AIM-HIGH - Wrong weapon, wrong target, or both?

Ever since the news broke earlier this year that the National Heart, Lung, and Blood Institute (NHLBI) would be terminating the AIM-HIGH trial of extended-release niacin early due to lack of efficacy,  clinicians have been eagerly awaiting release of the full results.  Following their publication last week in conjunction with the 2011 American Heart Association meeting in Orlando, FL, the results of AIM-HIGH [1] have been the subject of intense controversy.

Although statins have been shown to reduce the risks associated with coronary artery disease, some residual risk remains even in patients who achieve target concentrations of low-density lipoprotein (LDL).  Some of this risk has been attributed to low concentrations of high-density lipoprotein (HDL), which have been shown in epidemiologic studies to be an independent predictor of cardiovascular risk.  As a result, strategies shown to improve HDL concentrations, such as the use of niacin, have recently garnered significant attention.  The purpose of AIM-HIGH was to evaluate whether the addition of extended-release niacin (Niaspan^®) to intensive statin therapy (i.e., targeting LDL concentrations of 40-80 mg/dL) would further reduce cardiovascular events.

However, the study was terminated 18 months early when a data and safety monitoring board found that the threshold for inefficacy had been met and a small (but not statistically significant) increase in ischemic strokes was observed in the group randomized to niacin therapy.  The decision to terminate the trial early, as well as the conclusions reached by the trial investigators, have been fiercely contested ever since.

The most frequently cited criticism of the trial is that it was stopped far too early to demonstrate the 25% expected reduction in outcomes upon which the trial's statistical power was based.  Some have contended that such differences would require a decade or more to become statistically significant.  Other factors that may have impacted the study's power to detect a difference have also been the subject of debate, including the dropout rate (which exceeded 25% in the niacin group), the increase in HDL observed in the placebo group (which resulted in a net difference in HDL at three years of only 5 mg/dL between the two groups). Other criticisms include the use of small doses (< 50 mg) of niacin in the placebo group (i.e., to induce the common "flushing" effect that would have otherwise unblinded patients randomized to the niacin group) and differences in the dose-adjustments of simvastatin necessary to maintain target lipoprotein levels.

In summary, more questions than answers have emerged out of the publication of AIM-HIGH.  I also awaited the full results to see how it might influence my practice, as we use a fair amount of niacin at our institution.  In short, it probably won't have too much of an impact... for two reasons.

First, I have not routinely recommended the addition of niacin solely for the purpose of improving HDL.  In the vast majority of cases, there are clinical reasons that preclude its use (e.g., the patient is far from achieving a target LDL concentration) or there are more compelling patient-specific strategies for improving HDL (e.g., dietary modification, increased physical activity, moderation of alcohol intake, smoking cessation) rather than adding yet another drug. Where I have found myself recommending niacin is in patients without diabetes (or in whom hyperglycemia is well-controlled) who, despite having LDL concentrations at goal, have significantly elevated triglycerides (> 200 g/dL).  Triglycerides carry significant atherogenic risk and elevated concentrations have been shown to be an independent predictor of heart disease [2]; therefore, in patients with LDL concentrations at goal and in whom other risk factors are well-controlled, the addition of niacin to target isolated triglyceride elevations still seems reasonable.

The second reason why I have not recommended the use of niacin for improving HDL concentrations is that I am not convinced that targeting HDL with pharmacologic strategies (niacin or otherwise) is associated with compelling improvements in outcomes. From what limited data does exist, the exact role of HDL appears far more complex than we initially thought, making strategies for targeting it less intuitive than the management of other dyslipidemias.  However, ongoing focus has been placed on strategies to impact HDL, including the use of cholesteryl ester transfer protein (CETP) inhibitors (e.g., evacetrapib) and HDL infusions, both of which are still too early in studies to demonstrate significant results. There is still a part of me that believes HDL may not be as simple a surrogate marker to target as LDL -- instead, it may be a more global marker of atherogenic risk and perhaps our efforts would be better spent emphasizing those non-pharmacologic strategies known to improve HDL (e.g., physical activity, smoking cessation) but are also associated with independent benefits in reducing cardiovascular risk.

Critics of AIM-HIGH now await the results of HPS2-THRIVE, which recently completed patient enrollment and pairs niacin with laropiprant, a prostaglandin D antagonist that reduces the problematic cutaneous effects associated with niacin therapy. In the meantime, the two lessons I have taken away from AIM-HIGH are 1) statins remain the most proven strategy for impacting cardiovascular risk and should always be attempted first; and 2) no pharmacologic strategy targeting HDL in patients with LDL concentrations at goal has yet to be associated with overall improvements in cardiovascular risk.

References

1. The AIM-HIGH Investigators. Niacin in Patients with Low HDL Cholesterol Levels Receiving Intensive Statin Therapy. N Engl J Med. 2011 Nov 15.
2. Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81:7B-12B.