- Email: [email protected]
Heparin-Induced Thrombocytopenia
Hematol Oncol Clin N Am 21 (2007) 589–607
HEMATOLOGY/ONCOLOGY CLINICS OF NORTH AMERICA
Heparin-Induced Thrombocytopenia Theodore E. Warkentin, MDa,b,c,d,* a
Department of Pathology and Molecular Medicine, Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario, Canada b Department of Medicine, Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario, Canada c Transfusion Medicine, Hamilton Regional Laboratory Medicine Program, Hamilton, Ontario, Canada d Service of Clinical Hematology, Hamilton Health Sciences, Hamilton General Hospital, Hamilton, Ontario, Canada
H
eparin-induced thrombocytopenia (HIT) is an immune-mediated adverse drug reaction that is caused by heparin-dependent, plateletactivating IgG antibodies that recognize complexes of platelet factor 4 (PF4) bound to heparin [1,2]. HIT is a strong risk factor for venous and arterial thrombosis (odds ratio, 20 to 40) [3]. HIT often leads to thrombosis in large veins and arteries; however, HIT also is a risk factor for microvascular thrombosis, especially with warfarin treatment of HIT-associated deep-vein thrombosis (DVT) or overt disseminated intravascular coagulation (DIC) [4,5]. Thrombosis usually occurs in association with large platelet count declines [6,7], consistent with in vivo platelet activation. A central role for in vivo thrombin generation [4] provides a rationale for the use of anticoagulants that inhibit thrombin or its activity. DEFINITION AND TERMINOLOGY HIT can be defined as any clinical event (or events) best explained by plateletactivating anti-PF4/heparin antibodies (‘‘HIT antibodies’’) in a patient who is receiving, or who recently received, heparin [1]. Thrombocytopenia is the most common event in HIT, and it is observed in at least 90% of patients, depending on how thrombocytopenia is defined. Most patients who have HIT develop thrombosis [1,5–9].
Studies cited [3,4,6,7,9,11,12,14,17–19,23–25,27,28,30–37,50] were supported by the Heart and Stroke Foundation of Ontario (operating grants #A2449, #T2967, B3763, #T4502, and #T5207).
*Room 1-180A, Hamilton Regional Laboratory Medicine Program, Hamilton Health Sciences, Hamilton General Hospital, 237 Barton Street East, Hamilton, Ontario L8L 2X2, Canada. E-mail address: [email protected] 0889-8588/07/$ – see front matter doi:10.1016/j.hoc.2007.06.004
Ó 2007 Elsevier Inc. All rights reserved. hemonc.theclinics.com
590
WARKENTIN
HIT is a clinicopathologic syndrome [1]; thus, the diagnosis requires one or more clinical events (eg, thrombocytopenia, thrombosis) bearing a temporal association with pathologic HIT antibody formation. Thus, a patient suspected to have HIT but in whom antibodies cannot be detected does not have this diagnosis. Some disorders mimic HIT (‘‘pseudo-HIT’’), pointing to the importance of laboratory testing for HIT antibodies [10,11]. PATHOGENESIS Fig. 1 summarizes HIT pathogenesis [12]. The central concept is formation of heparin-dependent IgG that activates platelets by way of their FccIIa receptors [13]. IgG and IgM class antibodies are not pathogenic [14] because they cannot activate platelets by way of Fcc (IgG) receptors. The target antigen is a complex between (anionic) heparin and (cationic) PF4 [15], a tetrameric member of the CXC subfamily of chemokines. HIT antibodies recognize conformationally altered sites on PF4 resulting from its binding to heparin [16] or because of close approximation of PF4 tetramers by heparin charge neutralization [17]. Low molecular weight heparin (LMWH) is less likely to trigger antibodies and HIT compared with unfractionated heparin (UFH) [6,7], particularly in women receiving surgical thromboprophylaxis [18]. The pentasaccharide anticoagulant, fondaparinux, although similarly immunogenic as LMWH, does not form well the antigens on PF4 [19], suggesting that it has an even lower (perhaps negligible) risk for causing HIT. Factors leading to thrombosis in HIT include [12] the platelet-activating nature of HIT antibodies (including formation of procoagulant, platelet-derived microparticles), as well as ‘‘pancellular’’ activation (endothelium, monocytes) and neutralization of the anticoagulant effect of heparin by PF4 released from activated platelets. FREQUENCY Four factors influence the frequency of HIT: duration of heparin use, type of heparin, type of patient population, and patient gender (Table 1) [5–7,18,20,21]. The ‘‘typical’’ patient who has HIT is a woman receiving postsurgery thromboprophylaxis with UFH for more than 1 week (risk, 1%–5%). Despite the greater risk in women, HIT is rare in pregnancy [22]. HIT occurs in about 1% of patients receiving UFH for treatment of venous thrombosis and in about 0.1% to 0.5% of postoperative patients receiving LMWH prophylaxis [21]. CLINICAL PICTURE HIT is a distinct syndrome compared with other drugs that cause immune thrombocytopenia (eg, quinine, vancomycin, glycoprotein IIb/IIIa antagonists) (Fig. 2) [23]. Although these other agents produce severe thrombocytopenia and mucocutaneous hemorrhage, HIT usually results in a moderate thrombocytopenia, but with a greatly increased risk for thrombosis. Also, although classic drug-induced immune thrombocytopenia causes dramatic bleeding,
HEPARIN-INDUCED THROMBOCYTOPENIA
591
Fig. 1. Pathogenesis of HIT. The figure illustrates two potential explanations for thrombosis in HIT. Activation of platelets (plt) by anti-PF4/heparin IgG antibodies (‘‘HIT antibodies’’), leading to formation of procoagulant, platelet-derived microparticles and neutralization of heparin by PF4 released from activated platelets, results in a marked increase in thrombin generation (‘‘hypercoagulability state’’) that is characterized by increased risks for venous or arterial thrombosis and coumarin-associated microthrombosis (venous limb gangrene, skin necrosis). It also is possible that unique mechanisms operative in HIT explain unusual thromboses, such as arterial ‘‘white clots.’’ For example, HIT antibodies have been shown to activate endothelium and monocytes (leading to cell surface tissue factor expression, although this stimulation may be largely ‘‘indirect’’ through poorly defined mechanisms involving platelet activation and, possibly, formation of platelet-derived microparticles). Further, aggregates of platelets and polymorphonuclear (PMN) leukocytes have been described in HIT. The extent to which the cooperative interactions between platelets, platelet-derived microparticles, PMN leukocytes, monocytes, and endothelium lead to arterial (or venous) thrombotic events in HIT, either in large or small vessels, remains unclear. (From Warkentin TE. An overview of the heparin-induced thrombocytopenia syndrome. Semin Thromb Hemost 2004;30(3):275; with permission.)
WARKENTIN
592
Table 1 Risk factors for immune heparin-induced thrombocytopenia Risk factor
Odds ratio
Duration of heparin use (>1 week versus <1 day)
20–100
Type of heparin (UFH > LMWH > fondaparinux)
10–15
Type of patient (surgery > medical > pregnancy)
3–4
Gender (female > male)
1.5–2.0
Comment OR estimated as follows: risk for HIT postcardiac surgery (with UFH prophylaxis for >1 week) 2% versus risk for ‘‘delayedonset HIT’’ (without prophylaxis) 0.02–0.1% Difference in risk for HIT between heparin types is established for postsurgical thromboprophylaxis (UFH versus LMWH) and is more pronounced in women. ORs shown are for UFH versus LMWH (women, post-surgical prophylaxis) [18] Highest reported frequencies of HIT are in postsurgical thromboprophylaxis (OR shown is for surgery versus medical) [18] Difference in risk for HIT between women and men has only been established for UFH thromboprophylaxis [18]
Abbreviation: OR, odds ratio.
fatalities are rare, whereas in HIT, fatal or long-term sequelae of thrombosis are not uncommon [23]. Thrombocytopenia The median platelet count nadir in HIT is about 60 109/L [5,9,23]. Most patients evince a 50% or greater decrease in the platelet count [7]. In postoperative patients, the appropriate ‘‘baseline’’ platelet count is not the preoperative platelet count, but rather the highest postoperative platelet count preceding the HIT-associated platelet count decrease [5,7]. Timing HIT is a transient immune response that occurs most often when heparin is given intra- or perioperatively. ‘‘Typical-onset’’ HIT (70% of patients) manifests as a platelet count decrease that begins 5 to 10 days after starting a course of heparin (first day of heparin use ¼ day 0) [24], with a median of 2 days required to reach a threshold defining thrombocytopenia [5]. ‘‘Rapid-onset’’ HIT is the presenting feature in about one quarter of patients and is defined as a decrease in platelet count within 24 hours of administering a course of heparin [24]. This presentation occurs because the patient already has circulating HIT antibodies, a result of exposure to heparin within the past few weeks or months.
HEPARIN-INDUCED THROMBOCYTOPENIA
No. of Patients (arbitrary units, increasing from bottom to top)
Drug-induced immune thrombocytopenia 9
60 x 10 /L (median)
Bleeding
5
10
Heparin-induced thrombocytopenia 9
<10 x 10 /L (median)
3
593
Thrombosis
20
50
100
200
500
1000
Nadir Platelet Counts (x10-9/L) Shown on a Log10 Scale Fig. 2. Clinical picture of HIT versus classic drug-induced immune-mediated thrombocytopenia. ‘‘Classic’’ drug-induced immune-mediated thrombocytopenia (eg, caused by quinine, vancomycin, or glycoprotein IIb/IIIa antagonists, among other drugs) typically produces severe thrombocytopenia (median platelet count nadir, about 10 109/L) and associated mucocutaneous bleeding. In contrast, HIT typically results in mild-to-moderate thrombocytopenia (median platelet count nadir, 60 109/L) and associated venous or arterial thrombosis. Note that the heights of the two peaks are not drawn to scale, because HIT is much more common than all other causes of drug-induced immune-mediated thrombocytopenia combined. (From Warkentin TE. Drug-induced immune-mediated thrombocytopenia: from purpura to thrombosis. N Engl J Med 2007;356(9):891; with permission. Copyright Ó 2007, Massachusetts Medical Society.)
‘‘Delayed-onset’’ HIT refers to when the platelet count begins to decrease after all heparin has been stopped [25]. This results from unusually high levels of antibodies that can activate platelets even in the absence of pharmacologic heparin, perhaps because the antibodies recognize PF4 bound to plateletassociated chondroitin sulfate [26]. ‘‘Spontaneous’’ HIT is a rare disorder in which a patient develops a clinical profile suggestive of HIT and in whom high levels of HIT antibodies can be detected, even though no heparin has been administered previously [27]. Thrombosis and Other Sequelae Table 2 presents the complications of HIT [5]. The strong association between HIT and thrombosis means that this diagnosis should be considered in patients who develop symptomatic thrombosis during, or within several days after stopping, heparin use [28]. In some studies of postsurgical thromboprophylaxis with UFH, the development of symptomatic thrombosis indicated about a 50% risk for HIT being present [29,30]. Venous thrombosis is the most common sequela of HIT [1,5–9,21]. Venous limb ischemia usually indicates warfarin toxicity [4,5]. Adrenal hemorrhage is a sign of adrenal vein thrombosis, with secondary hemorrhage [28]; when
594
Table 2 Clinical sequelae of heparin-induce thrombocytopenia Comment
Venous thromboembolism
DVT (50%) and pulmonary embolism (25%) are the two most common sequelae of HIT, especially in postoperative patients Limb artery thrombosis is most common (10%–15%) followed by thrombotic stroke (5%–10%), myocardial infarction (3%–5%), and other (eg, mesenteric artery thrombosis, spinal artery thrombosis) Besides cerebral artery thrombosis, other causes of stroke include cardiac embolization and cerebral venous (dural sinus) thrombosis HIT is a strong risk factor for coumarin necrosis, especially venous limb gangrene (acral limb necrosis usually in the setting of DVT), but also ‘‘classic’’ skin necrosis (subdermal and dermal necrosis, usually in nonacral sites [eg, breast, abdominal wall, thigh, calf, or forearm]); characterized by microvascular thrombosis Associated with thrombosis of adrenal veins; unilateral adrenal hemorrhage presents as abdominal or flank pain; when adrenal hemorrhage is bilateral, adrenal crisis is likely to occur Necrotizing skin lesions at heparin injection sites are specific for HIT; erythematous plaques at injection sites also can indicate HIT; some patients with skin lesions do not evince thrombocytopenia One or more signs or symptoms that begin 5–30 min after IV heparin injection (or sc LMWH injection): cardiac (tachycardia, chest pain, hypertension, cardiac arrest), respiratory (dyspnea, tachypnea, chest pain, respiratory arrest), inflammatory (fever, chills, rigors, flushing), neurologic (pounding headache, transient global amnesia syndrome), gastrointestinal (diarrhea); these features are not that of anaphylaxis About 10%–20% of patients who have HIT have overt (decompensated) DIC (eg, hypofibrinogenemia, increased INR, positive protamine sulfate paracoagulation assay, microangiopathic hemolysis, or circulating normoblasts [nucleated red cells])
Arterial thrombosis Thrombotic stroke Coumarin necrosis
Adrenal hemorrhage Skin lesions at heparin injection site Anaphylactoid reaction
Overt DIC
Abbreviations: IV, intravenous; sc, subcutaneous.
WARKENTIN
Sequela
HEPARIN-INDUCED THROMBOCYTOPENIA
595
bilateral, death can result from adrenal failure. Arterial thrombosis most commonly affects large lower-limb arteries, with thrombotic stroke and myocardial infarction seen less commonly [5,9]. Limb ischemia can be classified based on whether pulses are palpable (or Doppler identifiable). Absent pulses suggest that limb ischemia is caused by artery occlusion, whereas pulses despite acral (distal extremity) ischemia suggest microvascular occlusion. The hallmark of coumarin-associated venous limb ischemia is an elevated international normalized ratio (INR; usually > 3.5), which is a surrogate marker for severely depleted protein C (in parallel with reduced factor VII) [4]. Rarely, limb ischemia occurs as a result of severe DIC alone [5]. Classic (nonacral) skin necrosis also has been reported in HIT [4,5]. Table 2 also summarizes other miscellaneous sequelae of HIT, including necrotizing and nonnecrotizing skin lesions at heparin injection sites [31], acute anaphylactoid reactions following intravenous heparin bolus administration [5,32], and overt DIC [5,25]. CLINICAL SCORING SYSTEM FOR HEPARIN-INDUCED THROMBOCYTOPENIA In evaluating a patient for possible HIT, a clinical scoring system can help [33]. One system, the 4 Ts (Table 3), evaluates Thrombocytopenia, its Timing, the presence of Thrombosis (or other sequelae of HIT), and whether oTher plausible explanations for thrombocytopenia or thrombosis are present. A low score (3 points) makes HIT unlikely (< 2%). In some settings, a high score predicts a high likelihood of HIT [33]. LABORATORY DETECTION OF HEPARIN-INDUCED THROMBOCYTOPENIA ANTIBODIES Detectability of HIT antibodies by the laboratory is a sine qua non of HIT diagnosis. There are two complementary ways to detect antibodies: platelet activation assays and PF4-dependent immunoassays [34]. In general, the more abnormal the test result, the more likely the patient is to have HIT, given a certain pretest probability of having this diagnosis [14,35,36]. Also, because HIT antibodies can be transient [24], it is important to test acute serum or plasma. Platelet Activation Assays HIT antibodies can be detected based on their highly characteristic heparindependent, platelet-activating properties. The typical pattern is strong platelet activation at pharmacologic (0.1 to 0.3 U/mL), but not at high (100 U/mL), heparin levels. By testing patient serum against donor platelets ‘‘washed’’ in apyrase-containing buffer (to maintain sensitivity to adenosine diphosphate—an important potentiator of HIT antibody–induced platelet activation) and by performing the reactions at physiologic concentrations of calcium and magnesium, test sensitivity is enhanced [37,38]. Further, by selecting platelets from donors whose platelets are known to react well to HIT antibodies and by using
596
Table 3 Estimating the pretest probability of heparin-induced thrombocytopenia: the ‘‘4 T’s’’ scoring system Points (0, 1, or 2 for each of 4 categories: maximum possible score ¼ 8) Date:
2
1
0
Thrombocytopenia score ¼ _____
>50% platelet decrease to nadir 20 109/L
<30% platelet decrease or nadir <10 109/L
Timinga of platelet count decrease, thrombosis, or other sequelae (first day of heparin course ¼ day 0)Score ¼ ______
Day 5–10 onseta or 1 day (with recent heparin exposure within past 5–30 days)
Thrombosis (including adrenal infarction) or other sequelae (eg, skin lesions)Score ¼ ______
Proven new thrombosis, or skin necrosis (at injection site), or post-IV heparin bolus anaphylactoid reaction
30%–50% platelet count decrease (or >50% directly resulting from surgery) or nadir 10–19 109/L Consistent with day 5–10 decrease, but not clear (eg, missing platelet counts), or 1 day (heparin exposure within past 31–100 days), or platelet decrease after day 10 Progressive or recurrent thrombosis, or erythematous skin lesions (at injection sites), or suspected thrombosis (not proven) Possible other cause is evident
OTher cause for No explanation for platelet count thrombocytopeniaScore ¼ decrease is evident ______ Total score ¼ ______ Pretest probability score: 6–8 ¼ high; 4–5 ¼ intermediate; 0–3 ¼ low
Platelet count decrease 4 days without recent heparin exposure
None
Definite other cause is present
WARKENTIN
Changes to score can occur, based upon new information (eg, further decrease in platelets, new thrombosis, other causes for platelet decrease). The scoring system shown has undergone minor modifications from previous publications. Abbreviation: IV, intravenous. a First day of immunizing heparin exposure considered day 0; the day the platelet count begins to decrease is considered the day of onset of thrombocytopenia (it generally takes 1 to 3 more days until an arbitrary threshold that defines thrombocytopenia is passed. Usually, heparin administered at or near surgery is the most immunizing situation).
HEPARIN-INDUCED THROMBOCYTOPENIA
597
‘‘weak’’ positive controls (to ensure the platelets are sufficiently reactive in any given experiment) [37], assay performance is optimized. Washed platelet activation assays have excellent operating characteristics (ie, high sensitivity-specificity tradeoff) [14]. Suitable platelet activation end points include serotonin release and platelet aggregation. Conventional platelet aggregometry assays using normal donor platelets in citrate-anticoagulated plasma have a low sensitivity (50%–80%) for HIT [34] and are no longer used widely. Enzyme Immunoassays PF4-dependent enzyme immunoassays (EIAs) use PF4/heparin or PF4/polyvinyl sulfonate as targets. PF4-dependent EIAs are sensitive for clinical HIT (99%); hence, a negative test essentially rules out the diagnosis [34,39]. Although EIAs are more sensitive than platelet activation assays for detecting anti-PF4/heparin antibodies, this is not an advantage because more clinically insignificant antibodies are detected (potential overdiagnosis). Rapid Immunoassays Two ‘‘rapid assays’’ for HIT have been developed based upon particle agglutination [34]. A ‘‘particle gel immunoassay’’ uses PF4/heparin complexes bound to red polystyrene beads; if anti-PF4/heparin antibodies are present, these often (but not always) result in agglutination of the polystyrene beads, which is evident when centrifugation is unable to cause migration of the beads through sephacryl gel. Unlike the EIAs, a negative test does not necessarily rule out HIT. This assay is not available in the United States. Another rapid assay, the ‘‘particle immunofiltration assay,’’ is being marketed in the United States; however, its operating characteristics for HIT are not well defined. Iceberg Model The frequency of forming anti-PF4/heparin antibodies is much greater than the risk for HIT [6,7,14,19,21]. Further, only a minority of anti-PF4/heparin antibodies is detectable by EIA activate platelets and, hence, potentially can cause HIT [14,39]. The interrelationship of antibody detectability via-a`-vis clinical HIT is illustrated by the ‘‘iceberg’’ model (Fig. 3) [21]. This is a fair proxy for HIT, because perhaps only 1 in 10 antibodies identified (by EIA) in prospective studies of heparin use confer the potential for HIT (just as one tenth of an iceberg protrudes above the waterline). Among patients in whom testing is performed because of some clinical suspicion for HIT, however, about half of ‘‘positive’’ tests do not indicate HIT, for two reasons: washed platelet activation assays yield negative results and the clinical data suggest a more plausible explanation for the thrombocytopenia. Interpretation of Heparin-Induced Thrombocytopenia Antibody Testing Results of HIT antibody tests must be interpreted in the appropriate clinical context of pretest probability and the specific test result obtained. Thus, a classic clinicopathologic view of HIT requires the pretest probability to be intermediate or high and a strong positive test for HIT antibodies, preferably using
WARKENTIN
598
Waterline
HIT (with or without thrombosis)
Commercial antiWashed PF4/heparin (or platelet Anti-PF4/ PF4/polyanion) EIA activation heparin assay (IgG, IgA, IgM EIA-IgG (e.g., SRA, antibody classes HIPA test) are detected)
Subclinical seroconversion
Fig. 3. Iceberg model of HIT. Only a minority of anti-PF4/heparin antibodies cause HIT; typically, such pathogenic antibodies evince platelet-activating properties. Hence, the diagnostic specificity of (washed) platelet activation assays, such as the serotonin-release assay (SRA) or the heparin-induced platelet activation (HIPA) test, is greater than that of anti-PF4/heparin EIAs. EIAs that detect only IgG class antibodies (ie, those that are potentially platelet-activating through the platelet FccIIa receptors) have greater diagnostic specificity than EIAs that detects antibodies of the three major classes (IgG, IgA, IgM).
a washed platelet activation assay (eg, >50% serotonin release [normal < 20% release]). In 80% to 90% of cases, however, an EIA result is informative because the diagnosis can be effectively ruled out (negative EIA) or ruled in (EIA >1.0 optical density [OD] units [normal, <0.40 OD units] in an intermediate/high pretest probability situation). When a positive EIA is obtained in a low probability setting or a weak-positive EIA result (0.4–1.0 OD units) occurs in a patient with a higher score, referral to a laboratory that performs a high-quality platelet activation assay can be useful. Is Heparin-Induced Thrombocytopenia Being Under- or Overdiagnosed? Only about half of all patient sera referred for testing for HIT antibodies in which a positive EIA is obtained also yield a positive result in a washed platelet activation assay [14,36,39]. This suggests two interpretations. The first is that the negative activation assay result excludes HIT; the second is that the platelet activation assay gave a false-negative result. Recent studies indicate that the former interpretation is correct, because these patients have a low risk for thrombosis and usually have an alternate explanation for thrombocytopenia. These observations also infer that HIT is overdiagnosed. Although some cases probably are missed (usually because of atypical presentations), HIT overdiagnosis is a growing issue. TREATMENT The strong association between HIT and thrombosis and the tendency for thrombotic events to occur early in the course of HIT, mean that many— perhaps most—patients have symptomatic thrombosis at the time that HIT is recognized [1,5–9]. Further, there is evidence that isolated HIT—defined as
HEPARIN-INDUCED THROMBOCYTOPENIA
599
HIT recognized because of a platelet count decline (without clinically apparent thrombosis)—subsequently is complicated by thrombosis in one third to one half of cases [1,14,40]. This indicates that a highly probable (or confirmed) diagnosis of HIT should be treated with an alternative nonheparin anticoagulant—at least pending platelet count recovery—even in the absence of clinically apparent thrombosis [40,41]. Box 1 summarizes these and other principles of HIT management. Alternative Nonheparin Anticoagulants Three alternative nonheparin anticoagulants—danaparoid, lepirudin, and argatroban (in order of market entry)—are approved for the treatment of HIT (although approvals vary by jurisdiction) (Table 4). Current trends and expert opinion support using dosing of lepirudin that is substantially less than that indicated in the manufacturer’s package insert. Only danaparoid was evaluated in a randomized controlled trial (with superior efficacy vis-a`-vis dextran-70) [42]. Data supporting efficacy of the two direct thrombin inhibitors (DTIs) lepirudin and argatroban, compared against historical controls, are presented in Table 5 (HIT-associated thrombosis) [43–47] and Table 6 (isolated HIT) [45–48]. Two other agents (bivalirudin and fondaparinux) are rational therapies for HIT [40,49], but controlled studies are lacking.
Box 1: Principles of heparin-induced thrombocytopenia management when it is strongly suspected or has been confirmed Two Do’s Do stop all heparin (including LMWH and heparin administered as ‘‘flushes’’). Do start an alternative nonheparin anticoagulant, generally in therapeutic doses (eg, danaparoid [not available in the United States], lepirudin, or argatroban [bivalirudin and fondaparinux are rational therapies, but supporting data are sparse and anecdotal, and these therapies are ‘‘off-label’’ for HIT]. Two Don’ts Don’t give warfarin (or other coumarins) during the acute phase of HIT; if warfarin has already been started when HIT is diagnosed, vitamin K (eg, 10 mg over 30 minutes by intravenous injection) should be administered. Don’t give prophylactic platelet transfusions. Two Diagnostics Diagnosis laboratory support: test for HIT antibodies. In 80% to 90% of cases, an EIA is sufficient to rule in or to rule out HIT in the appropriate clinical context; in the remaining cases, patient serum should be referred to a laboratory experienced in performed a washed platelet activation assay (eg, serotonin release assay). Diagnosis of DVT: perform duplex ultrasonography to investigate for lower-limb DVT (the most common thrombotic sequela of HIT).
600
Table 4 Comparison of three alternative anticoagulants used to treat heparin-induced thrombocytopenia
Drug type (molecular mass)
Anticoagulant action
Immunologic features
Half-life
Lepirudin
Argatroban
Heterogeneous, polydispersed mix of anticoagulant GAGs: heparan sulfate (84%), dermatan sulfate (12%), chondroitin sulfate (4%) (6000 Da [mean]) Indirect (antithrombin-dependent) inhibition of factor Xa and thrombin (anti-Xa/anti-thrombin ratio, 22) In vitro XR for HIT Abs in minority of patients (usually not clinically significant); in high therapeutic concentrations, inhibits HIT Ab-induced platelet activation 25 h (anti-Xa action) (assumes normal renal function) Bolus: 2250 U (1500 U for b.w. <60 kg; 3000 U for b.w. 75–90 kg; 3750 U for b.w. >90 kg); followed by 400 U/h 4 h, then 300 U/h 4 h, then 150–200 U/h thereafter
65-amino acid polypeptide (6980 Da); Small-molecule arginine derivative recombinant derivative of hirudin (leech (527 Da) anticoagulant)
High affinity (Ki ¼ 0.0001 nmol/L) binding to two sites on thrombin (ie, bivalent DTI) Neither promotes nor inhibits HIT Ab binding to PF4/polyanion complexes; lepirudin is immunogenic (allergic/ anaphylactic reactions are reported)
Moderate affinity (Ki ¼ 40 nmol/L) binding to active site of thrombin (univalent DTI) Neither promotes nor inhibits HIT Ab binding to PF4/polyanion complexes; argatroban is not immunogenic
80 min (assumes normal renal function)
40–50 min (assumes normal hepatobiliary function) No bolus; 1 to 2 lg/kg/min (adjusted according to aPTT)c; marked dose reduction if hepatobiliary dysfunctiond
Bolus (only in case of life-threatening thrombosis): 0.2–0.4 mg/kg; infusion, start at 0.10 mg/kg/h (adjust to aPTT)a,b; reduce dosing for renal dysfunctionb
WARKENTIN
Dosing regimen
Danaparoid
Renal dysfunction (minor danaparoid accumulation)
Laboratory monitoring Anti-factor Xa levels (target therapeutic range: 0.5–0.8 anti-Xa U/mL)
Renal dysfunction (major lepirudin accumulation) aPTT (estimate of drug levele)
Hepatobiliary dysfunction, reduced liver perfusion (eg, low cardiac output state) (major argatroban accumulation) aPTT (estimate of drug levele)
Abbreviations: Ab, antibody; aPTT, activated partial thromboplastin time; b.w., body weight; Da, Dalton; GAGs, glycosaminoglycans; Ki, dissociation constant of an inhibitor; U, unit; XR, cross-reactivity. a Note that this dosing regimen is much lower than that described in the manufacturer’s package insert. b The initial lepirudin infusion rate should be no higher than 0.10 mg/kg/h (serum creatinine <90 lmol/L), with lower infusion rates for patients with higher serum creatinine levels (90–140 lmol/L, start at 0.05 mg/kg/h; 140–400 lmol/L, start at 0.01 mg/kg/h; >400 lmol/L, start at 0.005 mg/kg/h). aPTT monitoring should be performed at 4-hour intervals until it is apparent that steady state within the therapeutic range (1.5–2.5 times patient baseline [or mean laboratory] aPTT) has been achieved. c Although manufacturer’s package insert recommends that dosing start at 2 lg/kg/min, many clinicians start at lower dosages (eg, 1 lg/kg/min), especially in patients who are critically ill or who have cardiac failure. d For patients who have moderate or greater liver dysfunction, the starting dosage is 0.5 lg/kg/min. e The aPTT is not reliable for anticoagulant monitoring in patients who have preexisting congenital or acquired coagulopathies, overt DIC, prolonged aPTT due to ‘‘lupus anticoagulant,’’ or effects of warfarin.
HEPARIN-INDUCED THROMBOCYTOPENIA
Drug accumulation
601
602
Table 5 Summary of studies describing efficacy of the direct thrombin inhibitors (lepirudin, argatroban) as treatment for thrombosis complicating heparin-induced thrombocytopenia Study
Intervention/dosing
Lubenow et al, Lepirudin: bolus 0.4 mg/kg; 2005 [43] 0.15 mg/kg/hb Controls: variable (danaparoid, n ¼ 24; coumarin, n ¼ 21; other, n ¼ 30) Lewis et al, [45–47]
Argatroban: 2 lg/kg/min (no bolus)c Controls: variable (eg, coumarin)
New thrombosis
Composite end pointa
Major bleeds
Lepirudin: 15/214 (7%) Controls: 19/75 (25%) RR ¼ 0.28 (95% CI: 0.15–0.52) P < .001 Argatroban: 58/373 (16%) Controls: 16/46 (35%) RR ¼ 0.45 (95% CI: 0.28–0.71) P ¼ .0032f
Lepirudin: 41/214 (19%) Controls: 30/75 (40%) RR ¼ 0.48 (95% CI: 0.32–0.71) P < .001 Argatroban: 158/373 (42%) Controls: 26/46 (57%) RR ¼ 0.75 (95% CI: 0.57–0.99) P ¼ .0830g
All patients tested positive Lepirudin: 33/214 for HIT Abs (by the HIPA (15%)c Controls: 5/75 (7%) test); data shown are pooled from 3 studies RR ¼ 2.31 (95% CI: 0.94–5.71) (HAT-1, -2, -3) versus historical controls P ¼ .0723 Argatroban: Patients did not require 30/373 (8%)d positive test for HIT Controls: 1/46 (2%) Abse; data shown RR ¼ 3.70 (95% are pooled from 3 CI: 0.52–26.50) studies (Arg-911, P ¼ .2309 -915, and -915X) versus controls
Comment
WARKENTIN
P values are Fisher’s exact test, 2-tailed, unless otherwise indicated. Abbreviations: Abs, antibodies; aPTT, activated partial thromboplastin time; Arg, argatroban trial; CI, confidence interval; HAT, heparin-associated thrombocytopenia; HIPA, heparin-induced platelet activation; RCT, randomized controlled trial; RR, risk ratio; XR, cross-reactivity. a Composite end point: all-cause mortality, all-cause limb amputation, or new thrombosis (each patient counted only once). b aPTT adjusted to 1.5 to 2.5 times baseline aPTT; mean lepirudin treatment duration was 15.8 days (for 214 patients who were treated with lepirudin). c aPTT adjusted to 1.5 to 3.0 times baseline aPTT; mean argatroban treatment duration was 6.6 days (for 373 patients who were treated with argatroban). d Major bleeding, expressed per treatment day, was 0.97% for lepirudin and 1.25% for argatroban. e 65% of patients who were treated with argatroban tested positive for HIT antibodies in Arg-911 study [45] (data not available for the Arg-915/915X studies; see Ref. [46]). f P < .001 by log-rank test (time-to-event analysis for proportion of patients event-free: note that 376 patients, rather than 373 patients, were used for this analysis [47], because 3 patients initially deemed ineligible subsequently were determined to be eligible for analysis). g P ¼ .014 and P ¼ .008 for Arg-911 and Arg-915/915X studies, respectively, by log-rank analysis [45,46]; P value for pooled data not available.
Study
Intervention/dosing
New thrombosis
Composite end pointa
Major bleeds
Comment
Lubenow et al, 2004 [48]
Lepirudin: No bolus; 0.10 mg/kg/hb Con: variable (no treatment, n ¼ 31; coumarin, n ¼ 11, ASA, n ¼ 5) Argatroban: 2 lg/ kg/min (no bolus)g Con: variable
Lepirudin: 18/91 (20%) Controls: 14/47 (30%) RR ¼ 0.66 (95% CI: 0.36–1.21) P ¼ .2161e Argatroban: 94/349 (27%) Controls: 57/147 (39%) RR ¼ 0.69 (95% CI: 0.53–0.91) P ¼ .0104
Lepirudin: 13/91 (14%)c Controls: 4/47 (9%) RR ¼ 1.68 (95% CI: 0.58–4.86) P ¼ .4189f Argatroban: 15/349 (4%) Controls: 12/147 (8%) RR ¼ 0.53 (95% CI: 0.25–1.10) P ¼ .0875
All patients tested positive for HIT Abs (by the HIPA test); data shown are pooled from 3 studies (HAT-1, -2, -3) versus historical controls
Lewis et al, [45–47]
Lepirudin: 4/91 (4%) Controls: 7/47 (15%) RR ¼ 0.30 (95% CI: 0.09–0.96) P ¼ .0451d Argatroban: 24/349 (7%) Controls: 33/147 (22%) RR ¼ 0.31 (95% CI: 0.19–0.50) P < .001
Patients did not require positive test for HIT Absh; data shown are pooled from 3 studies (Arg-911, -915, and -915X) versus historical controls
HEPARIN-INDUCED THROMBOCYTOPENIA
Table 6 Summary of studies describing efficacy of the direct thrombin inhibitors (lepirudin, argatroban) as treatment for isolated heparin-induced thrombocytopenia
P values are Fisher’s exact test, 2-tailed, unless otherwise indicated. Abbreviations: Abs, antibodies; aPTT, activated partial thromboplastin time; Arg, argatroban; ASA; acetylsalicylic acid (aspirin); CI, confidence interval; HAT, ; HIPA, heparin-induced platelet activation; RCT, randomized controlled trial; RR, risk ratio; XR, cross-reactivity. a Composite end point: all-cause mortality, all-cause limb amputation, or new thrombosis (each patient counted only once). b aPTT adjusted to 1.5 to 2.5 times baseline aPTT; also, mean lepirudin treatment duration was 13.9 days (pooled data). c Major bleeding, expressed per treatment day, was 1.03% for lepirudin and 0.84% for argatroban. d P ¼ .02 by log-rank test (time-to-event analysis). e P ¼ .0281 by log-rank test (time-to-event analysis). f P ¼ .5419 by log-rank test (time-to-event analysis). g aPTT adjusted to 1.5 to 3.0 times baseline aPTT; also, mean argatroban treatment duration was 5.2 days (pooled data). h 50% of patients who were treated with argatroban tested positive for HIT antibodies in the Arg-911 study (data not available for Arg-915 study).
603
604
WARKENTIN
Warfarin is Contraindicated During Acute Heparin-Induced Thrombocytopenia Acute HIT is a risk factor for warfarin (coumarin) necrosis, which can manifest as venous limb gangrene or classic skin necrosis [4,5,50]. The pathogenesis is microthrombosis due to depletion of the vitamin K–dependent natural anticoagulant, protein C, in the setting of increased thrombin generation from HIT (disturbed procoagulant–anticoagulant balance) [4]. Venous limb gangrene is characterized by an underlying hypercoagulability disorder such as HIT; acral (distal extremity) necrosis in a limb affected by DVT; and a supratherapeutic INR that is usually greater than 3.5 (surrogate marker for severe depletion in protein C). In contrast, skin necrosis is characterized by involvement of skin and subcutaneous tissues at central (nonacral) sites (eg, breast, abdomen, thigh, calf). Coumarins also prolong the activated partial thromboplastin time (aPTT), thus predisposing to underdosing of DTI therapy (because DTIs usually are monitored by the aPTT) [50]. The risk for coumarin necrosis in HIT is about 5% to 10% [4,40], which is much greater than the risk in populations that do not have HIT (0.01%). Vitamin K therapy is advised if a diagnosis of HIT is made when warfarin has already been started [40,49,50]. PREVENTION OF HEPARIN-INDUCED THROMBOCYTOPENIA AND ITS SEQUELAE Platelet count monitoring for HIT is rational in many clinical settings, such as higher risk situations (eg, postoperative UFH thromboprophylaxis beyond 4 days). Here, at least every-other-day platelet count monitoring until day 14 or stopping heparin (whichever comes first) is suggested [40]. Similar monitoring during the use of therapeutic-dose UFH also is appropriate. Less frequent monitoring (eg, twice or thrice weekly) is suggested when the risk for HIT is lower (eg, UFH prophylaxis in medical patients or LMWH prophylaxis in postoperative patients). In settings in which the risk is low (eg, LMWH prophylaxis in medical patients or during pregnancy), routine platelet count monitoring for HIT is not recommended. Regardless of the clinical situation, an urgent platelet count usually is appropriate when a patient who is receiving heparin develops symptomatic thrombosis [1,2,8,30,40]. The reduced risk for HIT with LMWH (or fondaparinux) prophylaxis, compared with UFH, suggests that preferring these former agents could reduce the risk for HIT in certain situations; however, it is uncertain whether the risk is reduced in internal medicine patients who are given LMWH rather than UFH. This is either because the absolute risk for HIT is sufficiently low that a true difference is difficult to prove or no difference exists. REPEAT HEPARIN EXPOSURE DESPITE PREVIOUS HEPARIN-INDUCED THROMBOCYTOPENIA For patients with a history of HIT and who have an important indication for heparin (eg, cardiac or vascular surgery), it is recommended to use heparin, provided that antibodies are no longer detectable or (in emergencies) that
HEPARIN-INDUCED THROMBOCYTOPENIA
605
sufficient time has elapsed from the episode of HIT (>2–3 months) that clinically significant levels of antibodies are unlikely to be present. The basis of this recommendation is that it takes a minimum of 5 days to form HIT antibodies— even in patients with previous HIT—thus permitting safe intraoperative exposure. Further, there is no evidence that a patient who had HIT previously is more likely to form strong levels of HIT antibodies upon reexposure. In situations in which an alternative anticoagulant agent is available and safe (eg, postoperative prophylaxis with danaparoid, fondaparinux, or warfarin), however, this is preferred. References [1] Warkentin TE. Heparin-induced thrombocytopenia: pathogenesis and management. Br J Haematol 2003;121(4):535–55. [2] Arepally GM, Ortel TL. Clinical practice. Heparin-induced thrombocytopenia. N Engl J Med 2006;355(8):809–17. [3] Warkentin TE. Management of heparin-induced thrombocytopenia: a critical comparison of lepirudin and argatroban. Thromb Res 2003;110(2–3):73–82. [4] Warkentin TE, Elavathil LJ, Hayward CPM, et al. The pathogenesis of venous limb gangrene associated with heparin-induced thrombocytopenia. Ann Intern Med 1997;127(9): 804–12. [5] Warkentin TE. Clinical picture of heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 21–66. [6] Warkentin TE, Levine MN, Hirsh J, et al. Heparin-induced thrombocytopenia in patients treated with low-molecular-weight heparin or unfractionated heparin. N Engl J Med 1995;332(20):1330–5. [7] Warkentin TE, Roberts RS, Hirsh J, et al. An improved definition of immune heparin-induced thrombocytopenia in postoperative orthopedic patients. Arch Intern Med 2003;163(20): 2518–24. [8] Greinacher A, Farner B, Kroll H, et al. Clinical features of heparin-induced thrombocytopenia including risk factors for thrombosis. A retrospective analysis of 408 patients. Thromb Haemost 2005;94(1):132–5. [9] Warkentin TE, Kelton JG. A 14-year study of heparin-induced thrombocytopenia. Am J Med 1996;101(5):502–7. [10] Warkentin TE. Pseudo-heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 261–82. [11] Warkentin TE. Venous limb gangrene during warfarin treatment of cancer-associated deep venous thrombosis. Ann Intern Med 2001;135(8 Pt 1):589–93. [12] Warkentin TE. An overview of the heparin-induced thrombocytopenia syndrome. Semin Thromb Hemost 2004;30(3):273–83. [13] Kelton JG, Sheridan D, Santos A, et al. Heparin-induced thrombocytopenia: laboratory studies. Blood 1988;72(3):925–30. [14] Warkentin TE, Sheppard JI, Moore JC, et al. Laboratory testing for the antibodies that cause heparin-induced thrombocytopenia: how much class do we need? J Lab Clin Med 2005; 146(6):341–6. [15] Greinacher A, Po ¨ tzsch B, Amiral J, et al. Heparin-associated thrombocytopenia: isolation of the antibody and characterization of a multimolecular PF4-heparin complex as the major antigen. Thromb Haemost 1994;71(2):247–51. [16] Suh JS, Aster RH, Visentin GP. Antibodies from patients with heparin-induced thrombocytopenia/thrombosis recognize different epitopes on heparin:platelet factor 4. Blood 1998;91(3):916–22.
606
WARKENTIN
[17] Greinacher A, Gopinadhan M, Guenther JU, et al. Close approximation of two platelet factor 4 tetramers by charge neutralization forms the antigens recognized by HIT antibodies. Arterioscler Thromb Vasc Biol 2006;26(10):2386–93. [18] Warkentin TE, Sheppard JI, Siguoin CS, et al. Gender imbalance and risk factor interactions in heparin-induced thrombocytopenia. Blood 2006;108(9):2937–41. [19] Warkentin TE, Cook RJ, Marder VJ, et al. Anti-platelet factor 4/heparin antibodies in orthopedic surgery patients receiving antithrombotic prophylaxis with fondaparinux or enoxaparin. Blood 2005;106(12):3791–6. [20] Warkentin TE, Eikelboom JW. Who is (still) getting HIT? Chest 2007;131:1620–2. [21] Lee DH, Warkentin TE. Frequency of heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-Induced Thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 67–116. [22] Fausett MB, Vogtlander M, Lee RM, et al. Heparin-induced thrombocytopenia is rare in pregnancy. Am J Obstet Gynecol 2001;185(1):148–52. [23] Warkentin TE. Drug-induced immune-mediated thrombocytopenia: from purpura to thrombosis. N Engl J Med 2007;356(9):891–3. [24] Warkentin TE, Kelton JG. Temporal aspects of heparin-induced thrombocytopenia. N Engl J Med 2001;344(17):1286–92. [25] Warkentin TE, Kelton JG. Delayed-onset heparin-induced thrombocytopenia and thrombosis. Ann Intern Med 2001;135(7):502–6. [26] Rauova L, Zhai L, Kowalska MA, et al. Role of platelet surface PF4 antigenic complexes in heparin-induced thrombocytopenia pathogenesis: diagnostic and therapeutic implications. Blood 2006;107(6):2346–53. [27] Warkentin TE, Jay RM, Makris M, et al. Platelet-activating anti-platelet factor 4/polyanion antibodies without preceding heparin therapy: a transient autoimmune disorder resembling heparin-induced thrombocytopenia (‘‘spontaneous HIT’’) [abstract]. Blood 2006;108(11): 311a–2a. [28] Warkentin TE. Think of HIT. Hematology Am Soc Hematol Educ Program 2006;408–14. [29] Greinacher A, Eichler P, Lietz T, et al. Replacement of unfractionated heparin by lowmolecular-weight heparin for postorthopedic surgery antithrombotic prophylaxis lowers the overall risk of symptomatic thrombosis because of a lower frequency of heparin-induced thrombocytopenia. Blood 2005;106(8):2921–2. [30] Warkentin TE. Think of HIT when thrombosis follows heparin. Chest 2006;130(3): 631–2. [31] Warkentin TE. Heparin-induced skin lesions. Br J Haematol 1996;92(2):494–7. [32] Warkentin TE, Hirte HW, Anderson DR, et al. Transient global amnesia following intravenous heparin bolus therapy is caused by heparin-induced thrombocytopenia. Thromb Haemost 1994;97(5):489–91. [33] Lo GK, Juhl D, Warkentin TE, et al. Evaluation of pretest clinical score (4 T’s) for the diagnosis of heparin-induced thrombocytopenia in two clinical settings. J Thromb Haemost 2006; 4(4):759–65. [34] Warkentin TE, Sheppard JI. Testing for heparin-induced thrombocytopenia antibodies. Transfus Med Rev 2006;20(4):259–72. [35] Warkentin TE. New approaches to the diagnosis of heparin-induced thrombocytopenia. Chest 2005;127(Suppl 2):35S–45S. [36] Lo GK, Sigouin CS, Warkentin TE. What is the potential for overdiagnosis of heparininduced thrombocytopenia? Am J Hematol, in press. [37] Warkentin TE, Hayward CPM, Smith CA, et al. Determinants of donor platelet variability when testing for heparin-induced thrombocytopenia. J Lab Clin Med 1992;120(3):371–9. [38] Greinacher A, Michels I, Kiefel V, et al. A rapid and sensitive test for diagnosing heparininduced thrombocytopenia. Thromb Haemost 1991;66(6):734–6. [39] Greinacher A, Juhl D, Strobel U, et al. Heparin-induced thrombocytopenia: a prospective study on the incidence, platelet-activating capacity and clinical significance of
HEPARIN-INDUCED THROMBOCYTOPENIA
[40]
[41] [42]
[43]
[44]
[45] [46] [47]
[48]
[49]
[50]
607
anti-PF4/heparin antibodies of the IgG, IgM, and IgA classes. J Thromb Haemost 2007; May 7 [epub ahead of print]. Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia: recognition, treatment, and prevention. The seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 2004;126(Suppl 3):311S–37S. Warkentin TE. Heparin-induced thrombocytopenia: diagnosis and management. Circulation 2004;110(18):e454–8. Chong BH, Gallus AS, Cade JF, et al. Prospective randomized open-label comparison of danaparoid with dextran 70 in the treatment of heparin-induced thrombocytopenia with thrombosis: a clinical outcome study. Thromb Haemost 2001;86(5):1170–5. Lubenow N, Eichler P, Lietz T, et al. Lepirudin in patients with heparin-induced thrombocytopenia—results of the third prospective study (HAT-3) and a combined analysis of HAT-1, HAT-2, and HAT-3. J Thromb Haemost 2005;3(11):2428–36. Greinacher A, Eichler P, Lubenow N, et al. Heparin-induced thrombocytopenia with thromboembolic complications: meta-analysis of two prospective trials to assess the value of parenteral treatment with lepirudin and its therapeutic aPTT range. Blood 2000;96(3): 846–51. Lewis BE, Wallis DE, Berkowitz SD, et al. Argatroban anticoagulant therapy in patients with heparin-induced thrombocytopenia. Circulation 2001;103(14):1838–43. Lewis BE, Wallis DE, Leya F, et al. Argatroban anticoagulation in patients with heparininduced thrombocytopenia. Arch Intern Med 2003;163(15):1849–56. Lewis BE, Wallis DE, Hursting MJ, et al. Effects of argatroban therapy, demographic variables, and platelet count on thrombotic risks in heparin-induced thrombocytopenia. Chest 2006;129(6):1407–16. Lubenow N, Eichler P, Lietz T, et al. Lepirudin for prophylaxis of thrombosis in patients with acute isolated heparin-induced thrombocytopenia: an analysis of 3 prospective studies. Blood 2004;104(10):3072–7. Greinacher A, Warkentin TE. Treatment of heparin-induced thrombocytopenia: an overview. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 283–317. Warkentin TE. Should vitamin K be administered when HIT is diagnosed after administration of coumarin? J Thromb Haemost 2006;4:894–6.
HEMATOLOGY/ONCOLOGY CLINICS OF NORTH AMERICA
Heparin-Induced Thrombocytopenia Theodore E. Warkentin, MDa,b,c,d,* a
Department of Pathology and Molecular Medicine, Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario, Canada b Department of Medicine, Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario, Canada c Transfusion Medicine, Hamilton Regional Laboratory Medicine Program, Hamilton, Ontario, Canada d Service of Clinical Hematology, Hamilton Health Sciences, Hamilton General Hospital, Hamilton, Ontario, Canada
H
eparin-induced thrombocytopenia (HIT) is an immune-mediated adverse drug reaction that is caused by heparin-dependent, plateletactivating IgG antibodies that recognize complexes of platelet factor 4 (PF4) bound to heparin [1,2]. HIT is a strong risk factor for venous and arterial thrombosis (odds ratio, 20 to 40) [3]. HIT often leads to thrombosis in large veins and arteries; however, HIT also is a risk factor for microvascular thrombosis, especially with warfarin treatment of HIT-associated deep-vein thrombosis (DVT) or overt disseminated intravascular coagulation (DIC) [4,5]. Thrombosis usually occurs in association with large platelet count declines [6,7], consistent with in vivo platelet activation. A central role for in vivo thrombin generation [4] provides a rationale for the use of anticoagulants that inhibit thrombin or its activity. DEFINITION AND TERMINOLOGY HIT can be defined as any clinical event (or events) best explained by plateletactivating anti-PF4/heparin antibodies (‘‘HIT antibodies’’) in a patient who is receiving, or who recently received, heparin [1]. Thrombocytopenia is the most common event in HIT, and it is observed in at least 90% of patients, depending on how thrombocytopenia is defined. Most patients who have HIT develop thrombosis [1,5–9].
Studies cited [3,4,6,7,9,11,12,14,17–19,23–25,27,28,30–37,50] were supported by the Heart and Stroke Foundation of Ontario (operating grants #A2449, #T2967, B3763, #T4502, and #T5207).
*Room 1-180A, Hamilton Regional Laboratory Medicine Program, Hamilton Health Sciences, Hamilton General Hospital, 237 Barton Street East, Hamilton, Ontario L8L 2X2, Canada. E-mail address: [email protected] 0889-8588/07/$ – see front matter doi:10.1016/j.hoc.2007.06.004
Ó 2007 Elsevier Inc. All rights reserved. hemonc.theclinics.com
590
WARKENTIN
HIT is a clinicopathologic syndrome [1]; thus, the diagnosis requires one or more clinical events (eg, thrombocytopenia, thrombosis) bearing a temporal association with pathologic HIT antibody formation. Thus, a patient suspected to have HIT but in whom antibodies cannot be detected does not have this diagnosis. Some disorders mimic HIT (‘‘pseudo-HIT’’), pointing to the importance of laboratory testing for HIT antibodies [10,11]. PATHOGENESIS Fig. 1 summarizes HIT pathogenesis [12]. The central concept is formation of heparin-dependent IgG that activates platelets by way of their FccIIa receptors [13]. IgG and IgM class antibodies are not pathogenic [14] because they cannot activate platelets by way of Fcc (IgG) receptors. The target antigen is a complex between (anionic) heparin and (cationic) PF4 [15], a tetrameric member of the CXC subfamily of chemokines. HIT antibodies recognize conformationally altered sites on PF4 resulting from its binding to heparin [16] or because of close approximation of PF4 tetramers by heparin charge neutralization [17]. Low molecular weight heparin (LMWH) is less likely to trigger antibodies and HIT compared with unfractionated heparin (UFH) [6,7], particularly in women receiving surgical thromboprophylaxis [18]. The pentasaccharide anticoagulant, fondaparinux, although similarly immunogenic as LMWH, does not form well the antigens on PF4 [19], suggesting that it has an even lower (perhaps negligible) risk for causing HIT. Factors leading to thrombosis in HIT include [12] the platelet-activating nature of HIT antibodies (including formation of procoagulant, platelet-derived microparticles), as well as ‘‘pancellular’’ activation (endothelium, monocytes) and neutralization of the anticoagulant effect of heparin by PF4 released from activated platelets. FREQUENCY Four factors influence the frequency of HIT: duration of heparin use, type of heparin, type of patient population, and patient gender (Table 1) [5–7,18,20,21]. The ‘‘typical’’ patient who has HIT is a woman receiving postsurgery thromboprophylaxis with UFH for more than 1 week (risk, 1%–5%). Despite the greater risk in women, HIT is rare in pregnancy [22]. HIT occurs in about 1% of patients receiving UFH for treatment of venous thrombosis and in about 0.1% to 0.5% of postoperative patients receiving LMWH prophylaxis [21]. CLINICAL PICTURE HIT is a distinct syndrome compared with other drugs that cause immune thrombocytopenia (eg, quinine, vancomycin, glycoprotein IIb/IIIa antagonists) (Fig. 2) [23]. Although these other agents produce severe thrombocytopenia and mucocutaneous hemorrhage, HIT usually results in a moderate thrombocytopenia, but with a greatly increased risk for thrombosis. Also, although classic drug-induced immune thrombocytopenia causes dramatic bleeding,
HEPARIN-INDUCED THROMBOCYTOPENIA
591
Fig. 1. Pathogenesis of HIT. The figure illustrates two potential explanations for thrombosis in HIT. Activation of platelets (plt) by anti-PF4/heparin IgG antibodies (‘‘HIT antibodies’’), leading to formation of procoagulant, platelet-derived microparticles and neutralization of heparin by PF4 released from activated platelets, results in a marked increase in thrombin generation (‘‘hypercoagulability state’’) that is characterized by increased risks for venous or arterial thrombosis and coumarin-associated microthrombosis (venous limb gangrene, skin necrosis). It also is possible that unique mechanisms operative in HIT explain unusual thromboses, such as arterial ‘‘white clots.’’ For example, HIT antibodies have been shown to activate endothelium and monocytes (leading to cell surface tissue factor expression, although this stimulation may be largely ‘‘indirect’’ through poorly defined mechanisms involving platelet activation and, possibly, formation of platelet-derived microparticles). Further, aggregates of platelets and polymorphonuclear (PMN) leukocytes have been described in HIT. The extent to which the cooperative interactions between platelets, platelet-derived microparticles, PMN leukocytes, monocytes, and endothelium lead to arterial (or venous) thrombotic events in HIT, either in large or small vessels, remains unclear. (From Warkentin TE. An overview of the heparin-induced thrombocytopenia syndrome. Semin Thromb Hemost 2004;30(3):275; with permission.)
WARKENTIN
592
Table 1 Risk factors for immune heparin-induced thrombocytopenia Risk factor
Odds ratio
Duration of heparin use (>1 week versus <1 day)
20–100
Type of heparin (UFH > LMWH > fondaparinux)
10–15
Type of patient (surgery > medical > pregnancy)
3–4
Gender (female > male)
1.5–2.0
Comment OR estimated as follows: risk for HIT postcardiac surgery (with UFH prophylaxis for >1 week) 2% versus risk for ‘‘delayedonset HIT’’ (without prophylaxis) 0.02–0.1% Difference in risk for HIT between heparin types is established for postsurgical thromboprophylaxis (UFH versus LMWH) and is more pronounced in women. ORs shown are for UFH versus LMWH (women, post-surgical prophylaxis) [18] Highest reported frequencies of HIT are in postsurgical thromboprophylaxis (OR shown is for surgery versus medical) [18] Difference in risk for HIT between women and men has only been established for UFH thromboprophylaxis [18]
Abbreviation: OR, odds ratio.
fatalities are rare, whereas in HIT, fatal or long-term sequelae of thrombosis are not uncommon [23]. Thrombocytopenia The median platelet count nadir in HIT is about 60 109/L [5,9,23]. Most patients evince a 50% or greater decrease in the platelet count [7]. In postoperative patients, the appropriate ‘‘baseline’’ platelet count is not the preoperative platelet count, but rather the highest postoperative platelet count preceding the HIT-associated platelet count decrease [5,7]. Timing HIT is a transient immune response that occurs most often when heparin is given intra- or perioperatively. ‘‘Typical-onset’’ HIT (70% of patients) manifests as a platelet count decrease that begins 5 to 10 days after starting a course of heparin (first day of heparin use ¼ day 0) [24], with a median of 2 days required to reach a threshold defining thrombocytopenia [5]. ‘‘Rapid-onset’’ HIT is the presenting feature in about one quarter of patients and is defined as a decrease in platelet count within 24 hours of administering a course of heparin [24]. This presentation occurs because the patient already has circulating HIT antibodies, a result of exposure to heparin within the past few weeks or months.
HEPARIN-INDUCED THROMBOCYTOPENIA
No. of Patients (arbitrary units, increasing from bottom to top)
Drug-induced immune thrombocytopenia 9
60 x 10 /L (median)
Bleeding
5
10
Heparin-induced thrombocytopenia 9
<10 x 10 /L (median)
3
593
Thrombosis
20
50
100
200
500
1000
Nadir Platelet Counts (x10-9/L) Shown on a Log10 Scale Fig. 2. Clinical picture of HIT versus classic drug-induced immune-mediated thrombocytopenia. ‘‘Classic’’ drug-induced immune-mediated thrombocytopenia (eg, caused by quinine, vancomycin, or glycoprotein IIb/IIIa antagonists, among other drugs) typically produces severe thrombocytopenia (median platelet count nadir, about 10 109/L) and associated mucocutaneous bleeding. In contrast, HIT typically results in mild-to-moderate thrombocytopenia (median platelet count nadir, 60 109/L) and associated venous or arterial thrombosis. Note that the heights of the two peaks are not drawn to scale, because HIT is much more common than all other causes of drug-induced immune-mediated thrombocytopenia combined. (From Warkentin TE. Drug-induced immune-mediated thrombocytopenia: from purpura to thrombosis. N Engl J Med 2007;356(9):891; with permission. Copyright Ó 2007, Massachusetts Medical Society.)
‘‘Delayed-onset’’ HIT refers to when the platelet count begins to decrease after all heparin has been stopped [25]. This results from unusually high levels of antibodies that can activate platelets even in the absence of pharmacologic heparin, perhaps because the antibodies recognize PF4 bound to plateletassociated chondroitin sulfate [26]. ‘‘Spontaneous’’ HIT is a rare disorder in which a patient develops a clinical profile suggestive of HIT and in whom high levels of HIT antibodies can be detected, even though no heparin has been administered previously [27]. Thrombosis and Other Sequelae Table 2 presents the complications of HIT [5]. The strong association between HIT and thrombosis means that this diagnosis should be considered in patients who develop symptomatic thrombosis during, or within several days after stopping, heparin use [28]. In some studies of postsurgical thromboprophylaxis with UFH, the development of symptomatic thrombosis indicated about a 50% risk for HIT being present [29,30]. Venous thrombosis is the most common sequela of HIT [1,5–9,21]. Venous limb ischemia usually indicates warfarin toxicity [4,5]. Adrenal hemorrhage is a sign of adrenal vein thrombosis, with secondary hemorrhage [28]; when
594
Table 2 Clinical sequelae of heparin-induce thrombocytopenia Comment
Venous thromboembolism
DVT (50%) and pulmonary embolism (25%) are the two most common sequelae of HIT, especially in postoperative patients Limb artery thrombosis is most common (10%–15%) followed by thrombotic stroke (5%–10%), myocardial infarction (3%–5%), and other (eg, mesenteric artery thrombosis, spinal artery thrombosis) Besides cerebral artery thrombosis, other causes of stroke include cardiac embolization and cerebral venous (dural sinus) thrombosis HIT is a strong risk factor for coumarin necrosis, especially venous limb gangrene (acral limb necrosis usually in the setting of DVT), but also ‘‘classic’’ skin necrosis (subdermal and dermal necrosis, usually in nonacral sites [eg, breast, abdominal wall, thigh, calf, or forearm]); characterized by microvascular thrombosis Associated with thrombosis of adrenal veins; unilateral adrenal hemorrhage presents as abdominal or flank pain; when adrenal hemorrhage is bilateral, adrenal crisis is likely to occur Necrotizing skin lesions at heparin injection sites are specific for HIT; erythematous plaques at injection sites also can indicate HIT; some patients with skin lesions do not evince thrombocytopenia One or more signs or symptoms that begin 5–30 min after IV heparin injection (or sc LMWH injection): cardiac (tachycardia, chest pain, hypertension, cardiac arrest), respiratory (dyspnea, tachypnea, chest pain, respiratory arrest), inflammatory (fever, chills, rigors, flushing), neurologic (pounding headache, transient global amnesia syndrome), gastrointestinal (diarrhea); these features are not that of anaphylaxis About 10%–20% of patients who have HIT have overt (decompensated) DIC (eg, hypofibrinogenemia, increased INR, positive protamine sulfate paracoagulation assay, microangiopathic hemolysis, or circulating normoblasts [nucleated red cells])
Arterial thrombosis Thrombotic stroke Coumarin necrosis
Adrenal hemorrhage Skin lesions at heparin injection site Anaphylactoid reaction
Overt DIC
Abbreviations: IV, intravenous; sc, subcutaneous.
WARKENTIN
Sequela
HEPARIN-INDUCED THROMBOCYTOPENIA
595
bilateral, death can result from adrenal failure. Arterial thrombosis most commonly affects large lower-limb arteries, with thrombotic stroke and myocardial infarction seen less commonly [5,9]. Limb ischemia can be classified based on whether pulses are palpable (or Doppler identifiable). Absent pulses suggest that limb ischemia is caused by artery occlusion, whereas pulses despite acral (distal extremity) ischemia suggest microvascular occlusion. The hallmark of coumarin-associated venous limb ischemia is an elevated international normalized ratio (INR; usually > 3.5), which is a surrogate marker for severely depleted protein C (in parallel with reduced factor VII) [4]. Rarely, limb ischemia occurs as a result of severe DIC alone [5]. Classic (nonacral) skin necrosis also has been reported in HIT [4,5]. Table 2 also summarizes other miscellaneous sequelae of HIT, including necrotizing and nonnecrotizing skin lesions at heparin injection sites [31], acute anaphylactoid reactions following intravenous heparin bolus administration [5,32], and overt DIC [5,25]. CLINICAL SCORING SYSTEM FOR HEPARIN-INDUCED THROMBOCYTOPENIA In evaluating a patient for possible HIT, a clinical scoring system can help [33]. One system, the 4 Ts (Table 3), evaluates Thrombocytopenia, its Timing, the presence of Thrombosis (or other sequelae of HIT), and whether oTher plausible explanations for thrombocytopenia or thrombosis are present. A low score (3 points) makes HIT unlikely (< 2%). In some settings, a high score predicts a high likelihood of HIT [33]. LABORATORY DETECTION OF HEPARIN-INDUCED THROMBOCYTOPENIA ANTIBODIES Detectability of HIT antibodies by the laboratory is a sine qua non of HIT diagnosis. There are two complementary ways to detect antibodies: platelet activation assays and PF4-dependent immunoassays [34]. In general, the more abnormal the test result, the more likely the patient is to have HIT, given a certain pretest probability of having this diagnosis [14,35,36]. Also, because HIT antibodies can be transient [24], it is important to test acute serum or plasma. Platelet Activation Assays HIT antibodies can be detected based on their highly characteristic heparindependent, platelet-activating properties. The typical pattern is strong platelet activation at pharmacologic (0.1 to 0.3 U/mL), but not at high (100 U/mL), heparin levels. By testing patient serum against donor platelets ‘‘washed’’ in apyrase-containing buffer (to maintain sensitivity to adenosine diphosphate—an important potentiator of HIT antibody–induced platelet activation) and by performing the reactions at physiologic concentrations of calcium and magnesium, test sensitivity is enhanced [37,38]. Further, by selecting platelets from donors whose platelets are known to react well to HIT antibodies and by using
596
Table 3 Estimating the pretest probability of heparin-induced thrombocytopenia: the ‘‘4 T’s’’ scoring system Points (0, 1, or 2 for each of 4 categories: maximum possible score ¼ 8) Date:
2
1
0
Thrombocytopenia score ¼ _____
>50% platelet decrease to nadir 20 109/L
<30% platelet decrease or nadir <10 109/L
Timinga of platelet count decrease, thrombosis, or other sequelae (first day of heparin course ¼ day 0)Score ¼ ______
Day 5–10 onseta or 1 day (with recent heparin exposure within past 5–30 days)
Thrombosis (including adrenal infarction) or other sequelae (eg, skin lesions)Score ¼ ______
Proven new thrombosis, or skin necrosis (at injection site), or post-IV heparin bolus anaphylactoid reaction
30%–50% platelet count decrease (or >50% directly resulting from surgery) or nadir 10–19 109/L Consistent with day 5–10 decrease, but not clear (eg, missing platelet counts), or 1 day (heparin exposure within past 31–100 days), or platelet decrease after day 10 Progressive or recurrent thrombosis, or erythematous skin lesions (at injection sites), or suspected thrombosis (not proven) Possible other cause is evident
OTher cause for No explanation for platelet count thrombocytopeniaScore ¼ decrease is evident ______ Total score ¼ ______ Pretest probability score: 6–8 ¼ high; 4–5 ¼ intermediate; 0–3 ¼ low
Platelet count decrease 4 days without recent heparin exposure
None
Definite other cause is present
WARKENTIN
Changes to score can occur, based upon new information (eg, further decrease in platelets, new thrombosis, other causes for platelet decrease). The scoring system shown has undergone minor modifications from previous publications. Abbreviation: IV, intravenous. a First day of immunizing heparin exposure considered day 0; the day the platelet count begins to decrease is considered the day of onset of thrombocytopenia (it generally takes 1 to 3 more days until an arbitrary threshold that defines thrombocytopenia is passed. Usually, heparin administered at or near surgery is the most immunizing situation).
HEPARIN-INDUCED THROMBOCYTOPENIA
597
‘‘weak’’ positive controls (to ensure the platelets are sufficiently reactive in any given experiment) [37], assay performance is optimized. Washed platelet activation assays have excellent operating characteristics (ie, high sensitivity-specificity tradeoff) [14]. Suitable platelet activation end points include serotonin release and platelet aggregation. Conventional platelet aggregometry assays using normal donor platelets in citrate-anticoagulated plasma have a low sensitivity (50%–80%) for HIT [34] and are no longer used widely. Enzyme Immunoassays PF4-dependent enzyme immunoassays (EIAs) use PF4/heparin or PF4/polyvinyl sulfonate as targets. PF4-dependent EIAs are sensitive for clinical HIT (99%); hence, a negative test essentially rules out the diagnosis [34,39]. Although EIAs are more sensitive than platelet activation assays for detecting anti-PF4/heparin antibodies, this is not an advantage because more clinically insignificant antibodies are detected (potential overdiagnosis). Rapid Immunoassays Two ‘‘rapid assays’’ for HIT have been developed based upon particle agglutination [34]. A ‘‘particle gel immunoassay’’ uses PF4/heparin complexes bound to red polystyrene beads; if anti-PF4/heparin antibodies are present, these often (but not always) result in agglutination of the polystyrene beads, which is evident when centrifugation is unable to cause migration of the beads through sephacryl gel. Unlike the EIAs, a negative test does not necessarily rule out HIT. This assay is not available in the United States. Another rapid assay, the ‘‘particle immunofiltration assay,’’ is being marketed in the United States; however, its operating characteristics for HIT are not well defined. Iceberg Model The frequency of forming anti-PF4/heparin antibodies is much greater than the risk for HIT [6,7,14,19,21]. Further, only a minority of anti-PF4/heparin antibodies is detectable by EIA activate platelets and, hence, potentially can cause HIT [14,39]. The interrelationship of antibody detectability via-a`-vis clinical HIT is illustrated by the ‘‘iceberg’’ model (Fig. 3) [21]. This is a fair proxy for HIT, because perhaps only 1 in 10 antibodies identified (by EIA) in prospective studies of heparin use confer the potential for HIT (just as one tenth of an iceberg protrudes above the waterline). Among patients in whom testing is performed because of some clinical suspicion for HIT, however, about half of ‘‘positive’’ tests do not indicate HIT, for two reasons: washed platelet activation assays yield negative results and the clinical data suggest a more plausible explanation for the thrombocytopenia. Interpretation of Heparin-Induced Thrombocytopenia Antibody Testing Results of HIT antibody tests must be interpreted in the appropriate clinical context of pretest probability and the specific test result obtained. Thus, a classic clinicopathologic view of HIT requires the pretest probability to be intermediate or high and a strong positive test for HIT antibodies, preferably using
WARKENTIN
598
Waterline
HIT (with or without thrombosis)
Commercial antiWashed PF4/heparin (or platelet Anti-PF4/ PF4/polyanion) EIA activation heparin assay (IgG, IgA, IgM EIA-IgG (e.g., SRA, antibody classes HIPA test) are detected)
Subclinical seroconversion
Fig. 3. Iceberg model of HIT. Only a minority of anti-PF4/heparin antibodies cause HIT; typically, such pathogenic antibodies evince platelet-activating properties. Hence, the diagnostic specificity of (washed) platelet activation assays, such as the serotonin-release assay (SRA) or the heparin-induced platelet activation (HIPA) test, is greater than that of anti-PF4/heparin EIAs. EIAs that detect only IgG class antibodies (ie, those that are potentially platelet-activating through the platelet FccIIa receptors) have greater diagnostic specificity than EIAs that detects antibodies of the three major classes (IgG, IgA, IgM).
a washed platelet activation assay (eg, >50% serotonin release [normal < 20% release]). In 80% to 90% of cases, however, an EIA result is informative because the diagnosis can be effectively ruled out (negative EIA) or ruled in (EIA >1.0 optical density [OD] units [normal, <0.40 OD units] in an intermediate/high pretest probability situation). When a positive EIA is obtained in a low probability setting or a weak-positive EIA result (0.4–1.0 OD units) occurs in a patient with a higher score, referral to a laboratory that performs a high-quality platelet activation assay can be useful. Is Heparin-Induced Thrombocytopenia Being Under- or Overdiagnosed? Only about half of all patient sera referred for testing for HIT antibodies in which a positive EIA is obtained also yield a positive result in a washed platelet activation assay [14,36,39]. This suggests two interpretations. The first is that the negative activation assay result excludes HIT; the second is that the platelet activation assay gave a false-negative result. Recent studies indicate that the former interpretation is correct, because these patients have a low risk for thrombosis and usually have an alternate explanation for thrombocytopenia. These observations also infer that HIT is overdiagnosed. Although some cases probably are missed (usually because of atypical presentations), HIT overdiagnosis is a growing issue. TREATMENT The strong association between HIT and thrombosis and the tendency for thrombotic events to occur early in the course of HIT, mean that many— perhaps most—patients have symptomatic thrombosis at the time that HIT is recognized [1,5–9]. Further, there is evidence that isolated HIT—defined as
HEPARIN-INDUCED THROMBOCYTOPENIA
599
HIT recognized because of a platelet count decline (without clinically apparent thrombosis)—subsequently is complicated by thrombosis in one third to one half of cases [1,14,40]. This indicates that a highly probable (or confirmed) diagnosis of HIT should be treated with an alternative nonheparin anticoagulant—at least pending platelet count recovery—even in the absence of clinically apparent thrombosis [40,41]. Box 1 summarizes these and other principles of HIT management. Alternative Nonheparin Anticoagulants Three alternative nonheparin anticoagulants—danaparoid, lepirudin, and argatroban (in order of market entry)—are approved for the treatment of HIT (although approvals vary by jurisdiction) (Table 4). Current trends and expert opinion support using dosing of lepirudin that is substantially less than that indicated in the manufacturer’s package insert. Only danaparoid was evaluated in a randomized controlled trial (with superior efficacy vis-a`-vis dextran-70) [42]. Data supporting efficacy of the two direct thrombin inhibitors (DTIs) lepirudin and argatroban, compared against historical controls, are presented in Table 5 (HIT-associated thrombosis) [43–47] and Table 6 (isolated HIT) [45–48]. Two other agents (bivalirudin and fondaparinux) are rational therapies for HIT [40,49], but controlled studies are lacking.
Box 1: Principles of heparin-induced thrombocytopenia management when it is strongly suspected or has been confirmed Two Do’s Do stop all heparin (including LMWH and heparin administered as ‘‘flushes’’). Do start an alternative nonheparin anticoagulant, generally in therapeutic doses (eg, danaparoid [not available in the United States], lepirudin, or argatroban [bivalirudin and fondaparinux are rational therapies, but supporting data are sparse and anecdotal, and these therapies are ‘‘off-label’’ for HIT]. Two Don’ts Don’t give warfarin (or other coumarins) during the acute phase of HIT; if warfarin has already been started when HIT is diagnosed, vitamin K (eg, 10 mg over 30 minutes by intravenous injection) should be administered. Don’t give prophylactic platelet transfusions. Two Diagnostics Diagnosis laboratory support: test for HIT antibodies. In 80% to 90% of cases, an EIA is sufficient to rule in or to rule out HIT in the appropriate clinical context; in the remaining cases, patient serum should be referred to a laboratory experienced in performed a washed platelet activation assay (eg, serotonin release assay). Diagnosis of DVT: perform duplex ultrasonography to investigate for lower-limb DVT (the most common thrombotic sequela of HIT).
600
Table 4 Comparison of three alternative anticoagulants used to treat heparin-induced thrombocytopenia
Drug type (molecular mass)
Anticoagulant action
Immunologic features
Half-life
Lepirudin
Argatroban
Heterogeneous, polydispersed mix of anticoagulant GAGs: heparan sulfate (84%), dermatan sulfate (12%), chondroitin sulfate (4%) (6000 Da [mean]) Indirect (antithrombin-dependent) inhibition of factor Xa and thrombin (anti-Xa/anti-thrombin ratio, 22) In vitro XR for HIT Abs in minority of patients (usually not clinically significant); in high therapeutic concentrations, inhibits HIT Ab-induced platelet activation 25 h (anti-Xa action) (assumes normal renal function) Bolus: 2250 U (1500 U for b.w. <60 kg; 3000 U for b.w. 75–90 kg; 3750 U for b.w. >90 kg); followed by 400 U/h 4 h, then 300 U/h 4 h, then 150–200 U/h thereafter
65-amino acid polypeptide (6980 Da); Small-molecule arginine derivative recombinant derivative of hirudin (leech (527 Da) anticoagulant)
High affinity (Ki ¼ 0.0001 nmol/L) binding to two sites on thrombin (ie, bivalent DTI) Neither promotes nor inhibits HIT Ab binding to PF4/polyanion complexes; lepirudin is immunogenic (allergic/ anaphylactic reactions are reported)
Moderate affinity (Ki ¼ 40 nmol/L) binding to active site of thrombin (univalent DTI) Neither promotes nor inhibits HIT Ab binding to PF4/polyanion complexes; argatroban is not immunogenic
80 min (assumes normal renal function)
40–50 min (assumes normal hepatobiliary function) No bolus; 1 to 2 lg/kg/min (adjusted according to aPTT)c; marked dose reduction if hepatobiliary dysfunctiond
Bolus (only in case of life-threatening thrombosis): 0.2–0.4 mg/kg; infusion, start at 0.10 mg/kg/h (adjust to aPTT)a,b; reduce dosing for renal dysfunctionb
WARKENTIN
Dosing regimen
Danaparoid
Renal dysfunction (minor danaparoid accumulation)
Laboratory monitoring Anti-factor Xa levels (target therapeutic range: 0.5–0.8 anti-Xa U/mL)
Renal dysfunction (major lepirudin accumulation) aPTT (estimate of drug levele)
Hepatobiliary dysfunction, reduced liver perfusion (eg, low cardiac output state) (major argatroban accumulation) aPTT (estimate of drug levele)
Abbreviations: Ab, antibody; aPTT, activated partial thromboplastin time; b.w., body weight; Da, Dalton; GAGs, glycosaminoglycans; Ki, dissociation constant of an inhibitor; U, unit; XR, cross-reactivity. a Note that this dosing regimen is much lower than that described in the manufacturer’s package insert. b The initial lepirudin infusion rate should be no higher than 0.10 mg/kg/h (serum creatinine <90 lmol/L), with lower infusion rates for patients with higher serum creatinine levels (90–140 lmol/L, start at 0.05 mg/kg/h; 140–400 lmol/L, start at 0.01 mg/kg/h; >400 lmol/L, start at 0.005 mg/kg/h). aPTT monitoring should be performed at 4-hour intervals until it is apparent that steady state within the therapeutic range (1.5–2.5 times patient baseline [or mean laboratory] aPTT) has been achieved. c Although manufacturer’s package insert recommends that dosing start at 2 lg/kg/min, many clinicians start at lower dosages (eg, 1 lg/kg/min), especially in patients who are critically ill or who have cardiac failure. d For patients who have moderate or greater liver dysfunction, the starting dosage is 0.5 lg/kg/min. e The aPTT is not reliable for anticoagulant monitoring in patients who have preexisting congenital or acquired coagulopathies, overt DIC, prolonged aPTT due to ‘‘lupus anticoagulant,’’ or effects of warfarin.
HEPARIN-INDUCED THROMBOCYTOPENIA
Drug accumulation
601
602
Table 5 Summary of studies describing efficacy of the direct thrombin inhibitors (lepirudin, argatroban) as treatment for thrombosis complicating heparin-induced thrombocytopenia Study
Intervention/dosing
Lubenow et al, Lepirudin: bolus 0.4 mg/kg; 2005 [43] 0.15 mg/kg/hb Controls: variable (danaparoid, n ¼ 24; coumarin, n ¼ 21; other, n ¼ 30) Lewis et al, [45–47]
Argatroban: 2 lg/kg/min (no bolus)c Controls: variable (eg, coumarin)
New thrombosis
Composite end pointa
Major bleeds
Lepirudin: 15/214 (7%) Controls: 19/75 (25%) RR ¼ 0.28 (95% CI: 0.15–0.52) P < .001 Argatroban: 58/373 (16%) Controls: 16/46 (35%) RR ¼ 0.45 (95% CI: 0.28–0.71) P ¼ .0032f
Lepirudin: 41/214 (19%) Controls: 30/75 (40%) RR ¼ 0.48 (95% CI: 0.32–0.71) P < .001 Argatroban: 158/373 (42%) Controls: 26/46 (57%) RR ¼ 0.75 (95% CI: 0.57–0.99) P ¼ .0830g
All patients tested positive Lepirudin: 33/214 for HIT Abs (by the HIPA (15%)c Controls: 5/75 (7%) test); data shown are pooled from 3 studies RR ¼ 2.31 (95% CI: 0.94–5.71) (HAT-1, -2, -3) versus historical controls P ¼ .0723 Argatroban: Patients did not require 30/373 (8%)d positive test for HIT Controls: 1/46 (2%) Abse; data shown RR ¼ 3.70 (95% are pooled from 3 CI: 0.52–26.50) studies (Arg-911, P ¼ .2309 -915, and -915X) versus controls
Comment
WARKENTIN
P values are Fisher’s exact test, 2-tailed, unless otherwise indicated. Abbreviations: Abs, antibodies; aPTT, activated partial thromboplastin time; Arg, argatroban trial; CI, confidence interval; HAT, heparin-associated thrombocytopenia; HIPA, heparin-induced platelet activation; RCT, randomized controlled trial; RR, risk ratio; XR, cross-reactivity. a Composite end point: all-cause mortality, all-cause limb amputation, or new thrombosis (each patient counted only once). b aPTT adjusted to 1.5 to 2.5 times baseline aPTT; mean lepirudin treatment duration was 15.8 days (for 214 patients who were treated with lepirudin). c aPTT adjusted to 1.5 to 3.0 times baseline aPTT; mean argatroban treatment duration was 6.6 days (for 373 patients who were treated with argatroban). d Major bleeding, expressed per treatment day, was 0.97% for lepirudin and 1.25% for argatroban. e 65% of patients who were treated with argatroban tested positive for HIT antibodies in Arg-911 study [45] (data not available for the Arg-915/915X studies; see Ref. [46]). f P < .001 by log-rank test (time-to-event analysis for proportion of patients event-free: note that 376 patients, rather than 373 patients, were used for this analysis [47], because 3 patients initially deemed ineligible subsequently were determined to be eligible for analysis). g P ¼ .014 and P ¼ .008 for Arg-911 and Arg-915/915X studies, respectively, by log-rank analysis [45,46]; P value for pooled data not available.
Study
Intervention/dosing
New thrombosis
Composite end pointa
Major bleeds
Comment
Lubenow et al, 2004 [48]
Lepirudin: No bolus; 0.10 mg/kg/hb Con: variable (no treatment, n ¼ 31; coumarin, n ¼ 11, ASA, n ¼ 5) Argatroban: 2 lg/ kg/min (no bolus)g Con: variable
Lepirudin: 18/91 (20%) Controls: 14/47 (30%) RR ¼ 0.66 (95% CI: 0.36–1.21) P ¼ .2161e Argatroban: 94/349 (27%) Controls: 57/147 (39%) RR ¼ 0.69 (95% CI: 0.53–0.91) P ¼ .0104
Lepirudin: 13/91 (14%)c Controls: 4/47 (9%) RR ¼ 1.68 (95% CI: 0.58–4.86) P ¼ .4189f Argatroban: 15/349 (4%) Controls: 12/147 (8%) RR ¼ 0.53 (95% CI: 0.25–1.10) P ¼ .0875
All patients tested positive for HIT Abs (by the HIPA test); data shown are pooled from 3 studies (HAT-1, -2, -3) versus historical controls
Lewis et al, [45–47]
Lepirudin: 4/91 (4%) Controls: 7/47 (15%) RR ¼ 0.30 (95% CI: 0.09–0.96) P ¼ .0451d Argatroban: 24/349 (7%) Controls: 33/147 (22%) RR ¼ 0.31 (95% CI: 0.19–0.50) P < .001
Patients did not require positive test for HIT Absh; data shown are pooled from 3 studies (Arg-911, -915, and -915X) versus historical controls
HEPARIN-INDUCED THROMBOCYTOPENIA
Table 6 Summary of studies describing efficacy of the direct thrombin inhibitors (lepirudin, argatroban) as treatment for isolated heparin-induced thrombocytopenia
P values are Fisher’s exact test, 2-tailed, unless otherwise indicated. Abbreviations: Abs, antibodies; aPTT, activated partial thromboplastin time; Arg, argatroban; ASA; acetylsalicylic acid (aspirin); CI, confidence interval; HAT, ; HIPA, heparin-induced platelet activation; RCT, randomized controlled trial; RR, risk ratio; XR, cross-reactivity. a Composite end point: all-cause mortality, all-cause limb amputation, or new thrombosis (each patient counted only once). b aPTT adjusted to 1.5 to 2.5 times baseline aPTT; also, mean lepirudin treatment duration was 13.9 days (pooled data). c Major bleeding, expressed per treatment day, was 1.03% for lepirudin and 0.84% for argatroban. d P ¼ .02 by log-rank test (time-to-event analysis). e P ¼ .0281 by log-rank test (time-to-event analysis). f P ¼ .5419 by log-rank test (time-to-event analysis). g aPTT adjusted to 1.5 to 3.0 times baseline aPTT; also, mean argatroban treatment duration was 5.2 days (pooled data). h 50% of patients who were treated with argatroban tested positive for HIT antibodies in the Arg-911 study (data not available for Arg-915 study).
603
604
WARKENTIN
Warfarin is Contraindicated During Acute Heparin-Induced Thrombocytopenia Acute HIT is a risk factor for warfarin (coumarin) necrosis, which can manifest as venous limb gangrene or classic skin necrosis [4,5,50]. The pathogenesis is microthrombosis due to depletion of the vitamin K–dependent natural anticoagulant, protein C, in the setting of increased thrombin generation from HIT (disturbed procoagulant–anticoagulant balance) [4]. Venous limb gangrene is characterized by an underlying hypercoagulability disorder such as HIT; acral (distal extremity) necrosis in a limb affected by DVT; and a supratherapeutic INR that is usually greater than 3.5 (surrogate marker for severe depletion in protein C). In contrast, skin necrosis is characterized by involvement of skin and subcutaneous tissues at central (nonacral) sites (eg, breast, abdomen, thigh, calf). Coumarins also prolong the activated partial thromboplastin time (aPTT), thus predisposing to underdosing of DTI therapy (because DTIs usually are monitored by the aPTT) [50]. The risk for coumarin necrosis in HIT is about 5% to 10% [4,40], which is much greater than the risk in populations that do not have HIT (0.01%). Vitamin K therapy is advised if a diagnosis of HIT is made when warfarin has already been started [40,49,50]. PREVENTION OF HEPARIN-INDUCED THROMBOCYTOPENIA AND ITS SEQUELAE Platelet count monitoring for HIT is rational in many clinical settings, such as higher risk situations (eg, postoperative UFH thromboprophylaxis beyond 4 days). Here, at least every-other-day platelet count monitoring until day 14 or stopping heparin (whichever comes first) is suggested [40]. Similar monitoring during the use of therapeutic-dose UFH also is appropriate. Less frequent monitoring (eg, twice or thrice weekly) is suggested when the risk for HIT is lower (eg, UFH prophylaxis in medical patients or LMWH prophylaxis in postoperative patients). In settings in which the risk is low (eg, LMWH prophylaxis in medical patients or during pregnancy), routine platelet count monitoring for HIT is not recommended. Regardless of the clinical situation, an urgent platelet count usually is appropriate when a patient who is receiving heparin develops symptomatic thrombosis [1,2,8,30,40]. The reduced risk for HIT with LMWH (or fondaparinux) prophylaxis, compared with UFH, suggests that preferring these former agents could reduce the risk for HIT in certain situations; however, it is uncertain whether the risk is reduced in internal medicine patients who are given LMWH rather than UFH. This is either because the absolute risk for HIT is sufficiently low that a true difference is difficult to prove or no difference exists. REPEAT HEPARIN EXPOSURE DESPITE PREVIOUS HEPARIN-INDUCED THROMBOCYTOPENIA For patients with a history of HIT and who have an important indication for heparin (eg, cardiac or vascular surgery), it is recommended to use heparin, provided that antibodies are no longer detectable or (in emergencies) that
HEPARIN-INDUCED THROMBOCYTOPENIA
605
sufficient time has elapsed from the episode of HIT (>2–3 months) that clinically significant levels of antibodies are unlikely to be present. The basis of this recommendation is that it takes a minimum of 5 days to form HIT antibodies— even in patients with previous HIT—thus permitting safe intraoperative exposure. Further, there is no evidence that a patient who had HIT previously is more likely to form strong levels of HIT antibodies upon reexposure. In situations in which an alternative anticoagulant agent is available and safe (eg, postoperative prophylaxis with danaparoid, fondaparinux, or warfarin), however, this is preferred. References [1] Warkentin TE. Heparin-induced thrombocytopenia: pathogenesis and management. Br J Haematol 2003;121(4):535–55. [2] Arepally GM, Ortel TL. Clinical practice. Heparin-induced thrombocytopenia. N Engl J Med 2006;355(8):809–17. [3] Warkentin TE. Management of heparin-induced thrombocytopenia: a critical comparison of lepirudin and argatroban. Thromb Res 2003;110(2–3):73–82. [4] Warkentin TE, Elavathil LJ, Hayward CPM, et al. The pathogenesis of venous limb gangrene associated with heparin-induced thrombocytopenia. Ann Intern Med 1997;127(9): 804–12. [5] Warkentin TE. Clinical picture of heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 21–66. [6] Warkentin TE, Levine MN, Hirsh J, et al. Heparin-induced thrombocytopenia in patients treated with low-molecular-weight heparin or unfractionated heparin. N Engl J Med 1995;332(20):1330–5. [7] Warkentin TE, Roberts RS, Hirsh J, et al. An improved definition of immune heparin-induced thrombocytopenia in postoperative orthopedic patients. Arch Intern Med 2003;163(20): 2518–24. [8] Greinacher A, Farner B, Kroll H, et al. Clinical features of heparin-induced thrombocytopenia including risk factors for thrombosis. A retrospective analysis of 408 patients. Thromb Haemost 2005;94(1):132–5. [9] Warkentin TE, Kelton JG. A 14-year study of heparin-induced thrombocytopenia. Am J Med 1996;101(5):502–7. [10] Warkentin TE. Pseudo-heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 261–82. [11] Warkentin TE. Venous limb gangrene during warfarin treatment of cancer-associated deep venous thrombosis. Ann Intern Med 2001;135(8 Pt 1):589–93. [12] Warkentin TE. An overview of the heparin-induced thrombocytopenia syndrome. Semin Thromb Hemost 2004;30(3):273–83. [13] Kelton JG, Sheridan D, Santos A, et al. Heparin-induced thrombocytopenia: laboratory studies. Blood 1988;72(3):925–30. [14] Warkentin TE, Sheppard JI, Moore JC, et al. Laboratory testing for the antibodies that cause heparin-induced thrombocytopenia: how much class do we need? J Lab Clin Med 2005; 146(6):341–6. [15] Greinacher A, Po ¨ tzsch B, Amiral J, et al. Heparin-associated thrombocytopenia: isolation of the antibody and characterization of a multimolecular PF4-heparin complex as the major antigen. Thromb Haemost 1994;71(2):247–51. [16] Suh JS, Aster RH, Visentin GP. Antibodies from patients with heparin-induced thrombocytopenia/thrombosis recognize different epitopes on heparin:platelet factor 4. Blood 1998;91(3):916–22.
606
WARKENTIN
[17] Greinacher A, Gopinadhan M, Guenther JU, et al. Close approximation of two platelet factor 4 tetramers by charge neutralization forms the antigens recognized by HIT antibodies. Arterioscler Thromb Vasc Biol 2006;26(10):2386–93. [18] Warkentin TE, Sheppard JI, Siguoin CS, et al. Gender imbalance and risk factor interactions in heparin-induced thrombocytopenia. Blood 2006;108(9):2937–41. [19] Warkentin TE, Cook RJ, Marder VJ, et al. Anti-platelet factor 4/heparin antibodies in orthopedic surgery patients receiving antithrombotic prophylaxis with fondaparinux or enoxaparin. Blood 2005;106(12):3791–6. [20] Warkentin TE, Eikelboom JW. Who is (still) getting HIT? Chest 2007;131:1620–2. [21] Lee DH, Warkentin TE. Frequency of heparin-induced thrombocytopenia. In: Warkentin TE, Greinacher A, editors. Heparin-Induced Thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 67–116. [22] Fausett MB, Vogtlander M, Lee RM, et al. Heparin-induced thrombocytopenia is rare in pregnancy. Am J Obstet Gynecol 2001;185(1):148–52. [23] Warkentin TE. Drug-induced immune-mediated thrombocytopenia: from purpura to thrombosis. N Engl J Med 2007;356(9):891–3. [24] Warkentin TE, Kelton JG. Temporal aspects of heparin-induced thrombocytopenia. N Engl J Med 2001;344(17):1286–92. [25] Warkentin TE, Kelton JG. Delayed-onset heparin-induced thrombocytopenia and thrombosis. Ann Intern Med 2001;135(7):502–6. [26] Rauova L, Zhai L, Kowalska MA, et al. Role of platelet surface PF4 antigenic complexes in heparin-induced thrombocytopenia pathogenesis: diagnostic and therapeutic implications. Blood 2006;107(6):2346–53. [27] Warkentin TE, Jay RM, Makris M, et al. Platelet-activating anti-platelet factor 4/polyanion antibodies without preceding heparin therapy: a transient autoimmune disorder resembling heparin-induced thrombocytopenia (‘‘spontaneous HIT’’) [abstract]. Blood 2006;108(11): 311a–2a. [28] Warkentin TE. Think of HIT. Hematology Am Soc Hematol Educ Program 2006;408–14. [29] Greinacher A, Eichler P, Lietz T, et al. Replacement of unfractionated heparin by lowmolecular-weight heparin for postorthopedic surgery antithrombotic prophylaxis lowers the overall risk of symptomatic thrombosis because of a lower frequency of heparin-induced thrombocytopenia. Blood 2005;106(8):2921–2. [30] Warkentin TE. Think of HIT when thrombosis follows heparin. Chest 2006;130(3): 631–2. [31] Warkentin TE. Heparin-induced skin lesions. Br J Haematol 1996;92(2):494–7. [32] Warkentin TE, Hirte HW, Anderson DR, et al. Transient global amnesia following intravenous heparin bolus therapy is caused by heparin-induced thrombocytopenia. Thromb Haemost 1994;97(5):489–91. [33] Lo GK, Juhl D, Warkentin TE, et al. Evaluation of pretest clinical score (4 T’s) for the diagnosis of heparin-induced thrombocytopenia in two clinical settings. J Thromb Haemost 2006; 4(4):759–65. [34] Warkentin TE, Sheppard JI. Testing for heparin-induced thrombocytopenia antibodies. Transfus Med Rev 2006;20(4):259–72. [35] Warkentin TE. New approaches to the diagnosis of heparin-induced thrombocytopenia. Chest 2005;127(Suppl 2):35S–45S. [36] Lo GK, Sigouin CS, Warkentin TE. What is the potential for overdiagnosis of heparininduced thrombocytopenia? Am J Hematol, in press. [37] Warkentin TE, Hayward CPM, Smith CA, et al. Determinants of donor platelet variability when testing for heparin-induced thrombocytopenia. J Lab Clin Med 1992;120(3):371–9. [38] Greinacher A, Michels I, Kiefel V, et al. A rapid and sensitive test for diagnosing heparininduced thrombocytopenia. Thromb Haemost 1991;66(6):734–6. [39] Greinacher A, Juhl D, Strobel U, et al. Heparin-induced thrombocytopenia: a prospective study on the incidence, platelet-activating capacity and clinical significance of
HEPARIN-INDUCED THROMBOCYTOPENIA
[40]
[41] [42]
[43]
[44]
[45] [46] [47]
[48]
[49]
[50]
607
anti-PF4/heparin antibodies of the IgG, IgM, and IgA classes. J Thromb Haemost 2007; May 7 [epub ahead of print]. Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia: recognition, treatment, and prevention. The seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 2004;126(Suppl 3):311S–37S. Warkentin TE. Heparin-induced thrombocytopenia: diagnosis and management. Circulation 2004;110(18):e454–8. Chong BH, Gallus AS, Cade JF, et al. Prospective randomized open-label comparison of danaparoid with dextran 70 in the treatment of heparin-induced thrombocytopenia with thrombosis: a clinical outcome study. Thromb Haemost 2001;86(5):1170–5. Lubenow N, Eichler P, Lietz T, et al. Lepirudin in patients with heparin-induced thrombocytopenia—results of the third prospective study (HAT-3) and a combined analysis of HAT-1, HAT-2, and HAT-3. J Thromb Haemost 2005;3(11):2428–36. Greinacher A, Eichler P, Lubenow N, et al. Heparin-induced thrombocytopenia with thromboembolic complications: meta-analysis of two prospective trials to assess the value of parenteral treatment with lepirudin and its therapeutic aPTT range. Blood 2000;96(3): 846–51. Lewis BE, Wallis DE, Berkowitz SD, et al. Argatroban anticoagulant therapy in patients with heparin-induced thrombocytopenia. Circulation 2001;103(14):1838–43. Lewis BE, Wallis DE, Leya F, et al. Argatroban anticoagulation in patients with heparininduced thrombocytopenia. Arch Intern Med 2003;163(15):1849–56. Lewis BE, Wallis DE, Hursting MJ, et al. Effects of argatroban therapy, demographic variables, and platelet count on thrombotic risks in heparin-induced thrombocytopenia. Chest 2006;129(6):1407–16. Lubenow N, Eichler P, Lietz T, et al. Lepirudin for prophylaxis of thrombosis in patients with acute isolated heparin-induced thrombocytopenia: an analysis of 3 prospective studies. Blood 2004;104(10):3072–7. Greinacher A, Warkentin TE. Treatment of heparin-induced thrombocytopenia: an overview. In: Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. 4th edition. New York: Informa Healthcare USA, Inc.; 2007. p. 283–317. Warkentin TE. Should vitamin K be administered when HIT is diagnosed after administration of coumarin? J Thromb Haemost 2006;4:894–6.