Abstract 020: NT-proBNP, hs-Troponin T and Incident Venous Thromboembolism: The Atherosclerosis Risk in Communities (ARIC) Study

Abstract

Background: NT-pro B-type natriuretic peptide (NT-proBNP) and high sensitivity cardiac troponin T (TnT) are important biomarkers for CHD and HF risk, but whether they are related to incidence of venous thromboembolism (VTE) is unknown. NT-proBNP is a marker of volume overload; volume overload may contribute to VTE risk via increased stasis and thrombogenesis. TnT might relate to VTE indirectly, as a marker of vascular inflammation or endothelial dysfunction. Hypotheses: NT-proBNP and TnT are associated positively with risk of incident VTE. Methods: Data from visit 4 (1996-1998) of the prospective population-based Atherosclerosis Risk in Communities study (n = 10,735; 53-74 yrs) were used. VTE events through 2005 were identified and validated by medical record review. Cox proportional hazards regression was used to explore the relation between the biomarkers and VTE. Model 1 adjusted for age, sex, and race. Model 2 additionally adjusted for HRT use, diabetes status, BMI, eGFR, CRP, aPTT and factor VIII. Results: A total of 227 incident VTE cases accrued over a median follow-up of 8.0 yrs. Relative to participants in the lowest quintile of NT-proBNP (≤26.7 pg/mL), those in the highest quintile (≥151.3 pg/mL) were at 2.28 (95% CI: 1.47-3.54) times greater risk of incident VTE (p-trend = 0.0002), adjusted via Model 1. This association was only slightly attenuated after adjustment in Model 2 [HR: 1.98 (95% CI: 1.25, 3.13); p-trend = 0.003]. The general trend was present for both unprovoked [HR: 1.62 (95% CI: 0.70, 3.75)] and provoked VTE [HR: 2.22 (95% CI: 1.28, 3.84)] in the fully-adjusted models, though power was low when stratifying by VTE type. Participants in the highest category (≥0.014 ug/L) of TnT were also at modestly greater risk of incident VTE in Model 1 [HR: 1.99 (95% CI: 1.23, 3.21); p-trend = 0.0002] and Model 2 [HR: 1.52 (95% CI: 0.89, 2.60); p-trend = 0.03], relative to those with undetectable TnT. The association was, however, present for only provoked VTE, with no evidence of a relation for unprovoked VTE. Discussion: In this large prospective cohort, both NT-proBNP and TnT were positively associated with risk of incident VTE. For NT-proBNP, the general trend was present for both provoked and unprovoked events, while for TnT the relation was isolated to provoked VTE events. Additional research is warranted to evaluate whether NT-proBNP and TnT are related to risk of VTE, or if the associations observed were the result of uncontrolled confounding.