AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Comparison of Lipoprotein(a) and Other Apolipoprotein B Containing Lipoproteins as Predictors of Major Adverse Cardiovascular Events in REDUCE-IT
By SZAREK M, Bhatt DL, Miller M et al.
Background: Lipoprotein(a) (Lp(a)) and other atherogenic lipoproteins, including low density lipoprotein (LDL), contain one molecule of apolipoprotein B (apoB) per particle. Recent studies indicate Lp(a) is more atherogenic and prognostic for major adverse cardiovascular events (MACE) than LDL-C on a per particle basis. However, because there are far fewer Lp(a) particles than LDL particles, the overall risk from LDL is usually much greater. In REDUCE-IT, icosapent ethyl (IPE) consistently reduced MACE across the distribution of baseline Lp(a) levels, including among those with clinically relevant elevations. The present post hoc analysis tested the hypothesis that, after LDL-C control with a statin, Lp(a) is more strongly associated with MACE on a per particle basis than non-Lp(a) apoB-containing particles. Methods: Baseline molar Lp(a), estimated as measured mass concentration х 2.5, and (molar) apoB were assessed in 6988 REDUCE-IT participants, randomized to receive 2g twice daily of IPE or matching placebo and followed for a median 4.9 years. Study inclusion criteria included statin therapy, LDL-C 41 to 100 mg/dL, established CV disease, or diabetes plus ≥1 additional risk factor, and baseline fasting TG 150 to 499 mg/dL. Non-Lp(a) apoB was calculated as total apoB - Lp(a), all in nmol/L. ApoB, non-Lp(a) apoB and Lp(a) in both treatment groups were related to risk of the study primary endpoint MACE with proportional hazards models. Results: Median (Q1, Q3) apoB, non-Lp(a) apoB and Lp(a) concentrations were 1591 (1397, 1804), 1528 (1328, 1738) and 29 (13, 93) nmol/L, respectively. Adjusted for treatment assignment, HRs (95% CI) for MACE per 50 nmol/L increment were 1.006 (0.998, 1.015) (p=0.13) for apoB alone and, in a joint model, 1.004 (0.996, 1.012) (p=0.36) for non-Lp(a) apoB and 1.062 (1.034, 1.091) (p<0.0001) for Lp(a). Results were similar for subgroups with Lp(a) ≥150 (n=1142) or <150 nmol/L (n=5846) (Figure). Applying different multipliers of Lp(a) mass to estimate molar concentration (2.25, 2.75) yielded comparable results. Analysis by quartile revealed Lp(a)-associated risk was restricted to the highest quartile (>93 nmol/L; Table). Conclusion: Among participants with statin-controlled LDL-C in REDUCE-IT, Lp(a) was related to risk of MACE whereas non-Lp(a) apoB was not prognostic, including those with clinically relevant elevations in Lp(a). These findings suggest Lp(a) is an important therapeutic target among those with LDL-C controlled on statins, but with elevated baseline TG, independent of IPE treatment.