Wednesday, September 17, 2008

Top Statistician Says Cancer Risk from Vytorin/Inegy 'Not Ruled Out'

The stuff doesn't work anyway. That you have to do statistical marvels to detect any effects shows how weak the effects concerned are. I would worry about its overall efficacy (effect on mortality) rather than anything else

One of the most prominent statisticians in the U.S. is taking issue with an analysis that claims there is no credible evidence the cholesterol drug Vytorin increases the risk of cancer death. "There are clinically important ["clinically important" is waffle. It can mean "not statistically significant"] increases in the risk of cancer-related death that are not ruled out by this data," writes Thomas Fleming of the University of Washington in an editorial published on the Web site of the New England Journal of Medicine.

Moreover, he argues in the editorial, the analysis arguing against a risk of cancer death for Vytorin should never have been made public because it could compromise ongoing studies of the drug. That would mean clear evidence about Vytorin's safety and effectiveness might never be obtained.

The editorial spells more trouble for Vytorin and its sister drug, Zetia--which share an active ingredient--and for their makers, Merck and Schering-Plough. Vytorin is a combo pill of Zetia and the generic drug Zocor. The pills generated $5 billion in sales last year, but U.S. prescriptions have dropped by a third as a result of the controversy over Zetia's effectiveness at preventing heart attacks.

In July, researchers hastily released a study, called SEAS, in which patients on Vytorin developed and died of cancer more often than those in a placebo group. But, at the same time, they presented an analysis by Oxford University's Sir Richard Peto, who argued there is "no credible evidence" of a link between the active ingredient in Vytorin and Zetia and cancer. Both the study and the analysis were published in the New England Journal on Sept. 2.

Many cardiologists who defend Vytorin and Zetia rest their arguments on Peto's brilliance. Eugene Braunwald, a prominent Harvard cardiologist who is running a Vytorin trial, called Peto "the best statistician in the universe" while telling reporters that any cancer risk with the drug is unlikely.

But Fleming is also highly regarded. A frequent adviser to the Food and Drug Administration, Fleming was a key player in sorting through the confusing data surrounding the link of arthritis drugs Vioxx and Bextra to heart risks. He also warned the FDA not to approve the cancer drug Iressa because of limited data; use of the medicine was substantially restricted after it failed to demonstrate a survival benefit in big trials.

In a statement, Merck and Schering-Plough said they believe the cancer finding "is likely to be an anomaly that, taken in the light of all the available data, does not support an association with Vytorin."

Fleming agrees with Peto on one point. Peto's analysis looked at data taken from two incomplete studies of Vytorin: SHARP, which compares Vytorin with a placebo in patients with kidney disease, and IMPROVE-IT, which compares Vytorin with Zocor. Some critics have said Peto should have lumped all three trials together. Fleming writes Peto's was the right approach, since the point was to see if the cancer risk that shows up in SEAS would show up elsewhere.

Doing the analysis this way showed no increase in the number of cancers, but an increase in the number of deaths from cancer. Fleming argues that this still leaves a possibility that the pills are linked to cancer death. In fact, the data are consistent with a 34% increase in cancer death, and don't rule out an 84% increase.

Fleming also argues that Peto's analysis of the IMPROVE-IT and SHARP studies raises "important concerns" about the integrity of those clinical trials. Studies are normally kept secret until they are completed. Only a special committee charged with monitoring patient safety is allowed to look at the unblinded data, so that a trial can be stopped if a drug turns out to be dangerous. (Forbes raised these issues this morning.)

Fleming writes there is a "serious risk" that data from incomplete studies will be misinterpreted. He also argues that making unblinded data available to the public could disturb the existing studies, potentially making the results less valid when the trials are eventually released. He writes the data in the Peto analysis should only have been made available to the committees monitoring the safety of the patients in SHARP and IMPROVE-IT.

He concludes that more data are needed to "adequately address the signal that [Vytorin] is associated with an increased risk of death from cancer." He adds that getting more data is "especially important" because Zetia has a safety signal of major illness or death, while evidence of its efficacy is limited to lowering cholesterol.


Small Packages May Lead to Overeating

Another counterproductive idea from the food Fascists

Tempting treats are being offered in small package sizes these days, presumably to help consumers reduce portion sizes. Yet new research in the Journal of Consumer Research found that people actually consume more high-calorie snacks when they are in small packages than large ones. And smaller packages make people more likely to give in to temptation in the first place.

Authors Rita Coelho do Vale (Technical University of Lisbon), Rik Pieters, and Marcel Zeelenberg (both Tilburg University, the Netherlands) found that large packages triggered concern of overeating and conscious efforts to avoid doing so, while small packages were perceived as innocent pleasures, leaving the consumers unaware that they were overindulging.

"The increasing availability of single-serve and multi-packs may not serve consumers in the long-run, but-because they are considered to be innocent pleasures-may turn out to be sneaky small sins," write the authors.

One fascinating aspect of the research is the difference between belief and reality. In an initial study, researchers found that consumers believe that small packages help them regulate "hedonistic consumption," where self-restraint is at stake. When participants were asked to choose phone plans, those who thought the plan was for social rather than work purposes tended to choose smaller plans.

The researchers then moved on to food. Participants in one group had their "dietary concerns" activated by completing a "Body Satisfaction scale," a "Drive for Thinness scale," and a "Concern for Dieting scale." They were then weighed and measured, in front of a mirror, to fully activate their awareness. Then those participants (and a control group, which didn't have its "dietary concerns" activated) watched episodes of Friends interspersed with commercials. They believed they were there to evaluate the ads.

But researchers were really monitoring their consumption of potato chips. Chips were available to participants in large packages or small ones. The study found that consumption was lowest when dieting concerns were activated and package size was large. People were less likely to open large packages, and participants deliberated longer before consuming from the larger packages.

"Maybe the answer lies in consumers taking responsibility for their consumption and monitoring internal cues of sufficiency, rather than letting package size take control," conclude the authors.


No comments: