Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: a European Atherosclerosis Society consensus statement

Abstract

This 2022 European Atherosclerosis Society lipoprotein(a) [Lp(a)] consensus statement updates evidence for the role of Lp(a) in atherosclerotic cardiovascular disease (ASCVD) and aortic valve stenosis, provides clinical guidance for testing and treating elevated Lp(a) levels, and considers its inclusion in global risk estimation. Epidemiologic and genetic studies involving hundreds of thousands of individuals strongly support a causal and continuous association between Lp(a) concentration and cardiovascular outcomes in different ethnicities; elevated Lp(a) is a risk factor even at very low levels of low-density lipoprotein cholesterol. High Lp(a) is associated with both microcalcification and macrocalcification of the aortic valve. Current findings do not support Lp(a) as a risk factor for venous thrombotic events and impaired fibrinolysis. Very low Lp(a) levels may associate with increased risk of diabetes mellitus meriting further study. Lp(a) has pro-inflammatory and pro-atherosclerotic properties, which may partly relate to the oxidized phospholipids carried by Lp(a). This panel recommends testing Lp(a) concentration at least once in adults; cascade testing has potential value in familial hypercholesterolaemia, or with family or personal history of (very) high Lp(a) or premature ASCVD. Without specific Lp(a)-lowering therapies, early intensive risk factor management is recommended, targeted according to global cardiovascular risk and Lp(a) level. Lipoprotein apheresis is an option for very high Lp(a) with progressive cardiovascular disease despite optimal management of risk factors. In conclusion, this statement reinforces evidence for Lp(a) as a causal risk factor for cardiovascular outcomes. Trials of specific Lp(a)-lowering treatments are critical to confirm clinical benefit for cardiovascular disease and aortic valve stenosis.

Keywords: Aortic stenosis; Cardiovascular risk; Clinical guidance; Consensus; Lipoprotein(a); Model of care; Testing; Treatment.

Graphical abstract

Key points from the 2022 Lp(a) consensus statement. Current evidence demonstrates a causal continuous association in different ethnicities between Lp(a) concentration and cardiovascular outcomes including aortic valve stenosis, but not for venous thrombotic events. A meta-analysis of prospective studies shows that very low Lp(a) levels are associated with increased risk of diabetes mellitus. For clinical practice, Lp(a) should be measured at least once in adults and results interpreted in the context of a patient's absolute global cardiovascular risk, with recommendations on intensified early risk factor management by lifestyle modification. The statement also reviews currently available and future possibilities to specifically lower Lp(a).

Figure 1

Structure and genetic variability of the LPA gene. The upper panel shows the topology of apolipoprotein(a) and the association of the Kringle-IV repeat polymorphism with lipoprotein(a) [Lp(a)] concentration, which explains 30%–70% of variation depending on ethnicity. The lower panel shows the structure of the LPA gene and the known single-nucleotide polymorphisms within the gene that have marked effects on Lp(a) concentration. The exons are numbered according to the domain that they encode (L. leader sequence, 1–10: KIV-1 to KIV-10, V: KV domain, P: protease domain, 5′: 5′UTR, 3′: 3′ UTR). Single-nucleotide polymorphisms associated with increased Lp(a) concentration are shown above the gene structure, and those associated with decreased Lp(a) concentrations (both causally or by association only) are shown below. Single-nucleotide polymorphisms that cause null alleles are underlined; however, Lp(a)-lowering single-nucleotide polymorphisms may cause null alleles if present on an allele already associated with low Lp(a) production. Single-nucleotide polymorphisms in the Kringle-IV Type-2 region are named according to their publication; they cannot be assigned a single rs-identifier as their location is not unique. Figure provided and adapted by Prof. Florian Kronenberg and Prof. Stefan Coassin based on reference.

Figure 2

Distribution of lipoprotein(a) [Lp(a)] concentration and association with risk for major cardiovascular events. Data from the UK Biobank show the typical distribution of Lp(a) concentrations in White (Panel A) and Black people (Panel B) and the linear relationship of Lp(a) concentration with risk for major cardiovascular events in White (Panel C), and Black people (Panel D). Panels A and B give the percentage of the population with an Lp(a) of 170, 190, 215, and 240 nmol/L or higher, respectively. Panels C and D show the smoothed adjusted hazard ratio (HR) and 95% confidence interval (95% CI) for lifetime risk for major cardiovascular events for a given Lp(a) concentration relative to the median Lp(a) in the population (19.7 nmol/L). These data were estimated using a Cox proportional hazards regression model adjusted for age at enrolment, sex, and the first 10 principle components of ancestry and modelled using cubic natural splines. Confidence intervals are wider in Black people due to the smaller sample size. Panel E shows the lifetime risk of major cardiovascular events with increasing Lp(a) concentrations among men of European ancestry in the UK Biobank (results were similar for women but with lower absolute event rates). Participants were partitioned into categories with increasingly greater median Lp(a) plasma concentrations; and the cumulative major cardiovascular event rates were plotted for each group up to age 80 years. Panel A and B are provided by Prof. Florian Kronenberg and Silvia Di Maio; Panel C-E are provided by Prof. Brian Ference and Prof. Alberico L. Catapano. For detailed methodological description, see Supplementary material online.

Figure 3

Risk of clinical outcomes with Lp(a) concentration. Absolute and relative risks of aortic valve stenosis, ischaemic stroke, myocardial infarction and heart failure as a function of increasing plasma Lp(a) concentration in the general population. Top panel shows the absolute risk per 10000 person-years, and the lower panel shows hazard ratios as solid red line with 95% confidence intervals as dotted black lines; when the lower 95% confidence interval no longer overlap the hazard ratios reference value of 1.0 for the median Lp(a) concentration, risk is significantly elevated. Based on data from 70 286 White individuals in the Copenhagen General Population Study with a median 7.4 years of follow-up. Data provided by Prof. Børge G. Nordestgaard and Dr. Anne Langsted.

Figure 4

Effect of Lp(a) concentration and LPA score copies on risk for major cardiovascular events and venous thrombotic events. Data are from 440 368 UK Biobank participants of European ancestry. The LPA score is defined as the number of minor alleles of LPA variants rs10455872 or rs3798220 inherited by each participant, with the reference group defined as participants with no copies of either minor allele. The measured median Lp(a) concentration is provided for each group. Panel A shows the effect of higher Lp(a) among participants who inherit one minor allele [median Lp(a) 146.3 nmol/L] or two minor alleles [median Lp(a) 261.9 nmol/L] on the risk of major cardiovascular events [defined as the composite of the first occurrence of fatal or non-fatal myocardial infarction, fatal or non-fatal ischaemic stroke, or coronary revascularization (percutaneous coronary intervention or coronary artery bypass graft surgery)] vs. the reference group [median Lp(a) 13.6 nmol/L]. Panel B shows the effect on the risk of venous thromboembolic events (VTE, defined as deep venous thrombosis or pulmonary embolism). Higher Lp(a) levels were strongly associated with increased risk of atherosclerotic cardiovascular disease but not VTE. The solid boxes represent point estimates for effect and the lines 95% confidence intervals. Data provided by Prof. Brian Ference and Prof. Alberico L. Catapano. For detailed methodological description, see Supplementary material online.

Figure 5

Association of Lp(a) concentration with diabetes mellitus. Random effects meta-analysis of seven prospective studies and one case–control study examining the association of lipoprotein(a) with risk of diabetes. Summary relative risks comparing bottom vs. top quintile were adjusted for clinical risk factors, with the size of the squares proportional to the number of cases in each study. Results were similar when excluding the Kamstrup et al. study (for potential overlap with the Tolbus et al. and Langsted et al. studies; Supplementary material online, Figure S1A) or when excluding the case–control study by Gudbjartsson et al. (Supplementary material online, Figure S1B). The summary relative risk P-values were all <0.001. (See also Supplementary material online). Data provided by Prof. Samia Mora, Dr. Olga Demler and Dr. Yanyan Liu.

Figure 6

Effect of increasing Lp(a) levels and estimated baseline absolute risk for major cardiovascular events. Panel A shows the estimated remaining lifetime risk of a major cardiovascular event [defined as the composite of the first occurrence of fatal or non-fatal myocardial infarction, fatal or non-fatal ischaemic stroke, or coronary revascularization (percutaneous coronary intervention or coronary artery bypass graft surgery)] among 415 274 participants of European ancestry in the UK Biobank for whom measured Lp(a) values were available. Participants are divided into categories of baseline estimated lifetime risk (5%, 10%, 15%, 20%, and 25%) calculated using the Joint British Societies (JBS3) Lifetime Risk Estimating algorithm (derived from a similar UK population). Within each baseline risk category, participants are then further divided into categories defined by baseline measured Lp(a) concentration. The incremental increase in risk caused by higher Lp(a) concentrations from 30 to 150 mg/dL (70 from 350 nmol/L) was estimated by adding Lp(a) as an independent exposure to the JBS3 risk estimating algorithm. The numbers at the upper end of each bar represent the increment of increased absolute risk above the estimated baseline risk caused by Lp(a). This information is presented in tabular form in Supplementary material online, Table S2. For example, for a person with a baseline risk of 5%, an Lp(a) concentration of 30 mg/dL increases the absolute remaining lifetime risk of a major cardiovascular event by 1.1% to 6.1% (vs. a person with an Lp(a) of 7 mg/dL). By contrast, for a person with a baseline risk of 25%, an Lp(a) concentration of 30 mg/dL increases the absolute risk of a major cardiovascular event by 5.5% to 30.5% (vs. a person with an Lp(a) of 7 mg/dL). For individuals with an Lp(a) concentration of 75 mg/dL, the corresponding absolute increases in risk are 3.3% and 16.3%, respectively. The figure illustrates that failure to consider a person’s Lp(a) level can lead to a substantial underestimate of their absolute risk of a major cardiovascular event. Data provided by Prof. Brian Ference and Prof. Alberico L. Catapano. Further details are provided in the Supplementary material online. Panel B provides the intervention strategies as a function of total cardiovascular risk and untreated Lp(a) concentration. In the absence of specific Lp(a)-lowering therapy, these focus on management of other cardiovascular risk factors. For detailed methodological description, see Supplementary material online, Table S3.

Figure 7

Low-density lipoprotein cholesterol reduction needed to reduce global cardiovascular risk to a similar extent as the risk attributable to high Lp(a). Panel A shows the cumulative absolute lifetime risk of a major cardiovascular event (defined as the composite of the first occurrence of fatal or non-fatal myocardial infarction, fatal or non-fatal ischaemic stroke, or coronary revascularization [percutaneous coronary intervention or coronary artery bypass graft surgery]) among 440 368 UK Biobank participants of European descent. For each sex, these were divided into three groups: a reference group with population average Lp(a) [16–17 nmol/L] and low-density lipoprotein cholesterol levels [3.5–3.6 mmol/L] who had no copies of either the rs10455872 or rs3798220 Lp(a)-increasing alleles; a group with higher lifetime exposure to Lp(a) [136–138 nmol/L) due to inheriting one copy of either the rs10455872 or rs3798220 Lp(a)-increasing alleles, but with population average low-density lipoprotein cholesterol levels [3.5–3.6 mmol/L]; and a group with BOTH higher lifetime exposure to Lp(a) [136–138 nmol/L) due to inheriting one copy of either the rs10455872 or rs3798220 Lp(a)-increasing alleles AND lifetime exposure to 0.5 mmol/L lower low-density lipoprotein cholesterol [3.0-3.1 mmol/L) due to inheriting a combination of low-density lipoprotein cholesterol lowering genetic variants. The Figure demonstrates that the increased risk of major cardiovascular event caused by lifetime exposure to approximately 120 nmol/L higher Lp(a) can be mitigated at all ages by a lifetime exposure to approximately 0.5 mmol/L lower LDL-C. This finding illustrates the potential to estimate how much intensification of risk factor modification, in this case low-density lipoprotein cholesterol lowering, is needed to mitigate the increased risk of major cardiovascular events caused by a person’s Lp(a) level. Panel B provides a quantitative estimate of the intensification of low-density lipoprotein cholesterol lowering needed to mitigate the increased risk of major cardiovascular event caused by increasingly higher Lp(a) levels, and the age at which low-density lipoprotein cholesterol lowering is initiated. Because the proportional reduction in risk produced by lowering low-density lipoprotein cholesterol decreases with decreasing duration of exposure, greater intensification of low-density lipoprotein cholesterol lowering is needed to mitigate a given Lp(a) level if low-density lipoprotein cholesterol lowering is started at a later age. For example, a person with an elevated Lp(a) level of 220 nmol/L has a 1.87-fold increased risk of major cardiovascular event as compared to a person with an Lp(a) level of 20 nmol/L (assuming all other risk factors are equal). This increased risk of major cardiovascular event can be mitigated by lowering low-density lipoprotein cholesterol by 0.8 mmol/L if started at age 30 years, but would require more intense low-density lipoprotein cholesterol lowering by 1.5 mmol/L if started at age 60 years. The data in the table of panel B elaborates on current clinical practice guidelines that recommend more intense risk factor modification among persons with elevated Lp(a) levels by providing specific quantitative guidance for how much low-density lipoprotein cholesterol lowering should be intensified to mitigate the increased risk of major cardiovascular event caused by increasingly higher Lp(a) levels. An easy-to-use online Lp(a) risk and benefit algorithm can provide convenient and specific guidance on how much intensification of low-density lipoprotein cholesterol lowering is needed to mitigate the risk caused by a person’s Lp(a) level depending on the age at which low-density lipoprotein cholesterol lowering is initiated. However, where the main part of the risk is substantial and mainly attributable to Lp(a), a lowering of traditional risk factors such as low-density lipoprotein cholesterol will be insufficient to mitigate this increased risk. In these cases specific Lp(a)-lowering therapies are urgently required. This information should motivate testing of Lp(a) and inform the clinical use of measured Lp(a) levels. The online Lp(a) risk and benefit algorithm is available at the European Atherosclerosis Society website (www.eas-society.org/LPA_risk_and_benefit_algorithm). Further details are provided in the Supplementary material online. For detailed methodological description, see Supplementary material online. Data provided by Prof. Brian Ference and Prof. Alberico L. Catapano.