“PCSK9 inhibitor” now names four different kinds of medicine — two injectable antibodies, a twice-yearly injection that works by gene silencing, and, in the United States, a daily tablet. The evidence, the prescribing route and the eligibility threshold differ between them in ways that are widely misunderstood, including by clinicians.
And it was followed up, which is the part usually missing. EBBINGHAUS-OLE followed 473 of those participants for a further median 5.1 years — up to 7.2 years from randomisation — and again found no cognitive signal. That is the strongest reassurance available on this question, and it is longer follow-up than most people asking about it realise exists.
The gap between eligibility and use is real. In one large English region, roughly 1,900 people were receiving inclisiran against roughly 3,700 estimated eligible — about half — years after the commissioning arrangement was built precisely to remove the access barriers. Checked August 2026.
What this page will not do. It will not tell you to start, stop or change anything. Every route on this page ends at a prescriber — for the antibodies, a lipid specialist; for inclisiran, usually your own GP. What it can do is make sure the conversation starts from an accurate map.
1. The biology, told through the people born without it
PCSK9 is a protein made mainly by the liver. Its day job is to bind the LDL receptor — the molecular hook that pulls LDL out of the blood — and mark it for destruction. More PCSK9 means fewer hooks and higher LDL; less PCSK9 means more hooks and lower LDL. It was discovered the way the best drug targets are: by finding the humans in whom the experiment has already run for a lifetime.
In 2003, French families with severe inherited high cholesterol were found to carry overactive PCSK9 variants — the third gene ever linked to familial hypercholesterolaemia. Three years later came the other half. In a large American cohort, roughly 1 in 40 Black participants carried a variant that switched one PCSK9 copy off. Their LDL ran about 28% lower — and their coronary heart disease risk was 88% lower, far more than the LDL difference alone predicts, because the exposure lasted from birth. A milder variant in white participants cut LDL by about 15% and coronary risk by 47%.
And a small number of people have been found with no working PCSK9 at all: LDL around 0.4 mmol/L, and — this is the part that matters for the safety section — healthy, with normal development, cognition and fertility as far as has been reported. Strong for the causal direction, and it is the same cumulative-exposure argument the ezetimibe chapter makes about NPC1L1.
2. The four ways to block it status checked 15 August 2026
| Medicine | What it is | Dosing | LDL-C | UK status |
|---|---|---|---|---|
| Evolocumab (Repatha) Alirocumab (Praluent) |
Monoclonal antibodies binding PCSK9 in the blood | Self-injected fortnightly; monthly options exist | ~60% on top of a statin | Available — specialist initiation only (TA394 / TA393) |
| Inclisiran (Leqvio) | Small interfering RNA — silences the liver's PCSK9 production | Day 0, month 3, then twice a year | ~50% | Available — can be started in primary care (TA733) |
| Enlicitide (Lipfendra) | Oral macrocyclic peptide — the first PCSK9 tablet | 20 mg once daily | 56% and 59% placebo-adjusted at 24 weeks | US only — FDA 16 July 2026; no UK licence, no NICE appraisal, no timetable |
| Lerodalcibep (Lerochol) | Small engineered binding protein | Monthly injection | Substantial, of the same order as the antibodies | US only — no UK availability |
And one thing that is not a medicine yet: a single-dose gene-editing treatment that rewrites one letter of the PCSK9 gene in liver cells is in early human trials, covered in section 8 of the ezetimibe chapter. Listed here for completeness, not as an option — it is years from any clinic, and that section explains why “permanent” is both the promise and the entire question.
3. What is proven, what is not — and the one that failed
The antibodies have cardiovascular outcome evidence. Everything else so far has LDL evidence. That distinction is the spine of this page, and this site applies it to friendly and unfriendly findings alike.
- Evolocumab — FOURIER (2017). Strong 27,564 people with established cardiovascular disease on statins, median 2.2 years. LDL fell from 2.4 to about 0.8 mmol/L, and major cardiovascular events fell by 15% (hazard ratio 0.85). No effect on mortality at that trial length — which is what a 2.2-year window predicts, not a verdict on the drug.
- Alirocumab — ODYSSEY OUTCOMES (2018). Strong 18,924 people one to twelve months after an acute coronary syndrome, median 2.8 years: events down 15%, an absolute reduction of about 1.6 percentage points.
- Inclisiran. Emerging for outcomes. LDL lowering is beyond doubt; ORION-4 and VICTORION-2P have not reported. This site uses inclisiran as its standing example of the difference between a marker and an outcome, and liking the mechanism does not change that.
- Enlicitide. Emerging for outcomes. Approved in the US on LDL reduction; CORALreef Outcomes has not reported. A second example does not strengthen the surrogate — it clarifies the pattern.
- Bococizumab — the cautionary tale. A third antibody was abandoned in 2016 after a large programme: patients developed antibodies against the drug, its effect faded, and injection reactions were common. It shows the class is not automatically interchangeable — bococizumab was part-mouse where the survivors are fully human, which is the likeliest explanation and the reason the others are watched for the same problem.
Supporting, not outcome, evidence. Imaging trials show these drugs change the plaque itself — regression of plaque volume, and thickening of the fibrous cap that separates plaque contents from the bloodstream. That is mechanistically encouraging and it rhymes with the discussion of plaque quality — but imaging is a surrogate too, and it sits below the event trials rather than beside them.
4. Safety — the questions clinicians actually ask
“Is LDL that low safe?” Three answers point the same way. The people born with little or no PCSK9, with lifelong LDL far below anything a tablet achieves, appear healthy. A dedicated cognitive sub-study of the evolocumab outcome trial — EBBINGHAUS, 1,974 enrolled and 1,204 in the primary analysis cohort, formally tested over time — found no difference in memory or executive function, including in those reaching very low LDL. And neither large outcome programme showed an excess of haemorrhagic stroke, cataract or neurocognitive events. The honest caveat is duration: the randomised evidence covers years, not decades; the genetics covers decades, but in people who never had high LDL to begin with.
“Do they cause diabetes like statins?” The trials say no — new-onset diabetes was not increased in either. The genetics says be a little humble: people carrying LDL-lowering PCSK9 variants show a slightly higher type 2 diabetes risk, much as carriers of statin-pathway variants do, suggesting a small glucose effect may travel with LDL lowering itself by any mechanism. If it exists for these drugs it is too small to see in trials of tens of thousands — and, as with statins, the cardiovascular arithmetic is not close.
What is actually seen: injection-site reactions — redness, soreness, usually mild and short-lived — for all the injectables, and occasional flu-like symptoms. Muscle symptoms are not a feature of this class, which is precisely why the pathway offers these drugs to people with genuine statin intolerance.
Two practical points that surprise clinicians
- No routine blood monitoring is required. No liver function tests, no creatine kinase. The follow-up blood test is a lipid panel to confirm the drug is working, typically around three months in.
- Pregnancy is a stop-and-discuss. These medicines are not for use in pregnancy or when planning one — the antibodies cross the placenta in late pregnancy as all antibodies do, and the data are thin. Plan it in advance with the prescriber rather than discovering it.
5. Who qualifies in the UK — the table nobody carries in their head
All figures are LDL-C, persistent despite maximum tolerated statin with or without ezetimibe, or despite other lipid-lowering therapy where statins are not tolerated or contraindicated. Criteria as checked August 2026; the technology appraisal outranks this table.
| Situation | Antibodies (TA393 / TA394) | Inclisiran (TA733) |
|---|---|---|
| History of cardiovascular events | LDL-C above 4.0 at high risk, or above 3.5 at very high risk — meaning recurrent events or disease in more than one vascular bed. Above 3.5 with heterozygous FH | LDL-C persistently 2.6 or more |
| No cardiovascular events | Heterozygous FH with LDL-C above 5.0 | Not recommended |
| Who starts it | Specialist services only | Primary care can initiate — appraised and commissioned for exactly that, with national funding arranged so local budgets are not the barrier |
If you are that person and have been told there is nothing more to add, the accurate question is: “My LDL is persistently above 2.6 despite maximum treatment and I have had a cardiovascular event — does NICE TA733 apply to me?” If your practice does not offer it, asking where locally does is a reasonable follow-up. None of this is a demand for the drug; it is a request for the assessment the guidance describes.
Where this sits in the ladder: statin → ezetimibe → bempedoic acid where statins are not tolerated → then this page. The full sequence, with what each step is worth, is in the ezetimibe chapter, and the clinical depth covers how statin intolerance is properly confirmed before the ladder is climbed.
6. Three things that rarely make the summary
- They lower Lp(a), a little. The antibodies reduce lipoprotein(a) by a modest amount, and inclisiran similarly. That is not enough to treat a genuinely high Lp(a), it is not an approved use, and the drugs designed for that job are still in trials — see the genetics chapter and the Lp(a)HORIZON readout expected before the end of 2026. But it is a real effect, and it may be part of why the outcome benefit slightly outruns the LDL arithmetic.
- They work without a statin. The LDL reduction is broadly preserved as monotherapy, which is the whole basis of the statin-intolerance route. Combined with a statin and ezetimibe, total reductions around 85% are achievable — numbers that were simply not available a decade ago.
- Non-response has a short list. If the lipid panel does not move: injection technique, storage — the antibodies live in the fridge — missed doses, and, rarely, anti-drug antibodies. It is a conversation, not a mystery.
7. What exists elsewhere status checked 15 August 2026
The oral option — enlicitide, approved in the US on 16 July 2026 — and the monthly protein lerodalcibep are covered with full status in section 7 of the ezetimibe chapter, and summarised in the table above. The one-line version for a UK reader: the tablet exists, it is not available here, no UK timetable has been announced, and its outcome trial has not reported — the same position inclisiran occupies, and the same discipline applies.
NICE is separately appraising obicetrapib, an oral cholesterol drug with a different mechanism, with guidance expected in October 2026. Mentioned here only so that “the new cholesterol pill” in the news is not mistaken for a PCSK9 inhibitor.