Merck "good" cholesterol drug meets trial goals

Merck believes the favorable results should sway patients, who now shun HDL-boosting drugs because of bothersome flushing, to stick with treatment that may cut their risk of heart attacks and other cardiovascular problems.

Analysts reckon that could make Cordaptive, which may get to market in 2008, an eventual multibillion-dollar-a-year seller.

The two-in-one drug contains an extended-release form of the vitamin niacin to raise levels of HDL, plus laropiprant to stop flushing, a dilation of blood vessels that causes redness and an unpleasant burning sensation on the neck and face.

HDL is considered "good" because it acts like a garbage truck, removing excessive levels of artery-clogging LDL cholesterol from the bloodstream.

For years, drugmakers and doctors have relied on drugs that cut LDL cholesterol. But increasingly the focus is shifting to HDL-raising drugs because a significant portion of heart attack patients have low levels of HDL, yet relatively normal LDL.

The Phase III trial reported at the annual European Society of Cardiology congress in Vienna involved 1,613 patients with high levels of LDL cholesterol or harmful blood fats known as triglycerides or low levels of heart-protective HDL.

About 65 percent were already taking statins, the popular class of LDL-lowering medicines that include Pfizer's (NYSE: - ) Lipitor and Merck's Zocor.

Patients started out taking either 1 gram of Cordaptive daily, 1 gram of extended-release niacin alone or a placebo. After four weeks, those receiving Cordaptive and niacin alone doubled their doses to 2 grams for another 20 weeks.

Levels of LDL cholesterol fell an average 19 percent during weeks 12 to 24 among patients taking 2 gram-doses of Cordaptive, compared with a fall of 0.5 percent among those on placebos.

Average HDL levels rose 19 percent in the Cordaptive group, versus a 1.2 percent drop in the placebo group. Moreover, triglyceride levels fell 22 percent, compared with an increase of 3.6 percent with placebo.

EFFICACY MATCHES NIASPAN

Dr. Richard Pasternak, a senior Merck research official who helped conduct the trial, said the findings could trigger a rethink of treatment strategies.

"It's a breakthrough because this comprehensive lipid therapy has not been available or achievable by most people. Physicians are reluctant to prescribe the agent (niacin) and patients are reluctant to take it," he said in an interview.

Importantly, the beneficial effects of Cordaptive matched those commonly seen with Niaspan -- a widely used niacin drug sold by Abbott Laboratories (NYSE: - ).

Although Niaspan was not used in the Merck trial, Pasternak said the cholesterol outcomes were "in exactly the same range," with Cordaptive avoiding the significant flushing problem.

An estimated 69 percent of patients given the initial 1 gram doses of Cordaptive reported either no flushing or mild flushing symptoms during the first week of treatment, compared to 44 percent of those who received extended-release niacin alone.

The frequency of moderate or greater flushing by week 24 was 0.2 days per week among patients taking 2-gram doses of Cordaptive or placebos, compared with 0.7 days among those given niacin without the flushing inhibitor.

The difference meant more than twice as many patients on niacin alone discontinued treatment because of flushing than patients taking Cordaptive -- 22 percent versus 10 percent.

Two-gram doses of niacin are typically needed to boost HDL levels but few patients taking current niacin products are willing to stick with such doses because of flushing. The most common prescribed dose is only half a gram.

Cordaptive, by making a full 2-gram dose of niacin tolerable, should allow patients to get the full benefit.

Merck said last week U.S. regulators will likely decide by the second quarter of 2008 whether to approve Cordaptive.

If cleared, it will compete head-on with Abbott's Niaspan, which had second-quarter sales of $170 million -- and Merck is banking on its anti-flushing power to give it a commercial edge.