170Lipoprotein subclasses associated with high-risk coronary atherosclerotic plaque: insights from the PROMISE clinical trial

Abstract

Abstract Background Greater than half of major adverse cardiovascular events (MACE) occur in individuals with non-obstructive coronary artery disease (CAD) and may be related to the presence of high-risk coronary atherosclerotic plaque (HRP). HRP can be detected by coronary computed tomographic angiography (CTA) and is associated with increased MACE risk. Biomarkers associated with HRP may therefore aid in risk stratification in individuals with chest pain and provide insight into the biology of HRP. Purpose To determine whether nuclear magnetic resonance (NMR)-measured lipoprotein particles are associated with HRP independent of risk factors. Methods This study included 1755 stable symptomatic outpatients enrolled in the PROMISE trial who had coronary CTA performed. Thirty-seven lipoprotein particle parameters were measured in plasma using NMR. Principal components analysis was used to reduce the number of correlated lipoproteins into a smaller number of uncorrelated factors. HRP cases (N=277) were defined as presence of HRP features (positive remodeling, low CT attenuation or napkin-ring sign) with or without obstructive CAD; controls had no HRP and no obstructive CAD (N=1478). Multivariable logistic regression models adjusting for age, sex, diabetes, hypertension, BMI, smoking status and statin use were used to test for association of lipoprotein factors with HRP case/control status; lipoproteins loaded within significant factors (p<0.05) were tested individually. Finally, because event rates in PROMISE were too low, we used a separate study (CATHGEN, a study of individuals undergoing coronary angiography; N=8707) to test for association of lipoproteins with MACE using multivariable Cox proportional hazard models. Results HRP cases were less likely to be female than controls (36.8% vs. 56.6%) and were more likely to have diabetes (23.5% vs 19.6%), but were of similar age (60.1 vs 60.4 years [SD±8]) and with similar statin use (45.2% vs 44.4%). In multivariable models, two lipoprotein factors were associated with HRP: an HDL factor (OR 1.29 [95% CI, 1.10–1.52], p=0.002) and a triglyceride (TG) factor (OR 1.19 [1.03–1.36], p=0.01). Within these two factors, H6P (a large HDL subclass; OR 0.67 [0.54–0.82], p<0.001), HDL size (OR 0.78 [0.66–0.92], p=0.003) and small TG-rich lipoproteins (OR 1.22 [1.07–1.39], p=0.002) were associated with HRP. H6P (HR 1.1 [1.07–1.13], p<0.0001) and HDL size (HR 1.28 [1.24–1.32], p<0.0001) were also associated with time to death or MI in CATHGEN, but in the opposite direction. Conclusions In a cohort of low-risk patients from the PROMISE trial, large HDL and small TG particles were associated with CTA-defined HRP independent of risk factors. Large HDL particles were also associated with incident MACE in a separate high-risk cohort. The discordant direction of association of HDL subclasses with HRP and MACE is likely due to underlying HDL biology and requires further study. Acknowledgement/Funding The PROMISE trial was funded by the National Heart, Lung, and Blood Institute (grants R01HL098237, R01HL098236, R01HL98305, and R01HL098235)