- Email: [email protected]
Prevalence of resistance against activated protein C resulting from factor V leiden is significantly increased in myocardial infarction: Investigation of 507 patients with myocardial infarction
Genetics
Prevalence of resistance against activated protein C resulting from factor V Leiden is significantly increased in myocardial infarction: Investigation of 507 patients with myocardial infarction Katharina Middendorf, MD,a Peter Go ¨ hring, MD,b Tanya Y. Huehns, MD,c Dietrich Seidel, Prof,b a Gerhard Steinbeck, Prof, and Sigrid Nikol, Profa,d Mu ¨ nster and Munich, Germany, and London, Great Britain
Background A point mutation in the gene encoding coagulation factor V is a cause of resistance against activated protein C. The presence of factor V Leiden is linked to 50% of congenital defects causing venous thrombosis. Its relationship to arterial thrombosis, particularly to myocardial infarction, has not been defined. Therefore, we performed a study on the role of factor V Leiden in patients with myocardial infarction. The study was carried out in Bavarians of German origin, a relatively homogeneous population. Methods and Results The study group consisted of 507 patients with documented myocardial infarction (77.5% (393/507) men, 22.5% (114/507) women), with a mean age of 56.1 (range 18-– 86) years. Strict criteria for patient selection and highly sensitive and specific functional tests for factor V Leiden were used. In addition, all patients with pathological test results were genotyped. The prevalence of factor V Leiden in patients with myocardial infarction was 8.7% (44/507), a significant increase in the prevalence of this mutation compared with the control group (3.7%, P ⫽ .0025). The odds ratio was 2.46 (95% CI 1.35– 4.50). Conclusions A significantly increased prevalence of factor V Leiden in patients with documented myocardial infarction was seen. Patients with this mutation appear to have a predisposition for myocardial infarction. (Am Heart J 2004; 147:897–904.) Resistance against activated protein C commonly occurs in the presence of a mutation producing the molecule termed factor V Leiden. Its presence leads to resistance against activated protein C: activated factor V resists inactivation by activated protein C, which results in a potentially procoagulant state.1 Factor V Leiden is associated with 50% of congenital defects that result in increased thrombosis in the venous system. This is often seen in combination with other initiFrom the aMedical Clinic I, Klinikum Großhadern, Ludwig Maximilian University, Munich, Germany, bInstitute of Clinical Chemistry, Klinikum Großhadern, Ludwig Maximilian University, Munich, Germany, cBritish Heart Foundation, Cardiovascular Medicine, Imperial College School of Technology and Medicine, Hammersmith Hospital, London, Great Britain, and dDepartment of Cardiology and Angiology, University of Mu¨nster, Mu¨nster, Germany. Submitted April 21, 2003; accepted October 17, 2003. Reprint requests: Prof Dr med Sigrid Nikol, Medizinische Klinik und Poliklinik C (Kardiologie und Angiologie), Universita ¨ tsklinikum Mu¨nster, Albert-Schweitzer-Str 33, D-48129 Mu¨nster, Germany. E-mail: [email protected] 0002-8703/$ - see front matter © 2004, Elsevier Inc. All rights reserved. doi:10.1016/j.ahj.2003.10.048
ating factors, such as during pregnancy and periods of immobilization following operations. The role of factor V Leiden in thrombosis in relation to arterial disease remains controversial. Patients who are heterozygous for factor V Leiden and have a myocardial infarction have less than the expected amount of coronary atherosclerosis, suggesting that the prevalence of factor V Leiden may predispose to myocardial infarction at an earlier stage of coronary disease.1 While some researchers have not found an association between resistance against activated protein C caused by the presence of factor V Leiden and myocardial infarction,2–18 others, especially in more recent publications, have suggested that there is a relationship.1,13,19 –27 The availability of improved functional tests in these more recent studies suggests that the relationship may be a real one. A total of 90% to 95% of patients with resistance against activated protein C have the FV506Q mutation (leading to the molecule termed factor V Leiden);28 –30 in the remaining 5% to 10%, increased lupus anticoagu-
American Heart Journal May 2004
898 Middendorf et al
Table I. Characteristics of 19 publications investigating the association between factor V Leiden and myocardial infarction—3 of the studies consisted of 2 to 4 single study groups Characteristics of patients with MI First author, year of appearance
Study design
Test for FVL
Patient number
Amowitz 19992 Ardissino 19963
Case-control Case-control
Genotyping Genotyping
36 100
Baranovskaya 199819
Case-control
Genotyping
Van Bockxmeer 19954 Van der Bom 19965 Cushman 19986
Case-control Case-control Prospective
Genotyping Functional, genotyping Genotyping
287 168/287 119/287 149* 114 149
Dacosta 19987
Case-control
Functional, genotyping
75
Demarmels Biasiutti 19958
Case-control
Not modified functional
134
Doggen 199820 Emmerich 19959
Case-control Case-control
Genotyping Genotyping
Eritsland 199510 Holm 199623
Case-control Case-control
Functional Functional, genotyping
560 643 197/643 96/643 207/643 143/643 281† 101
Holm 199924 Inbal 199912 Kontula 199513
Case-control Case-control Case-control
Genotyping Genotyping Genotyping
23‡ 112 122 51/122 71/122
Prohaska 199515
Case-control
Functional, genotyping
134§
Redondo 199916 Ridker 199517 Rosendaal 199727
Case-control Prospective Case-control
Modified functional, genotyping Genotyping Genotyping
177㛳 374 84
m/f
Age (y)
Prevalence(%)
Only f 96:4 96%:4% ?
Ø 39 Ø 40.2 (⬍45)
2.8 (1/36) 1.0 (1/100)
Ø 60 (27–91) ⬍65 ⱖ65
? 92:57 62%:38% 66:9 88%:12% 115:19 86%:14% Only m Only m
⬎55 Ø 74.3 (⬎65)
3.1 (9/287) 0.6 (1/168) 6.7 (8/119) 4.7 (7/149) 3.5 (4/114) 3.4 (5/149)
Ø 38.0 (⬍45)
5.3 (4/75)
Ø 54.9 (32–72)
2.2 (3/134)
Ø 56.2 25–64
6.8 (38/560) 5.3 (34/643) 2.5 (5/197) 4.2 (4/96) 10.6 (22/207) 2.1 (3/143) 1.8 (5/281) 17.8 (18/101)
79:22 78%:22% only f only m 107:15 88%:12% 49:2 96%:4% 58:13 82%:18%
Ø 44 (28–50)
⬍52
30.4 (7/23) 6.3 (7/112) 5.7 (7/122)
⬍45
2.0 (1/51)
⬍60
8.5 (6/71) 7.5 (10/134)
Only m Only f
? 40–84 (⬎40) Ø 39.6 (23–44)
6.2 (11/177) 6.1 (23/374) 9.5 (8/84)
FVL, factor V Leiden; MI, myocardial infarction; functional, functional test for resistance against APC which is not specific for the detection of factor V Leiden; m:f, ratio between men and women; ?, no more information given in the publication. *From a total collective of 222 patients with ischemic heart disease including 149 patients with MI; of all patients 87% were (193/222) men and 13% (29/222) women under the age of 50 years; the control group consisted of 126 persons. †From a total collective of 610 patients with coronary artery bypass surgery including 281 patients with MI; of all patients 87% were (531/610) men and 13% (79/610) women with a mean age of 60 years; the control group consisted of 265 persons. ‡From a total collective of 295 patients with acute coronary syndrome including 86 patients with MI and 210 patients with unstable angina, of all patients 64% were (190/ 295) men and 36% (105/295) women aged between 51 and 73 years; 23 of patients with MI were women; the control group consisted of 423 fertile women. §From a total collective of 317 patients with angiographically diagnosed coronary artery disease including 134 patients with MI, of all patients 85% were (269/317) men and 15% (48/317) women with a mean age of 58 years (range from 28 to 85 years). 㛳From a total collective of 200 patients with coronary artery disease including 177 patients with MI tested for factor V Leiden, of all patients 87% were (174/200) men and 13% (26/200) women; the control group consisted of 100 persons including 90 persons tested for factor V Leiden.
lants, increased factor VIII, or the effects caused by pregnancy or oral contraceptives are responsible.30 In the mutation for factor V Leiden, arginine at position 506 is replaced by glutamine (or guanine at position 1691 by adenine) in the factor V molecule.31 The mutation results in an alteration in the activated factor V activated protein C cleavage site. Hence, factor V Lei-
den is inactivated by proteolysis 10 to 20 times slower than the native form of factor V. The phenotype can be determined using the original resistance against activated protein C functional test,28 or a modified version that has the advantage of nearly 100% sensitivity for factor V Leiden.32,33 Additional genotyping allows confirmation of factor
American Heart Journal Volume 147, Number 5
Middendorf et al 899
Table I. Continued Characteristics of controls Patient number 36 100 483 373/483 110/483 126 222 482
53 100 646 726 178/726 148/726 193/726 207/726 265 101 423 187 137 50/137 87/137 190 90 704 388
m/f
Age (y)
Only f ? ?
? Ø 41.6
Prevalence (%)
Cardio-vascular risk factors Yes Yes No
Yes Yes Yes
0.31 (0.03–3.17) 0.49 (0.04–5.55) 1.08 (0.46–2.54) 0.17 (0.02–1.28) 7.86 (0.97–63.87) 1.19 (0.37–3.86) 0.70 (0.22–2.24) 0.46 (0.18–1.19)
OR (95% CI)
? 178:304 37%:63%
? Ø 72.3
8.3 (3/36) 2.0 (2/100) 2.9 (14/483) 3.5 (13/373) 0.9 (1/110) 4.0 (5/126) 5.0 (11/222) 7.1 (34/482)
46:7 87%:13% 87:13 87%:13% Only m ?
Ø 36.5
3.8 (2/53)
Yes
1.44 (0.25–8.15)
Ø 54.9 (32–74)
2.0 (2/100)
Yes
1.12 (0.18–6.84)
Ø 57.3 ?
5.0 (32/646) 4.7 (34/726) 5.6 (10/178) 0.7 (1/148) 8.8 (17/193) 2.9 (6/207) 2.6 (7/265) 10.9 (11/101)
Yes No
No No
1.40 (0.86–2.27) 1.14 (0.70–1.85) 0.44 (0.15–1.31) 6.39 (0.70–58.07) 1.23 (0.63–2.40) 0.72 (0.18–2.92) 0.67 (0.21–2.13) 1.77 (0.79–3.40)
Yes Yes Yes
3.87 (1.51–9.92) 0.97 (0.37–2.55) 2.02 (0.58–7.09)
Ø 11 (6–17) Ø 80
?
?
Only f Only m 95:42 69%:31% 42:8 84%:16% 53:34 61%:39% 116:74 61%:39% ? Only m Only f
fertile age ?
10.2 (43/423) 6.4 (12/187) 2.9 (4/137)
?
6.0 (3/50)
0.31 (0.03–3.12)
?
1.1 (1/87)
7.94 (0.93–67.57)
Ø 34 (18–65)
9.5 (18/190)
No
0.77 (0.34–1.73)
? 40–84 (⬎40) Ø 37.7 (19–44)
6.7 (6/89) 6.0 (42/704) 4.1 (16/388)
Yes Yes Yes
0.93 (0.33–2.60) 1.03 (0.61–1.75) 2.45 (1.01–5.92)
V Leiden and detection of homozygotes and heterozygotes. Patient numbers in previously published studies may have generally been too low to reach significance, partly because of high regional differences in the prevalence of the underlying genetic defect (1%–15%) and also because of the ethnic heterogeneity of enrolled patients. This study aimed to investigate the role of factor V Leiden in a more homogeneous group of patients with myocardial infarction, using a phenotyping test with high sensitivity, together with confirmation of the presence of the factor by subsequent genotyping. To interpret results of the study, data were related to a meta-analysis of previous publications.
Methods Patients The patient population comprised 507 patients who presented to a university hospital in Munich between September 1996 and March 2000. Inclusion criteria were documented prior myocardial infarction (increased creatine kinase [CK] with CK-MB ⬎10% or proof of increased troponin I, typical alteration in the ECG, and typical clinical symptoms), with the first myocardial infarction date accurately recorded, together with written consent for the molecular genetic analysis (required in Bavaria since March 1999; patients prior to this consented orally). One of the consecutive patients was excluded because he refused written consent. Patients with venous thromboembolisms and non-Germans were excluded (family name was used as a proxy-variable for German nationality).
American Heart Journal May 2004
900 Middendorf et al
Patients had a mean age of 56.1 years (range 18 to 86 years); 77.5% (393/507) were men and 22.5% (114/507) were women. Of the total, 76.3% (387/507) had survived a single myocardial infarction, and 23.7% (120/507) had survived multiple myocardial infarctions. Among the 666 myocardial infarctions experienced by patients in the study, 36.2% (241/ 666) were anterior, 36.5% (243/666) were posterior, and 27.3% (182/666) were myocardial infarctions without known location. The main cardiovascular risk factors, including hyperlipidemia, arterial hypertension, smoking, and family history were recorded in one third of patients. The genetic defect leading to the presence of factor V Leiden was diagnosed in the enrolled patient population as described below.
Control group The control population comprised 404 patients from the same geographical region who presented to a university hospital in Munich with other diagnoses not involving thrombotic events between September 1996 and March 2000. Patients with venous and arterial thromboembolisms were excluded. Controls in this group had a mean age of 54.4 years (range 12 to 88 years); 68.3% (276/404) were men and 31.7% (128/404) were women. The genetic defect leading to the presence of factor V Leiden was diagnosed in the enrolled control population in the same way as in the investigated group with myocardial infarctions, as described below.
Methods for analyses The same factor V Leiden test used for the control group was used for the patients with myocardial infarction enrolled in this study. The modified test used contains factor V deficient plasma for the dilution of the test plasma. It provides 100% sensitivity and about 98% specificity for FV:Q506 and is more sensitive than the activated protein C screening test used previously.28 The APCTM-Resistance V kit (COATEST, Chromogenix, Milan, Italy) was used in combination with a STA coagulation analyser (Roche/Stago, Mannheim, Germany). This gives a ratio of the activated partial thromboplastin time (aPTT) with and without addition of activated protein C. A ratio of ⬍2.1 was defined as resistance against activated protein C. Differentiation between normal and pathological phenotype is possible in nearly 100% of cases,32,33 and conclusions regarding the pathology of the genotype (factor V Leiden) can also be made.32 However, differentiation between the homozygous and heterozygous form is not always possible and therefore all patients with pathological results were then genotyped following this test. Genotyping was performed as follows: DNA was extracted from Na2EDTA-blood. A fragment of the factor V gene containing the Leiden mutation site was amplified using a polymerase chain reaction. The following oligonucleotide primers were used to amplify this fragment in exon 10 encoding the activated protein C cleavage site of the factor V molecule: 5⬘-GGAACAACACCATGATCAGAGCA-3⬘ and 5⬘-TAGCCAGGAGACCTAACATGTTC-3⬘.34 After restrictive digestion of the amplified DNA with Mnl I, the enzymatically digested fragments were separated by agarose gel electrophoresis and visualized with ethidium bromide.
Statistics Results are presented as means together with the standard error of the mean (SEM). The significance of the hypothesis
relating to the noncontinuous characteristics was investigated by 2 testing. The odds ratio (OR) with its logarithmus naturalis (ln) and 95% CI were calculated adjusted for age, gender, and nationality to reflect the risk of myocardial infarction in the presence of factor V Leiden compared with the risk in the absence of this factor.
Meta-analysis A search of the online database medline.de using the terms “factor V Leiden, resistance against activated protein C, myocardial infarction” yielded 200 publications. Inclusion criteria were documented myocardial infarction in the patient group, known prevalence of factor V Leiden in patients with myocardial infarction and in controls, and similar design characteristics. Publications were excluded if the controls recording the prevalence of factor V Leiden failed to originate from the same geographical region,1,11,14,18,22,25 or if missing data meant that calculation of the OR was impossible.26 The 19 studies included published between 1995 and 1999 are tabulated in Table I; 3 of the studies consisted of 2 to 4 single study groups. Ln (OR) and CI from each trial were graphically displayed by a Forrest plot to show heterogeneity and quality of published results.35
Results Factor V Leiden testing The prevalence of the genetic defect leading to factor V Leiden in the Bavarian control group was 3.7% (15/404). The modified functional test for factor V Leiden showed a mean activated protein C ratio of 2.3 (SEM ⫾ 0.3) in all patients presenting with myocardial infarction. In patients with myocardial infarction and genotyping positive for heterozygous or homozygous factor V Leiden, the average activated protein C ratio was 1.5 (SEM ⫾ 0.2). A pathological genotype was always associated with pathological values in the functional test. The prevalence of factor V Leiden in patients with myocardial infarction was 8.7% (44/507), with 2 homozygous and 42 heterozygous patients. The 2 test indicates a significant increased prevalence of factor V Leiden compared with the prevalence in the Bavarian control population (8.7% vs 3.7%, P ⫽ .0025; Figure 1). The crude OR was 2.46 (95% CI 1.35– 4.50); the adjusted OR for age, gender, and nationality is shown in Table II.
Patients with myocardial infarction and factor V Leiden Of these 44 patients, 68.2% (30/44) were men and 31.8% (14/44) were women, with a mean age of 54.5 (range 35– 84) years at the time of the first myocardial infarction; 16 patients were aged ⱕ50 years. A total of 75.0% (33/44) had survived a single myocardial infarction and 25.0% (11/44) had survived multiple myocardial infarctions. Among the 59 documented myocardial infarctions, 42.4% (25/59) were anterior, 30.5% (18/
American Heart Journal Volume 147, Number 5
Figure 1
Middendorf et al 901
Table II. Adjusted odds ratio and P value Adjusted OR
Factor V Leiden Age Male German nationality
OR
95% CI
2
P
2.66 1.01 1.61 0.84
1.45–4.88 1.00–1.02 1.20–2.18 0.47–1.50
9.91 3.63 9.82 0.36
.0016 .0569 .0017 .5487
Table III. Comparison of cardiovascular risk factors in all patients with myocardial infarction and patients with myocardial infarction and factor V Leiden Prevalence of factor V Leiden in patients with myocardial infarction and controls from the same geographical region.
59) were posterior, and 27.1% (16/59) were myocardial infarctions without known location. The average age at the time of the first myocardial infarction was 56.1 years in patients with myocardial infarction, and 54.5 years in the subgroup with both myocardial infarction and factor V Leiden. There was no significant association between factor V Leiden and the age of first myocardial infarction. There was no significant association between factor V Leiden and female sex in patients with myocardial infarction (31.8%) compared with all women with myocardial infarction (22.5%). There were no significant differences in identified cardiovascular risk factors in the patients with myocardial infarction and factor V Leiden compared with patients with myocardial infarction and no factor V Leiden (Table III). There was a trend towards a lower prevalence of hyperlipidemia, arterial hypertension, and smoking in patients with myocardial infarction and factor V Leiden compared with the total patient population with myocardial infarction.
Results of meta-analysis Table I summarizes the characteristics of the patients and controls included, and details the methods for detecting factor V Leiden used in each study. Comparison of the 19 studies demonstrated an OR total of 1.18 (95% CI 0.97–1.43) when investigating the risk of myocardial infarction in patients possessing factor V Leiden compared with the control patients. Using the 2 test, no significant heterogeneity in the publications (P ⫽ .3394) was found between the 19 studies. The Forrest plot compares our results with the 19 studies reported in the literature (Figure 2). The results show a random distribution, with the various studies showing mark-
14 Patients with 168 Patients MI and factor V with MI Leiden 2 Hyperlipidemia Arterial hypertension Smoking Family history No risk factors
66.7% (112/168) 61.9% (104/168) 47.0% (79/168) 29.8% (50/168) 3.0% (5/168)
50.0% (7/14) 50.0% (7/14) 35.7% (5/14) 21.4% (3/14) 0.0% (0/14)
P
1.91 .1671 0.92 .3380 0.78 .3759 0.51 .4763 0.47 .4937
edly different CIs. The study presented in this paper has one of the smallest CIs due to the large number of patients and controls.
Discussion Coagulation disorders that increase thrombin generation associated with resistance against activated protein C and produce occlusive coronary episodes may represent additional risk factors for myocardial infarction. The risk may vary throughout the world. Regional differences in the prevalence of factor V Leiden have been documented in different geographical and ethnic populations worldwide, varying between 0.0% and 13.4%.36 Factor V Leiden is mostly limited to Caucasians and almost never found in Asians or Africans. There is a prominent north-south gradient in Europe, with decreasing prevalences of factor V Leiden going from northern to southern countries (13.4% to 0.0%). In this study, the population was limited to Bavaria, a region of comparative ethnic homogeneity compared with more cosmopolitan regions. In patients with myocardial infarction, the study recorded a prevalence of factor V Leiden of 8.7%. This is a significant increase in the prevalence compared with the Bavarian control group, which had a prevalence of 3.7%. A previously published Bavarian study (BATERS) with 821 randomized women, in which the prevalence was 5.5% (45/821),37 as well as a further small independent study in Bavaria, with a prevalence of 4.3%
902 Middendorf et al
Figure 2
Forrest plot of previous publications. This plot includes 19 study groups and also our own study.
(2/47),36 confirm the prevalence in this population to be this low. However, the BATERS controls included only fertile women (aged 18 – 49 y) while the further small independent study gives no information about the distribution of gender. In comparison with that, our control group includes both genders and excludes patients with any venous or arterial thromboembolism. Compared with control data, we have clearly found a significant increased prevalence of factor V Leiden with resistance against activated protein C in a group of patients with confirmed myocardial infarction, using strict criteria to define this. Testing for factor V Leiden, with the mutation Arg 3 Gln, excluded the 5% to 10% of patients with resistance against activated protein C resulting from other underlying causes.30 The results suggest that the presence of factor V Leiden predisposes to myocardial infarction and arterial thrombosis. The age of the patients presenting with myocardial infarction was not significantly different in the presence of factor V Leiden; thus, additional initiating factors may be required to cause myocardial infarction, as described for venous thrombosis, although increased mortality from coronary heart disease in younger patients who are the parents of children with factor V Leiden has been recorded.21 The number of females was higher among patients with myocardial infarction together with factor V Leiden compared with the patients with myocardial infarction in the absence of the factor, although this did not reach statistical significance. It has previously been hypothesized that the etiology in myocardial infarction and the occurrence of myocardial infarction may vary according to gender.27 Endogenous and exogenous estrogen in women, independent of the presence of factor V Leiden, both increase resistance against activated protein C. Thus, estrogens lower the degree of inactivation of activated factor V by activated protein
American Heart Journal May 2004
C. This is one reason why factor V Leiden might be a more important predisposing risk factor for myocardial infarction in women than in men. Further prospective studies to investigate this issue are needed. We also examined the contribution of the main cardiovascular risk determinants to the occurrence of myocardial infarction. The existence of a positive family history for myocardial infarction was similar in patients with and without factor V Leiden. The pattern of inheritance of myocardial infarction is complex and may reflect many factors, including a possible relation to factor V Leiden, which is inherited autosomal dominantly; thus few further conclusions can be made from the family history results in this study. We found trends towards a lower prevalence of hyperlipidemia, arterial hypertension, and smoking in patients with myocardial infarction and factor V Leiden compared with the total patient population with myocardial infarction. This might suggest that traditional risk determinants active in the general factor V Leiden–negative population may be of less importance in patients with the mutation. This conclusion would be supported by other data: a 9-fold increased mortality from coronary heart disease in patients ⬍45 years of age who are the parents of children with factor V Leiden has been recorded.21 We carried out a meta-analysis of other publications reporting similar data, to put our own research into context. All publications presenting data on the prevalence of resistance against activated protein C resulting from the presence of factor V Leiden in patients with myocardial infarction were extracted from an online database. The comparison was restricted to studies in which the control population reflected the study population, making the OR calculations more valid. We identified several differences among the previously published studies. Amongst these were differences in patient selection, in the tests used for biochemical analyses, and in the statistics tests carried out. Not all studies included both genders, or recorded age at first myocardial infarction, or ethnic group(s) of the patients. The tests for the mutation also varied. In some studies, only unmodified functional tests for resistance against activated protein C were used as a marker for factor V Leiden, and therefore tests were not necessarily specific for factor V Leiden. Some studies carried out only the modified functional test and others performed only genotyping. Patient numbers with myocardial infarction were low in some studies, and there were wide variations in the prevalence of the underlying genetic defect in the regional study populations. Our finding that there is a significant increased prevalence of factor V Leiden with resistance against activated protein C in a group of patients with confirmed
American Heart Journal Volume 147, Number 5
myocardial infarction can be compared directly with other large studies individually. The biggest retrospective study,9 the multinational Etude Cas-Te´moin de l’Infarctus du Myocarde (ECTIM) study, which included 643 patients and 726 controls, did not demonstrate an increased prevalence of heterozygous factor V Leiden in patients with myocardial infarction (OR ⫽ 1.1363). Subgroups were investigated in each city where the study was performed. A significant association between factor V Leiden and myocardial infarction was found in 1 of the 4 subgroups (Lille, France), supported by a strong trend in another subgroup (Strasbourg, France). No association was found in centers in Belfast (Northern Ireland) or Toulouse (France). This supports the notion of possible regional differences in the association even within different mainly Caucasian descendants. Conclusions may show nonsignificance when patient groups and control groups are of mixed ethnogenicity, for example in some regions of the United States. The biggest prospective study, the Physicians’ Health Study by Ridker et al,17 which included 374 patients and 704 controls, has also not demonstrated an increased prevalence of factor V Leiden in patients with myocardial infarction (OR ⫽ 1.0328). In this study, performed on the East Coast of the United States, Caucasians were randomly assigned to 1 of 4 treatments (aspirin given on alternate days, beta carotene given on alternate days, both combined, or no treatment). One reason that this study showed essentially no association could be that the various treatments may have influenced the risk of myocardial infarction in patients with factor V Leiden. Another limitation may be the fact that study entrants were exclusively male Caucasians, although there is likely to be a mixed ethnic background within this population with differences in the prevalence of factor V Leiden varying between 13.4% and 0.0% in different European regions.36 These circumstances may have diluted a possible relationship between factor V Leiden and myocardial infarction. In the present study, we attempted to control for several sources of potential bias. First, bias resulting from regional differences in the prevalence of factor V Leiden was reduced by limiting the study to a region with a more homogeneous ethnic population than more cosmopolitan areas with long histories of immigration. Second, patients with myocardial infarction of both genders were included, without age limitations. Third, we chose a highly sensitive and specific modified functional test for activated protein C associated with factor V Leiden with consequent additional genotyping to avoid false-positive results. Using such measures, a significant association was found between factor V Leiden and myocardial infarction. However, we found that the age at the time of the myocardial infarction did not differ from that in
Middendorf et al 903
the overall myocardial infarction group; thus, in this study, the prevalence of factor V alone did not appear to be sufficient on its own to lead to myocardial infarction (a situation paralleled by observations in the venous system). For such cases in which there is a phenotype of resistance against activated protein C, additional risk factors may be required to lead to myocardial infarction when there is only minimal atherosclerosis. We suggest that a pathological genotype alone does not justify preventive measures such as systemic anticoagulation. However, subgroups of patients with the genotype may be at higher risk for local coronary thrombosis than patients without resistance against activated protein C, for example following thrombolysis or intervention, or in the case of progressive coronary heart disease. Further studies are needed to identify such patients, in whom preventative measures should be considered and other risk factors aggressively addressed. Clinical studies will demonstrate whether subsequent antiplatelet aggregation treatment is sufficient or whether warfarin may be necessary. Investigations into the predisposing role of factor V Leiden in cerebrovascular, visceral, and peripheral arterial occlusive disease may further elucidate the role of the mutation in other arterial thromboembolic complications. We would like to thank Dorothea Nagel for help with the statistical analyses, and Peter Lohse for performing the genotyping of patients.
References 1. Eskandari MK, Bontempo FA, Hassett AC, et al. Arterial thromboembolic events in patients with the factor V Leiden mutation. Am J Surg 1998;176:122–5. 2. Amowitz LL, Komaroff AL, Miletich JP, et al. Factor V Leiden is not a risk factor for myocardial infarction among young women. Blood 1999;93:1432–3. 3. Ardissino D, Peyvandi F, Merlini PA, et al. Factor V (Arg506 3 Gln) mutation in young survivors of myocardial infarction. Thromb Haemost 1996;75:701–2. 4. Van Bockxmeer FM, Baker RI, Taylor RR. Premature ischaemic heart disease and the gene for coagulation factor V. Nat Med 1995;1:185. 5. Van der Bom JG, Bots ML, Haverkate F, et al. Reduced response to activated protein C is associated with increased risk for cerebrovascular disease. Ann Intern Med 1996;125:265–9. 6. Cushman M, Rosendaal FR, Psaty BM, et al. Factor V Leiden is not a risk factor for arterial vascular disease in the elderly: results from the cardiovascular health study. Thromb Haemost 1998;79:912–5. 7. Dacosta A, Tardy-Poncet B, Isaaz K, et al. Prevalence of factor V Leiden (APCR) and other inherited thrombophilias in young patients with myocardial infarction and normal coronary arteries. Heart 1998;80:338 – 40.
904 Middendorf et al
8. Demarmels Biasiutti F, Merlo C, Furlan M, et al. No association of APC resistance with myocardial infarction. Blood Coagul Fibrinolysis 1995;6:456 –9. 9. Emmerich J, Poirier O, Evans A, et al. Myocardial infarction, Arg506 to Gln factor V mutation, and activated protein C resistance. Lancet 1995;345:321. 10. Eritsland J, Gjønnes G, Sandset PM, et al. Activated protein C resistance and graft occlusion after coronary artery bypass surgery. Thromb Res 1995;79:223– 6. 11. Hansson PO, Eriksson E, Welin L, et al. Prevalence of APC resistance and its relationship to arterial and venous thromboembolism in a general population sample of elderly Swedish men: the study of men born in 1913. J Intern Med 1999;245:593– 600. 12. Inbal A, Freimark D, Modan B, et al. Synergistic effects of prothrombotic polymorphisms and atherogenic factors on the risk of myocardial infarction in young males. Blood 1999;93:2186 –90. 13. Kontula K, Ylikorkala A, Miettinen H, et al. Arg506 Gln mutation (factor V Leiden) in patients with ischaemic cerebrovascular disease and survivors of myocardial infarction. Thromb Haemost 1995;73:558 – 60. 14. Lowe GDO, Rumley A, Woodward M, et al. Activated protein C resistance and the FV: R506 Q mutation in a random population sample—associations with cardiovascular risk factors and coagulation variables. Thromb Haemost 1999;81:918 –24. 15. Prohaska W, Mannebach H, Schmidt M, et al. Evidence against heterozygous coagulation factor V 1619 G 3 A mutation with resistance to activated protein C being a risk factor for coronary artery disease and myocardial infarction. J Mol Med 1995;73: 521– 4. 16. Redondo M, Watzke HH, Stucki B, et al. Coagulation factors II, V, VII, and X, prothrombin gene 20210 G 3 A transition, and factor V Leiden in coronary artery disease— high factor V clotting activity is an independent risk factor for myocardial infarction. Arterioscler Thromb Vasc Biol 1999;19:1020 –5. 17. Ridker PM, Hennekens CH, Lindpaintner K, et al. Mutation in the gene coding for coagulation factor V and the risk of myocardial infarction, stroke, and venous thrombosis in apparently healthy men. N Engl J Med 1995;332:912–7. 18. Samani NJ, Lodwick D, Martin D, et al. Resistance to activated protein C and risk of premature myocardial infarction. Lancet 1994;344:1709 –10. 19. Baranovskaya S, Kudinov S, Fomicheva E, et al. Age as a risk factor for myocardial infarction in Leiden mutation carriers. Mol Genet Metab 1998;63:155–7. 20. Doggen CJM, Cats VM, Bertina RM, et al. Interaction of coagulation defects and cardiovascular risk factors—increased risk of myocardial infarction associated with factor V Leiden or prothrombin 20210A. Circulation 1998;97:1037– 41. 21. Hille ETM, Westendorp RGJ, Vandenbroucke JP, et al. Mortality and causes of death in families with the factor V Leiden mutation (resistance to activated protein C). Blood 1997;89:1963–7.
American Heart Journal May 2004
22. Holm J, Zo ¨ ller B, Svensson PJ, et al. Myocardial infarction associated with homozygous resistance to activated protein C. Lancet 1994;344:952–3. 23. Holm J, Zo ¨ ller B, Berntorp E, et al. Prevalence of factor V gene mutation amongst myocardial infarction patients and healthy controls is higher in Sweden than in other countries. J Intern Med 1996;239:221– 6. 24. Holm J, Hillarp A, Zo ¨ ller B, et al. Factor V Q506 (resistance to activated protein C) and prognosis after acute coronary syndrome. Thromb Haemost 1999;81:857– 60. 25. Kiechl S, Muigg A, Santer P, et al. Poor response to activated protein C as a prominent risk predictor of advanced atherosclerosis and arterial disease. Circulation 1999;99:614 –9. 26. Ma ¨ rz W, Seydewitz H, Winkelmann B, et al. Mutation in coagulation factor V associated with resistance to activated protein C in patients with coronary artery disease. Lancet 1995;345:526 –7. 27. Rosendaal FR, Siscovick DS, Schwartz SM, et al. Factor V Leiden (resistance to activated protein C) increases the risk of myocardial infarction in young women. Blood 1997;89:2817–21. 28. Dahlba ¨ ck B, Carlsson M, Svensson PJ. Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Natl Acad Sci USA 1993;90: 1004 – 8. 29. Dahlba ¨ ck B, Hildebrand B. Inherited resistance to activated protein C is corrected by anticoagulant cofactor activity found to be a property of factor V. Proc Natl Acad Sci USA 1994;91:1396 – 1400. 30. Williamson D, Brown K, Luddington R, et al. Factor V Cambridge: a new mutation (Arg306 3 Thr) associated with resistance to activated protein C. Blood 1998;91:1140 – 4. 31. Bertina RM, Koeleman BPC, Koster T, et al. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64 –7. 32. Jorquera JI, Montoro JM, Ferna´ ndez MA, et al. Modified test for activated protein C resistance. Lancet 1994;344:1162–3. 33. Trossae¨rt M, Conard J, Horellou MH, et al. Modified APC resistance assay for patients on oral anticoagulants. Lancet 1994;344: 1709. 34. Zo ¨ ller B, Svensson PJ, He X, et al. Identification of the same factor V gene mutation in 47 out of 50 thrombosis-prone families with inherited resistance to activated protein C. J Clin Invest 1994;94: 2521– 4. 35. Egger M, Smith GD, Phillips AN. Meta-analysis – principles and procedures. BMJ 1997;315:1533–7. 36. Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995;346:1133– 4. 37. Spannagl M, Dick A, Assmann A, et al. Resistance to activated protein C in women using oral contraceptives. Semin Thromb Hemost 1998;24:423–30.
Prevalence of resistance against activated protein C resulting from factor V Leiden is significantly increased in myocardial infarction: Investigation of 507 patients with myocardial infarction Katharina Middendorf, MD,a Peter Go ¨ hring, MD,b Tanya Y. Huehns, MD,c Dietrich Seidel, Prof,b a Gerhard Steinbeck, Prof, and Sigrid Nikol, Profa,d Mu ¨ nster and Munich, Germany, and London, Great Britain
Background A point mutation in the gene encoding coagulation factor V is a cause of resistance against activated protein C. The presence of factor V Leiden is linked to 50% of congenital defects causing venous thrombosis. Its relationship to arterial thrombosis, particularly to myocardial infarction, has not been defined. Therefore, we performed a study on the role of factor V Leiden in patients with myocardial infarction. The study was carried out in Bavarians of German origin, a relatively homogeneous population. Methods and Results The study group consisted of 507 patients with documented myocardial infarction (77.5% (393/507) men, 22.5% (114/507) women), with a mean age of 56.1 (range 18-– 86) years. Strict criteria for patient selection and highly sensitive and specific functional tests for factor V Leiden were used. In addition, all patients with pathological test results were genotyped. The prevalence of factor V Leiden in patients with myocardial infarction was 8.7% (44/507), a significant increase in the prevalence of this mutation compared with the control group (3.7%, P ⫽ .0025). The odds ratio was 2.46 (95% CI 1.35– 4.50). Conclusions A significantly increased prevalence of factor V Leiden in patients with documented myocardial infarction was seen. Patients with this mutation appear to have a predisposition for myocardial infarction. (Am Heart J 2004; 147:897–904.) Resistance against activated protein C commonly occurs in the presence of a mutation producing the molecule termed factor V Leiden. Its presence leads to resistance against activated protein C: activated factor V resists inactivation by activated protein C, which results in a potentially procoagulant state.1 Factor V Leiden is associated with 50% of congenital defects that result in increased thrombosis in the venous system. This is often seen in combination with other initiFrom the aMedical Clinic I, Klinikum Großhadern, Ludwig Maximilian University, Munich, Germany, bInstitute of Clinical Chemistry, Klinikum Großhadern, Ludwig Maximilian University, Munich, Germany, cBritish Heart Foundation, Cardiovascular Medicine, Imperial College School of Technology and Medicine, Hammersmith Hospital, London, Great Britain, and dDepartment of Cardiology and Angiology, University of Mu¨nster, Mu¨nster, Germany. Submitted April 21, 2003; accepted October 17, 2003. Reprint requests: Prof Dr med Sigrid Nikol, Medizinische Klinik und Poliklinik C (Kardiologie und Angiologie), Universita ¨ tsklinikum Mu¨nster, Albert-Schweitzer-Str 33, D-48129 Mu¨nster, Germany. E-mail: [email protected] 0002-8703/$ - see front matter © 2004, Elsevier Inc. All rights reserved. doi:10.1016/j.ahj.2003.10.048
ating factors, such as during pregnancy and periods of immobilization following operations. The role of factor V Leiden in thrombosis in relation to arterial disease remains controversial. Patients who are heterozygous for factor V Leiden and have a myocardial infarction have less than the expected amount of coronary atherosclerosis, suggesting that the prevalence of factor V Leiden may predispose to myocardial infarction at an earlier stage of coronary disease.1 While some researchers have not found an association between resistance against activated protein C caused by the presence of factor V Leiden and myocardial infarction,2–18 others, especially in more recent publications, have suggested that there is a relationship.1,13,19 –27 The availability of improved functional tests in these more recent studies suggests that the relationship may be a real one. A total of 90% to 95% of patients with resistance against activated protein C have the FV506Q mutation (leading to the molecule termed factor V Leiden);28 –30 in the remaining 5% to 10%, increased lupus anticoagu-
American Heart Journal May 2004
898 Middendorf et al
Table I. Characteristics of 19 publications investigating the association between factor V Leiden and myocardial infarction—3 of the studies consisted of 2 to 4 single study groups Characteristics of patients with MI First author, year of appearance
Study design
Test for FVL
Patient number
Amowitz 19992 Ardissino 19963
Case-control Case-control
Genotyping Genotyping
36 100
Baranovskaya 199819
Case-control
Genotyping
Van Bockxmeer 19954 Van der Bom 19965 Cushman 19986
Case-control Case-control Prospective
Genotyping Functional, genotyping Genotyping
287 168/287 119/287 149* 114 149
Dacosta 19987
Case-control
Functional, genotyping
75
Demarmels Biasiutti 19958
Case-control
Not modified functional
134
Doggen 199820 Emmerich 19959
Case-control Case-control
Genotyping Genotyping
Eritsland 199510 Holm 199623
Case-control Case-control
Functional Functional, genotyping
560 643 197/643 96/643 207/643 143/643 281† 101
Holm 199924 Inbal 199912 Kontula 199513
Case-control Case-control Case-control
Genotyping Genotyping Genotyping
23‡ 112 122 51/122 71/122
Prohaska 199515
Case-control
Functional, genotyping
134§
Redondo 199916 Ridker 199517 Rosendaal 199727
Case-control Prospective Case-control
Modified functional, genotyping Genotyping Genotyping
177㛳 374 84
m/f
Age (y)
Prevalence(%)
Only f 96:4 96%:4% ?
Ø 39 Ø 40.2 (⬍45)
2.8 (1/36) 1.0 (1/100)
Ø 60 (27–91) ⬍65 ⱖ65
? 92:57 62%:38% 66:9 88%:12% 115:19 86%:14% Only m Only m
⬎55 Ø 74.3 (⬎65)
3.1 (9/287) 0.6 (1/168) 6.7 (8/119) 4.7 (7/149) 3.5 (4/114) 3.4 (5/149)
Ø 38.0 (⬍45)
5.3 (4/75)
Ø 54.9 (32–72)
2.2 (3/134)
Ø 56.2 25–64
6.8 (38/560) 5.3 (34/643) 2.5 (5/197) 4.2 (4/96) 10.6 (22/207) 2.1 (3/143) 1.8 (5/281) 17.8 (18/101)
79:22 78%:22% only f only m 107:15 88%:12% 49:2 96%:4% 58:13 82%:18%
Ø 44 (28–50)
⬍52
30.4 (7/23) 6.3 (7/112) 5.7 (7/122)
⬍45
2.0 (1/51)
⬍60
8.5 (6/71) 7.5 (10/134)
Only m Only f
? 40–84 (⬎40) Ø 39.6 (23–44)
6.2 (11/177) 6.1 (23/374) 9.5 (8/84)
FVL, factor V Leiden; MI, myocardial infarction; functional, functional test for resistance against APC which is not specific for the detection of factor V Leiden; m:f, ratio between men and women; ?, no more information given in the publication. *From a total collective of 222 patients with ischemic heart disease including 149 patients with MI; of all patients 87% were (193/222) men and 13% (29/222) women under the age of 50 years; the control group consisted of 126 persons. †From a total collective of 610 patients with coronary artery bypass surgery including 281 patients with MI; of all patients 87% were (531/610) men and 13% (79/610) women with a mean age of 60 years; the control group consisted of 265 persons. ‡From a total collective of 295 patients with acute coronary syndrome including 86 patients with MI and 210 patients with unstable angina, of all patients 64% were (190/ 295) men and 36% (105/295) women aged between 51 and 73 years; 23 of patients with MI were women; the control group consisted of 423 fertile women. §From a total collective of 317 patients with angiographically diagnosed coronary artery disease including 134 patients with MI, of all patients 85% were (269/317) men and 15% (48/317) women with a mean age of 58 years (range from 28 to 85 years). 㛳From a total collective of 200 patients with coronary artery disease including 177 patients with MI tested for factor V Leiden, of all patients 87% were (174/200) men and 13% (26/200) women; the control group consisted of 100 persons including 90 persons tested for factor V Leiden.
lants, increased factor VIII, or the effects caused by pregnancy or oral contraceptives are responsible.30 In the mutation for factor V Leiden, arginine at position 506 is replaced by glutamine (or guanine at position 1691 by adenine) in the factor V molecule.31 The mutation results in an alteration in the activated factor V activated protein C cleavage site. Hence, factor V Lei-
den is inactivated by proteolysis 10 to 20 times slower than the native form of factor V. The phenotype can be determined using the original resistance against activated protein C functional test,28 or a modified version that has the advantage of nearly 100% sensitivity for factor V Leiden.32,33 Additional genotyping allows confirmation of factor
American Heart Journal Volume 147, Number 5
Middendorf et al 899
Table I. Continued Characteristics of controls Patient number 36 100 483 373/483 110/483 126 222 482
53 100 646 726 178/726 148/726 193/726 207/726 265 101 423 187 137 50/137 87/137 190 90 704 388
m/f
Age (y)
Only f ? ?
? Ø 41.6
Prevalence (%)
Cardio-vascular risk factors Yes Yes No
Yes Yes Yes
0.31 (0.03–3.17) 0.49 (0.04–5.55) 1.08 (0.46–2.54) 0.17 (0.02–1.28) 7.86 (0.97–63.87) 1.19 (0.37–3.86) 0.70 (0.22–2.24) 0.46 (0.18–1.19)
OR (95% CI)
? 178:304 37%:63%
? Ø 72.3
8.3 (3/36) 2.0 (2/100) 2.9 (14/483) 3.5 (13/373) 0.9 (1/110) 4.0 (5/126) 5.0 (11/222) 7.1 (34/482)
46:7 87%:13% 87:13 87%:13% Only m ?
Ø 36.5
3.8 (2/53)
Yes
1.44 (0.25–8.15)
Ø 54.9 (32–74)
2.0 (2/100)
Yes
1.12 (0.18–6.84)
Ø 57.3 ?
5.0 (32/646) 4.7 (34/726) 5.6 (10/178) 0.7 (1/148) 8.8 (17/193) 2.9 (6/207) 2.6 (7/265) 10.9 (11/101)
Yes No
No No
1.40 (0.86–2.27) 1.14 (0.70–1.85) 0.44 (0.15–1.31) 6.39 (0.70–58.07) 1.23 (0.63–2.40) 0.72 (0.18–2.92) 0.67 (0.21–2.13) 1.77 (0.79–3.40)
Yes Yes Yes
3.87 (1.51–9.92) 0.97 (0.37–2.55) 2.02 (0.58–7.09)
Ø 11 (6–17) Ø 80
?
?
Only f Only m 95:42 69%:31% 42:8 84%:16% 53:34 61%:39% 116:74 61%:39% ? Only m Only f
fertile age ?
10.2 (43/423) 6.4 (12/187) 2.9 (4/137)
?
6.0 (3/50)
0.31 (0.03–3.12)
?
1.1 (1/87)
7.94 (0.93–67.57)
Ø 34 (18–65)
9.5 (18/190)
No
0.77 (0.34–1.73)
? 40–84 (⬎40) Ø 37.7 (19–44)
6.7 (6/89) 6.0 (42/704) 4.1 (16/388)
Yes Yes Yes
0.93 (0.33–2.60) 1.03 (0.61–1.75) 2.45 (1.01–5.92)
V Leiden and detection of homozygotes and heterozygotes. Patient numbers in previously published studies may have generally been too low to reach significance, partly because of high regional differences in the prevalence of the underlying genetic defect (1%–15%) and also because of the ethnic heterogeneity of enrolled patients. This study aimed to investigate the role of factor V Leiden in a more homogeneous group of patients with myocardial infarction, using a phenotyping test with high sensitivity, together with confirmation of the presence of the factor by subsequent genotyping. To interpret results of the study, data were related to a meta-analysis of previous publications.
Methods Patients The patient population comprised 507 patients who presented to a university hospital in Munich between September 1996 and March 2000. Inclusion criteria were documented prior myocardial infarction (increased creatine kinase [CK] with CK-MB ⬎10% or proof of increased troponin I, typical alteration in the ECG, and typical clinical symptoms), with the first myocardial infarction date accurately recorded, together with written consent for the molecular genetic analysis (required in Bavaria since March 1999; patients prior to this consented orally). One of the consecutive patients was excluded because he refused written consent. Patients with venous thromboembolisms and non-Germans were excluded (family name was used as a proxy-variable for German nationality).
American Heart Journal May 2004
900 Middendorf et al
Patients had a mean age of 56.1 years (range 18 to 86 years); 77.5% (393/507) were men and 22.5% (114/507) were women. Of the total, 76.3% (387/507) had survived a single myocardial infarction, and 23.7% (120/507) had survived multiple myocardial infarctions. Among the 666 myocardial infarctions experienced by patients in the study, 36.2% (241/ 666) were anterior, 36.5% (243/666) were posterior, and 27.3% (182/666) were myocardial infarctions without known location. The main cardiovascular risk factors, including hyperlipidemia, arterial hypertension, smoking, and family history were recorded in one third of patients. The genetic defect leading to the presence of factor V Leiden was diagnosed in the enrolled patient population as described below.
Control group The control population comprised 404 patients from the same geographical region who presented to a university hospital in Munich with other diagnoses not involving thrombotic events between September 1996 and March 2000. Patients with venous and arterial thromboembolisms were excluded. Controls in this group had a mean age of 54.4 years (range 12 to 88 years); 68.3% (276/404) were men and 31.7% (128/404) were women. The genetic defect leading to the presence of factor V Leiden was diagnosed in the enrolled control population in the same way as in the investigated group with myocardial infarctions, as described below.
Methods for analyses The same factor V Leiden test used for the control group was used for the patients with myocardial infarction enrolled in this study. The modified test used contains factor V deficient plasma for the dilution of the test plasma. It provides 100% sensitivity and about 98% specificity for FV:Q506 and is more sensitive than the activated protein C screening test used previously.28 The APCTM-Resistance V kit (COATEST, Chromogenix, Milan, Italy) was used in combination with a STA coagulation analyser (Roche/Stago, Mannheim, Germany). This gives a ratio of the activated partial thromboplastin time (aPTT) with and without addition of activated protein C. A ratio of ⬍2.1 was defined as resistance against activated protein C. Differentiation between normal and pathological phenotype is possible in nearly 100% of cases,32,33 and conclusions regarding the pathology of the genotype (factor V Leiden) can also be made.32 However, differentiation between the homozygous and heterozygous form is not always possible and therefore all patients with pathological results were then genotyped following this test. Genotyping was performed as follows: DNA was extracted from Na2EDTA-blood. A fragment of the factor V gene containing the Leiden mutation site was amplified using a polymerase chain reaction. The following oligonucleotide primers were used to amplify this fragment in exon 10 encoding the activated protein C cleavage site of the factor V molecule: 5⬘-GGAACAACACCATGATCAGAGCA-3⬘ and 5⬘-TAGCCAGGAGACCTAACATGTTC-3⬘.34 After restrictive digestion of the amplified DNA with Mnl I, the enzymatically digested fragments were separated by agarose gel electrophoresis and visualized with ethidium bromide.
Statistics Results are presented as means together with the standard error of the mean (SEM). The significance of the hypothesis
relating to the noncontinuous characteristics was investigated by 2 testing. The odds ratio (OR) with its logarithmus naturalis (ln) and 95% CI were calculated adjusted for age, gender, and nationality to reflect the risk of myocardial infarction in the presence of factor V Leiden compared with the risk in the absence of this factor.
Meta-analysis A search of the online database medline.de using the terms “factor V Leiden, resistance against activated protein C, myocardial infarction” yielded 200 publications. Inclusion criteria were documented myocardial infarction in the patient group, known prevalence of factor V Leiden in patients with myocardial infarction and in controls, and similar design characteristics. Publications were excluded if the controls recording the prevalence of factor V Leiden failed to originate from the same geographical region,1,11,14,18,22,25 or if missing data meant that calculation of the OR was impossible.26 The 19 studies included published between 1995 and 1999 are tabulated in Table I; 3 of the studies consisted of 2 to 4 single study groups. Ln (OR) and CI from each trial were graphically displayed by a Forrest plot to show heterogeneity and quality of published results.35
Results Factor V Leiden testing The prevalence of the genetic defect leading to factor V Leiden in the Bavarian control group was 3.7% (15/404). The modified functional test for factor V Leiden showed a mean activated protein C ratio of 2.3 (SEM ⫾ 0.3) in all patients presenting with myocardial infarction. In patients with myocardial infarction and genotyping positive for heterozygous or homozygous factor V Leiden, the average activated protein C ratio was 1.5 (SEM ⫾ 0.2). A pathological genotype was always associated with pathological values in the functional test. The prevalence of factor V Leiden in patients with myocardial infarction was 8.7% (44/507), with 2 homozygous and 42 heterozygous patients. The 2 test indicates a significant increased prevalence of factor V Leiden compared with the prevalence in the Bavarian control population (8.7% vs 3.7%, P ⫽ .0025; Figure 1). The crude OR was 2.46 (95% CI 1.35– 4.50); the adjusted OR for age, gender, and nationality is shown in Table II.
Patients with myocardial infarction and factor V Leiden Of these 44 patients, 68.2% (30/44) were men and 31.8% (14/44) were women, with a mean age of 54.5 (range 35– 84) years at the time of the first myocardial infarction; 16 patients were aged ⱕ50 years. A total of 75.0% (33/44) had survived a single myocardial infarction and 25.0% (11/44) had survived multiple myocardial infarctions. Among the 59 documented myocardial infarctions, 42.4% (25/59) were anterior, 30.5% (18/
American Heart Journal Volume 147, Number 5
Figure 1
Middendorf et al 901
Table II. Adjusted odds ratio and P value Adjusted OR
Factor V Leiden Age Male German nationality
OR
95% CI
2
P
2.66 1.01 1.61 0.84
1.45–4.88 1.00–1.02 1.20–2.18 0.47–1.50
9.91 3.63 9.82 0.36
.0016 .0569 .0017 .5487
Table III. Comparison of cardiovascular risk factors in all patients with myocardial infarction and patients with myocardial infarction and factor V Leiden Prevalence of factor V Leiden in patients with myocardial infarction and controls from the same geographical region.
59) were posterior, and 27.1% (16/59) were myocardial infarctions without known location. The average age at the time of the first myocardial infarction was 56.1 years in patients with myocardial infarction, and 54.5 years in the subgroup with both myocardial infarction and factor V Leiden. There was no significant association between factor V Leiden and the age of first myocardial infarction. There was no significant association between factor V Leiden and female sex in patients with myocardial infarction (31.8%) compared with all women with myocardial infarction (22.5%). There were no significant differences in identified cardiovascular risk factors in the patients with myocardial infarction and factor V Leiden compared with patients with myocardial infarction and no factor V Leiden (Table III). There was a trend towards a lower prevalence of hyperlipidemia, arterial hypertension, and smoking in patients with myocardial infarction and factor V Leiden compared with the total patient population with myocardial infarction.
Results of meta-analysis Table I summarizes the characteristics of the patients and controls included, and details the methods for detecting factor V Leiden used in each study. Comparison of the 19 studies demonstrated an OR total of 1.18 (95% CI 0.97–1.43) when investigating the risk of myocardial infarction in patients possessing factor V Leiden compared with the control patients. Using the 2 test, no significant heterogeneity in the publications (P ⫽ .3394) was found between the 19 studies. The Forrest plot compares our results with the 19 studies reported in the literature (Figure 2). The results show a random distribution, with the various studies showing mark-
14 Patients with 168 Patients MI and factor V with MI Leiden 2 Hyperlipidemia Arterial hypertension Smoking Family history No risk factors
66.7% (112/168) 61.9% (104/168) 47.0% (79/168) 29.8% (50/168) 3.0% (5/168)
50.0% (7/14) 50.0% (7/14) 35.7% (5/14) 21.4% (3/14) 0.0% (0/14)
P
1.91 .1671 0.92 .3380 0.78 .3759 0.51 .4763 0.47 .4937
edly different CIs. The study presented in this paper has one of the smallest CIs due to the large number of patients and controls.
Discussion Coagulation disorders that increase thrombin generation associated with resistance against activated protein C and produce occlusive coronary episodes may represent additional risk factors for myocardial infarction. The risk may vary throughout the world. Regional differences in the prevalence of factor V Leiden have been documented in different geographical and ethnic populations worldwide, varying between 0.0% and 13.4%.36 Factor V Leiden is mostly limited to Caucasians and almost never found in Asians or Africans. There is a prominent north-south gradient in Europe, with decreasing prevalences of factor V Leiden going from northern to southern countries (13.4% to 0.0%). In this study, the population was limited to Bavaria, a region of comparative ethnic homogeneity compared with more cosmopolitan regions. In patients with myocardial infarction, the study recorded a prevalence of factor V Leiden of 8.7%. This is a significant increase in the prevalence compared with the Bavarian control group, which had a prevalence of 3.7%. A previously published Bavarian study (BATERS) with 821 randomized women, in which the prevalence was 5.5% (45/821),37 as well as a further small independent study in Bavaria, with a prevalence of 4.3%
902 Middendorf et al
Figure 2
Forrest plot of previous publications. This plot includes 19 study groups and also our own study.
(2/47),36 confirm the prevalence in this population to be this low. However, the BATERS controls included only fertile women (aged 18 – 49 y) while the further small independent study gives no information about the distribution of gender. In comparison with that, our control group includes both genders and excludes patients with any venous or arterial thromboembolism. Compared with control data, we have clearly found a significant increased prevalence of factor V Leiden with resistance against activated protein C in a group of patients with confirmed myocardial infarction, using strict criteria to define this. Testing for factor V Leiden, with the mutation Arg 3 Gln, excluded the 5% to 10% of patients with resistance against activated protein C resulting from other underlying causes.30 The results suggest that the presence of factor V Leiden predisposes to myocardial infarction and arterial thrombosis. The age of the patients presenting with myocardial infarction was not significantly different in the presence of factor V Leiden; thus, additional initiating factors may be required to cause myocardial infarction, as described for venous thrombosis, although increased mortality from coronary heart disease in younger patients who are the parents of children with factor V Leiden has been recorded.21 The number of females was higher among patients with myocardial infarction together with factor V Leiden compared with the patients with myocardial infarction in the absence of the factor, although this did not reach statistical significance. It has previously been hypothesized that the etiology in myocardial infarction and the occurrence of myocardial infarction may vary according to gender.27 Endogenous and exogenous estrogen in women, independent of the presence of factor V Leiden, both increase resistance against activated protein C. Thus, estrogens lower the degree of inactivation of activated factor V by activated protein
American Heart Journal May 2004
C. This is one reason why factor V Leiden might be a more important predisposing risk factor for myocardial infarction in women than in men. Further prospective studies to investigate this issue are needed. We also examined the contribution of the main cardiovascular risk determinants to the occurrence of myocardial infarction. The existence of a positive family history for myocardial infarction was similar in patients with and without factor V Leiden. The pattern of inheritance of myocardial infarction is complex and may reflect many factors, including a possible relation to factor V Leiden, which is inherited autosomal dominantly; thus few further conclusions can be made from the family history results in this study. We found trends towards a lower prevalence of hyperlipidemia, arterial hypertension, and smoking in patients with myocardial infarction and factor V Leiden compared with the total patient population with myocardial infarction. This might suggest that traditional risk determinants active in the general factor V Leiden–negative population may be of less importance in patients with the mutation. This conclusion would be supported by other data: a 9-fold increased mortality from coronary heart disease in patients ⬍45 years of age who are the parents of children with factor V Leiden has been recorded.21 We carried out a meta-analysis of other publications reporting similar data, to put our own research into context. All publications presenting data on the prevalence of resistance against activated protein C resulting from the presence of factor V Leiden in patients with myocardial infarction were extracted from an online database. The comparison was restricted to studies in which the control population reflected the study population, making the OR calculations more valid. We identified several differences among the previously published studies. Amongst these were differences in patient selection, in the tests used for biochemical analyses, and in the statistics tests carried out. Not all studies included both genders, or recorded age at first myocardial infarction, or ethnic group(s) of the patients. The tests for the mutation also varied. In some studies, only unmodified functional tests for resistance against activated protein C were used as a marker for factor V Leiden, and therefore tests were not necessarily specific for factor V Leiden. Some studies carried out only the modified functional test and others performed only genotyping. Patient numbers with myocardial infarction were low in some studies, and there were wide variations in the prevalence of the underlying genetic defect in the regional study populations. Our finding that there is a significant increased prevalence of factor V Leiden with resistance against activated protein C in a group of patients with confirmed
American Heart Journal Volume 147, Number 5
myocardial infarction can be compared directly with other large studies individually. The biggest retrospective study,9 the multinational Etude Cas-Te´moin de l’Infarctus du Myocarde (ECTIM) study, which included 643 patients and 726 controls, did not demonstrate an increased prevalence of heterozygous factor V Leiden in patients with myocardial infarction (OR ⫽ 1.1363). Subgroups were investigated in each city where the study was performed. A significant association between factor V Leiden and myocardial infarction was found in 1 of the 4 subgroups (Lille, France), supported by a strong trend in another subgroup (Strasbourg, France). No association was found in centers in Belfast (Northern Ireland) or Toulouse (France). This supports the notion of possible regional differences in the association even within different mainly Caucasian descendants. Conclusions may show nonsignificance when patient groups and control groups are of mixed ethnogenicity, for example in some regions of the United States. The biggest prospective study, the Physicians’ Health Study by Ridker et al,17 which included 374 patients and 704 controls, has also not demonstrated an increased prevalence of factor V Leiden in patients with myocardial infarction (OR ⫽ 1.0328). In this study, performed on the East Coast of the United States, Caucasians were randomly assigned to 1 of 4 treatments (aspirin given on alternate days, beta carotene given on alternate days, both combined, or no treatment). One reason that this study showed essentially no association could be that the various treatments may have influenced the risk of myocardial infarction in patients with factor V Leiden. Another limitation may be the fact that study entrants were exclusively male Caucasians, although there is likely to be a mixed ethnic background within this population with differences in the prevalence of factor V Leiden varying between 13.4% and 0.0% in different European regions.36 These circumstances may have diluted a possible relationship between factor V Leiden and myocardial infarction. In the present study, we attempted to control for several sources of potential bias. First, bias resulting from regional differences in the prevalence of factor V Leiden was reduced by limiting the study to a region with a more homogeneous ethnic population than more cosmopolitan areas with long histories of immigration. Second, patients with myocardial infarction of both genders were included, without age limitations. Third, we chose a highly sensitive and specific modified functional test for activated protein C associated with factor V Leiden with consequent additional genotyping to avoid false-positive results. Using such measures, a significant association was found between factor V Leiden and myocardial infarction. However, we found that the age at the time of the myocardial infarction did not differ from that in
Middendorf et al 903
the overall myocardial infarction group; thus, in this study, the prevalence of factor V alone did not appear to be sufficient on its own to lead to myocardial infarction (a situation paralleled by observations in the venous system). For such cases in which there is a phenotype of resistance against activated protein C, additional risk factors may be required to lead to myocardial infarction when there is only minimal atherosclerosis. We suggest that a pathological genotype alone does not justify preventive measures such as systemic anticoagulation. However, subgroups of patients with the genotype may be at higher risk for local coronary thrombosis than patients without resistance against activated protein C, for example following thrombolysis or intervention, or in the case of progressive coronary heart disease. Further studies are needed to identify such patients, in whom preventative measures should be considered and other risk factors aggressively addressed. Clinical studies will demonstrate whether subsequent antiplatelet aggregation treatment is sufficient or whether warfarin may be necessary. Investigations into the predisposing role of factor V Leiden in cerebrovascular, visceral, and peripheral arterial occlusive disease may further elucidate the role of the mutation in other arterial thromboembolic complications. We would like to thank Dorothea Nagel for help with the statistical analyses, and Peter Lohse for performing the genotyping of patients.
References 1. Eskandari MK, Bontempo FA, Hassett AC, et al. Arterial thromboembolic events in patients with the factor V Leiden mutation. Am J Surg 1998;176:122–5. 2. Amowitz LL, Komaroff AL, Miletich JP, et al. Factor V Leiden is not a risk factor for myocardial infarction among young women. Blood 1999;93:1432–3. 3. Ardissino D, Peyvandi F, Merlini PA, et al. Factor V (Arg506 3 Gln) mutation in young survivors of myocardial infarction. Thromb Haemost 1996;75:701–2. 4. Van Bockxmeer FM, Baker RI, Taylor RR. Premature ischaemic heart disease and the gene for coagulation factor V. Nat Med 1995;1:185. 5. Van der Bom JG, Bots ML, Haverkate F, et al. Reduced response to activated protein C is associated with increased risk for cerebrovascular disease. Ann Intern Med 1996;125:265–9. 6. Cushman M, Rosendaal FR, Psaty BM, et al. Factor V Leiden is not a risk factor for arterial vascular disease in the elderly: results from the cardiovascular health study. Thromb Haemost 1998;79:912–5. 7. Dacosta A, Tardy-Poncet B, Isaaz K, et al. Prevalence of factor V Leiden (APCR) and other inherited thrombophilias in young patients with myocardial infarction and normal coronary arteries. Heart 1998;80:338 – 40.
904 Middendorf et al
8. Demarmels Biasiutti F, Merlo C, Furlan M, et al. No association of APC resistance with myocardial infarction. Blood Coagul Fibrinolysis 1995;6:456 –9. 9. Emmerich J, Poirier O, Evans A, et al. Myocardial infarction, Arg506 to Gln factor V mutation, and activated protein C resistance. Lancet 1995;345:321. 10. Eritsland J, Gjønnes G, Sandset PM, et al. Activated protein C resistance and graft occlusion after coronary artery bypass surgery. Thromb Res 1995;79:223– 6. 11. Hansson PO, Eriksson E, Welin L, et al. Prevalence of APC resistance and its relationship to arterial and venous thromboembolism in a general population sample of elderly Swedish men: the study of men born in 1913. J Intern Med 1999;245:593– 600. 12. Inbal A, Freimark D, Modan B, et al. Synergistic effects of prothrombotic polymorphisms and atherogenic factors on the risk of myocardial infarction in young males. Blood 1999;93:2186 –90. 13. Kontula K, Ylikorkala A, Miettinen H, et al. Arg506 Gln mutation (factor V Leiden) in patients with ischaemic cerebrovascular disease and survivors of myocardial infarction. Thromb Haemost 1995;73:558 – 60. 14. Lowe GDO, Rumley A, Woodward M, et al. Activated protein C resistance and the FV: R506 Q mutation in a random population sample—associations with cardiovascular risk factors and coagulation variables. Thromb Haemost 1999;81:918 –24. 15. Prohaska W, Mannebach H, Schmidt M, et al. Evidence against heterozygous coagulation factor V 1619 G 3 A mutation with resistance to activated protein C being a risk factor for coronary artery disease and myocardial infarction. J Mol Med 1995;73: 521– 4. 16. Redondo M, Watzke HH, Stucki B, et al. Coagulation factors II, V, VII, and X, prothrombin gene 20210 G 3 A transition, and factor V Leiden in coronary artery disease— high factor V clotting activity is an independent risk factor for myocardial infarction. Arterioscler Thromb Vasc Biol 1999;19:1020 –5. 17. Ridker PM, Hennekens CH, Lindpaintner K, et al. Mutation in the gene coding for coagulation factor V and the risk of myocardial infarction, stroke, and venous thrombosis in apparently healthy men. N Engl J Med 1995;332:912–7. 18. Samani NJ, Lodwick D, Martin D, et al. Resistance to activated protein C and risk of premature myocardial infarction. Lancet 1994;344:1709 –10. 19. Baranovskaya S, Kudinov S, Fomicheva E, et al. Age as a risk factor for myocardial infarction in Leiden mutation carriers. Mol Genet Metab 1998;63:155–7. 20. Doggen CJM, Cats VM, Bertina RM, et al. Interaction of coagulation defects and cardiovascular risk factors—increased risk of myocardial infarction associated with factor V Leiden or prothrombin 20210A. Circulation 1998;97:1037– 41. 21. Hille ETM, Westendorp RGJ, Vandenbroucke JP, et al. Mortality and causes of death in families with the factor V Leiden mutation (resistance to activated protein C). Blood 1997;89:1963–7.
American Heart Journal May 2004
22. Holm J, Zo ¨ ller B, Svensson PJ, et al. Myocardial infarction associated with homozygous resistance to activated protein C. Lancet 1994;344:952–3. 23. Holm J, Zo ¨ ller B, Berntorp E, et al. Prevalence of factor V gene mutation amongst myocardial infarction patients and healthy controls is higher in Sweden than in other countries. J Intern Med 1996;239:221– 6. 24. Holm J, Hillarp A, Zo ¨ ller B, et al. Factor V Q506 (resistance to activated protein C) and prognosis after acute coronary syndrome. Thromb Haemost 1999;81:857– 60. 25. Kiechl S, Muigg A, Santer P, et al. Poor response to activated protein C as a prominent risk predictor of advanced atherosclerosis and arterial disease. Circulation 1999;99:614 –9. 26. Ma ¨ rz W, Seydewitz H, Winkelmann B, et al. Mutation in coagulation factor V associated with resistance to activated protein C in patients with coronary artery disease. Lancet 1995;345:526 –7. 27. Rosendaal FR, Siscovick DS, Schwartz SM, et al. Factor V Leiden (resistance to activated protein C) increases the risk of myocardial infarction in young women. Blood 1997;89:2817–21. 28. Dahlba ¨ ck B, Carlsson M, Svensson PJ. Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Natl Acad Sci USA 1993;90: 1004 – 8. 29. Dahlba ¨ ck B, Hildebrand B. Inherited resistance to activated protein C is corrected by anticoagulant cofactor activity found to be a property of factor V. Proc Natl Acad Sci USA 1994;91:1396 – 1400. 30. Williamson D, Brown K, Luddington R, et al. Factor V Cambridge: a new mutation (Arg306 3 Thr) associated with resistance to activated protein C. Blood 1998;91:1140 – 4. 31. Bertina RM, Koeleman BPC, Koster T, et al. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64 –7. 32. Jorquera JI, Montoro JM, Ferna´ ndez MA, et al. Modified test for activated protein C resistance. Lancet 1994;344:1162–3. 33. Trossae¨rt M, Conard J, Horellou MH, et al. Modified APC resistance assay for patients on oral anticoagulants. Lancet 1994;344: 1709. 34. Zo ¨ ller B, Svensson PJ, He X, et al. Identification of the same factor V gene mutation in 47 out of 50 thrombosis-prone families with inherited resistance to activated protein C. J Clin Invest 1994;94: 2521– 4. 35. Egger M, Smith GD, Phillips AN. Meta-analysis – principles and procedures. BMJ 1997;315:1533–7. 36. Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995;346:1133– 4. 37. Spannagl M, Dick A, Assmann A, et al. Resistance to activated protein C in women using oral contraceptives. Semin Thromb Hemost 1998;24:423–30.