Plasma proteomics to identify people at risk of sudden heart attack

Background: Obstructive coronary artery disease (CAD) is the underlying cause of heart attack and stroke responsible for near 2/3 of global deaths. In 45% of people with obstructive CAD, the disease is silent. The current non-invasive diagnostic standard does not identify obstructive but silent CAD and inaccurately identify patients who present with symptoms of CAD.

As a consequence, the selection of patients for invasive diagnostics that accurately identifies obstructive CAD suffers from low sensitivity and specificity, leading to both under- and over-use of invasive coronary angiography in patients with and without obstructive CAD. Recent evidence suggest that the plasma proteome is a strong indicator of public health.

By analyzing the plasma proteome of the Stockholm-Tartu Atherosclerosis Reverse Engineering Network Task (STARNET) study, we provide compelling evidence that features of plasma proteins are powerful predictors of obstructive CAD. Specifically, machine learning analysis of the plasma proteome in STARNET identified 98 protein features that reliably predict the presence of obstructive CAD with receiver operating characteristics (ROC) of >90% area under the curve (AUC). In contrast, ROCs of Framingham heart and polygenetic risk scores of CAD fell short of 65% AUCs.

Objectives: To develop a simple blood test that accurately identifies people with obstructive CAD at risk of a sudden heart attack or stroke.

Work plan: Aim 1. Develop a diagnostic test panel (“Beat-IT”) of the most predictive plasma protein features for obstructive CAD allowing for large scale retrospective cross-sectional and prospective longitudinal clinical validation studies. Aim 2.

Retrospectively validate the Beat-IT panel in 14,000 patients with and without CAD from the Estonian Biobank. Aim 3. To prospectively validate Beat-IT by enrolling 1000 healthy volunteers and 1000 patients with suspected obstructive CAD followed by angiographic assessments in the PREDICT2 study at Huddinge Hospital in Stockholm.

Significance: With a simple blood test that accurately identifies patients with obstructive CAD in need of preventive care, the individual suffering and staggering cost to societies of CAD-related morbidity and mortality can be effectively addressed.