Prevencio’s HART PAD test was recently announced at European Society of Cardiology Congress in Munich. Utilizing biomarkers and clinical variables found and analyzed by AI, Prevencio was able to diagnose peripheral artery disease in diabetic patients with 95 percent accuracy and rule it out with 100 percent accuracy.
This technology can make a major implications for diabetics as it is a patient population in which PAD has traditionally been difficult to assess. The HART PAD panel consists of six biomarkers and one clinical variable generated and evaluated with AI. The system has the potential to improve patient outcomes with early diagnosis.
To discover more, Digital Journal connected with Prevencio’s CEO Rhonda Rhyne to discuss the data, the HART PAD test and its implications.
Digital Journal: How is artificial intelligence reshaping healthcare?
Rhonda Rhyne: More broadly, Artificial Intelligence is reshaping healthcare in everything from developing new drugs, selecting personalized therapies for a given patient, providing apps to help patients manage their conditions, optimizing billing, and more accurately diagnosing patients for given conditions, based on the mass amounts of data available. When it comes to accurately diagnosing cardiovascular disease—the number one killer of men and women—this is where Prevencio has successfully capitalized on the power of AI in healthcare.
DJ: How accurate is artificial intelligence in terms of making diagnoses?
Rhyne: As with anything, the quality of the inputs drives the quality of the outputs. When Artificial Intelligence platforms have access to well-characterized clinical data sets, those platforms can look for signatures that can diagnose or risk stratify patients with accuracy superior to many current standards of care including functional tests (e.g. exercise stress tests) or individual biomarkers.
In the case of Prevencio, we have published or presented data on four of our tests that demonstrate Positive and Negative Predictive values superior to current methods for diagnosing conditions such as Coronary Artery Disease (CAD), Peripheral Artery Disease (PAD), and Aortic Stenosis (AS). One example is our HART CAD blood test which has a 91 percent Negative Predictive Value (NPV) and a 93 percent Positive Predictive Value (PPV) in diagnosing CAD.1
DJ: What is peripheral artery disease and what is the association with diabetes?
Rhyne: Peripheral artery disease (PAD) is a narrowing of the peripheral arteries serving the legs, stomach, arms and head. PAD most commonly affects arteries in the legs. PAD is commonly due to atherosclerosis, a plaque buildup in the arteries which restricts blood flow.
The presence of diabetes mellitus (DM) greatly increases the risk of PAD, as well as accelerates its course, making these patients more susceptible to ischemic events and impaired functional status compared to patients without diabetes.
DJ: How did Prevencio develop its AI system for peripheral artery disease in diabetic patients?
Rhyne:Prevencio employs Artificial Intelligence to interrogate well-characterized clinical data sets to produce novel, multi-protein, algorithmically-scored tests. As a part of this approach, Prevencio analyzes proteins that are considered more reflective of the progression of disease states. While genomics has been investigated more extensively, genomic risk scores do not substantially improve risk discrimination over traditional risk factors. Proteins are more dynamic than genes and are much closer to disease development and expression.
Prevencio uses multi-analyte algorithmic protein tests which surpass the clinical value of any single analyte cardiac biomarker test. Single protein tests, specifically troponin and BNP, revolutionized patient care but, due to the multifactorial issues related to cardiac disease, single protein tests have inherent limitations.
DJ: What is the HART PAD panel?
Rhyne:The clinical/proteomic HART PAD panel consists of one clinical variable (history of hypertension) and concentrations of six biomarkers (midkine, kidney injury molecule-1, interleukin-23, follicle stimulating hormone, angiopoietin-1, and eotaxin-1).
DJ: How accurate is the system and how was this assessed?
Rhyne:In patients with DM, the HART PAD panel had excellent performance for diagnosis of PAD. Using a 5-point score, a score of 1 had a 100% Negative Predictive Value (NPV), a very low likelihood of PAD, and a score of 5 had a 95 percent Positive Predictive Value (PPV), a very high likelihood of PAD. Additionally, the HART PAD panel was highly accurate in predicting the need for revascularization in patients with PAD. These results were comparable to those patients without DM.
In a prospective cohort of 354 patients referred for diagnostic peripheral angiography and/or coronary angiography enrolled in the Catheter Sampled Blood Archive in Cardiovascular Diseases (CASABLANCA) Study (Clinical Trials.Gov NCT00842868), performance of this diagnostic panel was assessed in patients with and without DM using Monte Carlo cross validation.
DJ: How is the data captured and stored?
Rhyne:All of the initial research data from the CASABLANCA Study (see above) was captured and stored by Massachusetts General Hospital. Prevencio utilized several proprietary mechanisms for capturing, storing, and backing up our raw, processed and analyzed data.
DJ: Has the system been approved by regulators?
Rhyne:No, our tests are not approved for clinical purposes. However, Prevencio is able to offer testing services for Research Use to pharmaceutical companies and researchers. Our HART tests can assist investigators to confirm the clinical utility, safety and efficacy of treatment, screen/target patients to enroll or exclude from a study, and assess the potential for anticipated/unanticipated effects. Prevencio is conducting additional studies with plans to submit to various regulatory agencies once those studies are complete.
DJ: What has been the response from the medical profession?
Rhyne:We have received tremendous interest and support from medical professionals, including Dr. James Januzzi, practicing cardiologist at Massachusetts General Hospital and Professor of Medicine at Harvard Medical School, as well as other key opinion leaders.
Dr. James Januzzi has said there is a huge unmet need to accurately diagnose the presence of cardiovascular and peripheral arterial diseases. Though single biomarker tests have helped to refine diagnosis and prognosis, it is clear the future will be with a multiple biomarker approach supported by cutting-edge machine learning.
DJ: What other projects are you working on?
The five blood tests in our current product pipeline are: HART CAD, a blood test for obstructive coronary artery disease diagnosis; HART CVE a blood test for 1-year risk of heart attack, stroke or cardiac death; and HART PAD a blood test for peripheral artery disease diagnosis. Also, HART AS, a blood test for aortic valve stenosis, and HART AMP, a blood test for risk of amputation.
Rhyne: We are also working on tests to evaluate acute kidney injury risk, atrial fibrillation risk, and a diagnostic for transthyretin amyloidosis (ATTR) cardiomyopathy.