Impact of Apolipoprotein(a) Isoform Size on Lipoprotein(a) Lowering in the HPS2-THRIVE Study

Abstract

Background: Genetic studies have shown lipoprotein(a) (Lp[a]) to be an important causal risk factor for coronary disease. Apolipoprotein(a) isoform size is the chief determinant of Lp(a) levels, but its impact on the benefits of therapies that lower Lp(a) remains unclear.

Methods: HPS2-THRIVE (Heart Protection Study 2-Treatment of HDL to Reduce the Incidence of Vascular Events) is a randomized trial of niacin-laropiprant versus placebo on a background of simvastatin therapy. Plasma Lp(a) levels at baseline and 1 year post-randomization were measured in 3978 participants from the United Kingdom and China. Apolipoprotein(a) isoform size, estimated by the number of kringle IV domains, was measured by agarose gel electrophoresis and the predominantly expressed isoform identified.

Results: Allocation to niacin-laropiprant reduced mean Lp(a) by 12 (SE, 1) nmol/L overall and 34 (6) nmol/L in the top quintile by baseline Lp(a) level (Lp[a] ≥128 nmol/L). The mean proportional reduction in Lp(a) with niacin-laropiprant was 31% but varied strongly with predominant apolipoprotein(a) isoform size (P_Trend=4×10^-29) and was only 18% in the quintile with the highest baseline Lp(a) level and low isoform size. Estimates from genetic studies suggest that these Lp(a) reductions during the short term of the trial might yield proportional reductions in coronary risk of ≈2% overall and 6% in the top quintile by Lp(a) levels.

Conclusions: Proportional reductions in Lp(a) were dependent on apolipoprotein(a) isoform size. Taking this into account, the likely benefits of niacin-laropiprant on coronary risk through Lp(a) lowering are small. Novel therapies that reduce high Lp(a) levels by at least 80 nmol/L (≈40%) may be needed to produce worthwhile benefits in people at the highest risk because of Lp(a).

Clinical trial registration: URL: https://clinicaltrials.gov. Unique identifier: NCT00461630.

Keywords: apolipoprotein(a); clinical trial; coronary disease; isoform size; lipoprotein(a); therapeutics.

Figure 1.

Baseline lipoprotein(a) (Lp[a]) levels by the kringle IV domains of the predominantly expressed apolipoprotein(a) isoform, within the United Kingdom and China. Blue circles denote United Kingdom, and red squares denote China. The 11 kringle IV domain groups shown are ≤15, 16, 17, 18, 19, 20 to 21, 22 to 23, 24 to 25, 26 to 27, 28 to 30, and ≥31. Lp(a) indicates lipoprotein(a).

Figure 2.

Percentage and absolute reductions in lipoprotein(a) (Lp[a]) with niacin–laropiprant by quintiles of baseline Lp(a) levels and kringle IV domains. Percentage reduction panels show adjusted Lp(a) reductions based on modeling log_e Lp(a) (Methods) and are plotted with an x scale linear in the log_e Lp(a) difference. Absolute reductions (with adjustment by stratification) are plotted on linear scales. Usual mean Lp(a) is the mean Lp(a) in the placebo arm at 1 y.