Thrombophilia Screen

The thrombophilia screen tests for inherited and acquired disorders that predispose to venous thromboembolism.

The inherited panel comprises antithrombin, protein C and protein S activity, activated protein C resistance with factor V Leiden genotyping, and the prothrombin G20210A variant. The antiphospholipid tests (lupus anticoagulant and antibodies) are usually requested with it, since they are the commonest acquired cause.

It identifies a defect only. It cannot predict whether or when thrombosis will occur, and the result seldom changes the management of an acute event.

Heterozygous factor V Leiden affects 3 to 8% of people of European ancestry and is rare in African and East Asian populations. The prothrombin variant affects about 2% of Europeans and is also rare in other populations. Antithrombin, protein C and protein S deficiency each affect fewer than 1 in 500 people, and antithrombin deficiency is the rarest (about 1 in 2000 to 5000).

Relative risk of a first venous event is about 10 to 20 times for antithrombin deficiency, 5 to 10 times for protein C or S deficiency, 3 to 5 times for heterozygous factor V Leiden and 2 to 3 times for the prothrombin variant. Homozygous factor V Leiden and compound heterozygosity carry a higher risk.

Indications. Testing is selective. It is considered after venous thromboembolism under 50 years without a clear provoking factor, and in the following settings:

• thrombosis at an unusual site (cerebral venous, splanchnic, upper limb without catheter)
• recurrent events
• thrombosis with a strong family history
• neonatal purpura fulminans
• warfarin induced skin necrosis
• women with a first degree relative who has a known high risk defect when oestrogen therapy or pregnancy is planned

Routine testing after a provoked event, in older adults, or to decide the duration of anticoagulation is of low yield.

Specimen and timing.

• Citrated platelet poor plasma from a clean venepuncture, centrifuged twice, tested promptly or frozen at minus 70 °C.
• Functional assays are unreliable during acute thrombosis, heparin, warfarin or direct oral anticoagulant therapy, pregnancy, oestrogen use, liver disease and acute illness. Testing is ideally done at least 3 months after an event and 2 weeks after stopping warfarin, and a low result is always confirmed on a second sample.
• DNA tests (factor V Leiden, prothrombin variant) are unaffected by these factors and can be done at any time.

Reference ranges are method specific. Local ranges take precedence.

Antithrombin. Serine protease inhibitor that neutralises thrombin and factors Xa and IXa, and is the cofactor for heparin. Range: activity 80 to 120 IU/dL by chromogenic assay. Raised: warfarin therapy, vitamin K deficiency, cholestasis. Lowered: inherited deficiency; unfractionated heparin (by 10 to 30%), acute thrombosis, DIC, liver disease, nephrotic syndrome, pre-eclampsia, oestrogen, L-asparaginase, sepsis, extracorporeal circuits, neonates (about 50% at birth, reaching adult values by 6 months).

Type I deficiency lowers both activity and antigen. Type II lowers activity with normal antigen, and the heparin binding site variants carry little thrombotic risk. Heterozygous deficiency gives levels of about 40 to 70 IU/dL. Antithrombin deficiency is the highest risk inherited defect, with thrombosis often before 40 years and a high risk in pregnancy. It should be suspected when heparin gives a poor anticoagulant response. A chromogenic anti-Xa or anti-IIa assay is preferred, and direct oral anticoagulants falsely raise results.

Protein C. Vitamin K-dependent protein that, with thrombomodulin, inactivates factors Va and VIIIa. Range: activity 70 to 140 IU/dL. Raised: no clinical significance, and a false rise occurs in clot based assays with direct oral anticoagulants. Lowered: inherited deficiency; warfarin and vitamin K deficiency, liver disease, DIC, sepsis, acute thrombosis, L-asparaginase, neonates (about 35 IU/dL at birth, reaching adult values by adolescence).

Heterozygous deficiency gives levels of about 30 to 65 IU/dL and many carriers remain asymptomatic. Homozygous or compound heterozygous deficiency (levels under 1 IU/dL) causes neonatal purpura fulminans. Protein C has a half life of about 8 hours, so it falls first on starting warfarin, and heterozygotes given warfarin without heparin cover can develop skin necrosis. A chromogenic assay is preferred, since clot based assays are affected by lupus anticoagulant, raised factor VIII, factor V Leiden and direct oral anticoagulants.

Protein S. Cofactor for activated protein C and tissue factor inhibitor. About 60% circulates bound to C4b-binding protein and 40% is free and active. Range: free antigen 60 to 130 IU/dL in men and lower in women (about 55 to 120 IU/dL), activity 60 to 140 IU/dL. Raised: no clinical significance. Lowered: inherited deficiency; pregnancy (falling to about 30 to 50 IU/dL by the third trimester), oestrogen, acute thrombosis, inflammation, liver disease, warfarin, vitamin K deficiency, DIC, HIV, nephrotic syndrome, L-asparaginase, neonates.

Type I deficiency lowers total and free protein S, type II lowers activity with normal antigen, and type III lowers free protein S with normal total. It is the least reliable assay. Activity assays are falsely low with factor V Leiden or high factor VIII, so free antigen is preferred and sex specific ranges are needed. A diagnosis of deficiency in a woman who is pregnant or taking oestrogen cannot be made.

Activated protein C resistance and factor V Leiden. The G1691A variant (Arg506Gln) removes a cleavage site in factor V, so factor Va resists inactivation. The phenotypic assay is an aPTT-based ratio with and without activated protein C, and plasma pre-diluted in factor V-deficient plasma gives specificity near 100% for factor V Leiden. A ratio under the local cut off (about 1.7 to 2.0) is abnormal and is confirmed by PCR genotyping. Acquired resistance occurs in pregnancy, oestrogen use, raised factor VIII, antiphospholipid antibodies and inflammation.

Heterozygotes carry about a 3 to 5 fold risk of venous thrombosis, and homozygotes carry about a 10 fold or greater risk. The risk for arterial events and pregnancy loss is small. The result is reported as absent, heterozygous or homozygous.

Prothrombin G20210A. A 3' untranslated region variant that raises prothrombin concentration by about 30% in heterozygotes. It is detected by PCR genotyping, since prothrombin activity overlaps the normal range. Heterozygotes carry about a 2 to 3 fold risk of venous thrombosis, with a larger risk when combined with factor V Leiden.

Higher risk results. Antithrombin deficiency, protein C deficiency, protein S deficiency, homozygous factor V Leiden or prothrombin variant, compound heterozygosity (factor V Leiden with prothrombin variant), and persistent antiphospholipid antibodies, particularly triple positivity.

Lower risk results. Heterozygous factor V Leiden or prothrombin variant alone.

Other acquired contributors. Persistently raised factor VIII (over about 150 IU/dL) carries about a 2 fold risk. Antiphospholipid syndrome, malignancy, paroxysmal nocturnal haemoglobinuria (needs flow cytometry), myeloproliferative neoplasms (JAK2 mutation in splanchnic vein thrombosis), nephrotic syndrome and heparin induced thrombocytopenia are the main differentials for a prothrombotic state. Homocysteine measurement is no longer used to assess thrombotic risk.

Effect of the clinical setting on results.

Critical findings needing immediate action.

• Neonatal purpura fulminans with undetectable protein C or S, with skin necrosis and thrombosis.
• Warfarin induced skin necrosis.
• Antithrombin deficiency with thrombosis during pregnancy or before surgery.
• Cerebral or splanchnic vein thrombosis with a positive screen.
• Multiple thromboses with antiphospholipid antibodies (catastrophic antiphospholipid syndrome).

Interferences and pitfalls.

• Testing during acute thrombosis or anticoagulation produces false deficiency or false normal results, and this is the commonest error.
• Direct oral anticoagulants give false normal results in clot based protein C, protein S and activated protein C resistance assays, and an adsorbent that removes the drug, or delayed testing, is needed.
• Lupus anticoagulant and raised factor VIII interfere with clot based assays.
• Low results need confirmation on a second sample, and persistent low activity with a normal antigen suggests a type II defect.
• Phenotypic assays do not replace genotyping for factor V Leiden.
• A negative screen does not exclude inherited thrombophilia, since many causes are unknown.

Special groups.

• Neonates and children: protein C, protein S and antithrombin are physiologically low, so age specific ranges are needed and deficiency is confirmed by repeat testing and genetic analysis.
• Pregnancy: protein S falls and acquired activated protein C resistance appears, so functional results cannot be interpreted, and genotyping is used.
• Family testing: restricted to relatives in whom the result would change management, such as women considering pregnancy or oestrogen therapy with a known high risk defect.
• Ancestry: factor V Leiden and the prothrombin variant are uncommon in African and East Asian populations.

Follow up tests.

• Lupus anticoagulant, anticardiolipin and anti-beta-2 glycoprotein I antibodies.
• Factor VIII.
• Repeat antithrombin, protein C and protein S after 3 months.
• Sequencing of SERPINC1, PROC and PROS1.
• Flow cytometry for paroxysmal nocturnal haemoglobinuria, and JAK2 V617F.
• Full blood count and film, liver function tests and urine protein.
• Imaging of thrombosis, and family testing.