SATURN and AIM-HIGH: ‘Back down to planet Earth’
Again this year, clinical trials evaluating lipoprotein-modifying therapies were in the spotlight at the American Heart Association Scientific Sessions 2011, with concurrent publications in The New England Journal of Medicine.
The Study of Coronary Atheroma by Intravascular Ultrasound: Effect of Rosuvastatin Versus Atorvastatin (SATURN) is a prospective, randomized, multicenter, double blind clinical trial led by Stephen J. Nicholls, MD, Steven E. Nissen, MD, and colleagues. These investigators performed serial intravascular ultrasound (IVUS) at baseline and after 104 weeks of treatment in 1,039 patients with coronary heart disease.
Patients were randomly assigned to treatment with atorvastatin (Lipitor, Pfizer) 80 mg daily or rosuvastatin (Crestor, AstraZeneca) 40 mg daily. The rosuvastatin group attained lower LDL levels than the atorvastatin group (63 mg/dL vs. 70 mg/dL; P<.001) and slightly higher levels of HDL (50 mg/dL vs. 49 mg/dL; P=.01); however, the primary efficacy endpoint percent atheroma volume did not differ between the groups. Coronary atherosclerosis regressed with both treatment strategies: by 0.99% (95% CI, 1.19 to 0.63) with atorvastatin and by 1.2% (95% CI, 1.52 to 0.90) with rosuvastatin. Both agents induced regression in most patients: 63% with atorvastatin and 68.5% with rosuvastatin.
In the Atherothrombosis Intervention in Metabolic Syndrome with Low HDL Cholesterol/High Triglyceride and Impact on Global Health Outcomes (AIM-HIGH) trial, 3,414 patients with established cardiovascular disease were randomly assigned to extended-release niacin 1,500 mg to 2,000 mg daily or matching placebo with 50 mg of niacin per tablet to mask treatment identity. Both groups received simvastatin (Zocor, Merck) 40 mg to 80 mg and ezetimibe (Vytorin, Merck/Schering-Plough) 10 mg daily, as needed, to maintain LDL levels between 40 mg/dL and 80 mg/dL.
The trial was stopped for a lack of efficacy after an average follow-up of 3 years. Despite an improvement in the lipid profile, including a significant increase in median HDL from 35 mg/dL to 42 mg/dL, there was no difference between the groups in the incidence of the primary composite endpoint of CHD death, nonfatal myocardial infarction, ischemic stroke, hospitalization for an acute coronary syndrome (ACS) or symptom-driven coronary or cerebral revascularization.
It was clearly a challenge to end the AIM-HIGH trial after a mean follow-up of 3 years. The rise in strokes may have been a chance finding because many of the participants assigned niacin who sustained a stroke had stopped their niacin a number of months before. The Coronary Drug Project did not find an excess rate of stroke with niacin but did report an excess incidence of atrial fibrillation, a major risk factor for stroke. The rate of AF was not reported in the AIM-HIGH participants by treatment group. The other point to consider is that the event curves in the Cholesterol and Recurrent Events (CURE) study did not separate until after 2 years; this trial compared pravastatin (Pravachol, Teva Pharmaceuticals) vs. placebo. The 3-year follow-up may not have been long enough in AIM-HIGH because the trial mandated equal LDL levels that were accomplished by a greater use of ezetimibe in those not receiving niacin. Only HDL and triglyceride levels differed between treatment groups.
Trials viewed in tandem
SATURN and AIM-HIGH address opposite sides of the coin in some respects SATURN aimed low with potent statin therapy, whereas AIM-HIGH did (as its name implies) with add-on niacin. The primary endpoint of SATURN was imaging-based and the primary endpoint of AIM-HIGH was clinical outcomes. However, at the core, both trials assessed two competing lipoprotein management strategies for patients with atherosclerotic vascular disease, with modest differential results on patients lipid profiles, and found neutral effects on the primary endpoints.
Some may be surprised, and others may not. Either way, as both camps reconvene back down on planet Earth, two key questions are:
- How do we reconcile these studies with existing literature?
- What is the bottom line when it comes to managing our patients?
Reconciling data with existing literature
Reconciling these studies may have a lot to do with endpoints. Painting with broad strokes, there is a ladder of endpoints, starting with hard clinical points, running down to soft clinical endpoints, to imaging endpoints, on down to biomarker endpoints. Ultimately, we want to be confident that if we see a change in an endpoint in response to a therapy, then this means we are benefiting patients. As it turns out, this is easier said than done. SATURN and AIM-HIGH raise important questions about surrogate imaging endpoints because SATURN employs one and AIM-HIGH was immediately preceded by a positive trial based on one: Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 6-HDL and LDL Treatment Strategies (ARBITER-6-HALTS).
Regarding SATURN, as pioneered by Drs. Nicholls and Nissen, and accepted by the expert community, IVUS quantifies plaque burden by percent atheroma volume, subtracting lumen area from the total area within the external elastic membrane divided by the total area. If different calculations are used, as in the secondary endpoint of SATURN, then a signal is seen for rosuvastatin trumping atorvastatin in atherosclerosis regression, but the clinical significance is unclear. Also unclear is the fact that the Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial failed to show plaque regression with high-dose atorvastatin, whereas SATURN showed such an effect.
Moreover, from the perspective of the individual patient, we are left with the question: Why do some patients regress on potent statin therapy, whereas others only stabilize and others even progress? Also, how much does atherosclerosis regression actually have to do with increases in HDL (15% in ASTEROID and 4% in SATURN)? This was highlighted as a possible mechanism in the publication of the A Study to Evaluate the Effect of Rosuvastatin on Intravascular Ultrasound-Derived Coronary Atheroma Burden (ASTEROID) trial, but why was the same magnitude of HDL increase not seen in SATURN?
The limitations of IVUS in assessment of atherosclerosis progression were discussed by Navin Kapur, MD, and Roger S. Blumenthal, MD, in an editorial on the ASTEROID trial in the Journal of the American Medical Association in April 2006. Indeed, we simply do not know how reliably an IVUS measurement or serial IVUS measurements track with clinical outcomes.
The same can be said for serial carotid intima-media thickness data. Although ARBITER-6-HALTS showed a 0.014-mm improvement in carotid intima-media thickness with niacin, this did not translate into a clinical outcomes benefit during a 3-year period in AIM-HIGH. This is a disappointing result, but there were a number of reasons for tempered enthusiasm after ARBITER-6-HALTS, including its premature termination and the small number of patients studied, as laid out in an editorial by Erin Michos, MD, and Blumenthal accompanying publication of the original manuscript in NEJM. The discordance between ARBITER-6-HALTS and AIM-HIGH implies that we should not be relying on serial carotid intima-media thickness testing to determine the best therapies.
Rather, we should rely on clinical outcomes data. Although AIM-HIGH did not show benefit, this does not mean that the HDL hypothesis is dead. Indeed, several clinical trials of novel agents that raise HDL are undergoing investigation. AIM-HIGH also does not mean that niacin itself should dig an early grave.
It remains to be seen whether a modest incremental benefit can be detected in the much larger ongoing niacin trial, Treatment of HDL to Reduce the Incidence of Vascular Events (HPS-2-THRIVE), with results expected in 2013. Although a clinical outcomes trial, it is notable that AIM-HIGH patients were not the highest risk, and events were less than expected, so the original primary composite endpoint was amended to include hospitalization for an ACS (softened from high-risk ACS) or symptom-driven coronary or cerebral revascularization, which ended up accounting for most of the primary endpoints in the trial.
In some respects, the inclusiveness of the composite endpoint is a good thing because doctors, patients and payers care about hospitalizations and revascularizations; however, in other respects, it is a bad thing because it is a softer or less reliable endpoint because more subjectivity enters the equation.
It is not unrealistic to think that HPS-2-THRIVE could show a benefit. Even if it does, we must point out that it does not necessarily follow that the benefit is from HDL raising. Although niacin is associated with increased HDL concentrations, this does not tell us about reverse cholesterol transport and HDL functionality.
Niacin also lowers LDL or, perhaps more importantly, atherogenic lipoprotein particles. Through the particle lens, a clinical effect of niacin may be more clearly understood because the impact on atherogenic particles is modest, so a very large trial may be needed to detect an effect, especially in patients who are already aggressively treated.
The bottom line
At the end of the day, in our humble opinion, the message boils down to the same message that was loud and clear before these studies: In managing high-risk patients with atherosclerotic vascular disease, the primary emphasis should be on atherogenic lipoprotein-lowering through appropriately potent statin therapy. We predict that atorvastatin will continue to be the dominant statin given physicians and patients comfort with it, its compatibility with calcium channel blockers and because it approaches rosuvastatin in potency. Cost is a huge issue for patients and the health care system, especially today, and is tied to adherence, so the transition of atorvastatin to generic form will be another factor that makes this medication the dominant one in the statin market for many years to come.
As for niacin, for the time being, although it did not show a benefit as an add-on therapy in patients with low LDL in AIM-HIGH, it is still a reasonable agent for lipoprotein-lowering in patients who cannot tolerate a statin and as an add-on agent for patients above treatment targets despite maximally tolerated statin therapy.
Seth S. Martin, MD, is a cardiology fellow at The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease. Roger S. Blumenthal, MD, is professor of medicine and director of The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease.
For more information:
- Boden W. LBCT.04.
- Nicholls S. LBCT.04. Both presented at: American Heart Association Scientific Sessions 2011; Nov. 12-16; Orlando, Fla.
Disclosure: Drs. Martin and Blumenthal report no relevant financial disclosures.