#07 – Deep Dive: Lp(a) — what every doctor, and the 10-20% of the population at risk, needs to know

KEYPOINTS

• 01

  Lp(a) is an LDL particle with apo(a) bonded to it

  In this solo episode, with Bob Kaplan asking the questions, Attia describes Lp(a) as a garden-variety LDL particle of about 20 nm whose apoB-100 is covalently linked, through a strong disulfide bond, to an extra protein, apo(a). Which LDLs carry apo(a) is mostly genetically determined, inherited codominantly. Attia estimates somewhere between one in five and one in 10 people have elevated levels, and calls the distribution long-tailed to the right.

• 02

  Apo(a) resembles the clotting protein plasminogen

  Plasminogen has five 'kringle' folding domains. Apo(a) lacks kringles 1–3 but has the same kringle 5 and a kringle 4 with 10 subtypes. Kringle 4 type 2 varies most between people, from a couple of copies to about 40. That repeat count largely sets the particle's mass, which is why Attia spends time on what he admits is a very technical detail.

• 03

  Particle number beats Lp(a) mass

  Attia estimates 19 out of 20 Lp(a) tests measure mass, which counts the whole structure. Two people with the same mass can differ in risk: the one with shorter kringle 4 type 2 repeats has more particles and higher risk. A very low mass, under 5 mg/dL, makes a high particle count unlikely. He finds the Lp(a)-cholesterol test unhelpful and prefers particle number in nmol/L.

• 04

  Attia's thresholds and reference ranges

  For particle number he likes under 50 nmol/L, sees 50–100 as a grey zone, and worries above 100, certainly above 125. The highest he has seen is about 650–700, and a few patients sit at 400–500. Mass-based cut-offs vary: US labs call under 30 mg/dL normal, the European Atherosclerosis Society under 50. He believes the UK and Germany cover apheresis above 60.

• 05

  High Lp(a) may once have been an advantage

  Attia speculates on two evolutionary benefits. Because of its similarity to plasminogen, Lp(a) promotes clotting, which helps in trauma or childbirth bleeding. It may also scavenge oxidised phospholipids and carry them back to the liver. In today's more inflammatory setting, with hypertension and high insulin, some particles get stuck in the artery wall and bring their oxidised cargo with them.

• 06

  Three harms: atherosclerosis, aortic stenosis and blood clots

  Attia believes about two-thirds of aortic stenosis cases are explained by elevated Lp(a). The particle seems drawn to the high-pressure aortic valve, where it triggers bone-forming proteins and calcification. He puts the hazard ratio for aortic stenosis at about 2–4, median around 2.5, and for venous thromboembolism at about 3. He stresses that tripling a 1% risk means 3% absolute risk, not cause for panic.

• 07

  Statins don't lower Lp(a), but are still used

  Statins work mainly by switching on SREBP-2, which adds LDL receptors to the liver, but SREBP-2 also raises PCSK9, which breaks those receptors down. Tom Dayspring's explanation, as Attia relays it, is that Lp(a) is cleared 'last in line', after LDL and VLDL, so it never gets its turn. On statins alone, Lp(a) changes not at all or even rises. Attia is clear that people with high Lp(a) probably should take a statin, but to control apoB, not Lp(a).

• 08

  PCSK9 inhibitors lower Lp(a), especially combined with a statin

  By blocking PCSK9, these drugs protect LDL receptors and the other receptors, such as LRP, that clear Lp(a), and combined with a statin they lower Lp(a). Attia notes they are not FDA approved for that purpose. Niacin lowers Lp(a) by about a third, but whether it saves lives is contested. He thinks the trial that sank niacin was poorly designed, and says lipidologists are split roughly evenly on using it.

• 09

  Lifestyle barely moves Lp(a)

  Lp(a) levels are decided mainly by how much the liver makes, with clearance a secondary factor. LDL particle number, by contrast, depends on triglycerides, cholesterol synthesis, absorption and receptors, and diet can change all of them. That is why diet works on LDL particle number but not Lp(a).

• 10

  Antisense drugs are the first therapies designed for Lp(a)

  Antisense oligonucleotides block the liver's production of apo(a), cutting levels by about 70% to as much as 99% in early data. Attia says that could normalise someone at 200. They had completed phase 1 and 2 trials and were slowly enrolling phase 3. He jokes that three years earlier he predicted results in five years and still says five. He is wary of wiping something out entirely, but says only such a trial can prove causation.