PCSK9 Inhibitors: A Different Mechanism, the Same Unanswered Questions

Cardiologists are pushing PCSK9 inhibitors — Repatha (evolocumab), Praluent (alirocumab) — as the next generation of cholesterol-lowering therapy. Patients who can't tolerate statins or whose LDL won't budge are told these drugs are the answer. They are injectable monoclonal antibodies that cost thousands of dollars a year and drive LDL to levels the human body has never routinely operated at.

They are not statins. They don't block the mevalonate pathway. They don't suppress CoQ10 or K2 or hormone synthesis. That distinction matters and should be stated clearly. But a different mechanism does not mean a clean bill of health. The questions that follow are different from those on the statins page — but they are no less important.

How PCSK9 Inhibitors Work

PCSK9 (Proprotein Convertase Subtilisin/Kexin type 9) is an enzyme produced primarily by the liver. Its job is to bind to LDL receptors on hepatocyte surfaces and tag them for destruction. When PCSK9 is active, LDL receptors are degraded — fewer receptors survive, less LDL gets cleared from the blood, and circulating LDL rises.

PCSK9 inhibitors are monoclonal antibodies that bind to PCSK9 in the bloodstream, preventing it from reaching LDL receptors. The result: more LDL receptors survive on liver cells, the liver clears more LDL from circulation, and serum LDL drops dramatically — often by 50–60% on top of whatever a statin has already achieved.

What they don't do

Unlike statins, PCSK9 inhibitors:

Do not block HMG-CoA reductase — the mevalonate pathway remains open
Do not suppress CoQ10 synthesis — mitochondrial energy production is not directly impaired
Do not reduce K2 (MK-4) synthesis — the GGPP → UBIAD1 → MK-4 pathway is unaffected
Do not impair dolichol, heme A, or prenylation — the non-sterol branches are left intact

This is a genuine advantage over statins and should be acknowledged. The mechanism is targeted: increase LDL clearance without shutting down a multi-branch biosynthetic pathway.

What they do

They drive LDL to levels that are historically unprecedented in adult humans. The FOURIER trial achieved a median LDL of 30 mg/dL. Some patients reached levels below 20 mg/dL. The question is not whether they lower LDL — they do, dramatically. The question is what happens when you push a molecule that the body synthesizes, regulates, and uses for multiple purposes to near-zero circulating levels.

The FOURIER Trial: What It Found and What It Didn't

The FOURIER trial was the landmark outcomes study for evolocumab (Repatha). It enrolled 27,564 patients with established atherosclerotic cardiovascular disease who were already on statin therapy. The study ran for a median of 2.2 years.

What it found

LDL dropped from a median of 92 mg/dL to 30 mg/dL
The primary composite endpoint (cardiovascular death, heart attack, stroke, hospitalization for unstable angina, coronary revascularization) was reduced by 15% (relative risk reduction)
Heart attacks were reduced by 27% (relative)
Strokes were reduced by 21% (relative)

What it didn't find

No reduction in cardiovascular mortality. The drug reduced some nonfatal events but did not reduce death from cardiovascular causes.
No reduction in all-cause mortality. Despite driving LDL to 30 mg/dL in a high-risk population already on statins, there was no mortality benefit.

This is a critical finding. If LDL is truly the causal agent of cardiovascular disease — if the relationship is linear and "lower is better" as current guidelines assert — then a 70% additional reduction in LDL, in patients who already had events, should produce a mortality signal. The absence of one, after 2.2 years in nearly 28,000 patients, is not easily explained by the LDL hypothesis alone.

The NNT picture

Outcome	2-Year NNT
Primary composite endpoint	67
Heart attack alone	100
Stroke alone	200
Cardiovascular death	No benefit
All-cause death	No benefit

For context: this is a secondary prevention population — patients who have already had cardiovascular events. The NNTs are better than primary prevention statins because the baseline risk is higher. But the mortality result is the same: the drug reduces some nonfatal events without keeping anyone alive longer, at least within the trial timeframe.

ODYSSEY OUTCOMES

The alirocumab (Praluent) trial showed similar results. A post-hoc analysis of the highest-risk subgroup suggested a possible mortality signal, but post-hoc subgroup analyses are hypothesis-generating, not confirmatory. The overall trial did not demonstrate a statistically significant mortality benefit.

The Very Low LDL Question

PCSK9 inhibitors routinely push LDL below 30 mg/dL. Some patients reach the teens. This raises questions that don't exist at higher LDL levels.

Cholesterol is not just a transport particle

LDL particles carry cholesterol to peripheral tissues that need it. Cholesterol is:

The structural backbone of every cell membrane
The precursor for every steroid hormone — testosterone, estrogen, progesterone, cortisol, DHEA, aldosterone
The substrate for vitamin D synthesis — 7-dehydrocholesterol → pre-vitamin D3 in the skin
A precursor for bile acid synthesis — essential for fat digestion and nutrient absorption
Critical for myelin formation and maintenance in the nervous system

Driving circulating LDL toward zero does not just remove a risk factor. It reduces the delivery vehicle for a molecule that every cell in the body uses.

The U-shaped curve

Multiple observational studies have documented a U-shaped relationship between LDL levels and adverse outcomes:

A UK Biobank study of 147,000+ participants found that both very high and very low LDL were associated with worse cognitive performance. LDL mediated more than 50% of the observed association between statin use and cognitive decline.
Meta-analyses have found associations between very low LDL and increased risk of hemorrhagic stroke. The brain's vascular integrity depends in part on membrane cholesterol content.
Observational data in elderly populations consistently shows that low total cholesterol is associated with increased all-cause mortality — a finding that has been replicated across multiple cohorts and countries.

The standard response is that correlation is not causation, and that sick people have low cholesterol because of their illness, not the reverse. That may explain some of the signal. It does not explain why pharmacologically driving LDL to these same low levels fails to produce mortality benefits.

Hormones and substrate availability

PCSK9 inhibitors don't block cholesterol synthesis (that's what statins do). But they aggressively clear cholesterol-carrying LDL particles from circulation. Peripheral tissues that normally take up LDL for local cholesterol needs — the adrenals, the gonads, the skin — may have less substrate available.

This concern is theoretical at the population level — we don't yet have large-scale data on hormone levels in long-term PCSK9 inhibitor users. But the biochemistry raises a question that should be studied rather than assumed away. When you reduce LDL by 70% on top of a statin that already reduced synthesis, what happens to the tissues that needed that cholesterol?

The Brain at Very Low LDL

The brain synthesizes most of its own cholesterol behind the blood-brain barrier. This is the argument used to dismiss neurological concerns: "the brain makes its own, so circulating LDL doesn't matter."

Why it's not that simple

LDL receptors exist on the blood-brain barrier. While the BBB is largely impermeable to LDL particles, receptor-mediated transport does occur. The degree to which peripheral LDL contributes to brain cholesterol homeostasis — especially during periods of high demand like myelin repair — is not fully characterized.
The EBBINGHAUS substudy of FOURIER tested cognitive function in ~1,974 patients over 19 months and found no significant decline. This is routinely cited as reassurance. But 19 months is not a meaningful timeframe for detecting progressive cognitive effects. Neurodegenerative processes unfold over years to decades.
The FDA required neurocognitive monitoring in PCSK9 inhibitor trials after early post-marketing reports of cognitive complaints. The fact that monitoring was required suggests the concern was serious enough to warrant investigation. The fact that the investigation was short-term does not constitute long-term clearance.
The UK Biobank data showing a U-shaped LDL-cognition relationship doesn't distinguish by mechanism of LDL lowering. It suggests that the level itself may matter, regardless of how you got there.

What we don't have

We do not have 5-year, 10-year, or 20-year cognitive data on humans maintained at LDL levels below 30 mg/dL. The drugs have not existed long enough. The absence of long-term safety data is not the same as evidence of long-term safety.

The Stacking Problem

In practice, PCSK9 inhibitors are almost never prescribed alone. They are added on top of maximally tolerated statin therapy. This means the patient receives:

Mevalonate pathway suppression (from the statin) — CoQ10 depletion, K2 suppression, reduced prenylation, impaired hormone substrate
Aggressive LDL clearance (from the PCSK9 inhibitor) — driving whatever circulating cholesterol remains to historically low levels

The stacking compounds the substrate problem. The statin reduces cholesterol production. The PCSK9 inhibitor accelerates cholesterol clearance. The patient gets less made and more removed simultaneously. The downstream effects documented on the statins page — calcification, CoQ10 depletion, hormone disruption — are not mitigated by adding a PCSK9 inhibitor. They may be compounded by further reducing available cholesterol.

No major trial has studied whether PCSK9 inhibitor patients on statins have worse K2 status, lower CoQ10, or greater calcification progression than patients on statins alone. The question has not been asked.

The Cost and Incentive Structure

Pricing

Drug	Annual Cost (US)
Generic atorvastatin (statin)	~$48
Repatha (evolocumab)	~$5,800
Praluent (alirocumab)	~$5,800
Leqvio (inclisiran)	~$6,500

PCSK9 inhibitors are among the most expensive cardiovascular drugs ever marketed. The price has come down from the original ~$14,000/year after pushback from insurers and pharmacy benefit managers, but remains orders of magnitude more expensive than generic statins.

The marketing push

When a cardiologist recommends Repatha, that recommendation occurs in a context:

Pharmaceutical representatives detail PCSK9 inhibitors heavily to cardiologists
Speaker programs, advisory boards, and consulting fees create financial relationships
Direct-to-consumer advertising drives patient requests
Prior authorization requirements create the impression that insurers are withholding valuable treatment

None of this means the drug is ineffective. It means the enthusiasm for the drug should be evaluated in light of the financial ecosystem that generates it. A drug that costs $5,800/year and does not reduce mortality generates different commercial incentives than a generic that costs $48/year.

The inclisiran expansion

Inclisiran (Leqvio) is a newer approach — an siRNA (small interfering RNA) that silences PCSK9 production in the liver, rather than blocking it in the bloodstream. It requires only two injections per year. Novartis has positioned it for even broader use than monoclonal antibody PCSK9 inhibitors.

The same questions apply: LDL goes very low, there is no mortality data yet from outcomes trials (ORION-4 is ongoing), and the cost is higher than the drugs it stacks on top of.

The Natural PCSK9 Argument

Proponents of aggressive LDL lowering frequently cite people with naturally occurring PCSK9 loss-of-function mutations. These individuals have lifelong very low LDL and appear to have lower cardiovascular risk. The argument: "Nature already tested this — low PCSK9 is safe and protective."

Why the analogy breaks down

Lifetime adaptation vs. pharmacological intervention. A person born with low PCSK9 develops from conception with low LDL. Every cell, every tissue, every developmental process unfolds in that context. The body compensates. This is categorically different from taking a 60-year-old with normal LDL and pharmacologically crashing it by 70%.
These mutations are rare. The populations studied are small. Long-term outcomes data at scale don't exist for the mutations, let alone the drugs that mimic them.
Survivorship bias. We study people who survived to adulthood with these mutations. We don't know the full reproductive and developmental consequences across a large population.
Compensatory mechanisms. People with lifelong low LDL may upregulate alternative cholesterol delivery mechanisms. A drug that rapidly depletes LDL doesn't give the body decades to adapt.

Citing natural variants as proof of drug safety is like arguing that because some people thrive on 4 hours of sleep, pharmacologically suppressing sleep to 4 hours would be safe. Genetic adaptation is not the same as pharmacological intervention.

What Should Be Studied But Hasn't Been

Hormone panels in long-term PCSK9 inhibitor users — testosterone, estrogen, DHEA, cortisol, vitamin D status. If cholesterol substrate matters, these should show signal.
Cognitive function beyond 2 years — 19-month follow-up in EBBINGHAUS is insufficient for drugs intended for decades of use.
Interaction with statin-induced K2 depletion — does further LDL reduction compound the calcification mechanism?
Hemorrhagic stroke risk at very low LDL — the observational signal exists but has not been prospectively studied at LDL levels below 25 mg/dL.
Quality of life and functional outcomes — trials measure cardiovascular events. They don't routinely capture fatigue, cognitive fog, libido changes, or the subjective experience of patients maintained at extreme LDL levels.

Conclusion

PCSK9 inhibitors are not statins. They don't shut down the mevalonate pathway. That is a real and meaningful distinction — and for patients who cannot tolerate statins, it may be relevant.

But they share a foundational assumption with statins: that LDL is the enemy, and lower is always better. The FOURIER trial tested this aggressively — driving LDL to 30 mg/dL in 28,000 high-risk patients — and found no mortality benefit. The U-shaped observational data on very low LDL and cognitive function, hemorrhagic stroke, and all-cause mortality has not been reconciled with the "lower is better" claim.

The drugs reduce some nonfatal cardiovascular events. They do not keep people alive longer. They cost $5,800/year. They are prescribed on top of statins that are already suppressing a multi-branch biosynthetic pathway. And they push a molecule the body synthesizes, regulates, and uses for dozens of purposes to levels that have never been tested over a full human lifespan.

The right response to PCSK9 inhibitors is not the same as the right response to statins. The mevalonate pathway concerns are different. But the questions about what we're doing to cholesterol — why the body makes it, what it's for, and what happens when we aggressively remove it — are the same questions, asked at a deeper level.