The field of smoking cessation has been abuzz since the FDA approved Chantix (varenicline) in 2006. Pfizer’s studies showed better efficacy for Chantix than bupropion (Wellbutrin/Zyban), and the mechanism of action is fascinating (see below).
But several controversial issues have arisen to muddy the waters. Reports of psychiatric side effects surfaced, prompting the FDA to issue a public health advisory. And on an unrelated note, a new “cold turkey” quit smoking movement has arisen, filled with critics of pharmacological cures. Let’s look at some of these issues in more detail.
Is quitting “cold turkey” the best smoking cessation technique?
There’s no question that placebo-controlled trials have repeatedly shown a benefit of pharmacological approaches to quitting. For example, over 130 placebo-controlled trials of nicotine replacement therapy (NRT) have been done, and NRT produces an average one year quit rate of about 16% vs. 10% for placebo (Stead LF et al., Cochrane Database of Systematic Reviews 2008 Issue 1). While this may not look like such a huge effect, it means that NRT increases your chances of quitting by at least 50% over using a placebo, and given how dangerous smoking can be, this makes a big difference.
However, if you look at a different kind of study, namely, a retrospective survey, you get a different sense of things. Rather than starting with smokers and enrolling them in a clinical trial, these studies start with large groups of former smokers and ask them what technique finally allowed them to successfully kick the habit. In these studies, the most commonly endorsed techniques is neither NRT, nor bupropion, nor Chantix; rather, it is cold turkey, a relatively inexpensive technique that involves simply deciding never to smoke again. How low tech can you get!
For example, in a British study of 2069 smokers who had sought treatment, 25% of successful quitters recalled cold turkey as having been successful, while 15% reported that bupropion or NRT were successful (Ferguson J et al., Addiction 2005;100:59-69). Similarly, an Australian study found that 40% of former smokers endorsed cold turkey and 20% endorsed NRT or bupropion (Doran CM et al., Addictive Behaviors 2006;31:758-766). Findings such as these have fueled an increasingly vocal movement of cold-turkey boosters, with well-organized websites that often appear to have an ax to grind with pharmaceutical companies (see, for example, WhyQuit.com).
So what gives? One possibility is that many of the placebo-controlled NRT studies may have overestimated the effect of NRT, because they were not successful at blinding the placebo. Patients in such studies can often accurately guess whether they are randomized to an inactive patch, presumably because they experience more severe withdrawal symptoms than patients on the real thing (Mooney M et al., Addictive Behaviors 2004;29:673-684). This may have caused patients assigned to the placebo arm to experience a weaker placebo effect than the NRT group.
On the other hand, the survey data has its own methodological problems, the most glaring of which is that they are retrospective and therefore rely on that notoriously unreliable commodity, human memory (Shiffman S, Thorax 2007;62:930-931). Despite the findings of these survey studies, the overwhelming majority of experts in the field still strongly believe in the value of NRT, and the sheer number of positive studies is impressive. But it may be valuable to remind patients that many of their partners-in-smoking have benefited from the cold turkey approach, and it’s something to try when you are running out of options.
Tips for Prescribing Nicotine Replacement Treatment
Assuming that you believe in NRT, an extremely useful article was recently published by leading experts in this treatment, only some of whom have financial affiliations with manufacturers of these products (Kozlowski LT et al., Addictive Behaviors 32 (2007) 2140-2150).
These experts decry a paradoxical “regulatory imbalance” in NRT, namely that there are far more FDA-mandated health warnings on over-the-counter NRT preparations than there are on packs of cigarettes, which are – let’s face it – much more dangerous than the patch! Misconceptions abound, including the belief that nicotine can cause cancer (it doesn’t – other chemicals in tobacco are carcinogenic), and that nicotine in medication can cause an addiction (it doesn’t – it’s just a convenient way of tapering off an existing addiction).
The article is available free on the internet, and I posted it on The Carlat Psychiatry Report website. Some of the most useful pieces of practical advice include:
“When using nicotine gum or lozenges, patients should “park the gum” between their teeth for 2-3 min between chews, which optimizes the absorption of nicotine through the mucosa. Do not drink anything while chewing to avoid diluting the nicotine.
“While you shouldn’t smoke heavily while wearing the patch, it’s okay to have a few cigarettes if you can’t resist – don’t take the patch off just because you slip.
“You may need to combine the patch with the gum or lozenge to get you over the hump, or you may need to use more than one patch at a time. Let your own reactions be the guide of whether you are getting too much nicotine – typical side effects are dizziness and a racing pulse.
“Shop around. NRT costs vary widely and you can often find much more affordable NRT by shopping at online pharmacies."
What’s Happening with Chantix?
Chantix (varenicline) was approved by the FDA for smoking cessation in May of 2006. It has a fascinating dual mechanism of action: 1) It is a partial agonist at nicotinic receptors, mimicking nicotine effects on the brain; 2) It also blocks nicotine from binding to these receptors, thereby decreasing the reinforcing effect of smoking. So essentially it is a pill that fools your brain into believing you are smoking while at the same time preventing smoking from feeling very good.
The Pfizer-sponsored trials have compared Chantix with both Zyban (bupropion) 150 mg BID and placebo, but, strangely enough, not with NRT. In two of the major trials, patients received treatments for a total of 12 weeks, and the primary outcome measure was continuous abstinence from weeks 9-12. The combined results showed Chantix outperforming both placebo and Zyban, with quit rates of: Chantix (44%), Zyban (30%), and placebo (18%) (Gonzales D et al, JAMA 2006;296:47-55, and Jorenby DE et al, JAMA 2006;296:56-63). However, as you might predict, the abstinence rates for all patients declined over time. The continuous abstinence rate for weeks 9-52 were: Chantix (22.5%), Zyban (15.5%), and placebo (9%). One non-industry funded open label trial recently compared Chantix with NRT and found that it was only slightly more effective, and in fact no more effective than combination NRT (meaning the use of two methods simultaneously, such as the patch plus the gum). In this trial, patients were assigned to treatments based on their preferences (Stapleton JA, et al., Addiction 2007; 103, 146-154). While this was neither placebo-controlled nor double-blinded, it is nevertheless a useful study for clinicians, because it implies that those patients who are amenable to NRT can do just as well using it as they could by using Chantix, particularly good news given the recent FDA warning covered below.
How Dangerous Is Chantix?
Recently, the FDA issued a public warning about the psychiatric side effects of Chantix (http://www.fda.gov/cder/drug/ InfoSheets/HCP/vareniclineHCP.htm). There have been a number of patients reporting worsening depression, anxiety, and suicidal ideation. More alarmingly, there have been several reports of completed suicides: so far, a total of 34 suicides out of approximately 4 million users. That works out to roughly 1 suicide/100,000, which is much lower than the baseline U.S. suicide rate of 10/100,000. However, post-marketing adverse events are notoriously underreported, so the actual suicide rate on Chantix might be much higher.
While the risk of serious psychiatric effects appears to be quite low, psychiatrists should be particularly alert to this possibility, given that our patients are likely to be more susceptible to these symptoms.
One reassuring finding was reported in the Stapleton study mentioned above. Part of this report focused specifically on the effects of Chantix and NRT in 111 smokers with psychiatric disorders, primarily depression or bipolar disorder. Chantix led to no more side effects in psychiatric patients than in other subjects in the study, with the most prominent side effects being nausea (38% of all patients), insomnia (30%) and vivid dreams (13%). No suicidality was reported.
The bottom line is that Chantix appears to be an effective smoking cessations aid, but you should warn patients about the possibility of nausea, vivid dreams, insomnia, and worsening anxiety and depression.