UMEÅ UNIVERSITY MEDICAL DISSERTATIONSNew series No 386 - ISSN 0346 - 6612

From the Departments of Internal Medicine and Clinical Chemistry, University of Umeå, Umeå, Sweden

Serum Lipoprotein(a) in Relation to Ischemic Heart Disease and

Associated Risk Factors

by

Lisbeth Slunga

UMEÅ UNIVERSITY MEDICAL DISSERTATIONSNew series No 386 - ISSN 0346 - 6612

From the Departments of Internal Medicine and Clinical Chemistry, University of Umeå, Umeå, Sweden

Serum Lipoprotein(a) in Relation to Ischemic Heart Disease and

Associated Risk Factors

AKADEMISK AVHANDLING som med vederbörligt tillstånd av Rektorsämbetet vid Umeå Universitet

för avläggande av medicine doktorsexamen kommer att offentligen försvaras i sal B, 9 tr, Regionsjukhuset i Umeå, fredagen den 22 oktober 1993 kl 09.00

av

Lisbeth Slunga

Umeå 1993

ABSTRACT

Serum lipoprotein(a) in relation to ischemic heart disease and associated risk factorsLisbeth Slunga, Department of Medicine, University of Umeå,S-901 85 UMEÅ, Sweden

Lipoprotein(a) (Lp(a)) consists of an LDL-like particle and the specific protein apo(a), which is very similar to plasminogen. Apo(a) contains repeated kringle structures and a serine protease domain, which cannot be activated by t-PA. Lp(a) is considered to be a predictor for atherosclerotic disease. It has been found incorporated in atherosclerotic plaques and inhibits in vitro fibrinolysis.

Lp(a) was determined in 1527 randomly selected individuals participating in the Northern Sweden WHO-MONICA project. A weak but significant relation between Lp(a) and increasing age was found. Menopausal status was the strongest independent predictor of Lp(a) level in women. Fibrinogen was independently related to Lp(a) in both sexes. Only a minor fraction of Lp(a) variance could be explained for in a multiple regression model, which is in agreement with the contention that Lp(a) is highly genetically determined.

Lp(a) was determined in 1571 patients investigated with coronary angiography because of suspected severe coronary artery disease (CAD). Patients with proven CAD at elective angiography had significantly higher Lp(a) than patients without significant CAD or healthy controls. Lp(a) was found to be an independent discriminator of CAD in both sexes.

HLA-DR genotype 13 or 17 was found more frequently in 30 male patients with angiographic CAD at young age (< 50 years) than in 30 age matched controls. These genotypes were common in patients with high Lp(a) levels, which indicates that Lp(a) may be related to immunological processes.

The reaction of Lp(a) was investigated in 32 patients with acute myocardial infarction (AMI). Lp(a) increased during the first week, but the response was comparatively weak. Individual Lp(a) responses were heterogeneous and no correlations to infarct size or changes in the acute phase proteins were found.

In a randomized cross-over study on 36 hypercholesterolaemic patients treated with simvastatin/placebo during 12+12 weeks Lp(a) did not change significantly, but patients with high Lp(a) levels at baseline tended to develop further increased Lp(a).

To conclude, Lp(a) was found to be an independent predictor of angiographic CAD in both men and women. Lp(a) levels are primarily genetically determined and only a small fraction of Lp(a) variance could be explained by other factors in this study. Lp(a) may be related to HLA DR types and immunological processes involved in atherosclerotic disease. Lp(a) increased slightly during the first week of AMI, but was not related to changes in the acute-phase proteins. The effective LDL-lowering agent simvastatin did not influence Lp(a) significantly.

KEY WORDS: Lipoprotein(a), lipids, epidemiology, coronary artery disease, coronary angiography, HLA DR, acute-phase protein, myocardial infarction, HMG CoA reductase inhibitor, simvastatin

UMEÅ UNIVERSITY MEDICAL DISSERTATIONSNew series No 386 - ISSN 0346 - 6612

From the Departments of Internal Medicine and Clinical Chemistry, University of Umeå, Umeå, Sweden

Serum Lipoprotein(a) in Relation to Ischemic Heart Disease and

Associated Risk Factors

by

Lisbeth Slunga

A i v

Umeå 1993

Copyright © Lisbeth Slunga ISBN 91-7174-828-8

Printed in Sweden by the Printing Office of Umeå University

Umeå 1993

TABLE OF CONTENTS

ABSTRACT............................................................... .......................................... 5

ORIGINAL PAPERS.......................................................................................... 6

INTRODUCTION............................................................................................... 7

AIMS OF THE STUDIES................................................................................... 11

SUBJECTS, STUDY DESIGN AND METHODS............................................ 12

P ap e ri......................... 12II...................................................................................................... 13m .......................................................................................................14I V.................................................................................................... 14 V...................................................................................................... 15

Laboratory procedures.......................................................................... 16Statistical analyses.................................................................................. 17Ethical aspects......................................................................................... 18

RESULTS.............................................................................................................. 18

P ap eri....................................................................................................... 18I I...................................................................................................... 20II I.......................................................................................................26I V.................................................................................................... 26 V...................................................................................................... 27

DISCUSSION....................................................................................................... 28

Lp(a) in the general population........................................................... 28

Lp(a) in coronary artery disease.......................................................... 30

Pathogenic mechanisms in Lp(a) related atherosclerotic disease 34

Lp(a) in the acute phase reaction......................................................... 36

Treatment of elevated Lp(a) levels...................................................... 38

CONCLUSIONS................................................................................................... 39

ACKNOWLEDGEMENTS......................... 41

REFERENCES...................................................................................................... 43

ABBREVIATIONS

AMI acute myocardial infarctionBMI body mass indexCAD coronary artery diseaseCRP C-reactive proteinDBP diastolic blood pressureECG electrocardiogramHDL high density lipoproteinHRT hormonal replacement therapyIDDM insulin-dependent diabetes mellitusLDL low density lipoproteinLp(a) lipoprotein(a)NIDDM non-insulin-dependent diabetes mellitusNSAID nonsteroidal anti-inflammatory drugsPAI plasminogen activator inhibitorSBP systolic blood pressureSR sedimentation rateTG triglyceridestPA tissue plasminogen activator

ABSTRACT

Lipoprotein(a) (Lp(a)) consists of an LDL-like particle and the specific protein apo(a), which is very similar to plasminogen. Apo(a) contains repeated kringle structures and a serine protease domain, which cannot be activated by t-PA. Lp(a) is considered to be a predictor for atherosclerotic disease. It has been found incorporated in atherosclerotic plaques and inhibits in vitro fibrinolysis.

Lp(a) was determined in 1527 randomly selected individuals participating in the N orthern Sweden WHO-MONICA project. A weak but significant relation betw een Lp(a) and increasing age was found. M enopausal status was the strongest independent predictor of Lp(a) level in women. Fibrinogen was independently related to Lp(a) in both sexes. Only a minor fraction of Lp(a) variance could be explained for in a multiple regression model, which is in agreement with the contention that Lp(a) is highly genetically determined.

Lp(a) was determ ined in 1571 patients investigated w ith coronary angiography because of suspected severe coronary artery disease (CAD). Patients w ith proven CAD at elective angiography had significantly higher Lp(a) than patients w ithout significant CAD or healthy controls. Lp(a) was found to be an independent discriminator of CAD in both sexes.

HLA-DR genotype 13 or 17 was found more frequently in 30 male patients w ith angiographic CAD at young age (< 50 years) than in 30 age m atched controls. These genotypes were common in patients w ith high Lp(a) levels, which indicates that Lp(a) may be related to immunological processes.

The reaction of Lp(a) was investigated in 32 patients with acute myocardial infarction (AMI). Lp(a) increased during the first week, but the response was comparatively weak. Individual Lp(a) responses were heterogeneous and no correlations to infarct size or changes in the acute phase proteins were found.

In a random ized cross-over study on 36 hypercholesterolaemic patients treated w ith sim vastatin/placebo during 12+12 weeks Lp(a) did not change significantly, but patients with high Lp(a) levels at baseline tended to develop further increased Lp(a).

To conclude, Lp(a) was found to be an independen t p red ic tor of angiographic CAD in both men and women. Lp(a) levels are prim arily genetically determ ined and only a small fraction of Lp(a) variance could be explained by other factors in this study. Lp(a) may be related to HLA DR types and im m unological processes involved in atherosclerotic disease. Lp(a) increased slightly during the first week of AMI, but was not related to changes in the acute-phase proteins. The effective LDL-lowering agent simvastatin did not influence Lp(a) significantly.

KEY WORDS: Lipoprotein(a), lipids, epidemiology, coronary artery disease, coronary angiography, HLA DR, acute-phase protein, myocardial infarction, HMG CoA reductase inhibitor, simvastatin

ORIGINAL PAPERS

This thesis is based on the following papers, which are referred to in the text bytheir Roman numerals:

I. Slunga L, Asplund K, Johnson O, Dahlén GH. Lipoprotein(a) in a randomly selected 25-64-year-old population: the Northern Sweden MONICA Study. J Clin Epidemiol 1993; 46: 617-624.

II. Slunga L, Johnson O, Wester PO, Dahlén GH. Lipoprotein(a) in patients investigated with coronary angiography because of suspected severe coronary artery disease. Submitted for publication.

III. Dahlén GH, Slunga L, Holmlund G, Langö A, Lindblom B. Lp(a) lipoprotein and HLA-DR genotype in early coronary artery disease.Eur J Immunogenet 1993; 20: 95-102.

IV. Slunga L, Johnson O, Dahlén GH, Eriksson S. Lipoprotein(a) and acute- phase proteins in acute myocardial infarction. Scand J Clin Lab Invest 1992; 52: 95-101.

V. Slunga L, Johnson O, Dahlén GH. Changes in Lp(a) lipoprotein levels during the treatment of hypercholesterolaemia with simvastatin. Eur J Clin Pharmacol 1992; 43: 369-373.

INTRODUCTION

Atherosclerotic disease is the predominant cause of morbidity and mortality in industrialized countries. It usually starts early in life as a silent process, which involves accumulation in the arterial intima of increasing amounts of lipid and connective tissue matrix proteins and formation of plaques. Clinical symptoms; angina pectoris, AMI or stroke, most often appear decades later and are related to plaques complications with thrombosis. The etiology of the atherosclerotic process is complex and multifactorial with both genetic and environmental, life-style factors contributing and co-operating. Hypertension, smoking, diets rich in saturated fat and cholesterol, and diabetes are some factors known to increase the risk of atherosclerotic disease. A propensity to develop early manifest atherosclerotic disease is most often related to strong, genetic factors (1, 2). Familial hypercholesterolaemia, which is characterized by defective LDL- receptors, is one example of a genetic disturbance strongly associated w ith early atherosclerosis. It is well-known that individuals from the general population respond very differently in lipid levels to dietary modifications. Genes probably contribute to determining the limits w ithin which life-style factors can cause risk factor changes in a certain individual (3).

Kåre Berg reported in 1963 a new antigen in blood, the Lp(a) antigen, which was present in approximately 35% of healthy people and clearly under genetic control (4).

In 1971 Gösta Dahlén reported on a new lipoprotein fraction, which did not fit into the Fredricksson's classification of hyperlipemia (5). It was designated pre-betai-lipoprotein due to its electrophoretic mobility. This lipoprotein fraction was detected in approximately 20% of the reference population. A highly significant association between the occurrence of this lipoprotein and clinical angina pectoris was found. In collaborative studies Kåre Berg and Gösta Dahlén established that Lp(a) lipoprotein and pre-betai-lipoprotein were related, probably identical (6, 7).

Lp(a) lipoprotein/pre-betai-lipoprotein was originally considered to be a qualitative trait under strong genetic control and related to atherosclerotic disease (8, 9). When sensitive immunologic testing methods became available it was m ade clear that Lp(a) was present in virtually all individuals and constitu ted a continuous variable. In further case-control studies Lp(a) lipoprotein was found to be an independent predictor for myocardial infarction (10) and angiographic CAD (11). An Lp(a) level greater than 300 mg/1 was

7

reported to im part an increased risk for CAD (11, 12). Lp(a) has also been show n to be a significant predictor for stroke (13-15) and peripheral artery disease (16, 17). In two prospective studies Lp(a) constituted a significant predictor of CAD (18,19). In a nested case-control study of participants in the prim ary preventive Helsinki Heart Study Lp(a) was, however, not found to predict future coronary events (20).

Lp(a) consists of an LDL-like particle and the specific protein apo(a), which is linked to apoBlOO by a disulphide bridge (21) (Figure 1).

K4

apo B LDL

K5K4 K4

Figure 1. Structure of lipoprotein(a).

In 1987 McLean et al. revealed by sequencing of cloned hum an apo(a)-cDNA, that apo(a) was very similar to plasminogen, a plasma protein involved in the fibrinolytic process (22). Apo(a) contained a serine protease domain and two types of plasminogen-like kringle domains; a single kringle 5 and a sequence of 37 repeated kringle 4 (Figure 2). Kringles are triple-loop domains, which are common in proteins involved in the fibrinolytic system, clotting cascade and complement system. Apo(a) can not be activated by tPA or streptokinase due to a substitution of arginine to serine at the protease activation site.

8

S TPlasminogen Kl K2 K3 K4 K5

apo(a) K4 K4 K4 -Ib K4 K4 K4 K4 K4 K5

Figure 2. Comparison of the gene structures of plasminogen and apo(a). S = signal peptide, T = ”tail" region, Kl-5 = kringle domains, P = protease domain.

Apo(a) is a large glycoprotein w ith a molecular w eight ranging from approximately 350 to 900 kD (23). The size heterogeneity of apo(a) is due to a highly varied polypeptide chain size and a highly glycosylated structure. The interindividual variation in Lp(a) concentration is enormous. Serum Lp(a) concentration thus ranges from undetectable to more than 1000 mg/1. The distribution of Lp(a) in Caucasians is extremely skewed w ith most of subjects having serum Lp(a) concentrations at the low end of the range. In Africans the Lp(a) d istribu tion is closer to norm al, and the average Lp(a) level is considerably higher than in Caucasians (24, 25). A quantitative polymorphism at the apo(a) locus explains the highly varied polypeptide chain size and the large variance of the serum Lp(a) concentration (26,27). An inverse relationship betw een the num ber of repeated kringle 4 and the serum level of Lp(a) has been found (28, 29). Different numbers of apo(a) isoforms have been detected depending on the sensitivity of the phenotyping methods used. With a high- resolution SDS-agarose electrophoresis method followed by immunoblotting 23 different apo(a) isoforms were identified (30). According to Cohen et al. it can be estim ated from pulsed-field data combined w ith data from single-strand DNA conformation polymorphisms, that there may be more than 100 different alleles at the apo(a) locus (31). The apo(a) gene resides on chromosome six adjacent to the plasminogen gene (32). A close linkage between apo(a) isoforms and DNA polymorphisms at the plasminogen locus has been found (33).

The interest in Lp(a) has been extensive since the discovery of the structure of apo(a). Lp(a) may constitute a link between lipoproteins and thrombosis, probably having both atherogenic and thrombogenic properties. Apo(a) has been found incorporated in large amounts in arterial plaques (34) and in vein grafts (35) . Kringle 4, found in plasminogen and apo(a), has lysine-binding capacity and the potential of binding to fibrinogen, fibrin products and cell

9

receptors (36). In vitro studies have shown that Lp(a) competes w ith and inhibits binding of plasminogen to fibrin products (37-39) and to cell receptors (40-42). Lp(a) also interferes w ith streptokinase-m ediated activation of plasminogen (43,44) and the tPA-induced lysis of fibrin clots (38).

Apo(a) is primarily synthesized in the liver (45), but mRNA for apo(a) has also been found in testis and brain tissue (46). Lp(a) is probably synthesized independently of other apoB-containing lipoproteins (47, 48). The serum level of Lp(a) is determined by the rate of synthesis (49). The further metabolism and degradation of Lp(a) is unclear. Apo(a) fragments have been detected in urine, bu t no relation between serum creatinine or creatinine clearance and serum Lp(a) was found (50). Conflicting results have been obtained concerning the possible catabolism of Lp(a) via the LDL-receptor. Most studies have reported a considerably lower binding of Lp(a) to the LDL-receptor as com pared w ith LDL (51-54). However, in transgenic mice an overexpression of hum an LDL- receptors lead to an accelerated catabolism of Lp(a) (55).

The physiological roles of Lp(a) are obscure. Apart from in humans, Lp(a) has been detected in monkeys, principally Old World monkeys (56, 57), and in the hedgehog (58). It has been proposed that Lp(a) may be of importance for cell repair and wound healing (59).

The intraindividual level of Lp(a) has been claimed to remain very stable throughout life. Different disease states, however, affect Lp(a) levels. Increased Lp(a) levels have thus been reported in patients with end-stage renal disease or heavy proteinuria (60-63), in diabetics with nephropathy (64, 65), in patients w ith rheum atoid arthritis (66) and in cancer patients (67). Lp(a) levels are reduced in patients with liver cirrhosis (68, 69).

10

AIMS OF THE STUDIES

• To assess the distribution and variation of Lp(a) in the general population and to relate Lp(a) to other predictors for CAD (I).

• To assess the distribution and variation of Lp(a) in a population investigated with coronary angiography because of suspected severe CAD and the relation of Lp(a) to other predictors for CAD (II).

• To assess the value of Lp(a) as a predictor for CAD (II).

• To investigate whether any relations can be found between Lp(a) and HLA- DR antigens and susceptibility to early CAD (IE).

• To evaluate if Lp(a) can be regarded as an acute-phase protein in acute myocardial infarction (IV).

• To investigate if treatment with the HMG CoA reductase inhibitor simvastatin influences Lp(a) level (V).

11

SUBJECTS, STUDY DESIGN AND METHODS

Paper I

A screening of cardiovascular risk factors in the general population in northern Sweden was perform ed in January to April 1990 as a part of the WHO M ONICA Study. Originally 2000 individuals, 25-64-years old, stratified according to sex and age and randomly chosen from continuously updated population registers in the counties Norrbotten and Västerbotten, were invited by letter to participate in the study. In total, 1583 subjects (79.2%) accepted. Ind iv iduals w ho did not participate were contacted by telephone and interview ed concerning reasons for not attending, social background and cardiovascular risk factors. Results of Lp(a) analyses were available in 1527 participating subjects (76.4% of the originally invited subjects), who were subsequently enrolled in the present study.

Participants were asked to complete at home a questionnaire concerning cardiovascular risk factors, medical history and intake of drugs. Angina pectoris was diagnosed according to criteria established by Rose (70). Previous m yocardial infarction was diagnosed in individuals who stated previous treatm ent at hospital because of definite myocardial infarction.

Two mobile teams performed the screening procedure in the whole study area. Examinations and sampling were performed between 7 a. m. and 3 p. m. Subjects who were examined before noon were instructed to be fasting since 12 hours. Data concerning last intake of food or drink were collected. It was established that 56% of the study population was at the time of sam pling fasting since 12 hours. TG and calculated LDL values are only reported from fasting individuals. Total cholesterol, HDL, fibrinogen and Lp(a) are reported from all subjects irrespective of fasting time.

For calculation of the waist-to-hip ratio the smallest circumference of the waist and the thickest part of the hip in the standing position were measured. Blood pressure was m easured in the sitting position after 5 min rest by a mercury sphygmomanometer using the random zero method (71). The m ean of tw o repeated m easurem ents was registered. Subjects were classified as hypertensive if they were treated with drugs because of hypertension or if SBP >160 mm Hg or if DBP >90 mm Hg.

Fasting subjects were examined w ith an oral glucose tolerance test in accordance w ith WHO guidelines; 75 g of glucose in 300 ml of w ater was

12

ingested over 5 min. Venous plasma samples for determination of glucose were taken immediately before the test and after 2 hours. Subjects were classified as diabetics if they already had a diagnosis of diabetes or if fasting or 2-hour plasm a glucose equalled or exceeded 7.8 or 11.1 mmol/1, respectively, at an oral glucose tolerance test. An impaired glucose tolerance was diagnosed when fasting plasma glucose was below 7.8 mmol/1 and 2-hour plasma glucose in the interval 7.8-11.1 mmol/1.

Paper II

From November 1987 to August 1990 a total of 1951 patients were examined w ith coronary angiography at the University Hospital in Umeå, because of suspected severe CAD which probably required intervention w ith coronary bypass grafting or angioplasty. The main indication for investigation was severe angina pectoris with restricted functional capacity. Due to insufficient data (missing records or unavailable results of Lp(a) analysis) 143 patients (7.3%) were excluded. Patients on lipid-lowering drugs (n = 146) and patients previously treated w ith coronary bypass grafting (n = 80) or coronary angioplasty (n = 7) were also excluded. The final study population consisted of 1571 patients, 1216 men and 355 women. This population was further divided into three groups; 1 / stable patients investigated w ith elective angiography, who had at least one significant coronary artery stenosis, 2 / unstable patients investigated w ith acute angiography, who had at least one significant coronary artery stenosis and 3 / patients without any significant coronary artery stenosis at angiography.

Patients were sampled, after on overnight fast, in the morning of the day of the coronary catheterization for determination of lipids and Lp(a). Samples for determination of fibrinogen and sedimentation rate were taken in a non-fasting condition on the day preceding the angiography. Clinical data were collected retrospectively from the patients medical records.

Coronary angiography was perform ed using Judkins' technique w ith m ultip le projections of the coronary arteries. A coronary stenosis was considered to be significant if the cross-sectional area was reduced w ith at least 75%. Myocardial coronary obstruction scores were calculated according to a m ethod described by Brandt et al (72). It implies an evaluation of the myocardial mass distal of the vessel obstruction in addition to the degree of

13

coronary obstruction. Collateral scores were determined by a similar method. Left ventricular motion scores were assessed by a method described by Austen et al (73).

Cardiovascularly healthy subjects from the N orthern Sweden MONICA Study (579 men and 525 women) were used as referents. Subjects w ith a history of previous myocardial infarction or stroke were excluded as referents, as well as subjects on lipid-lowering drugs and individuals w ith diabetes or angina pectoris.

Paper III

M en younger than 50 and wom en younger than 60 years at the time of angiographically confirmed CAD, who lived in the central areas of Umeå or Skellefteå municipals, were considered for participation in the study. From January 1987 to August 1990 in total 84 patients investigated w ith coronary angiography fulfilled these inclusion criteria. They all had at least one significant coronary stenosis at angiography. Coronary angiography was perform ed according to methods described in paper II. The present follow-up and investigation w ith standardized sampling was performed in early 1991. At that time six of the original patients were deceased and six patients were lost. Seven patients were not included because of some recent complicating disease or cardiovascular event and five patients declined to participate. In total 60 patients (71.4% of the original population), 30 men and 30 women, agreed to participate and were included in the study. From the municipal census list of Umeå 60 sex and age matched referents were recruited. Subjects w ith known cardiovascular disease or diabetes were excluded.

Patients and controls were invited to the hospital in the m orning, in a fasting condition since 12 hours, for sampling. Information on clinical data, heredity, menopausal status and present medication was collected. Venous blood was draw n for analysis of Lp(a) and HLA-DR genotypes.

Paper IV

Thirty-two patients (19 male and 13 female) treated at the coronary care unit at the University Hospital in Umeå because of definite AMI during late 1987 and

14

early 1988 were included in the study. Myocardial infarction was diagnosed by observing clinical symptoms, ECG findings and serial total creatine kinase as well as CKMB activities.

Patients treated w ith thrombolysis were not considered for participation because of possible effects on Lp(a) level by the thrombolytic agents. During the study period only 18% of all patients treated at the coronary care unit because of m yocardial infarction received thrombolytic therapy. The m ain reasons for not administrating thrombolytic agents were a symptom duration exceeding 4 hours, indecisive diagnosis at admission or some contraindication to thrombolysis.

Venous blood for determ ination of Lp(a) and the acute-phase proteins haptoglobin, orosomucoid and ai-antitrypsin was draw n in study patients at the time of admission to the coronary care unit and in the morning, after fasting overnight, on the following 6 days. The last 21 patients included in the study were also sampled daily for analysis of lipids and plasma albumin.

Paper V

Subjects w ere recruited to the study from patients w ho a ttended the Outpatients' Clinic at the Medical Department of the University Hospital in Umeå because of hyperlipidem ia during the period Novem ber 1988 to February 1990. Thirty-six patients (29 male and 7 female) w ith total cholesterol above 6.5 mmol/1 and serum Lp(a) above 100 mg/1 finally participated in and com pleted the trial according to the protocol. Patients w ith a recent cardiovascular event, unstable angina, diabetes type 1, elevated liver enzymes or a secondary hypercholesterolaemia were not considered for participation as well as prem enopausal women. All study patients had at baseline fasting serum TG below 4 mmol/1. All lipid-lowering drugs were stopped at least 6 weeks before the start of the baseline period and patients were instructed to follow a diet comparable to the American Heart Association Recommended Diet (phase 1) (74) during the whole study period.

The study was double-blind, placebo controlled with cross-over design. The baseline period lasted 6 weeks. Patients were then random ly allocated to receive sim vastatin /p lacebo during a 12 week period. The daily dose of sim vastatin/placebo was 10 mg during week 1-4, 20 mg during week 5-8 and

15

40 mg during week 9-12. After the first 12-week period each patient switched to the alternative treatment not yet received.

Patients visited the Outpatients' Clinic in the morning, having fasted for at least 12 h, in weeks -6, -2, 0 (baseline) and after 4, 8,12,16, 20 and 24 weeks of treatment. Determinations of total cholesterol, TG and safety laboratory tests were carried out at each visit. In weeks -6, -2, 0, 12 and 24 an extended sampling was performed, which included Lp(a), HDL, apolipoprotein A-I and apolipoprotein B. On these occasions patients also had an ECG and a physical examination was performed.

Laboratory procedures

Lp(a) lipoprotein was determined by an enzyme linked im munosorbent assay (ELISA). The detection limit was 10 mg/1 and the assay range was 10 - 600 mg/1. The day-to-day CV, coefficient of variation, was 5.4%.

Total cholesterol and TG were determ ined by enzymatic m ethod kits (Boehringer M annheim GmbH Diagnostica, M annheim, Germany). HDL- cholesterol was m easured after precipitation of the other lipoproteins w ith sodium phosphow olfram ate/m agnesium chloride. LDL-cholesterol was calculated by Friedewald's equation (75). Only TG values less than 4.5 mmol/1 w ere used in the calculations. Lp(a)-adjusted total cholesterol and LDL- cholesterol were calculated with subtraction of the cholesterol present in Lp(a), (Lp(a) (mg/l)*0.3/386.7). Apoprotein A-I and apoprotein B were measured by radioim m unoassay research kits (RIA 100, Pharmacia Diagnostics, Uppsala, Sweden).

Fibrinogen was in paper I determined by the reagent kit Fibrinogen Kinetics from Boehringer Mannheim, using a Hitachi 717 analyser. Fibrinogen was in paper II determ ined w ith a throm bin reaction-rate m ethod (BioMerieux, France). Haptoglobin, orosomucoid and ai-antitrypsin were quantified by electroimmunoassay.

CKMB activities were determ ined by immuno-inhibition, based on the presence of CKM-subunit antibodies.

DNA was isolated from frozen EDTA-blood according to the SDS-urea m ethod by Lindblom and Holmlund (76). Three micrograms of DNA were cleaved w ith nine units of the restriction enzyme Taq I. The fragments were separated by electrophoresis in 20 cm long 0.7% agarose gels for 22 h at 40V in

16

Trisborate-EDTA buffer. A 1 Kb DNA ladder (BRL) was present in four lanes on each gel. The DNA was transferred to Zeta Probe membranes (BioRad) by vacuum blotting and hybridized with 32-P labelled probes. The HLA-DR types were identified by two cDNA probes; a DRB probe consisting of a 598 base pair (bp) HIND III/SacI fragment of the cDNA clone pll-ß-l (77) and a DQB probe consisting of a 627 bp AVAI fragment of the cDNA clone pll-ß-l (78).

Statistical analyses

Variables were tested for normality. The distribution of Lp(a) was highly skewed which necessitated either use of non-parametric tests or transformation to a norm al distribution. Logarithms of Lp(a) were approximately normally d istribu ted and hence used in param etric significance testing. TG and fib rinogen w ere also transform ed to logarithm s to p roduce norm al distributions.

Independent t-tests (IV) or the Mann-Whitney U test (I, III, IV) was used for comparison of continuous variables between groups. Analysis of variance was used w hen adjustment for age was required (I, II). Paired t-tests were used to compare repeated determinations of variables in the same individuals (IV, V).

Pearson's correlation coefficients were calculated for normally distributed variables (IV, V). M ultiple linear regression was used to study relations between variables (I, II, IV, V). A stepwise multiple linear regression model was used to assess the independent ability of different variables to predict Lp(a) level (I, II) and coronary obstruction score (II).

A discriminant analysis was carried out to assess the independent ability of lipid variables and BMI to predict patient or referent status (II),

Odds ratios for different strata of patients and referents were calculated w ith corresponding 95% confidence intervals (II).

The Chi-Square test w ith Yates' correction was used for com paring proportions (II, III).

Box-plots were constructed to illustrate serial distributions of Lp(a) (79) (IV).A tw o-tailed p-value of < 0.05 was chosen as the level of statistical

significance.Statistical calculations were perform ed in a com puterized statistical

program, SYSTAT® version 5.

17

Ethical aspects

Informed consent was obtained from all patients and referents who entered the studies. All study protocols were approved by the Hum an Ethics Committee of the Medical Faculty of Umeå University.

RESULTS

Paper I

Lp(a) level did not differ between subjects who had fasted for 12 hours and those who had not. The distribution of Lp(a) was as expected highly skewed and ranged from <10 mg/1 to 1291 mg/1 (Figure 3). Median Lp(a) in 745 men was 103 mg/1 and in 782 women 109 mg/1. An Lp(a) level exceeding 300 mg/1 was detected in 18.4% of men and in 21.7% of women, and an Lp(a) level exceeding 480 m g/1 in 7.8% of men and in 9.8% of women. No significant difference in Lp(a) between the sexes was found. A positive relation between Lp(a) and age was found in both men and women. Lp(a) increased almost continuously by increasing age in men, whereas in wom en a more distinct increase was observed in the oldest age group, 55-64 years (Table 1, I). Adjustment for different age was systematically performed in further analyses. Sexes were usually analysed separately. When numbers were small both sexes were analysed in common.

In men, an inverse relation between Lp(a) and waist-to-hip ratio was detected (standardized regression coefficient, ß = -0.086, p = 0.028), and in women, a positive relation between Lp(a) and fibrinogen (ß = 0.096, p = 0.011). Total cholesterol and LDL-cholesterol were significantly related to Lp(a) in both sexes, these relations however disappeared after adjustment for the content of cholesterol in Lp(a). Female smokers had significantly higher Lp(a) than non- smokers (median Lp(a) 130 and 100 mg/1, respectively, p = 0.017).

Individuals w ith a first-grade relative deceased before the age of 65 because of AMI had significantly higher Lp(a) than the rest of the study population (m edian Lp(a) 132 and 101 m g/1, respectively, p < 0.05). There was no difference in Lp(a) between individuals with a similar heredity for early death because of stroke and not. In the whole study population 29 individuals (1.9%) had clinical symptoms typical of angina pectoris, 21 individuals (1.4%) had a

18

history of previous AMI and 16 individuals (1.0%) a history of previous stroke. These groups, analysed separately or in common, did not have significantly different Lp(a) as compared with the rest of the population.

0 . 4 -

CO ° * 3 -c0

1 0 . 2 .001

0 . 1 -

0 300 600 900 1200 1500

Lipoprotein(a), mg/ml

Figure 3. Distribution of Lp(a) in 1527 individuals participating in the Northern Sweden MONICA Study.

A trend for lower Lp(a) in 45 diabetics as compared with 766 non-diabetics, who had been evaluated w ith an oral glucose tolerance test, was observed (m edian Lp(a) 73 and 106 mg/1, respectively, p = 0.051). The majority of diabetics had NIDDM (29/45, 64.4%). Lp(a) levels were similar in individuals w ith an impaired as compared with a normal glucose tolerance.

Lp(a) was non-significantly increased in hypertensive as com pared w ith normotensive subjects when both sexes were analysed together (median Lp(a) 115 and 102 mg/1, respectively, p = 0.051). Significance was not obtained w hen sexes were analysed separately and no significant relations between SBP or DBP and Lp(a) were found.

Of the whole study population 557/1527 (36.5%) were on treatm ent w ith some kind of drug (vitamin and mineral supplem ents and sex horm ones included). The most common drugs were aspirin, other analgesics or NSAID, beta blockers and diuretics. Subjects on diuretics (n - 50) had significantly

19

higher Lp(a) than the rest of the study population (median Lp(a) 176 and 103 m g/1, respectively, p = 0.005). In a subgroup analysis on hypertensives only, individuals on diuretics (n = 43) had significantly higher Lp(a) than those on other antihypertensive agents (p= 0.023). In a stepw ise m ultiple linear regression analysis with Lp(a) as the dependent variable use of diuretics was, however, only of borderline significance (p = 0.093), whereas sex and SBP were significant predictors (Table 4 ,1). As previously stated high Lp(a) levels were particularly common in the oldest age group of women. Of all hypertensives treated w ith diuretics 23/43 (53.5%) were postmenopausal women.

H ealthy individuals (n = 948), defined as subjects w ithout any kind of medication, w ithout diabetes, angina pectoris, previous AMI or stroke, had the same Lp(a) level as the rest of the population. In this subgroup of healthy individuals a significant positive relation between Lp(a) and age remained in both sexes. Cardiovascularly healthy individuals (n = 1297), defined as subjects w ithou t treatm ent w ith cardiovascular drugs including antihypertensive agents, w ithout diabetes, angina pectoris, previous AMI or stroke, had m edian Lp(a) 102 mg/1 as compared w ith 125 mg/1 in the rest of the population (p = 0.090).

In a stepwise multiple linear regression analysis on m en significant and independent predictors of Lp(a) were waist-to-hip ratio, age and fibrinogen (Table 5, I). In w om en m enopausal status, smoking and fibrinogen were independent predictors, whereas use of diuretics was of borderline significance (Table 6 ,1).

Paper II

Both m ale and female patients w ith significant CAD at elective coronary angiography had significantly higher Lp(a) than patients w ithou t any significant stenosis. Patients w ith documented CAD who were investigated w ith acute angiography had higher Lp(a) levels than those investigated electively (Table 1,2).

20

Table 1. Lp(a) levels in 1216 male patients, 31 - 80 years old, investigated withcoronary angiography

Significant CAD atelectiveangiography

Significant CAD at acute angiography

No significant CAD at angiography

Num ber ofsubjects 996 (81.9%) 170 (14.0%) 50 (4.1%)

Age mean ± SD 59.2 ± 8.1 60.318.5 54.319.5 ***

Lp(a) (mg/1) m ean ± SD 2381250 3341302 *** 1551172 *m edian 142 220 106

Percent of patients w ith Lp(a) > 300 m g /l

28.7 44.7 *** 16.0

Percent of patients w ith Lp(a) > 480 m g /l

15.9 29 4 *** 6.0

* p < 0.05, *** p < 0.001Comparisons were made between patients with CAD at acute angiography or patients w ith no significant CAD and patients with documented CAD at elective angiography.

21

Table 2. Lp(a) levels in 355 female patients, 23 - 77 years old, investigated withcoronary angiography

Significant CAD atelectiveangiography

Significant CAD at acute angiography

No significant CAD at angiography

Num ber of subjects 222 (62.5%) 63 (17.8%) 70 (19.7%)

Age mean ± SD 60.0 ± 8.0 61.5 ±8.4 52.2 ± 9.8 ***

Lp(a) (mg/1) m ean ± SD 309 ±316 354 ±304 186 ± 228 *m edian 192 236 90

Percent of patients w ith 36.0 42.9 20.0 *Lp(a) > 300 m g /l

Percent of patients with 24.3 30.2 11.4*Lp(a) > 480 m g /l

* p < 0.05, *** p < 0.001Comparisons were made between patients with CAD at acute angiography or patients with no significant CAD and patients with documented CAD at elective angiography.

22

The further reported analyses relate only to patients with significant CAD at elective angiography. Female patients had significantly higher Lp(a) than male patients (p = 0.004), but in patients below 55 years of age a trend for higher Lp(a) in men was observed (Table 3, II). Lp(a) showed a weak, but significant, decreasing trend w ith increasing age in male patients (p = 0.046), whereas in wom en a non-significant increasing trend with increasing age was found (p = 0.083).

In male patients Lp(a) was inversely related to BMI and TG and positively related to HDL and SR. In female patients Lp(a) was only inversely related to TG. All relations were rather modest with standardized regression coefficients below 0.16 (Table 4, II). No relations between Lp(a) and angiographic scores were found in either sex. Male patients who were current smokers had slightly lower Lp(a) than non-smokers. Lp(a) level was considerably higher in male patients w ith peripheral artery disease than in patients without. A similar trend was seen in female patients. Treatment with different kind of drugs did not seem to influence Lp(a) level (Table 5, II).

A stepwise multiple linear regression analysis on male patients showed that significant independent predictors of Lp(a) level were concomitant peripheral artery disease, HDL, age, current smoking, sedim entation rate and BMI in descending order. The model was highly significant, but explained only 5.8% of the variance of Lp(a) (Table 6, II). In female patients the only significant predictor of Lp(a) level was presence of diabetes. C urrent sm oking and peripheral artery disease were predictors of borderline significance. Age was the strongest independent predictor of the myocardial coronary obstruction score in both sexes (Table 7, E).

Cardiovascularly healthy individuals from the Northern Sweden MONICA Study were used as referents in this study. Hypertensive, bu t otherw ise healthy, individuals were also included as referents. Since the referents all were below 65 years of age, only patients below that age were included in the further analyses. Patients with documented CAD had in both sexes significantly higher Lp(a) than the referents (Table 3, 4). When different age groups were analysed separately a significant difference in Lp(a) was only obtained in men below 55 years of age, but the proportion of subjects with an Lp(a) level exceeding 480 mg/1 was increased also in patients over 55 years. A statistically significant difference in Lp(a) was obtained only in women older than 54 years, not in the younger women.

23

Odds ratios were calculated to study the effect of different Lp(a) levels on the risk for CAD (Table 10, II). In men below 55 years of age a significant increased risk by each Lp(a) interval was observed. In men 55-64 years old, and in wom en an increased risk for CAD was only seen when Lp(a) exceeded 480 m g/l.

In a discrim inant analysis Lp(a) emerged as a significant, independent predictor of CAD in both men and wom en (Table 11, II). The strongest predictors were TG and HDL-cholesterol.

Table 3. Lp(a) levels in male patients with documented CAD at elective angiography and in referents from the Northern Sweden MONICA Study

32-64 vears < 55 years > 55 years

Patients Controls Patients Controls Patients Controls

Num ber ofsubjects 718 579 270 414 448 165

Lp(a)(m g/l)

mean 240 172 *** 261 167 *** 227 183m edian 141 105 170 100 123 116

Percent ofpatients w ith Lp(a) > 300 m g /l

29.2 18.8 *** 33.0 17.9 *** 27.0 21.2

Percent ofpatients w ith Lp(a) > 480 m g /l

16.4 7.8 *** 18.1 8.2 *** 15.4 6.7 **

** p < 0.01, *** p < 0.001

24

Table 4. Lp(a) levels in female patients with documented CAD at elective angiography and in referents from the Northern Sweden MONICA Study

38-64 years < 55 years > 55._y.ears

Patients Controls Patients Controls Patients Controls

Number ofsubjects 153 525 55 334 98 191

Lp(a)(mg/1)

mean 298 202 * 253 174 323 250*m edian 172 116 145 102 204 153

Percent ofpatients w ith Lp(a) > 300 mg/1

33.3 23.2* 29.1 19.8 35.7 29.3

Percent ofpatients w ith Lp(a) > 480 mg/1

23.5 12.2 ** 16.4 8.7 27.6 18.3

* p < 0.05, **p < 0.01

25

Paper III

Characteristics of 30 male and 30 female patients with early CAD participating in this study are given in Table 1 (III). Male patients were as selected significantly younger than female patients. Due to technical reasons results of HLA-DR typing was missing in one male and two female patients and in four controls. HLA-DR frequencies in patients and controls are given in Table 4 (III). HLA-DR 15 specificity was less frequent in male patients than in male controls (p = 0.039). The most frequent genotypes in male patients were 4, 13a and 17, bu t in female patients and controls of both sexes they were 1, 4 and 15. HLA- DR 13 or 17 was significantly more frequent in male patients than in male controls (p = 0.012) and HLA-DR 1, 4 or 15 less frequent (p = 0.051). In female patients a slight trend towards an increase in HLA-DR 13 was found, bu t the frequencies of HLA-DR 17 were similar in patients and controls. Four male patients were diabetics. A separate analysis performed with diabetics excluded showed that HLA-DR type 13 or 17 still was significantly more frequent in male patients than in controls (p = 0.018). HLA-DR 13 or 17 was common in male patients w ith high or medium Lp(a) levels (Table 5, III). M edian Lp(a) in 15 male patients with HLA-DR 13 or 17 was 241 mg/1 and in 14 male patients with other genotypes 166 mg/1 (p = 0.167).

Paper IV

A significant positive relation was found between the relative values of Lp(a) and the time in days after the AMI event (p = 0.001). The increase in Lp(a) during the first week after AMI was, however, weak as compared w ith the classical acute-phase proteins; haptoglobin, orosomucoid and ai-antitrypsin (Figure 2, IV). The individual changes in the acute-phase proteins were highly correlated to each other and to the extent of myocardial damage estimated as m aximum values of CKMB. In contrast, Lp(a) did not correlate to changes in the acute-phase proteins or the CKMB values (Table III, in paper IV). The individual Lp(a) reactions were very heterogeneous. A linear regression model was used for each patient to characterize the Lp(a) reaction. An increasing Lp(a) by time was found in 10/32 (31.2%) patients, a decreasing Lp(a) in 4/32 (12.5%) and an indecisive reaction in 18/32 (56.3%). Admission values of Lp(a) did not differ between patients who showed an increase in Lp(a) and those who

26

did not. No relations between clinical course or drug intake and the change in Lp(a) could be detected either. Four patients were diabetics, two of them reacted w ith a decreasing Lp(a) level. Since glycémie control and possibly also infections might influence Lp(a) level (82), a separate analysis was perform ed w ith diabetics and patients w ith acute infections excluded. In this group consisting of 25 AMI patients no correlations could still be found between change in Lp(a) and change in the acute-phase proteins, and the increase in Lp(a) was still modest. A subset of 21 patients from the original study population was sam pled daily for lipids and plasma albumin. M ean total cholesterol decreased by 12.5% and mean HDL-cholesterol by 21.1% from adm ission to day 6 in these patients. The change in Lp(a) correlated significantly w ith change in total cholesterol and Lp(a)-adjusted LDL- cholesterol (Table V, in paper IV). Plasma albumin decreased only slightly and non-significantly during the infarct course. A positive correlation between change in Lp(a) and change in plasma albumin was, however, found.

Paper V

Characteristics of study patients are given in Table 1 (V). Ischemic heart disease had been diagnosed in 19/36 (52.8%) and hypertension in 10/36 (27.8%) of the study subjects. M edian Lp(a) at baseline was 447 mg/1. Median Lp(a) during treatm ent w ith placebo was 359 m g/I and during treatment w ith simvastatin 464 m g/1 (p = 0.103). Two outliers in respect of Lp(a) reaction were detected (Figure 1, V). They had very high baseline Lp(a) values, 1370 and 1470 m g/m l, respectively, and both reacted with a strong increase in Lp(a) during active treatment. To weaken the effect of these outliers the values of Lp(a) change w ere log transform ed prior to analysis w ith an approxim ate norm al distribution afterwards. In a multiple linear regression analysis, change in Lp(a) was significantly related to the baseline value of Lp(a), the baseline value of apoB, and the change in TG (Table 4, V). Patients with high baseline Lp(a) were inclined to develop further increased Lp(a).

27

DISCUSSION

Lp(a) in the general population

Previous population studies have shown that the distribution of Lp(a) in whites and Orientals is highly skewed, whereas in blacks it is more bell-shaped and norm al (24, 81, 82). Considerable differences between ethnic groups in Lp(a) levels and also in apo(a) phenotypes have been described in several studies (25, 83-86). We found in a randomly selected sample from the general population in northern Sweden, as expected, a highly positively skewed Lp(a) distribution. Comparisons of absolute Lp(a) levels obtained in different population studies are difficult since m ethods of Lp(a) analyses vary, and standardization of assays between different laboratories still is lacking.

We observed no significant difference in Lp(a) level between the sexes, which is in accordance w ith some previous reports (87, 88). However, we observed in the youngest (25-34 years) and in the oldest (55-64 years) age group a trend for higher Lp(a) in women than in men. In the Bogalusa Heart Study on children 8-17 years old a slightly higher Lp(a) level was detected in girls (82). Several authors have reported on increased Lp(a) levels in w om en after m enopause (88, 89). In the recently presented ARIC Study, which comprises a total of 14 524 subjects of both sexes, 45-64 years old, Lp(a) was found to be slightly higher in women than in men (90).

We observed an increasing Lp(a) level by increasing age in both men and women. In wom en a distinct increase in Lp(a) in the oldest age group, 55-64 years, was detected, whereas in men a more continuously increasing Lp(a) level by increasing age was found. Sundell et al. also found a positive relation betw een Lp(a) and age in a north Swedish population consisting of 260 individuals 30-60 years old (91). They did not analyse the sexes separately. Rhoads et al found an increased frequency of a "sinking" pre-beta lipoprotein band w ith increasing age in 1854 men of Japanese ancestry living in Honolulu (92). In the Framingham Offspring Study a trend for an increase in plasma Lp(a) between the ages of 20 and 59 years was observed (93), and also in the ARIC Study increasing age was associated with higher Lp(a) levels in both sexes (90). In two population studies on employees no relation could, however, be found betw een Lp(a) and age in men (88, 89). The relation we found between Lp(a) and age was modest. The lack of an association in some studies (88, 89, 94) may be explained by small sample sizes. Differences in selection of

28

study populations should also be considered. An acute-phase property of Lp(a) has been proposed (95). An increased frequency of diseases w ith an acute- phase reaction in elderly people could possibly explain the relation we found between Lp(a) and age. However, when we analysed only healthy individuals a significant relation between Lp(a) and age remained in both sexes.

M enopausal status was the most important independent predictor of Lp(a) level in women. Lp(a) is probably related to sex hormones in both m en and wom en (96-99). In the present study 12.8% of postmenopausal women were on horm one replacem ent therapy. A trend for lower Lp(a) in these w om en as com pared w ith postm enopausal women not on HRT was observed. Dahlén found decreasing Lp(a) levels in postmenopausal women, who treated w ith HRT (100). In the ARIC Study use of HRT was associated w ith a significantly decreased Lp(a) level (90). We could not detect any influence of horm onal con tracep tives on Lp(a). The num ber of w om en tak ing horm onal contraceptives was low. Relations between Lp(a) and other hormones; thyroid hormones and growth hormone, have been reported (101-103).

We found in both sexes a weak, but independent and significant, relation between Lp(a) and fibrinogen, which is in accordance w ith results from the PROCAM study on employees (89). We did not analyse relations between Lp(a) and other fibrinolytic variables. Most previous studies have not detected any relations between Lp(a) and the fibrinolytic variables PAI and tPA (91, 104, 105). However, in the PROCAM study an inverse relation between Lp(a) and PAI-1 was detected in multiple regression analysis.

Lp(a) has been shown to be mainly independent of other cardiovascular risk factors. In several studies no associations between Lp(a) and diet (106, 107), weight or body-fat distribution (84, 91, 108) were observed. A declining Lp(a) level in m en and women during weight reduction has been reported in one study (109), bu t initial Lp(a) values in that study did not correlate to overw eight or body fat distribution. Some authors have, however, found discrete relations between Lp(a) and body weight (88, 89, 92). We did find in men a discrete inverse relation between Lp(a) and waist-to-hip ratio. No similar relation was observed in women. We could not detect any relation between physical activity and Lp(a). Hellsten et al found declining Lp(a) levels in 16 w ell-trained men, who participated in a cross-country skiing tour in the Swedish mountains at winter time (110).

Most studies report no effect of smoking on Lp(a) (7, 84, 88, 89, 111). We observed no relation between smoking and Lp(a) in men, bu t in w om en

29

smoking was associated with a higher Lp(a) level. The increased Lp(a) level in female smokers may be explained by the anti-estrogen effect induced by smoking (112,113).

Discrete correlations between SBP and Lp(a) were observed in the two previously m entioned studies on employees (88, 89). We could not detect any significant relations between blood pressure levels and Lp(a). A trend for higher Lp(a) in hypertensive as compared w ith non-hypertensive individuals was, however, found. Subjects on diuretics had particularly high Lp(a). The reason for increased Lp(a) in individuals on diuretics is difficult to establish. The main indication for treatment with diuretics was hypertension. More than half of treated subjects were postmenopausal women. In a multivariate analysis on hypertensives only, sex and systolic BP were significant predictors of Lp(a), w hereas use of diuretics only was of borderline significance. There is very sparse inform ation in the literature concerning the effect of different cardiovascular drugs on Lp(a). Donders et al observed increased Lp(a) levels in 54 patients w ith essential hypertension and unsatisfactory blood pressure level (114).

Davies et a l found increased Lp(a) in subjects w ith an im paired glucose tolerance (115). We could not confirm that finding. The effect of diabetes on Lp(a) is somewhat controversial. The main evidence points towards normal Lp(a) levels in diabetics with good metabolic control and without nephropathy (116-121). We found slightly (non-significantly) lower Lp(a) in diabetics as compared with non-diabetics.

A small proportion of the general population in our study had clinical overt cardiovascular disease. These individuals did not differ significantly in Lp(a) level as compared with the rest of the study population. A strong heredity for early death because of AMI was, however, associated with a high Lp(a), which is in accordance w ith the concept that Lp(a) is strongly determined by genetic factors. In m ultiple regression analysis we also found that only very small proportions of Lp(a) variance could be explained by external or life-style factors.

Lp(a) in coronary artery disease

An increased Lp(a) level was found in patients w ith confirmed CAD at angiography both in comparison with patients without any significant CAD at

30

angiography and in comparison w ith healthy referents from the N orthern Sweden MONICA Study. These findings are in agreement w ith results from m ost previous angiographic studies (9, 122-125). However, Niem inen et a l could not detect any difference in Lp(a) level between 111 Finnish patients w ith angiographic CAD and 46 patients without CAD at angiography or 96 healthy controls (126). Their study included both males and females and also patients evaluated w ith angiography because of valvular heart disease. A different selection of study subjects and a rather small sample size may, at least in part, explain the divergent results obtained in this study. Ethnical differences in predictors of angiographic CAD may also be of importance.

We excluded from our study patients on lipid-lowering drugs and patients who previously had been treated with coronary bypass grafting or angioplasty. This probably im plies that individuals w ith the m ost pronounced lip id disturbances were left out. Treatment with lipid-lowering agents was, however, relatively uncom m on during the study period. In an Italian study on 1200 hyperlipidémie patients, an increased Lp(a) level was observed in subjects on lipid-lowering drugs, especially in patients receiving multiple lipid-lowering agents (127). We did find in our original material a trend for increased Lp(a) in patients who were excluded from the study because of treatm ent w ith lipid- lowering agents. An Lp(a) level exceeding 300 mg/1 was significantly more common in these patients (p < 0.05) than in patients w ith proven CAD at elective angiography who were not on lipid-lowering drugs. HMG Co A reductase inhibitors were the most common lipid-lowering drugs used in our patients. Patients excluded from our study because of previous coronary bypass surgery did not differ in Lp(a) level as com pared w ith included patients.

Patients who did not show any significant CAD at angiography constituted a heterogeneous group. A majority of them were females. It is well established that CAD is more difficult to establish from anamnesis and exercise test in females as com pared w ith males. Common diagnoses in patients w ithout significant CAD at angiography were hypertension, probable spasm angina, pains of non-coronary origin and suspected syndrome X. Some patients had discrete coronary sclerosis, but no significant stenoses.

The referents selected from the Northern Sweden MONICA Study were judged to be cardiovascularly healthy because of no history of a previous AMI or stroke, no current symptoms typical of angina pectoris and no treatm ent w ith lipid-lowering drugs. They were, however, not evaluated w ith either ECG,

31

exercise test or coronary angiography, which probably implies that some cases with asymptomatic CAD were included as referents. Asymptomatic CAD may be encountered as frequently as in 2.5 to 10 percent of middle-aged m en (128, 129). We excluded diabetics as referents since an increased prevalence of asymptomatic CAD has been reported in diabetics (130-132).

Patients w ith confirmed CAD at acute angiography had significantly higher Lp(a) than patients going through an elective evaluation. This may be due to increasing Lp(a) levels after AMI and possibly also during unstable ischemic events (133). An increased Lp(a) level may also predispose to acute, unstable courses.

Lp(a) levels in different age groups of patients w ith proven CAD at elective angiography differed between male and females. H igh Lp(a) levels were particularly common in young male patients (< 55 years) and in older female patients (> 55 years). Dahlén et ah found significantly elevated Lp(a) only in m en younger than 55 years, which is in agreement w ith our results (11). The incidence of CAD is low in women below the age of 55 years. It increases however rapidly thereafter. The pathogenic mechanisms of CAD in wom en before and after menopause may differ, probably because of alterations of sex hormones. Lp(a) is known to increase after menopause and, according to our results, seems to constitute a stronger predictor for CAD in women older than 55 years than in younger women.

We found, in agreement w ith previous reports (12, 124), that Lp(a) was m ainly unrela ted to other established predictors of CAD. Only small proportions of Lp(a) variance could be explained by other known factors in m ultip le regression analyses. However, in m en Lp(a) was w eakly, but independently, related to HDL-cholesterol and inversely to BMI. H igh Lp(a) levels accordingly seem to be more common in male patients w ithout derangem ents typical of the metabolic syndrome. In bivariate analysis a significant inverse relation between Lp(a) and TG was detected in both sexes. In multiple regression analysis the relation, however, did not remain. This was probably due to the well-known inverse relation between HDL and TG. In the Italian study on hyperlipidaemic patients a reduced Lp(a) level was observed in patients w ith hypertriglyceridemia as compared with other hyperlipidemias, and an inverse relation between TG and Lp(a) was detected in both sexes (127). These results are in accordance with data from our CAD patients. We could not detect any significant relation between TG and Lp(a) in the general population. The. observed inverse relation between TG and Lp(a) is probably m ost

32

prom inent in populations w here hypertriglyceridem ia is common. The mechanisms behind the association between Lp(a) and TG are not fully known. Apo(a) has been detected in TG-rich lipoprotein fractions postprandially and after fat ingestions (134-138). The association of apo(a) with TG-rich lipoprotein particles may influence the metabolism of Lp(a) and also affect the total serum concentration. We could not confirm the finding by Arm strong et a l that an increased LDL concentration markedly increased the risk of CAD due to elevated Lp(a) (12). Current smoking was associated with a slightly lower Lp(a) value in our patients, and in female patients also diabetes was associated w ith a lower Lp(a). These findings probably can be explained by a considerably increased risk for CAD in smokers and in diabetics even if the Lp(a) level is low or moderate. Werba et al found in hyperlipidémie subjects significantly lower Lp(a) in smokers than in non-smokers (127). Winocour et al have reported on low er Lp(a) in diabetic versus non-diabetic individuals w ith clinically diagnosed CAD (139).

Although Lp(a) levels are generally considered to be very stable through life, a reactive increase in Lp(a) may occur in diseases w ith an inflammatory com ponent (66, 95). In analogy to these findings Lp(a) could be reactively elevated also in atherosclerotic disease. Sandholzer et al have show n that apo(a) isoforms differ between CAD patients and controls in several different ethnic populations (140). In a study on individuals with a heredity for NIDDM, similar Lp(a) levels were found in diabetics with previous AMI and in their healthy relatives (121). These findings support the contention that Lp(a) is an inherited factor related to CAD. Even if Lp(a) is to some extent reactively increased in CAD patients it may nevertheless exert a harmful effect promoting atherosclerosis.

We found in elective male patients w ith established CAD a weak, but significant and independent, relation between Lp(a) and SR. Lp(a) is a huge lipoprotein particle, which could possibly in high concentrations lead to an increased SR. We could not detect any relations between Lp(a) and SR or fibrinogen in unstable male CAD patients, bu t in female CAD patients investigated acutely a comparatively strong relation between fibrinogen and Lp(a) was observed.

Lp(a) was not related to the angiographic scores. We used a coronary obstruction score which takes into consideration both the degree of obstruction and the m yocardial mass distal of the stenosis. The majority of our CAD patients had an extensive coronary sclerosis. Attempts to further grade the

33

disease in this highly selected population may be rather meaningless. We included in the analysis only individuals w ith significant CAD. In some prev ious studies both patients w ith and w ithout significant CAD at angiography were included in the analysis. Age was the strongest predictor of the coronary obstruction score in both sexes in our study. Hearn et a l found similar Lp(a) levels in subjects with different numbers of diseased coronary vessels (123).

A discriminant analysis was carried out to assess the independent ability of lipid variables to discriminate between CAD patients and healthy controls. A djustm ent for the effect of some possible confounders was perform ed sim ultaneously. Lp(a) constituted an independent discrim inator of CAD in both sexes of approxim ately the same strength as Lp(a)-adjusted LDL- cholesterol. The most potent discriminators were, however, TG and HDL- cholesterol. These results are in accordance with several previous reports (11, 123,124,141,142).

Pathogenic mechanisms in Lp(a) related atherosclerotic disease

The structure of Lp(a) suggests that it has both atherogenic and thrombogenic p ro p e rtie s . The exact pathogenic m echanism s in L p(a)-associated atherosclerotic disease are, however, not yet fully elucidated.

Several mechanisms through which Lp(a) may inhibit fibrinolysis have been described. In vitro studies have shown that Lp(a) competes w ith plasminogen for binding to endothelial cells and thereby inhibits fibrinolysis on cell surfaces (40-42). Lp(a) also inhibits activation of plasminogen by streptokinase (44) or tPA, and it competes with plasminogen for binding to fibrin (37-39). A relation between Lp(a) and the fibrinolytic capacity in vivo has been much more difficult to assess (105, 125, 143), and Lp(a) does not seem to affect the outcome of thrombolytic therapy (144,145).

Cushing et al detected high concentrations of apo(a) in resected vein grafts from patients undergoing coronary re-bypass surgery, but virtually no apo(a) in normal saphenous veins (35). Apo(a) is probably selectively retained in vein grafts since the apo(a)/apoB ratio was higher in graft tissue than in plasma. Beisiegel et a l perform ed biochemical and immunohistochemical studies of biopsies from fresh hum an arterial walls (146). They found an accumulation of intact apo(a) in the arterial intima, preferentially extracellularly in plaque areas,

34

where it was strongly co-localized with apoB and fibrin. In some cases they also detected apo(a) in foam cells. Krempler et a l showed that Lp(a)-dextran sulphate complexes were incorporated in the same extent as modified LDL in mouse peritoneal macrophages (147). The uptake of Lp(a) into the arterial intima is probably mediated by its strong binding to glycosaminoglycans (148, 149). Dahlén et al showed in an early in vitro study that binding of Lp(a) to glycosaminoglycans was minimal at physiological ionic strength, but increased considerably and exceeded that of LDL after addition of calcium ions (148). Lp(a) is the only lipoprotein which precipitates at physiological ionic strength and calcium ion concentration, a property which should facilitate uptake in m acrophages (150-152). Lp(a) has furtherm ore been show n to b ind to fibronectin (153), fibrin (38) and plasminogen receptors on endothelial cells (42). Lp(a) could thereby inhibit fibrinolysis at the endothelial surface, and both fibrin and Lp(a) would be internalised into the arterial intima.

Com ponents of the immune system are known to be involved in the atherosclerotic process (154). Activated T-lymphocytes, m acrophages and im m unoglobulins have been detected in large num bers in atherosclerotic plaques (155, 156). Several cytokines are also produced in atherosclerotic lesions (157). The scavenger receptor, which is responsible for cholesterol uptake in macrophages, is under cytokine control (158). Oxidized LDL particles and oxidation products are known to be strong antigens and m ay elicit production of autoantibodies (159). The protein part of lipoproteins may also be processed in macrophages and thereafter presented to T-cells in combination w ith HLA-DR antigens on the macrophage surface, which could lead to T-cell activation. Lp(a) has a strong tendency to aggregate, it can also form aggregates w ith LDL, preferably modified LDL (160). These aggregates could be taken up in m acrophages and constitute im portant initiators of the im munological process in the arterial intima.

We postu lated a hypothesis that certain HLA class II genotypes in conjunction w ith Lp(a) could influence T-cell activation and thus initiate or modulate the immune response present in atherosclerosis. The atherosclerotic process itself could also possibly stimulate the synthesis of Lp(a). We found in a pilot study on 30 men with angiographic CAD before the age of 50 an increased frequency of HLA-DR 13 or 17 as compared w ith 30 age m atched healthy controls. A trend for higher Lp(a) values in patients with these genotypes as com pared w ith other genotypes was found. The results of this pilot study should be interpreted w ith some caution, but if confirmed by larger studies

35

they could help to explain why some patients with high Lp(a) and cholesterol levels develop early atherosclerosis while others with comparable lipid levels do not. An over-representation of HLA-DR 13 or 17 could not be found in female patients w ith early CAD. The pathogenic mechanisms in early CAD in wom en may, at least partly, differ from those in men.

Individuals w ith the inherited disease homocystinuria are characterized by very high plasm a concentrations of the amino acid hom ocysteine and a strongly increased risk to develop cardiovascular disease at young age. Some recent studies have shown that individuals w ith a m oderate hyperhom o- cysteinemia also carry an increased risk to encounter cardiovascular disease (161). An interesting relation between Lp(a) and plasma homocysteine has been described by Harpel et al (162). They found that homocysteine, as well as other sulfhydryl-containing amino acids, enhanced the binding of Lp(a) to fibrin. This effect on Lp(a) could at least partly explain the atherogenic properties of homocysteine.

Lp(a) in the acute phase reaction

Plasma proteins which show fast and reversible changes in concentration during conditions characterized by cell injury are called acute-phase reactants. An acute-phase reaction may be encountered both in acute and chronic diseases. Im portant acute-phase proteins are CRP, ai-antitrypsin, orosomucoid, haptoglobin and fibrinogen. They are all synthesised in the liver. Different cytokines probably m ediate the increased synthesis of these proteins in conditions w ith cell injury. Several of the acute-phase proteins contain large am ounts of sialic acid. Lp(a) is also enriched in sialic acid and has been proposed to constitute an acute-phase protein in inflammatory conditions. Rantapää et a l observed increased Lp(a) levels in patients w ith rheum atoid arthritis (66). Maeda et al have previously reported on increasing Lp(a) levels during the course of AMI and after different kinds of surgical operations (95).

We investigated Lp(a) levels and the acute-phase proteins orosomucoid, oq- an titrypsin and haptoglobin in 32 patients treated because of AMI. We excluded patients treated w ith thrombolysis since the thrombolytic agents could possibly influence Lp(a). We found, in accordance w ith Maeda et a l , an increase in the relative Lp(a) levels, but the increase in Lp(a) was weak as com pared w ith increases in the classical acute-phase proteins. The change in

36

Lp(a) varied considerably between different patients. We could not detect any correlations between the change in Lp(a) and the changes in the acute-phase proteins or the extent of myocardial damage estimated as the maximum CKMB value. In 21 patients we also analysed lipids daily and in this subset of patients we found a significant correlation between the individual change in LDL- cholesterol (adjusted for Lp(a)) and the change in Lp(a). Inflam m atory conditions are known to be accompanied by decreases in HDL- and LDL- cholesterol (163). The mechanisms behind these transient effects on lipoproteins are not fully settled. Alterations in the synthesis or catabolism of lipoproteins or an altered intravascular metabolism are some suggested m echanisms. A changed distribution of HDL-lipoprotein particles between the intra- and extravascular space has also been suggested. Johansson et ah distinguished three different patterns of response in plasma proteins during AMI (164). CRP, haptoglobin, fibrinogen, orosomucoid and ai-antitrypsin showed a rapid increase within some days, whereas ceruloplasmin and C3 showed a moderate increase w ith a maximum level during the second week. Albumin, transferrin and im m unoglobulin G reacted w ith a rapid decrease the first week. We followed Lp(a) during six days after the admission to the coronary care unit. The peak in Lp(a) level was seen at day 5. According to other studies Lp(a) rises relatively slowly and reaches a peak 2 weeks after AMI (95, 165). The lack of correlation betw een change in Lp(a) and change in classical acute-phase proteins during the first week after AMI was anyway unexpected. Mechanisms other th an the acute-phase reaction also influencing Lp(a) should be considered. Plasma album in is known to decrease during AMI, probably because of a decreased synthesis in the liver and a shift to the extracellular space. We analysed plasma albumin in the subset of 21 patients. We only detected a weak, non-significant, decline in albumin during the infarct course, bu t we found a significant correlation between the individual change in albumin and the change in Lp(a). Change in albumin also correlated inversely to CKMB; patients with high CKMB tended to have larger declines in albumin. Linear regression analysis was therefore perform ed w ith Lp(a) as the dependent variable and with concomitant adjustment for albumin levels. Lp(a) d id show a slightly stronger increase by time in this analysis, but still no positive relations between the change in Lp(a) and the change in the acute- phase proteins could be found. The clinical course in AMI is of course very varied. Patients w ith small infarctions may develop unstable conditions afterwards and patients w ith large infarctions are inclined to develop heart

37

failure. Complications like infections are not uncommon. Glycémie control and fluid balance may also change. Patients are usually treated with many different drugs which m ight influence Lp(a). We could not detect any influence of some clinical variables on the Lp(a) reaction, but the numbers were of course small. Oshima et a l found increasing Lp(a) in patients treated because of unstable angina (133). Acute ischemic events w ithout cell injury may accordingly influence Lp(a). To conclude, we found that Lp(a) increased moderately during the first week after myocardial infarction, but it did not react in accordance w ith the classical acute-phase proteins. The Lp(a) reaction appeared instead to be related to the change in LDL-cholesterol and the change in plasma albumin.

Treatment of elevated Lp(a) levels

Different diets used to lower LDL-cholesterol do not influence Lp(a) levels significantly (106, 107). A diet high in trans-monounsaturated fatty acids has been reported to increase Lp(a) (166). The majority of lipid-lowering drugs used today do not seem to affect Lp(a). The bile acid séquestrants do not have any effect on Lp(a) (167,168). Most studies report no major effect of fibrates on Lp(a) (169-171). Maggi et ah, however, observed reduced Lp(a) levels in 21 patients treated w ith bezafibrate during six months (172). Nicotinic acid alone or in com bination w ith neom ycin reduces Lp(a) by 38-45% (173, 174). Treatm ent w ith nicotinic acid is, however, restricted because of troublesome adverse reactions.

The HMG CoA reductase inhibitors are potent LDL-lowering agents. They increase the clearance of LDL into the hepatocytes via the LDL-receptor. We investigated the effect of simvastatin on Lp(a) in 36 hypercholesterolemic patients in a randomized, placebo controlled cross-over study. A slight increase in Lp(a) was found during active treatment as compared w ith placebo, but the difference was not significant. Patients with high baseline Lp(a) values had a tendency to react w ith a further increase in Lp(a). Increased Lp(a) levels during treatm ent w ith reductase inhibitors have been reported previously (175, 176). The m ain evidence, however, points towards no major effects of the reductase inhibitors on Lp(a) (177-179). This finding indicates that catabolism via the LDL-receptor does not determine Lp(a) level to any major degree.

Probucol does not affect Lp(a) levels, but it protects Lp(a) against oxidative m odification and inhibits further uptake in m acrophages (180, 181).

38

Antioxidants may, according to these results, be promising agents for treatment of subjects w ith high Lp(a) values. Fish oil with high levels of co3 fatty acids do not seem to exert any major effects on Lp(a) (182-184). N-acetylcysteine is a reducing agent w hich like hom ocysteine has the potential of cleaving disulphide bonds. It was first reported that n-acetylcysteine had a profound decreasing effect on serum Lp(a) (185). In later experiments Scanu et al. found that n-acetylcysteine and other cysteine-containing compounds attenuated the immunoreactivity of Lp(a) (186). Harpel et al. found an increased affinity of Lp(a) to fibrin after treatment with n-acetylcysteine (162). N-acetylcysteine may thus, in a similar way as plasma homocysteine, exert an unfavourable effect on Lp(a) and promote atherosclerosis. Apheresis effectively lowers serum Lp(a) by 40-60% (187, 188). It m ay constitute a suitable treatm ent in fam iliar hypercholesterolemia and other conditions with severe lipid disturbances.

To conclude, the possibilities to treat elevated Lp(a) levels are today rather restricted as the majority of the traditional lipid-low ering agents do not influence Lp(a). No study has either thus far addressed the potential benefit in risk for atherosclerotic disease of lowering Lp(a). The general opinion today is that other risk factors should be treated vigorously when a high Lp(a) value is detected. Use of estrogens in hormonal replacement therapy to postmenopausal wom en is probably advantageous.

CONCLUSIONS

Only sm all proportions of the variation of serum Lp(a) in the general population, as well as in patients with CAD, could be explained by external or life-style factors, which is in accordance with the concept that serum Lp(a) is strongly determ ined by genetic factors. Lp(a) was mainly unrelated to other established cardiovascular risk factors. We did, however, in the general population find a discrete relation between Lp(a) and fibrinogen and a slightly increasing Lp(a) level by increasing age. Menopausal status was the strongest predictor of Lp(a) in women. Women who smoked had an increased Lp(a) level, which may be related to effects on estrogens. In male patients w ith CAD Lp(a) was weakly related to HDL-cholesterol and inversely to BMI. Lp(a) constituted an independent discriminator of angiographic CAD in both m en and women. High Lp(a) levels were particularly often found in young male patients and in elderly female patients with CAD. Lp(a) is probably a stronger

39

risk factor for CAD in women after m enopause than before. Patients w ith confirm ed CAD w ho w ere investigated w ith acute angiography had significantly higher Lp(a) than patients investigated electively.

Several plausible mechanisms to explain the atherothrom botic effect of Lp(a) have been proposed. In a pilot study we found preliminary support of a hypothesis that Lp(a) may be related to HLA-DR antigens and possibly constitutes an initiator or m odulator of the im mune response found in atherosclerotic processes. Lp(a) probably increases in conditions characterized by an acute-phase reaction, but we could not relate the changes in Lp(a) to changes in the classical acute-phase reactants. Lp(a) was not significantly influenced by treatment with the HMG CoA reductase inhibitor simvastatin. A trend for further increased Lp(a) in individuals with high baseline Lp(a) values was, however, found.

Both epidemiological and experimental studies thus support the concept that high Lp(a) levels predispose to atherosclerotic disease. However, to assess a confident causal relationship between Lp(a) and atherosclerosis further prospective and interventional studies are requested as well as basic research on pathogenic mechanisms. Regard should also be taken to the heterogeneity of Lp(a) lipoprotein particles both w ithin the same individual and betw een individuals. The same quantitative result of immunological apo(a) testing may signify different degrees of atherogeneicity in different individuals.

A large-scale screening of Lp(a) in the general population is probably not sensible today. In patients w ith an established increased risk for cardiovascular disease, especially with a hereditary component, Lp(a) should be determined. The m eans to treat elevated Lp(a) levels are rather restricted today, bu t hopefully new effective, interventional therapies will be developed in Lp(a)- associated atherosclerotic disease.

40

ACKNOWLEDGEMENTS

This study was carried out at the Departments of Internal Medicine and Clinical Chemistry, University Hospital, Umeå, Sweden.

I w ish to express my gratitude to all those who have helped me throughout this study.

Professor Per Olov Wester, Head of the Department of Medicine, for support and providing me with necessary working facilities.

Dr Owe Johnson and Dr Gösta Dahlén, my tutors, for their excellent guidance, encouragement and invaluable support throughout this study.

Dr Kjell Asplund, co-author, for stimulating advice and support.

Dr Sture Eriksson for invaluable statistical advice.

Bertil Lindblom, State Institute of Forensic Serology, Linköping University Hospital, co-author, for excellent co-operation.

Dr Karl-Anders Jacobsson, Head of the Cardiology Section, for generous support and encouragement.

All my friends and colleagues at the D epartm ent of Medicine and the Departm ent of Clinical Chemistry, who in various ways helped me during these years.

The nurses Britt Eländer, Kerstin Granberg and Carola Sundholm for invaluable assistance and support.

Karin Andersson for invaluable technical assistance with Lp(a) analyses.

Birgitta Stegmayr for excellent help w ith MONICA data and stim ulating exchange of experiences in research work.

41

Drs M ats Eliasson, Dan Lundblad and Per-Erik Evrin for help w ith the fibrinogen analysis in the MONICA study.

All patients and participants in the studies for their generous co-operation.

My family; my parents Julia and Nils, my brother Rikard, my sister Katarina and my husband Richard for love, patience and support.

This work has been generously supported by the Swedish National Association against Heart and Chest Diseases, the Swedish Medical Research Council (No. B88-19X-8314-01 and B91-19X-8314-04), the Swedish Heart-Lung foundation, King Gustav V's 80th Anniversary Fund, the Kempe Foundations, 'Förenade Liv' M utual Group Life Insurance Company, Stockholm, Sweden, the Joint Committee of the Northern Swedish Health Care Region, Merck Sharp & Dohme and the Faculty of Medicine, University of Umeå.

42

REFERENCES

1. Neufeld HN, Goldbourt U. Coronary heart disease: genetic aspects. Circulation 1983; 67: 943-954.

2. Hopkins PN, Williams RR. Human genetics and coronary heart disease: a public health perspective. Annu Rev Nutr 1989; 9: 303-345.

3. Berg K. Molecular genetics and genetic epidemiology of cardiovasculardisease and diabetes - Introductory remarks: risk factor levels and variability. Ann Med 1992; 24: 343-347.

4. Berg K. A new serum type system in man - the Lp system. Acta PatholMicrobiol Scand 1963; 59:369-382.

5. Dahlén GH, Ericson C. A new lipoprotein pattern in patientsw ith angina pectoris. Scand J Clin Lab Invest 1971; 118: 54.

6 . Berg K, Dahlén GH, Frick MH. Lp(a) lipoprotein and pre-ßi-lipoproteinin patients w ith coronary heart disease. Clin Genet 1974; 6 : 230-235.

7. Dahlén GH, Berg K, Ramberg U-B, Tamm A. Lp(a) lipoprotein and pre- ßl-lipoprotein in young adults. Acta med scand 1974; 196: 327-331.

8 . Dahlén GH, Berg K, Gillnäs T, Ericson C. Lp(a) lipoprotein/pre-ßi- lipoprotein in Swedish middle-aged males and in patients w ith coronary heart disease. Clin Genet 1975; 7: 334-341.

9. Frick MH, Dahlén GH, Berg K, Valle M, Hekali P. Serum lipids in angiographically assessed coronary atherosclerosis. Chest 1978; 73: 62-65.

10. Rhoads GG, Dahlén GH, Berg K, Morton NE, Dannenberg AL. Lp(a) lipoprotein as a risk factor for myocardial infarction. JAMA 1986; 256: 2540-2544.

11. Dahlén GH, Guyton JR, Attar M, Farmer JA, Kautz JA, Gotto AM Jr. Association of levels of lipoprotein Lp(a), plasma lipids, and other lipoproteins with coronary artery disease documented by angiography. Circulation 1986; 74: 758-765.

12. Armstrong VW, Cremer P, Eberle E, et al The association between serum Lp(a) concentrations and angiographically assessed coronary atherosclerosis. Atherosclerosis 1986; 62: 249-257.

13. Költringer P, Jürgens G. A dominant role of lipoprotein(a) in the investigation and evaluation of parameters indicating the development of cervical atherosclerosis. Atherosclerosis 1985; 58:187-198.

43

14. Zenker G, Költringer P, Boné G, Niederkorn K, Pfeiffer K, Jürgens G. Lipoprotein(a) as a strong indicator for cerebrovascular disease. Stroke 1986; 17: 942-945.

15. Asplund K, Olsson T, Viitanen M, Dahlén GH. Lp(a) lipoprotein in patients w ith acute stroke. Cerebrovasc Dis 1991; 1 : 90-96.

16. Norrgård Ö, Ängquist K-A, Dahlén GH. Lp(a) lipoprotein in patients w ith arterial insufficiency of the lower extremities. Eur J Vase Surg 1991; 5: 277-282.

17. Mölgaard J, Klausen IC, Lassvik C, Faergeman O, Gerdes LU, Olsson AG. Significant association between low-molecular-weight apolipoprotein(a) isoforms and intermittent claudication.Arteriosclerosis and Thrombosis 1992; 12: 895-901.

18. Rosengren A, Wilhelmsen L, Eriksson E, Risberg B, Wedel H. Lipoprotein(a) and coronary heart disease: a prospective case-control study in a general population sample of middle aged men. Br Med J 1990; 301:1248-1251.

19. Dahlén GH. Lipoprotein(a) as a risk factor for atherosclerotic diseases. Arctic Medical Research 1988; 47 (Suppl. 1): 458-461.

20. Jauhiainen M, Koskinen P, Ehnholm C, et al Lipoprotein(a) and coronary heart disease risk: a nested case-control study of the Helsinki Heart Study participants. Atherosclerosis 1991; 89: 59-67.

21. Gaubatz JW, Heideman C, Gotto AM Jr, Morrisett JD, Dahlén GH. H um an plasma lipoprotein(a) - structural properties. J Biol Chem 1983; 258: 4582-4589.

22. McLean JW, Tomlinson JE, Kuang W-J, et al cDNA sequence of hum an apolipoprotein(a) is homologous to plasminogen. Nature 1987; 300:132-137.

23. Gaubatz JW, Ghanem KI, Geuvara J Jr, Nava ML, Patsch W, Morrisett JD. Polymorphic forms of hum an apolipoprotein(a): inheritance and relationship of their molecular weights to plasma levels of lipoprotein(a). J Lipid Res 1990; 31: 603-613.

24. Guyton JR, Dahlén GH, Patsch W, Kautz JA, Gotto AM. Relationship of plasma lipoprotein Lp(a) levels to race and to apolipoprotein B. Arteriosclerosisl985; 5: 265-272.

44

25. Marcovina SM, Albers JJ, Jacobs DR Jr, et al Lipoprotein(a) concentrations and apolipoprotein(a) phenotypes in Caucasians and African Americans - the CARDIA Study. Arteriosclerosis and Thrombosis 1993; 13:1037-1045.

26. Utermann G, Menzel HJ, Kraft HG, Duba HC, Kemmler HG, Seitz C. Lp(a) glycoprotein phenotypes - Inheritance and relation to Lp(a)- lipoprotein concentrations in plasma. J Clin Invest. 1987; 80: 458-465.

27. Boerwinkle E, Leffert CC, Lin J, Lackner C, Chiesa G, Hobbs HH. Apolipoprotein(a) gene accounts for greater than 90% of the variation in plasma lipoprotein(a) concentrations. J Clin Invest 1992; 90: 52-60.

28. Koschinsky ML, Beisiegel U, Henne-Bruns D, Eaton DL, Lawn RM. Apolipoprotein(a) size heterogeneity is related to variable number of repeat sequences in its mRNA. Biochemistry 1990; 29: 640-644.

29. Lindahl G, Gersdorf E, Menzel HJ, Seed M, Humphries S, Utermann G. Variation in the size of human apolipoprotein(a) is due to a hypervariable region in the gene. Hum Genet 1990; 84: 563-567.

30. Kamboh MI, Ferrell RE, Kottke BA. Expressed hypervariable polymorphism of apolipoprotein(a). Am J Hum Genet 1991; 49:1063- 1074.

31. Cohen JC, Chiesa G, Hobbs HH. Sequence polymorphisms in the apolipoprotein(a) gene: evidence for dissociation between apolipoprotein(a) size and plasma lipoprotein(a) levels. J Clin Invest 1993; 91:1630-1636.

32. Frank SL, Klisak I, Sparkes RS, et al The apolipoprotein(a) gene resides on hum an chromosome 6q26-27, in close proximity to the homologous gene for plasminogen. Hum Genet 1988; 79: 352-356.

33. Drayna DT, Hegele RA, Hass PE, et al Genetic linkage between lipoprotein(a) phenotype and a DNA polymorphism in the plasminogen gene. Genomics 1988; 3: 230-236.

34. Rath M, Niendorf A, Reblin T, Dietel M, Krebber H-J, Beisiegel U. Detection and quantification of lipoprotein(a) in the arterial wall of 107 coronary bypass patients. Arteriosclerosis 1989; 9: 579-592.

35. Cushing GL, Gaubatz JW, Nava ML, et al Quantitation and localization of apolipoproteins (a) and B in coronary artery bypass vein grafts resected at re-operation. Arteriosclerosis 1989; 9: 593-603.

45

36. Scanu AM, Miles LA, Fless GM, et al Rhesus monkey lipoprotein(a) binds to lysine Sepharose and U937 monocytoid cells less efficiently than hum an lipoprotein(a): evidence for the dominant role of kringle 437.J Clin Invest 1993; 91: 283-291.

37. Harpel PC, Gordon BR, Parker TS. Plasmin catalyzes binding of lipoprotein(a) to immobilized fibrinogen and fibrin. Proc Natl Acad Sci USA 1989; 8 6 : 3847-3851.

38. Loscalzo J, Weinfeld M, Fless GM, Scanu AM. Lipoprotein(a), fibrin binding, and plasminogen activation. Arteriosclerosis 1990; 10: 240-245.

39. Rouy D, Grailhe P, Nigon F, Chapman J, Anglés-Cano E. Lipoprotein(a) impairs generation of plasmin by fibrin-bound tissue-type plasminogen activator: in vitro studies in a plasma milieu. Arteriosclerosis and Thrombosis 1991; 11: 629-638.

40. Gonzalez-Gronow M, Edelberg JM, Pizzo SV. Further characterization of the cellular plasminogen binding site: evidence that plasminogen 2 and lipoprotein(a) compete for the same site. Biochemistry 1989; 28: 2374- 2377.

41. Hajjar KA, Gavish D, Breslow JL, Nachman RL. Lipoprotein(a) m odulation of endothelial cell surface fibrinolysis and its potential role in atherosclerosis. Nature 1989; 339: 303-305.

42. Miles LA, Fless GM, Levin EG, Scanu AM, Plow EF. A potential basis for the thrombotic risks associated with lipoprotein(a). Nature 1989; 339: 301-302.

43. Karâdi I, Kostner GM, Gries A, Nimpf J, Romics L, Malle E. Lipoprotein(a) and plasminogen are immunochemically related. Biochim Biophys Acta 1988; 960: 91-97.

44. Edelberg JM, Gonzalez-Gronow M, Pizzo SV. Lipoprotein(a) inhibits streptokinase-mediated activation of hum an plasminogen. Biochemistry 1989; 28: 2370-2374.

45. Kraft HG, Menzel HJ, Hoppichler F, Vogel W, Utermann G. Changes of genetic apolipoprotein phenotypes caused by liver transplantation. J Clin Invest 1989; 83:137-142.

46. Tomlinson JE, McLean JW, Lawn RM. Rhesus monkey apolipoprotein(a) - sequence, evolution, and sites of synthesis. J Biol Chem 1989; 264: 5957- 5965.

47. Holmqvist L, Hamsten A, Dahlén GLI. Free apolipoprotein(a) in abetalipoproteinaemia. J Intern Med 1989; 225: 285-286.

46

48. Krempler F, Kostner G, Bolzano K, Sandhofer F. Lipoprotein(a) is not a metabolic product of other lipoproteins containing apolipoprotein B. Biochim Biophys Acta 1979; 575: 63-70.

49. Rader DJ, Cain W, Zech LA, Usher D, Brewer HB Jr. Variation in lipoprotein(a) concentrations among individuals with the same apolipoprotein(a) isoform is determined by the rate of lipoprotein(a) production. J Clin Invest 1993; 91:443-447.

50. Oida K, Takai H, Maeda H, et a l Apolipoprotein(a) is present in urine and its excretion is decreased in patients with renal failure. Clin Chem 1992; 38: 2244-2248.

51. Floren C-H, Albers JJ, Bierman EL. Uptake of Lp(a) lipoprotein by cultured fibroblasts. Biochemical and biophysical research communications 1981; 102: 636-639.

52. Maartmann-Moe K, Berg K. Lp(a) lipoprotein enters cultured fibroblasts independently of the plasma membrane low density lipoprotein receptor. Clin Genet 1981; 20: 352-362.

53. Knight BL, Perombelon YFN, Soutar AK, Wade DP, Seed M. Catabolism of lipoprotein(a) in familial hypercholesterolaemic subjects. Ahterosclerosis 1991; 87: 227-237.

54. Snyder ML, Polacek D, Scanu AM, Fless GM. Comparative binding and degradation of lipoprotein(a) and low density lipoprotein by hum an monocyte-derived macrophages. J Biol Chem 1992; 267: 339-346.

55. Hofmann SL, Eaton DL, Brown MS, McConathy WJ, Goldstein JL, Hammer RE. Overexpression of human low density lipoprotein receptors leads to accelerated catabolism of Lp(a) lipoprotein in transgenic mice. J Clin Invest 1990; 85:1542-1547.

56. Rainwater DL, Manis GS, Kushwaha RS. Characterization of an unusual lipoprotein similar to hum an lipoprotein(a) isolated from the baboon, Papio sp. Biochim Biophys Acta 1986; 877: 75-78.

57. Makino K, Abe A, Maeda S, Noma A, Kawade M, Takenaka O. Lipoprotein(a) in nonhuman primates: Presence and characteristics of Lp(a) immunoreactive materials using anti-human Lp(a) serum. Atherosclerosis 1989; 78: 81-85.

58 Laplaud PM, Beaubatie L, Rail SC Jr, Luc G, Saboureau M.Lipoprotein(a) is the major apoB-containing lipoprotein in the plasma of a hibernator, the hedgehog (Erinaceus europaeus). J Lipid Res 1988; 29: 1157-1170.

47

59. Rath M, Pauling L. Hypothesis: Lipoprotein(a) is a surrogate for ascorbate. Proc Natl Acad Sci USA 1990; 87: 6204-6207.

60. Karâdi I, Romics L, Pâlos G, et al. Lp(a) lipoprotein concentration in serum of patients with heavy proteinuria of different origin. Clin Chem 1989; 35: 2121-2123.

61. Haffner SM, Gruber KK, Aldrete G Jr, Morales PA, Stern MP, Tuttle KR. Increased lipoprotein(a) concentrations in chronic renal failure. J Am Soc Nephrol 1992; 3:1156-1162.

62. Dieplinger H, Lackner C, Kronenberg F, et al. Elevated plasma concentrations of lipoprotein(a) in patients with end-stage renal disease are not related to the size polymorphism of apolipoprotein(a). J Clin Invest 1993; 91: 397-401.

63. Cressman MD, Heyka RJ, Paganini EP, O'Neil J, Skibinski CI, Hoff HF. Lipoprotein(a) is an independent risk factor for cardiovascular disease in hemodialysis patients. Circulation 1992; 86 : 475-482.

64. Winocour PH, Bhatnagar D, Ishola M, Arrol S, Durrington PN. Lipoprotein(a) and microvascular disease in type 1 (insulin-dependent) diabetes. Diabetic Med 1991; 8 : 922-927.

65. Kapelrud H, Bangstad H-J, Dahl-Jorgensen K, Berg K, Hansen KF. Serum Lp(a) lipoprotein concentrations in insulin dependent diabetic patients w ith microalbuminuria. Br Med J 1991; 303: 675-678.

66 . Rantapää-Dahlqvist S, Wallberg-Jonsson S, Dahlen GH. Lipoprotein(a), lipids and lipoproteins in patients with rheumatoid arthritis. Ann Rheum Disl991; 50: 366-368.

67. Wright LC, Sullivan DR, Muller M, Dyne M, Tattersall MF1N, Mountford CE. Elevated apolipoprotein(a) levels in cancer patients. Int J Cancer 1989; 43: 241-244.

68 . Feely J, Barry M, Keeling PWN, Weir DG, Cooke T. Lipoprotein(a) in cirrhosis. Br Med J 1992; 304: 545.

69. Marth E, Cazzolato G, Bittolo Bon G, Avogaro P, Kostner GM. Serum concentrations of Lp(a) and other lipoprotein parameters in heavy alcohol consumers. Ann Nutr Metab 1982; 26: 56-62.

70. Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular Survey Methods, second edition. Geneva: WHO; 1982.

71. Wright BM, Dore CF. A random-zero sphygmomanometer. Lancet 1970; 1: 337-338.

48

72. Brandt PWT, Partridge JB, Wattie WJ. Coronary arteriography; m ethod of presentation of the arteriogram report and a scoring system. Clin Radiol 1977; 28: 361-365.

73. Austen WG, Edwards JE, Frye RL, et a l A reporting system on patients evaluated for coronary artery disease. Circulation 1975; 51: 6-40.

74. Grundy SM. Dietary treatment of hyperlipidaemia.In: Hypercholesterolemia and Atherosclerosis. Pathogenesis Prevention, Churchill Livingstone, New York, 1987, pp. 169-193.

75. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, w ithout use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499-502.

76. Lindblom B, Holmlund G. Rapid DNA purification for restriction fragment length polymorphism analysis. Gene Anal Techn 1988; 5: 97- 101.

77. Gustafsson K, Wiman K, Emmoth E, et al Mutations and selection in the generation of class II histocompatibility antigen polymorphism. EMBO Journal 1984; 3:1655-1661.

78. Larhammar D, Gustafsson K, Claesson L, et al Alpha chain of HLA-DR transplantation antigens is a member of the same protein superfamily as the immunoglobulins. Cell 1982; 30:153-161.

79. Velleman PF, Hoaglin DC. Applications, basics and computing of exploratory data analysis. Belmont, CA: Duxbury Press.

80 Bruckert E, Davidoff P, Grimaldi A, et a l Increased serum levels of lipoprotein(a) in diabetes mellitus and their reduction with glycémie control (letter). JAMA 1990; 263: 35-36.

81. Parra H-J, Luyéyé I, Bouramoué C, Demarquilly C, Fruchart J-C. Black- white differences in serum Lp(a) lipoprotein levels. Clin Chim Acta 1987; 167: 27-31.

82. Srinivasan SR, Dahlén GH, Jarpa RA, Webber LS, Berenson GS. Racial (Black-White) differences in serum lipoprotein(a) distribution and its relation to parental myocardial infarction in children: Bogalusa Heart Study. 1991; 84:160-167.

83. Abe A, Noma A. Studies on apolipoprotein(a) phenotypes. Part 1, Phenotype frequencies in a healthy Japanese population. Atherosclerosis 1992; 96:1-8.

49

84. Haffner SM, Gruber KK, et al. Lipoprotein(a) concentrations in Mexican Americans and Non-Hispanic Whites: the San-Antonio Heart Study. Am J Epidemiol 1992; 136:1060-1068.

85. Helmhold M, Bigge J, Muche R, et al Contribution of the apo(a) phenotype to plasma Lp(a) concentrations shows considerable ethnic variation. J Lipid Res 1991; 32:1919-1928.

8 6 . Sandholzer C, Hallman DM, Saha N, et al Effects of the apolipoprotein(a) size polymorphism on the lipoprotein(a) concentration in 7 ethnic groups. Hum Genet 1991; 86 : 607-614.

87. Albers JJ, Wahl P, Hazzard WR. Quantitative genetic studies of the hum an plasma Lp(a) lipoprotein. Biochem Genet 1974; 11: 475-486.

88 . Schriewer H, Assmann G, Sandkamp M. The relationship of lipoprotein(a) (Lp(a)) to risk factors of coronary heart disease: Initial results of the prospective epidemiological study on company employees in Westfalia. J Clin Chem Clin Biochem 1984; 22: 591-596.

89. Heinrich J, Sandkamp M, Kokott R, Schulte H, Assmann G. Relationship of lipoprotein(a) to variables of coagulation and fibrinolysis in a healthy population. Clin Chem 1991; 37:1950-1954.

90. Brown SA, Hutchinson R, Morrisett J, et a l Plasma lipid, lipoprotein cholesterol, and apoprotein distributions in selected US communities - the Atherosclerosis Risk in Communities (ARIC) Study. Arteriosclerosis and Thrombosis 1993; 13:1139-1158.

91. Sundell IB, Nilsson TK, Hallmans G, Hellsten G, Dahlén GH. Interrelationships between plasma levels of plasminogen activator inhibitor, tissue plasminogen activator, lipoprotein(a), and established cardiovascular risk factors in a north Swedish population.Atherosclerosis 1989; 80: 9-16.

92. Rhoads GG, Morton NE, Gulbrandsen CL, Kagan A. Sinking pre-beta lipoprotein and coronary heart disease in Japanese-American men in Hawaii. Am J Epidemiol 1978; 108: 350-356.

93. Jenner JL, Ordovas JM, Lamon-Fava S, et al Effects of age, sex, and menopausal status on plasma lipoprotein(a) levels - the Framingham Offspring Study. Circulation 1993; 87:1135-1141.

94. Albers JJ, Hazzard, WR. Immunochemical quantification of hum an plasma Lp(a) lipoprotein. Lipids 1973; 9:15-26.

50

95. Maeda S, Abe A, Seishima M, Makino K, Noma A, Kawade M. Transient changes of serum lipoprotein(a) as an acute phase protein. Atherosclerosis 1989; 78:145-150.

96. Henriksson P, Angelin B, Berglund L. Hormonal regulation of serum Lp(a) levels: opposite effects after estrogen treatment and orchidectomy in males w ith prostatic carcinoma. J Clin Invest 1992; 89:1166-1171.

97. van der Mooren MJ, Demacker PNM, Thomas CMG, Rolland R. Beneficial effects on serum lipoproteins by 17ß-oestradiol- dydrogesterone therapy in postmenopausal women; a prospective study. Eur J Obstet Gynecol Reprod Biol 1992; 47:153-160.

98. Crook D, Sidhu M, Seed M, O'Donnell M, Stevenson JC. Lipoprotein Lp(a) levels are reduced by danazol, an anabolic steroid. Atherosclerosis 1992; 92: 41-47.

99. Dionyssiou-Asteriou A, Katimertzi M. Endogenous testosterone and serum apolipoprotein(a) levels (letter). Atherosclerosis 1993; 100:123-126.

100. Dahlen GH. Clinical significance of Lp(a) lipoprotein. In: Berg K, Ritterstol N, Ref sum S, Eds. From phenotype to gene in common disorders. Copenhagen: Munksgaard; 1990:163-178.

101. Dahlén GH, Berg K. Further evidence for the existence of genetically determined metabolic differences between Lp(a+) and Lp(a-) individuals. Clin Genet 1976; 9: 357-364.

102. de Bruin TWA, van Barlingen H, van Linde-Sibenius Trip M, van Vuurst de Vries A-RR, Akveld MJ, Erkelens DW. Lipoprotein(a) and apolipoprotein B plasma concentrations in hypothyroid, euthyroid, and hyperthyroid subjects. J Clin Endocrinol Metab 1993; 76:121-126.

103. Eden S, Wiklund O, Oscarsson J, Rosén T, Bengtsson B-Å. Growth hormone treatment of growth hormone-deficient adults results in a m arked increase in Lp(a) and HDL cholesterol concentrations. Arteriosclerosis and Thrombosis 1993; 13: 296-301.

104. Szczeklik A, Radwan J, Kubicka A, et al Plasma fibrinolytic activity in healthy subjects with high and low lipoprotein(a) concentrations.Thromb Res 1992; 66 : 391-395.

105. Halvorsen S, Skjonsberg OH, Berg K, Ruyter R, Godai HC. Does Lp(a) lipoprotein inhibit the fibrinolytic system? Thromb Res 1992; 68 : 223-232.

51

106. Masarei JRL, Rouse IL, Lynch WJ, Robertson K, Vandongen R, Beilin LJ. Effects of a lacto-ovo vegetarian diet on serum concentrations of cholesterol, triglyceride, HDL-C, HDL2-C, HDL3-C, apoprotein-B, and Lp(a). Am J Clin Nutr 1984; 40: 468-479.

107. Mehrabian M, Peter JB, Barnard RJ, Lusis AJ. Dietary regulation of fibrinolytic factors. Atherosclerosis 1990; 84: 25-32.

108. De Pergola G, Giorgino F, Cospite MR, et a l Relation between sex hormones and serum lipoprotein and lipoprotein(a) concentrations in premenopausal obese women. Arteriosclerosis and Thrombosis 1993; 13: 675-679.

109. Sönnichsen AC, Richter WO, Schwandt P. Reduction of lipoprotein(a) by weight loss. Int J Obes 1990; 14: 487-494.

110 Hellsten G, Boman K, Hallmans G, Dahlén GH. Lipids and endurancephysical activity (letter). Atherosclerosis 1989; 75: 93-94.

111. Eliasson M, Lundblad D, Hägg E. Cardiovascular risk factors in young snuff-users and cigarette smokers. J Intern Med 1991; 230: 17-22.

112. Michnovicz JJ, Hershcopf RJ, Naganuma H, Bradlow HL, Fishman J. Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cigarette smoking. N Engl J Med 1986; 315:1305- 1309.

113. Jensen J, Christiansen C, Rödbro P. Cigarette smoking, serum estrogens, and bone loss during hormone-replacement therapy early after menopause. N Engl J Med 1985; 313: 973-975.

114. Donders SHJ, Lustermans FATh, van Wersch JWJ. Coagulation factors and lipid composition of the blood in treated and untreated hypertensive patients. Scand J Clin Lab Invest 1993; 53:179-186.

115. Davies M, Rayman G, Day J. Increased incidence of coronary disease in people w ith impaired glucose tolerance: link with increased lipoprotein(a) concentrations? Br Med J 1992; 304:1610-1611.

116. Csâszâr A, Dieplinger H, Sandholzer C, et al Plasma lipoprotein(a) concentration and phenotypes in diabetes mellitus. Diabetologia 1993; 36: 47-51.

117. Haffner SM, Morales PA, Stern MP, Gruber MK. Lp(a) concentrations in NIDDM. Diabetes 1992; 41:1267-1272.

52

118. Joven J, Vilella E. Serum levels of lipoprotein(a) in patients with well-controlled non-insulin-dependent diabetes mellitus (letter). JAMA 1991; 265:1113-1114.

119. Klausen IC, Schmidt EB, Lervang HH, Gerdes LU, Ditzel J, Faergeman O. Normal lipoprotein(a) concentrations and apolipoprotein(a) isoforms in patients w ith insulin-dependent diabetes mellitus. Eur J Clin Invest 1992; 22: 538-541.

120. Schernthaner G, Kostner GM, Dieplinger H, Prager R, M ühlhauser I. Apolipoproteins (A-I, A-II, B), Lp(a) lipoprotein and lecithin:cholesterol acyltransferase activity in diabetes mellitus. Atherosclerosis 1983; 49: 277-293.

121. Velho G, Erlich D, Turpin E, et ah Lipoprotein(a) in diabetic patients and normoglycemic relatives in familial NIDDM. Diabetes Care 1993; 16: 742- 747.

122. Sorrentino MJ, Vielhauer C, Eisenbart JD, Fless GM, Scanu AM, Feldman T. Plasma lipoprotein(a) concentration and coronary artery disease in black patients compared with white patients. Am J Med 1992; 93: 658- 662.

123. Hearn JA, DeMaio SJ, Roubin GS, Hammarström M, Sgoutas D. Predictive values of lipoprotein(a) and other serum lipoproteins in the angiographic diagnosis of coronary artery disease. Am J Cardiol 1990;66:1176-1180.

124. Genest J, McNamara JR, Ordovas JM, et al. Lipoprotein cholesterol, apolipoprotein A-I and B and lipoprotein(a) abnormalities in men with premature coronary artery disease. JACC 1992; 19: 792-802.

125. Alessi MC, Parra HJ, Joly P, et ah The increased plasma Lp(a): B lipoprotein particle concentration in angina pectoris is not associated w ith hypofibrinolysis. Clin Chim Acta 1990; 188:119-128.

126. Nieminen MS, Mattila KJ, Aalto-Setälä K, et ah Lipoproteins and their genetic variation in subjects with and without angiographically verified coronary artery disease. Arteriosclerosis and Thrombosis 1992; 12: 58-69.

127. Werba JP, Safa O, Gianfranceschi G, Michelagnoli S, Sirtori CR, Franceschini G. Plasma triglycerides and lipoprotein(a): inverse relationship in a hyperlipidémie Italian population. Atherosclerosis 1993; 101: 203-211.

128. Langou RA, Huang EK, Kelley MJ, Cohen LS. Predictive accuracy of coronary artery calcification and abnormal exercise test for coronary artery disease in asymptomatic men. Circulation 1980; 62:1196-1202.

53

129. Erikssen J, Thaulow E. Follow-up of patients with asymptomatic myocardial ischemia. In: Rutishauser W, Roskamm H, Eds. Silent myocardial ischemia. Berlin: Springer-Verlag; 1984, pp. 156-164.

130. Weinrauch L, D'Elia JA, Healy RW, Gleason RE, Christlieb AR, Leland OS Jr. Asymptomatic coronary artery disease: Angiographic assessment of diabetics evaluated for renal transplantation. Circulation 1978; 58: 1184-1189.

131. Abenavoli T, Rubier S, Fisher VJ, Axelrod HI, Zuckerman KP. Exercise testing with myocardial scintigraphy in asymptomatic diabetic males. Circulation 1981; 63: 56-64.

132. Langer A, Freeman MR, Josse RG, Steiner G, Armstrong LW.Detection of silent myocardial ischemia in diabetes mellitus. Am J Cardiol 1991; 67:1073-1078.

133. Oshima S, Uchida K, Yasu T, Uno K, Nonogi H, Haze K. Transient increase of plasma lipoprotein(a) in patients with unstable angina pectoris - does lipoprotein(a) alter fibrinolysis? Arteriosclerosis and Thrombosis 1991; 11:1772-1777.

134. Bersot TP, Innerarity TL, Pitas RE, Rail SC Jr, Weisgraber KH, Mahley RW. Fat feeding in humans induces lipoproteins of density less than 1.006 that are enriched in apolipoprotein(a) and that cause lipid accumulation in macrophages. J Clin Invest 1986; 77: 622-630.

135. Cohn JS, Lam CWK, Sullivan DR, Hensley WJ. Plasma lipoprotein distribution of apolipoprotein(a) in the fed and fasted states. Atherosclerosis 1991; 90: 59-66.

136. Dahlén GH, Ericson C, Furberg C. Variations in pre-beta lipoproteins after a test meal. Acta med scand 1972; (Suppl. 531): 21-24.

137. Pfaffinger D, Schuelke J, Kim C, Fless GM, Scanu AM. Relationship between apo(a) isoforms and Lp(a) density in subjects with different apo(a) phenotype: a study before and after a fatty meal. J Lipid Res 1991; 32: 679-683.

138. Scanu AM, Edelstein C, Fless GM, et al Postprandial lipoprotein(a) response to a single meal containing either saturated or co-3 poly­unsaturated fatty acids in subjects with hypoalphalipoproteinemia. Metabolism 1992; 41:1361-1366.

54

139. Winocour PH, Durrington PN, Bhatnagar D, et al A cross-sectional evaluation of cardiovascular risk factors in coronary heart disease associated with type 1 (insulin-dependent) diabetes mellitus. Diabetes Res Clin Pract 1992; 18:173-184.

140. Sandholzer C, Saha N, Kark JD, et al. Apo(a) isoforms predict risk for coronary heart disease - a study in six populations. Arteriosclerosis and Thrombosis 1992; 12:1214-1226.

141. Barbir M, Wile D, Trayner I, Aber VR, Thompson GR. High prevalence of hypertriglyceridaemia and apolipoprotein abnormalities in coronary artery disease. Br Heart J 1988; 60: 397-403.

142. Tollin C, Ericsson M, Johnson O, et al Decreased removal of triglycerides from the blood - a mechanism for the hypertriglyceridemia in male patients w ith coronary artery disease. Int J Cardiol 1984; 5:185-192.

143. Terres W, Krewitt M, Hamm CW. Effects of lipoprotein(a) on in vitro lysis of whole blood thrombi from healthy volunteers. Thromb Res 1993; 69: 479-487.

144. Armstrong VW, Neubauer C, Schütz E, Tebbe U. Lack of association between raised serum Lp(a) concentration and unsuccessful thrombolysis after acute myocardial infarction (letter). Lancet 1990; 336: 1077.

145. von Hodenberg E, Kreuzer J, Hautmann M, Nordt T, Kübler W, Bode C. Effects of lipoprotein(a) on success rate of thrombolytic therapy in acute myocardial inf aretino. Am J Cardiol 1991; 67:1349-1353.

146. Beisiegel U, Niendorf A, Wolf K, Reblin T, Rath M. Lipoprotein(a) in the arterial wall. Eur Heart J 1990; 11 (Suppl. E): 174-183.

147. Krempier F, Kostner GM, Roscher A, Bolzano K, Sandhofer F. The interaction of hum an apoB-containing lipoproteins with mouse peritoneal macrophages: a comparison of Lp(a) with LDL. J Lipid Res 1984; 25: 283-287.

148. Dahlén GH, Ericson C, Berg K. In vitro studies of the interaction of isolated Lp(a) lipoprotein and other serum lipoproteins with glycosaminoglycans. Clin Genet 1978; 14: 36-42.

149. Bihari-Varga M, Gruber E, Rotheneder M, Zechner R, Kostner GM. Interaction of lipoprotein Lp(a) and low density lipoprotein with glycosaminoglycans from hum an aorta. Arteriosclerosis 1988; 8 : 851-857.

55

150. Dahlén GH, Ericson C, Berg K. In vitro studies of the interaction of calcium ions and other divalent cations with the Lp(a) lipoprotein and other isolated serum lipoproteins. Clin Genet 1978; 14:115-124.

151. Gianturco SH, Bradley WA, Dahlén GH, et al Interaction of Lipoprotein(a) with murine pertoneal macrophages and cultured hum an fibroblasts (abstract). Arteriosclerosis 1983; 3: 500a.

152. Tertov W , Orekhov AN, Sobenin IA, et al. Three types of naturally occurring modified lipoproteins induce intracellular lipid accumulation due to lipoprotein aggregation. Circulation Research 1992; 71: 218-228.

153. Salonen E-M, Jauhiainen M, Zardi L, Vaheri A, Enholm C. Lipoprotein(a) binds to fibronectin and has serine proteinase activity capable of clearing it. EMBO J 1989; 8 : 4035-4040.

154. Libby P, Hansson GK. Biology of disease: Involvement of the immune system in hum an atherogenesis: Current knowledge and unanswered questions. Laboratory Investigation 1991; 64: 5-15.

155. Jonasson L, Holm J, Skalli O, Bondjers G, Hansson GK. Regional accumulations of T cells, macrophages, and smooth muscle cells in the hum an atherosclerotic plaque. Arteriosclerosis 1986; 6:131-138.

156. Hansson GK, Holm J, Jonasson L. Detection of activated T lymphocytes in the hum an atherosclerotic plaque. Am J Pathol 1989; 135:169-175.

157. Moyer CF, Sajuthi D, Tulli H, Williams JK. Synthesis of IL-1 a and IL-1 ß by arterial cells in atherosclerosis. Am J Pathol 1992; 138: 951-960.

158. Geng YJ, Hansson GK. Interferon-gamma inhibits scavenger receptor expression and foam cell formation in hum an monocyte-derived macrophages. J Clin Invest 1992; 89:1322-1330.

159. Parums DV, Brown DL, Mitchinson MJ. Serum antibodies to oxidized low-density lipoprotein and ceroid in chronic periaortitis. Arch Pathol Lab Med 1990; 114: 383-387.

160. Yashiro A, O'Neil J, Hoff HF. Insoluble complex formation of lipoprotein(a) with low density lipoprotein in the presence of calcium ions. J Biol Chem 1993; 268: 4709-4715.

161. Clarke R, Daly L, Robinson K, et al Hyperhomocysteinemia: an independent risk factor for vascular disease. N Engl J Med 1991;324:1149-1155.

162. Harpel PC, Chang VT, Borth W. Homocysteine and other sulfhydryl compounds enhance the binding of lipoprotein(a) to fibrin: A potential

56

biochemical link between thrombosis, atherogenesis, and sulfhydryl compound metabolism. Proc Natl Acad Sci USA 1992; 89:10193-10197.

163. Stubbe I, Gustafson A, Nilsson-Ehle P. Alterations in plasma proteins and lipoproteins in acute myocardial infarction: effects on activation of lipoprotein lipase. Scand J Clin Lab Invest 1982; 42: 437-444.

164. Johansson BG, Kindmark C-O, Trell EY, Wollheim FA. Sequential changes of plasma proteins after myocardial infarction. Scand J Clin Lab Invest 1972; 29 (Suppl. 124): 117-126.

165. Sonoda M, Sakamoto K, Miyauchi T, et al. Changes in serum lipoprotein(a) and C4b-binding protein levels after acute myocardial infarction. Jpn Circ J 1992; 56:1214-1220.

166. Mensink RP, Zock PL, Katan MB, Hornstra G. Effect of dietary cis and trans fatty acids on serum lipoprotein(a) levels in humans. J Lipid Res 1992; 33:1493-1501.

167. Stein E, Kumbla L, Fearn J, Hirth J. Comparison of lovastatin and resin effects on plasma apolipoproteins E and Lp(a), tocopherol, retinol and carotenoids (abstract). 8 th International Symposium Atherosclerosis, Rome 1988; p. 893.

168. Vessby B, Kostner G, Lithell H, Thomis J. Diverging effects of cholestyramine on apolipoprotein B and lipoprotein Lp(a). Atherosclerosis 1982; 44: 61-71.

169. Winocour PH, Durrington PN, Bhatnagar D, et al. Double-blind placebo- controlled study of the effects of bezafibrate on blood lipids, lipoproteins, and fibrinogen in hyperlipidaemic type 1 diabetes mellitus. Diabetic Med 1990; 7: 736-743.

170. Pelegri A, Romero R, Senti M, Nogués X, Pedro-Botet J, Rubiés-Prat J. Effect of bezafibrate on lipoprotein(a) and triglyceride-rich lipoproteins, including intermediate-density lipoproteins, in patients with chronic renal failure receiving haemodialysis. Nephrol Dial Transplant 1992; 7: 623-626.

171. Borresen A-L, Berg K, Dahlén GH, Gillnäs T, Ericson C. The effect of gemfibrozil on hum an serum apolipoproteins and on serum reserve cholesterol binding capacity. Artery 1981; 9: 77-86.

172. Maggi FM, Biasi GM, Catapano AL. Reduction of Lp(a) plasma levels by bezafibrate (letter). Atherosclerosis 1993; 100:127-128.

57

173. Carlson LA, Hamsten A, Asplund A. Pronounced lowering of serum levels of lipoprotein Lp(a) in hyperlipidaemic subjects treated with nicotinic acid. J Intern Med 1989; 226: 271-276.

174. Gurakar A, Hoeg JM, Kostner G, Papadopoulos NM, Brewer HB. Levels of lipoprotein Lp(a) decline with neomycin and niacin treatment. Atherosclerosis 1985; 57: 293-301.

175. Kostner GM, Gavish D, Leopold B, Bolzano K, Weintraub MS, Breslow JL. HMG CoA reductase inhibitors lower LDL cholesterol w ithout reducing Lp(a) levels. Circulation 1989; 80:1313-1319.

176. Jürgens G, Ashy A, Zenker G. Raised serum lipoprotein(a) during treatment w ith lovastatin (letter). Lancet 1989; II : 911-912.

177. Crook D, Sidhu M, Bruce R. Simvastatin and lipoprotein(a) (letter). Lancet 1992; 339: 313-314.

178. Leren TP, Hjermann I, Berg K, Leren P, Foss OP, Viksmoen L. Effects of lovastatin alone and in combination with cholestyramine on serum lipids and apolipoproteins in heterozygotes for familial hypercholesterolemia. Atheroclerosis 1988; 73:135-141.

179. Thiery J, Armstrong VW, Schleef J, Creutzfeldt C, Creutzfeldt W, Seidel D. Serum lipoprotein Lp(a) concentrations are not influenced by an HMG CoA reductase inhibitor. Klin Wochenschr 1988; 66 : 462-463.

180. Naruszewicz M, Selinger E, Dufour R, Davignon J. Probucol protects lipoprotein(a) against oxidative modification. Metabolism 1992; 41: 1225- 1228.

181. Maeda S, Okuno M, Abe A, Noma A. Lack of effect of probucol on serum lipoprotein(a) levels (letter). Atherosclerosis 1989; 79: 267-269.

182. Gries A, Malle E, Wurm H, Kostner GM. Letter to the Editors - Influence of dietary fish oils on plasma Lp(a) levels (letter). Thromb Res 1990; 58: 667-668.

183. Malle E, Sattler W, Prenner E, et al. Effects of dietary fish oil supplementation on platelet aggregability and platelet membrane fluidity in normolipemic subjects with and without high plasma Lp(a) concentrations. Atherosclerosis 1991; 88:193-201.

184. Schmidt EB, Klausen IC, Kristensen SD, Lervang H-H, Faergeman O, Dyerberg J. The effect of n-3 polyunsaturated fatty acid on Lp(a). Clin Chim Acta 1991; 198: 271-278.

58

185. Gavish D, Breslow JL. Lipoprotein(a) reduction by N-acetylcysteine. Lancet 1991; 337: 203-04.

186. Scanu AM, Pfaffinger D, Fless GM, Makino K, Eisenbart J, Hinm an J. Attenuation of immunologic reactivity of lipoprotein(a) by thiols and cysteine-containing compounds: structural implications. Arteriosclerosis and Thrombosis 1992; 12: 424.

187. Armstrong VW, Schleef J, Thiery J, et al Effect of HELP-LDL-apheresis on serum concentrations of human lipoprotein(a): kinetic analysis of the post-treatment return to baseline levels. Eur. }. Clin. Invest. 1989; 19: 235- 240.

188 Keller C. LDL-apheresis: results of longterm treatment and vascular outcome. Atherosclerosis 1991; 86:1-8.

59