Oxidized LDL, Lipoprotein (a) and Other Emergent Risk Factors in Acute Myocardial Infarction (FORTIAM study)

INTRODUCTION

Although the number of patients who suffer an acute myocardial infarction (AMI) without having a previous history of classical risk factors (RF) is low, these patients are an interesting group for studying other determinants of coronary heart disease. A meta-analysis including 14 clinical studies found that 15.4% of women and 19.4% of men with coronary heart disease did not have any of the classical RF.^1,2 Likewise, the IBERICA registry showed that 15% of the patients between 25 and 74 years of age did not have these factors, and furthermore, these patients might prove to have a poorer prognosis.³ Data such as those presented in the INTERHEART study have established the crucial role of classical RF in the development of heart disease, independently of geographical región⁴; even so, the collection of data obtained from the clinical history reported by the patient in the acute phase of AMI is not always sufficiently precise.

In recent years several emergent RF have been proposed as markers of atherosclerosis and the onset of clinical events, such as C-reactive protein (CRP), lipoprotein (a) [Lp(a)], fibrinogen, homocysteine, or Chlamydia pneumoniae, among others, often with controversial results.^5-7 More recently, oxidized low-density lipoprotein (oxLDL), which is directly involved in the formation of atheromatous plaques, has been associated with the process of instability and extent of coronary atherosclerosis.^8-11 However, few clinical data exist that associate oxLDL with the prognosis after an acute coronary syndrome.¹²

The aims of the present study were to determine the prevalence of patients with AMI who did not have classical RF, analyze whether they present a higher prevalence of emergent RF, ascertain whether a specific emergent RF adds prognostic information for patients who suffer an AMI, and if so, establish cut-points.

METHODS

Study Design

FORTIAM (Factores Ocultos de Riesgo Tras un Infarto Agudo de Miocardio) is a multicenter cohort study with a 6-month follow-up. The patients were recruited from the coronary care units of 15 Spanish hospitals.

The study included a total of 1371 patients between the ages of 25 and 74 years who were admitted to a coronary care unit within 24 hours of the onset of symptoms of a first AMI. For inclusion, AMI was defined according to the criteria jointly accepted by the European and Spanish Cardiology Societies (ESC-ACC clinical guidelines).¹³ Only patients who survived the first 24 hours after admission were included. Clinical management was carried out at the discretion of each center. With the objective of specifically determining the prevalence of AMI without classical RF, in 6 centers the inclusion was strictly consecutive (n=949). In the remaining coronary care units inclusion was done randomly to assure a sufficient sample size.

Definition of Classical RF

The following standardized definitions of classical RF were applied: dyslipidemia (a diagnosis or previous lipid-lowering treatment, low-density lipoprotein cholesterol [LDL-C] >160 mg/dL, high-density lipoprotein cholesterol [HDL-C] <35 mg/dL in men or <45 mg/dL in women); diabetes mellitus (a diagnosis or previous treatment, glycemia during admission ≥200 mg/dL, 2 fasting glycemia measurements during hospitalization >125 mg/dL); hypertension (a diagnosis or previous hypotensive drug treatment or systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg, measured 48 h before discharge in a stable hemodynamic state); and active smoker (a mean of 1 or more cigarettes per day); or ex-smoker for less than 1 year. Data were collected within 24 h of admission.

Laboratory Analysis

Blood samples were obtained within 24 h of admission and were centralized in the same laboratory. Plasma fibrinogen levels were measured using the coagulometric Clauss assay (Izasa, Barcelona, Spain). Serum glucose, total cholesterol, HDL-C, and triglycerides were measured using enzymatic methods (Roche, Basle, Switzerland). LDL-C was calculated using the Friedewald formula when triglycerides were <300 mg/dL. Serum Lp(a) was analyzed using immunoturbidimetry. Determination of oxLDL in EDTA-plasma was by sandwich ELISA (murine monoclonal antibody mAB-4E6—capture—and peroxidase-conjugated antibody against oxidized apolipoprotein B in the solid-phase [Mercodia, Uppsala, Sweden]). High sensitivity CRP in serum was determined using immunoturbidimetry (Horiba ABX, Montpellier, France). The serum homocysteine concentration was obtained using fluorescence polarization immunoanalysis (Abbott, IL). Serum IgG anti-Chlamydia pneumoniae antibodies were measured using ELISA (Vircell, Granada, Spain).

Measurements

Physical activity: measured with the Minnesota Leisure Time Physical Activity questionnaire, validated for use in both sexes in Spain.^14,15 Waist circumference was measured at the midpoint between the last rib and the iliac crest. The height and weight were measured with subjects wearing only undergarments. The blood pressure was obtained using a calibrated sphygmomanometer, after 10 min resting in a seated position. The mean of 2 measurements was calculated. A coronary angiography was performed in a randomized sample of 435 patients to determine the number of vessels with angiographically significant lesions (>70% stenosis). The ejection fraction (EF) was measured indiscriminately using either echocardiography, catheterization or isotopic ventriculography.

Follow-up and Endpoints

The combined endpoint included cardiovascular death, unstable angina, or infarction at the 6-month follow-up. Cardiovascular death was defined as unequivocal cardiovascular death according to medical registry or autopsy. Unstable angina included post-infarction angina (first 28 days after infarction) or readmission for unstable angina (need for hospitalization due to coronary pain with ischemic electrocardiographic changes without elevation of necrosis markers). Infarction was defined as readmission for coronary pain accompanied by an elevation in myocardial necrosis markers or reinfarction during the acute phase (28 days); periprocedural revascularization AMIs were not included. Clinical events after discharge were recorded in telephone interviews or personally and, in both the hospitalization phase as well as after discharge, were verified from the clinical registries. The study was approved by an ethics committee and all patients signed an informed consent.

Statistical Analysis

The Student t test or the Mann-Whitney U test was used for comparison of means between groups with and without RF and with and without each emergent RF, and the c² test for categorical variables. To analyze the relationship between the emergent RF and the event of interest 2 approaches were used: categorizing the emergent RF in tertiles and smoothing using 3-node splines in a Cox linear regression model to define the best cut-points. All the analyses were adjusted for possible confounding variables. The effect of each emergent RF was adjusted for classical RF (hypertension, diabetes, dyslipidemia, and smoking). The oxLDL was also adjusted for the use of statins. An adjustment was made for each hospital to confirm the consistency of the observed effects. The relationship between oxLDL and total LDL and between Lp(a) and oxLDL was studied using bivariate analysis, as was its correlation. The analyses were performed using the statistical package R (R Development Core Team, R Foundation for Statistical Computing, Vienna, Austria, 2008).

RESULTS

The prevalence of patients who suffered a first AMI with no classical RF was 13.2% when just the clinical history as reported by the patient was taken into account and 8.0% (95% confidence interval [CI], 6.3-9.7) when an active measurement of classical RF was made. Of all the patients, 981 (77.1%) had an ST-elevation AMI and 293 (22.9%), a non ST-elevation AMI. Of the patients with a non ST-elevation AMI, 44.0% underwent coronary angiography, and 30.0% of these underwent revascularization. Of the patients with an ST-elevation AMI, 643 (65.5%) received fibrinolysis; 182 (18.5%) primary angioplasty; and 156 (16.0%) did not receive reperfusion.

Of the whole sample of 1371 patients, 126 had a first AMI with no classical RF. Of all the patients, 21.8% had a cardiovascular event in the 6-month follow-up period (18.8% of those who did not present classical RF, with no statistically significant differences [NS]). Mortality was 3.1%, a new AMI in 4.3%, and unstable angina in 14.4%. Table 1 shows the characteristics and the classical RF profile in the patients who had at least 1 classical RF. The patients with no classical RF were a mean of 4.3 years older. Smoking was the most frequent classical RF noted, followed by hypercholesterolemia, hypertension, and diabetes. The clinical characteristics, energy expenditure, and EF were similar in both groups. The percentage of patients with main trunk or multivessel disease, in the patients who received coronary angiography, did not differ between the 2 groups: 43.2% of the patients with no classical RF as compared with 48.9% in the remainder (NS).

Emergent Risk Factors

Table 2 shows the characteristics of the emergent RF for the groups of patients with and without classical RF. Of note was the high percentage of positivity for anti-Chlamydia, with no significant differences between groups. Nor did the mean values of Lp(a), fibrinogen, homocysteine, or CRP differ between groups. The oxLDL values were seen to be higher in the group of patients who presented at least 1 classical RF. The AMI patients with no classical RF did not present more emergent RF.

Six-Month Follow-up

Follow-up was completed in 93% of the patients. The 6-month prognosis in patients with and without classical RF was similar (Table 3). The patients in the group with events during the follow-up were somewhat older and included more women. Low systolic and diastolic blood pressures on admission (greater hemodynamic instability), as well as a Killip III-IV were associated with a worse prognosis. The EF and the proportion of patients with multivessel or main trunk disease (in the subgroup who underwent angiography) were marginally superior in the group that suffered an event (51.5% as opposed to 47.1%; P=.285). The prevalence of classical RF was similar in both groups. No differences were observed in global or abdominal obesity. In relation to the emergent RF studied, higher levels of Lp(a) were found in the patients who suffered an event during the follow-up (Table 3).

Emergent RF and 6-Month Prognosis

Table 4 presents the hazard ratios (HR) of emergent RF in relation to the 6-month prognosis, divided into tertiles in a Cox model adjusted for classical RF, age, sex (and use of statins for oxLDL, which 14.7% of the patients received). A second model included the hospital as an adjustment variable and a third model analyzed the relationship between oxLDL and LDL, as well as Lp(a) and oxLDL.

Concerning the Lp(a) and oxLDL, the patients in the top tertile had an increased risk of developing events at 6 months. No significant differences were observed in the remaining emergent RF analyzed.

The association between CRP, Lp(a), oxLDL, and homocysteine with the combined end-point of cardiovascular morbidity and mortality is shown in Figure. Similar to the relationship seen in the tertile analysis of Lp(a) and oxLDL (Table 4), the values of the top tertile were associated with higher cardiovascular morbidity and mortality. The optimum cut-points were established using smoothing spline analysis: 60 mg/dL for Lp(a) and 74 U/L for oxLDL (Figure). The HR for the patients with values above these cut-points were 1.40 (95% CI, 1.06-1.84) for Lp(a) and 1.48 (95% CI, 1.06-2.06) for oxLDL. The number of events for patients with Lp(a) values >60 mg/dL was 87 (26.9%) and 69 (23.5%) for oxLDL values >74 U/L.

Figure 1. Relationship of emergent risk factors with prognosis at 6 months in patients with a first infarction. Spline analysis (adjusted for classical risk factors, age, and sex. OxLDL also adjusted for use of statins on admission).

The effect of oxLDL and Lp(a) was mutually independent (R²=0.001; P=.31). OxLDL correlated considerably with total LDL (R²=0.103; P<.001). Table 4 shows that, in spite of this, the magnitude of the effect of the variable oxLDL adjusted for LDL (model 3) did not change appreciably.

DISCUSSION

Prevalence of Classical and Emergent RF in the Patients Who Had an AMI

The results obtained in the present study show that the prevalence of patients with AMI but no classical RF out of all the AMI patients, when a strict measurement of these RF was performed in addition to asking about the history, was just 8%. The prevalence would have been 13.2% if only the history as reported by the patients were taken into account. This fact is surely a consequence of unawareness on the part of the patient or because no accurate measurement of classical RF was carried out prior to admission. This value is similar to that found in a meta-analysis² and in the INTERHEART study,⁴ where between 15%-20% of the patients did not present classical RF. These data confirm that the true proportion of AMI patients who lack classical RF is, in fact, low and somewhat lower than has been published in some studies.¹⁶ The patients with no classical RF were older, which is in agreement with previous studies where, in women over 75 years and in men over 65 years, the absence of classical RF was above 20%.¹ This fact can be explained by the decrease in the number of smokers in older patients. Indeed, in men who suffered an AMI who were under 55 years of age or in women under 65 years, an absence of classical RF was observed in only 10%.¹ Among the classical RF, smoking was the most common in young patients, as occurs in other studies where more than 70% of patients who suffered an AMI before reaching 45 years of age were active smokers.^5,6

In our study it was hypothesized that AMI patients with no classical RF could be more exposed to emergent RF, an aspect that has not been confirmed. The high proportion of positive Chlamydia pneumoniae serologies in both groups stands out, a fact also observed in other studies that identified the bacteria in atheromatous plaques.^17,18 Even so, the usefulness of antibiotic treatment in slowing the progression of coronary heart disease has not been demonstrated.¹⁸ The association found with Lp(a) is in agreement with various publications, among these a meta-analysis where this relationship was established independently of other lipid fractions.^6,19

Emergent RF and 6-Month Prognosis

The prognostic value of emergent RF is generally controversial, though the value of oxLDL has recently taken on special interest.¹² In our study only oxLDL and Lp(a) were independently related to the 6-month prognosis in patients with a low-risk AMI. OxLDL is one of the factors required for the formation of coronary plaque through the action of macrophages in the subendothelial space.^8,10 Various studies have shown that patients with ischemic cardiopathy present higher concentrations of oxLDL than healthy individuals,²⁰ and their serum concentrations are associated with extent and severity.^11,21 In middle-aged men, oxLDL values have shown the capacity to predict cardiovascular events¹¹; nevertheless, until now no data were available on the prognostic value of measurement of this marker of oxidative stress in patients who have had an AMI.¹² The use of statins markedly decreases the concentration of phospholipids associated with apolipoprotein B-100 in patients with coronary heart disease²² and various studies suggest these drugs show antioxidant capacity.²³ For this reason, oxLDL levels were adjusted for lipid-lowering therapy in our Cox model.

Lp(a), due to its chemotactic and thrombolysis inhibitory activity in damaged vascular zones,²⁰ has been associated with the presence and extent of coronary heart disease¹⁰ and has demonstrated its power to predict cardiovascular events, especially in diabetic patients.²⁴

In our study, CRP concentrations were similar in patients who had an AMI with or without classical RF. Furthermore, we observed no added prognostic value for CRP after adjusting for classical RF. Some publications have reported an increase in risk associated with elevated CRP values,^25-28 especially >3 mg/L. However, other studies found that CRP was a moderate or null risk predictor.²⁹ The lack of predictive power of CRP in our relatively low-risk cohort might have been influenced by the exclusion of patients who died within 24 h of admission.

Fibrinogen, a possible risk indicator both of coronary heart disease as well as cerebrovascular disease,⁵ was not elevated in our patients with no classical RF nor was a prognostic role seen. Patients treated with fibrinolytic drugs (in whom the fibrinogen concentration can drop abruptly) were excluded from the analysis. No differences were found in homocysteine concentrations between groups (with or without classical RF). Although studies exist that attribute a prognostic value due to their role in plaque activation, oxidative stress, endothelial dysfunction, hypercoagulability, or cell proliferation,⁷ no association with prognosis was observed in our study.

Limitations and Clinical Implications

The present study included patients who survived the first 24 h after infarction. This selection, necessary to obtain fasting blood samples as well as an active measurement of RF, might have influenced the evaluation of the prognostic capacity of RF. The selection of first infarctions reduces the possible confounding influence of a history of prior coronary events in the prognosis and in the management of RF. We highlight the fact that patients older than 74 years of age were excluded from our study, as their co-morbidity could lead to difficulties in the interpretation of oxidative stress.

Our results regarding oxLDL and Lp(a) may have clinical implications for the prognosis of patients who have had an AMI, particularly in those who do not present classical RF. Considering the pleiotropic effect as an antioxidant agent,²³ intensifying statin treatment may be justified in patients presenting an increase of one of these 2 risk indicators.

The absence of statistically significant differences in the 6-month prognosis with relation to EF and the extent of coronary heart disease reflects the low risk of our cohort of first AMI patients under 75 years of age and who survived the first 24 h. This meant that the number of events in each category was small, which could lead to extreme and unstable results. The fact that the study was centered on those patients who survived the first 24 h. from admission for an AMI means that the results cannot be extrapolated to all patients with an AMI, such as those that are not admitted to a hospital. The study did not include AMI related to percutaneous procedures.

The association of oxLDL and Lp(a) with the prognosis was significant and independent, as the analysis was adjusted for confounding variables; even so, the strength of this was modest (HR=1.40 and HR=1.48). Future studies should therefore attempt to replicate these findings.

CONCLUSIONS

The recording of RF in a patient with AMI based not only on the information provided by the patient but also on active measurement of the RF results in the proportion of AMI patients with no observed classical RF being, in reality, very low. These patients have a similar 6-month prognosis to those that do present classical RF. None of the emergent RF analyzed was more prevalent in the AMI patients with no classical RF, although the prognosis was worse in patients who presented with elevated levels of Lp(a) and oxLDL on hospital admission, independent of the presence of classical RF.

ABBREVIATIONS
AMI: acute myocardial infarction
CI: confidence interval
CRP: C-reactive protein
LDL: low-density lipoprotein
Lp(a): lipoprotein (a)
oxLDL: oxidized LDL
RF: risk factors

This study was financed by the Fondo de Investigación Sanitaria (FIS 01/0105/01-05), Red HERACLES (FIS G03/045), AGAUR, Generalitat de Catalunya (SGR 2005/00577), and the Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, (HERACLES Network RD06/0009), and by an unrestricted grant from Sanofi-Aventis.

Contract SNS-FIS-06 (CP06/00100) ISCIII. CIBER of Pathophysiology of Obesity and Nutrition (CB06/03/0028) is an initiative of the ISCIII.

---

Correspondence: Dr. J. Marrugat.
REPICA-ULEC-EGEC-IMIM.
Dr. Aiguader, 88. 08003 Barcelona. España.
E-mail: [e-mail]

Received July 11, 2008.
Accepted for publication January 5, 2009.