Clinical Presentation of HIT. Part 4
A number of investigators have adapted the enzyme-linked immunosorbent assay (ELISA) technique to devise diagnostic tests for HIT. The most recent of these uses immobilized PF4 bound to heparin as a target for the HIT antibody. Initial studies reported PF4 ELISA to be at least as sensitive as both the 14C-serotonin release and HIPA assays, although discrepant results occur in up to 25% of patients, indicating that the different assays recognize different patient cohorts. Two recent studies evaluating this assay in patients after cardiopulmonary bypass surgery have reported the heparin-associated antibody to be detectable in .50% of patients by the time of hospital discharge. The presence of antibody was not predictive of either thrombocytopenia or thromboembolic complications, suggesting that the increased sensitivity may be of limited clinical utility in this setting. Further evaluation of this assay will be required before its integration into routine clinical practice.
Prevention
The most important principle of prevention is to minimize the exposure to heparin. In patients with established venous thrombosis, early commencement of warfarin shortens the duration of heparin therapy, thus minimizing the likelihood of developing thrombocytopenia without affecting efficacy. Most investigators stress the importance of maintaining a high index of suspicion and performing regular platelet counts in all patients receiving heparin therapy. This has been predicated on the retrospective observation that prompt diagnosis of the condition with abrupt withdrawal of the drug can substantially reduce the complication and mortality rates. However, in a more recent retrospective analysis, patients with a diagnosis of HIT had a subsequent 30-day risk of thrombosis .50%, despite cessation of heparin therapy with or without institution of warfarin therapy, suggesting that additional alternative anticoagulant therapy may be necessary in this patient group.
Treatment
HIT type I. Asymptomatic patients developing mild thrombocytopenia in the absence of the heparin-dependent antibody do not require specific treatment. However, it may be difficult to distinguish this condition from early type II; these patients should be closely monitored, and if there is any doubt, heparin should be discontinued.
HIT type II. In patients who develop thrombocytopenia with a positive laboratory test for the heparin-dependent antibody, the cornerstone of therapy is the absolute discontinuation of heparin. It is imperative that all potential sources of heparin be avoided, a task that can be difficult given the ubiquity of the drug in the hospital setting. Platelet transfusions are not recommended, both because bleeding complications are uncommon and thrombotic events can follow the transfusions. Once heparin is discontinued, the platelet count should begin to increase within 24 to 48 h and reach normal levels by 4 to 5 days. For patients with confirmed venous or arterial thrombosis, the selection of substitute anticoagulant therapy is particularly problematic in the United States at present, as there is currently no readily available, effective, alternative antithrombotic drug. If long-term anticoagulation is required, treatment with warfarin should be commenced. However, warfarin may precipitate venous limb gangrene in patients with deep venous thrombosis complicating HIT by producing protein C deficiency without simultaneous inhibition of thrombin generation, so it should not be administered without concomitant use of a rapidly acting anticoagulant agent.
Clinical Presentation of HIT. Part 3
HIT should be suspected in any patient who develops thrombocytopenia (,150,000/ml) or a 50% or greater fall in the platelet count after 5 days of heparin therapy. The platelet count should be repeated and the blood film examined to exclude platelet clumping causing pseudothrombocytopenia. After confirmation of the low platelet count, the diagnosis of HIT may be made according to the following criteria:
1) occurrence of thrombocytopenia during heparin administration;
2) exclusion of other causes of thrombocytopenia such as infection, drugs and autoimmune thrombocytopenia;
3) resolution of thrombocytopenia after cessation of heparin therapy;
4) demonstration of a heparin-dependent platelet antibody by an in vitro test.
Criteria 1 and 2 only are required for the diagnosis of HIT type I. As in most hospitals, in vitro confirmation requires analysis of samples in a reference laboratory (see later discussion); the initial diagnosis of HIT is often made on clinical grounds (criteria 1, 2 and 3). In the absence of thrombocytopenia, HIT should also be considered if a patient receiving heparin experiences a new thrombosis or develops heparin resistance or (rarely) skin necrosis at sites of heparin administration.
Laboratory Studies for the Heparin-Dependent Platelet Antibody
A variety of laboratory tests for the detection of heparindependent platelet antibodies have been described. The most widely used is the platelet aggregation test, which measures the aggregation of normal donor platelets by patient serum or plasma in the presence of heparin. This test is popular because it is simple, inexpensive and based on a technique that is already in use in most hemostatic laboratories and can provide a result within 2 to 3 h. Although it has a specificity of ;90%, the sensitivity is reported to be between 30% and 50%, which, in the view of some investigators, limits its clinical usefulness. However, when performed under optimal conditions, with appropriate positive and negative controls, using donor platelets from individuals known to be highly reactive to the antibody, the sensitivity of the test is reported to exceed 80%. Using washed platelets (prepared as for the two-point 14C-serotonin release assay [see later discussion]) in place of platelet-rich plasma for the aggregation assay may increase its sensitivity and specificity (heparin-induced platelet aggregation [HIPA] test) The currently accepted reference standard for the laboratory diagnosis of HIT is the two point 14C-serotonin release assay. In this assay, radiolabeled, washed platelets from reactive donors are incubated with heat-treated patient serum in the presence of heparin. The test is positive if 14C-serotonin release occurs at therapeutic (0.1 U/ml) but not high (100 U/ml) heparin concentrations. It is technically demanding, uses radioactivity and is timing consuming, and therefore usually performed in a reference laboratory and used to confirm rather than to make the diagnosis of HIT.
Clinical Presentation of HIT. Part 2
Less common manifestations. Premonitory symptoms. Skin necrosis occurring at sites of subcutaneous administration, thought to be mediated by microvascular thrombosis, may precede the development of thrombocytopenia. Administration of intravenous heparin to patients with these lesions may be catastrophic, with the development of severe systemic thromboembolic complications. Acute systemic reactions (resembling acute febrile transfusion reactions) to intravenous heparin bolus therapy have been described in patients with preexisting dermal lesions at sites of subcutaneous heparin administration. There have also been several reports of unexplained abdominal or limb pain, thought to be ischemic in origin, heralding thrombosis.
Heparin resistance. Impending events may also be heralded by the development of increasing heparin requirements to maintain adequate anticoagulation. Although the mechanism underlying this heparin resistance has not been elucidated, it may reflect platelet release of PF4, which has heparin-neutralizing properties, or it may be mediated by the production of platelet-derived microparticles with procoagulant activity.
Hemorrhage. Despite the presence of thrombocytopenia, hemorrhage is less common than thromboembolic complications in these patients, and there is no relation between the degree of thrombocytopenia and the likelihood of bleeding complications. However, when they occur, bleeding complications may be severe and have included intracranial, retroperitoneal, gastrointestinal or adrenal hemorrhage. Patients undergoing bypass surgery may be more prone to the bleeding manifestations of HIT.
Impact of HIT on patients with cardiovascular disease.
There have been several prospective studies investigating the frequency of HIT in patients receiving heparin in the coronary care unit or for coronary artery disease. The reported combined frequency of HIT types I and II ranged from 0% to 24%, and none of the patients reported in these studies experienced thrombotic events. Several early reports cited patients who developed deep venous thrombosis or pulmonary embolus associated with thrombocytopenia while receiving heparin for thromboprophylaxis after acute myocardial infarction. Heparin therapy has been used more aggressively in recent years, with the recognition of the integral role of intracoronary thrombus formation in the development of the acute coronary syndromes. In a recent analysis of 1,000 patients enrolled in four consecutive prospective thrombolytic trials, thrombocytopenia defined as either platelet counts ,100,000/ml or less than half baseline, occurred in 16.4% of patients, and these patients had a higher in-hospital mortality rate and a more complicated hospital course than patients without thrombocytopenia. In a significant number of these patients the fall in the platelet count occurred after 5 to 10 days (median 4) of heparin therapy, suggesting that HIT contributed to some of the detected thrombocytopenia. Patients undergoing bypass surgery for coronary artery disease have usually been exposed to heparin previously and require large doses for thromboprophylaxis during bypass surgery. Retrospective studies have reported a frequency of HIT in these patients of between 0.5% and 1.9%. In one of these reports, late recognition of the condition (i.e., postoperatively) was associated with a high frequency of bleeding (53%), arterial and venous thromboembolic complications (44%) and death (33%). Heparin is also routinely used during percutaneous coronary interventions, and these patients, too, have usually been exposed to the drug previously. Several patients have been described in whom an acute myocardial infarction occurred during coronary angioplasty in association with the abrupt onset of HIT. An association between femoral artery or venous punctures for cardiac catheterization, angiography or insertion of intraaortic balloon counterpulsation devices and arterial or venous thrombosis in the instrumented limbs of patients with HIT has been reported.
Clinical Presentation of HIT
Thrombocytopenia. Thrombocytopenia usually commences between 6 and 12 days after instituting therapy. The platelet count typically drops to between 30,000 and 55,000/ml, a feature that distinguishes HIT from other forms of druginduced immune thrombocytopenia, which are typically associated with more precipitous falls in the platelet count. However, counts as low as 5,000/ml have been reported in patients with HIT
Thrombosis. Thromboembolic events are always accompanied by a fall in the platelet count, usually to thrombocytopenic levels, and the mean platelet count in patients with HIT who develop thromboses tend to be lower than that in those who do not develop thromboses . However, in any individual patient the risk of a thrombotic event cannot be predicted from the platelet count, and complications in the absence of absolute thrombocytopenia have been reported. The most common thrombotic complications attributed to HIT have been arterial events resulting in lower limb ischemia, cerebrovascular accident or myocardial infarction. Arterial thrombotic episodes at many other sites have also been reported, including arterial graft thrombosis, upper limb arterial thrombosis, renal artery thrombosis, mesenteric ischemia spinal artery thrombosis, aortic thrombosis splenic infarction, thrombosis of cardiac chambers and adrenal thrombosis resulting in acute adrenal insufficiency. These events, when they occur, often have serious sequelae; it has been estimated that patients with HIT and arterial thrombosis have a mortality rate of ;30% and have a risk of leg amputation of ;20% Extension of venous thrombosis with recurrent pulmonary embolism has frequently been described in patients after onset of HIT. Venothrombotic events may be severe—for instance, the progression of venous thrombosis to phlegmasia cerulea dolens, in some cases requiring amputation.
Other reported venous thrombotic events include renal vein thrombosis, cerebral venous thrombosis and thrombosis of the upper limb veins. Thromboses frequently occur at multiple sites, and arterial and venous events may occur together.
Predictors of risk and nature of thrombotic event. A number of studies have suggested that thrombotic events are more likely to occur at sites of preexisting pathology. In one reported series of 25 patients developing HIT and arterial or venous thrombosis, 19 had passage of an arterial or venous catheter or an intraaortic balloon ounterpulsation device in the affected extremity. Similarly, Singer et al. reported a series of four patients who developed thrombotic complications of HIT requiring limb amputation after cardiopulmonary bypass, all of whom had previous intravascular devices at those sites. A recent detailed review of 53 patients with HIT identified an association between cardiovascular disease (myocardial infarction or cardiovascular surgery) and arterial thrombotic events. These investigators also noted a recent operation to be strongly associated with venous thrombosis. In patients receiving heparin after a cerebrovascular accident, 7 of 21 patients with a significant fall in the platelet count experienced an arterial thrombotic event, either extension of their cerebrovascular accident or a myocardial infarction. There were no venous thrombotic events reported in this patient group.
Heparin-Induced Thrombocytopenia. Part 2
Platelet activation. In HIT, IgG binds to the antigenic complex (usually heparin–PF4) on the platelet surface, with the activation of platelets through the Fc receptor (FcgRII). Platelets from patients with HIT express increased numbers of Fc receptors on their surface. These platelets show an increased tendency to aggregate in the presence of heparindependent antibodies. Recent evidence suggests that genetic polymorphisms of the Fc receptor, specifically the presence of the Fcg RIIA allele, may be associated with a predisposition to HIT. Binding of the HIT-IgG to the Fc receptor results in platelet granule release, thromboxane generation, strong platelet aggregation and microparticle release (see late discussion).
A recent report suggests that both IgM and IgA can induce HIT complicated by thrombosis. As platelets are not known to carry a receptor for IgM or IgA Fc, this suggests that in these instances, platelet activation occurs through Fcindependent mechanisms.
Activation of coagulation. Depletions in the levels of the coagulation factor inhibitors protein C, heparin cofactor II and antithrombin III have been reported in patients during the thrombocytopenic phase of HIT and normalize after resolution of the thrombocytopenia. Disseminated intravascular coagulation is also seen, although not commonly
The heparin-associated platelet activation and aggregation consequent on production of the antibody are associated with the generation of platelet-derived microparticles with procoagulant activity, and this, possibly together with endothelial activation (see later discussion), may contribute to the generalized activation of the coagulation cascade.
Endothelial cell activation. In vitro studies have provided suggestive evidence that endothelial cell activation may contribute to the pathophysiologic sequelae of HIT. Serum samples from patients with HIT have increased amounts of IgG, IgM and IgA on endothelial cells, stimulating the release of tissue factor. The binding of both IgG and IgM HIT antibodies to endothelial cells requires the presence of PF4 but not heparin. Binding is reduced when the cells are incubated with heparinase, suggesting that this interaction is mediated in part by endogenous heparin-like molecules such as heparan sulfate found on the surface of endothelial cells.
However, the in vivo contribution of endothelial cell activation and procoagulant activity to the thrombotic sequelae of HIT has been challenged for several reasons.
Pathophysiology of HIT syndrome. In susceptible subjects, heparin bound to PF4 results in an antibody response. An IgG antibody directed against a heparin–PF4 complex binds platelets through the Fc receptor. This results in platelet activation and microparticle formation. Platelet-rich thrombi form at sites of preexisting pathology (arterial thrombotic events in patients with atherosclerotic vascular disease) or sites of endothelial injury (e.g., deep venous thrombosis in patients undergoing orthopedic surgery) risk of thrombosis generally decreases quickly after cessation of heparin, despite the fact that the antibody can often be demonstrated for several weeks. In addition, thromboemboli tend to be located at sites of vascular injury rather than being diffusely distributed throughout the vascular tree suggesting that endothelial disruption rather than integrity may provide the focus for the thrombogenic response.
Heparin-Induced Thrombocytopenia. Part I
Frequency of Thromboembolic Events
There is conflicting information in the published dataconcerning the frequency of thromboembolic events in patients with HIT. Early retrospective studies reported a frequency rate .60%, although this may have reflected the fact that the condition was often not recognized before complications occurred In prospective studies of HIT, only small numbers of patients have developed thrombotic complications. When data from the adequately designed studies are pooled, the risk of thrombosis in patients with HIT is estimated at 30% The frequency may vary dramatically with the baseline risk of thrombosis in the patient group. In the study of patients receiving heparin for thromboprophylaxis after hip surgery, 8 (88.9%) of the 9 patients with HIT had one or more thrombotic events as compared with 117 (17.8%) of the 656 patients without HIT. Data from a recent, carefully performed, retrospective study suggest that the risk of thrombotic events may persist for several weeks after exposure to heparin . In this study of 62 patients diagnosed initially with isolated HIT, .50% experienced thrombotic complications over the ensuing 30 days.
Pathophysiology
Type II HIT is an immune-mediated reaction most commonly caused by an immunoglobulin (Ig) G antibody that binds to platelets in the presence of heparin and causes platelet activation. Heparin-dependent IgM and IgA antibodies have also been described, and recent evidence suggests that they too may induce type II HIT
The antigen. Despite the important role of heparin in initiating and sustaining the syndrome, it has not been possible to directly demonstrate an interaction between the antibody and heparin alone. In 1992, it was shown that the majority of antibodies generated in HIT recognized a complex between heparin and platelet factor 4 (PF4) . The antibody is not specific for heparin, but reacts with other sulfated polysaccharides bound to PF4. This cross reactivity increases with size and negative charge of the polysaccharide In up to 25% of cases the antibody does not recognize the polysaccharide–PF4 complex, indicating that antigens other than PF4 are responsible for some cases of HIT. These antigens remain unidentified.
The antibody response. The heparin-dependent IgG generally becomes detectable after 5 to 7 days of heparin therapy. Patients preexposed to heparin may develop HIT sooner after institution of therapy than those receiving heparin for the first time, suggestive of an anamnestic response. However, almost all of these reported events have occurred within days to weeks of the index exposure, when the heparinassociated IgG is usually still detectable in the circulation and is presumably responsible for the thrombocytopenia. It is noteworthy that patients with cardiac disease requiring percutaneous or surgical interventions have frequently been exposed to heparin within this time frame. One case of recurrent HIT, documented years after the initial episode, reported a time course comparable to the index event, consistent with the mounting of a second primary immune response and absence of immunologic memory.
Heparin-Induced Thrombocytopenia. Introduction
Since it was first used to prevent venous thrombosis in surgicalpatients in 1937, heparin has been the anticoagulant therapy ofchoice for the prevention and early management of arterialand venous thromboembolic disease. Heparin plays a major role in cardiovascular therapy; it is the routine anticoagulant agent used in percutaneous transluminal coronary angioplasty and coronary artery bypass graft surgery and has confirmed value in the management of patients with unstable angina and as adjunctive therapy in patients receiving thrombolysis after myocardial infarction.
Because of the vast clinical experience with heparin, physicians are generally well aware of most of the practical limitations associated with its use, including the potential for bleeding and interpatient variability in anticoagulant response after administration. The focus of this report, however, is one of the potentially most devastating but least well recognized complications of heparin therapy—its propensity to cause thrombocytopenia with associated thrombosis. With the more frequent use of heparin in recent years, the prevalence of this potentially catastrophic complication has prominently increased
Frequency of Thrombocytopenia
Approximately 10% to 20% of patients will develop a transient mild thrombocytopenia after 1 to 4 days of heparin therapy. The platelet count rarely goes below 100,000/ml, often becomes normal despite further heparin administration and is generally without clinical sequelae. This benign form of heparin-induced thrombocytopenia (HIT), attributed to a direct interaction between administered heparin and circulating platelets, is classified as HIT type I The more sinister form of HIT generally develops after 5 to 10 days (mean 7 to 8 ) of heparin therapy. This form is antibody mediated, can be associated with potentially devastating thromboembolic sequelae and is classified as HIT type II and is the topic of this report. A recent prospective, randomized study of 665 patients receiving either subcutaneous unfractionated or low molecular weight (LMW) heparin for thromboprophylaxis after hip surgery reported a frequency rate of HIT of 2.7% in patients receiving unfractionated heparin and 0% in patients receiving LMW heparin. HIT was defined as a decrease in the platelet count ,150,000/ml occurring at least 5 days after beginning heparin therapy, accompanied by heparindependent antiplatelet antibodies detected by the sensitive 14C-serotonin release assay Unfortunately there have not been any comparably designed, prospective, large-scale trials investigating the frequency of HIT in patients receiving full-dose intravenous heparin. However, pooled analyses derived from less rigorous prospective studies suggest that HIT becomes more common after full-dose intravenous heparin than after subcutaneous heparin.
The reported frequency of HIT type II in this patient group ranges from 0% to 30%, reflecting the varying and varying definitions of thrombocytopenia. Although HIT appears to be more common after administration of larger doses of heparin, it may still occur after exposure to very small quantities of the drug. The condition has been described in patients receiving small amounts of heparin in flush solutions used to maintain the patency of indwelling peripheral catheters and in patients with heparin-coated pulmonary artery catheters A number of investigators have observed that patients previously exposed to heparin develop HIT earlier after institution of heparin therapy compared with those with no previous exposure, suggestive of an anamnestic response, although the objective evidence for the existence of immunologic memory in these patients is not strong (see later discussion).
The small, prospective studies to date have not demonstrated an association between previous treatment with heparin and an increased likelihood of developing HIT (21), and it is anticipated that ongoing, adequately sized, prospective trials will provide a clearer perspective of the impact of previous exposure to heparin on the frequency of this syndrome. Most prospective studies comparing porcine mucosal with bovine lung heparin have reported a greater frequency of HIT among patients receiving the latter Individual batches of heparin may have an increased propensity to cause thrombocytopenia even when derived from the same source, giving rise to “epidemics” of HIT Thrombocytopenia and thrombosis are not unique to heparin and have also been described after administration of other sulfated glycosaminoglycans such as LMW heparin, pentosan polysulfate and chondroitin sulfate. The frequency with which it occurs appears to be directly related to the size and degree of sulfation of the molecule.
Smokers paradox
Fonarow et al. have explored the ‘smoker’s paradox’ in a novel inception cohort of hospitalized heart failure patients included in the OPTIMIZE-HF registry, by determining the impact of current or recent smoking on outcomes during hospitalization and in the first 60–90 day period following hospital discharge.
Adjustments for known important potential confounders were performed. Smokers were more likely to have left ventricular systolic dysfunction (59 vs. 51%), were more likely to be male (61 vs. 46%), and had higher B-type natriuretic peptide (BNP) levels (895 vs. 789 pg/mL). However, smokers were much younger (63 vs. 75 years of age), were less likely to have a previous history of heart failure admission (84 vs. 89%), and were less likely to have a diagnosis of diabetes requiring insulin therapy (14 vs. 17%). Despite this, when compared with non-smokers, the risk of in-hospital mortality was ∼30% lower for current or recent smokers, while the risk for re-hospitalization or post-discharge mortality was similar between the two groups.
How could this be? Assuming the authors appropriately assembled an inception cohort, correctly classified smokers, and appropriately adjusted for important potential confounders, only one of two conclusions can be drawn: either smoking is somehow ‘protective’ in heart failure patients; or a substantial component of the adverse effects of smoking in heart failure is reversible and has a short half-life (i.e. days) so that ‘enforced’ in-hospital smoking cessation led to an imbalance of beneficial therapies in the smokers (usual therapy plus smoking cessation) compared with their non-smoking cohorts (usual therapies only). The former hypothesis (that smoking is protective) can probably be refuted immediately. There is overwhelming evidence that smoking increases the incidence of heart failure, as well as heart failure morbidity and mortality (including sudden death) for those who have heart failure.
It can be argued that there is substantially more merit in the second argument. Constituents of inhaled tobacco damage the cardiovascular system in numerous ways: producing endothelial and platelet dysfunction; enhancing coagulation; increasing heart rate and blood pressure; increasing myocardial oxygen demand and consumption; and inducing vasoconstriction. Cigarette smoking also significantly increases carboxyhaemoglobin production, impairing oxygen carriage and release, which has a negative inotropic effect, increasing left ventricular end-diastolic pressure. Thus the abrupt removal of these deleterious effects of cigarette smoking following hospital admission and enforced smoking cessation, in addition to the provision of the usual heart failure therapies in smokers, might well be expected to ‘outstrip’ the impact of usual therapies alone in non-smokers.