Abstract 3357: Superiority of ApoB/ApoA1 Ratio for Predicting Cardiovascular Risk in Pooled Analyses of the Incremental Decrease in Endpoints through Aggressive Lipid-Lowering (IDEAL) and Treating to New Targets (TNT) Trials
Whether apolipoprotein B/A1 ratio (apo B/A1) is superior to other lipid, lipoprotein, and apolipoprotein parameters as a predictor of major cardiovascular events (MCVE) has been debated. Because data in patients on statin therapy are limited, year 1 results of 17987 patients from IDEAL (8529 of 8888 randomized) and TNT (9458 of 10001 randomized) were pooled to evaluate the relative value of apoB/A1, apo B, LDL-C, LDL-C/HDL-C, and other parameters in predicting risk of MCVE (composite of nonfatal MI, CHD death, resuscitated cardiac arrest, stroke). In a univariate analysis, a Cox proportional hazards regression model (CPHRM) adjusted for trial, age, and gender, was used to assess relationship of each parameter to MCVE risk. In a multivariate analysis, selected pairs of predictors were put into a CPHRM to assess independence of relationship of 1 predictor to risk after relationship of the other predictor had been taken into account. In the univariate analysis (Table⇓), all year 1 values were predictive of MCVE risk (all p<0.001) with rank order (highest z values) of apoB/A1, TC/HDL-C and LDL-C/HDL-C as the strongest predictors. In the multivariate analyses (Table⇓), LDL-C and apoB were both significant predictors of risk, but apoB was clearly a stronger predictor. Moreover, at similar apoB levels, an increase of LDL-C was associated with decreased MCVE risk, indicating that larger LDL particle size was associated with decreased risk. When LDL-C/HDL-C and apoB/A1 were paired, apoB/A1 was clearly a stronger predictor of risk.