Physiological deficiencies Neonates	Vitamin K deficiency Neonates Gastrointestinal disease Vitamin K antagonists Biliary obstruction Liver disease Miscellaneous, e.g. cephalosporins, parenteral nutrition
Liver disease	Disseminated intravascular coagulation
Drug induced bleeding Anticoagulants Thrombolytic agents Anti-platelet agents Miscellaneous	Renal disease and renal failure
Massive blood transfusion	Cardiopulmonary bypass and extracorporeal circuits
Acquired inhibitors of coagulation Factor VIII inhibitors Other factor inhibitors Anti-phospholipid antibodies	Miscellaneous Snake venoms and other toxic agents Myeloproliferative disorders Malignancy Paraproteinemias

Physiological deficiencies

Neonates

The coagulation system of the newborn infant is complex and reflects hepatic immaturity. Most of the clotting factors are present in reduced concentration in the newborn infant apart from factors V, VIII and fibrinogen.¹^–³ These physiological deficiencies in clotting factors result in prolongation of the prothrombin time (PT) and activated partial thromboplastin time (APTT) and as a consequence of this, reference ranges reflecting both gestational and neonatal age must be used to assess coagulation in the neonate.¹^–³ The platelet count is normal in the neonate although there may be a qualitative platelet abnormality. Fibrinolysis in the neonate is similar to that of adults.

The pattern of bleeding seen in neonates – umbilical bleeding, cephalohematomas, bleeding after circumcision, oozing after venepuncture and bleeding into the skin – is different from that seen in adults.

Drug-induced bleeding disorders

Drugs are a common cause of an acquired bleeding disorder. In many cases the drug may be obvious, e.g. an anticoagulant, but in other cases it may be less clear, as with the inhibitory effect on vitamin K metabolism observed with some cephalosporins.⁴

Heparin

Unfractionated heparin (UFH), the low molecular weight heparins (LMWHs) and fondaparinux (a synthetic pentasaccharide) are anticoagulants that potentiate the action of antithrombin by increasing its inhibitory activity.⁵ The inhibitory activity of UFH is directed against both thrombin (IIa) and factor Xa whereas that of the LMWH is primarily against factor Xa.⁵ Fondaparinux has exclusively anti-Xa activity. Bleeding in patients receiving heparin is usually secondary to excessive anticoagulation. Heparin is metabolized by the liver and excreted by the kidneys and LMWHs may accumulate in patients with impaired renal function⁶ and a dosage adjustment may be required if the creatinine clearance is less than 30 ml/min.

In patients receiving unfractionated heparin intravenously the rate of major hemorrhage ranges from 0%–7%. Fatal bleeding with a 5–14 day course of heparin ranges from 0%–2%. The risk of hemorrhage is significantly increased if there is concomitant use of other anticoagulants particularly anti-platelet agents such as aspirin or clopidogrel. Other patient-specific factors, including impaired renal function, disordered liver function, thrombocytopenia and invasive procedures, significantly increase the risk of bleeding. In individuals who are actively bleeding, unfractionated heparin can be effectively neutralized by protamine sulphate, a strongly basic drug that binds to the heparin. A dose of 1 mg of protamine sulphate will neutralize approximately 100 units of heparin. In overdose, protamine sulphate can function as an anticoagulant and no more than 50 mg of protamine sulphate should be administered at any one time. Protamine sulphate neutralizes only 60% of the anti-Xa activity of the low molecular weight heparins and is, therefore, less effective in correcting the bleeding problems associated with their use.⁷ Protamine sulphate does not bind to fondaparinux and is, therefore, of no value in the management of patients on fondaparinux who are bleeding.

Laboratory monitoring ⁸

1. Unfractionated heparin: therapeutic anticoagulation with UFH is monitored by means of the APTT aiming to maintain the APTT at 1.5–2x the mid-point of the reference range. In patients in whom the APTT is prolonged prior to the initiation of anticoagulation therapy, e.g. factor XII deficiency or a lupus anticoagulant, the measurement of anti-Xa levels may be necessary.

2. LMWH and fondaparinux: LMWHs have primarily anti-Xa activity and have little effect upon the APTT unless given in overdose. For patients receiving therapeutic LMWHs, routine monitoring is not usually indicated but may be in children, obese or underweight patients, or those with renal disease, pregnancy or unexpected bleeding or thromboses while on treatment. In patients receiving thromboprophylaxis with a LMWH, routine monitoring is not usually necessary. Fondaparinux inhibits exclusively Xa and so its activity, when necessary, is monitored using an anti-Xa assay. The elimination of fondaparinux from the body is reduced in individuals with renal impairment as the drug is primarily excreted in the urine. The clearance of fondaparinux from the body is reduced by approximately 25%, 40%, and 55% in patients with mild, moderate, and severe kidney impairment, respectively.

Hirudin and bivalirudin

Hirudin was originally isolated from the medicinal leech Hirudo medicinalis but now is available in a recombinant form (r-hirudin or lepirudin). Bivalirudin is a small (MW 2180 Da) synthetic peptide modeled after hirudin, which contains 20 amino acids and two thrombin-binding domains. Bivalirudin and lepirudin are direct thrombin inhibitors and bind to both the catalytic site and the anion-binding exosite of circulating and clot-bound thrombin but do not require antithrombin for their anticoagulant activity.⁹^–¹¹ Lepirudin and bivalirudin inhibit the conversion of fibrinogen to fibrin but also other thrombin-catalyzed reactions, for example activation of clotting factors and thrombin-induced platelet aggregation.

Lepirudin is a potent anticoagulant but has a very narrow therapeutic window and plasma levels of lepirudin show high levels of inter-individual variability even when the dose is adjusted for body weight. Over-anticoagulation with lepirudin can lead to severe bleeding problems. Lepirudin has a short half-life and its natural clearance through the kidneys may be sufficiently rapid such that in cases of overdose, specific neutralization is not required. In bleeding patients rVIIa or prothrombin complex concentrates may be of value or alternatively plasmaphoresis or exchange transfusion may remove lepirudin from the circulation.

Newer anticoagulants

Dabigatran etexilate is a pro-drug that is converted to dabigatran by esterase-catalyzed hydrolysis in the liver and the plasma. Dabigatran is a direct thrombin inhibitor which inhibits both clot-bound and free thrombin. Dabigatran is cleared primarily by the kidneys (85%) and a smaller amount (6%) is excreted in the feces.

Dabigatran is licensed for the primary prevention of venous thromboembolism in adult patients who have undergone elective total hip or knee replacement surgery. Laboratory monitoring of dabigatran is not routinely indicated although it is possible that in some cases, for example in the bleeding patient, this may be of value. The APTT shows a non-linear response to increasing doses of dabigatran and appears less sensitive and less precise than an ecarin clotting time (ECT). The ECT may provide a method for monitoring patients on dabigatran. To date no simple mechanism exists for the rapid reversal of dabigatran in, for example, the bleeding patient.

Rivaroxaban is an oral and highly selective inhibitor of factor Xa which inhibits free Xa and Xa bound to the prothrombinase complex. Rivaroxaban is licensed for the primary prevention of venous thromboembolism in adult patients who have undergone elective total hip or knee replacement surgery. A specific antidote to rivaroxaban is not available. Monitoring of rivaroxaban is not generally indicated and although the APTT is prolonged it is not recommended as a method for laboratory monitoring.

Laboratory monitoring

Therapeutic anticoagulation with lepirudin or bivalirudin is commonly monitored by the activated partial thromboplastin time (APTT) but there is considerable inter-individual variability in the degree of prolongation of the APTT at identical plasma levels of these drugs. The ‘ecarin’ clotting time (ECT) has been suggested as a more accurate test for monitoring individuals receiving direct thrombin inhibitors ¹² including dabigatran if needed. Ecarin is isolated from the venom of the saw-scaled viper Echis carinatus and in the assay a known amount of ecarin is added to the plasma. Ecarin activates prothrombin to meizothrombin – this activity is inhibited by lepirudin but is unaffected by heparin. The meizothrombin induces clotting via fibrinogen cleavage to fibrin. This prolongation in the clotting time increases in a linear fashion with increasing concentrations of lepirudin but also with bivalirudin, dabigatran and argatroban, another direct thrombin inhibitor. The ecarin chromogenic assay employs a similar approach but the concentration of meizothrombin is measured using a chromogenic substrate.¹³

Warfarin and vitamin K antagonists

Warfarin is a 4-hydroxycoumarin derivative that exerts its action by blocking the regeneration of vitamin K from its epoxide. The major complication of all vitamin K antagonists is bleeding and this risk increases as the intensity of treatment, i.e. the INR, increases.^14,¹⁵ Independent risk factors for bleeding during long-term warfarin therapy include age greater than 65 years, a history of past gastrointestinal bleeding, stroke, atrial fibrillation and one or more of three co-morbid conditions: myocardial infarction, renal insufficiency and severe anemia.¹⁶ For any individual the risk of bleeding is related to the duration of anticoagulant therapy although the risk may be higher in the early phase of treatment. Most studies in unselected groups of patients suggest that the risk of major bleeding is ~3% per annum and that CNS hemorrhage occurs at a rate of 0.1% per annum.¹⁷

The anticoagulant action of warfarin is potentiated by many drugs and these include:

• Drugs that displace warfarin from its plasma protein binding sites, e.g. statins

• Drugs that inhibit the metabolic clearance of warfarin, e.g. cimetidine, omeprazole, amiodarone, allopurinol

• Drugs that interfere with vitamin K metabolism, e.g. cephalosporins, high dose salicylates

• Drugs that independently increase the anticoagulant action, e.g. clofibrate, anabolic steroids, erythromycin.

Minor bleeding episodes in patients receiving oral anticoagulants may be treated with local measures and withdrawal of the drug. In cases of severe or life-threatening hemorrhage, rapid reversal of anticoagulation is required and this is most effectively achieved by the use of a combination of vitamin K and clotting factor concentrates (containing factors II, VII, IX and X) and less effectively by vitamin K and fresh frozen plasma.¹⁸

Phenindione. Phenindione is a vitamin K antagonist but differs from warfarin in its chemical structure. The degree of anticoagulation induced by phenindione is monitored by the INR although the dosing is different from that of warfarin with a loading dose of 200 mg on day 1, 100 mg on day 2 with subsequent dosing based upon the INR. The normal maintenance dose lies between 50 and 150 mg/day. Phenindione is rarely used because of the higher risk of side-effects including skin rashes and abnormal liver function tests.

Laboratory monitoring

Warfarin and other vitamin K antagonists are monitored by measuring the International Normalised Ratio (INR). The INR is the ratio of the prothrombin time of a patient divided by the mean normal prothrombin time of a normal plasma pool corrected for the sensitivity of the tissue factor used in the test:

where the ISI (International Sensitivity Index) is a value derived by calibrating the tissue factor used in the assay against an international WHO standard, the ISI of which is 1.0. For an individual who is not on warfarin the INR is 1.0. The target INR for any patient varies depending upon the indication for treatment but is usually 2.5, 3.0 or 3.5. The risk of bleeding on any VKA increases as the INR increases and patients with a target INR of 3.5 have a significantly greater risk of hemorrhage than those with a target INR of 2.5.¹⁷

Thrombolytic agents

Thrombolytic drugs act by stimulating endogenous fibrinolysis. T-PA or U-PA convert plasminogen to plasmin, a potent proteolytic enzyme which breaks down both cross-linked and non-cross-linked fibrin. The currently available thrombolytic agents include:

• Recombinant human tissue plasminogen activator (rt-PA)

• Urokinase (U-PA)

• Reteplase – a recombinant non-glycosylated form of human t-PA that contains only 357 of the 527 amino acids of the original protein

• Tenecteplase – a recombinant fibrin-specific form of t-PA but engineered at three sites to confer a higher fibrin specificity than native t-PA and a greater resistance to inactivation by endogenous plasminogen activator type I (PAI-1) – the major inhibitor of t-PA

• Staphylokinase

• Streptokinase and its anisolyated derivative APSAC.

T-PA, urokinase, reteplase and tenecteplase produce their pharmacological actions by converting plasminogen to plasmin at the site of fibrin deposition. In contrast staphylokinase and streptokinase bind to free plasminogen in the plasma leading to systemic hyperfibrinolysis. Bleeding occurs in 3–40% of patients receiving thrombolytic therapy and this risk is greatly increased in patients who are also receiving anti-platelet drugs or other anticoagulants.¹⁹ Thrombolytic therapy predisposes to bleeding by depleting the plasma concentration of procoagulant proteins and by the generation of anticoagulant fibrin(ogen) degradation products. Thrombolytic therapy cannot distinguish between a pathological thrombus occluding a critical vessel, e.g. coronary artery, and a physiological thrombus that is preventing bleeding from a critical site, e.g. in the cerebral circulation. Platelet function in patients receiving thrombolytic therapy is also impaired because of inhibition of platelet aggregation by high levels of FDPs and also by impaired platelet adhesion by plasmin-induced proteolysis of glycoprotein Ib (GpIb) and von Willebrand factor (vWF).²⁰

For patients receiving thrombolytic therapy and who develop minor bleeding episodes the thrombolytic agent, together with any anticoagulant or anti-platelet agent, must be discontinued. For life-threatening bleeding episodes, a fibrinolytic inhibitor should be given e.g. tranexamic acid. Fresh frozen plasma and/or cryoprecipitate or a fibrinogen concentrate should be given to restore depleted clotting factors.¹⁹

Laboratory monitoring

Laboratory monitoring of thrombolytic therapy is often unnecessary when its administration is short-term. However, during a more prolonged infusion (>24 hours) sequential monitoring may be of value. Fibrinolytic therapy alters most laboratory tests of coagulation but few tests predict either the efficacy of thrombolysis or the risks of bleeding. The APTT is prolonged in patients receiving thrombolytic therapy because of depletion of fibrinogen, factors V and VIII and the generation of high levels of fibrin(ogen) degradation products. An APTT ratio of 1.5 indicates significant systemic fibrinolysis.¹⁹ Plasma fibrinogen concentration falls during thrombolytic therapy reflecting the presence of free plasmin within the circulation. The fibrinolytic activity of the plasma can be measured by means of the euglobulin clot lysis time (ELT) which is shortened in patients receiving thrombolytic therapy but this is rarely, if ever, used. The thromboelastogram (TEG) may be of value and is considerably easier to perform than the ELT.

Anti-platelet drugs

A wide variety of drugs are in common use that have potent anti-platelet actions and such drugs are often used in combination, for example aspirin and clopidogrel. The risk of hemorrhage is significantly increased when anti-platelet drugs are used in combination with other anticoagulants, for example warfarin and aspirin.

1. Aspirin and non-steroidal anti-inflammatory drugs inhibit platelet function by preferentially inhibiting platelet cyclooxygenase activity while maintaining the activity of the enzyme within the endothelial cells.²¹ In this way, there is a reduction in the production of platelet thromboxane A₂ (TxA₂), a potent inducer of platelet aggregation, but preservation in prostacyclin (PGI₂) synthesis by the endothelial cell. Omega-3 fatty acids can substitute for arachidonic acid in prostaglandin synthesis resulting in the synthesis of thromboxane A₃ (TxA₃), which has little effect upon platelet aggregation. However, within the endothelial cell, synthesis of a novel prostaglandin occurs (PGI₃) which has potent anti-platelet activity.

2. Several drugs such as dipyridamole result in a decrease in platelet intracellular calcium concentration by inhibiting cAMP degradation. Elevated levels of cAMP favor movement of Ca²⁺ into the dense bodies where it is inert and so effectively decreases intracellular Ca²⁺ levels.²¹

3. Drugs such as ticlopidine and clopidogrel selectively inhibit platelet aggregation by binding to and blocking the P2Y₁₂ platelet receptor thereby inhibiting ADP-induced aggregation.^22,²³

4. A number of drugs selectively bind to and block the platelet GpIIb/IIIa complex resulting in a loss of activity. These include abciximab, a monoclonal antibody, eptifibatide (a cyclic heptapeptide derived from a protein found in the venom of the southeastern pygmy rattlesnake, Sistrurus miliarius barbouri) and tirofiban a synthetic, non-peptide inhibitor. These drugs are potent inhibitors of platelet function that can result in severe bleeding problems.²⁴ In addition, in a small number of patients their use is associated with a profound thrombocytopenia which may exacerbate the bleeding tendency.^25,²⁶

5. Platelet function abnormalities can be induced by certain antibiotics, particularly the β-lactam antibiotics.^25,²⁶ Penicillin, particularly penicillin G, ticarcillin and carbenicillin, have been reported to inhibit platelet function and to cause a clinically significant bleeding tendency particularly when administered in high dose.^25,²⁶ Some cephalosporins also appear to interfere with vitamin K metabolism resulting in an additional and additive increased risk of bleeding.²⁷

6. Finally, a variety of drugs appear to have nonspecific effects upon platelet function. β-adrenergic agents have a weak but inconsistent effect on platelet function in vitro but are rarely if at all associated with a clinical bleeding tendency. Sodium valproate can result in thrombocytopenia but can also result in qualitative platelet abnormalities.²⁸

Laboratory monitoring

There is increasing interest in monitoring platelet function in patients receiving anti-platelet drugs to identify those individuals who demonstrate drug ‘resistance’ and may have a reduced benefit.²⁹^–³⁴ Platelet aggregation studies or the use of the platelet function analyzer 100 (PFA-100™) may be of value in identifying these patients.³⁵^–³⁷

Hemostatic defects associated with vitamin K deficiency

Vitamin K and vitamin K deficiency

Vitamin K₁ is a fat-soluble vitamin obtained primarily from green leafy vegetables. It is absorbed in the upper part of the small intestine and its absorption is dependent upon the presence of pancreatic lipases and bile. Most of the vitamin K absorbed from the gut is stored in the liver although the stores of vitamin K are only a few days. The normal daily requirement of vitamin K is 0.5–1.0 µg/kg. Vitamin K₂ is synthesized by the gut flora but cannot compensate for a total deficiency of vitamin K₁. Vitamin K₃ is a synthetic form of vitamin K.

Vitamin K oxidation to its epoxide form is essential for the post-translational gamma-carboxylation of the glutamic acid residues present in the N-terminal region of factors II, VII, IX, X, protein C and S. Efficient gamma-carboxylation allows the modified glutamic acid residues to bind calcium and subsequently to the phospholipid receptors on cell membranes allowing coagulation to proceed. In the absence of efficient gamma-carboxylation, partially carboxylated forms of the clotting factors are released into the circulation, co-called ‘PIVKAS’. During carboxylation, vitamin K is oxidized to vitamin K epoxide and recycled to its active form by reductases. Oral anticoagulants such as warfarin inhibit vitamin K epoxide reduction and prevent recycling of vitamin K to its active form, thereby limiting the activity of the carboxylase.

Hemostatic defects in liver disease

The liver is responsible for the synthesis of all the coagulation factors apart from von Willebrand factor (vWF). The liver also synthesizes either completely or in part many of the proteins involved in the regulation of coagulation – antithrombin, protein C, protein S, heparin co-factor II and those involved in fibrinolysis – plasminogen and α₂-antiplasmin. The liver is also responsible for the clearance of activated clotting factors that are generated by the clotting cascade and during fibrinolysis. Liver disease is therefore associated with a major disruption of the clotting system resulting in an increased risk of hemorrhage. Factors V and VII are sensitive markers of hepatic function and may be used as an index of severity.⁴²

Defective production of clotting factors arises because of a failure in hepatic synthetic function including gamma-carboxylation of the vitamin K-dependent clotting factors although there may also be reduced absorption of the fat-soluble vitamins including vitamin K as a result of cholestasis. Von Willebrand factor is often raised in patients with liver failure reflecting its extra-hepatic site of synthesis (endothelial cells and megakaryocytes) and its acute phase nature.

Thrombocytopenia is a common finding in liver disease and is often due to sequestration of platelets within the spleen – hypersplenism. Thrombocytopenia may also be seen in association with alcohol abuse, folate deficiency, DIC and in some cases of viral hepatitis where the causative virus may have a direct effect upon megakaryopoiesis or accelerate peripheral destruction.⁴³ A qualitative platelet abnormality is often seen in patients with liver failure which further exacerbates the bleeding tendency.

Fibrinogen is relatively well maintained in liver disease until the terminal stages when the levels may drop dramatically. In addition, as a result of an increased sialic acid content of fibrinogen, patients with liver failure may develop an acquired dysfibrinogenemia resulting in slow fibrin polymerization and a relatively unstable fibrin clot.^44,⁴⁵ Abnormal fibrinogens and non-carboxylated prothrombin are also synthesized by patients with primary hepatocellular carcinoma and have been used as markers of these disorders.⁴⁴

Many patients with liver disease have evidence of systemic fibrinolysis secondary to reduced synthesis of α₂-antiplasmin, reduced clearance of t-PA and low grade DIC.⁴⁶^–⁵⁰ Primary hyperfibrinolysis may result in severe bleeding problems following surgery in patients with liver disease where tissue damage results in the release of large amounts of plasminogen activators which swamp the impaired protective mechanisms of the liver resulting in systemic fibrinolysis.

Chronic low-grade DIC is a common feature of liver disease. This occurs secondary to release of tissue thromboplastin from the damaged hepatocytes, reduced synthesis of the inhibitors of coagulation – antithrombin, protein C and protein S – and reduced clearance of activated clotting factors. Ascitic fluid appears to contain a potent thromboplastin-like material and may result in severe DIC following creation of a peritovenous shunt in which large amounts of ascitic fluid are infused directly into the circulation.⁵¹