| |
Lipoprotein(a), PCSK9 Inhibition, and Cardiovascular Risk
|
| |
| |
Download the PDF here
Introduction
Editorial, see p 1493
Lipoprotein [Lp(a)] consists of an low-density lipoprotein (LDL)-like particle that also contains apolipoprotein(a) [apo(a)] linked to apolipoprotein B. Lp(a) plasma concentrations are highly heritable and predominantly controlled by the apo(a) gene (LPA).1 Several epidemiological studies have demonstrated an association between higher plasma Lp(a) concentrations and coronary risk2; however, the strength of the association in patients with well-controlled plasma LDL cholesterol (LDL-C) concentration has been inconsistent.3-5
It is important to note that genetic studies support that Lp(a) plays a causal role in the development of coronary atherosclerosis; in particular, data from 2 large Mendelian randomization studies demonstrated that genetic polymorphisms in the LPA gene are associated with Lp(a) concentration and future coronary risk.6,7
To date, few therapies are available to reduce the concentration of Lp(a), and it remains unclear whether lowering Lp(a) will translate into improved cardiovascular (CV) outcomes.8-10 PCSK9 (proprotein convertase subtilisin/kexin type 9) inhibitors may offer clinical utility because they have been shown in phase 2 trials to reduce Lp(a) concentration by ≈
25% to 30%.11-13 However, it remains unknown whether the effect of evolocumab on risk of coronary events may be modified by baseline Lp(a) concentrations. Therefore, as a prespecified analysis of the FOURIER trial (Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk), we assessed the relationship between Lp(a) levels, PCSK9 inhibition with evolocumab, and CV risk reduction.14
Insights From the FOURIER Trial
Originally published30 Nov 2018
Abstract
Background:
Lipoprotein(a) [Lp(a)] may play a causal role in atherosclerosis. PCSK9 (proprotein convertase subtilisin/kexin 9) inhibitors have been shown to significantly reduce plasma Lp(a) concentration. However, the relationship between Lp(a) levels, PCSK9 inhibition, and cardiovascular risk reduction remains undefined.
Methods:
Lp(a) was measured in 25 096 patients in the FOURIER trial (Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk), a randomized trial of evolocumab versus placebo in patients with established atherosclerotic cardiovascular disease (median follow-up, 2.2 years). Cox models were used to assess the independent prognostic value of Lp(a) and the efficacy of evolocumab for coronary risk reduction by baseline Lp(a) concentration.
Results:
The median (interquartile range) baseline Lp(a) concentration was 37 (13-165) nmol/L. In the placebo arm, patients with baseline Lp(a) in the highest quartile had a higher risk of coronary heart disease death, myocardial infarction, or urgent revascularization (adjusted hazard ratio quartile 4: quartile 1, 1.22; 95% CI, 1.01-1.48) independent of low-density lipoprotein cholesterol. At 48 weeks, evolocumab significantly reduced Lp(a) by a median (interquartile range) of 26.9% (6.2%-46.7%). The percent change in Lp(a) and low-density lipoprotein cholesterol at 48 weeks in patients taking evolocumab was moderately positively correlated (r=0.37; 95% CI, 0.36-0.39; P<0.001). Evolocumab reduced the risk of coronary heart disease death, myocardial infarction, or urgent revascularization by 23% (hazard ratio, 0.77; 95% CI, 0.67-0.88) in patients with a baseline Lp(a) >median, and by 7% (hazard ratio, 0.93; 95% CI, 0.80-1.08; P interaction=0.07) in those ≤median. Coupled with the higher baseline risk, the absolute risk reductions, and number needed to treat over 3 years were 2.49% and 40 versus 0.95% and 105, respectively.
Conclusions:
Higher levels of Lp(a) are associated with an increased risk of cardiovascular events in patients with established cardiovascular disease irrespective of low-density lipoprotein cholesterol. Evolocumab significantly reduced Lp(a) levels, and patients with higher baseline Lp(a) levels experienced greater absolute reductions in Lp(a) and tended to derive greater coronary benefit from PCSK9 inhibition.
Study Population and Design
The FOURIER trial was a randomized, double-blind, placebo-controlled clinical trial that enrolled 27 564 patients between 40 and 85 years of age who had established atherosclerotic CV disease, determined by a prior myocardial infarction (MI), prior nonhemorrhagic stroke, or symptomatic peripheral artery disease, in addition to predictors of high CV risk. Patients were required to have a fasting LDL-C concentration ≥
70 mg/dL (1.8 mmol/L) or a non-high-density lipoprotein cholesterol concentration ≥
100 mg/dL (2.6 mmol/L) while on a background of optimized lipid-lowering therapy, defined as preferably a high-intensity statin and a minimum dose of 20 mg atorvastatin daily or its equivalent, with or without ezetimibe. There were no entry criteria based on Lp(a) concentration. The study protocol was approved by all relevant ethics committees and all participating subjects provided informed consent. The data, analytical methods, and study materials will not be made universally available to other researchers for purposes of reproducing the results or replicating the procedure. However, we encourage parties interested in collaboration and data sharing to contact the corresponding author directly for further discussions.
Discussion
In patients enrolled in the FOURIER trial, higher baseline Lp(a) concentration was independently associated with an increased risk of major coronary events and evolocumab significantly reduced Lp(a) concentration by ≈
27%. Moreover, patients with higher Lp(a) concentration at baseline experienced greater absolute Lp(a) reductions and tended to derive greater clinical benefit in terms of evolocumab’s ability to reduce the risk of major coronary events. Patients who achieved lower levels of LDL-C and Lp(a) were found to be at lowest risk of subsequent major coronary events.
In FOURIER, we observed a modest but significant association between Lp(a) concentration and the risk of major coronary events. The magnitude of the observed association tended to be greatest for those patients with baseline concentrations above the 90th percentile (≥
230 nmol/L, or ≈
96 mg/dL). Although epidemiological studies have been conflicting,17 the association appeared to be strongest with major coronary events rather than stroke. This is perhaps related to the heterogeneous etiology and different pathobiology of ischemic stroke subtypes. Of interest, it was reported previously that Lp(a) confers CV risk predominantly when LDL-C levels are elevated.3,18,19 However, we found that the relationship between Lp(a) and coronary risk remained similar throughout the entire LDL-C range, thereby suggesting a consistent association of Lp(a) with CV risk in patients with established CV disease independent of concomitant baseline or achieved LDL-C levels. In support, patients who achieved lower levels of both LDL-C and Lp(a) were those who were at lowest risk of subsequent events.
Although Lp(a) is believed to be a risk factor for coronary disease,6,7 the therapeutic targeting of Lp(a) has proven difficult to date. Niacin has been shown to modestly reduce Lp(a) concentration in a dose-dependent manner20; however, its use may be associated with an increased risk of non-CV serious adverse events and there is no evidence that treatment with niacin reduces CV risk on a background of statin.10,21 The effects of statins on plasma Lp(a) concentration have been inconsistent.1,22 Cholesteryl ester transfer protein inhibitors reduce Lp(a), but are not approved for clinical use.8 Mipomersen, an antisense oligonucleotide directed at human apo B100, has been shown to reduce both LDL-C and Lp(a) and is approved in the United States for patients with homozygous familial hypercholesterolemia, but its clinical utility beyond LDL-C lowering remains unknown.23 The microsomal triglyceride transfer protein inhibitor, lomitapide, has also been shown to reduce both Lp(a) and LDL-C, but adverse effects include liver function abnormalities, gastrointestinal side effects, and hepatic fat accumulation.8 Tocilizumab, a humanized monoclonal antibody directed against the interleukin-6 receptor, has been shown to decrease Lp(a) concentrations without concomitant lowering of LDL-C. This effect is believed to be mediated by an interleukin-6-responsive element in the promoter region of the LPA gene leading to a reduction in apo(a) synthesis;8 however, its clinical efficacy for attenuating CV risk remains unknown. Lipoprotein apheresis is used in some countries to reduce Lp(a) concentration, although its CV benefit has only been evaluated in small-scale studies.8
Although their exact mechanism of action remains under study, recent evidence suggests that PCSK9 inhibitors may reduce Lp(a) concentration by both enhancing clearance24 and reducing its production.25,26 In a study of healthy volunteers, evolocumab monotherapy has been demonstrated to lower plasma Lp(a) concentration by decreasing production of Lp(a) particles.26 However, in the presence of a statin, evolocumab may also act to increase Lp(a) catabolism through marked upregulation of the LDL receptor, leading to enhanced Lp(a) holoparticle clearance.26 In Lp(a) uptake studies in human hepatocytes and dermal fibroblasts, secretion of apo(a) appears to increase briskly in the presence of the PCSK9 protein, and this effect is reversed in the presence of the PCSK9 inhibitor alirocumab.25 Discordance between LDL-C and Lp(a) reductions has been reported, which may argue against upregulation of LDL receptor as the sole mechanism for Lp(a) lowering by PCSK9 inhibitors.27 In the current study of >25 000 patients in the FOURIER trial, patients with higher baseline Lp(a) levels also tended to experience greater reductions in major coronary events with evolocumab. This observation is supported by a recent Mendelian randomization study that suggested that genetically mediated lower Lp(a) is associated with a lower risk of major coronary events, although the relationship is not as strong as LDL-C on a mg/dL basis.28 Qualitatively consistent with these findings, we observed that a 34 nmol/L (95% CI, 18.5-97 nmol/L) absolute reduction in Lp(a) may be required to translate into a 20% relative risk reduction in CV events, which approximates the median reduction in Lp(a) that was seen in patients in the top quartile of Lp(a) concentration. These findings suggest that the benefit of Lp(a)-lowering therapies might be largely restricted to patients with elevated levels at baseline, therapies that produce large reductions in Lp(a), or both. These observations may also help to explain why higher baseline Lp(a) concentration was useful for helping to identify individuals with greater clinical efficacy with evolocumab.
It is interesting to note that the percent reduction in Lp(a) with evolocumab tended to diminish with higher baseline levels of Lp(a), possibly because of the reduced clearance of smaller isoforms or other as yet to be defined mechanisms. However, dedicated therapies that markedly reduce Lp(a) concentration by directly targeting the apo(a) protein remain in development and may be able to directly and more robustly test the Lp(a) hypothesis because they can reduce levels by up to ≈
90%.29,30
Practically, the current study suggests that Lp(a) may be useful to help identify patients who derive greater absolute risk reduction from evolocumab and thereby a lower number needed to treat to prevent a major adverse CV event. In patients with an Lp(a) concentration greater than the median (>37 nmol/L), the number needed to treat to prevent one major coronary event over 3 years was 40 in comparison with 105 for patients with a baseline Lp(a) concentration less than the median. Prior analyses from the same patient population have demonstrated that high-sensitivity C-reactive protein may also predict CV risk and identify patients with a larger absolute benefit from evolocumab.31 Although high-sensitivity C-reactive protein is a well-established risk marker, it does not appear to be an independent risk factor,32 and evolocumab does not lower high-sensitivity C-reactive protein concentration. In contrast, Lp(a) appears to be a causal factor in coronary heart disease,6,7 and its concentration is lowered by evolocumab.
Although the current analysis was prespecified, all cut points should be viewed as exploratory. Because the current analysis was observational in nature, observed associations should not imply causality. Patients in FOURIER were not selected based on an elevated Lp(a) concentration; therefore, there was no enrichment on this basis. Nonetheless, ≈
33.1% of patients had a baseline concentration >120 nmol/L (or ≈
50 mg/dL) which is believed to be the 80th percentile in a general patient population.1 Although there remains disagreement about the optimal method to assess Lp(a) concentration, the current study used an assay system that was isoform independent, as supported by previous consensus panels.33 In addition, the current study population predominantly enrolled male and white participants; therefore, it will be of interest to further examine these findings in additional cohorts given the sex and racial differences that exist in Lp(a) concentrations. The study was not designed to achieve statistical significance within subgroups, and tests for interaction were relatively underpowered to achieve statistical significance; therefore, numbers needed to treat within subgroups should be considered exploratory. In addition, analyses that examined achieved values of Lp(a) and LDL-C are at risk of residual confounding and therefore do not directly imply causality.
In summary, the current findings suggest that plasma Lp(a) concentration is associated with the risk of CV events in patients with stable atherosclerotic disease regardless of LDL-C concentration. Furthermore, Lp(a) may be useful for identifying individuals with a greater absolute benefit from evolocumab and lend support to the study of additional therapies that can lead to marked reductions in Lp(a) concentration.
|
|
| |
| |
|
|
|
