Stratifying Cardiovascular Risk
Cardiovascular disease (CVD) accounts for one out of three deaths in the U.S., with an average of one death occurring every 40 seconds. The economic burden is staggering, with medical costs in excess of $300 billion each year; this does not include productivity losses. The latest projections estimate that by 2035, nearly half the U.S. population with have some form of CVD, with associated medical costs nearing $749 billion.
Many people (an estimated 15.5 million in the U.S. alone) live with stable coronary heart disease (CHD). For these patients, the goal is to prevent a second attack, and the recommended treatment options are largely the same — lifestyle changes (e.g. diet, exercise, smoking cessation) and medications (e.g. aspirin, nitroglycerin, beta blockers, angiotensin inhibitors, statins). Given the wide variation in disease severity among those with chronic CHD, and the advances in management options and surgical techniques, more precise stratification of cardiovascular risk in patients is needed to inform treatment decisions.
Researchers at the University of California, San Francisco and SomaLogic used the SomaScan Assay to measure the levels of 1130 proteins in approximately 2,500 plasma samples from two different cohorts of patients with stable CHD. These data were used to identify and validate a set of nine proteins whose levels could predict the four-year probability of a heart attack, stroke, congestive heart failure or death. The protein-based risk score proved better at stratifying patients than a model that uses traditional cardiovascular health indicators, including age, sex, cholesterol, blood pressure and smoking status.
Paired samples, in which a second sample from the same patient was collected five years after the first, were used to evaluate whether the nine-protein risk score changed over time. The risk scores became significantly worse for 139 patients who suffered an adverse cardiovascular event after the second sample than for 375 patients who remained event-free. The response of the protein-based risk score to future events offers a potential advantage over genetic risk prediction, which remains unchanged during lifetime.
In a follow-on study, researchers at Pfizer, the Karolinska Institute, the University of California, San Francisco and SomaLogic found that the nine protein-based risk score predicted problems with a cardiovascular drug candidate (torcetrapib) well before significant adverse effects were seen in a previously conducted clinical trial. Their analysis also identified changes in several proteins that give a better understanding of the biology behind the problems. This work shows how profiling proteins could provide early warning of off-target effects and help speed drug development.
Ganz, P et al. (2016) “Development and validation of a protein-based risk score for cardiovascular outcomes among patients with stable coronary heart disease.” JAMA 315(23):2532-2541.
Williams, SA et al. (2018) “Improving assessment of drug safety through proteomics: early detection and mechanistic characterization of the unforeseen harmful effects of torcetrapib.” Circulation 137(10):999-1010.