HIT/HITT and Alternative Anticoagulation: Current Concepts

British Journal of Anaesthesia 90 (5): 676±85 (2003) DOI: 10.1093/bja/aeg063

REVIEW ARTICLE HIT/HITT and alternative anticoagulation: current concepts

E. Pravinkumar* and N. R. Webster

Academic Unit of Anaesthesia and Intensive Care, Institute of Medical Sciences, Foresterhill, Aberdeen AB25 2ZD, UK *Corresponding author. Email: [email protected] Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021

Br J Anaesth 2003; 90: 676±85 Keywords: blood, coagulation; blood, anticoagulants, heparin Accepted for publication: July 25, 2002

Heparin is a widely used anticoagulant for the treatment and factors, such as factors IX, X, XI and XII and kallikrein, and prevention of thromboembolic disorders in medical and on the conversion of prothrombin to thrombin.50 62 Once surgical patients. Its importance as an anticoagulant has active thrombosis has developed, larger amounts of heparin been well established by its effectiveness, rapid onset of can inhibit further coagulation by inactivating thrombin and action, ease of laboratory monitoring and cost.18 Heparin is preventing the conversion of ®brinogen to ®brin. Heparin a member of the heterogeneous family of glycosaminogly- also prevents the formation of a stable ®brin clot by cans, which range in size from 3000 to 30 000 Da. inhibiting the activation of the ®brin-stabilizing factor. The non-branching, negatively charged chain structure of heparin consists of repeating disaccharide units. Heparin is an anticoagulant released by mast cells and basophils in the Thrombocytopenia and heparin process of clot formation, as well as a drug that is Thrombocytopenia can result from conditions that lead to administered to the same effect. Standard unfractioned increased platelet destruction or decreased platelet produc- heparin is usually derived from porcine intestinal mucosa or tion (Table 2).84057Heparin can cause thrombocytopenia bovine lung.53 via immune and non-immune mediated mechanisms.41 Heparin plays many roles in human physiology, such as: There are two types of heparin-induced thrombocytopenia (i) binding to antithrombin III and increasing the ef®cacy of that can result from heparin administration: type I, non- antithrombin III as an inhibitor of the activation of thrombin immune-mediated; and type II, immune-mediated.23 67 68 In and certain clotting factors; (ii) inhibiting platelet forma- the interest of standardization, the term `non-immune tion; (iii) increasing the permeability of vessel walls; (iv) heparin-associated thrombocytopenia' is recommended for inhibiting the proliferation of vascular smooth muscle cells; type I, a benign condition in which no heparin-dependent and (v) playing a role in the regulation of angiogenesis. antibodies are present. The term `heparin-induced throm- In clinical use heparin has numerous activities, but the bocytopenia' (HIT) is recommended for thrombocytopenia most important is its anticoagulant property (Table 1). It is in which pathogenic heparin-dependent antibodies are also used as a thrombolytic, a fat-clearing anti-athero- detectable. This term is the most widely accepted designa- sclerotic and as a potentially effective anti-in¯ammatory tion for HIT type II.67 agent. Heparin inhibits reactions that lead to clotting of Non-immune-mediated thrombocytopenia, also known as blood and formation of ®brin clots both in vitro and in vivo. heparin-associated thrombocytopenia (HAT) and heparin- It acts at multiple sites in the normal coagulation system. It induced thrombocytopenia type I (HIT type I), denotes the binds to antithrombin III (heparin cofactor), causing a absence of heparin-dependent antibodies. It is probably conformational change in the structure of antithrombin III. caused by direct non-immune platelet activation by heparin. This conformational change converts antithrombin III from Type I HIT is usually associated with larger doses of heparin a slow- to a fast-acting inhibitor of thrombin activation. The in contrast to HIT type II, which can occur with variable complex has a further inhibitory effect on other clotting doses. HIT type I occurs earlier in the treatment course,

Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2003 HIT/HITT and alternative anticoagulation within a period of 4 days, in 30% of patients receiving occur with warfarin treatment of HIT-associated DVT, and heparin. The platelet abnormality is usually mild and is characterized by a high international normalized ratio reversible even with the continuation of the heparin (INR), resulting from severe reduction in protein C caused (Table 3). The condition is self-limiting and there are no by warfarin. Adrenal vein thrombosis and cerebral sinus major complications associated with it. Heparin should be thrombosis are other unusual venous thromboses that continued despite the low platelet count.23 67 The clinical complicate HIT. Arterial thromboses associated with importance of the asymptomatic, self-limiting disease HIT HITT may result in ischaemic limb damage that often type I lies in the necessity of differentiating it from the more requires amputation. Myocardial infarction, ischaemic serious HIT type II. stroke and end-organ thromboses, such as mesenteric, Immune-mediated HIT type II exists in three separate renal, brachial, splenic and hepatic arterial thromboses, diagnosable forms: can also occur (Table 4).10 23 47 74 HIT develops in 3% of (i) latent: antibodies without thrombocytopenia; patients treated with unfractioned heparin for more than (ii) HIT: antibodies with thrombocytopenia; and 4 days.72 Without treatment, mortality in HIT patients

(iii) HITT: antibodies with thrombocytopenia and throm- with new thromboembolic complications is about 20±30%, Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 bosis. with equal morbidity caused by arterial and venous HIT/HITT (heparin-induced thrombocytopenia and he- thrombosis.11 20 75 The risk of venous thromboembolism parin-induced thrombocytopenia with thrombosis syn- from HIT is greater in high-risk surgical patients than in drome) is an immune-mediated adverse reaction to medical patients. Also, 40±50% of HIT patients will have a heparin that is often underdiagnosed and can result in thrombotic event when heparin is discontinued.25 65 74 76 venous and arterial thrombosis.6 56 67 69±73 79 The alternative name of HITT, white clot syndrome, refers to the gross pathology of the clots. The platelet±platelet adhesion Mechanism of heparin±PF4±IgG complex without erythrocyte involvement gives a classic appearance Most patients with HIT produce immunoglobulin G (IgG) of a white clot. Patients with HITT may suffer from venous antibodies, commonly IgG141755against macromolecular thrombosis, most often deep venous thrombosis (DVT), complexes of platelet factor 4 (PF4) and heparin which can be extensive and complicated by pulmonary (H±PF4).2112637416364The high binding af®nity between embolism. Venous limb gangrene has been reported to heparin and PF4 is probably attributable to a high concentration of C-terminus lysine residues that interact strongly Table 1 Clinical uses of heparin

Vascular Acute-phase myocardial infarction Table 2 Causes of thrombocytopenia Venous and arterial thromboembolism Pulmonary embolism Increased platelet destruction Peripheral arterial disease Non-immune Septicaemia/in¯ammation Atherosclerosis Disseminated intravascular coagulation (DIC) Atrial ®brillation with embolization Thrombotic thrombocytopenic purpura (TTP) Acute ischaemic stroke Immune Autoimmune: idiopathic or secondary immune Deep vein phlebitis thrombocytopenia Topical Flap survival/pruritus Alloimmune: post-transfusion purpura Leg ulcers Drug-induced: prothrombic (heparin), Surgery Open-heart surgery prohaemorrhagic (quinine, quinidine, gold, Vitreoretinal surgery sulpha antibiotics, rifampicin, vancomycin, Prophylaxis Pulmonary embolism NSAIDs) Deep vein thrombosis Decreased platelet production Alcohol, cytotoxic drugs Support systems Extracorporeal circulation Leukaemia, aplastic anaemia Renal replacement therapy Myelodysplasia Critical care Heparin-coated central venous and pulmonary Metastatic bone marrow involvement arterial catheters Infections Heparin ¯ushes in arterial and central venous Other causes Hypersplenism lines Haemodilution (infusion of blood products, Other (as anticoagulant) Laboratory samples colloids, crystalloids) Blood transfusion NSAIDs=non-steroidal anti-in¯ammatory drugs.

Table 3 Heparin-induced thrombocytopenia68

Non-immune-mediated HAT (type I HIT) Immune-mediated HIT (type II HIT)

Onset Within 4 days 5±14 days or sooner Platelet count Typically 100 000±150 000 ml±1 Typically 20 000±150 000 ml±1 Incidence 5±30% 1±3% Complications None Thromboembolic lesion Recovery 1±3 days 5±7 days Cause Benign: tiny platelet aggregates IgG-mediated platelet activation

677 Pravinkumar and Webster with the highly negatively charged heparin molecule. between days 5 and 15 after the introduction of heparin,72 Neither heparin nor PF4 is normally antigenic, but the but can develop earlier in patients exposed to heparin during formation of an H±PF4 complex induces a conformational the previous 3 months.41 Case reports have shown change in both molecules. Upon binding to the heparin thrombocytopenia developing within 10.5 h in patients molecule, PF4 exposes several antigenic epitopes, which exposed to heparin in the last 100 days. Thus, a patient with trigger the immune reaction and the production of a rapid decrease in platelets soon after commencing heparin immunoglobulin G.45 60 82 These antibodies activate therapy, who has not been exposed to heparin previously, platelets through their FcRIIA receptors, causing platelet probably has another aetiology for the thrombocytopenia. destruction and release of prothrombotic platelet-derived During recovery from thrombocytopenia, heparin- microparticles.12 Microparticles in turn promote thrombin dependent antibodies in the serum decrease to undetectable generation and contribute to a hypercoagulable state77 levels by 50±85 days.78 Those with undetectable (Fig. 1). PF4, a small peptide stored within the alpha antibody but a previous proven diagnosis of HIT type II granules of platelets, binds to heparin and is released into may take as long as 5 days to produce IgG antibodies and 3 the blood during treatment with it. In vitro, thrombocytopenia. Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 IgG±PF4±heparin complexes can activate platelets.11 38 49 64 Increased propensity to thrombosis in HIT is therefore probably mediated by thrombin generated as a result of Laboratory diagnosis in vivo platelet activation,13 77 from interaction between There are two main classes of assays for laboratory heparin±PF4±IgG immune complexes with Fc receptors on diagnosis of HIT. Activation (functional) assays include platelets.38 A minimum of 12±14 saccharides is required to form the antigenic complex with PF4. Heparin molecules the platelet aggregation assay and the serotonin release assay. The platelet aggregation assay involves the use of with a molecular weight >4000 Da have the potential to washed platelets rather than platelet-rich plasma derived cause HIT. Hence HIT is less common with low molecular from normal donors. This increases the reliability of weight heparin (LMWH) than with unfractioned activation assays,27 28 which are performed in the laboratory heparin.21 72 No speci®c laboratory or clinical characteris- tics can predict which patients will have isolated thrombocytopenia or have thrombocytopenia with thrombotic complications.

Diagnosis of HIT Manifestations of HIT include an otherwise unexplained decrease in platelet count and skin lesions at heparin injection sites, accompanied by HIT antibody formation. The platelet count during HIT can be variable, but only rarely decreases to <20 0003109 litre±1. The median platelet count nadir in several studies is ~50 0003109 litre ±1. HIT should also be suspected if the platelet count decreases by 30±50% after 5 days of heparin treatment (Table 3).68 78 The occurrence of thromboembolic complications during heparin therapy is another strong marker of HIT.11 23 Regardless of the degree of thrombocytopenia, the pre- dominant clinical feature is thrombosis and not bleeding. The decrease in platelet count almost always occurs

Table 4 Venous and arterial thromboses associated with heparin-induced thrombocytopenia and thrombosis syndrome (HITT)

Venous thrombosis Arterial thrombosis

Deep venous thrombosis Lower limb thrombosis Pulmonary embolism Myocardial infarction Venous limb gangrene associated Acute coronary syndromes with warfarin Adrenal vein thrombosis/infarction Stroke Cerebral sinus thrombosis End-organ thromboses (mesenteric, Fig 1 Mechanism of action of heparin±PF4±IgG complex and the vicious renal, brachial, splenic, hepatic) circle that leads to thrombus formation and platelet destruction. IgG=immunoglobulin G; PF4=platelet factor 4; GP=glycoprotein.

678 HIT/HITT and alternative anticoagulation with a speci®city >90%. The disadvantages are low Management of HIT sensitivity (35%) and the fact that assay reactivity varies 11 20 After the clinical diagnosis or suspicion of HIT, heparin between donor platelets. The serotonin release assay should be discontinued promptly. The platelet count should measures the release of serotonin from platelet aggregates. be monitored carefully and warfarin should not be used It relies on the aggregation of the platelets from the patient alone in patients with acute HIT.67 Management of the in the presence of heparin. This assay has high sensitivity patient with HIT is aimed at prevention of the thromboem- and speci®city and has been validated by blinded assess- bolic complications seen in 50±70% of patients. The ment of clinical trials.58 The disadvantages are that the assay mortality associated with isolated thrombocytopenia with- is technically demanding and involves the use of radioactive out thrombosis is 21%. Therefore, the principle of managing materials.27 28 Of the various activation assays available, HIT is the discontinuation of all forms of heparin exposure those that use washed platelets and platelet serotonin release and the institution of an alternative anticoagulant.43 74 The or heparin-induced platelet activation (HIPA) are the most diagnosis should be con®rmed as soon as practicable, but accurate.19 31 58 70 treatment should not be delayed for this reason. In the

The other class of assay is the antigen assay. The Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 absence of thrombosis, cessation of heparin has been the heparin±PF4 enzyme-linked immunosorbent assay (ELISA) traditional mode of therapy,65 but several studies have relies on the speci®city of HIT IgG antibodies for the shown that stopping heparin may be inadequate because of heparin±PF4 complex. This assay is 10 times more sensitive the high risk of overt thrombosis (50%) in the week after its than the serotonin release assay for detecting heparin- interruption.924296574 induced antibodies. However, the heparin±PF4 ELISA is expensive and time-consuming. The assay also responds to clinically insigni®cant antibodies more often than activation assays, and hence has a lower speci®city.20 64 Warfarin in HIT Most of the available laboratory tests for the diagnosis of Warfarin should not be used alone to treat acute HIT HIT are expensive, time-consuming, frequently contradict- complicated by DVT because of the paradoxical risk of ory to the clinical presentation, and vary in sensitivity and developing venous limb gangrene, which causes more speci®city. Testing a patient with a history of HIT before than half the limb loss attributable to HIT at one centre.75 expected heparin re-exposure should therefore be done with Rapid loading with warfarin can cause depletion of more than one sensitive test. Hence, the diagnosis of HIT the anticoagulant vitamin K-dependent protein C faster remains clinical. It should be suspected in patients who than it depletes procoagulant vitamin K-dependent develop thrombocytopenia with or without associated proteins.32 57 75 77. Higher warfarin doses cause a rapid arterial or venous thrombosis while on heparin.15 increase in the INR, but this is entirely a result of earlier reduction of factor VII rather than a reduction in factor II, a coagulable protein with a longer half-life. Initiating Prevention of HIT warfarin therapy with a relatively high loading dose for Currently there are no tests available to predict which rapid anticoagulation, which is a common practice, may patients are at high risk of developing HIT. Therefore, prove dangerous, particularly in HIT.46 75 Although other whenever heparin is clinically indicated, it is necessary to thrombin inhibitors have been introduced for continuing assess the risk± bene®t pro®le in all patients. The following anticoagulation in the place of heparin, warfarin still points should be considered in the prevention of HIT: remains an option for patients with HIT who require (i) history of previous sensitization to heparinÐ anticoagulation. Studies have shown that aiming for modest requires initiation of alternative anticoagulant levels of the INR with warfarin and a smaller loading dose therapy; prevents the development of a potential hypercoagulable (ii) limit heparin duration to less than 5 days whenever state caused by a precipitous decrease in the concentration possible; of protein C.32 66 (iii) early use of warfarin in patients requiring long-term It is therefore important to start treatment in all patients anticoagulation, except when HIT is diagnosed; with HIT who remain at risk of thrombosis, including (iv) initiating warfarin at the start of heparin treatment postoperative patients and those with sepsis. Thrombin in patients who require long-term anticoagulation; inhibitors (lepirudin or argatroban) should be used in these (v) use of LMWHÐa prospective study shows LMWH patients until the platelet count has recovered. Treatment is less likely to cause HIT and HIT±IgG anti- with thrombin inhibitors should also be considered for bodies;72 patients with acute HIT without thrombosis (isolated HIT), (vi) avoid unnecessary heparin ¯ushes; because there is a high risk of subsequent clinically evident (vii) use of porcine heparin (frequency of HIT, 1.3±8%) thrombosis. Warfarin given to patients adequately anti- rather than bovine heparin (frequency of HIT, coagulated with lepirudin appears safe in acute HIT, but it is 1.9±30.8%);11 68 prudent to delay starting warfarin until the platelet count has (viii) regular monitoring of platelet count. risen above 1003109 litre±1.33

679 Pravinkumar and Webster

Ancrod in HIT Ancrod, a ®brin inhibitor derived from the Malayan pit viper venom, is another treatment option for HIT. Ancrod has not been widely used and there is only one small study on 11 patients.16 Being a foreign protein, Ancrod induces resist- ance with prolonged use. There has also been increased thrombin generation and warfarin-induced limb gangrene in patients treated with it.

Prostacyclin analogues in HIT Prostacyclin is a natural vasodilator and inhibitor of platelet aggregation. Iloprost has been shown to be effective in

suppressing HIT-induced platelet aggregation and in pre- Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 venting thrombotic complications. Prostacyclin analogues have a short half-life of 15±30 min and have been preferred in cardiac and vascular surgery for anticoagulant coverage in patients with HIT. They are also used in intensive care Fig 2 Actions of thrombin. NO=nitric oxide; PgI2=prostaglandin I2; t-PA=tissue plasminogen activator. Adapted with permission from Fuster units for anticoagulation during renal replacement therapy. and Verstraete.22 The main side-effects are hypotension and bradycardia.11 23

Thrombin inhibitors in HIT Low molecular weight heparin and HIT Thrombin plays a central role in HIT-related thrombo- Low molecular weight heparin does not bind strongly to embolic complications (Fig. 2). Thrombin generation is plasma proteins and endothelial cells; its bioavailability is enhanced in HIT by the concomitant activation of plate- 13 much greater than with unfractionated heparin and there is lets, the generation of platelet microparticles and the 80 also a more predictable dose response. However, low alteration of endothelial cells. Thrombin inhibitors such as molecular weight heparins have a high in vitro cross- lepirudin and argatroban have found a greater role in the 39 reactivity rate with heparin-dependent antibodies, reaching management of HIT/HITT. 100% in some very sensitive assays. Low molecular weight Thrombin inhibitors currently available for the manage- heparins can cause HIT,72 and when used in its treatment ment of HIT and HITT are: â there is a signi®cant risk of recurrent or progressive (i) lepirudin (Re¯udan ); â â thrombocytopenia with or without thrombosis. Because of (ii) argatroban (Argatroban , Novastan ); â these drawbacks, and as there are other effective drugs (iii) hirulog and bivalirudin (Angiomax ) and â available, low molecular weight heparins are not recom- (iv) desirudin (ReVase ). mended for the management of patients with HIT.67 Lepirudin is currently licensed for the treatment of HIT/ HITT in the UK, the USA and Europe. Argatroban has recently been approved for use in the prophylaxis and treatment of thrombosis in HIT in the USA and Canada. Danaparoid in HIT Bivalirudin and desirudin are currently not licensed for use Danaparoid sodium (Orgaranâ) does not contain heparin in HIT/HITT, although their use in the treatment of this and is a low molecular weight heparinoid. It is a mix of condition has been reported. dermatan sulphate, glycosaminoglycans and chondroitin Hirudin is a direct inhibitor of thrombin and acts sulphate. The use of danaparoid is associated with a independently of cofactors such as antithrombin.52 Unlike favourable outcome in 90% of patients with HIT.11 The heparin, hirudin is not inactivated by PF4 and may be more main activity of danaparoid is against factor Xa, with the effective in the presence of platelet-rich thrombi.59 Hirudin anti-Xa:anti-IIa ratio of 22:1 resulting in inhibition of ®brin can also inhibit thrombin bound to ®brin or ®brin degrad- formation. The potential disadvantage is the prolonged half- ation products.80 81 life of anti-factor Xa activity (around 25 h). This is Lepirudin (Re¯udanâ) is a recombinant hirudin derived disadvantageous for patients who are at risk of bleeding from yeast cells and is a highly speci®c direct and and who might need a surgical procedure. There is a irreversible inhibitor of thrombin. One molecule of 10±20% cross-reactivity rate with HIT antibodies in vitro, lepirudin binds with one molecule of thrombin. The half- although it is less common in vivo.30 78 Danaparoid does not life for distribution and elimination is 10 min and 1.3 h interfere with INR measurements, and therefore allows respectively. Lepirudin is almost exclusively excreted in the concurrent and easy monitoring of warfarin anticoagulation. kidneys and hence systemic clearance of lepirudin is

680 HIT/HITT and alternative anticoagulation

Table 5 Comparison between lepirudin and argatroban7

Lepirudin Argatroban

Mechanism Direct thrombin inhibitor Direct thrombin inhibitor Irreversible inhibition Reversible inhibition Dosing 0.4 mg kg±1 bolus i.v. followed by Start at 2 mgkg±1. Dose can be increased infusion of 0.15 mg kg±1 h±1. Dose up to a maximum of 10 mgkg±1 min±1. adjustment required in renal insuf®ciency Dose adjustment required in hepatic insuf®ciency Monitoring aPTT, ecarin clotting time aPTT Half-life (min) 40±120 24±50 Reversal agent None None

Table 6 Dosing adjustment for lepirudin7 Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 Creatinine clearance Serum creatinine concentration Adjusted infusion rate

(ml min±1) (mmol l±1) (% of standard initial (mg kg±1 h±1) infusion rate)

45±60 140±176 50% 0.075 30±44 176.1±264 30% 0.045 15±29 264.1±528 15% 0.0225 <15 >528 Avoid or stop infusion Haemodialysis Give 0.1 mg kg±1 every other day only if the aPTT falls patients below the lower therapeutic limit of 1.5 dependent on the glomerular ®ltration rate. It should prothrombin-de®cient states without interference from therefore be used cautiously in patients with chronic renal heparin or aprotinin or from treatment with oral anti- failure and should be avoided in patients with acute renal coagulants.54 Published clinical data regarding the use of failure (Table 5). There is an increased risk of bleeding with this monitoring tool come mainly from case reports. Ecarin other anticoagulants, thrombolytics and antiplatelet drugs. clotting time values showed good correlation with plasma The usual adult dose is 0.4 mg kg±1 as a bolus over 15±20 s, concentrations of lepirudin. In these reports, an ECT of followed by 0.15 mg kg±1 h±1 initially; adjustments are then 350±400 s corresponds to lepirudin concentrations of 3.5± made on the basis of the activated partial thromboplastin 4 mgml±1.42 time (aPTT) (Table 5). However, the dose should not exceed The Heparin-Associated Thrombocytopenia Trials an infusion rate of 0.21 mg kg±1 h±1 without checking for (HAT-1 and HAT-2)24 25 are prospective clinical trials of coagulation abnormalities. In haemodialysis patients and lepirudin. The clinical outcomes of patients were compared patients with acute renal failure with creatinine clearance with those of a historical control group. A meta-analysis of <15 ml min±1 (normally 120 ml min±1) or serum creatinine these trial results29 represents the largest population of >528 mmol litre±1, the drug should be avoided. The infusion patients with HIT and thromboembolic complications rate should be adjusted in patients with renal impairment treated with lepirudin. Overall, 198 patients (82 in HAT-1, (Table 6). 116 in HAT-2) were treated with lepirudin and 182 Ecarin clotting time (ECT) is a suitable assay for bedside historical control patients were given other treatments. All real-time monitoring of lepirudin anticoagulation. Ecarin is except ®ve (one in HAT-1, four in HAT-2 group) a metalloproteinase from snake venom (Echis carinatum) prospective patients and all historical control patients that activates prothrombin to meizothrombin. Hirudin forms were diagnosed with HIT using the heparin-induced platelet a 1:1 complex with meizothrombin, thus preventing the activation assay or equivalent assays for testing. In total, conversion of ®brinogen to ®brin and the formation of 113 (54 in HAT-1, 59 in HAT-2) prospective patients thrombin. Therefore, the clotting time is prolonged with (lepirudin) and 91 historical control patients presented with increasing amounts of hirudin. The precision of the ECT thromboembolic complications at baseline (day of positive assay is directly related to the prothrombin and ®brinogen test result) and quali®ed for direct comparison of clinical concentration, but does not seem to be affected by the endpoints. presence of heparin or oral anticoagulants.54 The speci®city In the meta-analysis, the pooled lepirudin patients of the of the ECT-determined lepirudin concentration is not HAT-1 and HAT-2 studies who presented with thromboem- in¯uenced by treatment with oral anticoagulants. ECT is a bolic complications at baseline were compared with the rapid, sensitive clotting assay suitable for precise measure- respective historical control patients. Seven and 35 days ment of plasma concentrations of lepirudin in non- after the start of treatment, the cumulative risks of death

681 Pravinkumar and Webster

Table 7 Results of argatroban study: categorical ef®cacy analysis.44 Values are n (%). aAll-cause death, all-cause amputation or new thrombosis within 37- day study period; bseverity ranking: all-cause death > all-cause amputation > new thrombosis; patients with multiple outcomes counted once; cpatients counted only once if multiple events occurred. (Adapted with permission from Lewis et al.)44

HIT arm HITTS arm

Control Argatroban Control Argatroban Variable (n=147) (n=160) P-value (n=46) (n=144) P-value

Primary ef®cacy end-point Composite end-pointa 57 (38.8) 41 (25.6) 0.014 26 (56.5) 63 (43.8) 0.131 (OR=0.54, 95% CI 0.33±0.88) (OR=0.60, 95% CI 0.31±1.17) Secondary ef®cacy end-point Components by severityb Death (all causes) 31 (21.8) 27 (16.9) 0.311 13 (28.3) 26 (18.1) 0.146 Amputation (all causes) 3 (2.0) 3 (1.9) 1.000 4 (8.7) 16 (11.1) 0.787 New thrombosis 22 (15.0) 11 (6.9) 0.027 9 (19.6) 21 (14.6) 0.486 Death caused by thrombosis 7 (4.8) 0 (0.0) 0.005 7 (15.2) 1 (0.7) <0.001 Any new thrombosisc 3 33 (22.4) 13 (8.1) <0.001 16 (34.8) 28 (19.4) 0.044 Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021

OR=odds ratio; CI=con®dence interval. were 4.4 and 8.9% respectively in the lepirudin group, hypersensitivity to the agent. Bleeding is the major adverse compared with 1.4 and 17.6% in the historical control effect of argatroban, occurring in 2.3±14.4% of patients in group. The cumulative risks of limb amputation were 2.7 clinical trials. The effective dose is titrated against the aPTT and 6.5% in the lepirudin group compared with 2.6 and and activated clotting time (ACT). The drug has a predict- 10.4% in the historical control group. Most importantly, the able pharmacokinetic and pharmacodynamic pro®le, and cumulative risks of new thromboembolic complications there is dose-dependent response in aPTT and ACT. were 6.3 and 10.1% in the lepirudin group and 22.2 and Argatroban is given as an initial bolus of 2 mgkg±1 min±1 27.2% in the historical control group. The differences in the in adult patients without hepatic impairment. The infusion cumulative risk of death, limb amputation or new rate is titrated according to the desired level of anti- thromboembolic complications between the groups were coagulation. An aPTT ratio of 1.5±3.0 is usually achieved statistically signi®cant in favour of lepirudin in the analysis within 4 h of i.v. administration. of time to event (P=0.004). A study with argatroban enrolled 304 patients into two Bivalirudin (Angiomaxâ) is a direct thrombin inhibitor armsÐ160 patients in the HIT arm and 144 in the HITT arm and an analogue of the peptide fragment hirugen, a (Table 7). They were compared with 193 historical control compound derived from hirudin. Unlike lepirudin, the subjects with HIT (n=147) or HITT (n=46).44 Outcomes of binding of bivalirudin to thrombin is reversible. Bivalirudin these patients were compared with those of the historical binds speci®cally to the catalytic site and substrate-binding controls for a period of 37 days after administration of site of thrombin. Bivalirudin has been approved by the US argatroban in the treatment group and after diagnosis of HIT Food and Drug Administration (FDA) for use in patients or HITT in the control group. The primary ef®cacy end- undergoing coronary angioplasty with unstable angina who point was a composite of all-cause death, all-cause ampu- are also on aspirin therapy. tation and new thrombosis within 37 days of baseline. The Argatroban (Argatrobanâ, Novastanâ) is a direct com- secondary ef®cacy end-point included the individual petitive synthetic inhibitor of thrombin derived from L- components of the composite end-point together with arginine, with a molecular weight of 526. It binds reversibly death caused by thrombosis, any new thrombosis, achieve- to the thrombin catalytic site and thus inhibits reactions that ment of adequate anticoagulation, and resolution of are catalysed or induced by the presence of soluble and clot- thrombocytopenia. Argatroban was initiated at 2 mgkg±1 bound thrombin. Actions inhibited by the presence of min±1 and titrated to maintain aPTT at 1.5±3 times baseline argatroban include ®brin formation and activation of for 14 days. Treatment continued for an average of 5.3 days coagulation factors V, VIII and XIII and the natural in the HIT group and 5.9 days in the HITT group. In patients anticoagulant protein C.34 51 61 It is 100% bioavailable with HIT, argatroban therapy signi®cantly reduced new when given i.v. and is metabolized in the liver by thrombosis when compared with control subjects (8.1 vs cytochrome P450 enzymes. Unchanged drug is excreted in 22.4%, P<0.001). However, the reduction in new throm- the urine (16%) and faeces (14%). The primary metabolite boses in patients with HITT on argatroban therapy was more exerts an anticoagulant effect that is three to ®ve times less modest when compared with the reduction in the controls potent than that of the parent drug. It is eliminated as its (19.4 vs 34.8%, P=0.044). There was no decrease in death or metabolite in the faeces (65%), presumably through biliary amputation caused by all causes in the HIT and HITT secretion, and in the urine (22%). Argatroban is contra- groups, but a signi®cant reduction was seen in death caused indicated in patients with active bleeding or known by thrombosis in both groups (P=0.005 and P<0.001

682 HIT/HITT and alternative anticoagulation respectively). Argatroban was associated with more rapid serious thrombotic complications. Circulating antibodies resolution of thrombocytopenia and no increase in bleeding. bind to the transfused platelets and release platelet Argatroban has also been shown to be more effective than microparticles, leading to a thrombotic state. Treatment is heparin for coronary reperfusion after the use of recombi- intended to reduce the thromboembolic complications. nant tissue plasminogen activator (r-tPA) in acute myocar- There are several pharmaceutical agents available for the dial infarction.36 treatment of HIT. Before starting any of these agents, the The use of warfarin with argatroban results in prolonga- clinician should decide which of them is safe and effective tion of the INR beyond that produced by warfarin alone. For for the individual patient. Lepirudin is an irreversible, direct doses up to 2 mgkg±1 min±1, the equivalent INR on warfarin thrombin inhibitor that has shown effectiveness in treating alone (INRW) can be calculated from the INR for cotherapy patients with HIT type II. It does not cross-react with of warfarin and argatroban (INRWA) according to a heparin and can be readily monitored with the aPTT and published graph.48 The error associated with this prediction ECT. Disadvantages include a prolonged half-life, pro- is 60.4 units. For argatroban doses >2 mgkg±1 min±1, the longation of the prothrombin time, possible antibody relationship is less predictable and the error of prediction is formation, and the requirement for dose adjustments in Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 61 units. The equivalent INR on warfarin alone (INRW) patients with renal impairment. Argatroban is the newest cannot be predicted reliably from the INRWA at doses agent used for HIT type II. It binds reversibly to thrombin greater than 2 mgkg±1 min±1.48 and is monitored with the aPTT. Patients who require anticoagulation and have renal impairment may bene®t from the use of argatroban because of its hepatic metabol- Other alternatives in HIT ism. A disadvantage of this agent is prolongation of the There are no convincing data available for the use of the prothrombin time (necessitating additional monitoring in following alternatives. However, they may have a role as an patients who are also receiving warfarin), and dose adjust- 23 67 adjunct with other anticoagulants: ment in patients with hepatic impairment. Danaparoid, (i) intravenous immunoglobulins (IgG class); prostacyclin analogues and desirudin, although previously (ii) platelet transfusionÐcan increase new thrombo- used in HIT, are not licensed for the treatment of HIT or embolic complications and is not indicated currently HITT. even as an adjunct in the treatment of HIT; (iii) plasma exchangeÐits effectiveness is in doubt; (iv) fresh frozen plasmaÐsome use in HIT complicated by warfarin therapy; References (v) glycoprotein IIb/IIIa inhibitor. 1 Ahmad S, Walenga JM, Jeske WP, et al. Functional heterogeneity of antiheparin-platelet±factor 4 antibodies: implications in the pathogenesis of the HIT syndrome. Clin Appl Thromb Hemost Post-heparin thrombosis 1999; 5: S32±7 Studies have shown that HIT antibodies are capable of 2 Amiral J, Bridey F, Dreyfus M, et al. Platelet factor 4 complexed to heparin is the target for antibodies generated in heparin- activating platelets in a non-heparin-dependent manner, as a 1 induced thrombocytopenia. Thromb Haemost 1992; 68: 95±6 result of the generation of `superactive' HIT antibody. 3 Amiral J. Antigens involved in heparin induced thrombo- Using serum of patients with HITT, peripheral blood cytopenia. Semin Haematol 1996; 36 (Suppl 1): 7±11 monocytes and monoclonal antibodies speci®c for H±PF4, 4 Arepally G, McKenzie SE, Jiang X-M, et al.FcgRIIa H/R131 researchers have shown that monoclonal antibodies and polymorphism, subclass speci®c IgG antiheparin/platelet factor 4 antibodies from patients with HITT induce monocyte antibodies and clinical course in patients with heparin-induced activation with secretion of interleukin 8 and cell-surface thrombocytopenia and thrombosis. Blood 1997; 89: 370±5 procoagulant activity. The time-dependence of these 5 Arepally GM, Mayer IM. Antibodies from patients with heparin- changes in vitro (>6 h) far exceeds the short incubation induced thrombocytopenia stimulate monocytic cells to express tissue factor and secrete interleukin-8. Blood 2001; 98: 1252±4 period (<1 h) required for platelet activation by H±PF4 6 Babcock RB, Dumper CW, Scharfman WB. Heparin-induced antibodies. Hence, it is suggested that these changes in the thrombocytopenia. N Engl J Med 1976; 295: 237±41 monocyte could amount to the pathogenesis of thrombi after 7 Barcelona R. Type II HIT new treatment optionsÐCleveland the discontinuation of heparin.5 Clinic Center For Continuing Education. Pharmacother Update 2001; 4: http://www.clevelandclinicmeded.com/medical_info/ pharmacy/septoct2001/thrombocytopenia.htm Summary 8 Baughman RP, Lower EE, Flessa HC, et al. Thrombocytopenia in intensive care unit. Chest 1993; 104: 1243±7 It is important to have a high index of clinical suspicion in a 9 Boons DM, Michels JJ, Stibbe J, et al. Heparin-induced patient with a falling platelet count who is receiving thrombocytopenia and antithrombotic therapy. Lancet 1994; heparin. It is imperative to discontinue all sources of heparin 344: 1296 even before the laboratory con®rmation of HIT. Platelet 10 Brieger DB, Mak K-H, Kottke-Marchant K, et al. Heparin induced transfusion is contraindicated as this can exacerbate the thrombocytopenia. J Am Coll Cardiol 1998; 31: 1449±59

683 Pravinkumar and Webster

11 Chong BH. Heparin-induced thrombocytopenia. Br J Haematol analysis of 2 prospective trials to assess the value of parenteral 1995; 89: 431±9 treatment with lepirudin and its therapeutic aPTT range. Blood 12 Chong BH, Fawaz I, Chesterman CN, et al. Heparin-induced 2000; 96: 846±51 thrombocytopenia: mechanism of interaction of the heparin 30 Greinacher A, Drost W, Michels I, et al. Heparin associated dependant antibody with platelets. Br J Haematol 1989; 73: thrombocytopenia: successful therapy with the heparinoid 235±40 danaparoid sodium in a patient showing cross-reaction to 13 Chong BH, Murray B, Berndt M, et al. Plasma P-selectin is LMW heparins. Ann Haematol 1992; 64: 40±2 increased in thrombotic consumptive platelet disorder. Blood 31 Grif®ths E, Dzik WH. Assays for heparin induced 1994; 83: 1535±41 thrombocytopenia. Transfus Med 1997; 7: 1±11 14 Cines DB, Tomaski A, Tannenbaum S. Immune endothelial cell 32 Harrison L, Johnston M, Massicotte MP, et al. Comparison of 5 injury in heparin associated thrombocytopenia. N Engl J Med mg and 10 mg loading doses in initiation of warfarin therapy. Ann 1987; 316:58 Intern Med 1997; 15: 133±6 15 Deitcher SR, Carman TL. Heparin-induced thrombocytopenia: 33 Hirsch J, Anand S, Halperin J, et al. Guide to anticoagulant natural history, diagnosis, and management. Vasc Med 2001; 6: therapyÐheparin: a statement for healthcare professionals from 113±19 the American Heart Association. Circulation 2001; 103:

16 Demers C, Ginsberg JS, Brill-Edwards P, et al. Rapid 2994±3018 Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 anticoagulation using ancrod for heparin induced 34 Hursting M, Alford K, Becker J, et al. Novastan (brand of thrombocytopenia. Blood 1991; 78: 2194±7 argatroban): a small molecule direct thrombin inhibitor. Semin 17 Denomme GA, Warkentin TE, Horsewood P, et al. Activation of Thromb Haemost 1997; 23: 503±16 platelets by sera containing IgG1 heparin-dependant antibodies: 35 Januzzi JL, Jang IK. Fundamental concepts in the pathobiology of an explanation for the predominance of the FcgRIIa `low heparin-induced thrombocytopenia. J Thromb Thrombolysis 2000; responder' (his131) gene in patients with heparin induced 10: S7±11 thrombocytopenia. J Lab Clin Med 1997; 130: 278±84 36 Jang I, Brown D, Giugliano R, et al. A multicenter, randomized 18 Deykin D. The use of heparin. N Engl J Med 1969; 279: 937±8 study of argatroban versus heparin as adjunct to tissue 19 Eichler P, Budde U, Haas S, et al. First workshop for detection of plasminogen activator (TPA) in acute myocardial infarction: heparin-induced antibodies: validation of the heparin induced myocardial infarction with Novastan and TPA (MINT) study. J platelet activation test (HIPA) in comparison with a PF4/heparin Am Coll Cardiol 1999; 33: 1879±85 ELISA. Thromb Haemost 1999; 81: 625±9 37 Kelton J, Smith J, Warkentin T, et al. Immunoglobulin G from 20 Fondu P. Heparin associated thrombocytopenia: an update. Acta patients with heparin-induced thrombocytopenia binds to a Clin Belg 1995; 50: 343±57 complex of heparin and platelet factor 4. Blood 1994; 83: 3232±9 21 Funk S, Eichler P, Albrecht D, et al. Heparin-induced 38 Kelton J, Sheridan D, Santos A, et al. Heparin induced thrombocytopenia (HIT) in orthopaedic patients: a prospective thrombocytopenia: laboratory studies. Blood 1988; 72: 925±30 cohort trial comparing UFH and LMWH. Ann Haematol 2000; 79 39 Keng T, Chong B. Short report: heparin-induced (Suppl 1): A92 thrombocytopenia and thrombosis syndrome: in vivo cross- 22 Fuster V, Verstraete M. Haemostasis, thrombosis, ®brinolysis, reactivity with danaparoid and successful treatment with r- and cardiovascular disease. In: Braunwauld E, ed. Heart Disease: A hirudin. Br J Haematol 2001; 114: 394±6 Textbook of Cardiovascular Medicine. Philadelphia: W. B. Saunders, 40 Kibbe UR, Rhee RY. Heparin induced thrombocytopenia: 1997; 1809±42 pathophysiology. Semin Vasc Surg 1996; 9: 284±91 23 Greinacher A. Heparin associated thrombocytopenia. Biomed 41 King DJ, Kelton JG. Heparin-associated thrombocytopenia. Ann Prog 1994; 7: 53±6 Intern Med 1984; 100: 536±40 24 Greinarcher A, Volpel H, Janssens U, et al. for the HIT 42 Koster A, Hansen R, Kuppe H, et al. Recombinant hirudin as an Investigators Group. Recombinant hirudin (lepirudin) provides alternative for anticoagulation during cardiopulmonary bypass in safe and effective anticoagulation in patients with heparin- patients with heparin-induced thrombocytopenia type II: a 1 year induced thrombocytopenia: a prospective study. Circulation 1999; experience in 57 patients. J Cardiothorac Vasc Anesth 2000; 14: 99: 73±80 243±8 25 Greinarcher A, Janssens U, Berg G, et al. for the Heparin 43 Laster J, Cikrit D, Walker N, Silver D. The heparin induced Associated Thrombocytopenia Study (HAT) investigators. thrombocytopenia syndrome: an update. Surgery 1987; 102: Lepirudin (recombinant hirudin) for parenteral anticoagulation 763±70 in patients with heparin induced thrombocytopenia. Circulation 44 Lewis B, Wallis D, Mathai W for the ARG-911 study 1999; 100: 587±93 investigators. Argatroban anticoagulant therapy in patients with 26 Greinacher A, Potzsch B, Amiral J, et al. Heparin associated heparin induced thrombocytopenia. Circulation 2001; 103: 1838 thrombocytopenia: isolation of the antibody and characterisation 45 Mikhailov D, Young HC, Linhardt RJ, et al. Heparin of a multimolecular PF-4 heparin complex as the major antigen. dodecasaccharide binding to platelet factor 4 and growth Thromb Haemost 1994; 93: 81±8 related protein alpha. Induction of a partially folded state and 27 Greinacher A, Michels I, Kiefel V, et al. A rapid and sensitive test implications for heparin-induced thrombocytopenia. J Biol Chem for diagnosing heparin associated thrombocytopenia. Thromb 1999; 274: 25317±29 Haemost 1991; 66: 734±6 46 Mohiuddin SM, Hilleman DE, Destache CJ, et al. Ef®cacy and 28 Greinacher A, Amiral J, Dummel V, et al. Laboratory diagnosis of safety of early versus late initiation of warfarin during heparin heparin associated thrombocytopenia and comparison of platelet therapy in acute thromboembolism. Am Heart J 1992; 123: aggregation test, heparin-induced platelet activation test, and 729±32 platelet factor 4/heparin enzyme linked immunosorbent assay. 47 Nand S, Wong W, Yeun B, et al. Heparin induced Transfusion 1994; 34: 381±5 thrombocytopenia with thrombosis: incidence, analysis of risk 29 Greinacher A, Eichler P, Lubenow N, et al. Heparin induced factors, and clinical outcomes in 108 consecutive patients thrombocytopenia with thromboembolic complications: meta treated at a single institution. Am J Haematol 1997; 56: 12±6

684 HIT/HITT and alternative anticoagulation

48 Abbott Laboratories. New Drug Application 20-883. North 66 Wallis DE, Quintos R, Wehrmacher W, Messmore H. Safety of Chicago, IL: Abbott Laboratories, 2001; 4±27 warfarin anticoagulation in patients with heparin-induced 49 Newman PM, Chong BH. Heparin induced thrombocytopenia: thrombocytopenia. Chest 1999; 116: 1333±8 new evidence for dynamic binding of puri®ed anti-PF4-heparin 67 Warkentin TE, Chong BH, Greinacher A. Heparin-induced antibodies to platelets and the resultant platelet activation. Blood thrombocytopenia: towards consensus. Thromb Haemost 1998; 2000; 96: 182±7 79: 1±7 50 Ofosu FA, Esmon CT, Blajchman MA, et al. Unfractioned heparin 68 Warkentin TE, Kelton JG. Interaction of heparin with platelets, inhibits thrombin-catalysed ampli®cation reactions of coagulation including heparin induced thrombocytopenia. In: Bounmeaux H, more ef®ciently than those catalysed by factor Xa. Biochem J ed. Low Molecular Weight Heparins in Prophylaxis and Therapy of 1989; 257: 143±50 Thromboembolic Diseases. Fundamentals and Clinical Cardiology. 51 SmithKline Beecham. Package insert. Argatroban. North New York: Marcel Dekker, 1994; 75±127 Chicago, IL: SmithKline Beecham, 2000 69 Warkentin TE. Heparin induced thrombocytopenia: a 52 Parent F, Bridey F, Dreyfus M, et al. Treatment of severe venous clinicopathologic syndrome. Thromb Haemost 1999; 82 (Suppl): thromboembolism with intravenous hirudin (HBW 023): an 439±47 open pilot. Thromb Haemost 1993; 70: 386 70 Warkentin TE, Levine MN, Hirsch J, et al. Formation of heparin-

53 Medical Economics Company. Heparin. In: Physician's Desk induced thrombocytopenia IgG without thrombocytopenia: Downloaded from https://academic.oup.com/bja/article/90/5/676/269997 by guest on 28 September 2021 Reference. Montvale, NJ: Medical Economics Company, 1998: analysis of a clinical trial. Blood 1995; 86 (Suppl 1): 537 3044 71 Warkentin TE. Heparin induced skin lesions. Br J Haematol 1996; 54 Potzsch B, Hund S, Madlener K, et al. Monitoring of recombinant 92: 494±7 hirudin: assessment of a plasma-based ecarin clotting time assay. 72 Warkentin TE, Levine MN, Hirsch J, et al. Heparin induced Thromb Res 1997; 86: 373±83 thrombocytopenia in patients treated with low molecular weight 55 Pouplard C, May MA, Lochmann S, et al. Antibodies to platelet heparin or unfractioned heparin. N Engl J Med 1995; 332: factor 4±heparin after cardiopulmonary bypass in patients with 1330±5 unfractioned heparin or a low molecular weight heparin: clinical 73 Warkentin TE. Clinical picture of heparin induced implications for heparin induced thrombocytopenia. Circulation thrombocytopenia. In: Warkentin TE, Greinarcher A, eds. 1999; 99: 2530±6 Heparin Induced Thrombocytopenia. New York: Marcel Dekker, 56 Rhodes GR, Dixon RH, Silver D. Heparin induced 2000; 43±80 thrombocytopenia with thrombotic and haemorrhagic 74 Warkentin TE, Kelton JG. A 14-year study of heparin induced manifestations. Surg Gynecol Obstet 1973; 136: 409±16 thrombocytopenia. Am J Med 1996; 101: 502±7 57 Schulman LN, Branwauld E, Rosenthal DS. 75 Warkentin TE, Elavathil LJ, Hayward CP, et al. Pathogenesis of Haematological±oncological disorders and heart disease. In: venous limb gangrene associated with heparin-induced Braunwauld E, ed. Heart Disease: A Textbook of Cardiovascular thrombocytopenia. Ann Intern Med 1997; 127: 804±12 Medicine. Philadelphia: W. B. Saunders, 1997; 1786±808 76 Warkentin TE. Limitations of conventional treatment options for 58 Sheridan D, Carter C, Kelton JG. A diagnostic test for heparin induced thrombocytopenia. Blood 1986; 67: 27±30 heparin induced thrombocytopenia. Semin Haematol 1998; 35 59 Stringer KA, Linden®eld J. Hirudins: antithrombin anticoagulants. (Suppl 4): 17 Ann Pharmacother 1992; 26: 1535 77 Warkentin TE, Hayward CP, Boshkov LK, et al. Sera from 60 Suh JS, Aster RH, Visentin GP. Antibodies from patients with patients with heparin-induced thrombocytopenia generate heparin-induced thrombocytopenia/thrombosis recognise platelet-derived microparticles with procoagulant activity: an different epitopes on heparin platelet factor 4. Blood 1998; 91: explanation for the thrombotic complications of heparin induced 916±22 thrombocytopenia. Blood 1994; 84: 3691±9 61 Swan M, Hursting M. The pharmacokinetics and 78 Warkentin TE, Kelton JG. Temporal aspects of heparin induced pharmacodynamics of argatroban: effect of age, gender, and thrombocytopenia. N Engl J Med 2001; 344: 1286±92 hepatic or renal dysfunction. Pharmacotherapy 2000; 20: 318±29 79 Weismann R, Tobin RW. Arterial embolism occurring during 62 Villanueva GB, Danishefsky I. Evidence for a heparin-induced systemic heparin therapy. Arch Surg 1958; 76: 219±27 conformational change on antithrombin. Biochem Biophys Res 80 Weitz JL, Hudoba M, Massel D, et al. Clot bound thrombin is Commun 1977; 74: 803±9 protected from inhibition by heparin-antithrombin III but is 63 Visentin GP, Ford SE, Scott JP, et al. Antibodies from patients susceptible to inactivation by antithrombin III-independent with heparin-induced thrombocytopenia/thrombosis are speci®c inhibitors. J Clin Invest 1990; 86: 385 for platelet factor 4 complexed with heparin or bound to 81 Weitz JL, Leslie B, Hudoba M. Thrombin binds to soluble ®brin endothelial cells. J Clin Invest 1994; 93: 81±8 degradation products where it is protected from inhibition from 64 Visentin GP, Aster RH. Heparin induced thrombocytopenia and heparin-antithrombin but susceptible to inactivation by thrombosis. Curr Opin Haematol 1995; 2: 351±7 antithrombin independent inhibitors. Circulation 1999; 99:73 65 Wallis DE, Workman DL, Lewis BE, et al. Failure of early heparin 82 Ziporen L, Li ZQ, Park KS, et al. De®ning an antigenic epitope on cessation as treatment for heparin-induced thrombocytopenia. platelet factor 4 associated with heparin-induced Am J Med 1999; 106: 629±35 thrombocytopenia. Blood 1998; 92: 3250±9

685