Morin, Daniel P.; Chong-Yik, Ronald; Thihalolipavan, Sudarone; Krauthammer, Yoaav S.; Bernard, Michael L.; Khatib, Sammy; Polin, Glenn M.; Rogers, Paul A.
Aims Evidence links markers of systemic inflammation and heart failure (HF) with ventricular arrhythmias (VA) and/or death. Biomarker levels, and the risk they indicate, may vary over time. We evaluated the utility of serial laboratory measurements of inflammatory biomarkers and HF, using time-dependent analysis.; Methods and results We prospectively enrolled ambulatory patients with left ventricular ejection fraction (LVEF) <= 35% and a primary-prevention implanted cardioverter-defibrillator (ICD). Levels of established inflammatory biomarkers [C-reactive protein, erythrocyte sedimentation rate (ESR), suppression of tumourigenicity 2 (ST2), tumour necrosis factor alpha (TNF-alpha)] and brain natriuretic peptide (BNP) were assessed at 3-month intervals for 1 year. We assessed relationships between biomarkers modelled as time-dependent variables, VA, and death. Among 196 patients (6614 years, LVEF 238%), 33 experienced VA, and 18 died. Using only baseline values, BNP predicted VA, and both BNP and ST2 predicted death. Using serial measurements at 3-month intervals, time-varying BNP independently predicted VA, and time-varying ST2 independently predicted death. C-statistic analysis revealed no significant benefit to repeated testing compared with baseline-only measurement. C-reactive protein, ESR, and TNF-alpha, either at baseline or over time, did not predict either endpoint.; Conclusion In stable ambulatory patients with systolic cardiomyopathy and an ICD, BNP predicts ventricular tachyarrhythmia, and ST2 predicts death. Repeated laboratory measurements over a year's time do not improve risk stratification beyond baseline measurement alone.