The anticoagulant heparin is used for a wide variety of conditions and procedures, including surgical and interventional procedures, acute coronary syndromes, venous thromboembolism, atrial fibrillation, peripheral arterial occlusive disease, dialysis, and extracorporeal circulation.[1] It is estimated that about 12 million hospitalized Americans receive some form of heparin each year.[2] Approximately half of heparin recipients are medical patients and half are surgical/procedural patients.[3]
Heparin binds to platelet factor 4 (PF4) and exposes a previously masked immune epitope, leading to formation of heparin-PF4 immunoglobulin (Ig) G antibodies in some heparin-treated patients. In vitro study of the interaction of heparin and PF4[4] suggests that heparin promotes large antigen complexes through a charge neutralization mechanism, and that antibody generation is promoted by low concentrations of heparin. These antibodies bind to the heparin-PF4 complex, and the resulting circulating immune complex (CIC) activates platelets through binding to the Fc receptor. This is thought to mediate the process of thrombocytopenia and promote a state of hypercoagulability that can result in thrombosis. Platelet activation also leads to additional PF4 release, as well as endothelial cell injury, as depicted in Figure 1.[5] This increase in plasma PF4 could serve as a positive feedback mechanism in the presence of heparin for further stimulation of the immune reaction.
Figure 1. Model of pathogenesis of HIT[5]
Thrombocytopenia associated with heparin due to heparin-PF4 antibodies is referred to as heparin-induced thrombocytopenia (HIT). HIT may affect up to 5% of the patients receiving unfractionated heparin (UFH), but an accurate estimate is complicated by diagnostic difficulty. A substantial number of patients with HIT develop morbidity and mortality due to thrombotic complications.[6] Patients who are treated with heparin and develop HIT are at a markedly increased relative thrombotic risk (odds ratio [OR] = 20 to 40).[7] Some of the surgical and medical conditions associated with HIT are listed in Table 1.[1]
Table 1. Frequency of HIT[1]
The ORs for other HIT risk factors are given in Table 2.[8] The strongest risk factor is prior exposure to heparin, which increases the OR by 20 to 100 times.
*OR = odds ration; LMWH = low-molecular-weight heparin, UFH = unfractionated heparin
Table 2. Risk Factors for HIT[8]
HIT appears 5-10 days after initiation of heparin therapy.[9] Patients previously exposed to heparin within 100 days can experience "rapid onset" HIT as a result of the interaction of heparin with existing circulating antibodies, as illustrated in Figure 2.[9] Some patients with HIT antibodies can develop acute systemic reactions such as fever, chills, hypertension, tachycardia, or dyspnea.[10]
Figure 2. Representative Case of Typical-Onset HIT, Followed by a Rapid-Onset Episode 2[9]
In addition to thrombocytopenia, HIT complications include deep vein thrombosis (DVT), pulmonary embolism (PE), myocardial infarction, thrombotic stroke, and occlusion of a limb artery requiring amputation.[10] The rate of venous thrombosis is 4 times higher than arterial thrombosis in surgical patients,[6] although the rate of arterial events is higher in patients receiving heparin for cardiovascular disease.[11] In HIT patients with thrombosis, approximately 10% require a limb amputation and the mortality rate is estimated at 17% to 30%.[1]
Multiple clinical studies show that low-molecular-weight heparin (LMWH) poses less of a risk of HIT than UFH.[12] This model is consistent with the notion that higher molecular-weight molecules and polymers are more antigenic and, therefore, more likely to elicit a robust antibody response. However, a recent meta-analysis challenges the superiority of LMWH.[13] This analysis of 13 randomized controlled trials (RCT) focused on treatment of PE and/or DVT patients and found no evidence to support the suggestion that UFH is associated with a higher rate of complications than LMWH. As noted in an accompanying editorial,[12] this population is at a lower risk than surgical patients receiving prophylactic heparin, and heparin treatment patients typically receive higher doses than prophylaxis patients. Higher heparin doses may be less antigenic. The activation of platelets and endothelial injury is dictated by multiple factors, including the state of the immune system, extent of trauma, antibody titer, stoichiometry of heparin:PF4 complex formation, and perhaps the multimeric nature of the heparin. Another meta-analysis of RCTs focused largely on patients after orthopedic surgery and concluded that the risk of HIT with LMWH is considerably lower than with UFH.[14] Finally, a meta-analysis[15] of VTE after prophylactic or therapeutic administration of UFH or LMWH found a 13% incidence of HIT-related VTE after subcutaneous or IV UFH vs 0.7% for those given LMWH. Thus, most evidence suggests that from a medical standpoint, LMWH poses no greater risk of HIT than UFH, and may have less risk for some patients.
The frequency of HIT varies considerably in different settings and between different indications, as noted in Table 1. A recent review of HIT occurrence in the intensive care unit[16] found that HIT is uncommon (0.3 to 0.5%), whereas thrombocytopenia from other causes is very common (30% to 50%). This poses a danger of overdiagnosis and suggests the use of strict criteria to enhance diagnostic accuracy. Heparin-PF4 antibodies were assayed in patients presenting to the emergency department (ED) with thrombosis.[17] Of 324 ED patients with chest pain or symptoms of thrombosis, there was no difference in the seropositivity rate between 196 patients recently hospitalized and the 128 patients not recently hospitalized (P = .19). Nor was there a difference in patients with chest pain and those with other thrombosis (P = .64). Of 22 seropositive patients retested, 8 (7 recently hospitalized) had platelet-activating antibodies. The overall heparin-PF4 antibody prevalence was 7.4% in ED patients with chest pain or thrombosis, with approximately 1 out of 3 seropositive patients having platelet-activating antibodies.
In a study of 598 consecutive hospitalized patients treated with subcutaneous UFH for prophylactic (n = 360) or therapeutic (n = 238) indications, the incidence of clinically diagnosed HIT was about 0.8%.[18] Although antibody titers were available in only 3 of the 5 HIT subjects, all 3 sera were positive for heparin-PF4 antibodies. Three of the patients suffered thromboembolism and 2 died. This prevalence of thromboembolic complications in patients with HIT (60%) was higher than that observed in the remaining 593 patients (3.5%); OR = 40.8 (95% CI, 5.2-162.8). All observed cases of HIT occurred in the heparin prophylaxis group. Care must be exercised in drawing conclusions from a study with a small number of cases.
In the setting of cardiopulmonary bypass surgery, 25% to 50% of postcardiac surgery patients develop heparin-dependent antibodies during the next 5 to 10 days. The risk of HIT is 1% to 3% if UFH is continued throughout the postoperative period.[19] Although postcardiac surgery patients with HIT have thrombosis rates typical of HIT, arterial thrombi predominate, unlike other populations at risk for the development of HIT. HIT-associated mortality was 28% in a large observational study of cardiac surgery patients.[20]
Hemodialysis presents a special risk for HIT because patients are repeatedly exposed to heparin. A recent prospective study of 419 asymptomatic hemodialysis patients showed that heparin-PF4 IgG antibody formation is associated with increased mortality.[21] Elevated mortality in this population cannot be attributed to HIT, however, because the increase in major cardiovascular events in patients with heparin-PF4 antibodies is a trend that does not reach statistical significance.
Warkentin[22] summarized the risk of HIT in various settings (Figure 3). The three major determinants of HIT risk are duration of heparin exposure, the type of heparin (bovine UFH > porcine UFH > porcine LMWH) and the type of patient (surgical > medical > obstetric). Cardiac surgery patients have approximately half the risk of HIT than orthopedic surgery patients, though the HIT antibody positivity rate is 3-fold higher in the former group; the grounds for this discrepancy are not known. In a retrospective study of cardiac transplant patients, 45 patients were identified with thrombocytopenia. Of these, 11 had positive results in a HIT ELISA test, and 5 of these were diagnosed with HIT with thrombosis syndrome.[23]
Figure 3. Iceberg model of HIT
A. The iceberg model is built on antibody positivity. Some of the patients have activating antibodies, a subset has thrombocytopenia (dark grey), and a minority has HIT with thrombosis (black).
B. Frequency of HIT antibody formation and of clinical HIT in different patient populations. Although the frequency of HIT antibody formation is higher in cardiac than in orthopedic surgery patients, the risk of developing clinical HIT is higher in orthopedic surgery patients than in cardiac surgery patients.[22]
Diagnosis
Thrombocytopenia is the hallmark of HIT. "Typical-onset" HIT occurs within 5-10 days of heparin initiation and is characterized by a ≥ 50% fall in platelet count. This is the most common presentation, occurring in approximately 70% of HIT patients.[22] For approximately 90% of HIT patients, the platelet count bottoms out at 15-150 X 109 platelets/L.[24] In "rapid-onset" HIT associated with prior heparin exposure, thrombocytopenia occurs within 24 hours of heparin initiation. The short time course is due to the presence of existing antibodies, rather than the stimulation of new antibody formation. "Delayed-onset" HIT, usually presenting 1-6 weeks after heparin exposure, is a severe condition characterized by high-titer HIT antibodies that are not dependent on ongoing heparin stimulation, and can occur after discontinuation of heparin therapy.
Although thrombocytopenia should raise the suspicion of HIT, and laboratory-verified antibodies can occur in the absence of true HIT, the combination of clinical and laboratory evidence in a patient exposed to heparin should raise a strong clinical suspicion for HIT. Table 3 details the characteristics of clinical thrombocytopenia and the various laboratory tests that have been developed for the heparin-PF4 antibodies.[22]
Table 3. Clinical and pathologic features of HIT[22]
*Thrombocytopenia usually defined as a ≥ 50% fall in platelet count. In a few patients, thrombocytopenia is associated with thrombosis, skin lesions, or other clinical sequelae and HIT antibodies.†Usually associated with coumarin therapy (variant of coumarin-induced skin necrosis).
‡Reported to be associated with adrenal vein thrombosis, with subsequent adrenal hemorrhage.
§Occurs in 10% to 20% of patients who develop HIT while receiving subcutaneous heparin injections (eg, erythematous plaques and skin necrosis).
¡Occurs in 25% of HIT patients who receive an IV heparin bolus at a time when they have formed HIT antibodies.
¶Indicates prolongation of international normalized ratio, hypofibrinogenemia, or RBC fragmentation attributable to HIT.
#In a few patients, an HIT-mimicking syndrome could be explained by antibodies directed against chemokines other than PF4 (eg, interleukin-8 or neutrophil activating peptide-2).
**Detected by flow cytometry.
††These tests are performed infrequently due to limited sensitivity and specificity for HIT.
Often, thrombosis is the first sign of HIT. DVT (50%) and PE (25%) are the most common forms.[10] Arterial thrombosis usually occurs at sites of vascular damage and frequently manifests as acute lower limb ischemia. Thus, cardiovascular surgery patients have a higher risk of arterial thrombosis.[2]
A useful framework for clinical evaluation of HIT is the "4T" system.[7] Although the presence of HIT antibodies is key for confirming the diagnosis of HIT, the acute presentation of thrombocytopenia requires quick and judicious action on the part of the clinician. The 4T system permits pretest evaluation of the likelihood of HIT and recommends appropriate action. As shown in Table 4, the 4T system assigns point values: 0, 1, or 2 points according to the severity of clinical evaluation in the four categories. The four categories are: thrombocytopenia; timing of the decrease in platelet count; thrombosis or other sequelae; and other causes of thrombocytopenia. In the first category, there are two alternate criteria for thrombocytopenia; either a percent drop from preheparin levels or a nadir value of platelet count. The second category accounts for the timing of prior heparin exposure, if appropriate. The third category assigns 2 points for proven thrombosis, skin necrosis or an acute systemic reaction, 1 point for lesser symptoms, and 0 points for no symptoms. The fourth category considers other causes for thrombocytopenia, which are summarized in Table 5. A recent publication by Lo et al[25] emphasized the importance of a clinical tool such as the 4T system. The study found that using only an ELISA assay led to overdiagnosis of HIT by ~100%. The patients with positive ELISA results, irrespective of the pretest probability of HIT, tended to have weak antibodies and low risk for thrombosis, suggesting non-HIT explanations for thrombocytopenia.
Table 4. Clinical estimation of the probability of HIT: The 4Ts[7]
A total score is calculated by adding the results from the four categories. A high probability of HIT is suggested by scores of 6-8, and the recommended immediate action is replacement of heparin with a direct thrombin inhibitor. A score of 4 or 5 indicates a moderate risk for HIT, and the course of action is left to physician judgment. A lower score suggests a low risk for HIT.
Table 5: Differential Diagnosis of Thrombocytopenia[26]
Another source of thrombocytopenia is drug-induced thrombocytopenia (DIT). This disorder can be a consequence of decreased platelet production (bone marrow suppression) or accelerated platelet destruction.[27] Immune-mediated platelet consumption is often mediated by antibodies that recognize platelet glycoproteins in the presence of a particular drug. Some instances of DIT are described in Table 6.[28]
Table 6. Drug-induced immune thrombocytopenia: pathogenetic mechanisms[28]
Two types of laboratory tests are available to confirm a diagnosis of HIT. ELISA assays measure the presence of antibodies to the heparin-PF4 complex. The assays are sensitive, but because HIT antibodies can be present without clinical disease, their specificity is low. Functional assays include a heparin-induced platelet activation (HIPA) test and a serotonin release assay (SRA). The HIPA assay has variable sensitivity in different laboratories, but is relatively easy to perform. The SRA has high sensitivity and specificity, however it is technically demanding to perform and not widely available.[5] Because the results of such tests are usually not available until hours (or days) after clinical diagnosis, it is recommended that the healthcare provider act on the basis of the clinical presentation, obtain the laboratory tests in a timely manner, and adjust the management plan in accordance with all of the evidence when it becomes available.
A recent study examined the value of repeat ELISA antibody testing as a predictor for HIT.[29] Negative initial antibody tests were followed within 3 weeks with secondary antibody tests. Optical densities (OD) in the negative range (OD < 0.4) were subdivided into low (OD = 0-0.132), medium (OD = 0.133-0.267), and high (OD = 0.268-0.399) groups. Figure 4 shows that in the initial negative result group, 40% of the patients with high titers converted to subsequent positivity (SP) in the follow-up test, while 16% of the medium-titer patients and 7% of the low-titer patients converted. Most of the patients with a second positive test were exposed to heparin in the interim. When the 4T scores were calculated for the initially negative-test group, 77% of patients had a low pretest clinical score, 23% had an intermediate pretest clinical score, and no patient had a high pretest clinical score.
Figure 4. Subsequent positivity (SP) of patients with negative first ELISA tests.
Of the patients with a secondary negative test, 80% (89/111) had a low pretest clinical score, while only 20% had an intermediate pretest clinical score. For the subsequent positive-test group, 62% (13/21) had a low pretest clinical score and 38% had an intermediate pretest clinical score. The one-month mortality rate was similar between groups; 7/22 patients (33%) in the subsequent positive group and 30/115 (26%) in the subsequent negative group (P = .58). This study found that follow-up 4T evaluation is more useful than repeating the ELISA test for monitoring patients with initial negative ELISA results. Further testing that encompasses both positive and negative ELISA results, as well as subsequent thrombosis, will be useful to define the positive and negative predictive values of the ELISA test for different types of patients.
A retrospective study of samples tested in a university hospital laboratory investigated the agreement of ELISA and serotonin release assays.[30] These tests were done on 1017 consecutive samples from patients suspected of having HIT based on clinical evidence. Using the ELISA manufacturer's criteria for reading results, there was no serological evidence of HIT in 83% of the samples. A positive ELISA and a negative SRA were seen in 12% of the samples, and slightly more than 4% of the samples showed a positive ELISA antigenic test and a positive SRA functional test. A small percentage (< 1%) of samples were negative in the ELISA test but were positive in the specific SRA test. These results could be artifactual false-positives, or could result from the presence of cross-reacting antibodies to proteins homologous to PF4.
Therapy
Although discontinuation of heparin is a critical element of managing HIT, alternative anticoagulation therapy is important to minimize the risk of thrombosis. Both UFH and LMWH should be discontinued because of their antigenic similarity and ability to stimulate further immune reaction. Patients can be inadvertently exposed to heparin through its use in heparinized catheters and heparin lock flushes. Care should be taken to eliminate all exposure to heparin.
There are several direct thrombin inhibitors (DTI) that are useful as anticoagulation therapy in HIT. Natural hirudin is produced in trace amounts as a family of highly homologous isopolypeptides by the leech Hirudo medicinalis. Lepirudin is a recombinant hirudin, produced in yeast cells, and is a highly specific direct inhibitor of thrombin. This polypeptide is composed of 65 amino acids. Hirudin inhibits thrombin by irreversibly binding both to the catalytic site and to the anion-binding exosite of circulating and clot-bound thrombin. The bioengineered lepirudin molecule is identical to natural hirudin except for the substitution of leucine for isoleucine at the N-terminal end of the molecule and the absence of a sulfate group on the tyrosine at position 63.[31]
Two similar clinical studies (HAT-1 and HAT-2) examined the use of lepirudin for HIT.[31] Overall, 198 (HAT-1: 82, HAT-2: 116) patients were treated with lepirudin and 182 historical control patients were treated with other therapies. All prospective patients and all historical control patients except 5 were diagnosed with HIT using the HIPA or equivalent assays. In total, 113 prospective patients ("lepirudin") and 91 historical control patients ("historical control") presented with thromboembolic complications (TEC) at baseline (day of positive test result) and qualified for direct comparison of clinical endpoints.
Initiation of treatment was set as the starting point for the analyses. For the historical control group, the first treatment selected within 2 days of laboratory confirmation of HIT was used for reference. Seven days after the start of treatment, the cumulative risk of death, limb amputation, or new TECs was 3.7% in the HAT-1 lepirudin patients and 16.9% in the HAT-2 lepirudin patients, compared with 24.9% in the historical control group. At 35 days, when approximately 10% of patients were still at risk, the cumulative risk was 13.0% in the HAT-1 lepirudin patients and 28.9% in the HAT-2 lepirudin patients, compared with 47.8% in the historical control group.
To consolidate the data from the two trials, a meta-analysis was done with the pooled data from lepirudin patients of the HAT-1 and HAT-2 studies who presented with TECs at baseline, compared with the respective historical control patients. At 7 and 35 days after start of treatment, the cumulative risks of death were 4.4% and 8.9% in the lepirudin group, compared with 1.4% and 17.6% in the historical control group. The cumulative risks of limb amputation were 2.7% and 6.5% in the lepirudin group, compared with 2.6% and 10.4% in the historical control group. Most important, the cumulative risks of new TECs were 6.3% and 10.1% in the lepirudin group, compared with 22.2% and 27.2% in the historical control group. As shown in Figure 5, lepirudin substantially reduced the risk of time-to-event of cumulative risk of death, limb amputation, or new TECs in comparison with a historical control group (P = .004 log-rank test). In another analysis, the Haemostasis and Thrombosis Task Force of the British Committee for Standards in Haematology concluded that lepirudin, at doses adjusted to achieve an activated partial thromboplastin time (aPTT) ratio of 1.5 to 2.5, reduces the risk of reaching the composite endpoint of limb amputation, death, or new thrombosis in patients with HIT and HITT.[32]
Figure 5. Cumulative risk of death, limb amputation, or new TECs after start of treatment[31]
Antihirudin antibodies develop in 40% or more of HIT patients treated with lepirudin. These antibodies may increase the anticoagulant effect of lepirudin, possibly due to stabilization of the molecule by the antibody and delayed renal elimination of active lepirudin-antihirudin complexes. Therefore, strict monitoring of the aPTT is necessary during prolonged therapy. Increased and refractory bleeding as well as anaphylaxis on re-exposure can be complications of lepirudin therapy.[33] Since lepirudin is excreted through the kidneys, patients with renal impairment must be monitored carefully. Lepirudin is indicated for anticoagulation in patients with heparin-associated thrombocytopenia and associated thromboembolic disease in order to prevent further TECs.
Argatroban is a synthetic direct thrombin inhibitor derived from L-arginine.[34] It binds reversibly to the thrombin active site, and does not require the cofactor antithrombin III for antithrombotic activity. Argatroban is highly selective for thrombin with an inhibitory constant (Ki) of 0.04 mcM. At therapeutic concentrations, argatroban has little or no effect on related serine proteases (trypsin, factor Xa, plasmin, and kallikrein). Argatroban is capable of inhibiting the action of both free and clot-associated thrombin. It exerts its anticoagulant effects by inhibiting thrombin-catalyzed reactions, including fibrin formation; activation of coagulation factors V, VIII, and XIII; activation of protein C; and platelet aggregation. For infusion doses of up to 40 mcg/kg/min in healthy volunteers and cardiac patients, argatroban increases the following in a dose-dependent fashion: the aPTT; the activated clotting time (ACT); the prothrombin time (PT) and International Normalized Ratio (INR); and the thrombin time (TT).
Thrombotic outcomes in 882 HIT patients (697 patients receiving argatroban for 5 to 7 days, plus 185 historical control subjects) from prospective studies were retrospectively analyzed (Table 7).[35] Time-to-event analyses were conducted of the composite primary end point and its individual components (death due to thrombosis, amputation secondary to HIT-associated thrombosis, or new thrombosis within 37 days). Argatroban significantly reduced the thrombotic composite endpoint compared with the historical control risk (HIT hazard ratio = 0.33; P < .001; HIT with thrombosis hazard ratio = 0.39; P < .001). More argatroban-treated patients than control subjects remained free of thrombotic events during follow-up, regardless of whether baseline thrombosis was absent (91% vs 73%, respectively) or present (72% vs 50%, respectively). Argatroban significantly reduced new thrombotic events (P < .001) and death due to thrombosis (P < .001). Major bleeding was similar between groups (6% to 7%, P = .74). This study is consistent with the contention that argatroban provides effective antithrombotic therapy in patients with HIT without significantly increasing the risk of bleeding complications.
Argatroban is indicated in the United States as an anticoagulant for prophylaxis or treatment of thrombosis in patients with HIT, and also as an anticoagulant in patients with (or at risk for) HIT undergoing percutaneous coronary interventions (PCI). Since it is metabolized in the liver and excreted in feces, it may be a strategic alternative to lepirudin in patients with renal impairment; on the other hand, patients with hepatic insufficiency and normal renal function should be considered for treatment with lepirudin.
Table 7. Patients event-free of the thrombotic composite or its individualized components by time-to-event analysis[35]
Figure 6.Time-to-event analysis of new thrombosis by HIT presentation. Circles indicate the days that patients were censored. Upper panel: argatroban-treated patients (n = 321, 299 censored) and control patients (n = 139, 108 censored) with HIT (hazard ratio, 0.29; P < .001). Lower panel: argatroban-treated patients (n = 376, 320 censored) and control patients (n = 46, 30 censored) having HIT with thrombosis (hazard ratio, 0.32; P < .001)[35]
Bivalirudin is a synthetic, 20 amino acid peptide that is a specific and reversible direct thrombin inhibitor.[36] Bivalirudin inhibits thrombin by binding both to the catalytic site and to the anion-binding exosite of circulating and clot-bound thrombin. The binding of bivalirudin to thrombin is reversible, as the inhibitor is a substrate for the protease. Thrombin slowly cleaves the bivalirudin-Arg3-Pro4 bond, resulting in recovery of thrombin proteolytic function. In in vitro studies, bivalirudin inhibited both soluble (or free) and clot-bound thrombin and was not neutralized by products of the platelet release reaction. It prolonged the aPTT, TT, and PT of normal human plasma in a concentration-dependent manner. Bivalirudin is indicated in the United States for the following uses:
- As an anticoagulant in patients with unstable angina undergoing percutaneous transluminal coronary angioplasty (PTCA)
- As an anticoagulant in patients undergoing percutaneous coronary intervention (PCI) with provisional use of glycoprotein IIb/IIIa inhibitor (GPI)
- For patients with, or at risk of, HIT/HITT undergoing PCI.
Fondaparinux, a synthetic pentasaccharide, is an indirect Factor Xa inhibitor that acts by causing conformational changes in antithrombin, thereby accelerating the rate of factor Xa inhibition. In vitro studies have shown that fondaparinux does not cause platelet activation in the presence of HIT antibodies. However, the rate of heparin-PF4 antibody formation was the same in orthopedic surgery patients treated with LMWH or fondaparinux, despite the expectation that fondaparinux would be less antigenic because of its size (less than 8-10 subunits).[37] This was observed in both ELISA and IgG-mediated heparin-dependent platelet activation tests. Although fondaparinux was associated with the formation of antiheparin-PF4 antibodies that are indistinguishable from those generated during LMWH therapy, these antibodies do not bind well to PF4/fondaparinux. Despite the induction of these antibodies and the similarity in sugar structure to heparin, there have been no documented cases of HIT associated with fondaparinux use.
A recent case report documents strong serum-induced platelet activation, thrombocytopenia, and antiheparin-PF4 antibodies in a patient treated with fondaparinux in the absence of heparin.[38] This suggests that on rare occasions, fondaparinux can cause a disorder resembling HIT. Fondaparinux may present an alternative therapy to heparin in cases where HIT or HITT are suspected, though it is not currently approved for this indication. Fondaparinux is indicated for prophylaxis of DVT in various surgical settings as well as treatment in conjunction with warfarin of acute DVT and acute PE.
Properties and recommended uses of these DTI drugs as well as fondaparinux are summarized in Table 8.
Table 8. Nonheparin anticoagulants and direct thrombin inhibitors in HIT[2,33,39](Click to enlarge).
IV, intravenous; SQ, subcutaneous; aPTT, activated partial thromboplastin time; INR, international normalized ratio; ACT, activated clotting time; ECT, ecarin clotting time
Table 9 summarizes some of the misconceptions about the incidence, diagnosis, and treatment of HIT.
Table 9. Misconceptions about HIT[40]
Warfarin is an anticoagulant that acts by inhibiting vitamin K-dependent coagulation factors.[41] Vitamin K is an essential cofactor for the postribosomal synthesis of these factors. The vitamin promotes the biosynthesis of gamma-carboxyglutamic acid residues in the proteins, which are essential for biological activity. Warfarin is approved for the prophylaxis and/or treatment of several thrombotic disorders, including venous thrombosis, pulmonary embolism, and thromboembolic complications associated with atrial fibrillation.
Warfarin can cause major or fatal bleeding, and carries a black box warning. Bleeding is more likely to occur during the starting period and with a higher dose (resulting in a higher INR). Risk factors for bleeding include high intensity of anticoagulation (INR > 4.0), age ≥ 65, highly variable INRs, history of gastrointestinal bleeding, hypertension, cerebrovascular disease, serious heart disease, anemia, malignancy, trauma, renal insufficiency, concomitant drugs and long duration of warfarin therapy.
Treatment of HIT-associated DVT with warfarin alone can contribute to venous limb gangrene.[7] This typically occurs with discontinuation of other anticoagulants in patients with a high INR (usually > 3.5) and is due to a severe reduction in factor VII and protein C with persisting thrombin generation. Patients with isolated confirmed HIT who are managed by simple discontinuation of heparin or substitution of heparin with warfarin have a risk of symptomatic thrombosis from 25% to 50%, including an overall risk of fatal thrombosis of approximately 5%.[19]
Warfarin is not recommended for patients with strongly suspected or confirmed HIT until after the platelet count has substantially recovered. It should be administered with at least a 5-day overlap with an alternative anticoagulant (usually a DTE), and begun at a relatively low dose. The alternative anticoagulant should not be stopped until the INR is within the target therapeutic range for at least the last 2 days.[7] For HITT with thrombosis, warfarin should be continued for at least 3 months.
Future Prospects for Therapy
The use of fondaparinux as an anticoagulant in HIT is a potential alternative to the parenteral direct thrombin inhibitors, but formal clinical trials in this disorder have not been undertaken. Several oral direct Factor Xa inhibitors for the prevention and treatment of thrombosis are being developed. These include rivaroxaban (BAY 59-7939), apixaban (BMS), YM-150 (Astellas), DU-176b (Daiichi), LY517717 (Lilly), and PRT054021 (Portola).[42] These compounds could potentially be applied to patients with HIT.
Clarification of the basic mechanisms of heparin-PF4 antibody induction and the interaction of these antibodies with the thrombotic pathway and platelet activation may reveal novel targets for intervention in the steps of antibody formation, as well as in HIT and HIT-associated thrombosis.
Summary
Immune HIT is a rare but serious consequence of heparin administration. Different patient populations are at variable risk for HIT, although the grounds for these discrepancies are not clear. The "4T" system of clinical diagnosis gives a framework for accurate diagnosis as well as acute treatment recommendations. Accurate diagnosis of HIT, however, depends on laboratory confirmation of the presence of antibodies to the heparin-PF4 complex.
Removal of heparin is a critical but insufficient management practice, and alternative anticoagulants should be carefully used to prevent thrombosis. The renal and hepatic health of the patient should be considered when choosing a therapy because different drugs are cleared by these organs. Small molecule anticoagulation drugs are in development for prevention of thrombosis and may be of use in HIT after evaluation in clinical trials.