Advances in Genetics Systems-Based Approaches to Cardiovascular Biomarker Discovery

The pace of discovery of cardiovascular biomarkers seems to be slowing. Although important advances in the identification of molecular biomarkers have been made during the past decade, it is becoming apparent that their clinical relevance is limited and that advanced discovery methods are needed. There are 27 000 articles on cardiac biomarkers in PubMed. However, only a small number of these biomarkers are in clinical use. In patients with heart failure (HF), numerous biomarkers have been evaluated (Table), but only brain natriuretic peptide and its precursor molecules have been widely applied.1 In patients with stable coronary artery disease, C-reactive protein (CRP) is a potential prognostic biomarker. Although there is conflicting evidence about the ability of CRP to augment the predictive accuracy of traditional clinical risk factors,2,3 targeting patients with elevated CRP for lipid lowering therapy with rosuvastatin was associated with improved cardiovascular outcomes in 1 randomized controlled trial.4 However, a recent meta-analysis of 83 studies including 60 000 patients has questioned its value.5 The key hurdle to the introduction of new “omic” biomarkers has been the inability to demonstrate their clinical relevance and validity. This refers not only to meeting analytic validity and independent evaluation criteria, but also to the biomarkers’ capacity to guide the improvement of patient outcomes. We believe that a more integrated, unbiased approach is essential to discover novel diagnostic and prognostic models of heart disease. Biomarker discovery research has traditionally emphasized the study of individual molecular indicators of clinical condition. A major limitation of this approach is that the informational complexity underpinning many clinical states are not adequately taken into account. The relevance of hypothesis-driven, reductionist approaches to biomarker discovery is indisputable. However, an overreliance on this strategy may limit the translation of fundamental research into new clinical applications. Limitations to traditional research should be seen as logical consequences of the limited ability to interrogate the multivariate and combinatorial characteristics of cellular networks implicated in multi-factorial common diseases. Biological networks sense, integrate and compute responses to thousands of signals in an integrative fashion. Moreover, the networks controlling physiological responses are dynamic, cell-specific and organized into complex functional hierarchies. Systems approaches can guide the identification of single biomarkers based on a deeper understanding of their underlying biology. In addition, they can enable strategies in which multiple biomarkers are assayed together, as a system, either in blood or tissue. In this context, biomarkers become clinical state indicators computationally inferred from networks of interacting molecular entities. Thus, biomarkers can represent measurements of network activity comprising not only the expression of biochemical substances (or their combinations), but also information on clinically meaningful biological interactions. In cardiovascular research, the potential of systems-based methodologies is being demonstrated in several application domains. They include atherosclerosis, HF, and other cardiovascular pathology measured in either tissue or blood.6,7 Such biomarkers have been implicated in biological processes as disparate as angiogenesis, inflammation, and oxidative stress. The discovery of such widely applicable biomarkers has been facilitated by advances in high-throughput molecular profiling, including transcriptomics and proteomics. In this review we discuss how the integration of data arising from different molecular profiling techniques are enabling a more detailed and biologically meaningful characterization of putative biomarkers. Such an integration of technologies and disciplines promises a more accurate classification of clinical conditions, while presenting opportunities for novel therapeutic interventions.