When a Drug Fixes the Biomarker but Not the Outcome

Why does lowering a biomarker not always reduce heart attacks?

A surrogate endpoint is a laboratory or imaging measure used in place of a clinical event such as heart attack, stroke or death. Regulators sometimes accept surrogates for approval when the link to outcomes is well established, and the approach speeds up trials. The risk is that a marker can be associated with disease without being a cause of it, or that a drug can change the marker through a mechanism that does not alter the underlying disease process.

Even when a marker is causal, the size and timing of the benefit matter. Atherosclerosis develops over decades, so lowering a risk factor for a few years in older patients with established plaque may produce a smaller effect than the same change would across a lifetime. Trials that run too briefly, or that enroll patients who have already accumulated large plaque burden, can therefore miss a benefit that would appear with earlier or longer exposure.

Which cardiovascular trials show the gap between markers and outcomes?

The clearest success is LDL cholesterol. Meta-analyses by the Cholesterol Treatment Trialists' Collaboration showed that statins reduce major vascular events in proportion to the LDL reduction achieved, and later trials of ezetimibe and PCSK9 inhibitors such as evolocumab supported the same principle. That body of evidence is why LDL-C is widely accepted as a causal target.

HDL cholesterol went the other way. Epidemiology showed that higher HDL was associated with lower risk, yet the ILLUMINATE trial of torcetrapib, a CETP inhibitor, was stopped early because of excess deaths and cardiovascular events despite a large rise in HDL. Niacin added to statins in the AIM-HIGH and HPS2-THRIVE trials raised HDL but did not reduce cardiovascular events on top of statin therapy. These results reshaped how developers think about biomarkers.

What does this mean for newer drugs and for patients?

Newer therapies aimed at inherited risk markers, including Lp(a), face the same question: is the marker a bystander or a driver, and is the trial long and large enough to detect benefit? Results presented at the American Heart Association meeting showing that one Novartis Lp(a)-lowering drug reduced the marker without lowering heart attack or stroke risk illustrate why developers run large outcome trials rather than relying on laboratory changes alone. Such results do not necessarily doom the broader strategy, since different drugs, different degrees of lowering, and different populations may produce different results, but they do temper expectations.

For patients, the practical message is to continue proven preventive measures, including LDL lowering, blood pressure control, smoking cessation and physical activity, and to discuss any new therapy in terms of the outcomes it has been shown to improve. Clinicians should be cautious about off-label use of marker-lowering treatments before outcome data are available.