Reclassification of cardiovascular risk using integrated clinical and molecular biosignatures: Design of and rationale for the Measurement to Understand the Reclassification of Disease of Cabarrus and Kannapolis (MURDOCK) Horizon 1 Cardiovascular Disease Study Svati H. Shah, MD, MS, MHS, a,b Christopher B. Granger, MD, a,c Elizabeth R. Hauser, PhD, b William E. Kraus, MD, a Jie-Lena Sun, MS, c Karen Pieper, MS, c Charlotte L. Nelson, MS, c Elizabeth R. Delong, PhD, c,d Robert M. Califf, MD, a,e and L. Kristin Newby, MD, MHS a,c for the MURDOCK Horizon 1 Cardiovascular Disease Investigators f Durham, NC Background Clinical predictive models leave gaps in our ability to stratify cardiovascular risk. High-throughput molecular profiling promises to improve risk classification. Methods Horizon 1 of the Measurement to Understand the Reclassification of Disease of Cabarrus and Kannapolis (MURDOCK) Study was conceived to apply emerging molecular techniques to existing data sets to characterize mechanistic diversity underlying complex human diseases, response to therapy, and prognosis. No previous studies have applied multiple, complementary molecular techniques in combination with well-developed clinical risk models to refine cardiovascular risk prediction. The MURDOCK Cardiovascular Disease Study will assess molecular profiles integrated with clinical data in “clinomic” profiles for cardiovascular risk classification. Conclusion Herein, we describe the design of and rationale for the MURDOCK Cardiovascular Disease Study. (Am Heart J 2010;160:371-379.e2.) Randomized clinical trials have identified several therapies that applied broadly to at-risk populations reduced the risk of myocardial infarction (MI) or death. Although such trials are integral to determining the safety and efficacy of drugs and devices, they define “average” responses over large cohorts of patients. “Personalized medicine,” or perhaps better stated “stratified medicine,” focuses on understanding the specific treatments and interventions most beneficial (and least harmful) in refined subgroups of patients. A more refined description of death or MI risk could facilitate tailored therapy and fuel discovery of new treatments. Blood-based biomar- kers may serve this purpose. Novel molecular techniques (eg, genomics, proteomics, and metabolomics) facilitate high-throughput measurement of diverse biomarkers. Application and integration of biomarkers measured by these techniques may yield more refined risk assessment than traditional risk factors alone, helping to achieve the goal of stratified medicine and providing clearer under- standing of underlying disease processes. Current models for risk prediction are insufficient The ability to discriminate risk using clinical character- istics and routine clinical laboratory testing is limited. The discriminative ability of the Framingham Risk Score (FRS), the most widely used clinical model for quantification of coronary heart disease (CHD) risk in the general population, is modest (c-index 0.69 in men and 0.72 in women). 1 The addition of high-sensitivity C-reactive protein (CRP) to the FRS model yieled a c-index of 0.81. 2 In the GRACE cohort, the c-index of a post–acute coronary syndrome (ACS) risk model that included From the a Division of Cardiovascular Medicine, Duke University Medical Center, Durham, NC, b Duke Center for Human Genetics, Duke University Medical Center, Durham, NC, c Duke Clinical Research Institute, Duke University Medical Center, Durham, NC, d Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC, and e Duke Translational Medicine Institute, Duke University Medical Center, Durham, NC. f See online Appendix for a complete listing of MURDOCK Horizon 1 Cardiovascular Disease Investigators. Submitted February 7, 2010; accepted June 24, 2010. Reprint requests: L. Kristin Newby, MD, MHS, Duke Clinical Research Institute, PO Box 17969, Durham, NC 27715-7969. E-mail: newby001@mc.duke.edu 0002-8703/$ - see front matter © 2010, Mosby, Inc. All rights reserved. doi:10.1016/j.ahj.2010.06.051 Trial Design