|
Journal Publication |
||||||||||||||||||||||
The Role of Activated Protein C Resistance in the Pathogenesis of Venous Thrombosis by W. Craig Hooper and Bruce L. Evatt
Following myocardial infarction and stroke, venous thromboembolism (VTE) is the third most common cardiovascular disease in the United States. The mortality and morbidity of VTE is significant with an annual incidence of approximately 1:1000 individuals and pulmonary embolism is a leading cause of in-patient hospital deaths (1,2). It has been estimated that venous thrombosis is responsible for between 300,000-600,000 hospitalizations and up to 100,000 deaths annually (3,4). The clinical consequences of venous thrombosis such as chronic venous insufficiency with skin ulceration affects up to 500,000 individuals per year. Studies have reported that as many as 90% of patients with venous thrombosis suffer significant disabilities at 2-5 year follow-up intervals (5). The current health costs associated with VTE are significant and are expected to rise as the prevalence of VTE increases in the aging population (6,7). Venous thromboembolism (VTE) is usually a consequence of either acquired or inherited alterations in hemostatic regulatory proteins which are predominantly those of the protein C/protein S natural anticoagulant pathway. Acquired deficiencies in this pathway are frequently a consequence of other clinical entities such as cancer, AIDS and diabetes, while inherited deficiencies can be responsible for venous thrombosis in an otherwise healthy individual. In contrast to the acquired state, inherited deficiencies which are frequently referred to as thrombophilia or primary hypercoagulability, are a consequence of genetic abnormalities in either the anticoagulant proteins or other interacting proteins. Early genetic analyses of families with a history of idiopathic venous thrombosis have demonstrated that DNA mutations or deletions in genes encoding for protein C, protein S, and antithrombin III were responsible for approximately 5-10% of these events. These mutations were phenotypically expressed at the protein level either by low plasma levels or impaired function. In addition to the anticoagulant proteins, DNA mutations in other hemostatic/coagulation proteins listed in Table I have also been associated with VTE. Recently the activated partial thromboplastin time (APTT) assay demonstrated a poor anticoagulant response to exogenously added activated protein C (APC) in a patient with idiopathic venous thrombosis (8). This previously unknown coagulation abnormality termed activated protein C- resistance (APC-R), was also detected in family members thus suggesting that this response was inherited (8). The significance of this finding quickly became apparent when it was found that APC-R was present in approximately 40-60% of referred thrombosis patients (9,10). The underlying molecular basis of APC-R was soon shown to be a single DNA point mutation at an APC cleavage site in factor V that resulted in the replacement of the amino acid arginine by glutamine at codon 506 (11,12,13). This DNA defect in factor V sometimes referred to as FV Q506, FV 1691 G->A but more commonly as factor V Leiden (FV Leiden) (Table II) has been found in approximately 90-95% of patients with APC-R. In contrast to the multiple DNA mutations that have been identified in each of the anticoagulant proteins or in the other regulatory proteins, FV Leiden is unique because only one mutation has been identified, thus allowing for rapid genetic analysis. The purpose of this review is to briefly re-acquaint the reader with the pathobiology of the anticoagulant protein system and to review the clinical implications of APC-R. In normal hemostasis, the procoagulant and anticoagulant proteins are balanced to favor anticoagulation, but when injury occurs, the balance shifts to the procoagulant forces in order to form a physiological clot. However, when pathological conditions either alone or in conjunction with injury induce a shift to a procoagulant state, an increased risk for a pathologic thrombosis can exist. The significance of this shift in balance toward anticoagulation was first proposed over a century ago as an attempt to define thrombosis. Known as Virchows Triad, it was postulated that three elements acting in concert, stasis, vessel wall injury and hypercoagulability were responsible for thrombosis (14). Since that revolutionary insight, it has been recognized that the coagulation system consists of a series of complex biochemical reactions that initiate thrombin formation which in turn converts fibrinogen into insoluble fibrin, the major component of venous thrombi. Paradoxically, thrombin also interacts with the anticoagulant proteins to control or limit clot formation. Although coagulation has classically been divided into the intrinsic and extrinsic pathways, it is probable that these pathways have little physiologic relevance when defined as such. However, for the sake of simplicity we will refer to the coagulation cascade using these terms. The intrinsic pathway which includes the proteins prothrombin, fibrinogen, prekallikrein, kininogen and factors XII, XI, IX, VIII, X, and V becomes activated by contact with a negatively charged surface, whereas the extrinsic pathway becomes functional following vascular injury. The extrinsic pathway which includes tissue factor, prothrombin, fibrinogen, factors V, VII, and X is believed by many to be the most physiologically relevant pathway in physiological hemostasis while proteins assigned to the intrinsic pathway are thought to function as positive feedback amplifiers (15,16). It has been proposed that, in the extrinsic pathway, coagulation is initiated when tissue factor becomes exposed to blood following vascular injury. The binding of tissue factor to factor VII (FVII) initiates a series of sequentially proteolytic steps that results in the formation of thrombin with the subsequent conversion into insoluble fibrin. Once activated by tissue factor, FVIIa in turn activates factors X and IX. Factors Xa and IXa then bind to specific platelet receptors and induce amplification of the coagulation cascade (15,16). Factors V and VIII activated by thrombin (Va and VIIIa) during the initiation of coagulation also bind to specific platelet receptors. On the surface of the platelet Va is localized in the vicinity of Xa while VIIIa is found next to lXa The localization of these factors form both the tenase (IXa-VIIIa) and the prothrombinase complex (Va-Xa) which serves to propagate coagulation through the conversion of X into Xa and the exponential amplification of thrombin. In order to limit clot formation, a tight regulatory network has evolved with the protein C/protein S anticoagulant pathway being a central component. The major function of this pathway which includes the proteins thrombomodulin, protein C, and protein S, is to inactivate factors Va and VIIIa (16). This pathway becomes activated when thrombin binds to thrombomodulin which in turn activates protein C (APC). APC inactivates both membrane bound Va and VIIIa which in turn inhibits prothrombinase activity. Protein S functions as a catalyst for activated protein C (APC). A biological alteration in the APC- mediated inactivation process is the underlying molecular mechanism of APC-R. In essence, a DNA mutation in factor V results in replacement of the amino acid arginine by glutamine at the APC codon 506 cleavage site (16). This mutation does not destroy but rather slows APC mediated inactivation and leads to excess thrombin formation. APC-R and Factor V Leiden: Clinical Features Prevalence in Venous Thrombosis Since the first description of APC-R, there have been over 500 publications pertaining to the subject. Consequently there is little doubt about the clinical significance of this coagulation abnormality in the pathogenesis of venous thrombosis. It also is becoming clear that additional risk factors for venous thrombosis such as protein C deficiency occurring in tandem with APC-R, increases the risk of thrombosis over that of APC-R alone. However as will be discussed later, the clinical significance of APC-R appears to be primarily limited to individuals of European descent. There is also evidence demonstrating that the APC-R prevalence can vary within the various European-derived populations. In the initial clinical studies conducted in Sweden and the U.S. using a highly selected patient population, it was found that approximately 40-60% of VT cases exhibited a poor anticoagulant response to APC (9,10). But in the first large-scale population- based case-control study in which 301 unselected patients with a confirmed VT diagnosis were investigated, only 21% of the patients presented with an impaired anticoagulant response to APC (17). This study which was conducted in the Netherlands also reported that 5% of the controls exhibited APC-R (17). Soon after the publications of these seminal papers, the DNA mutation responsible for approximately 90-95% of APC-R cases, Factor V Leiden, was reported (11,12,13 ). Similar to those early reports, other subsequent studies conducted within the Western Hemisphere using either the plasma based APC-R assay or DNA analysis reported a high prevalence of APC-R. In a retrospective French study Cadroy et al. (18) reported a prevalence rate of 19% in individuals with a history of VTE while in a large prospective series of unselected French patients, the prevalence of APC-R was found to be 14.5% in patients with VTE, and 3.5% in the controls (19). Similar findings were found in a large Italian series in which the prevalence of APC-R was found to be in 15% of patients with VTE and in 2% of the controls (20). In a reflection of differences of the APC-R prevalence rate in unselected patients within the European community, a recent Swedish report found that 28% of unselected patients with a confirmed history of VTE were APC-R positive as compared to 11% of the controls (21). In a large U.S. clinical/epidemiological study conducted among 14,1916 healthy predominantly white male physicians, Ridker et al (22) found the prevalence of APC-R to be 6% in those free of vascular disease. However the prevalence increased to approximately 25% in men who had their first primary VTE after the age of 60. Despite the importance of this finding, this study failed to address the relationship between VTE and APC-R in women or in minorities residing in the U.S. We have recently, however, reported a 1.2% prevalence rate in African-Americans with a confirmed history of VTE and a corresponding 1.3% rate in the controls (23). This low prevalence rate was not due to early mortality as the birth prevalence of APC-R in African-Americans was similar to that of the controls used in our study (24). Other studies have found a low prevalence rate in this population as well (25). It should be pointed out that this low prevalence rate does not necessarily imply that APC-R has no role in VTE in African- Americans but rather it is considered not to be a significant population-based risk factor. The differences in the APC-R prevalence have not only been attributed to ethnic origin but to the selection criteria of the study population (26). It also has been suggested that the prevalence of APC-R could have been underestimated in studies which excluded patients on coumarin because those patients were probably the ones who were most likely to have presented with recurrent thrombosis (27). Population Prevalence The prevalence of the FV Leiden genotype does vary among the populations of the world (Table III). In Europe where the highest prevalence is found, the rate among healthy individuals can range from 2-11%. A similar prevalence rate can also be found in countries outside Europe where the majority of the population is of European descent. Among the endogenous populations of Africa, Eastern Asia and the Americas, the prevalence is significantly lower and is rarely seen in individuals from Eastern Asia (28,29). This population difference is reflected in the U.S. In a study of over 4000 subjects without a history of cardiovascular disease from either the U.S. Physicians Health Study or the Womens Health Study, Ridker et al. (25) reported the FV Leiden frequency to be 5.27% for Caucasians, 1.23% for African Americans and 0.45 in Asian Americans (Table III) Influence of age on thrombosis in persons with FV Leiden Since FV Leiden is an inherited trait, it was thought that it would be a significant risk factor for thrombosis for those under the age of 45. Although several studies have reported as such, other studies have found that it could also be a significant risk factor for older adults. In a large study which examined at over 300 individuals from 50 families with a history of APC-R, Swedish investigators determined that the average age for the first thrombotic event in Factor V Leiden heterozygotes was 36, while it was 25 for those who were homozygous (30). Other large family studies have reported similar data ( ). However, in addition to family history, when other risk factors such as pregnancy, oral contraceptives and injury (surgical or trauma) are excluded, there is data to support that FV Leiden is also a risk factor for older adults (31). In the U.S. Physicians Health Study it was found that that the prevalence of FV Leiden was 25% in men who reached at least the age of 60 before their first thrombotic event (22). In a more detailed analysis of this U.S. Physicians Health Study data, Ridker et al (31) reported that in men under the age of 50, the incidence of venous thrombosis was the same irrespective of their FV Leiden status, whereas after the age of 50, the incidence of venous thrombosis increased in men with FV Leiden as compared to those without the mutation (31). Several studies have reported that FV Leiden may also play a major role in the development of both venous and arterial thrombosis in children (32,33). Clinical expression of thrombosis Perhaps the most important clinical determinants of FV Leiden expression are genotype (heterozygous or homozygous) and presence or absence of other genetic risk factors (Tables IV, V). In a study that looked at 471 consecutive patients, which presented with their first episode of venous thrombosis before the age of 70, investigators estimated that the relative risks for individuals heterozygous for FV Leiden was 7- fold and 80-fold for those who were homozygous (34). In this series of patients, it was of interest to the authors to note that not only were there no thrombotic events prior to adulthood reported in the homozygous individuals but also several of the homozygotes experienced other risk factors (i.e. pregnancy) without the occurrence of thrombosis (35). As illustrated in another report, age of the first VTE in homozygotes can be variable (35). In the investigation of a family in which 4 of 5 offspring were homozygous for FV Leiden, Greengard et al. (35) found that the 2 homozygous sons and the heterozygous son had their first thrombotic event between the ages of 16-23 years. In contrast, the two homozygous daughters at ages of 28 and 33 years had yet to experience symptoms of thromboembolic disease. This observation and other reports (35) have indicated that FV Leiden homozygosity is a relatively more benign disorder as compared to protein C and protein S homozygosity. The presence of other inherited genetic risk factors such as DNA mutations in protein C or protein S can have a profound effect on the risk of thrombosis for FV Leiden carriers (36, 37, 38). Furthermore, the co-existence of either homocystinuria or hyperhomocysteinemia with FV Leiden can significantly increase the risk of thrombosis (39,40). In a recent Dutch study that looked at 18 protein C deficient families, Koeleman et al. (36) found that the coexistence of FV Leiden with a heterozygous protein C genotype significantly increased the risk of thrombosis as compared to that of a carrier of a single defect. Likewise Brenner et al. (41) reported using data from a large Arab family with recessive protein C deficiency that 5 of the 11 protein C heterozygous members with FV Leiden had a history of VTE whereas 13 members heterozygous for only protein C had yet to present with VTE. Another study which predominantly consisted of individuals of British and Scandinavian descent reported a similar overall finding (37). However, the major exception was that the prevalence of FV Leiden was not as high in symptomatic protein C deficient individuals (37) as reported for the Dutch families and the one Arab family. Koeleman et al. (38) have also found that individuals heterozygous for protein S deficiency with FV Leiden also have significantly higher risk of thrombosis. Similar findings have observed in individuals who were both antithrombin III deficient and FV Leiden positive (42). Using data from the U.S. Physicians Health Study, Ridker et al (39) observed that subjects with the coexistence of hyperhomocysteimia and FV Leiden when compared to men without either abnormality had a 20-fold increase in risk for idiopathic VTE. It is quite clear from this data that the presence of dual genetic defects can be synergistic with respect to an increased risk for thrombosis. Risk of recurrent thrombosis Despite important clinical implications, the role that FV Leiden plays in recurrent thrombosis is less clear. In a relatively small retrospective study, Rintelen et al. (43) found that among heterozygotes there was no difference in the probability of recurrent thrombosis as compared to those without the mutation. The data did however suggest that the chance of recurrence may be higher in homozygotes (43). In a larger prospective study with a mean follow-up period of 20 months, Eichinger et al. (44) reported similar results. However they did not rule out that FV Leiden carriers may have a higher recurrence rate further out in the follow-up period (44). Ridker et al (45) with a mean follow-up time of 68 months did find a three-fold increase in recurrent thrombosis, and after a follow-up time up to 8 years (mean was 3.9 years) Simioni et al. (46) also found a significant increase of recurrent thrombosis in FV Leiden carriers. This data is suggestive that not only does the probability of recurrent venous thrombosis FV Leiden carriers increase as a function of time after the first event, but also that the chance of recurrence may depend on whether or not the individual is heterozygous or homozygous for the mutation. Arterial thrombosis Although the association of FV Leiden and arterial thrombosis is not clear, the evidence does suggest that its overall effect may be relatively minor both in myocardial infarction and stroke (22,47,48). However there may be specific associations. For example, one report has found an association with cryogenic stroke in the young (49), while another study has found an association with MI in young women who smoke (50). Pregnancy Since pregnancy and the use of oral contraceptives are well known risk factors for VTE, the clinical significance of the FV Leiden genotype is women is particularly important to discern. The importance is highlighted by the fact that in the U. S., VTE is the leading cause of maternal morbidity and mortality in both the ante partum and post partum period (51,52,53,54,55) The incidence of VTE in pregnant women increases approximately 3-5 fold following a cesarean section as compared to a vaginal delivery. It has been reported that approximately 15-24% of obstetric patients with an untreated deep vein thrombosis may develop a pulmonary embolism which is associated with a reported 12-15% mortality rate (55, 56). Hellgren et al. (57) recently reported a prevalence rate of approximately 59% APC-R in Swedish women who thrombosed during pregnancy. Another study in the U.S. which looked at 407 unselected nonconsecutive gravid women found that 28% of FV Leiden carriers had a documented VTE during pregnancy (54). It is of interest to note that in the latter study no recurrent thrombotic events were seen in subsequent pregnancies. A recent prospective study which looked at a highly selected group of Jewish-Arab women who were admitted to a high risk pregnancy unit found that the 46% of the patients were FV Leiden carriers (58). All carriers presented with a VT during pregnancy and when compared to the total study population the prevalence of FV Leiden in women who presented with VTE was 78% (58). It has been estimated that a pregnant FV Leiden carrier has an approximate 1 out 400 chance of having a thrombotic event (59). Complications of pregnancy Recent studies have made it clear that FV Leiden is also associated with other complications of pregnancy such as preeclampsia, fetal loss, and intrauterine growth retardation (IUGR) (60,61,62,63,64,65,66,67). In an analysis of Dutch women, Dekker et al. (67) found an APC-R prevalence rate of 16% in women with early onset preeclampsia, while in the U.S., Dizon-Townson et al. (64) reported a rate of approximately 9% in their patient series. While this discrepancy could be due to the ethnic origins of the two study populations, it could also be due to the classification criteria used for patient inclusion in each study (64). Other studies have suggested the possibility that the presence of a second mutation in tandem with FV Leiden such as MTHFR may also contribute to preeclampsia (68.69). Not only has recurrent fetal loss been attributed to placental infarcts, but it has also been suggested that many of these infarcts may be a consequence of thrombophilia. In an analysis of 128 pregnancies in 28 FV Leiden carriers or those with acquired activated protein C resistance, Brenner et al. (65) found over 50% of the women had first trimester abortions, 17% had late abortions and fetal death occurred in 47%. Perhaps the most revealing finding was that only 18% of the pregnancies resulted in a live birth (65). In a case-control study, Dizon-Townson et al. (61) reported a 2-fold increase in FV carrier frequency in women who presented with a spontaneous miscarriage and a 10-fold increase in those with a greater than 10% placental infarction. Oral contraceptives There have been several reports of an increased risk for VTE in female FV Leiden carriers who were on oral contraceptives (57, 70,71). It was estimated that those who were heterozygous had a 35-50 fold increase in risk, while the risk to the homozygous female was increased several hundred-fold (70). This increased risk may also extend to the third generation progestagen (71). Oral contraceptives may also decrease the APC-R plasma ratio (see Laboratory Testing section) in the absence of the FV Leiden mutation (72,73,74). Once off oral contraceptives, the ratio usually returns to normal within 1-2 months (72). It is unclear however whether or not this acquired APC-R effect is a risk factor for thrombosis. Myocardial infarction Although no significant gender difference appear to exist between FV Leiden and VT, there is data that suggests that FV Leiden may be a risk factor for MI in young women who smoke (50). Analysis of data from the U.S. Physicians Health Study found no association between FV Leiden and MI in older men (22). Other reports likewise have found no such association between FV Leiden and MI in other studies (48). It should be emphasized however that as yet no study has addressed the role of FV Leiden and MI in young male smokers. Significance of APC-R Oncology Patients It has been long recognized that venous thrombosis is a frequent occurrence in cancer patients. Rather than being due to an inherited trait, most thrombotic events in these patients are thought to be a consequence of host and treatment related events (75). In a study of 353 consecutive unselected predominantly white oncology patients, Otterson et al. (75) found a FV Leiden prevalence rate of 5.4% but no significant association with venous thrombosis. They did find however that the 2 patients with the FV Leiden genotype who thrombosed had a positive family history of venous thrombosis. It should be noted however that although the average follow-up was approximately 69 months, the occurrence of VTE further out in time cannot be excluded at the present time. Furthermore since this study included different histological subtypes, the suggestion that the FV Leiden genotype may have a positive association with certain types of cancer also cannot be ruled out. As pointed out by Green et al (76) who likewise found no association between FV Leiden and VTE in cancer patients, APC-R which can be found in cancer patients as measured by the coagulation assay is probably due to high factor VIII and fibrinogen levels and is not a predictor of thrombosis. The clinical laboratory can determine APC-R either by using a coagulation-based assay to determine the APC-R ratio or by using a DNA-based test for detection of the FV Leiden mutation. Although both tests can be and are frequently performed, many laboratories because of cost considerations, perform only the DNA test if an abnormal coagulation result is obtained. The numerical values from the coagulation-based assay can be used to identify individuals with APC-R by either using the APC-R ratio or the normalized ratio. The APC-R ratio is derived by determining the APTT with the addition of APC divided by the APTT (APC-R = APTT + APC), while the normalized ratio is obtained by dividing the APC-R ratio of the patient by the APC-R ratio standard (77). The difference between these two measurements with respect to distinguishing the APC-R phenotype is that the APC-R ratio measurement is less sensitive but more specific while the converse is true for the normalized ratio. This laboratory assay is not an indication of enzyme activity but rather a reflection of a response. This distinction is important in that a standard curve cannot be utilized but rather a reference range must be established for each clinical laboratory. It is advisable that the practitioner obtain the reference range directly from the laboratory it which the assay was performed before a decision regarding diagnosis is made. Although the FV Leiden DNA mutation can be detected by many different molecular techniques, the most commonly used method involves the use of the polymerase chain reaction to amplify the segment of the DNA where the mutation is located. Following amplification, a restriction enzyme is used to cut the amplified DNA segments into fragments. Once separated by gel electrophoresis, the DNA fragment pattern is used to determine whether or not the mutation is present. In order to minimize costs associated with genetic testing, the Coagulation and Molecular Diagnostic Laboratory of the University of Minnesota Medical Center has developed a testing algorithm for APC-R. Because of its sensitivity, the laboratory uses the normalized ratio to determine APC-R (77). If the normalized ratio was below 0.85, DNA testing was initiated. If the value was above 0.85, APC-R was excluded as a diagnosis and the thrombophilia work-up was extended to include other diagnostic possibilities. However since heparin can interfere with the coagulation assay, these investigators felt that DNA analysis should directly be done on individuals who are currently on heparin therapy (77). However there are cases when a APC-R phenotype as indicated by the coagulation assay is not associated with the FV Leiden genotype (77). In these particular cases, the phenotype is referred to as acquired APC-R. In many instances acquired APC-R can be linked to high VIII levels, presence of the lupus anticoagulant, anticardolipin antibodies, oral contraceptives, age, pregnancy and the female sex. Several groups have found that females have a lower APC-R ratio than men (72,73,77) and it has been advocated that separate APC-R reference ranges should be established for men and women (77). It would be of further use to establish similar reference ranges for the various ethnic groups. Clinical Implications and Management . In the U.S. Ridker et al. (25) have estimated that over 11 million Americans are FV Leiden carriers and 143,000 of these are homozygous for the mutation. Because of this high prevalence rate of FV Leiden in the U.S as well as in Europe, questions regarding the cost-benefit of genetic screening have been raised. Screening has specifically been advocated by some for women who are either using oral contraceptives or who wish to become pregnant (78,79). However others citing the lack of clinical data demonstrating a favorable benefit to risk ratio following anticoagulation prophylaxis feel that screening prior to oral contraceptive use or pregnancy may not be warranted (25). Ridker et al. (25) has further pointed out that the clinical risks of an unintended pregnancy may be greater than the risks associated with FV Leiden and oral contraceptives. Homozygosity or heterozygosity for FV Leiden in the absence of any symptomatology does not necessarily indicate that any type of prophylaxis treatment is required. This mutation even in the homozygous state does not appear to cause disease early in life as seen with protein S and protein C homozygosity. Furthermore FV Leiden does not appear to be associated with early mortality (80,81). Currently it is felt that the clinical management of thrombotic patients with APC-R should be no different than that of other thrombophilia patients. However prophylaxis treatment may be considered when a FV Leiden carrier encounters a high-risk situation such as surgery.
|
||||||||||||||||||||||