Purpose: Heart failure is a condition associated with high morbidity and mortality. Our current pharmacotherapeutic interventions can at best slow down the progression of the disease. At present, a proper molecular characterization of the disease process is lacking. Biomarkers that provides insight into the molecular mechanisms underlying the malfunctioning of the heart could be helpful in this respect. Because inflammation contributes to the symptoms of heart failure, we determined the circulating levels of three related chemokines, CXCL-9, -10 and -11, as well as NT-proBNP in serum from subjects with left ventricular dysfunction (LVD) and controls.
Methods: Subjects with either subclinical (n=17) or advanced (N=14) LVD as well as age- and sex-matched controls (n=32), were recruited from the large-scale family-based study on the genetic epidemiology of cardiovascular phenotypes (FLEMENGHO). Mean arterial blood pressure was significantly higher in the subclinical (104.8±8.6 mmHg) and advanced (101.1±11.7 mmHg) LVD groups vs. control (90.5±4.5 mmHg). One-way ANOVA, Odds ratios, Integration Discrimination Improvement and Net Reclassification Improvement were determined using SAS software.
Results: Increased serum levels of CXCL-9, -10 and -11 were observed in the LV dysfunction groups (1.5, 1.3 and 1.9-fold in the subclinical LVD and 4.1, 2.3 and 2.7-fold in the advanced LVD group, respectively; p<0.01, <0.001 and <0.01, respectively). NT-proBNP was 1.6 and 2.0-fold higher in the subclinical and advanced LVD groups; p<0.01). Adjustment for sex, age, BMI and mean arterial pressure resulted in a loss of significance for CXCL10 and NT-proBNP, whereas de difference for CXCL9 remained significant (p=0.03) and for CXCL11 a trend towards significance was observed (P=0.06). Doublings of CXCL-9, -10, -11 and NT-proBNP were associated with a 103%, 112%, 113% and 172% higher risk of subclinical and advanced LVD (p<0.019), respectively. Sensitivities were 0.58, 0.65, 0.58 and 0.76 and specificities 0.88, 0.69, 0.84 and 0.69 for CXCL-9, -10 and -11 and NT-proBNP, respectively. When corrected for age and BMI, adding the dichotomized biomarkers significantly enhanced the net reclassification improvement (p<0.006).
In conclusion, including CXCL-9, -10 and -11 levels improves the discrimination of risk prediction models for LVD. The results of this study underscore the importance of using a panel of biomarkers to better characterize subjects with LVD.