter. Bupropion SR does not appear to have sustained relapse prevention effects, but seems to provide temporary benefits (e.g. in craving reduction and the maintenance of positive affect, motivation, and self-efficacy) that help smokers establish and maintain short-term abstinence and perhaps also recover from lapses. Greater attention to the mechanisms of change activated by bupropion SR and other treatments may further suggest a rational basis for matching individual smokers to treatment and for combining treatments in complementary ways. As noted above, this area of research has been largely neglected in previous bupropion SR research but has the potential to yield important information about what bupropion SR does and does not do to support abstinence.

A second priority with bupropion SR and all other smoking cessation treatments is to better understand why they are underutilized and how best to promote the dissemination of treatments by providers and promote uptake and adherence among smokers. The surveillance data regarding use of bupropion SR to date are disappointing, but research also suggests relatively simple interventions that can promote treatment acceptance. ⁹⁷ If results of clinical trials continue to suggest that varenicline is superior to bupropion SR and that bupropion SR confers little benefit as an alternative to or adjunct to nicotine replacement therapies, the place for bupropion SR in clinical practice may be narrowed so that bupropion SR is offered as a cessation aid to those who refuse or respond poorly to varenicline and nicotine replacement therapies, rather than as a first-line treatment. Dissemination policies will thus hopefully reflect our growing understanding of the changing place of bupropion SR in smoking cessation treatment.

Conflict of Interest Statement

Douglas E. Jorenby has received research support from the National Institute on Drug Abuse, Nabi Biopharmaceutical, and Pfizer, Inc., and consulting fees from Nabi Biopharmaceutical.