Glycoprotein IIb/IIIa is the major platelet receptor for fibrinogen and plays an important role in platelet-mediated thrombosis and the development of coronary heart disease. A number of studies have examined the role of the PLA2 variant of glycoprotein IIIa (Leu/Pro33) in the occurrence of myocardial infarction (MI), but the association remains controversial. A second common genetic glycoprotein IIb/IIIa variant, glycoprotein IIb Ile/Ser843, has not been associated with acute coronary thrombotic events in several studies involving predominantly middle-aged to older men (Hato et al, 1997; Böttiger et al, 1999, 2000). We previously demonstrated the importance of the interactions of two inherited prothrombotic mutations, factor V Arg506Gln and prothrombin G20210A, with other known cardiovascular risk factors in women under the age of 45 years with MI (Rosendaal et al, 1997). In this report, we assessed the association of the two common platelet glycoprotein IIb/IIIa polymorphisms, along with three other common platelet glycoprotein variants (glycoprotein Ibα Thr/Met145, glycoprotein Ia Glu/Lys505 and glycoprotein Ia C807T), in young women with MI, and examined potential interactions with traditional cardiovascular risk factors.
Several platelet glycoprotein polymorphisms have been associated with an increased risk of myocardial infarction (MI) in studies that included predominantly men. In a population-based sample of 68 Caucasian women < 45 years old with non-fatal MI and 346 demographically similar control subjects, we found an increased risk of MI among women who possessed at least one copy of the glycoprotein IIb Ser843 allele compared with those lacking the Ser843 allele (odds ratio 1·85; 95% confidence interval = 1·03–3·33). The increased risk was present only in subgroups of women who smoked cigarettes, had hypercholesterolaemia or who had a family history of early onset MI. The Ser843 variant of glycoprotein IIb may be associated with an increased risk of MI in young women with other cardiovascular risk factors. Additional studies involving larger numbers of subjects are needed to confirm this preliminary finding.
Patients and methods
Detailed methods of this population-based case–control study of acute MI in young women are described elsewhere (Rosendaal et al, 1997). Briefly, the MI patients included women aged 18–44 years, residing in Western Washington State, who suffered a non-fatal acute MI between 1991 and 1995. Controls were women 18–44 years of age without a history of cardiovascular disease from the same geographical area frequency, matched to the age distribution of the cases.
Current smokers were defined as subjects who reported smoking at least five cigarettes per week for at least six consecutive months. Subjects who were former smokers as well as those who had never or seldom smoked cigarettes were classified together as ‘non-smokers’. Obesity was defined as a body mass index (BMI) ≥ 30·0 kg/m2. A woman was classified as hypertensive or diabetic if she reported ever receiving the diagnosis by a physician. Hypercholesterolaemia was defined as a cholesterol measurement > 5·2 mmol/l. Family history of early onset MI was defined as having at least one first-degree relative who suffered an acute MI before the age of 55 years. Genotype analysis for five common platelet glycoprotein polymorphisms (included in Table I) was performed by polymerase chain reaction (PCR) amplification of genomic DNA, followed by restriction enzyme digestion (PCR-RFLP) (Reiner et al, 2000).
|Allele frequency||Cases (n = 68)||Controls (n = 346)|
The association of platelet glycoprotein genotypes with MI was examined using unconditional logistic regression adjusted for age, and expressed as odds ratios (OR) and 95% confidence intervals (CI95). We assessed the extent to which associations with glycoprotein polymorphisms were modified by other cardiovascular risk factors through stratified analyses. We then used logistic regression models to estimate the OR for persons possessing both the genotype of interest and the risk factor, persons possessing only the genotype of interest but not the risk factor, and persons possessing only the risk factor but not the genotype, for each combination of platelet glycoprotein genotype and cardiovascular risk factor, all relative to persons lacking both the genotype and risk factor. We measured the strength of the effect modification by calculating the relative excess risk due to interaction (RERI), which takes on the null value of 0 when there is no difference in the risk of MI associated with the genotype according to the level of the cardiovascular risk factor. The statistical precision of the magnitude of interaction was assessed by calculating 95% confidence intervals on the RERI (Hosmer & Lemeshow, 1992).
Of the initial 112 MI patients and 525 control women that were interviewed for the case–control study, DNA samples for platelet glycoprotein genotyping were available for 68 Caucasian MI patients and 346 Caucasian controls. The 68 MI patients had a mean age of 39·7 years (range 23–44 years) and the 346 control women had a mean age of 37·8 years (range 19–44 years). The majority of both MI cases (90%) and control subjects (96%) were premenopausal. Only 4% of MI cases and 11% of control subjects were current users of oral contraceptives. Traditional cardiovascular risk factors were more common in MI patients than in controls: current smoking (71% vs. 21%), obesity (44% vs. 16%), hypercholesterolaemia (75% vs. 36%), hypertension (35% vs. 9%) and diabetes (15% vs. 3%). There were no significant differences in demographic characteristics or cardiovascular risk factors between the women who had DNA samples available for analysis and those who did not (data not shown).
The distribution of glycoprotein genotypes among the controls subjects was in Hardy–Weinberg equilibrium. As shown in Table I, MI cases (73·5%) were more likely to carry at least one copy of the less common Ser843 allele of glycoprotein IIb than control subjects (59·5%), although the frequencies of Ser843 homozygotes were similar (16·2% vs. 14·5%). There was a significantly increased risk of MI for women who carried the glycoprotein IIb Ser843 allele (Ser843/Ser843 or Ile843/Ser843) compared with the Ile843/Ile843 genotype (OR 1·85; CI95 = 1·03–3·33). The OR comparing the Ile843/Ser843 genotype with the Ile843/Ile843 genotype was 1·90 (CI95 = 1·03–3·50). The OR comparing the Ser843/Ser843 genotype with the Ile843/Ile843 genotype was 1·69 (CI95 = 0·74–3·85). There was no association between glycoprotein IIb genotype and smoking, obesity, hypertension, diabetes or hypercholesterolaemia among either cases or controls (data not shown). Of the four other platelet glycoprotein polymorphisms analysed, the genotype distributions were similar among cases and controls (Table I).
In analyses stratified according to the presence or absence of other cardiovascular risk factors, the glycoprotein IIb Ser843 allele was associated with a ≈three- to fourfold increased risk of MI in the subgroup of women who smoked cigarettes, who had a serum cholesterol level > 5·2 mmol/l and who had a family history of early onset MI in first-degree relatives (Table IIA). In contrast, the glycoprotein Ser843 allele was not associated with increased risk of MI in women who were non-smokers, who had a serum cholesterol level ≤ 5·2 mmol/l and who had a negative family history. No significant effect modifications were observed when the risk of MI associated with the glycoprotein IIb Ser843 allele was stratified according to obesity, hypertension, diabetes, or the presence or absence of the factor V Leiden and prothrombin G20210A mutations (data not shown).
|Patients (n = 68)||Controls (n = 346)|
|Number (%)||Number (%)||Number (%)||Number (%)||OR||(C195)|
|Current non-smokers||10 (50·0)||10 (50·0)||162 (59·1)||112 (40·9)||0·71||(0·28–1·77)|
|Current smokers||40 (83·3)||8 (16·7)||44 (61·1)||28 (38·9)||2·95||(1·19–7·30)|
|≤ 5·2 mmol/l||10 (58·8)||7 (41·2)||133 (61·0)||85 (39·0)||0·91||(0·33–2·48)|
|> 5·2 mmol/l||40 (78·4)||11 (21·6)||69 (57·0)||52 (43·0)||2·60||(1·21–5·60)|
|Family history of early MI|
|No||29 (69·0)||13 (31·0)||182 (61·9)||112 (38·1)||1·37||(0·68–2·75)|
|Yes||19 (79·2)||5 (20·8)||20 (44·4)||25 (55·6)||4·43||(1·33–14·69)|
Analysis of the separate and combined effects of the glycoprotein IIb Ser843 allele with smoking (Table IIB) indicates that smoking by itself was associated with a threefold increased risk of MI compared with non-smokers without the glycoprotein IIb Ser843 allele, while the glycoprotein IIb Ser843 allele was not associated with MI in the absence of smoking. However, the combination of smoking and the glycoprotein IIb Ser843 allele was associated with a 10-fold increase in risk compared with women who did not smoke and did not carry the glycoprotein IIb Ser843 allele; the effect of smoking on the association of the glycoprotein IIb allele with non-fatal MI was greater than predicted by the addition of the risks associated with smoking alone and the glycoprotein IIb Ser843 allele alone (RERI = 6·77, CI95 = 0·43, 13·11). Qualitatively similar but less pronounced effect modifications were observed for the combinations of the glycoprotein IIb Ser843 allele and plasma cholesterol or family history of MI (Table IIB). The results of this analysis were altered only slightly by additional adjustment for the confounding variables of smoking and hypercholesterolaemia.
|Glycoprotein IIb genotype||Risk factor||OR||CI95||RERI (CI95)|
|Ser843-positive||Current||9·86||4·50–21·60||6·77 (0·43, 13·11)|
|Ser843-negative||≤ 5·2 mmol/l||1||–|
|Ser843-positive||≤ 5·2 mmol/l||0·91||0·33–2·48|
|Ser843-negative||> 5·2 mmol/l||2·43||0·88–6·70|
|Ser843-positive||> 5·2 mmol/l||6·33||2·65–15·16||3·99 (0·18, 7·81)|
|Family History of MI|
|Ser843-positive||Yes||7·95||3·35–18·84||5·79 (−0·28, 11·86)|
Our results suggest a possible association between the Ser843 variant of platelet glycoprotein IIb and the risk of MI in young women. The increased risk of MI associated with glycoprotein IIb Ser843 was only present among women who smoked cigarettes, had elevated serum cholesterol levels or had a positive family history of MI, and the interaction with smoking appeared strong. Women who possessed the glycoprotein IIb Ser843 allele and who were current smokers had an ≈10-fold increased risk of MI, whereas there was no increased risk associated with the glycoprotein IIb Ser843 allele in women who were non-smokers.
A recent in vitro study indicated that the glycoprotein IIb Ser843 variant is associated with both increased platelet aggregation and decreased clot retraction compared with the Ile843 variant (Bray et al, 1999). Whether the Ile/Ser843 substitution is causally responsible for these effects or whether the dimorphism is genetically linked to another causative locus in the glycoprotein IIb gene or elsewhere remains to be determined. Further functional characterization of the glycoprotein IIb Ile/Ser843 polymorphism may contribute to our understanding of the role of the platelet glycoprotein IIb/IIIa receptor in the pathogenesis of arterial thrombosis and potentially allow risk stratification of MI patients (e.g. the identification of individuals particularly likely to benefit from glycoprotein IIb/IIIa inhibitor therapy). Cigarette smoking is associated with increased glycoprotein IIb/IIIa receptor activation (Markovitz et al, 1999), as well as with increased circulating levels of fibrinogen (Folsom, 1999), the ligand that binds to glycoprotein IIb/IIIa. These findings provide possible mechanisms underlying the synergism we observed between the genetic variant of platelet glycoprotein IIb and smoking in the pathogenesis of acute MI.
In contrast to our results, several larger studies involving predominantly male patients of a wider age range have not observed an association between the Ser843 variant of glycoprotein IIb and the risk of MI (Hato et al, 1997; Böttiger et al, 2000) or coronary stent occlusion (Böttiger et al, 2000). However, the genetic influence on risk of MI may be stronger in younger women than in older women or men (Marenberg et al, 1994). Furthermore, premenopausal women exhibit increased platelet glycoprotein IIb/IIIa activation in response to platelet agonists compared with young men (Faraday et al, 1997). Of note, we recently observed an association between glycoprotein IIb Ser843 homozygosity and ischaemic stroke in young women with other cardiovascular risk factors (Reiner et al, 2000). In contrast, Carter et al (1999) found no relationship between the glycoprotein IIb polymorphism and the occurrence of ischaemic stroke in elderly men and women, but did report an association between the Ser843 variant and a decreased risk of stroke-related death.
The major limitation of our study is the relatively small number of cases (MI is a rare event in women < 45 years of age), which may increase the likelihood of a false-positive association. The performance of multiple subgroup analyses may also increase the probability that an observed association may have arisen by chance. Furthermore, while our results do not support an association between several other platelet glycoprotein polymorphisms and MI in young women, our sample size may not have sufficient power to detect a small effect, particularly for the less frequent polymorphisms. Thus, analysis of a larger number of patients will be required to exclude these other platelet glycoprotein polymorphisms as risk factors for MI in young women. In addition, as our study included only women who survived an acute MI, it is possible that an association may have been missed if a particular genotype is also associated with early post-MI mortality. The lack of an increased risk of MI in women who were homozygous for the glycoprotein IIb Ser843 variant in our study may suggest the possibility of a spurious association. Alternatively, the lack of association in Ser843 homozygous women could be as a result of either the relatively small number of cases or an over-representation of the Ser843/Ser843 genotype among the fatal MI cases. Thus, confirmation of our preliminary association between the glycoprotein IIb polymorphism and the risk of MI in young women will require further study involving a larger number of subjects, as well as patients who suffer fatal MI.
We thank Fran Chard, Karen Graham, Carol Handley-Dahl, Judy Kaiser, Marlene Bengeult, Carol Ostergard, Denize Horlander, Barb Twaddell, Sandy Tronsdal and Jill Ashman for assistance with the study, Benjamin Siscovick for providing technical assistance and Ida McCormick for assistance in preparation of the manuscript.