Association of high-sensitivity troponin T with outcomes in asymptomatic non-severe aortic stenosis: a post-hoc substudy of the SEAS trial

Abstract

Background: High-sensitivity Troponin T (hsTnT), a biomarker of cardiomyocyte overload and injury, relates to aortic valve replacement (AVR) and mortality in severe aortic stenosis (AS). However, its prognostic value remains unknown in asymptomatic patients with AS. We aimed to investigate if an hsTnT level >14 pg/mL (above upper limit of normal 99th percentile) is associated with echocardiographic AS-severity, subsequent AVR, ischaemic coronary events (ICE), and mortality in asymptomatic patients with non-severe AS.

Methods: In this post-hoc sub-analysis of the multicentre, randomised, double-blind, placebo-controlled SEAS trial (ClinicalTrials.gov, NCT00092677), we included asymptomatic patients with mild to moderate-severe AS. We ascertained baseline and 1-year hsTnT concentrations and examined the association between baseline levels and the risk of the primary composite endpoint, defined as the first event of all-cause mortality, isolated AVR (without coronary artery bypass grafting (CABG)), or ICE. Multivariable regressions and competing risk analyses examined associations of hsTnT level >14 pg/mL with clinical correlates and 5-year risk of the primary endpoint.

Findings: Between January 6, 2003, and March 4, 2004, a total of 1873 patients were enrolled in the SEAS trial, and 1739 patients were included in this post-hoc sub-analysis. Patients had a mean (SD) age of 67.5 (9.7) years, 61.0% (1061) were men, 17.4% (302) had moderate-severe AS, and 26.0% (453) had hsTnT level >14 pg/mL. The median hsTnT difference from baseline to 1-year was 0.8 pg/mL (IQR, -0.4 to 2.3). In adjusted linear regression, log(hsTnT) did not correlate with echocardiographic AS severity (p = 0.36). In multivariable Cox regression, a hsTnT level >14 pg/mL vs. hsTnT ≤14 pg/mL was associated with an increased risk of the primary composite endpoint (HR, 1.41; 95% CI, 1.18-1.70; p = 0.0002). In a competing risk model of first of the individual components of the primary endpoint, a hsTnT level >14 pg/mL was associated with ICE risk (HR 1.71; 95% CI, 1.23-2.38; p = 0.0013), but not with isolated AVR (p = 0.064) or all-cause mortality (p = 0.49) as the first event.

Interpretation: hsTnT level is within the reference range (≤14 pg/mL) in 3 out of 4 non-ischaemic patients with asymptomatic mild-to-moderate AS and remains stable during a 1-year follow-up regardless of AS-severity. An hsTnT level >14 pg/mL was mainly associated with subsequent ICE, which suggest that hsTnT concentration is primarily a risk marker of subclinical coronary atherosclerotic disease.

Funding: Merck & Co., Inc., the Schering-Plough Corporation, the Interreg IVA program, Roche Diagnostics Ltd., and Gangstedfonden. Open access publication fee funding provided by prof. Olav W. Nielsen and Department of Cardiology, Bispebjerg University Hospital, Denmark.

Keywords: AS, Aortic valve stenosis; Aortic valve stenosis; Biomarkers; CAD, Coronary artery disease; Coronary artery disease; High-sensitivity troponin T; ICE, Ischaemic coronary events; Ischaemic coronary events; hsTnT, High-sensitivity troponin T.

Fig. 1

High-sensitivity Troponin T concentration stratified by echocardiographic severity of aortic stenosis. The boxplots show hsTnT concentrations at baseline (grey) and year 1 (bluish) stratified by the severity of aortic stenosis (AS) in non-severe (left) and moderate-severe AS (right). The red dashed lines demarcate hsTnT level >14 pg/mL (above the upper limit of normal 99th percentile) and very high hsTnT level >50 pg/mL. p values reflect a two-sided paired sign test of medians for baseline vs. year 1 and a two-sample unpaired t-test for comparisons of non-severe AS vs. moderate-severe AS, respectively. Abbreviations: AS, aortic stenosis; hsTnT, high-sensitivity Troponin T.

Fig. 2

Multivariable Cox regression models. The figure shows Forest plot with event rates [95% CI] and Hazard Ratios [95% CI] based on Cox regression models examining the association between baseline high-sensitivity Troponin T concentrations ≤14 pg/mL and >14 pg/mL and outcomes: the primary composite end point and its individual components. Models were adjusted for standardised baseline values of age, sex, creatinine, and mean aortic gradient at baseline, stratified by center and lipid-lowering treatment. Follow-up from baseline to year 5. ^aThe first event of All-cause mortality, AVR without CABG, or ICE. ^bThe first event of CABG+AVR, myocardial infarction, or PCI. Abbreviations: AVR, aortic valve replacement; CABG, coronary artery bypass grafting; CI, confidence interval; hsTnT, high-sensitivity Troponin T; ICE, ischaemic coronary events; PCI, percutaneous coronary intervention.

Fig. 3

Stacked cumulative incidence function plots. The figure shows cumulative incidence of all-cause mortality (ACM) as the first and only event (grey), aortic valve replacement (AVR) without revascularisation (violet), and ischaemic coronary events (ICE) (pink) during follow-up from baseline to year 5 in patients with asymptomatic aortic stenosis, stratified by (a) baseline high-sensitivity Troponin T (hsTnT) levels ≤14 pg/mL and (b) hsTnT levels >14 pg/mL. Abbreviations: ACM, all-cause mortality; AVR, aortic valve replacement; hsTnT, high-sensitivity Troponin T; ICE, ischaemic coronary events.