Pathophysiology and Diagnosis of Heparin-Induced Thrombocytopenia

Heparin is one of the most widely used parenteral drugs in modern medicine, with approximately 12 million patient exposures annually in the United States alone. Although generally safe and effective, heparin use is associated with several side effects, the most potentially serious of which is heparin-induced thrombocytopenia (HIT). This condition affects between 0.1% and 3% of patients exposed to heparin products.

The variable frequency is due to several factors, including the type of heparin used (low-molecular-weight heparin [LMWH] is less immunogenic than unfractionated forms), route of administration (IV heparin is more likely to cause HIT than that given subcutaneously), and the type of laboratory test used to confirm the diagnosis.

HIT is an immune-mediated reaction caused by an antibody to the complex formed between heparin (H) and platelet factor 4 (PF4) that is released from activated platelets (Kelton et al. Blood. 1994;83[11]: 3232). IgG forms of the H-PF4 antibody bind to platelets via their FcIIa receptors, resulting in intense platelet activation, release of highly procoagulant microparticles, and, ultimately, intravascular thrombin formation. Thrombin causes additional platelet activation and fibrin clot formation and a worsening cycle of serious hypercoagulability that demands early recognition and prompt and appropriate treatment.

Despite the apparent etiology of HIT, the condition is not a typical immune response. For example, the IgG antibody response is relatively rapid and may occur without IgG class precedence; the antibody persistence is limited, and repeat heparin exposure often does not restimulate antibody production (Selleng et al. Transfusion. 2009;49[9]:1812). H-PF4 antibodies may be generated in patients with lupus and antiphospholipid syndrome, and HIT-like syndromes with positive laboratory tests have been reported in patients with acute infectious diseases without heparin exposure. Indeed, there is evidence that H-PF4 antibodies are part of an innate bacterial defense system (Greinacher J. Thromb Haemost. 2009;7[suppl 1] 9).

The clinical diagnosis of HIT should be considered if the platelet count falls by 50% or greater, with or without venous or arterial thrombosis, in a patient currently or recently receiving heparin therapy. Such patients may have other reasons for thrombocytopenia or thrombosis and do not have HIT. Nevertheless, since HIT can present in several diverse ways, it is important to include it in the differential diagnosis.

HIT may present systemically or locally with skin lesions at the heparin injection site. Most HIT cases (~70%) present 4 to 10 days after heparin exposure (Warkentin TE. Semin Hematol 1998;35[4 suppl 5]:9), but a rapid onset (within 24 h) form of HIT (~30% of cases) may occur in patients with an existing circulating H-PF4 antibody (Warkentin and Kelton. N Engl J Med. 2001;344[17]:1286).

Because these antibodies are usually transient—90% have disappeared from circulation within 3 months—rapid-onset HIT is typically associated with recent heparin exposure. Indeed, circulating H-PF4 antibodies are found in 10% of patients presenting to the ED with chest pain or symptoms of thrombosis and a history of recent (within 6 months) hospitalization (Francis et al. Am J Emerg Med. 2007;25[3] :279). Such patients are at higher risk of developing rapid onset HIT. A minority of patients present 2 to 6 weeks after heparin therapy with delayed-onset HIT (Rice et al. Ann Intern Med. 2002;136[3]: 210) that ‘can be particularly difficult to recognize, as a history of heparin exposure may be elusive.

The recognition of HIT largely depends on regular monitoring of the platelet count (Warkentin et al. Chest. 2008;133[suppl]:340S). Platelets should be measured at least every other day between days 4 and 14, in postoperative patients receiving therapeutic unfractionated heparin, while earlier monitoring is recommended for patients with a history of recent (within 3 months) heparin exposure, because of the risk of rapid onset HIT.

Postoperatively, patients receiving LMWH require monitoring every 2 to 3 days after day 4, while medical/obstetrical patients receiving LMWH do not require regular monitoring.

Using thrombocytopenia as a marker of HIT is problematic in the cardiac surgery setting, as significant platelet decreases occur in most patients. The platelet count typically begins to rise by the second or third day after cardiac surgery, but most patients do not develop the H-PF4 antibody until day 4 or 5. Thus, this early postoperative fall is not a sign of HIT without other supportive clinical signs.

In contrast, a secondary fall in the platelet count, or a failure of the platelet count to recover, may indicate HIT (Pouplard et al. Br J Haematol. 2005; 128[6]:837). Thrombocytopenia is also common in the ICU, although the incidence of HIT is less than 1%, despite positive H-PF4 antibody test results in 10 to 30% of medical, neurotrauma, or shock-trauma patients in the ICU (Levine et al. J Thromb Thrombolysis. [published online ahead of print, November 13, 2009]).

The laboratory diagnosis of HIT rests on the demonstration of a circulating H-PF4 antibody. The most widely available test is the enzyme-linked immunoassay (ELISA), although this assay carries a high rate of false-positive results, the scale of which depends on the clinical scenario. For example, 40 to 60% of postcardiac surgery patients develop a positive ELISA in the absence of clinical HIT (Francis et al. Ann Thorac Surg. 2003;75[1]:17).

Thus, the specificity of the ELISA for clinical HIT is relatively low in this setting, and, as discussed below, the physician must interpret the result in the light of the pretest clinical findings. On the other hand, its high sensitivity for H-PF4 antibodies means that a negative test result virtually excludes the diagnosis.

Recently, rapid immunoassays suitable for the near-patient setting have become available. One of these, the PIFA® heparin/PF4 immunoassay is approved for use in the United States, but, in the author’s experience, correlates very poorly with the ELISA.

Many laboratories report ELISA test results simply as positive or negative. However, the diagnostic specificity is increased by knowing the actual optical density value, as high values (eg, >1.5) are more likely to be associated with a platelet-activating antibody and, thus, clinical HIT. Most commercially available ELISAs detect H-PF4 antibodies of IgG, IgA, and IgM specificities. However, since platelet activation in HIT is dependent on their FcIIa (IgG) receptors, detection ‘of only IgG antibodies provides better diagnostic information without loss of sensitivity (Bakchoul et al. J Thromb Haemost. 2009;7[8]:1260).

The highest specificity for HIT is yielded by the serotonin release assay, which is based on the ability of H-PF4 antibodies to activate normal donor platelets in the presence of therapeutic amounts of heparin. Although the serotonin release assay is the “gold standard” of HIT tests, it is a complicated assay that is available in only a few centers.

Because of their low specificity, the ELISA results must be interpreted in light of the pretest probability of HIT.

The “4-T” score is a clinically validated assessment tool that takes into account the presence and timing of thrombocytopenia and thrombosis and whether other causes of these are present (Warkentin. Br J Haematol. 2003; 121[4]:535). Testing should only be performed when there is clinical evidence of HIT, since there is a danger of overdiagnosis if a positive ELISA is considered confirmatory in the absence of supportive clinical findings (Lo et al. Am J Hematol. 2007;82[12]:1037).

In patient populations where thrombocytopenia and heparin use are common, for example, in the ICU, the 4-T score, preferably coupled with knowledge of the ELISA optical density, an IgG-specific immunoassay, or serotonin release assay test, will be helpful in the differential diagnosis.

Because of its higher specificity, a strongly positive serotonin release assay is far more likely to indicate HIT than a weakly positive ELISA result, although a negative ELISA result has a high negative predictive value.

Repeating the ELISA in patients with low pretest probabilities of HIT is also problematic, as a significant number with an initially negative test result subsequently retest as positive (Chan et al. Am J Hematol. 2008;83[3]:212). Although repeat laboratory testing is justified if the pretest probability increases, there is a danger of overdiagnosis in scenarios where a positive ELISA is common without clinical symptoms of HIT (ie, open heart surgery).

In our experience, requests for HIT testing often coincide with the platelet nadir that occurs 2 to 3 days after heart surgery, yet most (>90%) of these test results are negative. However, if the test is repeated 2 to 3 days later, many patients will test positive, even if the platelet count is recovering, because of the natural history of H-PF4 antibody formation in this setting.

There is a similar risk of overdiagnosis in patients in the noncardiac surgical ICU, where the frequency of positive ELISA test results may increase almost threefold within 7 days of admission (Levine et al., Thrombosis and Haemostasis, in press).

In summary, HIT is a clinicopathologic syndrome resulting from the formation of antibodies against a neoantigen formed when heparin binds to PF4. This results in one of the most hypercoagulable states known to clinical medicine. Early recognition and prompt treatment are the cornerstones of effective management. The most widely available laboratory tests for HIT lack specificity and should always be inter-preted with the clinical picture in mind.

Dr. John L. Francis
Director, Florida Hospital Center for Thrombosis Research
Professor of Medical Education
University of Central Florida
College of Medicine
Orlando, FL