• Cardiology Genomics
• Heart Failure

Studies have suggested using a panel of biomarkers that measure diverse biological processes as a prognostic tool for heart failure (HF). Thomas P. Cappola, MD, ScM, University of Pennsylvania, Philadelphia, Pennsylvania, USA, presented the Penn Heart Failure Study based on the hypothesis that multiple biomarkers considered together are superior to clinical risk stratification in patients with chronic HF [Ky B et al. Circ Heart Fail 2012]. The aim of the study was to derive a biomarker score in ambulatory HF patients that predicts time to transplant, left ventricular assist device (LVAD), or death and to compare its performance to the Seattle Heart Failure Model (SHFM).

A total of 1513 patients with HF were evaluated with biomarker analysis on banked serum, plasma, and DNA; 2D echocardiogram; and detailed clinical covariates (SHFM). Eight candidate biomarkers that measure distinct biological processes and are individually associated with adverse outcomes were evaluated using high-quality assays: troponin I (TnI) for myocyte injury, creatinine for renal function, soluble toll-like receptor-2 (ST2) for myocyte stress, soluble fms-like tyrosine kinase receptor-1 (sFlt-1) for vascular remodeling, B-type natriuretic peptide (BNP) for neurohormones, myeloperoxidase and uric acid for oxidative stress, and high-sensitivity C-reactive protein (hsCRP) for inflammation.

At a median 2.5 years follow-up, 317 events had been reported, including 31 LVADs, 99 transplants, and 187 deaths. After biomarker evaluation using multiple methods, 7 markers remained in the multimarker score: BNP, sFlt-1, hsCRP, ST2, TnI, uric acid, and creatinine. The multimarker score was a strong predictor of adverse outcomes. The hazard ratio for adverse outcomes was 4.7 in patients with a moderate multimarker score and 15 in patients with a high multimarker score. The multimarker score was a stronger predictor of adverse outcomes than the SHFM score (AUC, 0.798; 95% CI, 0.763 to 0.833; p<0.01). Adding the multimarker score to the SHFM led to a significantly improved AUC of 0.803 (95% CI, 0.769 to 0.837; p<0.01). After SHFM risk stratification, the multimarker score improved risk assignment in 20% of patients (95% CI, 4.8% to 35.1%; p=0.01).

The derived multimarker score comprised of 7 biomarkers is an accurate predictor of adverse outcomes and has improved predictive accuracy compared to a clinical risk score. The same results were obtained using multiple analytic approaches. Broader screens using unbiased technologies are needed, as are clinical trials to prove the utility of biomarkers for risk prediction.

• © 2012 MD Conference Express®