|
| |||||||||||||||||||||||||||||||||||||||||||||
Background High blood levels of coagulation factor VII are associated with a risk of ischemic vascular disease. Although factor VII levels may be genetically determined, the relation between genetic polymorphisms of factor VII, factor VII blood levels, and the risk of myocardial infarction has not been established.
Methods We performed a casecontrol study of 165 patients with familial myocardial infarction (mean [±SD] age, 55±9 years) and 225 controls without a personal or family history of cardiovascular disease (mean age, 56±8 years). The polymorphisms involving R353Q and hypervariable region 4 of the factor VII gene were studied. Factor VII clotting activity and antigen levels were also measured.
Results Patients with the QQ or H7H7 genotype had a decreased risk of myocardial infarction (odds ratios, 0.08 [95 percent confidence interval, 0.01 to 0.9] and 0.22 [95 percent confidence interval, 0.08 to 0.63], respectively). For the R353Q polymorphism, the RR genotype was associated with the highest risk, followed by the RQ genotype and then by the QQ genotype (P<0.001). For the polymorphism involving hypervariable region 4, the combined H7H5 and H6H5 genotypes were associated with the highest risk, followed in descending order by the H6H6, H6H7, and H7H7 genotypes (P<0.001). Patients with the QQ or H7H7 genotype had lower levels of both factor VII antigen and factor VII clotting activity than those with the RR or H6H6 genotype. Patients with the lowest level of factor VII clotting activity had a lower risk of myocardial infarction than those with the highest level (odds ratio, 0.13; 95 percent confidence interval, 0.05 to 0.34).
Conclusions Our findings suggest that certain polymorphisms of the factor VII gene may influence the risk of myocardial infarction. It is possible that this effect may be mediated by alterations in factor VII levels.
Factor VII blood levels are influenced by both environmental and genetic factors. Triglyceride levels are a major determinant of factor VII levels in blood.6 Age, body-mass index, oral contraceptive use, and menopausal status have also been related to the levels and activity of factor VII.7 Recent studies have demonstrated that the levels of factor VII and their response to environmental stimuli are genetically determined. Green et al.8 reported a strong association between a common polymorphism (R353Q) of the factor VII gene and plasma factor VII levels; however, there was no association between the polymorphism and the risk of ischemic vascular disease.9 Another common polymorphism has been described in hypervariable region 4 of intron 7 of the factor VII gene and reported to be associated with levels of factor VII.10,11
We investigated whether the risk of myocardial infarction is associated with polymorphisms in the factor VII gene, and whether these polymorphisms are associated with factor VII levels. To emphasize the contribution of genetic factors, we studied patients with a family history of thrombosis.
Methods
Study Population
Case patients were patients over 45 years of age who had had a myocardial infarction and who were reported to have at least one first-degree relative who had had a myocardial infarction or a stroke (or both) before the age of 65 years.12 The patients were selected from the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico trial population13 on the basis of an interview regarding their family history of thrombosis.14 Of the 239 case patients identified, 180 patients could be contacted and were willing to participate.
Controls were consecutive patients over the age of 45 years who were hospitalized for any clinical reason except myocardial infarction, stable or unstable angina, stroke, or transient ischemic attacks. Among the 309 control patients interviewed, 37 who reported a personal or family history of thrombosis or who had definite defects of the hemostatic system were excluded. Patients with chronic diseases were not considered unless they were hospitalized for reasons unrelated to these conditions.
We used a structured questionnaire to characterize both case patients and controls.14 The subjects were all Italian: 40 percent of both case patients and controls were from northern Italy, 15 percent from central Italy, 40 percent from southern Italy, and 5 percent from Sardinia. All patients provided written informed consent. The study was performed according to the Declaration of Helsinki of 1975 and was approved by the Mario Negri Sud ethics committee.
Laboratory Measurements and Techniques
The case patients were assessed five to seven months after their most recent ischemic event. Blood was collected from the patients between 8 and 10 a.m. after an overnight fast and after a 20-minute rest in the supine position. Blood was not collected from nine patients who were receiving oral anticoagulant drugs. Blood samples for DNA and biochemical analyses were available for 165 of 171 case patients (96 percent) and 225 of 272 controls (83 percent).
Hypervariable region 4 of intron 7 of the factor VII gene was amplified according to a modification of the method of Marchetti et al.15 Three alleles were identified: a common allele (H6) of 443 bp with six monomers, a less frequent allele (H7) of 480 bp with seven monomers of 37 bp, and a very rare allele (H5) of 406 bp with five monomers (Figure 1).
|
Factor VII clotting activity was determined by a one-stage clotting assay as follows: 100 µl of plasma was diluted 1:10 in TRIShydrochloride buffer (50 mmol of TRIS per liter and 100 mmol of sodium chloride per liter, pH 7.4), mixed with 100 µl of human factor VIIdeficient plasma (Sigma, St. Louis), and incubated for five minutes at 37°C in a coagulometer (model KC 4A, Amelung, Lemgo, Germany). Then, 200 µl of a prewarmed (37°C) 1:1 mixture of recombinant human thromboplastin (Hemoliance, Cologno Monzese, Italy) and calcium dichloride (25 mmol per liter) was added, and the clotting time was automatically recorded on the coagulometer. Values were expressed as a percentage of the value in pooled normal plasma.
Plasma factor VII antigen was measured as total factor VII by a double-antibody enzyme-linked immunosorbent assay (Boehringer Mannheim Italia, Milan, Italy) that reacts with all species of factor VII.
Statistical Analysis
The variables factor VII clotting activity and antigen and triglyceride levels were analyzed on a logarithmic scale to remove positive skewness. The means were compared by analysis of variance or the KruskalWallis test. Chi-square analysis or Fisher's exact test was used to compare discrete variables. The frequencies of the alleles and genotypes among case patients and controls were determined and compared by the chi-square test with the values predicted on the basis of the assumption of HardyWeinberg equilibrium. The coefficient of gametic linkage disequilibrium was calculated by likelihood methods in the control sample.16 The coefficient (D') is reported as the ratio of the unstandardized coefficient to its maximal value. Odds ratios were calculated, together with their 95 percent confidence intervals. Dummy variables were created to assess the association of the various genotypes with familial myocardial infarction. Multiple logistic-regression analysis was performed with the Logistic procedure; confounding variables included in the model were age, sex, smoking status, and history of hyperlipidemia, hypertension, and diabetes. In a separate analysis, the effect of the genotype on the risk of myocardial infarction was also adjusted for factor VII activity. The effects of the H7 and H5 alleles, with H6 chosen as the reference allele, and the effect of the Q allele, with R chosen as the reference allele, were analyzed by the introduction of three dummy variables (0, 1, and 2) coding respectively for the number of H7, H5, and Q alleles. This model specifies a codominant multiplicative effect of the H7, H5, and Q alleles in determining the risk of myocardial infarction.
The association of factor VII levels with the genotypes was tested by one-way analysis of variance and covariance (with age, sex, and triglyceride levels as covariates) with use of the general linear model procedure. The TukeyKramer approach was used for multiple comparisons. Data for continuous variables were expressed as means ±SD; a two-tailed P value of less than 0.05 was considered to indicate statistical significance. All computations were carried out with the SAS statistical package (SAS Institute, Cary, N.C.).17
Results
Characteristics of the Study Population
The characteristics of the case patients and controls are shown in Table 1. There were no significant differences in age between the groups. The patients with familial myocardial infarction had a higher prevalence of common risk factors for atherosclerotic disease, as well as significantly higher levels of both factor VII clotting activity and factor VII antigen, than the control group.
|
The genotypes and alleles at the factor VII loci in case patients and controls are shown in Table 2. The distribution of the factor VII R353Q and hypervariable region 4 genotypes was virtually identical to that predicted on the basis of HardyWeinberg equilibrium, both in case patients and in controls (chi-square with 1 df = 3.1, P = 0.08, and chi-square with 1 df = 0.01, P = 0.9, respectively, for R353Q; chi-square with 2 df = 1.5, P = 0.5, and chi-square with 2 df = 0.6, P = 0.7, for hypervariable region 4). The two polymorphisms were in linkage disequilibrium, with a D' value of 0.65 (P<0.001).
|
The allele frequencies of both polymorphisms also differed between case patients and controls (Table 2). Allele Q of the R353Q polymorphism was less frequent among case patients than controls (chi-square with 1 df = 4.3, P = 0.04). Allele H7 of the hypervariable region 4 polymorphism was significantly less frequent among case patients (chi-square with 1 df = 6.9, P = 0.008), whereas the relative frequencies of H6 (chi-square with 1 df = 4.0, P = 0.05) and H5 (P = 0.04 by Fisher's exact test) were higher among case patients.
Odds Ratios for Familial Myocardial Infarction
The specific odds ratios for myocardial infarction with each R353Q and hypervariable region 4 genotype, with RR and H6H6 as the reference groups, are shown in Table 3. Homozygosity for the Q or H7 allele conferred significant protection from myocardial infarction. Indeed, patients with the QQ or H7H7 genotype had a decreased risk of myocardial infarction of 92 percent and 78 percent, respectively, after adjustment for age, sex, smoking status, and history of hyperlipidemia, diabetes, and hypertension. However, when these analyses were also adjusted for factor VII activity levels, the cumulative effect of factor VII genotypes in decreasing the risk of myocardial infarction was no longer significant (odds ratio for the QQ genotype, 0.18; 95 percent confidence interval, 0.01 to 2.17; odds ratio for the H7H7 genotype, 0.58; 95 percent confidence interval, 0.21 to 1.93), suggesting a role for factor VII in mediating the effect. For the R353Q polymorphism, the QQ genotype was associated with the lowest risk, followed by the RQ genotype, with the RR genotype as the reference group, in univariate (P = 0.02) and multivariate (P< 0.001) analyses. For the polymorphism involving hypervariable region 4, the highest risk was associated with the combined H7H5 and H6H5 genotypes, followed in descending order by the H6H6 (odds ratio, 1.0; the reference group), H6H7, and H7H7 genotypes in univariate (P = 0.008) and multivariate (P<0.001) analyses.
|
Sixty-eight percent of the patients reported a family history of myocardial infarction alone, 25 percent had a family history of both myocardial infarction and stroke, and the remaining 7 percent had a family history of stroke alone. When the patients with a family history of stroke were excluded, the QQ genotype (odds ratio, 0.13; 95 percent confidence interval, 0.02 to 1.05) and the H7H7 genotype (odds ratio, 0.42; 95 percent confidence interval, 0.20 to 0.88) were both associated with a lower risk of myocardial infarction. This relation was also found when only patients with a family history of myocardial infarction were analyzed (QQ genotype: odds ratio, 0.21; 95 percent confidence interval, 0.02 to 1.13; H7H7 genotype: odds ratio, 0.40; 95 percent confidence interval, 0.16 to 0.99).
Table 4 shows the effect of various combinations of genotypes for hypervariable region 4 and R353Q as compared with the RRH6H6 combination. The H7H7 genotype decreased the risk of myocardial infarction in combination with any R353Q genotype; homozygosity for both the Q and H7 alleles afforded the greatest protection. However, the small number of QQ carriers did not allow a formal statistical analysis to verify the effect of such combinations of genotypes.
|
In a multivariate analysis of variance, with age, sex, and triglyceride levels as covariates, the factor VII polymorphisms involving R353Q and hypervariable region 4 were significantly associated with the levels of factor VII clotting activity and antigen (Table 5). The effect was more evident for factor VII clotting activity in controls, owing to the small number of case patients with the QQ and H7H7 genotypes. Patients with the QQ or H7H7 genotype had the lowest levels of factor VII clotting activity and antigen, whereas those with the RR or H6H6 genotype had the highest levels of factor VII clotting activity. Analysis for trend and multiple comparisons showed a codominant effect of the R and Q alleles, whereas the effect of the H7 allele was recessive. The association between factor VII levels and the combination of polymorphisms showed the same trend (Table 4): factor VII levels were the lowest in patients with the QQH7H7 combination and highest in those with the RRH6H6 combination.
|
|
We found that polymorphisms of the factor VII gene are associated with the risk of familial myocardial infarction. The frequencies of the Q allele of R353Q and of the H7 allele of polymorphisms involving hypervariable region 4 were significantly lower in patients with a family history of thrombosis than in controls, suggesting that these alleles protect against familial myocardial infarction. The presence of the H5 allele seems to be associated with an increase in the risk of familial myocardial infarction, but because of the very low frequency of this allele, a larger sample is necessary to confirm this association.
Patients with a family history of myocardial infarction or stroke (or both) were selected for this study because genetic variants related to the development of thrombosis should be found more frequently in these patients than in patients without such a family history. To increase the contrast further, subjects with a personal or family history of vascular disease were excluded from the control group. These selection criteria, together with ethnic differences, can explain the discrepancy between our results and the inconclusive results of other studies.9,18,19
The distribution of the genotypes in the control group was similar to that recently reported by Bernardi et al.20 in the general population in a region of central Italy and to that which we found in a larger sample of the Italian nonhospitalized population.21 However, it differed significantly from that found in other populations, such as in the Netherlands, Denmark, and the United Kingdom, where the frequencies of the rare alleles of both polymorphisms were lower.9,18,22 Since these rare alleles are associated with low levels of factor VII, the finding of lower frequencies could be in agreement with the higher incidence of myocardial infarction in such countries as compared with Italy,23 further supporting the role of these polymorphisms in the development of or protection against myocardial infarction.
Both polymorphisms were significantly related to factor VII clotting activity and antigen levels, suggesting that the association between factor VII genotypes and myocardial infarction may occur as a result of the modulation of factor VII blood levels. Indeed, patients with the QQ or H7H7 genotype, who had the lowest risk of myocardial infarction, also had the lowest levels of factor VII. Moreover, the association of factor VII genotypes with a decreased risk of myocardial infarction was no longer significant once the multivariate analyses were adjusted for factor VII clotting activity, further supporting a role for the activation of factor VII in mediating the effect.
The way in which the factor VII polymorphisms involving R353Q and hypervariable region 4 modulate the levels of factor VII is only a matter of speculation. The amino acid substitution caused by the R353Q polymorphism has been shown to influence the correlation between plasma triglyceride levels and factor VII levels.24 Moreover, the R353Q polymorphism is in strong linkage disequilibrium with a polymorphism in the factor VII promoter, whose functional relevance has been demonstrated by transfection experiments in hepatoma cells.25
It is conceivable that the polymorphism involving hypervariable region 4 is in linkage disequilibrium with the R353Q polymorphism. However, this DNA locus contains a consensus splice sequence at the 5' repeat, which, even if it is not translated, could be important in regulating the splicing of messenger RNA as it is formed.26,27
In conclusion, our study supports the possibility of genetic control of myocardial infarction. Certain polymorphisms of the factor VII gene may protect against familial myocardial infarction, possibly by modifying factor VII blood levels. Prospective clinical trials are needed to investigate the clinical and therapeutic implications of our results.
Supported in part by a contract (960104375) with the Italian National Research Council (Progetto Strategico "Myocardial Infarction"). Dr. Di Castelnuovo is the recipient of a fellowship from the Centro di Formazione e Studi per il Mezzogiorno Progetto Speciale Ricerca Scientifica e Applicata nel Mezzogiorno.
We are indebted to our colleagues on the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico Scientific Committee, who shared in the initial planning of the study; to Dr. Giovanni de Gaetano for his helpful suggestions; to Dr. Anna Falanga for the evaluation of factor VII antigen; and to Prof. P. Brakman, Dr. Iacoviello's thesis advisor at the University of Leiden.
Source Information
From the Unit of the Genetics of Vascular Risk Factors, Angela Valenti Laboratory of Thrombosis Pharmacology, Department of Vascular Medicine and Pharmacology (L.I., A.D.C., A.D'O., C.A., M.B.D.), and the Laboratory of Drug Epidemiology, Department of Clinical Pharmacology and Epidemiology (R.A.), Istituto di Ricerche Farmacologiche Mario Negri, Santa Maria Imbaro, Italy; and Gaubius Laboratory (L.I., P. de K., C.K.) and Medical Center of Genetics Department of Human Genetics, Leiden University (P. de K.), Leiden, the Netherlands.
Address reprint requests to Dr. Iacoviello at Angela Valenti Laboratory of Thrombosis Pharmacology, Department of Vascular Medicine and Pharmacology, Consorzio Mario Negri Sud, 66030 Santa Maria Imbaro, Italy.
References
The following clinical centers and investigators participated in the study: Asti General Hospital, Asti (M. Alciati); Avellino General Hospital, Avellino (G. Amoroso); Barletta General Hospital, Barletta (A. Messina); Sant'Orsola Hospital, Bologna (G. Palareti); Bozzolo General Hospital, Bozzolo (E. Franzì); Cagliari General Hospital (M. Sias) and Cagliari University Medical School (F. Marongiu), Cagliari; Casale Monferrato General Hospital, Casale Monferrato (M. Pezzana); Casarano General Hospital, Casarano (S. Ciricugno); Caserta General Hospital, Caserta (R. Di Sarno); Cento General Hospital, Cento (L. Orselli); Colleferro General Hospital, Colleferro (E. Venturini); Copertino General Hospital, Copertino (A. Calcagnile); Desio General Hospital, Desio (G. Iacuitti); Fidenza General Hospital, Fidenza (S. Callegari); Grosseto General Hospital, Grosseto (A. Cresti); Guastalla General Hospital, Guastalla (V. Manicardi); Lanciano General Hospital, Lanciano (A. Valerio); Legnago General Hospital, Legnago (P. Todesco); Leno General Hospital, Leno (A. Lanzini); Lodi General Hospital, Lodi (C. Pezzi); Lugo General Hospital, Lugo (T. Tognoli); Magenta General Hospital, Magenta (R. Turato); Mantova General Hospital, Mantova (A. Izzo); Mestre General Hospital, Mestre (G. Gasparini); Sacco Hospital (E. Rossi) and Niguarda Hospital (C. Corsini), Milan; Mirano General Hospital, Mirano (A. Zanocco); Monza General Hospital, Monza (F. Achilli); Cardarelli Hospital (F. Piantadosi) and Second University Medical School (D. De Lucia), Naples; Novi Ligure General Hospital, Novi Ligure (L. Fascioli); Nuoro General Hospital, Nuoro (G. Tupponi); Cervello Hospital (A. Ledda, I. Greco) and Benfratelli Hospital (R. La Malfa), Palermo; Perugia General Hospital, Perugia (S. Brando); Pescara General Hospital, Pescara (T. Bonfini); Pescia General Hospital, Pescia (L. Iacopetti); Piombino General Hospital, Piombino (S. Bechi); Pisa General Hospital, Pisa (U. Conti); Rieti General Hospital, Rieti (S. Orazi); Rimini General Hospital, Rimini (F. Bologna); First University Medical School, Rome (P. De Paolis); San Donà di Piave General Hospital, San Donà di Piave (P. Della Valentina); Savona General Hospital, Savona (A. Gandolfo); Casa Sollievo della Sofferenza, San Giovanni Rotondo (A. Villella); Termoli General Hospital, Termoli (M. Esposito); Maria Vittoria Hospital, Turin (L. Mussano); Treviso General Hospital, Treviso (S. Perissinotto); Udine General Hospital, Udine (G. Feruglio, C. Fresco); Vasto General Hospital, Vasto (E. Bottari); and Fondazione S. Maugeri, Veruno (F. Soffiantino, M. Gattone).
| |||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |