Lipoprotein(a)

Abstract

HomeCirculationVol. 129, No. 6Lipoprotein(a) Free AccessEditorialPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessEditorialPDF/EPUBLipoprotein(a)There’s Life in the Old Dog Yet Florian Kronenberg, MD Florian KronenbergFlorian Kronenberg From the Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Innsbruck Medical University, Innsbruck, Austria. Search for more papers by this author Originally published11 Feb 2014https://doi.org/10.1161/CIRCULATIONAHA.113.007256Circulation. 2014;129:619–621When the first prospective studies on lipoprotein(a) [Lp(a)] were published in the early 1990s, the career of Lp(a) as a risk factor for cardiovascular disease (CVD) was almost stopped right at its beginning. A few of these prospective studies were negative1 and resulted in a major discussion whether Lp(a) had its best times already behind it. However, numerous later studies and meta-analysis of the data strengthened the association between high Lp(a) concentrations and CVD.2 Finally, genetic research on Lp(a) became the basis for what revolutionized epidemiological research during the last decade in general. This is based on the observation that Lp(a) concentrations are largely determined by a copy number variation in the LPA gene, which results in an interindividually highly variable and large number of so-called kringle-IV repeats of the apolipoprotein(a) [apo(a)] protein. This polymorphism explains between 30% and 70% of the Lp(a) concentrations, which makes Lp(a) to the lipoprotein with the strongest genetic determination.3 Individuals with a low number of kringle-IV repeats (small isoform carriers) have on average markedly higher Lp(a) concentrations.4 Because the number of kringle-IV repeats are subject of inheritance, it is already determined at the time of conception whether one inherits apo(a) isoforms associated with high or low Lp(a) concentrations. Consequently, those who inherit small isoforms are exposed life-long to high Lp(a) concentrations and are expected to have a high CVD risk. This was indeed the case when the first studies using this concept were published in the early 1990s. They demonstrated that individuals carrying small isoforms have indeed a high CVD risk.5 Studies using SNPs that tag a subfraction of the small apo(a) isoforms took the same line.6 This concept using genetic variants that determine an important fraction of the risk factor under suspicion to demonstrate an association with the outcome of interest was later called Mendelian Randomization.7 It was a major step not only for Lp(a) research but for epidemiology in general because usually epidemiological studies can hardly prove causality. However, the Mendelian randomization concept can provide a very strong support of causality. With the help of this concept several risk factors have been reevaluated and some of them changed their status from a risk factor to a risk marker that is not necessarily involved in the causal chain of the disease. For Lp(a) it was even a starting point because all of a sudden it became an interesting object for intervention.Article see p 635In the past it was discussed controversially whether high Lp(a) concentrations are only a risk factor when also other well-known CVD risk factors are present in an individual. In this issue of Circulation, Khera et al8 investigated whether Lp(a) is a significant determinant of residual CVD risk in the JUPITER trial that recruited individuals without a history of CVD at baseline, with low-density lipoprotein (LDL)-cholesterol <130 mg/dL and a high-sensitivity C-reactive protein (hsCRP) of 2.0 mg/L or more (representing a low-grade inflammation). Patients were randomized to statin therapy with rosuvastatin or placebo and the intervention reduced LDL-cholesterol by 50% and hsCRP by 37%. The study demonstrated that Lp(a) was a significant determinant of residual risk not only in the entire group. Lp(a) showed in both the statin and the placebo-treated patient groups very similar risk estimates. These associations were the same when Lp(a) measured at baseline or 1 year after start of treatment was used for risk calculations. Furthermore, the observed associations were similar in various risk factor subgroups.8 Analogous results were reported in the recently published Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglyceride and Impact on Global Health Outcomes (AIM-HIGH) Trial that observed Lp(a) to be associated with an increased CVD risk in both treatment groups (simvastatin plus placebo or simvastatin plus extended-release niacin).9 These 2 studies underscore that Lp(a) acts as a CVD risk factor that is independent from LDL-cholesterol and also other CVD risk factors. This argues for interventional targeting of high Lp(a) concentrations. However, the findings of a similar association of Lp(a) with CVD events independently of the LDL-cholesterol values are in contrast to a very recently published meta-analysis of secondary prevention studies that observed an effect modification by LDL-cholesterol: the association between Lp(a) concentrations and CVD events was only observed in studies in which the average LDL-cholesterol was ≥130 mg/dL at baseline and not in those below this value.10 A pronounced heterogeneity in the study populations, inclusion criteria, therapeutic regimens, and duration of observation periods might have contributed to the differences in results between the hitherto published studies.It is at the first glance surprising that Lp(a) should act independently from LDL-cholesterol on CVD risk because Lp(a) is structurally very similar to LDL. In fact it is an LDL particle that contains an additional apolipoprotein(a) protein. Therefore, the LDL-receptor was 1 of the first candidates to be studied in the context of Lp(a) catabolism. This idea was supported by the observations that Lp(a) levels are markedly elevated in patients with familial hypercholesterolemia that are caused by LDL-receptor mutations. Moreover, Lp(a) concentrations are decreased in transgenic mice overexpressing the LDL-receptor.11 On the other hand, statins, which cause an upregulation of the LDL-receptor markedly decrease levels of LDL-cholesterol but not Lp(a). Even the opposite might be the case as already suggested a long time ago by Kostner et al12 and very recently by O’Donoghue et al10: both studies observed an increase in Lp(a) concentrations after successful lowering of LDL-cholesterol by statins. This is in line with the results of the JUPITER Study in the present issue of Circulation8: patients under rosuvastatin treatment showed even a statistically significant positive shift in the overall Lp(a) distribution that was not observed in the placebo group.The authors made a further interesting observation: when they stratified patients by the median of hsCRP values (4 mg/L) they observed a significant association between Lp(a) levels at baseline and the incident primary end point in the group with hsCRP below the median (hazard ratio, 1.32; 95% confidence interval, 1.10–1.59; P=0.003), which was not the case in the group with hsCRP above the median (hazard ratio, 1.05; 95% confidence interval, 0.88–1.26; P=0.58). This might be explained by a secondary elevation of Lp(a) in the state of a low-grade inflammation. Because this Lp(a) elevation was not present life-long, the currently measured Lp(a) concentration might have lost some of its predictive ability when hsCRP levels are above 4 mg/L. Similar observations were made in patients with end-stage renal disease, a condition with an impressive secondary increase in Lp(a) concentrations. In these patients the ability to predict an event is less pronounced for Lp(a) concentrations than for the presence of a small apo(a) isoform. The apo(a) isoforms are not influenced by the disease state and reflect the life-long burden associated with genetically determined elevations of Lp(a) concentrations. Therefore, the apo(a) isoform size might be more predictive for CVD events in such conditions compared with Lp(a) concentrations.It is a misfortune that despite 50 years of research on this atherogenic particle, we still observe a large number of blank spots on the map of Lp(a). We still do not know the physiological function and which machinery of ligands, receptors, cofactors, or enzymes is involved in the metabolism of Lp(a). This makes it hard to find a promising strategy for an interventional lowering of Lp(a). Despite the opposite effect of statins on Lp(a) levels, the LDL-receptor machinery with its relatives gets never out of the focus of interest. Other suspects of receptors are the very-low-density lipoprotein–receptor/megalin13 or scavenger receptor B-I.14 Even a hitchhiking-like process was proposed whereby Lp(a) attached to LDL is removed by the LDL-receptor pathway.15 One of the most intriguing observations is proprotein convertase subtilisin kexin type 9 (PSCK9) that is synthesized and secreted by hepatocytes. PCSK9 binds to the LDL-receptor thereby making it to a target of lysosomal degradation. Inhibitors of PCSK9 decrease Lp(a) levels up to 32%16 and belong to the very promising drugs for future lipid-lowering therapies in case it will be demonstrated that they also lower clinical endpoints.Therapies that have a significant Lp(a)-lowering effect result in many cases also in substantial changes of other risk factors. That makes it hard to disentangle who is the real culprit. A typical example is niacin, which was for a long time considered to be kind of a magic bullet because it influences besides LDL- and high-density lipoprotein cholesterol also triglycerides and Lp(a). Nevertheless it turned out recently to be not effective in lowering CVD events despite these favorable lipoprotein changes.9 It will have to be demonstrated how effective PCSK9-inhibitors will be that have been shown to lower besides LDL-cholesterol also Lp(a) concentrations.16 Even if they are effective in terms of CVD events, it will not be clear whether the concomitant lowering of Lp(a) contributes to the changes in the CVD risk or not. The same holds true for other strategies which unspecifically target apoB-containing lipoproteins. At least from a scientific standpoint, one wishes a therapy that specifically targets high Lp(a) concentrations (by eg, specific antisense oligonucleotides directed against apo(a)). This will allow to follow the ideas of Koch’s postulate that the removal of high concentrations of the atherogenic Lp(a) will result in a decrease of CVD endpoints. It has to be seen whether this kind of interventional therapy is a luxury that results only in a cosmetic make-up of laboratory values or whether it indeed decreases the CVD risk in those patients who currently are exposed to an elevated risk due to dramatically increased Lp(a) levels. Support for the idea of an Lp(a)-lowering therapy comes from 3 interesting studies: 2 studies were performed in patients with very high concentrations of Lp(a) and already optimal lipid-lowering as well as other risk factor therapies. Both studies demonstrated a dramatic decrease in major coronary events after compared to the time before start of lipid apheresis therapy.17,18 A recent small study investigated the effect of a specific Lp(a) apheresis in patients with CVD verified by angiography, an Lp(a) level above 50 mg/dL and LDL-cholesterol ≤2.5 mmol/L.19 They used immunabsorption columns with sheep polyclonal monospecific antibodies directed against human apo(a), resulting in a 73% decrease of Lp(a) without other major changes in lipid profile. A quantitative coronary angiography revealed a significant improvement in the median percent diameter stenosis and the minimal lumen diameter of the coronary arteries during a 18 months treatment period in the Lp(a) apheresis group when compared with a parallel atorvastatin control group. These findings are convincing enough to prolong the search for Lp(a)-lowering therapies.In summary, the study by Khera at el8 has demonstrated the Lp(a) concentrations can be used for risk prediction even after risk reduction by successful statin therapy. This and other studies underscore Lp(a) concentrations as independent CVD risk factor and argue for interventional targeting of high Lp(a) concentrations. That means there’s life in the old dog yet, and he will be even rejuvenated as soon as a therapeutical lowering of Lp(a) demonstrates a concomitant decrease in CVD events. However, because most therapeutic strategies currently result in a concerted action on various risk factors (eg, Lp(a) and LDL- or high-density lipoprotein cholesterol) it will be hard if not impossible to disentangle whether a future therapeutic success is transmitted by a lowering of Lp(a) or by the other factors influenced by these therapies.DisclosuresNone.FootnotesThe opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.Correspondence to Florian Kronenberg, MD, Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Innsbruck Medical University, Schöpfstr 41, A-6020 Innsbruck, Austria. E-mail [email protected]