Skip to main content
Premium Trial:

Request an Annual Quote

Family Heart Foundation Looks to Spur Lp(a) Testing With FIND Lp(a) Machine Learning Tool

Premium
Technician preparing human blood sample for screening in the lab

NEW YORK – The Family Heart Foundation is hoping that a new program it is launching will increase testing for a cardiovascular biomarker that, while well established as a risk factor for heart disease, isn't routinely performed.

The nonprofit advocacy organization recently launched FIND Lp(a), a program designed to identify patients suspected to have elevated levels of the lipoprotein(a) biomarker based on data in their medical records. The goal of the program, which has been rolled out at a half-dozen hospital systems, is to encourage these patients to get a simple blood test that can reveal whether they have elevated Lp(a).

FIND Lp(a) is part of the Family Heart Foundation's FIND initiative, which stands for Flag, Identify, Network, and Deliver, and is the second such program the organization has developed. A previous effort focused on identifying patients likely to have familial hypercholesterolemia, an inherited disorder that causes high cholesterol.

Lp(a) is a type of low-density lipoprotein cholesterol that, when elevated, is a known risk factor for atherosclerotic cardiovascular disease (ASCVD). Heart disease and stroke are linked to this common inherited risk factor, and Diane MacDougall, VP of science and research at the Family Heart Foundation, estimated that upwards of 20 percent of the population has high Lp(a). That represents more than 1 billion people globally. 

"Despite the fact that it is common, [Lp(a)] is screened for very infrequently," MacDougall said. 

According to a recent analysis by the Family Heart Foundation, only about 1 percent of adults in the US have been tested for Lp(a). A lack of clear guidelines on how to manage patients with high Lp(a) — a gap groups like the American Heart Association are trying to fill — is likely one reason physicians haven't prioritized testing, and another is that Lp(a) testing isn't widely reimbursed. In addition to guidelines bodies updating their recommendations, convincing payors to routinely cover testing also will be critical to driving testing adoption. 

The Family Heart Foundation hopes to spur testing with the FIND Lp(a) program by homing in on patients with established cardiovascular disease as those with the "greatest need and most likely to benefit," MacDougall noted, but long term, the foundation hopes that every patient will be tested for elevated Lp(a). FIND Lp(a) is "intended to bridge a gap and accelerate change within our complex healthcare system," she said.

Testing could also pick up as drugmakers advance antisense oligonucleotides, siRNAs, and other new therapeutic candidates designed to lower Lp(a) in patients with high levels of the biomarker, which are promising new ways of managing risk factors beyond lifestyle and dietary changes.

Anticipating new Lp(a)-targeting drugs, some guidelines bodies have begun to advise broader screening for high Lp(a). The National Lipid Association earlier this year released an updated scientific statement on Lp(a), in which it recommended that doctors test all adults at least once. Previously, NLA had only recommended testing patients with certain medical histories, such as those with a personal or family history of premature ASCVD and suspected familial hypercholesterolemia. 

Since Lp(a) levels are predominantly determined by genetics, specifically by an inherited variation in the LPA gene, patients typically only need to be tested for it once in their life.

As hospitals prepare to test more patients for Lp(a), the Family Heart Foundation hopes that the FIND Lp(a) program will help to prioritize who to screen first. The foundation developed a machine-learning model for the program that combs through data in patients' health records, including demographic, medical, and laboratory data, and pinpoints signs that they're likely to have elevated Lp(a), and could benefit from Lp(a) testing and more intensive cardiovascular care management.

MacDougall described the FIND Lp(a) model as finding "clues" in patient health records to spot those who haven't had Lp(a) measured but are likely to have high levels of the biomarker.

The foundation developed the model using data from the Family Heart Database, a longitudinal dataset of anonymized medical claims and laboratory data from more than 300 million patients. Researchers analyzed data on those known to have elevated Lp(a) and those who didn't to better understand what factors are associated with the biomarker.

Using machine learning, researchers identified various features, including being female, Black, or having certain medical conditions, that were associated with a greater chance that a person has high Lp(a). The Family Heart Foundation said it's working to better understand the relative contribution of these various factors to the likelihood of having high Lp(a).

Researchers retrospectively validated the model using a subset of data in the Family Heart Database that they didn't use to develop the tool. They found that 60 percent of the patients who were flagged by the model as likely having high Lp(a) actually had it. As the model is implemented at hospital systems, the Family Heart Foundation will work with these partners to prospectively validate it.

The FIND Lp(a) program is funded by Novartis, one of the companies developing an Lp(a)-lowering drug. Its antisense oligonucleotide drug, pelacarsen, is undergoing pivotal Phase III clinical testing. The Family Heart Foundation noted, however, that the company is not involved in the strategy or execution of the FIND Lp(a) program or its implementation at hospital systems.

The foundation has corralled a group of hospital systems, which it refers to as the Collaborative Learning Network, that are implementing the FIND Lp(a) model to identify patients with high Lp(a). The hospital systems will collaborate and share information on how they're using the model and treating elevated Lp(a).

The participating hospital systems will send de-identified patient data to the Family Heart Foundation, which will run the FIND Lp(a) model on the dataset. The data are scrubbed of personally identifying information but attached to a code so that patients can be later re-identified by the hospital systems. A report on those deemed high risk by the model will be sent back to the hospital system.

The hospitals can then use this information to "focus" their outreach for Lp(a) testing, said Katherine Wilemon, founder and CEO of the Family Heart Foundation. The Family Heart Foundation also provides educational materials, care navigation services, and other resources for patients with elevated Lp(a).

Emory Healthcare in Atlanta is one of the hospitals participating in the program, along with OhioHealth, Stanford Health Care, and the Medical University of South Carolina.

Once a list of at-risk patients is generated, the FIND Lp(a) team at Emory will reach out to each patient's primary care doctor and cardiologist, if they have one, and ask permission to contact the patient. The FIND Lp(a) team will then notify the patient through the patient portal about their risk of elevated Lp(a) and offer to schedule them for testing. 

Patients who agree to testing will have the opportunity to schedule a visit with a preventive cardiologist. There, physicians will discuss the cardiovascular risks associated with high Lp(a) and steps the patient can take to manage their risk, such as making lifestyle changes or enrolling in drug trials. Physicians will also recommend cascade screening for first-degree relatives of patients with high Lp(a).

Since elevated Lp(a) is primarily an inherited risk factor, it is likely that patients with high Lp(a) have family members who also have it. In fact, for someone with high Lp(a) levels, each of their children and siblings has a 50 percent chance of also having this risk factor, said Ijeoma Isiadinso, medical director of the preventive cardiology clinic at Emory Healthcare.

"While there are currently no FDA-approved medications to lower elevated Lp(a) specifically, it still is important that patients and clinicians are aware … of the increased risk of cardiovascular disease associated with elevated Lp(a) so that appropriate steps can be taken to modify and treat cardiovascular risk factors," Isiadinso said.

The FIND Lp(a) team at Emory will also educate primary care physicians about the risks associated with high Lp(a), since they might not be aware of them.

Isiadinso said she hopes participating in the FIND Lp(a) program will enable Emory to screen more patients for the risk factor. In addition to using the FIND Lp(a) model, cardiologists at Emory will continue to prioritize Lp(a) testing for patients who have had recurrent or unexplained atherosclerotic cardiovascular events and use the NLA guidance to expand into other patients.

There are patients with elevated Lp(a) who are undiagnosed and at high risk of future cardiovascular events, Isiadinso said. It's a "missed opportunity" to wait until a patient has had a cardiovascular event, rather than testing for elevated Lp(a) proactively to identify at-risk patients early and, hopefully, avoid a cardiovascular event from happening at all, she added.

FIND Lp(a) builds off of an Amgen-sponsored program that the Family Heart Foundation launched a decade ago, dubbed FIND FH, which also comprised a machine-learning model that identified patients with probable familial hypercholesterolemia from data in their health records. 

Only about 0.4 percent of the population has familial hypercholesterolemia, and the genetic condition is underdiagnosed. If left untreated it can lead to cardiac issues, including atherosclerosis and heart attack. Through the FIND FH program, which is still ongoing, about 230 patients have been diagnosed with familial hypercholesterolemia across five hospital systems, including at Emory. 

The Family Heart Foundation was launched in 2011 by Wilemon, who sought answers after surviving a heart attack at 39 years old despite not having many risk factors. She was later diagnosed with familial hypercholesterolemia and elevated Lp(a).

Wilemon said she was alarmed by the lack of awareness about genetic dyslipidemias as a common cause of premature and generational cardiovascular disease. "I decided to dedicate my professional life to increasing understanding of familial hypercholesterolemia because we had all the tools to screen and manage it," she said.

Although initially launched as the FH Foundation, the organization decided to expand its programming to equally focus on familial hypercholesterolemia and elevated Lp(a). With a broadening focus on genetic lipid disorders associated with premature heart disease, the nonprofit changed its name to the Family Heart Foundation in 2021. 

The programs the foundation spearheads have also grown in step. Last month, for example, the Family Heart Foundation launched Cholesterol Connect, a program offering patients free at-home lipid screening kits that they can request online and use to test for cardiovascular risk factors like total cholesterol, LDL cholesterol, and Lp(a). The program is sponsored in part by Amgen.

The foundation expects to share best practices on Lp(a) screening and tools from the FIND Lp(a) program in mid-2026. Until then, the FIND Lp(a) Collaborative Learning Network will develop and refine the best practices for implementing the model and develop other resources to encourage adoption of Lp(a) testing and care management into routine care.

This focus on implementation is needed to ensure not only that the program identifies patients at risk but also that these patients actually receive proper treatment and support. "Both of these components are critical to the success of programs like this," MacDougall said. 

The machine-learning model alone won't facilitate Lp(a) testing, and best practices are needed to ensure follow through with patients. "It's our hope that this [FIND Lp(a) program] is a real driver and accelerator for Lp(a) screening across the US," MacDougall said.