The modern approach to lipid-lowering therapy is rooted in the event-driven, randomised clinical trial (RCT). From the Lipid Research Clinics Coronary Primary Prevention Trial to the Coronary Drug Project to the modern statin trials, guidelines have traditionally been informed by large RCTs driven by the outcomes that matter most—heart attacks and strokes.1 Based on a wealth of data from such RCTs, statins are now established as first-line drugs for nearly everyone considered to be at high risk of developing atherosclerotic cardiovascular disease (CVD).

Unfortunately, in the current era, large event-driven RCTs have become increasingly hard to conduct.2 The incidence of CVD is declining in developed nations, requiring either very large sample sizes (a problem compounded by the well-documented decline in RCT participation) or a largely international trial population with limited generalisability. Given the known efficacy of statins, new lipid-lowering therapies must be tested as add-ons to statins, limiting their marginal efficacy and again necessitating larger sample sizes. The global economic slowdown has also limited non-pharmaceutical and pharmaceutical funding for large RCTs, particularly for existing therapies.

The imaging-based surrogate endpoint trial was thought to be the solution to this problem.3 If one could document the benefit of a therapy on a highly sensitive atherosclerosis imaging endpoint, the theory states, one might need far fewer patients to prove the efficacy of a drug. While early statin surrogate endpoint studies showed plaque regression in close correlation with event reduction, studies of other drugs have failed to link surrogate endpoint changes to observed clinical benefits. One example was the ARBITER6-HALTS trial, which randomised patients to ezetimibe versus long-acting niacin as add-on therapy.4 This generally well-conducted trial of 363 patients was stopped early on the basis of a reduction in carotid intima-media thickness in the niacin arm. Although published in the …