November 7th, 2014

What You Need to Know About IMPROVE-IT

The IMPROVE-IT trial will be big news when its results are finally presented on November 17 during the annual meeting of the American Heart Association. The results of the trial — underway for nearly a decade — have been long and eagerly awaited by everyone interested in cardiovascular medicine.  The trial could impact the future sales of a key Merck drug, ezetimibe; because it is nearing the end of its patent life the commercial significance is somewhat limited. However, IMPROVE-IT will also have very important implications beyond its specific effect on one drug franchise and could influence the fate of several new drugs now being investigated and may even alter the entire drug development and evaluation process.

Here’s some background information and links to useful resources:


Ezetimibe blocks the absorption of cholesterol in the intestine. This leads to an increase in LDL-cholesterol receptors in the liver. These receptors then remove LDL from the blood. When added to statin therapy, ezetimibe generally lowers cholesterol by an additional 20-25% over the statin alone. In the U.S. it is marketed by Merck as Zetia (ezetimibe alone) and Vytorin (ezetimibe in combination with simvastatin).

Ezetimibe was originally developed by Schering Plough and co-marketed by Merck, which ultimately acquired Schering Plough. The drug was first approved by the FDA in 2002. It quickly became a blockbuster, with sales reaching more than $5 billion a year within a few years. Critics of the drug noted that, despite its success, there were no trials showing that the drug conferred any actual clinical benefit besides lowering LDL. This stood in sharp contrast to the already established statins, as a long series of trials had already firmly established their clinical benefits.


But there were no serious challenges to the drug’s commercial success until the ENHANCE trial controversy. ENHANCE was not a cardiovascular outcomes trial. It was a smaller trial that compared the effect in people with familial hypercholesterolemia of ezetimibe or placebo in combination with simvastatin on a surrogate endpoint, in this case the progression of atherosclerosis in the carotid and femoral arteries as measured by ultrasound. The trial found no advantage for ezetimibe. By itself this might have been only minor news and likely would only have had a small effect on the commercial market, but the importance of the trial was greatly magnified due to the enormous amount of prepublication publicity that came about because of revelations that the sponsor, Schering Plough, had interfered with the trial and delayed its presentation. When the results were finally presented they had a dramatic effect on the drug’s reputation but a more modest effect on sales. Although ezetimibe remains a successful drug to this day — worldwide sales still total more than $4 billion each year — the ENHANCE trial put an end to the drug’s explosive growth, and provoked an explosion of criticism and concern that has yet to entirely die down.


The IMProved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT) is a true cardiovascular outcomes trial and was designed to test the effect of adding ezetimibe to simvastatin in patients with an acute coronary syndrome (ACS). Earlier this year the IMPROVE-IT investigators reported that the trial had randomized 18,144 patients with ACS. These patients had LDL levels below 125 mg/dL if they were not already taking a cholesterol drug or below 100 mg/dL is they were taking a cholesterol drug. By itself the simvastatin dose was chosen to achieve an LDL level below 70 mg/dL, the target cholesterol goal according to the cholesterol guidelines at the time the trial was started. Trial enrollment began in 2005 and was completed in 2010.

The primary endpoint of IMPROVE-IT is cardiovascular death, nonfatal MI, nonfatal stroke, rehospitalization for unstable angina (UA), and coronary revascularization.

Along the way the trial design has been modified several times. The most important change was the increase in the number of patients, from 10,000 to 18,000. This was because of a lower than expected number of primary endpoint events. The trial was designed to accumulate 5,250 primary endpoint events after a minimum 2.5 years of followup for all patients. These parameters were chosen in order to detect a 9.3% reduction in the primary endpoint with ezetimibe, based on an anticipated 15 mg/dL drop in LDL in the ezetimibe group.

In addition to closely monitoring the safety of the patients in the study (for a while during the trial there was a concern that ezetimibe might raise the risk of cancer — this concern has been since dispelled) the data and safety monitoring board has performed three efficacy analyses. The protocol included stopping guidelines for “overwhelming efficacy” so it is clear that the results, at best, will be somewhat modest. But the protocol did not contain stopping rules for futility, so it is possible, as some think, that the DSMB has known for some time that the trial will not show a significant benefit for ezetimibe.

A distinct possibility, raised by the study investigators and others, is that because LDL levels will be low in both groups and because the difference in LDL between the groups will be modest, the results at best will show only small and unimpressive clinical differences. A good example of the attempt to prematurely discount the eventual result of IMPROVE-IT is an opinion piece published earlier this year in the European Heart Journal  which concludes that the likely negative results of the trial will not help understand whether ezetimibe is beneficial in people with higher cholesterol levels at baseline.


The IMPROVE-IT authors also discussed the implications of the trial beyond what it will tell us about ezetimibe:

… the practical value of a positive trial will serve as a definitive proof of concept that LDL-C lowering with compounds other than statins can affect outcomes. The trial may provide support for driving LDL-C to levels below those previously recommended and will raise the issue of whether other molecules — such as PCSK9 antibodies — could have similar proportional benefit as an adjunct to statins.”

I have argued in the past that the fate of the new PCSK9 inhibitors (potent LDL-lowering antibodies under development by Amgen and Sanofi/Regeneron) may depend on the results of IMPROVE-IT. If the trial meets its primary endpoint then the FDA will be reassured and will likely accept LDL-lowering as a surrogate endpoint, leading to early approval of the PCSK9 inhibitors without evidence from cardiovascular outcomes trials (which are now taking place). However, if IMPROVE-IT does not meet its primary endpoint then it seems to me that the FDA and its advisors will need to take a much more critical look at LDL as a surrogate endpoint. Failure of IMPROVE-IT will almost certainly encourage critics who will condemn the FDA for approving a new class of cholesterol-lowering drugs without evidence of significant clinical benefits.

Others have argued that the far greater potency of the PCSK9 inhibitors means that any comparison with ezetimibe is inappropriate.

If the results of IMPROVE-IT are positive they will also be used to support the use of ezetimibe in people who can’t tolerate high-dose statins and in people with persistently high LDL levels who are already taking statins. If it does not meet the primary endpoint, the authors say, the results “could certainly affect beliefs about the ‘lower is better’ hypothesis.”

The Bottom Line

In the end there are a few likely scenarios.

Negative Effect — Ezetimibe treatment results in worse cardiovascular outcomes than placebo. This is probably the least likely of the likely scenarios, because there is little suggestion of harm in previous studies, but the possibility should not be entirely dismissed. A true negative effect would have a stunning impact not only on the ezetimibe franchise but on the future of LDL as a surrogate endpoint and the development and approval pathway of future cardiovascular and metabolic drugs.

Positive Effect — A strong positive effect is perhaps slightly more likely than a strong negative effect. This would give a big boost to the final years of the ezetimibe franchise and would be welcomed by developers of the PCSK9s and other LDL lowering drugs, providing ammunition for their argument that LDL is a reliable surrogate that can be used to gain FDA approval.

Small, Mixed, or Ambiguous Effect — This is what most observers probably expect. There may be a small, beneficial effect that either just reaches or just misses statistical significance. Or there might be a significant effect, negative or positive, in one or another of the primary endpoint components, but no overall significance or clear signal. The possible variations of confusing outcomes are nearly innumerable.




Comments are closed.