(VP117) THE USE OF HSTNI AS AN INDEPENDENT RISK FACTOR IN A LARGE CANADIAN COHORT: NEW INSIGHTS IN CARDIOVASCULAR RISK ASSESSMENT
Friday, October 27, 2023
13:30 – 13:40 EST
Location: ePoster Screen 10
Disclosure(s):
Daniel Esau, MD, FRCPC: No relevant disclosure to display
Background: High-sensitivity troponin I (hsTnI) improves risk stratification when added to traditional CV risk scores. However, hsTnI is not yet implemented widely in the ambulatory setting and it is unclear if it provides unique information on risk that is not already captured by traditional risk factors. The aim of this study was to explore a large contemporary Canadian cohort with hsTnI and assess predictors of CV risk in the real world.
METHODS AND RESULTS: 29832 patients had a hsTnI routinely measured as part of an executive health screen between January 2021 and April 2023. Information on patient demographics, lipid panel, Lp(a), CRP, and other lab markers was gathered through the clinic EMR. Patients were divided into low, medium, and high risk based on hsTnI values using sex specific cut offs recommended by the troponin assay manufacturer. Univariate analysis comparing hsTnI values to Framingham Risk Score (FRS) was performed with a Chi-squared test. Multivariate analysis using linear and ordered logistic regression with backwards stepwise removal was performed. The variation in repeat troponin measurement was examined.
Average age was 55.6 years (range 20-97). 81.7% of women and 56.9% of men had undetectable hsTnI. In the subset of patients with repeat troponin measurements, the majority did not show significant change and repeat measurement rarely led to change in the estimated risk category. hsTnI was independent of the FRS in univariate analysis. For multivariate analysis, male sex, elevated Lp(a), elevated systolic BP, and a lower non-HDL-C was associated with increased hsTnI in both linear and ordered logistic regression, while increasing age was associated with increased hsTnI only in ordered logistic regression. Both linear and ordered logistic regression models had poor fit, with only 1.3% to 3.6% of the variation in hsTnI explained by the covariates.
Conclusion: In this relatively young population, more than 50% of patients did not have detectable hsTnI. HsTnI was a stable biomarker over 2 years. Although hsTnI was dependent on age, sex, Lp(a), systolic BP, and non-HDL-C in regression analysis, model fit was poor. FRS was not associated with hsTnI. hsTnI therefore appears to provide unique information on cardiovascular risk that is not otherwise captured by traditional risk factors. This supports the use of hsTnI as an additional marker for CV risk in ambulatory settings.