Prothrombotic Genetic Risk Factors in Young Survivors of Myocardial Infarction

This case-control study involved consecutive patients who had survived their first episode of acute myocardial infarction occurring before the age of 45. They had been admitted between 1994 and 1997 to the Coronary Care Units of the IRCCS, Policlinico San Matteo, Pavia; the Niguarda Hospital, Milan; the San Camillo Hospital, Rome; and the Ospedale di Bentivoglio and the Giovanni Battista Morgagni Hospital, Forli’, Italy. Acute myocardial infarction was defined as resting chest pain lasting more than 30 minutes, accompanied by ST-segment elevation evolving into pathological Q waves and confirmed by an increase in total creatinine-kinase or the MB fraction to more than twice the upper normal limit. Control subjects were healthy individuals who attended the same hospitals within the frame of a primary prevention study of cardiovascular risk factors, matched with cases for sex, age, and area of origin. Control subjects were evaluated during the same period, and their clinical histories, physical examinations, and negative exercise stress tests excluded the presence of ischemic heart disease.

The collected demographic and laboratory data included age, sex, and traditional risk factors, such as family history of ischemic heart disease, smoking history, blood pressure, total serum cholesterol level, diabetes, obesity, and a history of previous coronary events. For the case subjects, these data were taken from the medical records at the time of their first myocardial infarction; for the control subjects, they were collected at the time of hospital evaluation for enrollment in the primary prevention study. Both the case and the control subjects who agreed to participate in the study were requested to give blood samples for DNA analysis. In the 171 patients who had undergone coronary arteriography, the absence of any narrowing in diameter was considered evidence of a normal coronary artery; a narrowing of less than 50% was considered to be nonsignificant coronary artery stenosis; and a narrowing of more than 50% was considered to be significant coronary stenosis. The patients were classified as having 0-, 1-, 2-, or 3-vessel disease, according to the number of vessels with significant coronary stenoses. The study was approved by the Institutional Review Boards of the corresponding hospitals of the subjects (Pavia, Milan, Rome, Forlı̀, Bentivoglio) and the subjects gave their informed consent.

DNA was extracted from white blood cells using the salting-out method.8 

The guanine insertion/deletion polymorphism is located at nucleotide 675 in the promoter region of the PAI-1 gene and is associated with high plasma PAI-1 levels. A polymerase chain reaction (PCR) and restriction analysis method for detecting this polymorphism has been developed.9 Briefly, a mutated oligonucleotide was synthesized, and a site for the BslI enzyme was inserted in the amplification product, thus making it possible to identify the extra G base by restriction analysis on 4.5% agarose gel electrophoresis.

To detect the C to T transition responsible for the PIA1/A2 polymorphism at nucleotide 1565 in exon 2 of the glycoprotein IIIa gene, we used PCR to amplify the genomic DNA with primers flanking exon 2, as previously described by Weiss et al.10 The PCR products were digested with MspI and NciI restriction enzymes, and the resulting fragments were analyzed on 3% agarose gel.

The 587-bp fragment of the GPIbα gene involving the C3550T polymorphism11 was amplified by PCR and the PCR product was digested with BsaH I followed by 7% acrylamide gels electrophoresis as previously described.12 

The G to A transition at nucleotide 10976 of exon 8 in the factor VII gene responsible for the replacement of arginine by glutamine at codon 353 was determined as described by Green et al.13 In this method, the relevant DNA fragment was amplified by PCR and digested with MspI. The G to A transition results in loss of the recognition site and is designed as the M2 allele and the wild type as the M1 allele.

The mutation was determined by amplifying a region of exon 10 and the adjacent intron by PCR, as previously described.4 The 220-bp fragment was digested by MnII at 37°C and analyzed on 2% agarose gel. The genotypes are designated as GG, GA, and AA.

The C677T transition was detected by PCR.14 The 198-bp fragment underwent HinfI restriction enzyme analysis and subsequent electrophoresis on 4% agarose gel.

For the direct identification of the nucleotide transition G20210A in the prothrombin gene, genomic DNA was specifically amplified by using the 5′ primer in exon 14 and a mutagenic primer in the 3′-untranslated region, as previously described.5 

Descriptive statistics include mean values and standard deviations for continuous variables, and proportions for categorical data. The strength of the association between traditional and prothrombotic genetic risk factors and acute myocardial infarction at young age was estimated by calculating the odds ratios and 95% confidence intervals (CI). The presence of interactions between traditional and prothrombotic genetic risk factors was estimated by calculating “the attributable proportion of the disease” caused by the interaction, with 95% CI.15 

The demographic characteristics and prevalence of traditional risk factors for coronary artery disease in the 200 young survivors of myocardial infarction and in the 200 control subjects are shown in Table 1. As the two groups were age- and sex-matched, there were no significant differences in these variables. Of the other traditional risk factors examined, a family history of ischemic heart disease was associated with the highest odds ratio, followed in order by hypertension, smoking, obesity, diabetes mellitus, and hypercholesterolemia (Table 1). Coronary arteriography was performed in 171 cases (86%), 22 of whom (13%) had normal coronary arteries, 79 (46%) had 1-vessel disease, 42 (25%) had 2-vessel disease, and 28 (16%) had 3-vessel disease. There was no association between traditional risk factors and the degree of coronary artery disease at angiography (data not shown).

Table 2 shows that there was no difference between the young survivors of myocardial infarction and controls in terms of genotype or allele frequencies for the following polymorphisms: 4G/5G of the PAI-1 gene, C3550T of the platelet glycoprotein Ib gene, G10976A of the factor VII gene, C677T of the methylenetetrahydrofolate reductase gene, G1691A of the factor V gene, and G20210A of the prothrombin gene. However, there was a statistically significant difference in the frequency of the PIA1/PIA2 polymorphism of the platelet glycoprotein IIIa gene. Among the young survivors of myocardial infarction, 141 (70.5%) carried the PIA1/PIA1 genotype, 54 (27%) carried the PIA1/PIA2 genotype, and 5 (2.5%) carried the PIA2/PIA2 genotype, so that the overall frequency of the PIA1 and PIA2 alleles in patients was 84% and 16%. In controls, the distribution of the PIA1/PIA1, PIA1/PIA2, and PIA2/PIA2 genotypes was respectively 81.5%, 16.5%, and 2%, with an allele frequency of 89.8% for PIA1 and 10.2% for PIA2. The prevalence of PIA2 genotypes (ie, the percentage of patients who were either heterozygous [PIA1/PIA2] or homozygous [PIA2/PIA2]) was higher in the young survivors of myocardial infarction (29.5%) than in the controls (18.5%) (odds ratio 1.84; 95% CI 1.12 to 3.03), as was frequency of the PIA2 allele (16% v 10.2%) (odds ratio 1.67; 95% CI 1.07 to 2.59). The combination of two or more genetic risk factors for thrombosis was not significantly associated with an increased risk for premature myocardial infarction (data not shown). There was no association between the genetic risk factors for thrombosis and the degree of coronary artery disease at angiography (data not shown).

There was an interaction between smoking and the presence of the PIA2 polymorphism: among patients who did not smoke, the PIA2 polymorphism was not associated with an increased risk of premature myocardial infarction (odds ratio 1.64; 95% CI 0.55 to 4.63); whereas, in the presence of smoking, it was (odds ratio 2.03; 95% CI 1.04 to 4.01). In comparison with the patients who did not smoke and did not carry the PIA2 allele, smokers carrying the PIA2 polymorphism had a 13-fold increase in their risk of premature myocardial infarction (Table3). In smokers who carried the PIA2 allele, 46% (95% CI: 11% to 81%) of premature myocardial infarctions was attributable to the interaction of the two risk factors. The risk of premature myocardial infarction associated with smoking was also modified by the G1691A mutation in the factor V gene. In the absence of the mutation, smoking was associated with a sevenfold increase in the risk of premature myocardial infarction, whereas, in its presence, smoking was associated with a 12-fold increase in the risk (Table4). However, smoking and the mutation in the factor V gene showed only a trend towards an interaction: 31% (95% CI: 1% to 145%) of premature myocardial infarction was attributable to the interaction of the two risk factors. There was no interaction between smoking or other traditional risk factors and the remaining five polymorphic alleles.

Myocardial infarction usually occurs because an occlusive acute thrombus develops at the site of a ruptured atheromatous plaque in an epicardial coronary artery.16,17 Both atherosclerosis and thrombosis do, therefore, contribute towards the occurrence of myocardial infarction, although the relative importance of these two processes varies from patient to patient and is different at different ages. Younger patients with premature myocardial infarction tend to have less coronary atherosclerosis and a higher prevalence of normal or near-normal coronary arteriograms.6,7 The importance of hypercoagulability in the pathogenesis of acute myocardial infarction has been well established by the finding that plasma levels of proteins involved in the hemostatic mechanism (such as fibrinogen, factor VII, tissue plasminogen activator antigen and its primary inhibitor) have been associated with an increased risk of myocardial infarction.18-20 The contribution of thrombosis to the development of myocardial infarction may be particularly important at a young age, and premature myocardial infarction provides a unique model to investigate the role of prothrombotic risk factors.

With this background, this study was designed to assess the degree of association between genetic markers that previous studies have linked to an increased thrombotic tendency and the occurrence of myocardial infarction in a selected group of young survivors of myocardial infarction, mainly men. More than half of our patients had nonsignificant coronary artery stenosis or single vessel disease, identifying a cohort with relatively little coronary atherosclerosis, in which prothrombotic genetic factors are more likely to play an important pathogenic role. The presence of the PIA2 allele of the platelet glycoprotein IIIa gene was the only genetic factor associated with the risk of developing premature myocardial infarction. The presence of the PIA2 allele seems to be associated with an increase in platelet function. Feng et al21 showed that platelets from patients with the PIA2 allele require lower concentration of epinephrine and adenosine diphosphate (ADP) for aggregation than platelets with the PIA1 allele and also that sensitivity to ADP correlates with the number of copies of the PIA2 allele. Zotz et al22 have shown that PIA2 platelets bind more fibrinogen than PIA1 platelets across a wide range of ADP concentrations. A few clinical studies also found an association between the PIA2 allele and ischemic heart disease. In 318 patients who had undergone successful stent placement, patients with the PIA2 allele experienced a more than fivefold increase in the risk of stent thrombosis (9.5% v1.9%).23 In the Cholesterol and Recurrent Events (CARE) trial, the PIA2 allele was associated with an excess of recurrent coronary events after myocardial infarction, and pravastatin seems to reduce this excess risk substantially.24 Although the majority of these studies are only published in abstract form, on the whole they give biological and clinical plausibility to our findings that the PIA2 allele is a marker of risk for coronary artery disease.

The role of the PIA2 allele as a marker of risk of coronary artery disease is currently controversial. Ridker et al,25 who studied subjects from the Physician Health Study with an exceptionally low prevalence of smoking, found no association between the PIA2 allele and the development of myocardial infarction, whereas Weiss et al,10 who studied a younger population with a much higher prevalence of smoking (72% v 15.7%), did find such an association. There are several other studies with negative results as to the association between the PIA2 allele and myocardial infarction. The largest and most recent one is that of Gardemann et al,26 who studied 1,191 patients. The mean age of their patients is much older (61.3 years) than that of our patients, and the interaction between the polymorphism and age was not fully evaluated. Furthermore, differences in selection criteria may explain the different findings; for example, our cases included only patients who survived a Q wave myocardial infarction, and our controls were selected on the basis of negative exercise test. Perhaps the choice of controls with a negative exercise test explains the fact that in our study, the prevalence of the PIA2 allele was slightly lower than that reported by most studies taking as controls unselected “healthy” subjects from the general population, who may include cases with latent ischemic heart disease.

None of the remaining genetic markers evaluated in this study were associated with the occurrence of premature myocardial infarction. In particular, we failed to observe the previously reported “protective” effect of the polymorphic allele of the factor VII gene identified as M2 or Gln353 allele.27 However, the population selected for the previous study was substantially different from ours,28 because patients were more than 45 years old at the time of infarction.27 In addition, patients were selected on the basis of a positive family history for cardiovascular disease.27 The analysis of our subgroup with these features (93 patients and 11 controls) showed no significant difference (data not shown), but the small size of the sample and large beta error might account for the negative findings.

In patients with venous thromboembolism, the carriership of more than one genetic prothrombotic risk factor has a clear additive effect.28 In myocardial infarction, no additive effect was observed by us in patients carrying more than one genetic prothrombotic risk factor. However, the presence of the PIA2 allele did interact with smoking. As expected, smoking by itself was associated with an increased risk of developing myocardial infarction at a young age, but this risk was doubled in the presence of the PIA2 allele, and 46% of the disease was attributable to the interaction between the two factors. One possible explanation for this interaction may be related to the fact that smoking per se is a prothrombotic stimulus, because it is able to increase fibrinogen, other coagulation proteins, and platelet aggregability.20 It is therefore conceivable that the combined effect of prothrombotic factors stands out most sharply in young individuals with nonsignificant coronary atherosclerosis, in whom the prevailing pathogenetic mechanism of myocardial infarction is hypercoagulability. This view is consistent with the present findings and with our previous observation29 showing that, in populations mainly composed of young men, the factor V mutation by itself is not associated with an increased risk of early myocardial infarction, although in this study there was a nonsignificant trend for an interaction between the mutation and smoking. Moreover, Eikelboom et al30 have shown that in young survivors of myocardial infarction, predominantly men, the prothrombin mutation was not associated with an increased risk. Only in a highly selected population of young women who survived myocardial infarction, both the factor V and prothrombin gene mutations interacted with smoking.31,32 Young women have usually even less coronary atherosclerosis than young men, emphasizing once more that prothrombotic genetic factors play a role particularly in the individuals with less abnormal coronary arteries.

Based on the results of this study, we surmise that in a complex, polygenic, and multifactorial disorder such as myocardial infarction, the role of an inherited predisposition to thrombosis is relatively weak in itself and weaker than the role of traditional risk factors. Yet some prothrombotic genetic factors may be important contributors in combination with environmental factors or acquired prothrombotic stimuli, particularly with smoking. Prospective studies are needed to examine the clinical value of the interaction between environmental factors and the currently available or newly found prothrombotic genotypes.