Heart failure is a leading cause of disability and death in the U.S., but many cases are diagnosed late due to limited access to echocardiography, the primary method of detecting the condition. To address that need, the U.S. FDA has granted breakthrough device designation to Eko Devices Inc., of Berkeley, Calif., for an electrocardiogram (ECG)-based algorithm that could serve as an easily accessible screening tool for heart failure during routine physical exams.
Heart failure affects about 6.5 million adults in the U.S. and factored in one in every eight deaths in 2017, according to the U.S. Centers for Disease Control and Prevention. In 2012, the combined costs of health services, medication and missed days of work associated with heart failure totaled $30.7 billion.
Eko’s algorithm parses 15 seconds of ECG data collected from the Eko Duo digital stethoscope during an exam in the doctor’s office and helps to identify reduced left ventricular ejection fraction (LVEF), a common measure of heart failure.
Broad unmet need
Connor Landgraf, Eko’s co-founder and CEO, said the technology addresses a broader unmet need for screening tools in cardiovascular disease, given the number of people with heart disease and those that miss early detection.
“It’s kind of a novel concept because we didn’t know that there was a correlation between ECG and low ejection fraction previously, and ejection fraction is typically measured through more invasive imaging studies,” he told BioWorld MedTech. This technology “can help screen patients more accurately for diseases like heart failure or other diseases that cause poor heart function using a very simple, quick, noninvasive test that can make sure patients get identified early and treated early.”
Eko’s goal is that patients at risk for heart disease would regularly be screened using the ECG algorithm during office visits.
“A low ejection fraction means that the heart pump is weak, which can lead to shortness of breath, swelling, exercise intolerance or sudden death, so it is important to identify, as many treatments exist,” said Paul Friedman, chair of the Department of Cardiovascular Medicine at Mayo Clinic in Rochester, Minn., which collaborated on the development of the algorithm. “This technology gives physicians a tool to detect heart disease earlier, and before it develops into a more serious illness. In effect, by imbedding the technology in a commonly used clinical tool – the stethoscope – all caregivers carry some of the diagnostic prowess of an expert cardiologist with them.”
Eko’s AI and machine learning algorithms are designed to distinguish healthy from abnormal heart sounds using a deep neural network model trained on multiple datasets of heart sounds. The ECG algorithm was initially trained on 44,959 patients at Mayo Clinic to identify patients with ventricular dysfunction, defined as having an ejection fraction equal to or lower than 35%, using ECG data alone. It has been updated several times since based on different ECG factors, Landgraf said.
For the breakthrough designation, the company submitted data from two studies. In the first, published in January 2019 in Nature Medicine, the algorithm demonstrated area under the curve (AUC) values, sensitivity, specificity and accuracy of 0.93, 86.3%, 85.7% and 85.7%, respectively, in a cohort of 52,870 patients. For patient without ventricular dysfunction, those with a positive screen were four times more likely to develop ventricular function than those with a negative screen.
In a second study of 100 patients, the Duo stethoscope plus algorithm were able to detect ejection fraction lower than 35% with an AUC of 0.90. The results were presented at the American Heart Association Scientific Sessions in November.
Several other studies are ongoing, including validation studies and studies to train the algorithm on diverse populations, Landgraf said. He declined to estimate a time frame for regulatory approval, noting the submission will be a de novo application, which can take significantly more than a year in some situations. “We want to be very judicious with the FDA and go through evaluating all of the safety and efficacy considerations with the device,” he said.
More in the pipeline
The ECG is the first in a pipeline of algorithms Eko is developing. An algorithm to detect heart murmur, using Duo, is currently being tested at two clinical sites. The study aims to show that Eko’s stethoscope and AI algorithms can accurately interpret heart sounds to help screen for heart murmurs and valvular heart disease. Eko is also working on some arrhythmia screening algorithms, Landgraf said.
As for the ECG algorithm, Eko would like to extend it beyond heart failure to diseases like asymptomatic left ventricular dysfunction, a precursor to full-blown heart failure. With more than 6 million heart failure patients in the U.S. and hundreds of thousands of new cases diagnosed each year, there are many cases of heart function decline that fall outside the normal patterns, Landgraf said.
“One of the things that we’d like to study is what happens if we detect patients who are in the asymptomatic phase first and start them on a treatment program or start monitoring them more closely,” he added.
While Eko takes advantage of FDA breakthrough status to advance its product through the regulatory process, its cardiac home monitoring app, called Eko Home, is starting to gain traction with physician practices and health systems. Launched in July, the app allows clinicians to remotely track cardiac function of their patients in real time using ECG and heart sounds. The device can also be used to create drug-data combinations to demonstrate real-world efficacy for pharmaceutical trials, while capturing data outside the clinical environment.
Priced from $200 to $349, Eko’s handheld ECG and stethoscope devices aren’t likely to break the budget. Landgraf said the plan is to roll out the algorithm technology, once it is approved, via a software license.