Hypercoagulability Syndromes

REVIEW ARTICLE Hypercoagulability Syndromes

Robert H. Thomas, MD

ypercoagulability can be defined as the tendency to have thrombosis as a result of certain inherited and/or acquired molecular defects. Clinical manifestations of hypercoagulability can be devastating and even lethal. In the past 20 years, the origin of most of these diverse hypercoagulability syndromes has been elucidated. Cur- rently,H hypercoagulability disorders can be correctly diagnosed in approximately 80% to 90% of patients. Defining the cause of hypercoagulability may determine the type and duration of treatment for the associated thrombosis. The discovery of an occult carcinoma allows for the possibil- ity of early and possibly curative treatment. Finding a genetic defect in coagulation allows for testing of asymptomatic family members as well. The purpose of this review is to provide internists with a logical approach to the identification and treatment of hypercoagulability syndromes. Arch Intern Med. 2001;161:2433-2439

Most commonly, thrombosis is the result depending on the ethnic backgrounds of of more than one “hit.” For example, pa- persons in a particular geographic area. tients with the factor V Leiden defect may be asymptomatic until they start taking oral Disorder Incidence, % contraceptives. Patients with antithrom- Antiphospholipid antibody 28 syndrome bin deficiency may go on without inci- Activated protein C resistance 25 dent until they undergo a hernia repair. Elevated coagulation factor VIII 25 Also, multiple genetic defects predispose levels one to thrombosis much more than does Malignancy 15 a single defect.1,2 Some defects are known Sticky platelet syndrome 14 to be more powerful predictors than oth- Protein C deficiency 10 ers. Therefore, hypercoagulability is not Protein S deficiency 10 Homocystinemia 10 a uniform disease process but rather a host Prothrombin G20210A 5-10 of predisposing conditions that may or may Plasminogen deficiency 2-3 not be expressed as thrombosis, depend- Dysfibrinogenemia 1.5 ing on environmental insults and the Plasminogen activator inhibitor 1-3 strength and number of predisposing fac- increase tors. Tissue plasminogen 1 activator deficiency INDIVIDUAL SYNDROMES Antiphospholipid Antibody Syndrome The following is a list of the disorders that cause hypercoagulability and their ap- The antiphospholipid antibody syn- proximate incidences. Since these were de- drome is probably the most common of the rived from different studies, percentages hypercoagulable disorders. It is caused by cannot be exact. Also, incidences may vary, a heterogeneous family of immunoglobu- lins that bind to plasma proteins that have From the Department of General Medicine, University of Miami School of Medicine, an affinity for phospholipid surfaces. These Miami, Fla. antigens include B2 glycoprotein I, pro-

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2433

Downloaded From: https://jamanetwork.com/ on 09/28/2021 thrombin, high- and low-molecular- clude valvular abnormalities, livedo 80-fold.17 Heterozygous factor V weight kininogens, annexin V, ac- reticularis, superficial thrombophle- Leiden is, therefore, a relatively mild tivated protein C, and activated pro- bitis, ulcers, adrenal hemorrhage, fe- risk factor for thrombosis.15,18 The an- tein S. It is usually acquired and can tal wastage, chorea, transverse my- nual rate of thrombosis is 0.28%.19 Six be divided into the lupus anticoagu- elopathy, and thrombocytopenia. percent of patients will have a throm- lant syndrome and the anticardio- The treatment of patients with bosis by the age of 65 years.20 Sixty lipin antibody syndrome. Both of antiphospholipid antibody syn- percent of patients who experience these syndromes may be associated drome who have had thrombosis is thrombosis have a predisposing event, with other disorders, such as colla- long-term anticoagulation until the such as oral contraceptive use or preg- gen vascular diseases or infections, antibody has been absent for at least nancy.21 The presence of this muta- but are more often primary. Anti- 6 months.8 The drug of choice is low- tion does not appear to affect life ex- phospholipid antibody syndrome molecular-weight heparin sodium pectancy, and many patients will can also be associated with use of the since in 65% of patients warfarin so- remain asymptomatic. Therefore, pa- following medications: phenytoin, dium therapy fails8 and the interna- tients with no history of thrombosis quinidine, hydralazine, procain- tional normalized ratio is unreliable should not be treated prophylacti- amide hydrochloride, phenothi- in monitoring the intensity of therapy. cally with long-term anticoagula- azines, interferon, cocaine, qui- If warfarin must be used, an interna- tion.22 Functional tests for activated nine, and the combination product tional normalized ratio target range protein C resistance should be used of pyrimethamine and sulfadoxine. of 3 to 4 should be sought.9 The treat- to screen for the disorder, and posi- Usually, a patient will have one syn- ment of fetal wastage syndrome is be- tive results should be confirmed with drome or the other but not both. yond the scope of this review. There polymerase chain reaction for the ge- Multiple mechanisms as to the rea- is no clear indication for therapy in netic mutation. However, patients son for hypercoagulability have been asymptomatic persons; however, as- with phenotypic resistance to acti- postulated, but the exact cause is un- pirin therapy would be reasonable in vated protein C have an increased risk known at this time. The risk of this population because the risk of of thrombosis even if it is not due to thrombosis is 5.5% per year for thrombosis is higher than normal. factor V Leiden.20 Functional tests symptomatic patients.3 Other treatments, such as corticoste- may still be performed while pa- The lupus anticoagulant is di- roids, cyclophosphamide, and plasma tients undergo anticoagulation. rected against phospholipids, which exchange, have been used for se- then causes an in vivo prolongation verely symptomatic disease, but their Elevated Coagulation in the prothrombin time (PT), par- roles in routine management are not Factor VIII Levels tial thromboplastin time (PTT), or the well established. Russell viper venom time. These val- Elevated coagulation factor VIII lev- ues do not correct with normal Activated Protein C Resistance els appear to be nearly as common a plasma. However, the addition of risk factor for thrombosis as factor V phospholipids will correct the ab- Activated protein C resistance (eg, fac- Leiden. The Leiden Thrombophilia normality. Despite the prolonged co- tor V Leiden) is the most common in- Study found an 11% incidence in agulation times, thrombosis is the herited disorder that causes hyper- healthy controls and a 25% inci- predominant feature of this syn- coagulability. Factor V Leiden is dence in patients with venous throm- drome. The PT and PTT are not sen- present in 5% of whites but virtually bosis. The odds ratio for thrombosis sitive enough to be used as a screen- absent in Africans and Asians. How- was 4.8 for subjects with levels greater ing tool for the lupus anticoagulant. ever, 1% of African Americans have than 150 IU/dL vs those with levels Instead, the Russell viper venom time the mutation, reflecting racial mix- less than 100 IU/dL.23,24 For every 10- must be used. Venous thrombosis is ing.10 It results from a point muta- IU/dL rise in levels, the risk for a single much more common than arterial tion in the factor V gene, which causes episode of deep venous thrombosis thrombosis in these patients.4 the substitution of glutamine for ar- (DVT) increases 10% and the risk for The anticardiolipin antibody ginine at position 506.11 (Several other recurrent DVT increases 24%.25 Lev- syndrome is 5 times more common rare factor V gene mutations that can els of coagulation factor VIII are not than the lupus anticoagulant syn- lead to activated protein C resis- elevated because of the acute-phase drome.5 Antibodies can be detected tance have also been described.12-14) reaction but appear to be constitu- by enzyme-linked immunosorbent as- Consequently, 1 of 3 activated pro- tively increased in most patients with say. Both IgG and IgM are associated tein C cleavage sites is lost. The re- thrombosis, since coagulation factor with thrombosis.4,5 A total of 1% to sult is an impaired inactivation of fac- VIII levels are elevated indepen- 7% of asymptomatic individuals have tor V by activated protein C. Venous dently of C-reactive protein and fi- low titers of these antibodies.6 Even thromboses and fetal wastage may oc- brinogen, and 94% of patients con- asymptomatic persons have a 1% risk cur. It is not an important risk factor tinue to have high levels throughout per year of thrombosis. This in- for arterial disease except in the pres- long-term follow-up.26,27 Pregnancy creases to 6% in those with high ti- ence of smoking or other known risk and oral contraceptive use may also ters.7 This contrasts with a 0.1% risk factors.15,16 Those with factor V Leiden raise levels. The use of oral contra- per year in the general population. Ve- have a 2- to 3-fold risk for venous ceptives in patients with increased co- nous and arterial thrombi are equally thrombosis compared with healthy agulation factor VIII levels raises the common.4 Other manifestations in- subjects. The risk in homozygotes is risk of thrombosis 10-fold over pa-

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2434

Downloaded From: https://jamanetwork.com/ on 09/28/2021 tients with neither risk factor.28 The should be continued indefinitely un- function. Deficiency of protein C oc- genetic basis for increased coagula- til the patient is cured of the malig- curs in 1 of 250 controls.35 Protein tion factor VIII levels is not well un- nancy and is no longer receiving che- C is made in the liver and is vita- derstood at this time; however, one motherapy. If anticoagulation is min K dependent. It acts to inacti- small study25 showed high concor- contraindicated as with cerebral or vate factor V and factor VIII:C. It re- dance rates for first-degree adult fam- pericardial metastases, primary brain quires factor S as a cofactor and is ily members. tumors, or severe thrombocytope- activated by thrombin, when throm- nia, an inferior vena cava filter may bin is bound to thrombomodulin. Malignancy be placed. Long-term treatment may In families with thromboses be with low-molecular-weight hep- and protein C deficiency, thrombo- Cancer is the second most common arin or warfarin, although anec- ses begin in the late teens.36 Seventy- acquired cause of hypercoagulabil- dotal evidence suggests that hepa- five percent of affected individuals ity, accounting for 10% to 20% of rin may lead to fewer thrombotic will have 1 or more events.38 The spontaneous DVTs. Indeed, 15% of recurrences than warfarin.32 Cer- relative risk is 7.3.18 The annual in- patients with cancer have clinical tainly, warfarin failure should lead cidence is 1%.19 Seventy percent of thromboses and about 50% have to a switch to heparin. If heparin episodes are spontaneous.37 Both thromboses on autopsy.29 Cancer not fails, an inferior vena cava filter DVT and pulmonary embolism are only causes hypercoagulability but should then be placed. Thrombo- the most common manifestations. may also produce endothelial injury lytic agents should only be used in Superficial thrombophlebitis is also and venous stasis. Hypercoagulabil- patients with cancer who have a common.38 Arterial events are rare. ity is especially frequent in mucin- good prognosis and either pulmo- The optimal time to investigate secreting adenocarcinomas, brain tu- nary embolism with hemodynamic is at least 10 days after warfarin mors, acute promyelocytic leukemia, compromise or severe iliofemoral therapy is stopped, since both war- and myeloproliferative disorders. thrombosis of less than 4 days’ du- farin and acute thrombosis decrease Arterial thrombosis is much less ration. Of course, aggressive DVT protein C levels. Levels below 55% common than venous thrombosis prophylaxis with low-dose subcu- of normal are likely to be geneti- and is most often the result of non- taneous heparin or low-molecular- cally deficient; 55% to 65% is bor- bacterial thrombotic endocarditis or weight heparin (depending on se- derline. Abnormal results should al- disseminated intravascular coagula- verity and number of risk factors) ways be repeated for confirmation tion. Ninety percent of patients with should be carried out in patients with and family studies performed.39 cancer have clotting abnormalities, cancer who are hospitalized, immo- The short-term management of such as increased fibrinogen, clot- bilized, or undergoing surgery. thrombosis is with heparin or low- ting factors, fibrin degradation prod- molecular-weight heparin. Warfa- ucts, and platelets.30,31 Overt dissemi- Sticky Platelet Syndrome rin may be used for long-term treat- nated intravascular coagulation is ment; however, doses should be rare. There is no consensus as to the The sticky platelet syndrome is an started low and titrated upward value of measuring coagulation mark- autosomal dominant disorder that slowly only after heparin is thera- ers in predicting thrombosis in indi- results in platelets that are hyper- peutic because of the risk of warfa- vidual patients with cancer. aggregable to epinephrine and/or rin necrosis.40 In fact, one third of pa- Some cancers underlying spon- adenosine diphosphate. Venous or tients with warfarin necrosis have an taneous DVT are occult, early stage, arterial thrombosis may occur.33 Epi- underlying protein C deficiency. and curable. However, there is no sodes are more common during proof that aggressive diagnostic test- emotional stress. Retinal vascular Protein S Deficiency ing leads to improvement in sur- thrombosis appears to be associ- vival. Most experts recommend a ated with this entity. Fetal wastage Protein S is vitamin K dependent and thorough history and physical exami- may also occur. It is diagnosed with is synthesized by hepatocytes and nation, routine blood tests, chest x- platelet aggregation studies. Treat- megakaryocytes. It acts as a cofac- ray examination, urinalysis, and age- ment is with low-dose aspirin (81 tor for protein C. Fifty percent cir- and sex-specific screening, such as mg). If platelet aggregability does not culates free and 50% circulates prostate-specific antigen, Papanico- normalize, aspirin, 325 mg, may be bound to C4b binding protein. De- laou smear, lower endoscopy, mam- tried.34 If there is still no response, ficiency is transmitted autosomally mography, and fecal occult blood test- then clopidogrel (an adenosine di- dominant and can be quantitative or ing. Suspicious findings should be phosphate receptor antagonist simi- qualitative. aggressively evaluated. In addition, lar to but better tolerated than ticlo- Seventy-four percent of pa- patients without evidence of cancer pidine hydrochloride) may be used. tients develop DVT; 72% develop su- should be followed up closely for the perficial thrombophlebitis.41 The ensuing 2 years, during which time Protein C Deficiency relative risk of thrombosis is 8.5.18 virtually all occult cancers will be- The annual incidence is 1%19; 56% come clinically apparent. Protein C deficiency is an autoso- of episodes are spontaneous. Arte- The initial treatment of throm- mal dominant trait that may be rial events are uncommon. One half boses is the same as in patients with- caused by a decrease in absolute lev- of patients who develop thrombo- out cancer. However, treatment els of protein C or a decrease in its ses do so by the age of 25 years.41

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2435

Downloaded From: https://jamanetwork.com/ on 09/28/2021 Short-term therapy is stan- Antithrombin Deficiency activator deficiency, (3) increased dard. Long-term therapy is with war- plasminogen activator inhibitor, (4) farin or low-molecular-weight hep- Antithrombin is made in the liver congenital dysfibrinogenemia, and arin. Since warfarin necrosis may and endothelial cells. It inactivates (5) factor XII deficiency. Long-term occur, therapy should be started with thrombin and other serine prote- treatment may be with warfarin or warfarin at low doses and in- ases. Deficiency is an autosomal low-molecular-weight heparin for all creased slowly after heparin has been dominant disorder and occurs in 1 patients. administered. While the patient is of 5000 healthy blood donors.48 The Congenital plasminogen defi- undergoing warfarin therapy, pro- protein may be absent or dysfunc- ciency is a rare autosomal dominant tein C and S levels decrease by 50% tional. The normal concentration is disorder caused by either absent or within 48 hours and then increase 150 µg/mL. Thrombosis may occur dysfunctional plasminogen. Clini- to 70% of usual levels after 2 weeks. at less than 75% of this amount. Pa- cally, it mimics protein C and S de- Therefore, levels below 60% of nor- tients may present with DVT or ficiencies. Symptoms usually begin in mal while taking warfarin in the long pulmonary embolism. Mesenteric the late teens. Most commonly, it pre- term are suspicious for deficiency.8 vessels appear to be particularly sus- sents with DVT or pulmonary em- ceptible. Arterial events are rare. bolism. Arterial events are uncom- Homocystinemia Fifty percent of patients are asymp- mon. Events usually occur when tomatic. Thromboses occur early in plasminogen levels are less than 40% Elevated levels of homocysteine are life, with two thirds of patients pre- of the normal values. The results of known to be a risk factor for arterial senting by the age of 35 years. Forty routine coagulation studies are nor- and venous thrombosis and fetal percent of thromboses are sponta- mal.53 Treatment is standard. wastage. Homocysteine is an inter- neous.49 The relative risk of throm- Congenital deficiency of tis- mediate of methionine metabolism bosis is 8.1,18 and the annual inci- sue plasminogen activator and con- and, therefore, elevated levels may re- dence of thrombosis is 1%.19 genital increases of plasminogen ac- sult from cystathionine ␤-synthase Acute thrombosis, heparin, and tivator inhibitor are exceedingly rare. deficiency, homozygous expression other systemic diseases may de- Acquired abnormalities are more of the thermolabile form of methy- crease antithrombin levels.8 Warfa- common. They may occur with dia- lenetetrahydrofolate reductase, or rin may raise deficient levels into the betes mellitus, inflammatory bowel 50 from B12 or folic acid deficiency. Mild- normal range. Therefore, low lev- disease, and coronary atherosclero- to-moderate increases in homocys- els in a patient during acute throm- sis.54-56 teine occur in 5% to 10% of the popu- bosis or while taking heparin should Most congenital dysfibrinogen- lation.42 The relative risk of be confirmed when the patient is not emias occur in asymptomatic indi- thrombosis is 2.6.43 undergoing therapy. Likewise, nor- viduals (55% of patients) or cause Elevated homocysteine levels mal levels while the patient is tak- mild hemorrhagic disorders (20%). are thought to cause thromboses via ing warfarin should be confirmed Only 20% are associated with throm- several mechanisms, including (1) when the patient is not undergoing bosis.57 Venous thrombosis is most decreased protein C activation, (2) in- therapy. common but arterial events may oc- creased factor V activity, (3) induc- Treatment of acute thrombo- cur. They are usually autosomal tion of endothelial cell tissue factor sis is with low-molecular-weight dominant. They may be detected activity, (4) inhibition of thrombo- heparin because deficiency may with abnormal thrombin times or modulin expression and activation, cause resistance to unfractionated reptilase clotting times. Treatment (5) decreased antithrombin activ- heparin.51 In fact, heparin resis- of thrombosis consists of heparin or ity, and (6) enhanced affinity of li- tance may be a clue to the presence low-molecular-weight heparin fol- poprotein(a) and fibrin.44-46 of this deficiency. Lifelong therapy lowed by warfarin. Measurement of homocyste- should be considered for spontane- Factor XII deficiency is inher- ine levels is not well standardized, ous or recurrent thromboses. Pro- ited in autosomal dominant fash- and acute thrombosis may raise ho- phylactic treatment of asymptom- ion. It is involved in plasmin genera- mocysteine levels. Dietary supple- atic individuals is controversial but tion. Thus, patients will have a mentation with vitamin B6,B12, and usually is limited to high-risk situ- prolonged PTT, yet have a throm- folic acid can lower homocysteine ations, such as pregnancy or sur- botic diathesis. Arterial and venous levels.47 However, reduction of ho- gery. Antithrombin concentrate may thromboses and fetal wastage are mocysteine levels has not been be considered for situations in which common. Approximately 8% of de- shown to reduce thrombotic com- both thrombosis and bleeding may ficient subjects develop thrombo- plications. Folate supplementation occur, such as labor and delivery, ses.58 Factor XII deficiency should be (400 µg/d) may decrease levels by where anticoagulation might be con- suspected when a patient with throm- 52 30% to 42%. B12 supplementation traindicated. bosis has a prolonged PTT that cor- (100 µg/d) may decrease levels by rects with the addition of normal 15%. B6 supplementation (3 µg/d) Dysfibrinolysis plasma. A factor XII assay should then only reduces levels if there is a pre- be performed. Treatment is with low- existing deficiency. Thrombosis is There are 5 major forms of dysfibri- molecular-weight heparin followed treated in standard fashion in addi- nolysis: (1) congenital plasminogen by warfarin or continuation of low- tion to vitamin supplementation. deficiency, (2) tissue plasminogen molecular-weight heparin. Stan-

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2436

Downloaded From: https://jamanetwork.com/ on 09/28/2021 dard unfractionated heparin should not be used because of difficulties in Table 1. Genetic Disorders That Cause the Hypercoagulable Syndrome monitoring the PTT. Common Uncommon Prothrombin G20210A Sticky platelet syndrome Plasminogen activator inhibitor excess Protein C deficiency Heparin cofactor II deficiency Prothrombin G20210A mutation is Protein S deficiency Wein-Penzing defect Antithrombin deficiency Dysfibrinogenemia a relatively recently discovered de- Homocystinemia Tissue plasminogen activator deficiency fect in which there isaGtoAtran- Prothrombin G20210A Plasminogen deficiency sition at nucleotide position 20210. Increased factor VIII Tissue factor pathway inhibitor This mutation increases prothrom- Factor V Leiden bin activity and levels.59 It is found in 2.3% of healthy controls. The incidence is twice as high in people from southern Europe than from Table 2. Costs of Hypercoagulable Workup at the University of Miami northern Europe, and it is rare in Af- ricans and Asians.60 It may be de- Test Cost, $ tected through DNA analysis. At this Complete blood cell count with platelets 30.00 time, it must be considered a very Prothrombin and partial thromboplastin time 64.50 mild risk factor for venous and arte- Anticardiolipin antibodies 130.00 Lupus anticoagulant 240.00 rial thrombosis. The relative risk is Activated protein C resistance 75.00 approximately 2 to 3 times that of in- Confirmatory polymerase chain reaction for factor V Leiden 175.00 61-63 dividuals without the mutation. Factor VIII 90.00 Prothrombin G20210A 175.00 Other Hypercoagulable Platelet aggregation studies 360.00 Syndromes Protein C (functional) 105.00 Antigen 110.00 Protein S (functional) 105.00 Heparin cofactor II inhibits throm- Antigen 100.00 bin by mimicking the cleavage sites Free 130.00 of thrombin and forming a stable Homocysteine 324.00 complex with it, thus acting as a “suicide” substrate. Deficiency is rare and could theoretically cause throm- boxane levels. Thus, platelets are in All patients should have a com- botic potential, but its exact role is a state of increased activation. plete blood cell count performed, controversial. Heparin is effective in including platelets, to exclude my- the presence of heparin cofactor II INVESTIGATION OF eloproliferative disorders. Abnor- deficiency. HYPERCOAGULABILITY malities in PT and PTT suggest ei- Tissue factor pathway inhibi- ther the lupus anticoagulant or factor tor is a plasma component that binds Various clinical features should sug- XII deficiency. Antiphospholipid an- and inhibits factor Xa directly. This gest hypercoagulability, including tibodies should be obtained. Screen- complex then binds to the tissue fac- thrombosis at an early age (Ͻ50 ing for cancer as outlined earlier tor–factor VIIa complex, blocking its years), family history of thrombosis, should be performed. Patients in activity as well. Unstimulated plasma recurrent idiopathic thrombosis, whom this evaluation is negative and levels do not appear to be related to thrombosis at an unusual site (ex- all patients with a positive family his- thrombosis. However, plasma lev- cept for effort-related upper extrem- tory of thrombosis should undergo els measured 10 minutes after in- ity DVT66), spontaneous thrombosis testing for the common genetic dis- travenous heparin, 7500 U, is ad- or only mild provocation, unex- orders (Table 1). Table 2 pro- ministered correlate with venous plained spontaneous abortions, mas- vides a list of the approximate costs thrombosis.64 The role of tissue fac- sive thrombosis, and warfarin- of the various tests for hypercoagu- tor pathway inhibitor and its inci- induced skin necrosis. Information lability. Laboratory investigation for dence in thrombophilia are cur- from the history and physical exami- these disorders is generally unreli- rently unknown. nation determines the likelihood of able during acute thrombosis and Thrombomodulin mutations the underlying disorder. For ex- while undergoing anticoagulant have also been implicated in throm- ample, a young patient who pre- therapy. Thus, studies are optimally bophilia but prevalence and degree sents with a strong family history of performed while the patient is not of risk are unknown.65 thrombosis suggests a genetic disor- taking anticoagulants and is in the The Wein-Penzing defect is an der. A patient with systemic lupus ery- asymptomatic state. If tests are per- extremely rare deficiency of the li- thematosus is likely to have the an- formed while the patient is taking an- poxygenase metabolic pathway that tiphospholipid antibody syndrome. ticoagulants, knowledge of the alter- results in the compensatory in- An older patient with weight loss, ation of the individual factors by the crease of the cyclooxygenase path- early satiety, and epigastric pain is specific anticoagulant is essential. De- way and, therefore, elevated throm- likely to have a gastric carcinoma. pending on the level of suspicion for

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2437

Downloaded From: https://jamanetwork.com/ on 09/28/2021 Thrombosis Hypercoagulability Suspected?

Yes No Hypercoagulability No Further Workup Workup Standard Therapy

No Common Malignancy Antiphospholipid Sticky Platelet Syndrome Prothrombin G20210A >1 Abnormality Homocystinemia Protein C Deficiency Disorders Identified Antibody Syndrome (Family Studies) Factor V Leiden (Family Studies) Protein S Deficiency Are Other Family Elevated Factor Antithrombin Deficiency Members Affected? VIII Levels (Family Studies) (Family Studies)

Yes No Anticoagulate Anticoagulate Lifelong Antiplatelet Spontaneous or Consider Lifelong 3-6 mo of Consider Lifelong Evaluate for Rare No Further Workup Until Cancer Is Indefinitely Therapy Recurrent Anticoagulation Anticoagulation Anticoagulation No Longer Present Unless Antibodies Thrombosis? Genetic Defects Standard Therapy Confirm Effect With Lifelong Vitamin Persistently Negative Repeated Platelet Supplementation for > 6 mo Aggregation Studies

Thrombosis Yes No Recurrent? Consider Lifelong 3-6 mo Anticoagulation Anticoagulation Prophylaxis in High-Risk Situations

Yes No Consider Indefinite Treat 3-6 mo Therapy Prophylaxis in High-Risk Situations

Approach to treatment of hypercoagulability.

a genetic defect, referral to a hema- indefinite therapy in anyone with an coagulability cannot be general- tologist for testing of the rarer de- unprovoked thrombosis was termi- ized at this time. The clinician must fects may be indicated if the prior nated prematurely. Bauer68 divides pa- consider many patient factors with workup is unrewarding. A useful tients with hereditary defects into 2 statistical probabilities to deter- mnemonic for the common causes of groups: high risk (Ն2 spontaneous mine what conditions should be in- hypercoagulability is CALMSHAPES: episodes, 1 spontaneous life- vestigated. When a hypercoagu- protein C deficiency, Antiphospho- threatening thrombosis, 1 thrombo- lable syndrome is diagnosed, further lipid antibody syndrome, factor V sis at an unusual site, or 1 thrombo- judgment must be exercised to then Leiden; Malignancy, protein S defi- sis in the presence of Ͼ1 defect) and decide the best course of treat- ciency, Homocystinemia, Antithrom- moderate risk (asymptomatic indi- ment. Oversimplifications on the bin deficiency, Prothrombin viduals or 1 thrombosis in response evaluation and treatment of hyper- G20210A, increased factor VIII to a prothrombotic stimulus). In the coagulable syndromes are not help- (Eight), Sticky platelet syndrome. high-risk group, he recommends in- ful and may result in harm to indi- definite anticoagulation. In the mod- vidual patients. TREATMENT erate-risk group, he recommends vig- orous prophylaxis only for high-risk Accepted for publication April 9, 2001. When considering a patient for in- situations. Since no long-term stud- I would like to thank Barry Mat- definite therapy, many factors must ies have been performed comparing erson, MD, for his excellent advice and be considered: (1) the number, site, lifetime anticoagulation treatment assistance in editing the manuscript. and severity of thromboses; (2) with short-term anticoagulation Corresponding author and re- spontaneous vs provoked throm- therapy, definitive recommenda- prints: Robert H. Thomas, MD, Depart- bus; (3) the sex and lifestyle of the tions cannot be made at this time. Un- ment of General Medicine, University patient; (4) the strength of the pre- til these studies are performed, the of Miami School of Medicine, 1475 NW dictive value for thrombosis of the Figure may be used as a guide to the 12th Ave, Third Floor, Miami, FL particular hypercoagulable disor- evaluation and management of hy- 33136 (e-mail: [email protected]). der; (5) the compliance of the pa- percoagulable disorders. tient; and (6) the patient’s personal REFERENCES value construct. CONCLUSIONS Few guidelines exist for indefi- 1. Salomon O, Steinberg DM, Zivelin A, et al. Single nite therapy in hypercoagulable pa- The evaluation and treatment of a and combined prothrombotic factors in patients tients. A recent study67 that favored patient suspected of having hyper- with idiopathic venous thromboembolism: preva-

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2438

Downloaded From: https://jamanetwork.com/ on 09/28/2021 lence and risk assessment. Arterioscler Thromb vated protein C). Blood. 1997;89:1963-1967. 47. Franken DG, Boers GHJ, Blom HJ, et al. Treat- Vasc Biol. 1999;19:511-518. 23. van der Meer FJM, Koster T, Vandenbroucke JP, ment of mild hyperhomocysteinemia in vascular 2. DeStefano U, Martinelli I, Mannucci P, et al. The et al. The Leiden Thrombophilia Study. Thromb disease patients. Arterioscler Thromb Vasc Biol. risk of recurrent deep venous thrombosis among Haemost. 1997;78:631-635. 1994;14:465-470. heterozygous carriers of both factor V Leiden and 24. Koster T, Blann AD, Briet E, et al. Role of clotting 48. Tait RC, Walker ID, Perry DJ, et al. Prevalence of the G20210A prothrombin mutation. N Engl J Med. factor VIII in effect of von Willebrand factor on oc- antithrombin deficiency in the healthy popula- 1999;341:801-806. currence of deep vein thrombosis. Lancet. 1995; tion. Br J Haematol. 1994;87:106-112. 3. Finazzi G, Brancaccio U, Moia M, et al. Natural his- 345:152-155. 49. Thaler E, Lechner K. Antithrombin III deficiency tory and risk factors for thrombosis in 360 pa- 25. Kraaijenhagen R, Pieternella S, Koopman M, et al. and thromboembolism. Clin Haematol. 1981;10: tients with antiphospholipid antibodies: a four- High plasma concentrations of factor VIII:C is a 369-390. year prospective study from the Italian Registry. major risk factor for venous thromboembolism. 50. Marciniak E, Farley CH, DeSimone PA. Familial Am J Med. 1996;100:530-536. Thromb Haemost. 2000;83:5-9. thrombosis due to antithrombin III deficiency. 4. Bick RL. Antiphospholipid thrombosis syn- 26. O’Donnell J, Tuddenham EGD, Manning R, et al. Blood. 1974;43:219-231. dromes: etiology, pathophysiology, diagnosis and High prevalence of elevated factor VIII levels in 51. Francis J. Laboratory investigation of hyperco- management. Int J Hematol. 1997;65:193-213. patients referred for thrombophilia screening: role agulability. Semin Thromb Hemost. 1998;24:111- 5. Bick RL. The antiphospholipid thrombosis syn- of increased synthesis and relationship to the acute 126. dromes: fact, fiction, confusion and controversy. phase reaction. Thromb Haemost. 1997;77:825- 52. Menache D. Replacement therapy in patients with Am J Clin Pathol. 1993;100:477-480. 828. hereditary antithrombin III deficiency. Semin He- 6. Petri M. Epidemiology of the antiphospholipid syn- 27. O’Donnell J, Mumford A, Manning R, et al. Eleva- matol. 1991;28:31-38. drome. In: Asherson RA, Cervera R, Piette JC, tion of FVIII:C in venous thromboembolism is per- 53. Mammen EF. Plasminogen abnormalities. Se- Schoenfeld Y, eds. The Antiphospholipid Syn- sistent and independent of the acute phase re- min Thromb Hemost. 1983;9:50-51. drome. Boca Raton, Fla: CRC Press; 1996:13-28. sponse. Thromb Haemost. 2000;83:10-13. 54. Kirschsteln W, Simianer S, Dempfle CE. Impaired 7. Bick RL, Jakway J, Baker WF Jr. Deep vein throm- 28. Bloemenkamp KWM, Helmerhorst FM, Rosendaal fibrinolytic capacity and tissue plasminogen acti- bosis: prevalence of etiologic factors and results FR, Vandenbroucke JP. Venous thrombosis, oral vator release in patients with restenosis after PTCA. of management in 100 consecutive patients. Se- contraceptives and high factor VIII levels. Thromb Thromb Haemost. 1989;62:772-775. min Thromb Hemost. 1992;18:267-274. Haemost. 1999;82:1024-1027. 55. DeJong E, Porte RJ, Knot EA. Disturbed fibrinol- 8. Bick RL, Kaplan H. Syndromes of thrombosis and 29. Luzzato G, Schafer AI. The prethrombotic state in ysis in patients with inflammatory bowel dis- hypercoagulability. Med Clin North Am. 1998;82: cancer. Semin Oncol. 1990;17:147-159. ease. Gut. 1989;30:188-194. 409-458. 30. Rickles FR, Levine M, Edwards RL. Hemostatic al- 56. Juhan-Ungue I, Roul C, Alessi MC, et al. In- 9. Moll S, Ortel TL. Monitoring warfarin therapy in terations in cancer patients. Cancer Metastasis Rev. creased plasminogen activator inhibitor activity in patients with lupus anticoagulants. Ann Intern Med. 1992;11:237-248. non-insulin dependent diabetic patients– 1997;127:177-185. 31. Sun NC, McAfee WM, Hum GJ, et al. Hemostatic relationship with plasma insulin. Thromb Hae- 10. DeStefano U, Chirrolo P, Paciaroni K, et al. Epide- abnormalities in malignancy, a prospective study most. 1989;61:370-373. miology of factor V Leiden: clinical implications. Se- in one hundred eight patients, part I: coagulation 57. Ebert R. Index of Variant Human Fibrinogens. Boca min Thromb Hemost. 1998;24:367-379. study. Am J Clin Pathol. 1979;71:10-16. Raton, Fla: CRC Press; 1994. 11. Dahlback B. New molecular insights into the genet- 32. Sack GH, Levin J, Bell W. Trousseau’s syndrome 58. Goodnough LT, Saito H, Ratnoff OD. Thrombosis ics of thrombophilia: resistance to activated pro- and other manifestations of chronic dissemi- or myocardial infarction in congenital clotting fac- tein C caused by Arg 506 to Gln mutation in factor nated coagulopathy in patients with neoplasms: tor abnormalities and thrombocytopenias: a re- V as a pathogenic risk factor for venous thrombo- clinical, pathologic, and therapeutic features. Medi- port of 21 patients and a review of previously resis. Thromb Haemost. 1995;74:139-148. cine (Baltimore). 1977;56:1-37. ported cases. Medicine (Baltimore). 1983;62:248- 12. Chan WP, Lee CK, Kwong YL, et al. A novel mu- 33. Mammen EF. Sticky platelet syndrome. Semin 255. tation of Arg 306 of factor V gene in Hong Kong Thromb Hemost. 1999;25:361-365. 59. Simioni P, Tormene D, Manfrin D, et al. Prothrom- Chinese. Blood. 1998;91:1135-1139. 34. Mammen EF. Ten years’ experience with the “sticky bin antigen levels in symptomatic and asymp- 13. Williamson D, Brown K, Luddington R, et al. Fac- platelet syndrome.” Clin Appl Thromb Hemost. tomatic carriers of the 20210A prothrombin vari- tor V Cambridge: a new mutation (Arg 306 to Thr) 1995;1:66-72. ant. Br J Haematol. 1998;103:1045-1050. associated with resistance to activated protein C. 35. Tait RC, Walker ID, Reitsma PH, et al. Prevalence Blood. 1998;91:1140-1144. 60. Rosendaal FR, Doggen CJ, Zivelin A, et al. Geo- 14. Alhenc-Gelas M, Nicaud V, Gandrille S, et al. The of protein C deficiency in the healthy population. graphic distribution of the 20210 G to A prothrom- factor V gene A4070G mutation and the risk of ve- Thromb Haemost. 1995;73:87-93. bin variant. Thromb Haemost. 1998;79:706-708. nous thrombosis. Thromb Haemost. 1999;81: 36. Griffin JH. Clinical studies on protein C. Semin 61. Ridker PM, Hennekens CH, Miletich JP. G20210A 193-197. Thromb Hemost. 1984;10:162-166. mutation in prothrombin gene and risk of MI, 15. Ridker PM, Hennekens CH, Lindpainter K, et al. 37. Broekmans AW, Veltkamp JJ, Bertina RM. Con- stroke, and venous thrombosis in a large cohort Mutation in the gene coding for coagulation fac- genital protein C deficiency and venous throm- of US men. Circulation. 1999;99:999-1004. tor V and the risk of myocardial infarction, stroke boembolism: a study of three Dutch families. 62. Brown K, Luddington R, Williamson D, et al. Risk and venous thrombosis in apparently healthy N Engl J Med. 1983;309:340-344. of venous thromboembolism associated with a G Ј young men. N Engl J Med. 1995;332:912-917. 38. Broekmans AW. Hereditary protein C deficiency. to A transition at position 20210 in the 3 - 16. Siscovick DS, Schwartz SM, Rosendaal FR, et al. Hemostasis. 1985;15:233-240. untranslated region of the prothrombin gene. Br Thrombosis in the young: effect of atheroscle- 39. Bauer K. Hypercoagulable states. In: Hoffman R, J Haematol. 1997;98:907-909. rotic risk factors on the risk of myocardial infarc- Furie B, Cohen H, Benz E Jr, Silberstein L, Shattil 63. Hillarp A, Zoller B, Svensson PJ, et al. The 20210A tion associated with prothrombotic factors. S, eds. Hematology: Basic Principles and Prac- allele of the prothrombin gene is a common risk Thromb Haemost. 1997;78:7-12. tice. 2nd ed. Camden Town, London, England: factor among Swedish outpatients with verified 17. Rosendaal FR, Koster T, Vandenbroucke JP, et al. Churchill Livingstone Inc; 1995:1781-1794. deep venous thrombosis. Thromb Haemost. 1997; High risk of thrombosis in patients homozygous 40. Baker W Jr, Bick R. Treatment of hereditary and 78:990-992. for factor V Leiden. Blood. 1998;85:1504-1508. acquired thrombophilic disorders. Semin Thromb 64. Ariens R, Alberio G, Moia M, Mannucci P. Low lev- 18. Martinelli I, Mannucci PM, DeStefano V, et al. Dif- Hemost. 1999;25:387-406. els of heparin-releasable tissue factor pathway in- ferent risks of thrombosis in four coagulation de- 41. Engesser L, Broekmans AW, Briet E, et al. Heredi- hibitor in young patients with thrombosis. Thromb fects associated with inherited thrombophilia: a tary protein S deficiency: clinical manifestations. Haemost. 1999;81:203-207. study of 150 families. Blood. 1998;92:2353- Ann Intern Med. 1987;106:677-682. 65. Doggen CJ, Kunz G, Rosendaal FR, et al. A mu- 2358. 42. VanCott EM, La Posata M. Laboratory evaluation tation in the thrombomodulin gene, 127G to A cod- 19. Simioni P, Sanson BJ, Prandoni P, et al. Inci- of hypercoagulable states. Hematol Oncol Clin ing for Ala25Thr, and the risk of myocardial in- dence of venous thromboembolism in families with North Am. 1998;12:1141-1166. farction in men. Thromb Haemost. 1998;80:743- inherited thrombophilia. Thromb Haemost. 1999; 43. den Heijer M, Rosendaal FR, Blom HJ, Gerrits WB, 748. 81:198-202. Bos GM. Hyperhomocysteinemia and venous 66. Heron E, Lozinguez O, Alhenc-Gelas M, et al. Hy- 20. Rodeghiero F, Tosetto A. Activated protein C re- thrombosis: a meta-analysis. Thromb Haemost. percoagulable states in primary upper extremity sistance and factor V Leiden mutation are inde- 1998;80:874-877. DVT. Arch Intern Med. 2000;160:382-386. pendent risk factors for venous thromboembo- 44. Rees MM, Rodgers GM. Homocysteinemia: as- 67. Kearon C, Gent M, Hirsh J, et al. A comparison of lism. Ann Intern Med. 1999;130:643-650. sociation of a metabolic disorder with vascular dis- three months of anticoagulation with extended an- 21. Faioni EM, Razzari C, Martinelli I, et al. Resis- ease and thrombosis. Thromb Res. 1993;71:337- ticoagulation for a first episode of idiopathic ve- tance to activated protein C in unselected pa- 359. nous thromboembolism. N Engl J Med. 1999; tients with arterial and venous thrombosis. Am J 45. Harpel PC. Homocysteine, atherogenesis and 340:901-907. Hematol. 1997;55:59-64. thrombosis. Fibrinolysis. 1997;11:77-80. 68. Bauer KA. Management of patients with heredi- 22. Hille ET, Westenderp RG, Vandenbroucke JP, et 46. Bellamy MF, McDowell IFW. Putative mecha- tary defects predisposing to thrombosis includ- al. Mortality and causes of death in families with nisms for vascular damage by homocysteine. ing pregnant women. Thromb Haemost. 1995; the factor V Leiden mutation (resistance to acti- J Inherit Metab Dis. 1997;20:307-315. 74:94-100.

(REPRINTED) ARCH INTERN MED/ VOL 161, NOV 12, 2001 WWW.ARCHINTERNMED.COM 2439

Downloaded From: https://jamanetwork.com/ on 09/28/2021