Cardiology is about to run one of its oddest sequences in decades. In March, the American College of Cardiology and the American Heart Association published new guidelines telling doctors to measure a cholesterol particle called lipoprotein(a), or Lp(a), at least once in the lifetime of every adult. Nothing in the standard cholesterol panel has ever shown this particle. Fewer than 1 percent of Americans have been tested for it. And there is still no approved medicine that lowers it. The measurement arrived before the treatment, which is the reverse of how medicine usually works, and it happened for a reason: the treatment is expected to arrive within a year.
The race behind that expectation involves Novartis, Amgen, and Eli Lilly, the three biggest late-stage programs, and a market analysts describe in the tens of billions of dollars. But the more consequential story is the risk factor itself, which sat invisible inside the most routine blood test in American medicine while researchers spent two decades building the case that it matters more than almost anything else on the panel.
The number the standard test skips
Lp(a) is a cousin of LDL, the so-called bad cholesterol, with an extra protein attached. That extra piece makes it stickier: it promotes plaque and inflammation in arteries and interferes with the body's ability to dissolve clots. Particle for particle, many researchers consider it more atherogenic than LDL. Roughly one in five people worldwide carries it at a level that raises cardiovascular risk, and the level is set almost entirely by genetics, roughly 70 to 90 percent inherited. A person's Lp(a) is fixed by childhood. Diet, exercise, and weight loss barely move it, which makes it unusual among heart risk factors and also unusually knowable: one measurement in early adulthood tells you your number for life.
That one measurement is exactly what almost nobody gets. A standard lipid panel reports total cholesterol, LDL, HDL, and triglycerides. Lp(a) is not among them, and the consequences show up in the risk that remains after everything else is treated. In a pooled analysis of six statin trials covering more than 27,000 people, patients with high Lp(a) stayed at elevated risk even when statins drove their LDL down to the lowest ranges, below 77 milligrams per deciliter. Statins barely touch Lp(a), and the drugs that do move it modestly, the PCSK9 inhibitors, lower it by 20 to 30 percent without any approval for that purpose. Every 50 nanomoles per liter of extra Lp(a) translates into roughly 11 percent higher cardiovascular risk, and the risk climbs continuously. The gap in the panel was not a rounding error. It was the part of the test that mattered for one in five people.
The particle's reach extends beyond heart attacks. Lp(a) is also a driver of calcific aortic valve disease, a narrowing of the heart's main outflow valve that eventually requires surgery and has no drug treatment of its own. A person with high Lp(a) is not simply at elevated risk of a blocked artery. They carry a genetic setup that wears on the heart's plumbing in several ways at once, and until now the health system has had no standard way of finding them before the damage shows up. That is what makes the new testing guidance more than a technical adjustment to a blood panel. It is an invitation to finally see a risk that was always there and always skipped.
We stopped measuring what we could not treat
The neglect had a logic to it, even if the logic aged badly. Doctors are trained, reasonably, to avoid measuring what they cannot act on. A test result with no treatment attached produces anxiety without a plan, so through the statin era the clinical focus stayed on LDL, the risk factor that drugs could move. Lp(a) became the known unknown: genetic, causal in the epidemiological data, confirmed by Mendelian randomization studies as a driver of heart disease and aortic valve disease, and useless to measure because nothing could be done with the answer. The absence of a therapy justified the absence of a test, and the absence of a test kept the risk invisible to the people carrying it.
That reasoning started to erode from both directions at once. The guidelines flipped first. The European Society of Cardiology's 2025 update called Lp(a) a risk marker rather than a treatment target but still recommended measuring it once in every adult, and the 2026 American guideline went further, endorsing universal once-in-a-lifetime testing along with cascade testing of family members, since a high number is inherited and relatives of a carrier have nearly double the odds of carrying it themselves. The measurement is now ahead of the medicine, a deliberate bet by the guideline writers that the medicine is coming. Some populations have more riding on the bet than others: average Lp(a) levels run about four times higher in Black adults, a disparity the standard panel has been masking for decades.
The biomarker has burned the field before
The caution is written into cardiology's recent history. The field has twice celebrated drugs that moved a risk marker beautifully and failed to prevent a single event. Niacin raised HDL cholesterol and, in two landmark trials, did nothing for outcomes. The CETP inhibitors raised HDL by as much as 70 percent or more, and one of them, torcetrapib, was abandoned when it appeared to increase deaths. The lesson the industry learned at enormous expense is that a laboratory number is not a heart attack prevented, and Lp(a) now has to pass the same test that humiliated its predecessors.
The first exam is Novartis's HORIZON trial of pelacarsen, a monthly injection that silences the gene behind the particle and lowered Lp(a) by roughly 80 percent in earlier studies. The trial enrolled 8,323 patients with established heart disease and high Lp(a), and it ends only when 993 patients have suffered a major cardiovascular event. Results are expected later this year, and the market is openly nervous. Analysts at JPMorgan have put the probability of failure near 60 percent, Morgan Stanley has leaned the other way, and Jefferies has described market confidence as low. Behind them sit Amgen's olpasiran, an injection that cuts Lp(a) by more than 95 percent and has its own outcomes trial running, along with Lilly's two entries: lepodisiran, another injectable, and muvalaplin, the only pill in the class. This is the strongest pipeline the risk factor has ever had, and the first readout will price the entire field.
What separates Lp(a) from niacin and the CETP drugs is the quality of its causal evidence. Mendelian randomization, which reads genetics like a natural experiment, has made the case that high Lp(a) causes heart disease, not merely accompanies it. No failed biomarker therapy ever had a genetic case this strong. That is why the three drugmakers, and the guideline writers, are willing to stake so much on a number the standard test has ignored for forty years.
The machine has to turn around
If the trials succeed, the harder problem arrives the next day. Testing rates would need to go from under 1 percent to universal, laboratories would need to add the assay to routine panels, and tens of millions of people would learn they carry a genetic risk just as a treatment for it exists, which is a coordination problem American medicine has not solved gracefully. The drugmakers are already preparing the second wave: primary prevention trials testing the therapies in people who have never had a heart attack, a market larger than the first by an order of magnitude. Cost-effectiveness models published this year suggest the economics work, at least in secondary prevention, though a monthly injection priced like the PCSK9 drugs would strain health plans and invite the same access debates those drugs triggered.
Until the readouts land, the guidance for a person who learns their number is less satisfying. The recommended response to high Lp(a) today is to do everything else harder: lower LDL more aggressively with statins, control blood pressure, stop smoking, treat diabetes. The interim plan does not touch the particle itself, which is precisely the gap the trials are testing whether the new drugs can close. For the one in five people carrying the number, the next twelve months decide whether their result becomes something a doctor can finally act on, or just a worry the panel was right to keep hidden.
The sequence still runs backward to the end. Usually a drug is approved, and then the tests and guidelines catch up over a decade. Here the guidelines ran ahead on purpose, betting that a genetic risk factor with a once-in-a-lifetime number and a pipeline of silencing drugs would finally close the gap the standard panel opened. Whether the bet pays will be decided not by the measurement but by the events, the heart attacks and strokes the trials are counting right now. Either way, the one-in-five number has become visible for good, and the people carrying it are about to learn their own.
Primary sources
- The STAT report by Andrew Joseph for the framing of the three-company race, the 1-in-5 prevalence, and the absence of an approved therapy.
- The Current Atherosclerosis Reports review for the pipeline details, dosing, and reduction figures for pelacarsen, olpasiran, lepodisiran, and muvalaplin.
- The 2026 ACC/AHA guideline discussion in Circulation and the Cleveland Clinic Journal of Medicine review for the testing recommendations, the genetic inheritance figures, the statin trial pooled analysis, and the population disparities. ClinicalTrials.gov supplied the HORIZON and OCEAN(a) trial designs.