The ₂ Gene Coding Sequence T₈₀₇/A₈₇₃ of the Platelet Collagen Receptor Integrin ₂β₁ Might Be a Genetic Risk Factor for the Development of Stroke in Younger Patients

THE DEVELOPMENT OF stroke is known to be related to vascular risk factors such as hypertension, smoking, diabetes mellitus,1,2 and possibly high cholesterol.3 Besides these conventional risk factors, genetic predispositions may enhance the risk of developing a cerebrovascular thrombotic event.4-6 This is particularly true with stroke, as demonstrated by clustering of stroke cases in certain families and by ethnic differences in stroke incidence.7 8 

Platelet-dependent thromboembolism is an underlying mechanism in the pathogenesis of stroke. Several platelet receptors are known to be polymorphic9; however, no platelet glycoprotein-receptor polymorphism has yet been associated with an increased risk for stroke.10 

The platelet integrin α₂β₁, also known as glycoprotein (GP) Ia-IIa, is one of the major collagen receptors present on platelets.11 Via GPIa-IIa, platelets adhere to collagen exposed in subendothelial structures and subsequently become activated.12-14 Recently, two silent point mutations on GPIa (α₂) were identified, C₈₀₇T and G₈₇₃A.15 These mutations were shown to be associated with the expression density of GPIa-IIa on the platelet surface,15 thus explaining the previously reported difference in expression density of GPIa-IIa.16 The two polymorphisms are linked with each other. GPIa T₈₀₇/A₈₇₃ is associated with a higher expression of the receptor and GPIa C₈₀₇/G₈₇₃with a lower expression density.15 Furthermore, the expression density of GPIa-IIa could be correlated to the rate of platelet attachment to collagen type I, even under high shear rates.17 

The polymorphism increasing the density of GPIa-IIa on the surface of platelets might present an inherited risk factor for platelet-dependent thromboembolic complications. Studies in patients with myocardial infarct18 and stroke4 indicate that genetic risk factors may be especially relevant to the development of arterial vascular disease in younger patients. In this case-control study, we assessed the GPIa C₈₀₇T genotype distribution in 227 stroke patients divided into different age groups (≤50 and >50 years of age) and compared it with those of two control groups: inpatients without cerebrovascular disease (non-CVD) and healthy blood donors.

This case-control study was conducted in accordance with the good clinical practice guidelines of the European community and the Declaration of Helsinki. The study protocol was approved by the local ethics committee of the University of Greifswald (Greifswald, Germany). All patients and control subjects gave informed consent to participate in the study.

Blood samples were collected from 227 stroke patients admitted to the stroke service of the Department of Neurology. We included only patients with focal neurological symptoms due to transient ischemic attack or complete stroke and excluded patients with an intracerebral or subarachnoidal hemorrhage. Patients with stroke resulting from vasculitis or migraine were also excluded. All stroke patients underwent ultrasound sonography (both extracranial and transcranial), computed tomography (CT) or magnetic resonance imaging (MRI), and electrocardiogram (ECG). Transesophageal echocardiography and digital substraction angiography were optional. On the basis of clinical features, CT/MRI findings, and cardiac diagnostic criteria, the patients were assigned to the following CVD categories: (1) large-vessel disease—stroke due to significant (>70%) stenosis of the extracranial symptomatic artery; (2) lacunar stroke—due to small deep infarctions, with the location of the lacunae seen in the CT/MRI matched to the clinical symptoms; (3) cardiac embolism—source positively diagnosed by ECG or transesophageal echocardiography; or (4) other—eg, arterial dissection or unknown pathogenesis despite intensive investigation.

The stroke patients were compared with a control group of 170 neurological inpatients showing no acute or recent signs of CVD or any other condition in which vascular pathology plays a major role, such as migraine or arteriovenous malformations. Further exclusion criteria for the control group were acute myocardial infarction or venous thrombosis.

A second control group of 184 healthy blood donors was assessed to establish the GPIa-IIa C₈₀₇T allele distribution in the regional population. All participants were whites living in the same geographical region.

DNA was prepared from 5 mL whole blood according to Miller et al19 or from 250 μL buffy coat with the QIAamp blood kit (Qiagen GmbH, Hilden, Germany). Genotyping was performed in accordance with a newly developed sequence-specific primer (SSP)-polymerase chain reaction (PCR) method.20 In brief, 5 μL genomic DNA (20 μg/mL) was added to a 50 μL reaction mixture containing 10 mmol/L Tris (pH 8.3), 50 mmol/L KCl, 1.5 mmol/L MgCl₂, 0.125 mmol/L dNTP, 0.25 μmol/L each of sense primer #92V (5′-GAC AGC CCA TTA ATA AAT GTC TCC TCT G) and sequence-specific antisense primer #96C (5′-CCT TGC ATA TTG AAT TGC TAC G) or #96T (5′-CCT TGC ATA TTG AAT TGC TAC A), 0.125 μmol/L each of internal control primers HGH I (5′-CAG TGC CTT CCC AAC CAT TCC CTT A) and HGH II (5′-ATC CAC TCA CGG ATT TCT GTT GTG TTT C), and 2.5 U TaqGold (Perkin Elmer, Vaterstetten, Germany). After initial denaturation at 96°C for 10 minutes, 35 cycles of amplification were performed (93°C for 50 seconds, 56°C for 30 seconds, and 72°C for 15 seconds). The PCR products were analyzed on 1.5% agarose gels and visualized by ethidium bromide staining.

The allele frequencies and genotypes determined for stroke patients, non-CVD patients, and blood donors were compared by χ²analysis in 2 × 2 and 3 × 2 fields, respectively. Fisher’s exact test in 2 × 2 fields was used to compare other nominal data. The significance level was established at P ≤ .05. Odds ratios with the corresponding 95% confidence interval (CI) were determined for conventional risk factors (hypertension, diabetes, smoking [current or ex-smoker], and high cholesterol [defined as known high levels of plasma cholesterol, current lipid-lowering medication use, and/ or plasma cholesterol ≥6.5 mmol/L]), positive family history (stroke), and the presence of the GPIa T₈₀₇allele. The statistical analyses were performed using SAS software (SAS Institute, Cary, NC). A logistic regression analysis performed on SPSS software (SPSS Inc, Chicago, IL) included gender, hypertension, diabetes mellitus, smoking, high cholesterol, and the GPIa genotype in the age-defined patient groups.

The 227 stroke patients included 107 men (47.1%) and 120 women (52.9%). Their mean age was 62.2 ± 14.3 years. The control group of 170 neurological non-CVD inpatients comprised 78 men (45.9%) and 92 women (54.1%) with a mean age of 59.4 ± 14.3 years. Forty-five stroke patients (19.8%) and 41 non-CVD patients (24.1%) were ≤50 years of age. Patients’ and control subjects’ characteristics are shown in Table 1. One hundred and eighty-four healthy blood donors (104 men and 80 women) were included in the study to reflect the GPIa C₈₀₇T genotype distribution in a Middle European population. Their mean age was 30.3 ± 7.1 years.

Forty-two patients (18.5%) presented with a transient ischemic attack and 185 patients (81.5%) suffered a complete stroke. In the stroke patients, the pathogenesis were large vessel atherosclerosis in 70 patients (30.8%), lacunar stroke in 34 patients (15%), and cardiac embolic sources in 53 patients (23.3%). More than one of these causes was found in 16 patients (7%), 12 (5.3%) participants suffered from a dissection, and the pathogenesis of stroke remained unknown in 42 patients (18.5%).

There were no differences between non-CVD patients and healthy blood donors in GPIa C₈₀₇T allele frequencies (0.59 C₈₀₇, 0.41 T₈₀₇v 0.61 C₈₀₇, 0.39 T₈₀₇, respectively; P = .53) and genotype distributions (34.1% CC₈₀₇, 48.8% CT₈₀₇, 17.1% TT₈₀₇v 37.5% CC₈₀₇, 46.7% CT₈₀₇, 15.8% TT₈₀₇, respectively; P = .8).

The distribution of predefined risk factors, as well as the proposed risk factor, GPIa T₈₀₇, were compared between (1) stroke patients (n = 45) and non-CVD patients (n = 41) ≤50 years of age and (2) stroke patients (n = 182) and non-CVD patients (n = 129) more than 50 years of age (Table 1). In the group ≤50 years of age, none of the conventional risk factors was overrepresented in stroke patients. The only difference between the groups of patients ≤50 years of age was the overrepresentation of GPIa T₈₀₇ in stroke patients (odds ratio, 3.02; 95% CI, 1.20 to 7.61). In contrast, in the patient group more than 50 years of age, hypertension (P < .001), diabetes mellitus (P = .009), and smoking (P = .001) were overrepresented in stroke patients compared with non-CVD patients. High cholesterol (P = 1.000), a positive family history (P = .767), and GPIa T₈₀₇ (P = .712) were not overrepresented in the group of older stroke patients compared with control patients (Table 1).

A logistic regression model (Table 2) confirmed the importance of the GPIa T₈₀₇ allele in the young stroke patients. The presence of GPIa T₈₀₇ was identified as the only independent risk factor in this group (P= .0207). In contrast, in the patient group >50 years of age, hypertension (P < .0001), diabetes (P = .0110), and smoking (P = .0075) were independent risk factors for stroke.

A comparison of stroke pathogenesis in younger and older CVD patients is presented in Table 3. There were no lacunar strokes in younger patients, but dissection and unknown pathogenesis were more frequent in this group than in the older patients. The genotype distribution relative to CVD pathogenesis is also given in Table 3. The GPIa T₈₀₇ allele was most frequent in young CVD patients with stroke due to carotid atherosclerosis (86%) or due to cardiac embolic sources (100%).

In stroke patients with a positive family history, 27 of 41 (65.8%) carried the GPIa T₈₀₇ allele, but 64 of 92 (70.0%) of the stroke patients without strokes in the family history also carried the GPIa T₈₀₇ allele (P = .69). Of stroke patients ≤50 years of age, the family history was available in 25 cases, and 7 of these had a positive family history. Of the 7 with positive family histories, 5 (71.4%) carried the GPIa T₈₀₇ allele. Of the remaining 18 stroke patients ≤50 years of age without a positive family history, 12 (66.7%) also carried the GPIa T₈₀₇allele.

Results of this pilot case-control study show that the polymorphism T₈₀₇/A₈₇₃ of the α₂ gene coding sequence of platelet integrin α₂β₁, also known as GPIa-IIa, may be an inherited risk factor for the development of stroke in patients ≤50 years of age (odds ratio, 3.02; 95% CI, 1.20 to 7.61).

Platelet adhesion to subendothelial tissue is largely mediated via the collagen receptor GPIa-IIa.11 The GPIa gene is polymorphic at position 807 (C-T) and this point mutation is associated with GPIa-IIa expression density on the platelet surface. Platelets from individuals homozygous or heterozygous for the GPIa T₈₀₇ allele show an increased expression of GPIa-IIa; consequently, an enhanced adhesion to collagen compared with platelets homozygous for GPIa C₈₀₇.15-17 

In patients ≤50 years of age, GPIa T₈₀₇ was the only significant predefined risk factor for stroke; moreover, logistic regression analysis showed GPIa T₈₀₇ to be an independent risk factor for stroke in this patient group. However, the effect of the GPIa T₈₀₇ allele as a risk factor for stroke was not pronounced, which may be due to the small number of patients in our pilot study. Nevertheless, this effect might be important for subgroups of stroke patients, because the GPIa T₈₀₇ allele was particularly prominent in younger patients who had a stroke due to carotid atherosclerosis or cardiac embolism (Table 2). In fact, all patients ≤50 years of age who suffered a stroke likely due to cardiac embolism carried the GPIa T₈₀₇ allele. In patients more than 50 years of age, the GPIa T₈₀₇ allele was not as frequent and its potential effect appears to be overshadowed by conventional risk factors and general vessel disease.

Atherosclerosis is usually less frequent in younger patients than in older ones, but in our study 31% of the stroke patients ≤50 years of age and 31% of stroke patients more than 50 years of age presented with carotid artery atherosclerosis. Unfortunately, we could not calculate the effect of this risk factor in the multiple regression analysis model because patients in the control group were not systematically assessed for atherosclerosis of the carotid artery.

Surprisingly, we found no clear correlation between the presence of the GPIa T₈₀₇ allele and a positive family history of CVD, which one would expect with a genetic risk factor. Again, this finding might be related to the small sample size of the study.

Results of this pilot study concur with findings of Santoso et al (manuscript submitted), who investigated the impact of GPIa C₈₀₇T polymorphism on the development of coronary artery disease and myocardial infarction. These investigators assessed 2,237 male patients admitted for angiography and identified a strong association of the GPIa T₈₀₇ allele with nonfatal myocardial infarction in younger patients.20 Conversely, if the presence of the GPIa T₈₀₇ allele is a potential thromboembolic risk factor, the C₈₀₇ allele may be associated with an increased bleeding risk in certain populations. DiPaola et al22 found an overrepresentation of the GPIa C₈₀₇ allele among type I von Willebrand’s disease patients with bleeding complications.

Although there is increasing evidence that the GPIa T₈₀₇allele is an important genetic factor in the predisposition for thromboembolic events, due to the small number of patients in this pilot case-control study, our findings remain preliminary and require confirmation by larger epidemiological studies. Such studies could provide a rationale for determining the GPIa C₈₀₇T polymorphism to stratify at-risk patients for different antiplatelet treatments. Furthermore, GPIa-IIa itself might be an interesting target for long-term pharmacologic inhibitors of platelet function in high-risk patient populations.

The skillful technical assistance work of C. Blumentritt and the biometric advice of Dr B. Jäger and Dr M. Wodny are highly appreciated. We thank Prof Dr R. Walther for support in purification of patient DNA. The study is part of the Community Medicine program of the Medical Faculty of the Ernst-Moritz-Arndt-University Greifswald. The language editorial help of S. Owens is appreciated.