A tiny biotech may have found its heart drug could save millions of lives and make billions of dollars. Unfortunately, it had to do a scientific no-no to get there, holding off on announcing the results of a finished study until they could be completely re-analyzed.
The study, called Cart-2, was initially done at the Montreal Heart Institute, and was designed to measure plaque in patients' arteries. But AtheroGenics sent the data to the Cleveland Clinic to be re-analyzed. Steven Nissen, a leading cardiologist who was not originally part of the study, discarded 30% of patients because of poor imaging data. Among the 133 patients whose artery scans were usable, the analyses from both the Montreal Heart Institute and the Cleveland Clinic show plaque shrinking, a major win for AtheroGenics, and an indication that its drug could be a important weapon in the arsenal of cardiologists. Full results are due by year-end.
'It's not the final word'
"It in fact is promising," says Nissen, "there is no other interpretation. But it's not the final word."
AtheroGenics' heart drug, known by the code-name AGI-1067, was the first pill designed specifically to target inflammation in the arteries. Scientists have recently come to believe inflammation in blood-vessel walls maybe as important as high cholesterol in causing heart attacks. It is inflammation that leads bad cholesterol to bundle up into plaque in the arteries, and inflammation that causes plaque to become unstable, shearing off and triggering a heart attack. AGI-1067 was designed to halt this process, and an earlier trial had offered hints that the drug might clear plaque from the arteries.
That earlier trial had been conducted by Jean-Claude Tardif at the Montreal Heart Institute. In a trial dubbed Cart-1, Tardif had used a technique called intravascular ultrasound, or IVUS, to measure whether AGI-1067 stopped arteries from re-closing after being propped open with balloons in angioplasty surgery. Looking at videotape from his ultrasound probes, Tardif had seen hints that the drug was actually stopping plaque buildup in areas untouched by the balloon. Working with AtheroGenics, he redesigned another angioplasty study, called Cart-2, to try to prove the drug halted or even reversed atherosclerosis.
Such proof would have been a major win for AtheroGenics. Pfizer spent $1.3 billion to buy an injectible drug--a form of HDL, or good cholesterol -- that can clear plaque from the arteries. And that drug giant is spending more than $800 million to test its own HDL-raising pill. That drug, it has been said, could be the biggest medicine ever launched. But AGI-1067 would actually be closer to market--a late-stage trial is set to be finished in 2005, though detractors say it is too short and too small. If Cart-2 came out positive, that would be a big boost to AtheroGenics and its potential for snagging a lucrative deal with a big pharmaceutical company.
The Cart-2 results showed a reduction of 3.3% for patients taking AGI-1067, but plaque volume also decreased in patients who got placebo. (That result wasn't statistically significant.) AtheroGenics Chief Executive Russell Medford and Rob Scott, a former Pfizer executive who had come to AtheroGenics to develop AGI-1067, wanted a second opinion. The IVUS technique has taken the drug industry by storm. It showed Pfizer's Lipitor to be more effective than a lower dose of a rival drug from Bristol-Myers Squibb, and IVUS results led Pfizer to purchase the injectible HDL drug and its maker, Esperion Therapeutics.
But those results hadn't come from Tardif, but from The Cleveland Clinic's Nissen, who had pioneered the IVUS technique. Nissen was not initially involved in Cart-2, but he had worked closely with Scott when Scott was at Pfizer. AtheroGenics decided to have Nissen re-analyze all of the IVUS data Tardif had produced.
Nissen and his group at the Cleveland Clinic set about arduously going over the video Tardif's researchers had taken of the inside of patients' arteries. It takes one technician per day to look at one patient's IVUS result. Nissen found that a great many of the tapes didn't match his standards. He threw out a third of them, meaning that only 133 patients were used in the re-analysis. That includes patients getting both placebo and AGI-1067. At the end of the trial, it is expected that 260 patients will be evaluable out of the original 500.
But although the study was shrinking, Nissen's re-analysis provided a fairly clear result. In a year, AGI-1067 reduced artery plaque by 3.8%. That's comparable to what Esperion's injectible HDL achieved in six weeks. The reduction was statistically significant, but the difference between placebo and drug was not. If those results hold up, AGI-1067 would be the first pill ever shown to reduce artery plaque.
Some problems with the final Cart-2 result were originally designed as an angioplasty study. Moreover, investors may feel they were left out of the loop while AtheroGenics was going about re-analyzing data. In a sense, the company is now exactly where it started. Both Cart-1 and Cart-2 are studies that had to be re-analyzed to get at the meat of their results. That's an incredibly unusual step in medical studies.
Viewed in another way, though, AtheroGenics is in a wildly good position. There is a very good chance that AGI-1067 works, and that could encourage a big pharmaceutical firm to partner with the tiny biotech. Pfizer is betting big on its HDL-raising pill being the first to reduce artery plaque. If AGI-1067 looks like competition, Pfizer may be very tempted to buy AtheroGenics outright, rather than risking a competitor such as Merck, AstraZeneca, or Bristol-Myers Squibb getting a hold of the drug.
AtheroGenics has a real shot at an amazingly successful product. But the company needs to be careful that future studies come out right the first time.