HEPARIN INDUCED THROMBOCYTOPENIA (HIT) Who gets it How and why they get it How to recognize it How to treat it

HEPARIN-INDUCED THROMBOCYTOPENIA Causative agents Unfractionated heparin (UFH) (beef > pork) Continuous iv infusion Cardiopulmonary bypass Low dose sq Heparin flushes Heparin-bonded catheters Low molecular weight heparin More likely to cause HIT if pt previously exposed to UFH Most cases associated with administration of UFH for treatment or prophylaxis

HEPARIN-INDUCED THROMBOCYTOPENIA Clinical manifestations Isolated thrombocytopenia (“Isolated HIT”) Arterial or venous thrombosis (HITT) DVT, PE, MI, stroke, peripheral arterial occlusion DIC, microangiopathic hemolytic anemia Skin necrosis (at injection sites or distant) Venous limb gangrene (? Role of warfarin) Sudden death ARDS Hemorrhagic adrenal infarction

HEPARIN-INDUCED THROMBOCYTOPENIA Epidemiology UFH > LMWH >> Fondaparinux Duration of heparin treatment ≥ 6 days Rarely occurs in patients < 40 years old 2-3 fold higher incidence in women Surgical > medical > obstetric patients Incidence in trauma patients proportional to severity of trauma Related to degree of platelet activation? Blood 2012; 119: 2209

HEPARIN-INDUCED THROMBOCYTOPENIA Incidence and presenting features Warkentin and Kelton, Am J Med 1996;101:502 Presenting with thrombosis (n=65) Presenting with no thrombosis (n=62) Total (n=127) Age 67 ± 10.7 66.7 ± 12.3 67.0 ± 11.4 Male/Female 27/38 33/29 60/67 SURGICAL PTS 51 33 84 (66.1%) Orthopedic 25 15 40 Cardiovascular 10 9 19 Oncology 7 6 13 General 2 8 Neurosurgery 3 1 4 MEDICAL PTS 14 29 43 (33.9%) Cardiac 16 DVT or PE 11 Other 12 In retrospective studies, HIT most often recognized in surgery pts, esp ortho surgery ? Because they are at higher risk for thrombosis to start with

THROMBOTIC COMPLICATIONS IN HIT Am J Med 1996;101:502 Type of thrombosis Pts presenting with thrombosis (n=65) Pts presenting with only thrombocytopenia (n=62) VENOUS (n=78) 54 24 DVT (n=61) 40 21 New 35 Progression 4 Recurrence 1 PE (n=32) 26 6 25 5 ARTERIAL (n=18) 12 Limb 7 2 Myocardial infarct 3 Thrombotic stroke Other (n=3) Sudden death Adrenal hemorrhage NO THROMBOSIS (n=30) NA 30 Most thromboses venous Pts diagnosed with “isolated HIT” have high likelihood of subsequent thombosis

ISOLATED HIT IS ASSOCIATED WITH A HIGH RISK OF SUBSEQUENT THROMBOSIS Caveat: retrospective study, potential for diagnostic selection bias But it seems clear from this and other studies that isolated HIT does create a very high risk for subsequent thrombosis Am J Med 1996;101:502 Over 50% of patients presenting with “isolated HIT” had a subsequent thrombotic episode within 30 days Substitution of warfarin for heparin after the onset of thrombocytopenia did not prevent thrombosis

UNFRACTIONATED HEPARIN IS MORE LIKELY TO CAUSE HIT THAN LMWH Randomized trial in pts having hip surgery Warkentin et al, NEJM 1995;332:1330 3-4% incidence of HIT in hip surgery, probably the highest risk group

Odds ratio for thrombosis 37 times higher in HIT pts THE FREQUENCY OF THROMBOSIS AFTER HIP SURGERY IS MUCH HIGHER IN PATIENTS WITH HIT THAN IN THOSE WITHOUT HIT Warkentin et al, NEJM 1995;332:1330 Odds ratio for thrombosis 37 times higher in HIT pts

THE HIT INCIDENCE IN MEDICAL PATIENTS TREATED WITH LOW DOSE SQ UFH IS ABOUT 2% Girolami et al, Blood 2003;101:2955 But medical pts at substantial risk as well All cases in patients receiving UFH as prophylaxis HIT associated with 40-fold increase in risk of thrombosis

LMWH IS MORE LIKELY TO CAUSE HIT IN PATIENTS WITH PRIOR UFH EXPOSURE Prandoni et al, Blood 2005;106:3049 Prospective cohort study, 1754 medical pts 0.8% overall incidence of HIT 0.3% incidence if no prior UFH exposure 1.7% incidence if prior UFH exposure All cases in first 2 weeks Prevalence of thromboembolism 16.6x higher in patients with HIT (29% vs 2.4%) Incidence of HIT in pts getting LMWH after prior exposure to UFH roughly comparable to incidence with UFH

Lubenow, N. et al. Blood 2010;115:1797-1803 Development of HIT antibodies is more common in major surgery than minor surgery, and more common with UFH than LMWH Results of a prospective trial Lubenow, N. et al. Blood 2010;115:1797-1803

THROMBOSIS IN HIT MAY OCCUR WITH NORMAL PLATELET COUNT Warkentin et al, NEJM 1995;332:1330

THE PLATELET COUNT DROPS PRIOR TO THROMBOSIS IN HIT Warkentin et al, NEJM 1995;332:1330 * Platelet count normally rises steadily for at least a week after hip surgery. Note that all pts with HIT and normal plts had at least two days of dropping plts before thrombotic event *Thrombotic episode

Recent heparin exposure may cause “rapid onset” HIT Warkentin and Kelton, NEJM 2001;344:1286 HIT virtually never happens less than 4 days after starting heparin UNLESS there has been prior exposure to heparin

Rapid-onset HIT is associated with re-exposure to heparin within 90 days Warkentin and Kelton, NEJM 2001;344:1286 And this exposure usually has been within a month, almost always within 3 months

Heparin-dependent antibodies usually disappear within 90 days an episode of HIT Warkentin et al, NEJM 1995;332:1330 Because the antibodies that cause the problem disappear during this time period – functional assay being more predictive of the condition

DELAYED ONSET HIT Ann Intern Med 2002;136:210 Describes 14 patients treated with heparin, discharged, and later re-hospitalized with thromboembolism and positive tests for HIT antibodies Most patients got heparin during cardiac surgery 12/14 had mild thrombocytopenia (66-145K) at time of thrombotic episode Median time between discharge and readmission 14 days, maximum 40 days 11 patients re-treated with heparin: all had clinical deterioration and worsening thrombocytopenia 3 patients died We have said that HIT can occur without thrombocytopenia. It can also occur when a patient is no longer getting heparin. This is a particularly difficult form of HIT to diagnose, and the consequences of not diagnosing it can be dire

PATHOPHYSIOLOGY OF HIT

HIT IS CAUSED BY ANTIBODIES AGAINST A HEPARIN-PLATELET FACTOR 4 COMPLEX 3 Antibody binds heparin-PF4 complex Fab FC Antibody binding to platelet FC receptor activates platelet 4 Heparin-induced thrombocytopenia: Platelet factor 4 (PF4) released by activated platelet. This binds heparin, creating a potential neoantigen. Antibody binds the complex of heparin-PF4. The antigen antibody complex then binds to the FC receptor on another platelet, causing platelet activation. This may account for the association between HIT and thrombosis in some patients. 2 Heparin PF4 binds heparin 1 PF4 Activated platelet secretes PF4 FC receptor Platelet membrane

PATHOPHYSIOLOGY OF HIT Warkentin, Brit J Haematol 2003;121:535 Heparin-PF4 complexes stimulate antibody production Ag-Ab complex binds to and activates platelets, monocytes Size of immune complex is critical, varies with PF4 and heparin concentrations Inhibited by high heparin concentrations Activated platelets release procoagulant microparticles Activated monocytes produce tissue factor Antibodies may cross-react with PF4 bound to endothelial cell heparan sulfate → vessel wall injury Some HIT antibodies can activate platelets in the absence of heparin

Heparin concentration affects the size and charge of heparin:PF4 complexes and their ability to activate platelets Low heparin:PF4 ratio → small complexes High heparin:PF4 ratio → small complexes 1:1 heparin:PF4 → large complexes Charge of complexes Heparin conc→ Blood 2007;110:4253

Smaller heparin molecules → less platelet activation by HIT antibodies NEJM 2013;368:737

Clinical factors may help determine the likelihood of developing HIT Healthy volunteers given heparin or LMWH make IgM antibodies to heparin/PF4 Pathologic HIT antibodies are usually IgG Concomitant immune stimulus necessary to promote IgG HIT antibody formation? Higher PF4 levels after surgery or acute illness may promote formation of larger immune complexes Formation of DNA-PF4 complexes following tissue injury may initiate immune response, antibodies can cross-react with heparin-PF4 complexes Blood 2007;110:4253 J Thromb Haemost 2012;10:1446 Blood 2013;122:272

HIT Antibodies can activate platelets in the absence of heparin J Thromb Haemost 2005;3:2168

Why is HIT so different from other immune-mediated drug reactions? Disease feature Possible explanation Higher incidence after surgery or trauma More platelet activation and PF4 release→ more/larger immune complexes Higher incidence with unfractionated heparin Larger heparin molecules → larger immune complexes Paradoxical thrombosis Activation of platelets & monocytes by immune complexes Onset may follow heparin discontinuation Damage to endothelial cells by antibodies bound to PF4/heparan sulfate complexes? Heparin-independent antibodies? Lack of recurrence with heparin re-exposure Complex pathophysiology, antibody formation necessary but not sufficient to cause clinical disease

DIAGNOSIS OF HIT

DISTINGUISHING IMMUNE FROM NON-IMMUNE HEPARIN INDUCED THROMBOCYTOPENIA Many patients have a transient decrease in platelets within 24 hours of receiving heparin. This is not an antibody-mediated effect and not associated with thrombosis How can it be distinguished from HIT? By the time course By the clinical picture By serology and other lab tests

Severe thrombocytopenia is rare in HIT Warkentin, Brit J Haematol 2003;121:535 Median platelet nadir 55K 15% had nadir >150K (diagnosed because platelet count fell more than 50% or because of clinical events) The severity of thrombocytopenia did not predict thrombotic events 15% are not thrombocytopenic at all. Rarely does plt count drop below 20K No connection between severity of thrombocytopenia and clinical course

Clincal features that favor a diagnosis of HIT Thrombocytopenia usually but not always present and typically not severe Blood 2012;119:2209

The 4 T score predicts a positive HIT antibody test J Thrombos Haemost 2006;4:759 A corollary – more likely to get false + test if 4T score low – so don’t order test willy-nilly Score % Testing positive <4 0.8% 4-5 11% >5 34%

LABORATORY DIAGNOSIS OF HIT Immunoassay for heparin-PF4 antibodies (EIA) Very high sensitivity, rapid turnaround “High positive” results typical of HITT Rapid “bedside” assay Many false positive results, not recommended C14 Serotonin release assay (SRA) Measures heparin-dependent platelet activation Best predictor of thrombotic risk Limited availability (Blood Center of SE Wisconsin), slower turnaround A positive EIA test in a patient with a low pre-test probability of HIT according to 4T rule is likely to be a false positive ELISA a very good screening test and it’s all you need if the clinical picture fits Consider SRA when clinical picture cloudy or when risk of giving alternative anticoagulant high

The serotonin release assay predicts thrombosis in HIT SRA and EIA negative EIA positive, SRA negative SRA and EIA positive Am J Hematol 2007;82:1037

The incidence of HIT antibody formation and risk of HIT varies among different patient populations The “Iceberg” model Cardiac patients have high incidence of serologic abnormalities but low incidence of clinical HIT Ortho patients getting UFH have highest risk of clinical HIT Warkentin, Brit J Haematol 2003;121:535

A STRONGLY POSITIVE EIA RESULT IS ASSOCIATED WITH HIGHER RISK OF SUBSEQUENT THROMBOSIS We don’t routinely report OD here OD values in HIT vs HITT patients Thrombosis-free survival vs OD J Thrombos Haemost 2004;2:2133-7

The dilemma of HIT testing in critically ill patients Most critically ill patients receive heparin products Up to 20% have at least one clinical criterion for HIT Only about 1% actually develop HIT Prediction rules (4T score, etc) do not function well in this patient population A diagnosis of HIT in a critically ill patient should always be confirmed by SRA testing Crowther et al, Blood 2011;118:198

TREATMENT OF HIT

TREATMENT OF HIT Discontinue all heparin, including flushes LMWH may cross-react with HIT antibodies, should not be used If thrombosis present: give alternative thrombin inhibitor Consider treating even if thrombosis absent (high risk of thrombosis in patients with isolated HIT) Treatment alternatives: Direct inhibitors Lepirudin Bivalirudin (little data, but approved for HIT patients having PCI) Argatroban Dabigatran? Indirect inhibitors Fondaparinux Do not give warfarin (risk of venous gangrene)

DIRECT THROMBIN INHIBITORS Argatroban (Novastan®) Synthetic arginine derivative Clearance mainly hepatic (can use in renal failure); halflife 40-50 min Given by continuous iv infusion, monitoring aPTT Coagulopathic patients (long baseline aPTT) difficult to monitor No antidote No data yet for oral thrombin and Xa inhibitors Most of the data on the efficacy of these drugs is on lepirudin

Argatroban therapy in HIT Pooled data from 2 prospective non-randomized trials Hazard ratio with argatroban 0.3 vs historical controls No difference in major bleeding vs controls Chest 2006;129:1407

ARGATROBAN IN HIT ACCP RECOMMENDATIONS Bolus: None Continuous infusion: Normal organ function: 2 mcg/kg/mon Liver dysfunction, post cardiac surgery, anasarca: 0.5-1.2 mcg/kg/mon Adjust aPTT to 1.5-3.0 x baseline Check aPTT q 4h Argatroban prolongs PT/INR, making transition to warfarin tricky

FONDAPARINUX (Arixtra®) Synthetic polysaccharide, inhibits factor Xa preferentially Does not typically cross-react with HIT antibodies Long half-life (17-20 h), no antidote SQ administration Monitoring unnecessary Not FDA-approved for HIT treatment Rare reports of fondaparinux-associated HIT (NEJM 2007; 356:2653)

Fondaparinux appears to be effective and safe in patients with HIT Reference N New thrombosis Major Bleeding Kuo & Kovacs Thromb Haemost 2005 5 0/5 Lobo et al Thromb Haemost 2007 7 0/7 Grouzi et al Clin Appl Thromb Haemost 2009 24 0/24 Pooled Data 36 0/36 Warkentin, Hematol Oncol Clin N Am 2010; 24:755

VENOUS GANGRENE Arch Intern Med 2004;164:66 Tissue death Starting warfarin too soon in HIT may promote this process Arch Intern Med 2004;164:66

WARFARIN MAY PROMOTE VENOUS GANGRENE IN HIT Ann Intern Med 1997;127:804-12 Retrospective review of 158 cases of HIT 8/8 patients with venous limb gangrene treated with warfarin, vs 3/10 with arterial thrombosis (p=.004) Median INR 5.8 in patients with venous gangrene vs 3.1 in those who did not (p<.001) Elevated ratio of thrombin-antithrombin complex to protein C activity in patients with venous gangrene vs controls Conclusion: warfarin treatment of DVT associated with HIT may cause venous limb gangrene, possibly because of acquired defect in protein C pathway Do not start warfarin treatment until HIT resolves (platelet count returns to normal) Retrospective study in which all of the HIT patients who developed VG had been treated with W Long INR not protective Biochemical evidence that warfarin’s effect on protein C levels may mediate this effect

Platelet transfusions are associated with worse outcomes in HIT & TTP Blood 2015;125:1470

How long should anticoagulation continue after diagnosis of HIT? HIT with thrombosis: 3-6 months Isolated HIT (no thrombosis): at least until platelets normal, consider continuing for 30 days

Can patients with a history of HIT be given heparin again? Heparin should not be given while tests for heparin antibodies remain positive If cardiac surgery cannot be delayed, use alternative anticoagulant (e.g., bivalirudin) HIT recurrence or secondary antibody response uncommon in patients with “remote HIT” and negative HIT antibody test Heparin administration should be limited to the intraoperative period

SUMMARY-1 HIT typically occurs after 5+ days of exposure to unfractionated heparin Suspect HIT if platelet count falls by > 50% during heparin administration, or if new thrombotic event occurs within 2-3 weeks of heparin exposure Onset may be earlier if there was prior exposure to heparin within past 100 days Onset may follow discontinuation of heparin LMWH rarely causes HIT but may perpetuate it Risk of thrombosis in HIT is high even if patient does not have thrombosis at time of diagnosis

SUMMARY-2 HIT is caused by production of antibodies to heparin-PF4 complex that activate platelets HIT is unlikely if tests for heparin-PF4 antibodies are negative Patients with HIT should generally be treated with a thrombin or Xa inhibitor other than heparin or LMWH Warfarin treatment should be delayed until platelet count is normal