Tuesday, July 5, 2011

Taking Chances on Chantix

The smoking cessation drug Chantix (generic name varenicline) has been associated with several potentially serious side effects. In 2009, the FDA issued a “boxed warning” for the product, based on this rationale: “Some people who have taken Chantix or Zyban to help them quit smoking have reported experiencing unusual behavior changes, have become depressed or have had their depression worsen, or have had thoughts about suicide or dying; some have attempted suicide either while using one of these drugs or after they stopped taking them.” (FDA advisory here).

Now a bigger warning flag has been raised. On July 4 a meta-analysis, published in the Canadian Medical Association Journal (available here), concluded that Chantix was associated with “an increased risk of serious adverse cardiovascular events,” based on an odds ratio of 1.72 (95% Confidence Interval = 1.09 – 2.71).

The FDA may have been aware of the new analysis; on June 16, 2011, the agency issued a drug safety announcement under this headline: “Chantix (varenicline) may increase the risk of certain cardiovascular adverse events in patients with cardiovascular disease.” (read it here ). It is a fascinating document.

First, what is a “certain cardiovascular adverse event”? The answer is important: a nonfatal heart attack (myocardial infarction). It is strange that the FDA titled the announcement so vaguely, when the report was primarily about the risk of heart attacks. The first time myocardial infarction is mentioned is halfway through the notice.

The announcement’s “Data Summary” section contains information upon which the FDA based its decision. The agency reviewed a clinical trial involving over 700 patients, about half of which (n = 355) received Chantix for 12 weeks, while the other half (n = 359) received a placebo. All had been previously diagnosed with stable cardiovascular disease (other than high blood pressure). Both groups “received smoking cessation counseling throughout the study,” so it appears that they got the best treatment available. The follow-up was complete at one year.

Of patients getting the placebo, only three, or 0.86%, suffered heart attacks. Seven, or about 2%, of the patients in the Chantix group had heart attacks. The numbers are small, so this result is not statistically significant. But the relative risk of heart attack in the Chantix group is 2.3, which was undoubtedly worrisome to FDA officials.

So what were the quitting rates? In the placebo group, only 7% were continuously abstinent at one year. That’s 26 out of the 359 who enrolled. This means that an astounding 93%, or 333 smokers, did not respond to “smoking cessation counseling.” In other words, the claim that safer tobacco substitutes are not needed because smoking cessation counseling works is utter nonsense.

In the Chantix group, the abstinence rate at one year was 19%. That represents 68 of the 353 smokers who started the study. Although the FDA says that this is “significantly higher” than the placebo group, 285 participants were still smoking. Tobacco harm reduction could help them, too.

Strangely, it appears that the FDA was trying to downplay the announcement; they did not identify the research source. In fact, the study was published in 2010 in Circulation (abstract here ), by first author Nancy Rigotti (Harvard) and, among others, co-author Neal Benowitz, who is a member of the FDA Tobacco Products Scientific Advisory Committee. The study was conducted in 15 countries. Pfizer, manufacturer of Chantix, sponsored the research and paid consultation fees to Rigotti and Benowitz.

Another oddity: the authors of the study were not concerned about the adverse cardiovascular events related to Chantix. Although they acknowledged that “trial size and duration preclude a definitive conclusion about [safety],” they asserted that “this study provides reassurance to physicians that varenicline appears to be safe to use in smokers with stable cardiovascular disease.” Given its June 16 warning, the FDA clearly disagrees with Rigotti and colleagues about the potential risks of Chantix.

In a previous post I discussed the American Heart Association’s rejection of smokeless tobacco as a substitute for cigarettes (here), based on almost no significant research findings; that position was also co-authored by Neal Benowitz. It is evident that some health professionals who reject tobacco harm reduction are willing to tolerate serious adverse events associated with the medicines they prescribe. If they accept relative risk for prescription drugs, why reject relative risk for tobacco products?


Anonymous said...

I assume that everyone realized that TPSAC suppressed evidence that menthol cigarettes were associated with lower lung cancer risk compared to non-menthol cigarettes in their report. A lot of that was probably Samet, who was in charge of the disease risk section, but his NEJM editorial co-author Benowitz wouldn't have helped much.

Randall said...

Smoking cessation counseling does not work ? It might be because medical understanding of what helps is not appropriate. Medicine is - and should - rational ; motivations are not...

Brian said...

Benowitz is scary. He recently published a paper claiming that African-American heavy smokers (20+ cpd) had no greater nicotine exposure than African-American light smokers (< 10 cpd). He had data for 15 people in each group.

Of course, biomarker data from national health surveys such as NHANES confirm that this is nonsense.