Amgen’s small interfering RNA (siRNA) drug olpasiran, which tackles the most common inherited risk factor for cardiovascular disease — lipoprotein(a) — recently passed an important test. Preliminary results from a phase 2 trial demonstrate the drug’s ability to tackle lipoprotein(a). Although this low-density lipoprotein (LDL) particle has received scant attention compared with other cardiovascular disease drug targets, a wealth of genetic data has established it as an independent and causal risk factor for death, heart attack, stroke and aortic stenosis. This emerging drug class will cut the risk of heart disease in a small proportion of people with genetically driven high lipoprotein(a) — also known as ‘Lp little a’ or Lp(a). But beyond that, companies are hoping a small proportion of the overall population could benefit too. Even if it is a fraction of the 40% of adults affected by high cholesterol, this would represent a lucrative commercial market.

Few people know about the risk of heart disease from high Lp(a). Diet and exercise will not help because genes control how much Lp(a) a person makes. Credit: Craig Keifer/ Amgen

How much Lp(a) circulates in a person’s bloodstream depends on their genes — diet and exercise have no impact. Targeting this molecule, therefore, could benefit a minority of patients who are born with elevated Lp(a), some of whom are otherwise healthy and with no other known cardiovascular risks. “This is like a genetic disease,” says Sekar Kathiresan, co-founder and CEO of Verve Therapeutics, who has published widely on the genetics of cardiovascular disease. Several development programs aimed at lowering Lp(a) particles are underway (Table 1).

Table 1 Lipoprotein(a)-lowering therapies in development

Lp(a) is, Kathiresan says, “a cousin” of the atherosclerosis-inducing LDL. “It’s literally just an LDL, but there’s an extra protein attached.” Whereas LDL’s protein fraction comprises apolipoprotein B100 (apoB100) only, Lp(a) also contains apolipoprotein(a). This extra protein prevents the Lp(a) molecule from interacting with LDL receptors that remove cholesterol from the circulation, thus resulting in Lp(a)’s long circulating half-life. Such persistence, coupled with Lp(a) particles’ high concentration of inflammatory oxidized phospholipids, make them highly atherogenic.

Developing an understanding of precisely which patients will — and will not — benefit from lipoprotein(a)-lowering therapies is still a work in progress, as other factors that contribute to the pathological build-up of arterial plaque also need to be considered to estimate an individual’s overall risk. “You have to calculate somebody’s risk, just like you do with LDL or blood pressure lowering, and then determine the extra risk caused by the Lp(a). That gives you an estimate of how much they would benefit from lowering Lp(a),” says Brian Ference, of the University of Cambridge. Kári Stefánsson, CEO of Amgen’s Reykjavik, Iceland-based Decode Genetics subsidiary, takes a more bullish view: “About 15% of the population has a very significant risk of coronary artery disease from Lp(a),” he says. There is, however, no agreed Lp(a) concentration threshold above which therapy is recommended. “It’s somewhat arbitrary as to where you draw that line,” says Kathiresan. Ference argues against the idea of any threshold, on the basis that there is a continuous relationship between increasing Lp(a) levels and cardiovascular disease risk. Even so, those with the highest levels and highest risk will be immediate candidates for therapy. “About 1 or 2 % of the population have markedly elevated Lp(a) levels, conveying a lifetime risk of cardiovascular disease almost equivalent to FH, familial hypercholesterolemia,” he says. “For everyone else, it will become more nuanced, just as it is with LDL and blood pressure levels.” David Reese, Amgen’s executive vice president of R&D, argues that the genetic evidence points to a larger number of people who have residual cardiovascular disease risk that is not addressed by current therapies. “About 20% of people are estimated to have high Lp(a) levels, and data point toward a possible benefit to reducing Lp(a),” he says. The introduction of Lp(a)-lowering therapies will be gradual, however. “Similar to how statins were developed for LDL cholesterol, investigating Lp(a) reduction starts with patients that need therapy most urgently; in this case, secondary prevention patients with high Lp(a) levels. As we learn more about the potential benefit in these patients, we can translate that understanding to patients at other levels of cardiovascular risk.”

Dissecting out Lp(a)’s contribution to cardiovascular disease risk has been the work of several decades. Because Lp(a) is markedly elevated only in a small percentage of the general population, observational studies conducted in the late 1990s failed to reveal its clinical significance. Prospective epidemiological studies uncovered the association, and genetic and phenotypic analyses — both at a population level and at the level of individuals with either unusually high or low Lp(a) levels — deepened scientists’ understanding of its effects.

As part of this effort, Decode mined its large Icelandic genetic database to settle the controversy over whether Lp(a) particle size — which can be highly variable due to the presence of different numbers of so-called ‘kringle’ domain repeats within the apolipoprotein(a) (apo(a)) protein — or absolute concentration was more important in conferring risk. It turns out there is an inverse correlation between the two: those with high levels of Lp(a) generally have a low number of repeats and a smaller apo(a) isoform, in contrast to those with low Lp(a) levels, who generally have a larger apo(a) isoform. “It is clear that what matters is the molar concentration of the particle rather than the particle size,” Stefánsson says. “You know what it is that you need to measure because in this case the molecule that you’re trying to contain — or the target that you’re going to prevent from being made — is the biomarker.”

The investigational agents that are now in clinical development have so far proven potent. Furthest advanced is pelacarsen, an apo(a) mRNA-targeting 2′-methoxyethyl gapmer chimeric antisense oligonucleotide (ASO) covalently attached to triantennary N-acetylgalactosamine (GalNAc3) developed by Novartis, of Basel, Switzerland, in collaboration with Akcea Therapeutics (a former subsidiary of ASO developer Ionis Pharmaceuticals). This agent prevents translation of LPA mRNA into apo(a), which then combines with LDL to make Lp(a). The molecule is undergoing a phase 3 cardiovascular outcomes trial, called Lp(a) Horizon. In a dose-finding phase 2 study, the ASO demonstrated a maximum 80% reduction in circulating Lp(a), compared with a 6% reduction in the placebo group. The study enrolled patients with median Lp(a) concentrations of 204.5–246.6 nmol l–1 into six dose groups. (In contrast, median levels in individuals who do not have elevated Lp(a) are 17 nmol l–1 in Chinese people, 19 nmol l–1 in non-Hispanic people of European ancestry, 29 nmol l–1 in Hispanic people and 59 nmol l–1 in African American people.)

London-based Silence Therapeutics has reported up to 98% reductions in Lp(a) compared with a 10% reduction in the placebo group after a single dose of its blunt-ended 19-mer 2′-O-methyl (2′-O-Me) and unmodified siRNA drug SLN360 in a phase 1 study. Participants’ baseline Lp(a) ranged from 171 to 285 nmol l–1 across six dose groups.

Amgen recently reported data from a phase 1 study of olpasiran, a blunt-ended 21-mer 2′-O-Me and 2′-deoxy-2′-fluoro modified siRNA containing phosphorothioate linkages in its backbone to confer stability and conjugated to GalNAc to increase liver uptake. Olpasiran demonstrated up to 97% lipoprotein reductions in participants with 70–199 nmol l–1 Lp(a) at enrollment; in contrast, for individuals with >200 nmol l–1 at the outset, the drop was up to 91%. A phase 2 study, which is evaluating several doses given either at 12-week or 24-week intervals, will report results later this year, but Amgen disclosed in a press release that the drug slashed Lp(a) by 90% or more from a median baseline of ~260 nmol l–1.

Data from the Novartis Lp(a) Horizon trial of pelacarsen is still about three years away, but a large-scale Mendelian analysis that Ference and colleagues conducted has already provided insights into its likely outcome. The main finding is that a very large absolute reduction in Lp(a) — of the order of 100 mg dl–1 — is needed to achieve a meaningful reduction in risk. If a patient’s Lp(a) levels are low to begin with, cutting them even by as much as 80% is unlikely to be useful. Because the Novartis study enrolled people with a minimum Lp(a) level of just 70 mg dl–1, the risk reduction may only be about 15–16%, Ference says, assuming a median baseline level of around 85–90 mg dl–1. “That’s going to be, I think, slightly sobering to the medical community, because they’re expecting much larger reductions than that,” he says. “Amgen, by contrast, will almost certainly enroll patients with higher baseline levels and get larger absolute reductions and larger proportional reductions in risk. And that’s a strategic advantage because the siRNAs will have more convenient dosing schedules, and if they can show a larger proportional reduction in risk, they’ll cannibalize the market.” Amgen has not yet disclosed its plans for an outcomes trial of olpasiran, but a baseline level of 70 mg dl–1 (equivalent to 175 nmol l–1) “is not necessarily a very low level,” says Narimon Honarpour, who is vice president of global clinical development, general medicine, at the company.

Silence Therapeutics appears to have similar advantages to Amgen. As a small company, it will not be in a position to fund and run a cardiovascular outcomes study itself, so the London-based company is seeking a partner to take the SLN360 program through the later stages of development. “If people are interested in cardiovascular risk and in treating cardiovascular disease, they have to be in this area,” says Giles Campion, chief medical officer at the company. “It’s a matter of when they jump in.” Verve, which is pursuing single-course gene-editing therapies for genetically defined cardiovascular diseases, also has an Lp(a) program in the works, but it remains at a research stage at present. (As Nature Biotechnology went to press, its lead program, VERVE-101, a base editor designed to inactivate the gene encoding proprotein convertase subtilisin kexin type 9 (PCSK9), was about to enter a first-in-human trial in patients with heterozygous familial hypercholesterolemia.)

“If people are interested in cardiovascular risk and in treating cardiovascular disease, they have to be in this area”

It will be at least three years before any Lp(a)-lowering therapies can reach the market. Clinical guidelines on managing Lp(a)-associated risk in the absence of targeted therapies will become available shortly, says Ference. “There is a huge educational effort required,” says Campion. And now that effective therapies are on the horizon (pun intended), the question of whether to screen for high Lp(a) is becoming increasingly relevant. “If there’s a rationale and motivation to screen for familial hypercholesterolemia, there should be a similar rationale to screen for markedly elevated Lp(a), which is probably the most common inherited risk factor for atherosclerotic cardiovascular disease — it’s much more common than familial hypercholesterolemia,” says Ference.