INTRODUCTION

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Virchow identified hypercoagulability as a predisposing factor for thrombosis over 140 years ago, but until recently an abnormality affecting coagulation could be pinpointed in only a small minority of patients with thrombosis. Heritable deficiencies of the endogenous anticoagulants protein C, protein S, and antithrombin have been recognized for decades but they are uncommon, even in patients with familial thrombosis (1- 3). In the past several years, however, tremendous progress has been made in our understanding of the heterogeneity of thrombosis risk in the general population and in our ability to identify a specific, inherited predisposing factor in patients with thrombosis.

The most dramatic advance was the discovery of resistance to activated protein C (APC) by Dahlback in 1993 (4). This abnormality, caused by a point mutation in the factor V gene (5), is much more common than all previously recognized forms of heritable thrombophilia, combined. A subsequently described variant in the prothrombin gene, the 20210A allele, is also an important cause of heritable thrombophilia, though it is neither as prevalent nor associated with as high a thrombosis risk as is factor V Leiden (6). In addition, there is increasing evidence that mild hyperhomocystinemia, which results from the effects of both genetic factors and lifestyle on homocysteine metabolism, influences thrombosis risk (3, 7).

A genetic abnormality predisposing to thrombosis can now be identified in up to one-third of unselected patients with venous thromboembolism (VTE) and more than one-half of patients with familial thrombosis (8, 9). These newly identified abnormalities can coexist with one another as well as with previously recognized coagulation disorders, and multiple defects can often be found in patients with the most marked thrombotic predispositions (8, 10). Our increased ability to identify an underlying risk factor in many patients with thrombosis has raised important questions about testing for these abnormalities.

	RESISTANCE TO ACTIVATED PROTEIN C (FACTOR V LEIDEN)

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Protein C is an endogenous anticoagulant protein that, in its activated form (APC), cleaves and inactivates the activated forms of factors V and VIII (Figure 1). In 1993 Dahlback found that some persons with clinical hypercoagulability were resistant to APC, and this phenotype appeared to be inherited as an autosomal dominant trait (4). The molecular defect responsible for APC resistance was soon identified by investigators from Leiden, the Netherlands as a single point mutation in the factor V gene (5). This missense mutation causes an arginine to glutamine substitution in one of the protein's cleavage sites and renders activated factor V relatively resistant to cleavage, and thus inactivation, by APC. This single genotype, factor V Leiden, accounts for the phenotype of APC resistance in nearly all cases. It is now recognized as the most common cause of heritable thrombophilia.

View larger version (22K): [in this window] [in a new window]   Figure 1.   Simplified coagulation cascade and protein C anticoagulation system. Thrombin, in combination with thrombomodulin, activates protein C. Thrombin activates factors V and VIII and converts fibrinogen to fibrin. Activated protein C, with protein S as a cofactor, cleaves and inactivates factors Va and VIIIa. Abnormalities of the constituents marked with the star are the most common causes of heritable thrombophilia (factor V Leiden mutation, prothrombin G20210A mutation, deficiency of protein C, deficiency of protein S). Deficiency of antithrombin (not shown) is also a cause of heritable thrombophilia. APC = activated protein C; APC-PS = activated protein C with cofactor protein S; V = factor V; Vi = inactivated factor V; VIII = factor VIII; Vi = inactivated factor VIII; X = factor X; Xa = activated factor X.

The prevalence of factor V Leiden varies widely among ethnic groups. Overall, approximately 3 to 5% of whites are carriers of the mutation, whereas it is very rare in native African and Asian populations (11). Deep venous thrombosis (DVT) is the most common clinical manifestation of the thrombophilia associated with factor V Leiden. The mutation is found in approximately 20% of unselected patients with DVT and 40 to 60% of the selected patients referred to coagulation centers for evaluation (3, 9, 12). Based on case-control studies the risk of DVT appears to be increased 5- to 10-fold for heterozygous carriers of factor V Leiden, and approximately 80-fold for doubly affected homozygotes (3, 9).

Although carriers of the mutation are at increased risk of DVT, most affected persons will never have an episode of thrombosis. For a heterozygote from a thrombophilic kindred the risk for thrombosis by mid-adulthood is approximately 20% (12). Both additional genetic defects and clinical risk factors are important codeterminants of thrombosis risk. Thrombosis before adulthood is rare, and the incidence of venous thromboembolism (VTE) actually increases with age, likely reflecting increasing exposure to clinical risk factors (13). Pregnancy and the use of oral contraceptives both cause an acquired functional resistance to APC that, added to the resistance caused by the mutation, results in a multiplicative effect on thrombosis risk (3). The risk for thrombosis is higher, and the age at first thrombosis lower, in heterozygotes with a strong family history of thrombosis than for affected persons identified by population-based screening (10), likely due to coexisting heritable abnormalities in these kindreds. The incidence of recurrent DVT may be higher in factor V Leiden carriers, but studies have yielded conflicting results (14, 15). In the Physicians' Health Study the rate of recurrent DVT was higher only in the subset of patients whose initial thrombosis was classified as idiopathic (16). Limited evidence from family and population studies does not demonstrate a decrease in life expectancy in those with factor V Leiden (17, 18).

	PROTHROMBIN 20210A MUTATION

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Prothrombin is the precursor molecule of thrombin, which activates factors V and VIII and converts fibrinogen to fibrin (Figure 1). In 1996 investigators sequenced the prothrombin genes of probands from 28 families with unexplained thrombophilia and identified a G to A transition at nucleotide position 20210, in the 3'-untranslated region of the gene, that is associated with an increased risk for venous thrombosis (6). Patients with the 20210A allele have significantly higher levels of plasma prothrombin, which is believed to mediate the procoagulant effect. The gene variant is assumed to cause the elevation in prothrombin levels, but the mechanism is unknown (6).

The 20210A allele of the prothrombin gene has been confirmed to be one of the most prevalent genetic factors associated with venous thrombosis. It is present in 5.0 to 6.2% of unselected patients with venous thrombosis and 0.7 to 2.6% of control subjects (6, 8, 19, 20). Heterozygotes for the 20210A allele have a 2- to 5-fold increase in thrombosis risk compared with unaffected control subjects (6, 19, 20). Coinheritance of the 20210A variant appears to increase thrombosis risk in patients with other forms of thrombophilia. Patients with factor V Leiden or deficiency of protein C, protein S, or antithrombin who have had VTE are significantly more likely than control subjects to also carry the 20210A allele. The number of thrombotic events per patient is also greater for such compound heterozygotes (8).

	HYPERHOMOCYSTINEMIA

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Homocysteine is a sulfhydryl-containing amino acid derived from the metabolism of methionine. Hyperhomocystinemia can be caused by genetically or nutritionally determined abnormalities of homocysteine metabolism (7). VTE, accelerated atherosclerosis, and arterial thrombosis are well-recognized manifestations of severe hyperhomocystinemia (homocystinuria), usually caused by homozygous deficiency of the enzyme cystathione -synthase. Modest hyperhomocystinemia may be due to other genetic abnormalities, nutritional deficiencies of vitamins involved in homocysteine metabolism (B₆, B₁₂, and folate), or a combination of genetic and nutritional factors. The genetic abnormality that most commonly results in modest homocystinemia is homozygosity for a thermolabile, mutant form of the enzyme methylenetetrahydrofolate reductase (MTHFR). Modest hyperhomocystinemia is an established risk factor for coronary artery disease and occlusive arterial vascular disease, and there is accumulating evidence it is associated with increased risk for VTE, especially in patients with another thrombophilic condition.

Many studies with various experimental designs have found a correlation between modest elevation of plasma homocysteine and VTE, as recently reviewed (7). The best data come from the Leiden Thrombophilia Study, which compared 269 consecutive patients with a first episode of DVT to age- and gender-matched control subjects (21). Ten percent of the patients had homocysteine levels that were above the 95th percentile of the control group. The relative risk for thrombosis for the hyperhomocystinemic group was 2.5, and it increased with increasing concentration of homocysteine with what appeared to be a threshold effect. The effect of homocysteine on VTE risk appears to increase with age and to be greater in women than men (3). Coexisting hyperhomocystinemia increases the risk for thrombosis in patients with factor V Leiden (22). In the Physicians' Health Study patients with both disorders had a relative risk for venous thrombosis of nearly 10 compared with patients with neither disorder, while the risk for patients with only one abnormality was intermediate (22).

It is not clear how homocysteine affects thrombosis risk. In vitro, homocysteine has multiple potentially thrombogenic effects, including injury to vascular endothelium and antagonism of the synthesis and function of nitric oxide (7). Interestingly, the thermolabile MTHFR gene variant is not independently associated with thrombosis, highlighting the complex relationship between genetic and environmental factors in determining total homocysteine levels and, likely, thrombosis risk.

The relationship between plasma homocysteine levels, nutritional and lifestyle factors, and venous thrombosis requires further study. Because supplementation with folate and vitamins B₆ and B₁₂ can substantially lower homocysteine levels in patients with either genetic or nutritional causes of hyperhomocystinemia (21), the role of dietary therapy in reducing thrombosis risk warrants investigation. Hyperhomocystinemia may prove to be the most easily treated of the thrombophilias.

	DEFICIENCIES OF PROTEIN C, PROTEIN S, AND ANTITHROMBIN

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Deficiencies of the endogenous anticoagulant proteins protein C, protein S, and antithrombin (AT) were the first identified genetic causes of thrombophilia. Until several years ago they accounted for most of the few cases in which a heritable cause of thrombosis could be identified. In unselected patients presenting with DVT the aggregate prevalence of these three conditions is approximately 7% (1, 2). Even in the highly select subgroup of patients with recurrent VTE at a young age and a family history of thrombosis, less than one-third will have a deficiency of either protein C, protein S, or AT (1). In contrast to APC resistance and prothrombin 20210A, which are monogenic disorders, deficiency of each of these three proteins may be caused by a number of different mutations, resulting in either quantitative or functional deficiency. In general, these deficiencies have an autosomal dominant pattern of inheritance and homozygotes usually die in utero or shortly after birth. The clinical expression of heterozygosity varies widely, and it is increasingly clear that the coexistence of other genetic abnormalities causing a predisposition to thrombosis contributes to this phenotypic diversity (8, 10). Earlier estimates of thrombosis risk associated with deficiencies of these proteins were largely derived from case-control studies and investigation of thrombophilic kindreds. These estimates are now recognized as excessive, with some of the apparent high risk being attributed to the presence of multiple genetic defects in those kindreds.

Protein C

Heterozygous protein C deficiency has been associated with DVT, pulmonary thromboembolism, superficial venous thrombophlebitis and, rarely, arterial thrombosis. Some protein C- deficient heterozygotes remain asymptomatic. An analysis of protein C antigen levels in more than 5,400 normal, unselected adults suggested a prevalence of heterozygotes of 1:200 to 1:300 (23). Affected persons identified by such population-based screening do not appear to have a substantially increased risk for thrombosis. None of the cases had personal or family histories of thrombosis when identified, and a longitudinal study of 23 such asymptomatic heterozygous blood donors followed for 5 yr found that the incidence of VTE was minimal (24). In contrast, protein C-deficient patients who have had a thromboembolic event have a comparatively greater risk for thrombosis, as do their family members (25, 26). One small prospective study of protein C-deficient patients from a thrombophilic kindred, but with no personal history of thrombosis, found a 2.5% annual incidence of thromboembolic events (27). By the age of 45 yr, the heterozygous family members of a symptomatic patient with heterozygous protein C deficiency have a 50% chance of having VTE (25), with half or more of these events unassociated with situations of increased clinical risk. Pregnancy is well recognized as a high-risk clinical situation for protein C-deficient individuals. Thrombosis-prone and asymptomatic heterozygotes may have similar serum protein C antigen levels (25) or even the same mutation in the protein C gene (28). The coexistence of additional thrombophilic conditions in a subset of patients probably explains much of this clinical variability (8, 10, 29). Factor V Leiden has a high prevalence among symptomatic protein C-deficient persons, and individuals who are heterozygous for both conditions have a more than twofold risk of thrombosis when compared with family members with protein C deficiency alone (29).

Protein S

Protein S is a cofactor for APC (Figure 1). The prevalence of protein S deficiency in a large sample of unselected subjects has not been determined. Two large studies of consecutive patients with DVT found a prevalence of protein S deficiency of 1 to 2% (1, 2). Family members of symptomatic protein S-deficient heterozygotes are at increased risk for thromboembolic disease, and are more likely to have recurrent, juvenile, and idiopathic thromboembolism (30). A prospective case-control study of 24 asymptomatic protein S-deficient patients found the incidence of thromboembolic complications to be 3.5% per year (27). In a study of 12 kindreds with protein S deficiency, the probability of remaining free of thrombosis at age 35 was only 32% (30). Pregnancy-related thromboembolic events occur at an increased rate that is comparable to that seen in protein C-deficient patients (31). The prevalence of factor V Leiden (10) and the prothrombin 20210A allele (8) are increased in thrombophilic kindreds with protein S deficiency, and the risk of thrombosis is increased for persons with both defects.

Antithrombin

AT is an endogenous anticoagulant that binds with and blocks the biologic activity of thrombin and other activated coagulation proteins involved in the clotting cascade. As with deficiencies of proteins C and S, there is striking phenotypic variability among AT-deficient patients. In a cohort of 28 asymptomatic heterozygotes identified by screening of blood donors, only one suffered VTE over a 5-yr follow-up period (24). In contrast, thromboembolic events are common among AT-deficient subjects identified on the basis of a personal or family history for thrombosis. About half of these selected AT-deficient patients develop thrombosis before the age of 40 and about half of events occur without provocation (32). The incidence of pregnancy-related thrombosis in AT-deficient women is greater than that of either protein C- or protein S-deficient women, at 44 to 68% (31). The prevalence of factor V Leiden is increased in thrombophilic kindreds with AT deficiency, and limited evidence suggests that the likelihood of thromboembolism is substantially increased in patients who are heterozygous for both conditions (10).

	TESTING FOR HERITABLE THROMBOPHILIA

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Technical Aspects of Testing

APC resistance and factor V Leiden. Both phenotyping and genotyping can be used to screen for APC resistance or factor V Leiden, respectively. For phenotype testing an activated partial thromboplastin time (aPTT)-based assay is done with and without the addition of APC, and the results expressed as a ratio. The magnitude of APC resistance, as reflected by a lower ratio, appears to correlate with thrombosis risk, with the lowest ratios found in persons homozygous for factor V Leiden. Excellent sensitivity and specificity have been reported with a modified version of Dahlback's original functional assay (33), though meticulous specimen handling is important. Alternatively, factor V Leiden can be easily identified by molecular analysis of genomic DNA. The accuracy of genotyping is not affected by anticoagulation with either heparin or warfarin, and heterozygotes and homozygotes can be clearly distinguished (5). However, molecular analysis is considerably more expensive than functional testing. Both the phenotype and genotype provide complementary information, and the APC resistance assay is usually used as a screening test, with genotyping being used to confirm abnormal phenotypes.

Prothrombin 20210A. The prothrombin 20210A allele can be detected by molecular genetic analysis. There is no functional assay for prothrombin 20210A.

Hyperhomocystinemia. Homocysteine circulates in plasma in both free and protein-bound forms, with the protein-bound form predominating. Total plasma homocysteine should be measured in the fasting state. Normal plasma homocysteine concentration (50th percentile of the population) is approximately 10 µmol/L (21). Levels above the 95th percentile of an appropriate control group are usually considered to be elevated, generally corresponding to levels greater than 17.25 to 18.5 µmol/L (21, 22). Control subjects should be matched for age and gender as levels tend to be higher in men and to increase with age. Measurement of total homocysteine before and after a methionine load increases the sensitivity for detection of more subtle metabolic abnormalities but such provocative testing is more complex and costly. Molecular analysis for mutations affecting the metabolic pathways for homocysteine can be performed, but it is primarily a research tool (7).

Protein C, protein S, and antithrombin deficiency. The variety of abnormalities causing the protein C, protein S, and AT deficiency makes testing particularly complex. Screening tests for these disorders usually consist of combinations of functional assays and quantitative testing (35). The lack of a single molecular defect causing each of these deficiencies makes molecular analysis impractical for routine use. Anticoagulant therapy may interfere with testing for these deficiencies.

Whom to Test

Before 1993 testing for thrombophilia was recommended only for patients with recurrent, juvenile thrombosis and a family history of thrombosis (1, 34). The subsequent discoveries of several more prevalent forms of thrombophilia have led to suggestions that testing be more broadly applied. No consensus has emerged regarding whom to test for the "newer" thrombophilias, and the issue of expanded criteria for testing requires careful consideration. For example, in patients with similar genotypes, the risk for thrombosis varies; the relative risks derived from case-control studies may not apply to asymptomatic individuals identified through screening (24, 35). Screening persons who have not had thrombosis will detect more "low-risk" affected individuals and result in more false-positive test results. In addition, although individuals with APC resistance or the prothrombin 20210A allele have an increased relative risk for VTE, their absolute risk is low. Anticoagulation of patients who have not had a thrombotic event is not justified. Although aggressive risk factor avoidance and DVT prophylaxis in affected persons may be reasonable, performing screening solely to enhance prevention is unlikely to be cost-effective. Therefore screening of individuals who have not suffered a thrombotic event should be limited to family members of a patient with heritable thrombophilia of any type, who should be offered both testing and counseling (35).

For patients with an initial episode of VTE, routine testing for thrombophilia would be warranted if an alteration in type or duration of initial anticoagulant therapy, or of subsequent prophylactic therapy during periods of clinical risk, reduced the incidence of recurrent DVT. The relatively mild thrombotic tendency caused by factor V Leiden or the prothrombin 20210A allele does not justify lifelong anticoagulation after a single episode of DVT. However, extending the period of initial anticoagulant therapy beyond the standard 3 mo might be of benefit, and this approach is being studied in ongoing prospective trials. Until there is evidence to support such a change in management based upon the presence of either APC resistance or the prothrombin 20210A allele, testing for these abnormalities should not be routinely performed in patients with a first episode of DVT. Similarly, there are no data to support altering the duration or type of therapy for a thrombotic event in a patient with mild hyperhomocystinemia. If the results of clinical trials show that therapy with folate and/or the B vitamins reduces not only plasma homocysteine but also the vascular consequences of hyperhomocystinemia, measurement of homocysteine will likely become standard practice for all patients with VTE.

At present, then, substantial broadening of the population tested for thrombophilic disorders cannot be recommended. Patients with recurrent thrombosis before age 40 yr and a family history of thrombosis should be tested for factor V Leiden, prothrombin 20210A, and specific factor deficiencies, because multiple defects may coexist and the presence of more than one abnormality has important implications. Ongoing and future research may define situations in which it is appropriate to either test a broader range of patients or to perform a more limited array of studies on those patients selected for testing.

Given that there is phenotypic variability among patients with similar genotypes, and that in a significant minority of cases of heritable thrombosis the causative abnormality can-not yet be identified, phenotype, rather than genotype, should remain the most important determinant of duration of anticoagulant therapy. Patients with recurrent thrombosis or thrombosis of unusual site or severity have demonstrated a propensity to thrombosis and should be considered candidates for prolonged anticoagulant therapy regardless of the presence or absence of laboratory evidence of thrombophilia. Ultimately, decisions regarding testing and therapy must be individualized (35). In some cases, the presence or absence of one of these abnormalities may be a helpful factor to consideralong with severity of thrombosis, family history, comorbid conditions, lifestyle factors, and patient preferencein making a recommendation on the duration of anticoagulant therapy for a patient with DVT.

	CONCLUSION

TOP ABSTRACT INTRODUCTION RESISTANCE TO ACTIVATED PROTEIN PROTHROMBIN 20210A MUTATION HYPERHOMOCYSTINEMIA DEFICIENCIES OF PROTEIN C, TESTING FOR HERITABLE CONCLUSION REFERENCES

Our understanding of heritable thrombophilia has improved tremendously over the last several years. A heritable abnormality can now be found in many patients with VTE. Both APC resistance and the prothrombin 20210A variant differ from previously described heritable thrombophilias (protein C, protein S, and antithrombin deficiencies) in that they are surprisingly prevalent and monogenic. Hyperhomocysteinemia is unique in arising from a combination of genetic and lifestyle factors. There is considerable heterogeneity in the clinical expression of these disorders. Many affected patients will never have an episode of thrombosis, and those patients with the greatest propensity to thrombosis often have more than a single abnormality. Genetic abnormalities are best considered risk factors for VTE rather than causes of it, and they act in concert with other genetic and circumstantial risk factors. Despite the higher prevalence of the "newer" forms of thrombophilia and the relative ease with which they can be identified in the laboratory, broader use of testing cannot be recommended until we have data supporting changes in the management of those patients found to be affected.