This article is more than 5 years old. Information may no longer be current.
Protein-based system may predict CV outcomes in patients with stable CHD
Researchers have developed a system based on circulating proteins to anticipate risk for CV outcomes that performs better than the refit Framingham secondary-risk score in patients with stable CHD, according to study findings published in JAMA.
Peter Ganz, MD, from the department of medicine, University of California, San Francisco, and colleagues cautioned, however, that the system showed modest discrimination, and more research is needed to judge efficacy in a population at lower risk for CHD.
“In this case [when trying to predict risk for an adverse CV outcome] ... clinicians are typically less interested in discriminating risk between two patients and [are] more interested in being able to better calibrate a particular patient’s risk of adverse events,” Marc S. Sabatine, MD, MPH, from the TIMI study group, division of vascular medicine, Brigham and Women’s Hospital, Boston, wrote in an editorial about the study. “Ganz [and colleagues] showed the ability of their nine-protein model to do this. ... These data would be important to physicians and patients alike.”
Marc S. Sabatine
American, Norwegian cohorts
The researchers analyzed data from two populations: a derivation cohort of 938 baseline plasma samples from the Heart and Soul study consisting of patients (median age, 67 years; 82% men) with stable CHD from 12 clinics in the San Francisco area, and a validation cohort of 971 baseline plasma samples from HUNT3, consisting of patients (median age, 70 years; 72% men) with stable CHD from Nord-Trøndelag County in Norway.
Patients were excluded from either cohort if they had MI in the prior 6 months or were unable to walk one city block.
Proteins identified
The nine-protein model utilized by the researchers included the following: angiopoietin-2; matrix metalloproteinase-12; chemokine (C-C motif) ligand 18; complement 7; a1-antichymotrypsin complex; angiopoietin-related protein 4; troponin I; growth differentiation factor 11/8, and a2-antiplasmin. All were positively or negatively associated with the primary outcome of MI, stroke or transient ischemic attack, HF hospitalization or all-cause mortality.
The researchers compared the nine-protein model with the Framingham secondary event model, which was refit for the two cohorts in the present study.
For the San Francisco-area cohort, the C statistic measured 0.66 for the refit Framingham secondary-risk model, 0.74 for the nine-protein model and 0.75 for a combination of both models. For the Norwegian cohort, the C statistic was 0.64 for refit Framingham, 0.7 for the nine-protein model and 0.71 for a combination of both models.
Adding the nine-protein score to the refit Framingham model increased the C statistic by 0.09 (95% CI, 0.06-0.12) in the derivation cohort and by 0.05 (95% CI, 0.02-0.09) in the validation cohort, according to the researchers.
When compared with refit Framingham, the integrated discrimination index for nine-protein was 0.12 (95% CI, 0.08-0.16) for the San Francisco-area cohort and 0.08 (95% CI, 0.05-0.1) for the Norwegian participants.
“If some of these proteins are true risk factors rather than simply risk markers, then they could serve as targets for future therapies,” Sabatine wrote in JAMA. – by James Clark
Disclosure: SomaLogic provided the funding for and execution of the protein assay. Ganz reports no relevant financial disclosures. Please see the full study for a list of the other researchers’ relevant financial disclosures. Sabatine reports financial relationships with Abbott Laboratories, Accumetrics, Alnylam, Amgen, AstraZeneca, Bristol-Myers Squibb, Critical Diagnostics, Cubist, CVS Caremark, Daiichi-Sankyo, Duke Clinical Research Institute, Eisai, Genzyme, Gilead, GlaxoSmithKline, Intarcia, Merck, MyoKardia, Nanosphere, Novartis, Pfizer, Poxel, Quest Diagnostics, Roche Diagnostics, Sanofi-Aventis, Takeda and Zeus Scientific.