Critical Lessons From the ENHANCE Trial

The unusual release on January 14, 2008, in the news media and on a drug company Web site,¹ of a portion of the Effect of Ezetimibe Plus Simvastatin Versus Simvastatin Alone on Atherosclerosis in the Carotid Artery (ENHANCE) trial data resulted in numerous articles and commentaries in the lay media. The availability of only fragmentary information created massive confusion and raised many more questions than answers for patients, physicians, pharmaceutical companies, and regulators. A full report of the ENHANCE trial in a peer-reviewed medical journal is not expected for months, and the first public presentation of the study's findings in a medical setting will not occur before late March 2008.

The purpose of this Commentary is to review key features of the ENHANCE study and to point to lessons that should be learned by drug companies, clinical researchers, clinicians, and health agencies from this unfortunate medical research experience.

ENHANCE was a multinational, randomized, double-blind, active comparator trial that used a surrogate end point of vascular ultrasound rather than a clinical end point such as myocardial infarction or stroke. A total of 357 heterozygous familial hypercholesterolemia patients were randomized to receive a combination therapy of 10 mg of ezetimibe and 80 mg of simvastatin, and 363 familial hypercholesterolemia patients were randomized to receive 80 mg of simvastatin alone. Approximately 80% of the patients enrolled in the ENHANCE trial had previously been treated with statins. Single-frame ultrasound images were analyzed from the right and left carotid arteries at 3 sites (the common carotid, the internal carotid, and the carotid bulb) and at several time points (baseline, 6, 12, 18, and 24 months). The end point of ENHANCE was the change in carotid intimal-medial thickness. While intimal-medial thickness is a recognized surrogate marker for coronary heart disease risk,² it is not regarded by the US Food and Drug Administration (FDA) as an end point qualifying for drug approval or new drug indications.

According to the Merck/Schering-Plough Web site,¹ there was no statistically significant difference between treatment groups on the primary end point of change in carotid intimal-medial thickness, considered as the composite measure including all carotid sites. The reported change from baseline in the mean carotid intimal-medial thickness was 0.0111 mm for the ezetimibe/simvastatin group vs 0.0058 mm for the simvastatin alone group (P = .29). At baseline, the mean carotid intimal-medial thickness measurement for ezetimibe/simvastatin was 0.68 mm and for simvastatin alone was 0.69 mm. There also was no statistically significant difference between the treatment groups for each of the components of the primary end point, including the common carotid intimal-medial thickness.

During the trial, there was a significant difference in low-density lipoprotein cholesterol lowering achieved between the treatment groups (58% reduction at 24 months in the ezetimibe/simvastatin group vs 41% reduction at 24 months in the simvastatin alone group; P < .01). This difference was entirely expected, as it was the basis for the FDA's approval of the combination drug for cholesterol lowering. The incidence of cardiovascular clinical events in ENHANCE for the ezetimibe/simvastatin group and simvastatin alone group, respectively, were reported as follows: cardiovascular death, 2/357 vs 1/363; nonfatal myocardial infarction, 3/357 vs 2/363; nonfatal stroke, 1/357 vs 1/363; and revascularization, 6/357 vs 5/363 (all nonsignificant differences given the small numbers of events). There were no noncardiovascular deaths or resuscitated cardiac arrests in the ENHANCE trial. This surrogate end point study was not powered nor designed to assess cardiovascular clinical event outcomes, a key issue for the lessons to be learned from this study.

The ENHANCE trial reportedly used a “novel non-invasive methodology” to assess intimal-medial thickness using digital single-frame ultrasound imaging of the arteries.¹ The Merck/Schering-Plough Web site states that “examination of these images was a challenging process and the data analysis took significantly longer than expected.”¹ The Web site also stated that numerous steps were taken in 2006 and 2007 to address quality issues and finalize the data analysis.

Key questions about the ENHANCE trial include: What was this study designed to prove? How would the information be used once available? Would a “negative” outcome cause a different clinical impression than a “positive” outcome? Is it possible that the ENHANCE trial was designed for marketing purposes only?

First, it is unclear what the ENHANCE study was designed to prove. Specifically, a negative outcome most likely would not lead to a different clinical conclusion than a positive outcome, and it is unlikely that the ENHANCE results could have affected approval of ezetimibe by the FDA for any indication. Indeed, ENHANCE was begun at approximately the same time that ezetimibe received FDA approval for cholesterol lowering. Ezetimibe was approved in October 2002 while the fixed-dose combination of ezetimibe and simvastatin (known in the United States as Vytorin) was approved by the FDA in July 2004. ENHANCE was started in October 2002, and the last patient visit occurred in April 2006. A clinical end point study, known as IMPROVE-IT (Examining Outcomes in Subjects With Acute Coronary Syndrome: Vytorin [Ezetimibe/Simvastatin] vs Simvastatin),³ was started in December 2004 and will not be available until approximately 2011, 9 years after FDA approval of ezetimibe as an adjunct to statins for cholesterol lowering. IMPROVE-IT is a multicenter, randomized, double-blind active comparator study that includes approximately 10 000 patients with acute coronary syndrome. Patients are randomized to receive either a combination therapy of 10 mg of ezetimibe and 40 mg of simvastatin or 40 mg of simvastatin alone and will be followed up for at least 30 months. Thus, data from ENHANCE and IMPROVE-IT will not affect FDA approval of these drugs for use in cholesterol lowering, unless significant unanticipated adverse events occur.

ENHANCE also was not conducted for the purpose of seeking a new drug indication—such as the role of ezetimibe in reducing cardiovascular events—because the study was readily acknowledged as underpowered for clinical end points, the only acceptable outcome for FDA approval of a new indication for these drugs. IMPROVE-IT should provide the necessary clinical outcomes data, but will not be completed for another 3 years at least. Surrogate end point data, such as this study was collecting, would not avoid the need to do the larger, long-term clinical trial as the drug's proponents recognized in starting IMPROVE-IT before the ENHANCE results were known.

Given these facts, one reasonable conclusion is that ENHANCE may have been undertaken primarily or exclusively as a marketing tool—with high expectations that the results would be positive and provide the company's sales force evidence of vascular benefit, even if not actual clinical benefit, to show physicians who were skeptical about the additive effect of ezetimibe combined with simvastatin. Indeed, some physicians have been critical of ezetimibe's clinical benefits for years,⁴ and the approval of Vytorin was predicted to initiate a major marketing battle in the cholesterol-lowering market due to the lack of clinical outcomes data for several of the new drugs in the market, including ezetimibe and rosuvastatin.

Further in support of the conclusion that the study was performed for marketing purposes is the fact that the study was completed in April 2006 but not reported in any form until January 2008. It seems that a much more immediate release of results would have accompanied the positive outcome that the sponsors expected. Late-breaking trial results, whether positive or negative, are always topics of major interest at cardiology meetings and in the peer-reviewed literature. Further, if the results are important for scientific reasons, they deserve to be reported as soon as possible. This issue has been widely discussed in the medical literature and has led to efforts such as routine clinical trials registration as a way of avoiding not reporting negative results.⁵ Given the circumstances surrounding this trial, the fallout for patients has been highly negative. It has provided an opportunity for much misinformation to circulate in the public media including articles questioning the entire validity of cholesterol lowering^{6 - 7} despite overwhelming evidence to support the concept as a cornerstone of cardiovascular disease prevention.^{8 - 10}

In summary, no result for the ENHANCE trial could have had any scientific or clinical importance, and the apparent delay in reporting the negative study results has raised additional concerns about the trial. If the drug company argues now in the face of negative or neutral results that a surrogate end point trial such as ENHANCE should not influence clinical practice negatively, it should be equally true that a positive result for the same type of outcome should have no positive effect on clinical medicine. In light of these facts, it appears that this trial was undertaken for marketing purposes only.

Trials designed for marketing purposes should not be conducted. Such trials divert attention and resources away from the critically important information needed from clinical end point studies, pose a serious risk to the integrity of the scientific process in clinical research, and may place patients at unnecessary risk. Physician-scientists must avoid the temptation to participate in trials that do not add to the science base in meaningful ways. Drug companies should stop conducting trials that cannot improve the state of science and cannot meaningfully add to the evidence base needed to make critical clinical decisions such as the choice of one drug over another for prevention of a clinical event. Many other similar studies are performed, and patients, physicians, and drug companies all stand to lose from studies that seem to be valuable only to drug company marketing efforts and only reported when the desired result is achieved.¹¹

The news media coverage of ENHANCE had several important factual errors. For instance, ENHANCE was a small-scale, surrogate end point trial, not a clinical end point trial. Yet, despite widespread claims to the contrary, nothing new (positive or negative) was learned about the prevention of clinical events through cholesterol lowering. The results were not negative—they were not statistically significant, or null. The end point was intimal-medial thickness, not “fatty plaque” as was misstated in some media reports. In addition, many experts cited by the news media appeared to be unaware of the actual results of the study because their statements reflected uncertainty, confusion, or incorrect assumptions.

Some newspapers published contradictory reports,^{12 - 14} confusing the public even more, whereas other sources extrapolated from this one study to paint the entire cholesterol hypothesis as wrong.⁶ In this case, the news media apparently fanned a lot of the frenzy and then seemed to stand back, unable to clear up the confusion or in some instances, added to the confusion.⁷ It is difficult to understand how such a study could mushroom in importance to the point that 50 years of serious and logical research has been damaged and defamed for no good purpose. The lesson is clear: the news media must be scrupulous and balanced in its search for factual information in medical studies and should be especially wary of results reported by companies via news releases. Likewise, medical professionals who are called on to comment on stories such as this one must carefully avoid creating confusion both by being well aware of the facts, and by not commenting without the availability of all of the relevant data.

In the aftermath of the release of the ENHANCE trial results, patients and physicians sought expert and unbiased advice about the trial. Both the American College of Cardiology (ACC)¹⁵ and the American Heart Association (AHA)¹⁶ issued statements with the goal of interpreting and explaining the incomplete and confusing results of this surrogate end point trial. These efforts, to some extent, backfired due to accusations of conflict of interest. An article with the harshest criticism¹² indicated that the House Energy and Commerce Committee was investigating both the AHA and the ACC in regard to their relationships with Merck/Schering-Plough, but also reported that both organizations stated that industry financing, which they acknowledged having received, had “nothing to do with their statements.”¹²

As past and present volunteers for both the ACC and AHA, we believe that both organizations are capable of evaluating the science alone and making truly evidence-based decisions and recommendations. Both organizations advise their volunteers and consultants to avoid conflict of interest and to report conflict of interest thoroughly. Writing groups for both organizations attempt to avoid conflict of interest wherever possible, and this is stated policy for both groups.

However, these steps are not enough, as the recent events show. These organizations must distance themselves from industry support to retain their credibility with physicians and health consumers. Without that level of credibility, the marketplace for truth is up for grabs, with no one capable of claiming the high road. As this issue has shown, a crack in the armor is capable of promoting an unwarranted reexamination of the entire topic of cholesterol lowering rather than only the pinpoint focus on a single study.^{1 ,6 - 7 ,12 - 14} Organizations such as the AHA must have credibility to advise the public in situations like this one and can only retain this credibility when there is not even a hint of conflict of interest.