Cholesterol (And New Ways To Reduce It)

improving heart health

Cardiovascular disease kills more humans than anything else. Of the 3.07MM US deaths in 2024, 919k were from cardiovascular disease (29.9% of total).

Year	Total US deaths	Heart disease deaths	Total CVD deaths	CVD % of total
2013	2,596,993	611,105	801,000	30.8%
2014	2,626,418	614,348	808,000	30.8%
2015	2,712,630	633,842	838,548	30.9%
2016	2,744,248	635,260	840,768	30.6%
2017	2,813,503	647,457	850,125	30.5%
2018	2,839,205	655,381	868,662	30.6%
2019	2,854,838	659,041	874,613	30.6%
2020	3,390,029	696,962	924,741	27.3%
2021	3,464,231	695,547	931,578	26.9%
2022	3,279,857	702,880	941,652	28.7%
2023	3,090,582	699,659	916,932	29.7%

Swipe to see more →

According to a study between 1970-2022 for adults >25 years old, there’s been a crossover: declining ischemic mortality vs rising non-ischemic mortality. Note: Ischemic mortality is death caused by restricted blood flow (usually from heart disease or stroke). Non-ischemic causes include heart failure, arrhythmias, and damage from chronic high blood pressure.

Acute myocardial infarction (heart attack) deaths dropped from 54% to 29% of the ischemic total, while chronic ischemic heart disease rose from 46% to 71%.

This is a paradox of medical progress and success. It’s what effective lipid cholesterol management (statins) and revascularization (stents, bypasses) produced. Patients are more likely to survive an acute index event but go on to accumulate chronic cardiac burden over years, sometimes decades.

(A) Absolute and (B) age-adjusted mortality for ischemic and other heart disease in the United States (1970 – 2022)

cholesterol

Not all cholesterol is bad. A normal healthy person has ~2 grams cholesterol per liter plasma.

We need it. It helps to control the stability and fluidity of cell membranes and is an important starting ingredient for making hormones such as testosterone and estrogen.

Cholesterol isn’t dissolved in our blood. It floats around in packages called lipoproteins.

Lipoproteins vary in size and composition. The key ones to know are:

• LDL-C (stands for low-density lipoprotein). In this form, cholesterol is transported from the liver to the rest of the body.
• HDL-C (stands for high-density lipoprotein). In this form, cholesterol is transported back to the liver from the body’s tissues for recycling / disposal. It is broken down there and then flushed out of the body with the bile fluid. HDL has a higher ratio of protein:cholesterol vs LDL:

Cholesterol Particles Vary in Size and Density

(Above) Chemical composition of lipoproteins

how cholesterol causes damage – blood vessels

Cholesterol causes damage when ApoB-containing particles penetrate the intima (the inner lining of the artery wall). Once inside, the sequence is deterministic: the particles get trapped, become oxidized, trigger inflammation in the endothelium, attract immune cells, and gradually build plaque. (The immune cells become engorged and die, forming “foam cells.” This attracts more immune cells and causes smooth muscle cells in the artery wall to multiply, gradually building a plaque capped by a thin fibrous layer.) This is called atherosclerosis.

A fibrous scab (made of smooth muscle cells called foam cells) forms over the plaque, preventing bits from breaking off into the bloodstream. As plaque grows, it gradually narrows the artery’s opening (lumen), restricting blood flow.

The soft plaque under this scab is very dangerous. A spike in blood pressure can rupture the cap, exposing the plaque to blood. This forms a clot as the body tries to seal the wound, but a large clot can completely block the vessel, causing a heart attack or stroke.

Atherosclerosis develops slowly, often beginning in childhood, and typically causes symptoms only once a vessel is >70% narrowed.

The key variable is LDL-C: the amount of cholesterol carried inside LDL particles. By reducing the amount of cholesterol entering arterial walls, plaque formation is reduced.

how cholesterol also makes your heart weaker

Beyond plaque buildup, excess cholesterol weakens the heart itself over time.

When arteries become narrower and less flexible due to plaque buildup, the heart must pump harder to push blood through them. This raises blood pressure. And high blood pressure, in turn, damages the artery walls further, making it even easier for more cholesterol to lodge in them. It’s a terrible cycle.

The heart weakens through multiple mechanisms:

• Left ventricle hypertrophy: When a heart has to consistently pump against higher pressure, the muscle wall of the left ventricle thickens (think like a bicep growing from lifting heavy weights). But unlike a bicep, a thickened heart muscle becomes stiffer and less efficient. It needs more oxygen to function. Its chambers may not fill properly between beats.
• Reduced coronary blood flow: The coronary arteries (the ones that feed the heart muscle itself) also develop plaques. So the heart is working harder while simultaneously receiving less blood and oxygen. This mismatch can cause chest pain or, if a plaque ruptures and a clot forms, a heart attack that kills part of the muscle.
• Progression to heart failure: Over time, an overworked and underfed heart can’t keep up. The thickened muscle may begin to stretch and dilate, losing its ability to pump effectively. This leads to gradual heart failure.

cholesterol reduction

Across a very large, randomized evidence base, it was shown that each ~1 mmol/L (~39 mg/dL) LDL‑C reduction yields ~22% proportional reduction in major vascular events. (170k participants across 26 trials). Benefits are broadly consistent across demographics and baseline risk profiles.

There have been a few key milestones in cholesterol and cholesterol-lowering research:

• In 1913, Nikolay Anitschkow provided the first causal relationship between cholesterol and atherosclerosis. He found that feeding rabbits cholesterol induced atherosclerosis.
• In 1970s, NIH researchers Michael Brown and Joseph Goldstein demonstrated that hepatic LDL receptors regulate plasma cholesterol. This won them the 1985 Nobel Prize in Medicine. They were inspired by seeing a 7 year old girl in 1969 who had 8x normal cholesterol levels. We found that the liver actively clears LDL from blood using surface receptors. It also provided a conceptual target: by increasing the density of LDL receptors on the liver, you could pull more LDL out of circulation.
• The 1st statin: Akira Endo isolated compactin (mevastatin) from the fungus Penicillium citrinum in 1973, proving that HMG-CoA reductase (the rate-limiting enzyme in cholesterol synthesis) could be pharmacologically inhibited. Merck subsequently isolated lovastatin from Aspergillus terreus, which became the first FDA-approved statin in 1987 (brand name: Mevacor).
• In 1994, the Scandinavian Simvastatin Survival Study (4S) (4,444 patients) demonstrated that lowering LDL-C with simvastatin (Merck’s Zocor) reduced all-cause mortality by 30% and major coronary events by 34%. It proved the lipid lowering hypothesis. Statins became the foundational pillar of preventive cardiology.
• In 2004, the PROVE IT–TIMI 22 trial (4.1k patients on 80mg Lipitor vs 40mg statin) codified that more intensive lipid lowering produces better outcomes. High-intensity statin dose became the standard of care for high-risk patients. Cholesterol-lowering medication use rose from 3% of US adults in 1988 to ~45% by 2013.
• Merck’s ezetimibe (Zetia), a once-daily oral pill, was approved in 2002. It inhibits intestinal cholesterol absorption (a mechanism complementary to statins). The IMPROVE-IT trial (2015; 18.1k patients) showed that ezetimibe + simvastatin reduced cardiovascular events vs simvastatin alone. This made it the first nonstatin add-on to demonstrate improved outcomes. It’s now generic and widely accessible.
• PCSK9 inhibitors (FOURIER, 2017). This was a large cardiovascular outcomes study (27k people). It showed a reduction in composite cardiovascular events by 15%.

Prior to the development of PCSK9 inhibitors, statins and ezetimibe were the only effective medicines available to help lower LDL-C.

statin comparison

Statins are given as once-daily orals. Their typical efficacy at LDL reduction is below:

Statin	10 mg	20 mg	40 mg	80 mg	HDL-C ↑	Half life (h)	Bioavailability %	Protein bonding %
Crestor (rosuvastatin)	−46%	−52%	−55%	Not approved	+7.7–+9.6%	14-30	12	>98
Lipitor (atorvastatin)	−37%	−43%	−48%	−51%	+2.1–+6.8%	19-20	20	88–90
Zocor (simvastatin)	−28%	−35%	−39%	−46% (pulled for safety)	+5–+7%	2-3	<5	>95
Pravachol (pravastatin)	−20%	−24%	−30%	N/A	+3–+6%	1.3-2.8	18	43–55
Lescol (fluvastatin)	N/A	−21%	−27%	−33%	+3–+5%	0.5-2.3	19–29	>99
Mevacor (lovastatin)	−21%	−27%	−31%	N/A	+5–+7%	2-4	<5	>95
Livalo (pitavastatin)	−31%	−36%	−43%	N/A	+5–+9%	12	43–51	96

Swipe to see more →

Short-half-life statins (simvastatin, lovastatin, fluvastatin) must be administered in the evening to coincide with peak nocturnal cholesterol synthesis.

The extended half-lives of atorvastatin and rosuvastatin allowed for potent inhibition regardless of dose timing.

Statins remain the first line because they are cheap, effective, and deeply validated. However, they are not the best option out there. This crown belongs to a class of drugs called PCSK9 inhibitors.

the new class: pcsk9 inhibitors

In 2003, researchers in Montreal discovered the PCSK9 enzyme. (It stands for proprotein convertase subtilisin/kexin type 9). They published its sequence.

Simultaneously, researchers in Paris were studying multiple generations of a single French family that exhibited unusually high levels of LDL-C. They identified that the family’s history of heart disease and high LDL-C was correlated with a mutation in a chromosome affecting PCSK9.

The Canadian and French team collaborated to publish a joint paper in Jun-2003. It established the link between PCSK9 and cholesterol.

PCSK9 binds to LDL receptors on the surface of liver cells (hepatocytes) and promotes their degradation.

More PCSK9 means fewer LDL receptors survive to clear LDL from the blood.

PCSK9 is expressed mainly in the liver, intestine, kidney, skin and the nervous system.

pcsk9 inhibitor drugs

PCSK9 is now targeted by at least three clinically meaningful modalities.

Monoclonal antibodies remove circulating PCSK9 (Amgen’s Repatha, Sanofi’s Praluent); siRNA blocks PCSK9 synthesis (Novartis’s Leqvio). New oral agents, like Merck’s enlicitide, block PCSK9 function in a more convenient pill form.

Total PCSK9i Sales

Year	Repatha (Amgen)	Praluent (Sanofi)	Leqvio (Novartis)	Total ($ MM)
2015	10	10	—	20
2016	141	111	—	252
2017	319	206	—	525
2018	550	299	—	849
2019	661	290	—	951
2020	887	319	—	1,206
2021	1,117	247	12	1,376
2022	1,296	402	112	1,810
2023	1,635	467	355	2,457
2024	2,222	500	754	3,476
2025	3,016	617	1,198	4,831

Swipe to see more →

Sanofi / Regeneron’s Praluent was the 1st FDA-approved PCSK9 inhibitor in Jul-2015. It was launched at a high price ($14,000 per year). There was substantial payer resistance. By Feb-2019, price was cut to $5,850 per year.

Amgen’s Repatha – The Current Market Leader

Amgen began their PCSK9 research program began in 2005. They discovered that PCSK9 binds directly to LDL receptors on the liver, which results in the breakdown of the LDL receptor. This prevents the liver from clearing LDL-C from the blood. Once this was understood, Amgen’s efforts focused on ways to inhibit PCSK9 protein binding to LDL receptors, so that LDL-C clearance by the liver could be increased.

Amgen launched their antibody program in 2006, with the 1st proof of concept in 2008. 1st in-human trials for Repatha started in 2009. In Aug-2015, the FDA approved Amgen’s Repatha as a biologic. In Jul-2016, the FDA approved Repatha in a single monthly injectable dose.

Repatha is a human monoclonal antibody that inhibits PCSK9. It’s self-injected (subcutaneous) every 2 weeks or monthly.

According to Amgen on Mar 28th 2026, Repatha has “been used by more than 8 million patients globally”. (In Aug-2025, they mentioned this figure as 5 million patients). It’s well-covered by insurance. As per IQVIA data, ~80% of US patients pay <$50 out of pocket monthly.

It’s approved in 74 countries. It has been studied for 15 years in 51 clinical trials with 57k+ patients. US LOE is Aug-2029.

Novartis Leqvio

Leqvio (Inclisiran) was FDA approved in Dec-2021. It is a small interfering RNA (aka siRNA). It acts as a PCSK9 inhibitor.

The FDA considers it to have a unique MOA vs Repatha. Its LOE is 2036.

It was approved in China in Aug-2023 (EU: Dec-2020, UK: Aug-2021). For bridging data for Chinese approval, Novartis ran a 345 person, Phase 3 trial in Asia. It had 232 Chinese patients (2/3rds). For the Chinese patients, LDL-C dropped 61%.

As it uses RNA, there is no impact on DNA. It has a half-life of 9 hours. Within 48 hours of dosing, it is undetectable. Within this time, the drug loads onto the RNA-induced silencing complex (RISC). Once loaded, it repeatedly destroys PCSK9 mRNA (the recipe card for PCSK9). This inhibits PCSK9 production.

It comes as a single-dose prefilled syringe (see image below). It doesn’t require any refrigeration. It has a shelf life of 3 years. Its dosing shifts from 284mg initially, then at 3 months, and then every 6 months. This makes adherence easier.

It was initially developed by Cambridge, MA-based Alnylam Pharmaceuticals (Nasdaq: ALNY).

Trials showed ~2.5% of patients discontinued. Most common cause was injection site reactions (think redness, rash, pain).

It’s injected in the abdomen, upper arm, or thigh. It must be done by an HCP:

In 2025, the FDA expanded Leqvio’s indication, which pushes it earlier into the treatment pathway.

The following is an infographic Novartis uses try and convince doctors:

(L) Leqvio, (R) Repatha

pcsk9 vs high-dose statins

Factor	High-Intensity Statins	PCSK9i
LDL-C reduction (monotherapy)	Crestor 40mg: -55%, Lipitor 80mg: -51% at 52 weeks.	Repatha: -61% at 12 weeks.
LDL-C reduction (when used on top of statin)	-	Repatha: -59% from statin-treated benchmark. Merck’s new oral does even better at –64.6% from statin benchmark.
Lp(a) effect	Negligible	Repatha: -27%, Merck’s oral: -28.2%
ApoB reduction	Crestor 40mg: -40%, Lipitor 80mg: -35% at 52 weeks.	Repatha: -44.4%
Non-HDL-C reduction	Crestor 40mg: -45%, Lipitor 80mg: -40% at 52 weeks.	Repatha: -51% at 48 weeks, Merck’s oral: -53.4%
Dose-response	Diminishing (2x dose → ~6% more)	Consistent ~55–65% additional
Muscle side effects	High	None vs placebo
Liver enzyme elevation	Dose-dependent increase in ALT/AST	None significant
New-onset diabetes risk	+9% increased risk	No signal
Annual cost (US)	$30–300 (generic)	$5,800 - 6,500 (injectable)

Swipe to see more →

a metric statins don’t touch: lp(a)

Lp(a), or lipoprotein(a) is an LDL-like particle with an extra protein (called apolipoprotein(a), or apo(a)) bolted on. Apo(a) supercharges it vs a normal LDL particle; it makes it more likely to get trapped in the artery wall. It also makes the trapped cholesterol more likely to break off.

(Left) LDL (Right) Lp(a) Source: Amgen

Lp(a) is simultaneously:

• pro-atherogenic (builds plaque)
• pro-inflammatory (promotes inflammation)
• pro-thrombotic (increases clotting risk)

Lp(a) is a triple threat. Worse, Lp(a) is genetically determined. As per Amgen, Lp(a) levels are 70-90% genetically determined. (more details: key driver is the KIV type-2 repeat polymorphism in the Lp(a) gene, which creates a size variation in the apo(a) protein)

Diet does not lower it. Exercise does not lower it. Statins do not lower it.

This is a very good paper detailing the history of Lp(a).

For decades, this was a major blind spot in cholesterol treatment. The only thing physicians could do, in extreme cases, was to regularly administer a blood-cleaning procedure called apheresis.

PCSK9 inhibitors finally address this pharmacologically. And there is now a big regulation change too.

Newer US Guidelines Will Increase Lp(a) Testing

Historically, Lp(a) was infrequently measured in routine clinical practices.

Last weekend, at ACC 26, the ACC and AHA updated the guidelines for universal Lp(a) testing. Your diagnostic tests for lipids will show this metric going forward. New testing recommendations include universal Lp(a) measurement at least once in adulthood.

In addition to Lp(a), the guidelines include CAC (Coronary Artery Calcium) scanning for men >40 and women >45 at borderline risk, and apoB measurement for in certain cases.

The guidelines also bring back specific LDL targets (<100 mg/dL for borderline risk, <70 for high risk, and <55 for very high-risk ASCVD). There’s a stronger push for earlier intervention, with earlier pharmacotherapy for young adults with LDL-C ≥160.

Continued Limitation In Testing Quality

Lp(a) particles vary in size. Traditional lab tests, which count number of particles, do not accurately assess risk from Lp(a).

Current tests measure the total mass of Lp(a) in the blood in milligrams per deciliter (mg/dL). A more accurate approach is to measure Lp(a) in nanomoles per liter (nmol/L). This allows particle count to be specified.

pcsk9 comparison

	Enlicitide (Merck)	Repatha (Amgen)	Praluent (Regeneron/Sanofi)	Leqvio (Novartis)
Drug Class	Macrocyclic peptide	Monoclonal antibody	Monoclonal antibody	siRNA
Administration Route	Oral (daily pill)	Subcutaneous injection (every 2 weeks or monthly)	Subcutaneous injection (every 2 weeks or monthly)	Subcutaneous injection (After initial doses, every 6 months)
LDL-C Reduction (vs. Placebo)	56–60% (63% on-treatment)	59–66%	57–66%	50–52%
ApoB Reduction	50–55%	~49%	~48%	~41%
Lp(a) Reduction	26–28%	25–30%	25–30%	18–22%
Non-HDL-C Reduction	53–58%	~52%	~50%	~45%
Cardiovascular Outcomes	Pending (n=14,500 trial)	15% MACE ↓ (FOURIER trial)	15% MACE ↓ (ODYSSEY trial)	Pending (ORION-4 trial)
Safety Profile	Comparable to placebo	Injection-site reactions (2–5%)	Injection-site reactions (6–7%)	Injection-site reactions (~5%)
Adherence Rates	97–98% (Trial setting)	~80% (Real-world, 1 year)	~50–55% (Real-world, 1 year)	Anticipated higher (due to infrequent dosing)
FDA Approval Status	Being filed	Approved (2015)	Approved (2015)	Approved (2021)
Estimated US Cost/Year	TBD (expected lower)	~$5,850	~$5,800	~$6,500

Swipe to see more →

merck’s new oral PCSK9 inhibitor

Merck’s enlicitide (MK-0616) is a once-daily oral PCSK9 inhibitor. It’s deemed a macrocyclic peptide. It delivers antibody-equivalent LDL reductions.

Last weekend, at ACC 26, Merck presented 2 key Phase 3 datasets:

• In the Phase 3 CORALreef Lipids trial (2,909 patients), enlicitide delivered a -55.8% LDL-C difference vs placebo at week 24. 67.5% of patients achieved an LDL <55 mg/dl with a >50% reduction.
• In the Phase 3 CORALreef AddOn trial, enlicitide, added to statin therapy, reduced LDL-C by 64.6% from baseline at 8 weeks. This combination therapy is interesting, and also helps me lead into the next section.

Merck has effectively pulled off the difficult trick of getting once-daily oral efficacy close to the injectable.

Eli Lilly CEO Dave Ricks told me at JPM HC 2025 that ~25% of the population was needle-phobic. A pill is also easier to manufacture than the injectable. Both these facts will accelerate PCSK9i adoption.

Merck will file for FDA approval soon. I would expect a US launch by Q1-2027.

using statins with pcsk9 drugs

Statins and PCSK9i work through complementary mechanisms of action.

Statins block cholesterol synthesis in the liver, which causes the liver to upregulate LDL receptors on its surface to pull more LDL from the blood. But statins also increase PCSK9 expression, which degrades those same receptors. Adding a PCSK9i blocks this effect, preserving the extra LDL receptors that statins create. The net result is greater LDL-C reduction than either drug alone.

At ACC 26 last week, Amgen published results from a 3,655 patient (cardiovascular risk + T2D), 5 yr Phase 3 trial. It showed that Repatha + a statin reduced the risk of a major cardiovascular event (MACE) by 31%. This builds on an earlier trial (VESALIUS-CV, 12k patients) which showed a 25% reduction in MACE and a 36% reduction in 1st heart attacks.

why don’t we increase hdl too?

We know more HDL is good. Pharma invested substantial time and money into finding drugs to raise HDL. Every effort failed:

• Torcetrapib (ILLUMINATE trial, N=15,067): This was Pfizer’s blockbuster bet (~$800MM invested). But it increased mortality due to off-target hypertension and aldosterone activation. Was never approved. ILLUMINATE trial was halted in 2006.
• Dalcetrapib (dal-OUTCOMES, N=15,871): This was Roche’s attempt to learn from Pfizer’s mistake. In trials, it showed HDL went up. But this did not correspond to any visible risk reduction. Drug failed for futility.
• Evacetrapib (ACCELERATE, N=12,092): This was Lilly’s entry. It was more pharmacologiccaly potent than the previous ones (HDL jumped by +130%). However, the drug showed no benefit against its primary endpoint in the trial. Drug failed for futility.
• Anacetrapib (REVEAL, N=30,449): This was Merck’s attempt. It showed a modest benefit (-9% relative risk reduction) but was commercially abandoned due to prolonged tissue accumulation and marginal clinical utility.

← The Library