News · Heart & Metabolic
A cholesterol drug cut deaths from any cause by 20% in people who had never had a heart attack
Cholesterol drugs are judged on heart attacks and strokes, and rarely move the count of deaths at all. Across 12,301 people with no prior event, this one moved it, and moved deaths that had nothing to do with the heart.
- People given the drug were 20% less likely to die of anything over about five years.
- That is rare: cholesterol drugs usually move heart events without moving deaths.
- Deaths unrelated to the heart also fell, which nobody can currently explain.
- None of these patients had previously had a heart attack or stroke.
- 12,301 people randomly assigned, followed a median of 4.6 years.
Cholesterol trials almost never move the death count. They move heart attacks, they move strokes, they move the composite endpoints built out of both, and when the researchers get to the line reporting how many people died in each group, the two numbers usually sit close together.
There is a mundane reason. Deaths are rarer than heart attacks, so a trial sized to detect a difference in events is too small to detect one in deaths. Absence of a mortality result is normally absence of statistical power rather than absence of benefit.
Which is why the analysis published in Circulation stands out. Among patients at high cardiovascular risk with no previous myocardial infarction or stroke, all-cause mortality rates were 20% lower with evolocumab than with placebo.
What the numbers actually are
Over a median of 4.6 years, 973 patients died. The split was 434 deaths on the drug against 539 on placebo.
Expressed as a five-year rate, that is 7.9% against 9.7%. Just under two percentage points, or roughly one death avoided for every 55 people treated across five years.
For primary prevention, which is treating people who have not yet had an event, that is a substantial figure. Most of what is offered to this group is justified on the events it prevents rather than on the lives it saves.
The result nobody has an explanation for
Underneath the headline the deaths split into categories, and this is where it gets strange.
Cardiovascular deaths fell, at 156 against 195, which is exactly what the drug is supposed to do. Non-cardiovascular deaths also fell, 229 against 268, and there is no accepted mechanism for that.
Lowering LDL cholesterol acts on arteries. It is not supposed to reduce deaths from causes that have nothing to do with arteries, and a long history of cholesterol research has specifically looked for and failed to find such effects.
The honest reading is caution. The non-cardiovascular interval reaches up to no effect at all, so this may simply be the play of chance in a subgroup, and 13% of the deaths were of undetermined cause, which blurs the categories. It is the single most replication-hungry number in the paper.
Why this population is the interesting one
Evolocumab already had good evidence in people who had survived a heart attack. The harder question has always been whether to use it in people who have not had one.
That group is larger, healthier, and further from the point where aggressive treatment obviously pays. Their absolute risk is lower, so the same relative benefit saves fewer people per hundred treated, and the drug is expensive and injected rather than swallowed.
A mortality signal changes that calculation, because mortality is the outcome nobody argues about. Softer endpoints can be redefined; deaths cannot.
What the drug does
PCSK9 inhibitors are a different class from statins. Cholesterol medicines work in several ways, and this one is an injected antibody blocking a protein that would otherwise destroy the liver’s own LDL receptors.
Leave those receptors intact and the liver keeps pulling LDL out of the blood. The result is LDL reduction far beyond what a statin achieves alone, which is the reason these drugs exist and the reason they are reserved for people whose numbers stay high despite everything else.
The caveats worth carrying
This is a prespecified analysis of one trial, not a separate trial. Prespecified matters, because it means the question was written down before the data were seen rather than found by searching afterwards, but it remains one dataset.
Trial populations are also not the general population. People enrolled in cardiovascular trials are screened, monitored and adherent in ways ordinary patients are not, and the benefit measured under those conditions is an upper bound on what routine care delivers.
And a five-year window says nothing about twenty years on an injected antibody.
Where it leaves things
Not with a recommendation for anyone in particular. The people who might take this drug are identified by their risk and their cholesterol, and that assessment has not changed.
What has changed is the weight of the argument for treating high-risk people before their first event rather than after it. A 20% reduction in dying, from a drug tested in people who had never had a heart attack, is the kind of result that moves guidelines rather than prescriptions.
People also ask
What did the analysis find?
All-cause mortality was 20% lower with evolocumab than placebo: 434 deaths (5-year Kaplan-Meier rate 7.9%) versus 539 (9.7%), hazard ratio 0.80 (95% CI 0.70-0.91; P = 0.0005). Cardiovascular death was lower (HR 0.79; 0.64-0.98) and so was non-cardiovascular death (HR 0.85; 0.71-1.01).
Why is a mortality result unusual here?
Because cholesterol-lowering trials are usually powered for heart attacks and strokes, not deaths. Deaths are rarer, so most such trials cannot detect a difference even when one exists. A clear mortality signal in primary prevention is uncommon.
What is evolocumab?
A PCSK9 inhibitor: an injected antibody that stops the liver breaking down its own LDL receptors, so more LDL cholesterol is cleared from the blood. It lowers LDL far further than a statin alone.
Why does the non-cardiovascular death result raise eyebrows?
Because there is no accepted mechanism by which lowering cholesterol should reduce deaths from causes unrelated to blood vessels. Its interval brushes no effect, so it may be chance, and it is the part of the result that most needs replication.
Who were the patients?
People at high cardiovascular risk who had not yet had a heart attack or stroke. That is primary prevention, a group where the absolute benefit of intensive cholesterol lowering has been harder to demonstrate than in people who already have had an event.
How large is the benefit in practice?
The five-year death rate was 7.9% on the drug against 9.7% on placebo, a difference of about 1.8 percentage points, or roughly one death avoided for every 55 people treated over five years.
Does this mean everyone should take it?
No. These drugs are expensive, injected, and reserved for people whose risk is high or whose cholesterol stays high on other treatment. Who should take one is a clinical decision. This is general information rather than medical advice.