MyoKardia today announced the publication of an article titled, “Machine Learning Detection of Obstructive Hypertrophic Cardiomyopathy (oHCM) Using a Wearable Biosensor,” in the Nature Partner Journal, Digital Medicine. This research is part of MyoKardia’s efforts aimed at improving the detection, diagnosis and treatment of HCM, and holds the potential to help physicians more easily identify people who may be at risk of disease.
Results from an exploratory study provided encouraging evidence of the potential for a wrist-worn biosensor to screen for obstructive hypertrophic cardiomyopathy. The study demonstrated that continuous monitoring using a wrist-worn photoplethysmography (PPG) digital health device, similar to the optical sensors that monitor heart rate on commercially available fitness trackers, revealed differences in arterial pulse wave patterns between oHCM patients and those of individuals without oHCM. MyoKardia’s proprietary machine learning algorithm identified individuals with oHCM with a sensitivity of 0.95 and a specificity of 0.98. The digital health substudy was conducted by MyoKardia as part of the company’s Phase 2 PIONEER-HCM trial of mavacamten.
“Hypertrophic cardiomyopathy, or HCM, is associated with an increased risk for heart failure, stroke, and sudden death, even in asymptomatic patients. However, HCM is vastly underdiagnosed, as only approximately 15 percent of the estimated 1 in 500 people with HCM receive the correct diagnosis. The symptoms of HCM, such as shortness of breath, exercise intolerance, or fatigue, are nonspecific, and our research indicates that it can take patients up to three or more years after symptom onset to receive a diagnosis. The results of this study suggest that broadly available wrist-worn biosensor technologies may be able to identify undiagnosed oHCM patients who could be at risk of developing serious cardiac complications,” said Marc Semigran, M.D., Senior Vice President of Medical Sciences at MyoKardia, and senior author of the article. “As part of our commitment to improving the lives of patients, we have integrated digital health technologies into our interventional clinical studies, with the aim of advancing the diagnosis, treatment and monitoring of cardiomyopathies.”
Detecting oHCM Using a Wearable: Substudy Design and Findings
Using an investigational wearable PPG device, vascular pulse wave data were collected at baseline from 19 oHCM patients enrolled in MyoKardia’s Phase 2 PIONEER-HCM clinical trial. The pulse wave patterns of oHCM patients were compared to 64 healthy volunteers with no evidence of left ventricular hypertrophy, obstruction of the left ventricular outflow tract (LVOT), or other cardiovascular disease. A beat-by-beat machine learning model was developed to identify digital signatures of oHCM. The algorithm combined features corresponding to known hemodynamic abnormalities in HCM with novel morphological patterns extracted from the PPG signal. The pulse wave patterns of patients with oHCM were consistent with the turbulent blood flow and beat-to-beat variability associated with LVOT obstruction characteristic of oHCM. When patterns were examined across multiple beats, pulse wave traces from oHCM patients showed more frequent irregularly shaped beats and greater variability from beat to beat than those of the healthy volunteers. After training and cross-validation, the model achieved a C-statistic for oHCM detection of 0.99 (95% CI:0.99–1.0). At an operating threshold that optimizes the sum of sensitivity (95%) and specificity (98%), the model correctly classified 18/19 patients with oHCM and 63/64 healthy volunteers (98% accuracy).
Data from this PPG substudy were previously presented at the 2017 American Heart Association Scientific Sessions in a presentation titled, “Machine Learning Detection of Obstructive Hypertrophic Cardiomyopathy Using a Wearable Biosensor.”
About Obstructive HCM
Hypertrophic cardiomyopathy is the most common genetic cause of heart disease in which the walls of the heart thicken and prevent the left ventricle from expanding, resulting in a reduced pumping capacity. HCM is a chronic disease and for the majority of patients, the disease is progressive and can be extremely disabling. In approximately two-thirds of HCM patients, or an estimated 65,000 of the 100,000 people diagnosed in the U.S., the path followed by blood exiting the heart, known as the left ventricular outflow tract (LVOT), becomes obstructed by the enlarged and diseased muscle, restricting the flow of blood from the heart to the rest of the body. Mild exertion can quickly result in fatigue or shortness of breath, and a patient’s ability to participate in normal work, family or recreational activities can be substantially curtailed. Patients with oHCM are at an increased risk of severe heart failure and death. HCM can also cause stroke or sudden cardiac death.