Are varenicline (Chantix) and bupropion safe for smoking cessation in persons with mental disorders?

Even in 2018, there are large numbers of people who still smoke. This, of course, leads to multiple serious, bad health outcomes.

This is even more important for us as mental health clinicians because, overall, persons with mental disorders are more likely to smoke than others.

The problem

In addition to nicotine replacement products, other medications that are known to help with smoking cessation are varenicline (brand name Chantix®) and bupropion (brand name Zyban®).

I got a call today from a cardiologist I know saying that she had a patient in her office who was on sertraline for the treatment of “depression” and wanted varenicline to be prescribed to help the patient stop smoking. The cardiologist asked if it was OK to prescribe varenicline to a patient who was on sertraline for the treatment of depression. What did this cardiologist mean? What was her concern?

Case reports and post-marketing data had suggested that both varenicline and bupropion might be associated with neuropsychiatric adverse events (depression, suicidality, etc). So, in the past, clinicians were reluctant to prescribe varenicline (and bupropion) for smoking cessation in persons with mental health problems.

There was a very important and influential study called the EAGLES study. Do you know about this? This study addressed the issue of whether the use of varenicline or bupropion for smoking cessation is associated with an increased risk of neuropsychiatric adverse events. The findings of this study have provided us with important guidance in this matter. So, it is worth it for us to know about this study, at least briefly.

The EAGLES study

Even after these medications were already approved for smoking cessation, the FDA had asked the manufacturers of varenicline (Chantix®) and of bupropion (Zyban®) to do a post-marketing randomized controlled trial to further evaluate the risk of serious neuropsychiatric adverse events in persons who start on these medications for smoking cessation.

This important study, called the EAGLES study, was designed in consultation with the FDA (Anthenelli et al., 2016).

The EAGLES study answered the question about whether the safety of varenicline and bupropion for smoking cessation is different in persons with and without mental disorders.

Secondarily, it also provided the first definitive comparison of the relative efficacy of these different smoking cessation medications when used in smokers with mental disorders.

How was the EAGLES study conducted?

The study was a randomized, double-blind, clinical trial in smokers who were motivated to quit. On an average, the subjects smoked ten or more cigarettes per day.

Over 8000 participants–half with and half without a mental disorder–participated in this study.

The smokers belong to two groups: those with and without mental disorders. Which mental disorders? Participants in the group with mental disorders suffered from mood disorders, anxiety disorders, psychotic disorders, borderline personality disorder, etc. So, it wasn’t one specific mental disorder.

But there were two caveats about the kinds of mental disorders the participants had that are important for us to know when we are trying to apply the findings of this important study to our own patients.

1. The mental disorder had to be stable. We should not assume from this study that medications for smoking cessation have been shown to be safe in persons whose mental disorder is not currently stable.

2. No other substance use disorder in the previous 12 months.

The participants were randomly allocated to the following treatment groups:

1. Varenicline 1 mg twice a day

2. Bupropion 150 mg twice a day

3. Nicotine patch 21 mg per day with a gradual taper of the dose

4. Placebo

Brief smoking-cessation counseling was also provided at each visit.

Treatment was provided for 12 weeks and a follow-up was done after another 12 weeks.

What is a composite outcome measure?

Which neuropsychiatric adverse event should the study look at? It could be one of many. When many different outcomes of treatment are relevant, this becomes a problem statistically because when many statistical tests are done, it increases the chances that one of them might be statistically significant even when there is no real difference between the groups.

In such situations, one possible solution is to use a composite measure—a combination of several possible outcomes into one. One advantage of using a composite measure is that just one statistical test needs to be done for the main analysis. Another advantage of using a composite measure is that any one outcome (in this case, neuropsychiatric adverse event) may be rare, but combining them increases the total number of outcome events, increasing the ability of the study to detect a difference if there is one.

In the EAGLES study, the primary outcome measure for safety was the incidence of a composite measure of moderate and severe neuropsychiatric adverse events.

The primary outcome measure for efficacy was continuous abstinence from smoking during the last four weeks of the study. Abstinence from smoking was determined not only by self-report but was also confirmed by testing the participants for exhaled carbon monoxide.

Were the treatments safe to use?

The incidence of moderate or severe neuropsychiatric adverse events was as follows:

	Without mental disorders	With mental disorders
Varenicline	1%	7%
Bupropion	2%	7%
Nicotine patch	3%	5%
Placebo	2%	5%

What do you notice in these numbers (rounded to the nearest whole number)?

Within the group of those without a mental disorder, the incidence of moderate to severe neuropsychiatric adverse events in persons receiving any one of the three active treatments (nicotine patch, bupropion, or varenicline) was not statistically significantly different from that on placebo.

The same was true for participants with a mental disorder. Note: the 7% incidence in persons receiving bupropion or varenicline was not statistically significantly different from the 5% incidence in those receiving a placebo. What does “not statistically significantly different” mean? It means that there is a 5% or more probability that this difference could have been due to random chance.

Another way of thinking about these data is that if there really was a difference in the incidence of moderate to severe neuropsychiatric adverse events between persons who receive bupropion or varenicline and those who receive a placebo, the difference was so small that even with more than 8000 patients, we can’t identify it with statistical confidence.

Neuropsychiatric adverse events in those with or without a mental disorder

Of course, we see from the table that persons with mental disorders were more likely than those without mental disorders to have moderate to severe neuropsychiatric adverse events. But, this difference seemed to be related to their having had a mental disorder before entering the study and not to being in any of the four treatment groups (including placebo).

Which treatment worked better?

All three of the active treatments worked better than placebo. The odds ratio for being continuously abstinent during the last four weeks of the 12-week study were as follows (rounded to the nearest whole number):

Varenicline (odds ratio 4)

Bupropion (odds ratio 2), and

Nicotine patch (odds ratio 2)

From these odds ratios, it looks like varenicline worked better than bupropion or the nicotine patch. This was confirmed when varenicline was directly compared to bupropion and to the nicotine patch.

Efficacy in those with or without a mental disorder

Were these three treatments for smoking cessation more helpful in those with or those without mental disorders?

Answer: No difference.

“Black box” warning for varenicline dropped

Mainly as a result of the EAGLES study, the boxed warning (so-called “black box” warning) in the varenicline Prescribing Information was deleted in 2016.

But, varenicline still carries a Warning (less serious than a boxed warning) as follows (as of June 2018):

“Neuropsychiatric Adverse Events: Postmarketing reports of serious or clinically significant neuropsychiatric adverse events have included changes in mood (including depression and mania), psychosis, hallucinations, paranoia, delusions, homicidal ideation, aggression, hostility, agitation, anxiety, and panic, as well as suicidal ideation, suicide attempt, and completed suicide. Observe patients attempting to quit smoking with CHANTIX for the occurrence of such symptoms and instruct them to discontinue CHANTIX and contact a healthcare provider if they experience such adverse events.”

Bupropion (brand name Zyban) continues to carry a boxed warning, but this is because it is an antidepressant. The boxed warning notes “Increased risk of suicidal thinking and behavior in children, adolescents, and young adults taking antidepressants.”

Bottom-line

To summarize and review, here are the findings from the EAGLES study:

1. Persons with mental disorders were much more likely to experience neuropsychiatric adverse events during smoking cessation.

2. BUT, and this is the MAIN finding of the EAGLES study, the percentage of persons who experienced a moderate to severe neuropsychiatric adverse event (the composite primary outcome measure) was NOT statistically significantly different between those on varenicline, bupropion, nicotine patch, or placebo.

This means that in this large, systematic study, there was no evidence to suggest that varenicline or bupropion were any more likely than placebo to be associated with moderate to severe neuropsychiatric adverse events.

3. Of the three treatments, varenicline, bupropion, and nicotine patch, varenicline worked the best.

4. Also, the three treatments worked equally just as well in persons with a mental disorder as in those without.

What I told her

I told the cardiologist who called me that if her patient had been stable and free of significant depression while on sertraline, she should probably not withhold varenicline for smoking cessation just because the patient had a history of depression.

But, I told her that:

1. Patients, especially those with a history of mental disorders, can develop psychiatric symptoms during smoking cessation. So, all patients who stop smoking should be monitored during that period for any psychiatric symptoms.

2. And, patients with a history of a mental disorder who stop smoking should be monitored even more closely during the period of smoking cessation.

3. Patients should be told that they may develop emotional or psychological symptoms during smoking cessation, which can be a difficult period. They should immediately report any such symptoms to the treating clinician.

(By the way, I also told her that in June 2018, a warning about increased risk of cardiovascular adverse events was added to the Chantix Prescribing Information.)

A call to action

Please note, dear colleagues, that smoking is a very serious health problem. As an ex-smoker myself, I can tell you that can be very hard to stop smoking. We should not only encourage patients to stop smoking, we should try to do anything and everything we can do to help them to stop smoking.

I remind you of what I said earlier–smoking is much more common in persons with mental disorders. Our patients should not be deprived of the potential benefits of smoking cessation medications just because they have a mental disorder.

Varenicline (Chantix®): Basic Information

Chantix Prescribing Information (external link)

Zyban Prescribing Information (external link)

What are the BEST books on each topic related to psychiatry/ mental health?

References

Anthenelli RM, Benowitz NL, West R, St Aubin L, McRae T, Lawrence D, Ascher J, Russ C, Krishen A, Evins AE. Neuropsychiatric safety and efficacy of varenicline, bupropion, and nicotine patch in smokers with and without psychiatric disorders (EAGLES): a double-blind, randomised, placebo-controlled clinical trial. Lancet. 2016 Jun 18;387(10037):2507-20. PubMed PMID: 27116918.

Zawertailo L. Safety of smoking cessation drugs for mentally ill patients. Lancet. 2016 Jun 18;387(10037):2481-2. PubMed PMID: 27116919.

Disclaimer: The content on this website is provided as general education for medical professionals. It is not intended or recommended for patients or other laypersons or as a substitute for medical advice, diagnosis, or treatment. Patients must always consult a qualified health care professional regarding their diagnosis and treatment. Healthcare professionals should always check this website for the most recently updated information.