Published online 2 September 2008 | Nature | doi:10.1038/news.2008.1076


Cholesterol-lowering drug given cancer all-clear

Full trial results and meta-analysis both contradict preliminary scare.

pillsCancer fears about a cholesterol-lowering drug have turned out to be a 'statistical blip'.Getty

A class of cholesterol-lowering drug that had been implicated in a cancer scare has been cleared by epidemiologists using the results from two large ongoing clinical trials.

That 'meta-analysis', led by Richard Peto of the University of Oxford, UK, is published online today by the New England Journal of Medicine1, along with full results of the trial that first raised the scare, which were also presented today at the European Society of Cardiology congress in Munich, Germany.

Preliminary results from the relatively small SEAS (Simvastatin and Ezetimibe in Aortic Stenosis) study had triggered the cancer scare in July. The trial was designed to investigate whether a combination of these two cholesterol-lowering drugs would help in cases of aortic stenosis, where a key heart valve is partially blocked.

Simvastatin, a statin that has already been used in other drugs, works by blocking cholesterol synthesis in the body. The newer ezetimibe works by blocking the absorption of cholesterol from the gut. A unusual pre-publication announcement from the SEAS study reported that, while the therapy did not help very much in aortic stenosis, cancer appeared more frequently in the treated groups.

“If you interrogate your study hard enough, under torture it will confess – but that confession won't be worth the paper it is written on.”

Rory Collins
University of Oxford

In the study, 1,873 people in seven European countries with mild or moderate aortic stenosis were treated with either a combination of both drugs or a placebo, and followed for at least four years. The progress of stenosis appeared unaffected by the choice of treatment, although fewer of those treated needed bypass surgery for acute ischaemic heart disease over this period. However, 105 people from the treated groups developed a cancer within the four years, whereas only 70 in the placebo group did. There were also more cancer deaths in the treatment group.

Whenever a clinical trial throws up a major safety concern, the sponsors – in this case the companies Merck and Schering-Plough, who market the drug combination under the names Vytorin and Inegy — are required to inform the US Federal Drug Administration (FDA) and other national drug agencies within 15 days. So the managers of SEAS reported the cancer observations to the FDA and the European Medicines Authority on 21 July.

"By coincidence, this was also the day that Merck and Schering were obliged to update the stock exchange on the progress of their joint trial," says Terje Pedersen, a cardiologist from the Ulleval University Hospital in Oslo, Norway, and principal investigator in SEAS. "We estimated that there wasn't much hope of retaining confidentiality, so we decided it was probably best to go public with the data ourselves, without waiting for the full trial to be published." He says he wasn't surprised by the subsequent media alarm. "Cancer is a sensitive issue."

No surprises

But Pedersen says he was also unsurprised when the apparent association turned out to be a statistical blip2, as the sample size in the SEAS study was far too small to detect modest increases in risk with any certainty.

The media alarm and associated political pressure prompted the management teams of two large ongoing double-blind clinical trials involving ezetimibe to take the unusual step of 'unblinding' their trials several years before they are due to end. They provided data on cancer incidence for analysis.

Those two trials were the UK-based Study of Heart and Renal Protection (SHARP) study, with 9,264 patients, and the US-based Improved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT) study, with 11,353 patients so far. The link with cancer did not emerge in these larger groups.

"I think many of us are starting to find that things are going too far," says Oxford epidemiologist Rory Collins, an author on the NEJM paper that combined the results from SHARP and IMPROVE-IT. "A few years ago data-watch committees would never have agreed to unblind a clinical study." But, he notes, clinical trails have recently become much more sophisticated, with clinicians trying to wring as much information as possible from them – and they sometimes become victims of their success. "If you interrogate your study hard enough, under torture it will confess – but that confession won't be worth the paper it is written on."

Other small studies have raised cancer alarms with lipid-lowering drugs, he says, and the associations have always disappeared in statistically robust larger studies.

Nevertheless, an associated editorial3 in the NEJM does not give a whole-hearted 'all-clear' to ezetimibe. It raises the issue that this class of drug could theoretically block absorption of other molecules, in addition to cholesterol, and that could affect cancer growth.

Cardiologist Dan Atar, from Aker University Hospital in Oslo, who was not involved in the trial, comments that lipid-lowering drugs have a major clinical role in controlling cardiovascular disease, but not everyone can tolerate the well-tested class of drugs known as statins, of which simvastatin is a member. 

  • References

    1. Peto, R. et al. N. Engl. J. Med. published online 2 September 2008 (doi:10.1056/NEJMsa0806603).
    2. Rossebø, A.B. et al. N. Engl. J. Med. published online 2 September 2008 (doi:10.1056/NEJMoa0804602).
    3. Drazen, J.M. et al. N. Engl. J. Med. published online 2 September 2008 (doi:10.1056/NEJMe0807200).
Commenting is now closed.