SEARCH

SEARCH BY CITATION

Keywords:

  • apolipoprotein(a);
  • coronary heart disease;
  • isoform size;
  • lipoprotein(a)

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Objectives

Observational and genetic studies have shown that lipoprotein(a) [Lp(a)] levels and apolipoprotein(a) [apo(a)] isoform size are both associated with coronary heart disease (CHD) risk, but the relative independence of these risk factors remains unclear. Clarification of this uncertainty is relevant to the potential of future Lp(a)-lowering therapies for the prevention of CHD.

Methods

Plasma Lp(a) levels and apo(a) isoform size, estimated by the number of kringle IV (KIV) repeats, were measured in 995 patients with CHD and 998 control subjects. The associations between CHD risk and fifths of Lp(a) levels were assessed before and after adjustment for KIV repeats and, conversely, the associations between CHD risk and fifths of KIV repeats were assessed before and after adjustment for Lp(a) levels.

Results

Individuals in the top fifth of Lp(a) levels had more than a twofold higher risk of CHD compared with those in the bottom fifth, and this association was materially unaltered after adjustment for KIV repeats [odds ratio (OR) 2.05, 95% confidence interval (CI) 1.38–3.04, < 0.001]. Furthermore, almost all of the excess risk was restricted to the two-fifths of the population with the highest Lp(a) levels. Individuals in the bottom fifth of KIV repeats had about a twofold higher risk of CHD compared with those in the top fifth, but this association was no longer significant after adjustment for Lp(a) levels (OR 1.13, 95% CI 0.77–1.66, = 0.94).

Conclusions

The effect of KIV repeats on CHD risk is mediated through their impact on Lp(a) levels, suggesting that absolute levels of Lp(a), rather than apo(a) isoform size, are the main determinant of CHD risk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Recent genetic studies demonstrating strong support for lipoprotein(a) [Lp(a)] as a causal risk factor for coronary heart disease (CHD) have prompted a resurgence of interest in Lp(a) as a therapeutic target for prevention of CHD [1-4]. Prediction of the potential impact of Lp(a)-modifying treatments for CHD prevention requires reliable evidence of the relationships between Lp(a) and apolipoprotein(a) [apo(a)] isoform size and their independent relevance for CHD risk. In particular, it is important to determine whether it is high Lp(a) levels or the number of kringle IV (KIV) repeats that account for the increased risk of CHD associated with increased Lp(a) levels.

Lp(a) comprises a single low-density lipoprotein cholesterol (LDL-C) particle linked to a highly polymorphic apo(a) protein [5]. Plasma levels of Lp(a) vary by almost 1000-fold between individuals, with about 20–30% of the population having extremely increased Lp(a) levels [4]. Blood levels of Lp(a) are highly heritable and are chiefly determined by copy number variation at the LPA locus on chromosome 6 [6]. This copy number variation in the KIV type 2 protein domain, encoding apo(a) isoforms of varying size, results in a varying number of copies (up to about 40) of identical KIV type 2 repeats. Apo(a) isoform size is difficult to determine, but measurement of the number of KIV repeats by gel electrophoresis can provide informative estimates for each expressed allele. Lp(a) levels are correlated with the number of KIV repeats, with higher levels of Lp(a) being associated with smaller isoform size as represented by fewer KIV repeats, and vary considerably for a given apo(a) isoform size [7].

Many studies have demonstrated that increased plasma levels of Lp(a) and fewer KIV repeats are associated with increased risk of CHD [4]. However, variation in the methods adopted for measuring Lp(a) levels and the effects of Lp(a) levels on CHD risk observed in genetic, prospective and case–control studies has resulted in some uncertainty about the importance of Lp(a) levels, as well as the relevance of apo(a) isoform size, for prediction of CHD. For example, studies in individuals with genetically increased Lp(a) levels have suggested that Lp(a) is a strong risk factor for CHD [1-3]. By contrast, a meta-analysis of prospective studies has suggested that differences in plasma levels of Lp(a) have only a modest effect on CHD risk [8]. Further uncertainty about the relative impact of Lp(a) and apo(a) isoform size has been raised by a meta-analysis that suggested a higher risk of CHD in those with smaller versus larger apo(a) isoform size (corresponding approximately to 22 or fewer KIV repeats versus more than 22 repeats), but was unable to assess the extent to which this relationship depended on Lp(a) levels [9]. Few studies, most involving only small numbers of cases, have explored the relationships between CHD and both apo(a) isoform size and Lp(a) levels to determine whether the associations with smaller apo(a) isoform size are independent of plasma Lp(a) levels and vice versa [2, 10-14]. The aims of this study were to compare the pattern and strength of the associations between CHD risk and both plasma levels of Lp(a) and apo(a) isoform size (measured by KIV repeats) in about 1000 patients with CHD (cases) and 1000 individuals with no history of CHD (controls) from the PROCARDIS study and to assess the independent relevance of Lp(a) levels and apo(a) isoform size for CHD risk.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Study population

CHD cases were recruited from four European countries (UK, Italy, Sweden and Germany) [1]. Cases had been previously diagnosed with CHD and also had a sibling diagnosed with CHD, before the age of 66 years. Ascertainment criteria for PROCARDIS probands were myocardial infarction or symptomatic acute coronary syndrome. Controls, recruited from the same population as the cases, had no personal or sibling history of CHD before 66 years of age. The protocol was approved by the ethics committee at each participating centre, and all participants provided written informed consent.

Laboratory methods

The number of KIV repeats was estimated by isoelectric focusing followed by immunoblotting [15, 16]. Sodium dodecyl sulphate-agarose gel electrophoresis was performed over 12 h to fractionate the reduced plasma proteins according to their size before immunoblotting with an apo(a)-specific antibody. Apo(a) size was determined relative to a human apo(a) isoform standards (Immuno, Vienna, Austria) and a serum pool with predetermined apo(a) isoforms, and band size was evaluated using aida software (Straubenhardt, Germany). A step-by-step account of the method, reagents, buffer concentrations and standardization procedures used to determine the number of KIV repeats is provided elsewhere [15]. A typical immunoblot is shown in Fig. S1. This approach [in contrast to other approaches such as real-time polymerase chain reaction (qPCR)] provides information for each expressed allele separately, although this technique is limited in its ability to detect nonexpressed or poorly expressed alleles. For each individual, the measurements were classified as a null, single-band or double-band result. A null result was defined as two nondetectable KIV repeat alleles. A single-band result was defined as two non-null alleles with the same number of KIV repeats and indicates either a heterozygote with a single null allele or a homozygote with two alleles coding for the same apo(a) isoform size. A double-band result was defined as two non-null alleles with different numbers of KIV repeats and indicates a heterozygote with two alleles coding for different apo(a) isoform sizes. The true number of KIV repeats is believed to be within about two repeats of the measured value.

Lp(a) level was measured using a latex-enhanced immunoturbidimetric assay (Randox Laboratories, Crumlin, Co. Antrim, UK) based on a highly specific polyclonal rabbit anti-apo(a) antibody on an ADVIA 1800 autoanalyser (Siemens, Erlangen, Germany) and updated following International Federation of Clinical Chemistry recommendations [17, 18]. The Lp(a) assay has been validated against the enzyme-linked immunosorbent assay reference method and uses five calibrators derived from World Health Organisation reference material SRM2B. Laboratory analyses of lipid fractions, including high-density lipoprotein cholesterol, directly measured LDL-C, triglycerides, apolipoprotein A1 and apolipoprotein B, were performed using standard methods on automated analysers. All assays were performed in EDTA plasma samples that had been stored (for up to 10 years) in liquid nitrogen or in −80 °C freezers prior to analysis.

Statistical methods

Apo(a) isoform size, as measured by the number of KIV repeats, was measured successfully in 995 CHD cases and 998 controls matched for age, sex and country of recruitment for whom plasma levels of Lp(a) were available. After exclusion of individuals with null results (5%), data were available for a total of 947 cases and 946 controls. Correlations were assessed using Spearman correlation coefficients (ρ). KIV repeats and Lp(a) levels were classified into fifths based on the distribution of values in controls. Lp(a) levels were positively skewed and were natural-log-transformed for further analyses. Logistic regression models were used to assess the associations between CHD risk and fifths of Lp(a) before and after adjustment for conventional risk factors and fifths of KIV repeats. Similarly, associations between CHD and fifths of KIV repeats were examined before and after adjustment for conventional risk factors and fifths of Lp(a). Empirical estimates of variance were used in all models to account for familial clustering. Chi-squared statistics (based on the score test) were used to assess model fit. All odds ratios (ORs) are reported relative to the group with the lowest CHD risk unless otherwise stated. The method of floating absolute risks is used where indicated to enable comparisons between different groups. Analyses were performed in sas version 9 (Cary, NC, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

Baseline characteristics

Selected characteristics of the 998 controls and 995 CHD cases with measured KIV repeats and Lp(a) levels are shown in Table 1. As anticipated, the proportions of individuals who were current smokers had hypertension or diabetes or were statin users were higher amongst CHD cases than controls. In addition, the geometric mean Lp(a) levels were higher amongst CHD cases than controls [13.57 mg dL−1 (95% CI 12.48–14.46) vs. 10.56 mg dL−1 (95% CI 9.89–11.26)]. There was no difference in the proportion of cases and controls with nondetectable apo(a), although there was a slightly higher proportion of cases with a double-band result for KIV repeats. Furthermore, the CHD cases had slightly fewer KIV repeats for both the shorter– (27 vs. 25) and longer– (30 vs. 29)isoform alleles, compared with the controls. The 5% of individuals who were excluded due to nondetectable apo(a) had geometric mean Lp(a) levels that were much lower than those in the individuals included in the present results (1.11 mg dL−1 vs. 13.57 mg dL−1).

Table 1. Baseline characteristics in controls and CHD cases
 Controls (= 998)CHD cases (= 995)
  1. CHD, coronary heart disease; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; Lp(a), lipoprotein(a); KIV, kringle IV.

  2. Values are mean (SD) for continuous variables and n (%) for categorical variables, unless otherwise stated. Data on current smoking were only available for 994 of the 998 controls and for 978 of the 995 CHD cases.

Characteristics
Age (years)60.9 (10.0)63.0 (6.9)
Male691 (69.2%)676 (67.9%)
Diabetes15 (1.5%)161 (16.2%)
Hypertension236 (23.7%)514 (51.7%)
Body mass index (kg m−2)26.7 (4.0)28.6 (4.8)
Current smoker133 (13.3%)305 (30.7%)
Statin use0 (0.0%)659 (66.2%)
Biochemistry
LDL-C (mmol L−1)3.33 (0.79)2.84 (0.82)
HDL-C (mmol L−1)1.37 (0.37)1.15 (0.33)
Apolipoprotein B (mg dL−1)103.88 (21.91)94.41 (22.64)
Apolipoprotein A1 (mg dL−1)170.86 (26.69)162.76 (26.99)
Log (triglycerides mmol L−1)0.37 (0.54)0.62 (0.55)
Fibrinogen (g L−1)3.82 (0.89)4.35 (0.97)
Log (C-reactive protein mg L−1)0.31 (1.08)0.75 (1.13)
Lp(a) levels
Geometric mean (95% CI)10.56 (9.89–11.26)13.57 (12.48–14.46)
Range (mg dL−1)0.01–129.480.01–190.54
KIV repeats
Null52 (5.2%)48 (4.8%)
Single band620 (62.1%)569 (57.2%)
Double band326 (32.7%)378 (38.0%)
Lowest repeat (excluding null results)
Median (range)27 (12–40)25 (12–42)
Highest repeat (excluding null results)
Median (range)30 (13–43)29 (12–44)

KIV repeats and prediction of Lp(a) levels and CHD risk

Table 2 shows a comparison of the informativeness of the shorter- and longer-isoform alleles and of the different methods used to classify KIV repeats for prediction of CHD risk and Lp(a) levels. Fifths of KIV repeats based on the lowest KIV repeat number observed in an individual (i.e. the size of the shorter-isoform allele) appeared to be the strongest predictor of both Lp(a) levels and of CHD risk (= 3.9 × 10−106 and = 9.7 × 10−5, respectively). Consequently, this measure of KIV repeats was used in all subsequent analyses.

Table 2. Predictive strength of alternative representations of KIV repeat measurements for Lp(a) levels and CHD risk
Alternative independent variablesP-value for prediction of
Log Lp(a) levelCHD risk
  1. KIV, kringle IV; Lp(a), lipoprotein(a); CHD, coronary heart disease.

  2. P-values are based on generalized score statistics showing the significance of independent variables for log Lp(a) level and CHD risk.

Total KIV repeats (sum of both alleles)7.1 × 10−825.5 × 10−4
Fifths based on total KIV repeats8.9 × 10−801.4 × 10−3
Lowest KIV repeat number (shorter-isoform allele)3.3 × 10−996.8 × 10−5
Fifths based on lowest KIV repeat number3.9 × 10−1069.7 × 10−5

Relationships between Lp(a), KIV repeats and baseline characteristics

Amongst controls, plasma levels of Lp(a) were inversely correlated with KIV repeats (ρ = −0.5). This correlation was slightly stronger in CHD cases (ρ = −0.6). The distribution of Lp(a) levels and KIV repeats within fifths in the control group and the cross-classification of Lp(a) levels by KIV repeats are shown in Tables S1 and S2. More than 25% of CHD cases were in the fifth with highest Lp(a) levels and lowest number of KIV repeats, albeit the correlation between fifths of Lp(a) levels and fifths of KIV repeats was less marked in the remaining individuals. Plasma levels of Lp(a) were inversely associated with fifths of KIV repeats in controls (Tables S3 and S4) and showed a similar pattern in CHD cases and controls (Fig. S2). In addition, Lp(a) levels were weakly inversely associated with the prevalence of diabetes (= 0.02) and with body mass index (= 0.05) and were positively associated with LDL-C (= 0.02) and apolipoprotein B (= 0.04). The number of individuals with diabetes was very small, and the trends with body mass index, LDL-C and apolipoprotein B were not consistent across fifths of Lp(a). However, because the LDL-C assay includes the cholesterol component of the Lp(a) particle, it would be expected that higher measured LDL-C levels would be associated with higher Lp(a) levels. Whilst the proportion of men and the LDL-C levels were correlated with fifths of KIV repeats (= 0.001 and = 0.04, respectively), other risk factors were unrelated to fifths of Lp(a) levels or KIV repeats (Tables S3 and S4).

Associations between CHD risk and both Lp(a) levels and KIV repeats

The patterns of the associations between CHD risk and both Lp(a) levels and KIV repeats are shown in Fig. 1. Differences in geometric mean Lp(a) levels between successive fifths varied considerably and, as a result of this distribution of Lp(a) levels, almost all of the excess risk of CHD associated with Lp(a) levels was restricted to individuals in the two-fifths with highest Lp(a) levels. By contrast, most of the excess risk associated with KIV repeats was restricted to the fifth of individuals with the fewest KIV repeats.

image

Figure 1. Odds ratio for risk of CHD by fifths of Lp(a) level (a) and KIV repeats (b). Estimates are adjusted for age, sex and country and plotted as the geometric mean Lp(a) level in cases and controls within each fifth (as defined in controls only) (a) and as the median KIV repeats in cases and controls within each fifth (as defined in controls only) (b). CIs are based on floating absolute risks. A fitted linear regression line through the points is shown. Ranges of the fifths of Lp(a) levels and KIV repeats are shown in Table S1. CHD, coronary heart disease; Lp(a), lipoprotein(a); KIV, kringle IV; CI, confidence interval.

Download figure to PowerPoint

Individuals in the top fifth of Lp(a) levels had a twofold higher risk of CHD (OR 2.04, 95% CI 1.52–2.74, < 0.0001) compared with those in the bottom fifth (Fig. 2). After further adjustment for conventional risk factors, including diabetes, hypertension and smoking, and also for fifths of KIV repeats, this effect was essentially unaltered (OR 2.05, 95% CI 1.38–3.04, < 0.0001).

image

Figure 2. Association between CHD and Lp(a) before and after adjustment for KIV repeats (smallest isoform) and between CHD and KIV repeats before and after adjustment for Lp(a) levels. Chi-squared (χ2) and P-values are shown for generalized score statistics assuming a linear trend across fifths. CHD, coronary heart disease; Lp(a), lipoprotein(a); KIV, kringle IV.

Download figure to PowerPoint

Individuals in the bottom fifth of KIV repeats also had an approximately twofold higher risk of CHD (OR 1.87, 95% CI 1.40–2.49, < 0.0001) compared with those in the top fifth. After adjustment for conventional risk factors, KIV repeats remained strongly and significantly associated with CHD risk (OR 1.84, 95% CI 1.34–2.53, = 0.0007). However, after further adjustment for fifths of Lp(a) levels, the effect of KIV repeats on CHD risk was substantially attenuated (OR 1.13, 95% CI 0.77–1.66, = 0.94) and was no longer statistically significant.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

The findings of the present study demonstrate the relative importance of Lp(a) levels and apo(a) isoform size for CHD risk. Higher Lp(a) levels were associated with a higher risk of CHD independent of KIV repeats. Furthermore, whilst fewer KIV repeats were associated with higher plasma levels of Lp(a) and higher risk of CHD, these associations were attenuated substantially after adjustment for Lp(a) levels. The pattern of the association between Lp(a) and CHD in the present study suggests that most of the excess CHD risk is restricted to the two-fifths of individuals with highest Lp(a) levels, relating to about a twofold excess risk of CHD in those with Lp(a) levels above about 20 mg dL−1 compared with individuals who have lower Lp(a) levels. The pattern of the association between KIV repeats and CHD suggests that most of the excess CHD risk is in the fifth of individuals with the fewest KIV repeats, relating to about a twofold excess risk of CHD in individuals with 22 or fewer KIV repeats compared with individuals with more KIV repeats. The present results indicate that the impact of KIV repeats on CHD risk is chiefly mediated through their effect on Lp(a) levels, consistent with the findings from previous studies involving considerably fewer CHD cases [10, 11, 13, 14].

In a prospective study involving 145 CHD cases, Lamon-Fava et al. [11] found that apo(a) isoform size contributed only modestly to the association between Lp(a) and CHD and was not an independent predictor of CHD risk. Likewise, Sandholzer et al. reported that alleles at the apo(a) locus determine their impact on CHD risk through effects on Lp(a) levels [13, 14]. A large Danish study demonstrated that KIV repeats, as estimated by qPCR assay, explained about 20–25% of the variation in Lp(a) levels and that the association between KIV repeats and CHD was only partially attenuated after adjustment for Lp(a) levels [2]. By contrast, in the present study, KIV repeats, as estimated by immunoblotting, explained about twice as much of the variation in Lp(a) levels (i.e. 40–50%), and the effect of KIV repeats on CHD risk was substantially attenuated after adjustment for Lp(a) levels. The discrepant findings of the PROCARDIS and the Danish studies may reflect the limitations of the qPCR assay for estimating apo(a) isoform size. In addition, differences in the assays used to measure Lp(a) levels or apo(a) isoform size and in study populations may account for between-study variation.

The results of the present study show, for comparable differences in Lp(a) levels, much stronger associations with CHD risk than those reported by the Emerging Risk Factor Collaboration meta-analysis of prospective studies, which concluded that increased levels of Lp(a) were at best only a modest risk factor for CHD [8]. These discrepant findings may reflect differences in the Lp(a) assays used in the various studies that contributed to the meta-analysis and in the present study. Therefore, further large-scale studies using standardized techniques are needed to compare the relevance of Lp(a) in different populations. Additional research to assess the biological relevance of KIV repeats in different ethnic groups, the influence of different Lp(a) assay methods (e.g. polyclonal vs. monoclonal assays) and possible mechanisms underlying the excess risks (e.g. the impact of oxidized phospholipids [19]) is also needed to better understand how Lp(a) levels cause vascular disease. Furthermore, because measurement of KIV repeats is technically difficult, time-consuming, costly and has limited precision, future studies involving alternative approaches to quantifying apo(a) isoform size that can address some of these issues and refine the measurement techniques may help to provide additional insight into potential differences in the associations between Lp(a) levels and disease risk in particular subgroups, such as those with diabetes [20] or renal disease [21]. In the present study, information on the most strongly expressed apo(a) isoform was not available. However, as previously demonstrated, the smaller allele dominates or codominates in ~75% of individuals and as only about one-third of individuals in the present study showed two bands, use of the most strongly expressed apo(a) isoform as an alternative parameterization would be unlikely to materially affect the results [22]. Furthermore, despite an apparent lack of sensitivity of the apo(a) immunoblotting method in the present study, the strong associations between CHD risk and apo(a) isoform size clearly demonstrate the informativeness of this approach.

Genetic studies indicate a causal relevance of Lp(a) for CHD [1-3]; therefore, treatments aimed at achieving reductions in plasma levels of Lp(a) may offer therapeutic benefit, especially in people with high levels of Lp(a). Screening to identify those individuals with increased Lp(a) levels who may benefit from the use of Lp(a)-lowering treatments might also be warranted [23]. Niacin lowers Lp(a), but its lack of efficacy for lowering vascular risk as well as the adverse effects reported in large-scale randomized trials limits any potential uses [24-26]. Furthermore, anacetrapib, a cholesterol ester transfer protein inhibitor, lowers Lp(a) levels by about 30–40% [27] in addition to effects on other lipids; a large-scale trial is currently underway to assess its impact on vascular risk [28]. Examining the effects of any Lp(a)-lowering treatment in those with very high Lp(a) levels should be particularly informative. For example, Jaeger et al. [29] demonstrated that plasma apheresis substantially reduced both Lp(a) levels and CHD events in high-risk individuals with increased levels of Lp(a). Moreover, there is a need to discover, and assess the effects on vascular risk of, more efficacious and safe treatments that lower Lp(a) levels independently of their effects on other lipids. As apo(a) isoform size is unaltered by therapy, the results of the present study demonstrating that the absolute levels of Lp(a) provide most of the relevant information for risk prediction suggest that measurements of apo(a) isoform size are unlikely to alter treatment decisions. Therefore, lowering absolute levels of Lp(a) should be the main objective of any Lp(a)-lowering strategy for prevention of CHD in addition to modification of conventional risk factors.

Conflict of interest statement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

The PROCARDIS study was supported by the British Heart Foundation (BHF), the European Community Sixth Framework Programme (LSHM-CT-2007-037273), AstraZeneca, the Wellcome Trust, the UK Medical Research Council, the Swedish Heart–Lung Foundation, the Swedish Medical Research Council, the Knut and Alice Wallenberg Foundation and the Karolinska Institutet. The Clinical Trial Service Unit and Epidemiological Studies Unit (CTSU) at the University of Oxford receives core support from the UK Medical Research Council, Cancer Research UK and BHF as well as funding from the BHF Centre for Research Excellence, Oxford. The CTSU, University of Oxford, is the trial sponsor (in collaboration with Merck) for the HPS2-THRIVE trial of extended-release niacin laropiprant and the HPS3-TIMI55-REVEAL trial of anaceptrapib. The CTSU has a policy of not accepting honoraria or other payments from the pharmaceutical industry, except for reimbursement of costs to participate in scientific meetings. MF, TK, AG and HW acknowledge support from the Wellcome Trust core award (090532/Z/09/Z). JCH, MF, TK, HW, RCo and RCl acknowledge support from the BHF Centre for Research Excellence, Oxford.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information

We gratefully acknowledge expert technical assistance from Bertram Tambyrajah at Leibniz-Institut für Arterioskleroseforschung an der Universität Münster, Münster, Germany.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Clarke R, Peden JF, Hopewell JC et al. Genetic variants associated with Lp(a) lipoprotein level and coronary disease. N Engl J Med 2009; 361: 251828.
  • 2
    Kamstrup PR, Tybjaerg-Hansen A, Steffensen R, Nordestgaard BG. Genetically elevated lipoprotein(a) and increased risk of myocardial infarction. JAMA 2009; 301: 23319.
  • 3
    Tregouet DA, Konig IR, Erdmann J et al. Genome-wide haplotype association study identifies the SLC22A3-LPAL2-LPA gene cluster as a risk locus for coronary artery disease. Nat Genet 2009; 41: 2835.
  • 4
    Kronenberg F, Utermann G. Lipoprotein(a): resurrected by genetics. J Intern Med 2013; 273: 630.
  • 5
    Berg K. A new serum type system in man–the Lp system. Acta Pathol Microbiol Scand 1963; 59: 36982.
  • 6
    Utermann G. The mysteries of lipoprotein(a). Science 1989; 246: 90410.
  • 7
    Berglund L, Ramakrishnan R. Lipoprotein(a): an elusive cardiovascular risk factor. Arterioscler Thromb 2004; 24: 221926.
  • 8
    Erqou S, Kaptoge S, Perry PL et al. Lipoprotein(a) concentration and the risk of coronary heart disease, stroke, and nonvascular mortality. JAMA 2009; 302: 41223.
  • 9
    Erqou S, Thompson A, Di Angelantonio E et al. Apolipoprotein(a) isoforms and the risk of vascular disease: systematic review of 40 studies involving 58,000 participants. J Am Coll Cardiol 2010; 55: 21607.
  • 10
    Farrer M, Game FL, Albers CJ et al. Coronary artery disease is associated with increased lipoprotein(a) concentrations independent of the size of circulating apolipoprotein(a) isoforms. Arterioscler Thromb 1994; 14: 127283.
  • 11
    Lamon-Fava S, Marcovina SM, Albers JJ et al. Lipoprotein(a) levels, apo(a) isoform size, and coronary heart disease risk in the Framingham Offspring Study. J Lipid Res 2011; 52: 11817.
  • 12
    Rifai N, Ma J, Sacks FM et al. Apolipoprotein(a) size and lipoprotein(a) concentration and future risk of angina pectoris with evidence of severe coronary atherosclerosis in men: the Physicians' Health Study. Clin Chem 2004; 50: 136471.
  • 13
    Sandholzer C, Boerwinkle E, Saha N, Tong MC, Utermann G. Apolipoprotein(a) phenotypes, Lp(a) concentration and plasma lipid levels in relation to coronary heart disease in a Chinese population: evidence for the role of the apo(a) gene in coronary heart disease. J Clin Invest 1992; 89: 10406.
  • 14
    Sandholzer C, Saha N, Kark JD et al. Apo(a) isoforms predict risk for coronary heart disease. A study in six populations. Arterioscler Thromb 1992; 12: 121426.
  • 15
    Langer C, Tambyrayah B, Nowak-Gottl U. Testing for apolipoprotein(a) phenotype using isoelectric focusing and immunoblotting technique. Methods Mol Biol 2013; 992: 40712.
  • 16
    Marcovina SM, Hobbs HH, Albers JJ. Relation between number of apolipoprotein(a) kringle 4 repeats and mobility of isoforms in agarose gel: basis for a standardized isoform nomenclature. Clin Chem 1996; 42: 4369.
  • 17
    Siekmeier R, Marz W, Scharnagel H. Determination of lipoprotein(a): comparison of a novel latex enhanced immunoturbidimetric assay and an immunoradiometric assay. Lab Medizin 1996; 20: 2948.
  • 18
    Marcovina SM, Koschinsky ML, Albers JJ, Skarlatos S. Report of the National Heart, Lung, and Blood Institute Workshop on Lipoprotein(a) and Cardiovascular Disease: recent advances and future directions. Clin Chem 2003; 49: 178596.
  • 19
    Tsimikas S, Brilakis ES, Miller ER et al. Oxidized phospholipids, Lp(a) lipoprotein, and coronary artery disease. N Engl J Med 2005; 353: 4657.
  • 20
    Mora S, Kamstrup PR, Rifai N, Nordestgaard BG, Buring JE, Ridker PM. Lipoprotein(a) and risk of type 2 diabetes. Clin Chem 2010; 56: 125260.
  • 21
    Kronenberg F, Utermann G, Dieplinger H. Lipoprotein(a) in renal disease. Am J Kidney Dis 1996; 27: 125.
  • 22
    Rubin J, Paultre F, Tuck CH et al. Apolipoprotein [a] genotype influences isoform dominance pattern differently in African Americans and Caucasians. J Lipid Res 2002; 43: 23444.
  • 23
    Nordestgaard BG, Chapman MJ, Ray K et al. Lipoprotein(a) as a cardiovascular risk factor: current status. Eur Heart J 2010; 31: 284453.
  • 24
    Boden WE, Probstfield JL, Anderson T et al. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. N Engl J Med 2011; 365: 225567.
  • 25
    Carlson LA. Nicotinic acid: the broad-spectrum lipid drug. A 50th anniversary review. J Intern Med 2005; 258: 94114.
  • 26
    HPS2-THRIVE Collaborative Group. HPS2-THRIVE randomized placebo-controlled trial in 25 673 high-risk patients of ER niacin/laropiprant: trial design, pre-specified muscle and liver outcomes, and reasons for stopping study treatment. Eur Heart J 2013; 34: 127991.
  • 27
    Cannon CP, Shah S, Dansky HM et al. Safety of anacetrapib in patients with or at high risk for coronary heart disease. N Engl J Med 2010; 363: 240615.
  • 28
    HPS3-TIMI55-REVEAL. Randomized Evaluation of the Effects of Anacetrapib Through Lipid-modification. http://clinicaltrials.gov/ct2/show/NCT01252953.
  • 29
    Jaeger BR, Richter Y, Nagel D et al. Longitudinal cohort study on the effectiveness of lipid apheresis treatment to reduce high lipoprotein(a) levels and prevent major adverse coronary events. Nat Clin Pract Cardiovasc Med 2009; 6: 22939.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
joim12187-sup-0001-FigS1-S2.pdfapplication/PDF131K

Figure S1. Example of an immunoblot from the present study that was used to estimate the number of KIV repeats.

Figure S2. Association between geometric mean Lp(a) levels and KIV repeats in CHD cases and controls.

joim12187-sup-0002-TableS1-S4.docWord document154K

Table S1. Distribution of (a) Lp(a) levels and (b) KIV repeats (smallest isoform) within fifths (defined in controls).

Table S2. Classification by fifths of KIV repeats (smallest isoform) and Lp(a) levels shown as the proportion of (a) all individuals (b) CHD cases and controls separately.

Table S3. Characteristics by fifths of Lp(a) levels amongst controls.

Table S4. Characteristics by fifths of KIV (smallest isoform) repeats amongst controls.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.