May 2009: Volume 35, Number 5
How Lab Results Contribute to Clinicopathologic Diagnosis
Kristi J. Smock, MD, and George M. Rodgers, MD, PhD
Heparin has been used as an anticoagulant medication since the late 1930s and is one of the most prescribed drugs in the U.S. However, clinicians recognized that some patients developed a syndrome of immune-mediated thrombocytopenia and thrombosis, which came to be called heparin-induced thrombocytopenia (HIT). HIT occurs when IgG antibodies develop against neoantigens created by multimolecular heparin/platelet factor 4 (PF4) complexes. The antibodies activate platelets, resulting in thrombocytopenia and thrombosis. In HIT, therefore, thrombocytopenia can be considered an indicator of platelet activation and a procoagulant state. Inflammatory cytokines and tissue factor released from activated endothelium and white blood cells may also contribute to the thrombotic tendency.
But diagnosis of HIT is not simple or as straightforward as physicians expect. Consider this scenario. A 70-year-old cardiothoracic surgery patient developed thrombocytopenia during heparin therapy. His physician orders an ELISA test for heparin-induced platelet antibodies. The quantitative ELISA results just exceed the cutoff for positivity. The physician calls the lab to ask, “Does my patient have heparin-induced thrombocytopenia?” As you will see, the answer to this question requires additional information. This brief review addresses the use and interpretation of lab tests used for HIT diagnosis.
Since HIT is a clinicopathologic diagnosis, lab results must be interpreted in conjunction with clinical information. Figure 1 demonstrates the pathogenesis of HIT. Thrombotic events occur in up to 50% of affected patients and can be venous or arterial. Lab testing for HIT falls into two general groups: immunoassays that identify heparin/PF4 antibodies and functional assays that evaluate the capacity of heparin/PF4 antibodies to activate platelets. Patients with HIT receive anticoagulation therapy with direct thrombin inhibitors rather than heparin or related drugs.
PF4 is a chemokine that plays a role in immune responses and angiogenesis. Heparin/PF4 antibodies usually form 5 to 10 days after initiation of heparin therapy. Although many hospitalized patients are treated with heparin, formation of heparin/PF4 antibodies is more common in certain patient subgroups. Surgical patients develop the antibodies more often than medical or obstetric patients because surgery results in platelet activation and PF4 release from platelet alpha granules. In general, antibodies can be identified in up to 15% of orthopedic surgical patients and up to 50% of cardiopulmonary bypass patients.
However, not all antibodies activate platelets, and most patients with heparin/PF4 antibodies do not have clinical HIT. HIT occurs in approximately 2 to 3% of orthopedic surgery patients and approximately 1% of cardiopulmonary bypass patients.
Unfractionated heparin (UFH) stimulates formation of the antibodies that cause HIT more so than low molecular weight heparin (LMWH). Therapeutic heparin doses are not required to stimulate antibody production; even small amounts of heparin can result in antibody formation and clinical HIT. Most antibodies formed in response to UFH will cross-react with LMWH.
Physicians usually obtain patients’ baseline platelet counts before initiating heparin therapy. A falling platelet count during therapy raises the suspicion of HIT, and in the usual presentation, the platelet count begins to drop after 5 to 10 days of heparin therapy. Patients with recent heparin exposure (within 30 days) may have platelet counts that fall before day 5, often within the first 24 hours. Delayed-onset cases also occur but are uncommon.
Platelet count should not dip below 15,000–20,000/μL. Decreases below 50% of the baseline count are typical, even if the patient does not develop absolute thrombocytopenia. These relative decreases are an important consideration because platelets are an acute phase reactant, and some patients with suspected HIT have received platelet transfusions. However, patients often have a variety of potential reasons for a falling platelet count other than HIT, including bleeding, infection, medications, and hemodilution. In HIT, other routine coagulation tests, such as PT, fibrinogen, and D-dimer, are generally normal to mildly abnormal. PTT is often elevated due to the heparin.
Clinical Scoring Systems
The complexity of HIT diagnosis has led to the development of clinical scoring systems. One example is the 4T scoring system to assess the pre-test probability of HIT in at-risk patients. This system considers four main factors: 1) Thrombocytopenia (the percent platelet decrease and nadir platelet count); 2) Timing of platelet fall; 3) presence of Thrombosis or other sequelae (skin necrosis at heparin injection sites, or acute systemic reactions to heparin bolus); and 4) other potential causes for Thrombocytopenia. Each feature is given a score and patients are classified as high, intermediate, or low probability. Such clinical scoring systems have good negative predictive value for patients with low scores; however, patients with moderate or high clinical probability usually require lab testing.
Diagnosis of HIT may involve a variety of lab tests.
Immunoassays. Heparin/PF4 antibodies, often called heparin-induced platelet antibodies, are usually detected by ELISA. Complexes of heparin, or a heparin-like substance, and PF4 are immobilized to the ELISA plate and bind the patient’s antibodies. Most commercially available assays detect IgG, IgM, or IgA antibodies, and some assays include a confirmatory step that involves inhibition of a positive reaction in the presence of high heparin concentrations that interfere with antibody binding to the heparin/PF4 complexes.
ELISA results can be reported qualitatively (positive/negative) or quantitatively (optical density, OD), with OD referring to the strength of the optical signal. Labs typically run ELISA tests in batches, with turnaround times up to a day or more depending on the lab’s schedule.
While ELISA assays are sensitive for antibody detection, they are not specific for clinical HIT for a variety of reasons. For instance, most assays detect IgG, IgM, and IgA antibodies, but only IgG antibodies are capable of activating platelets through their FcγRIIa receptors. Whether IgM or IgA antibodies cause the HIT syndrome is controversial. Newer ELISA kits are available that detect only IgG antibodies, but these have not been extensively studied and not all IgG antibodies result in platelet activation.
Researchers have shown that higher OD values in ELISA assays are more likely to represent pathologic antibodies capable of platelet activation, while low positive values closer to most ELISA cutoffs are less likely to be pathologic. In other words, the specificity of ELISA assays is improved if the OD value is considered.
In general, HIT is unlikely in a patient with low clinical pre-test probability and a negative ELISA result. The probability of HIT in patients with a high clinical pre-test probability and a strong positive (high OD) ELISA result is strong. Patients that do not fall into either of these two categories require additional testing by functional platelet activation methods.
Functional assays. Often termed “confirmatory tests”, functional assays evaluate the ability of heparin/PF4 antibodies to activate platelets, resulting in greater specificity for diagnosis of clinical HIT. The serotonin release assay (SRA) is considered the gold standard functional assay. In simple terms, the SRA involves addition of patient serum to washed donor platelets labeled with 14C-serotonin in the presence of both low and high heparin concentrations. The patient serum must be heat-inactivated to destroy residual thrombin, which is a potent platelet activator. At low heparin concentrations, positive sera cause platelet activation and degranulation, detected by release of radioactivity into the supernatant. This platelet activation is inhibited in the presence of excess heparin. Non-radioactive methods to detect released serotonin from donor platelets include high performance liquid chromatography (HPLC) and ELISA. These latter methods are described in the literature, but are not widely available at this time.
The definition of positive and negative SRA results varies depending on the lab performing the test. In the original description of the test, positive patients had at least 20% release (usually much more) in the presence of low-dose heparin, and less than 20% release with excess heparin; but borderline and indeterminate results are not uncommon. Borderline refers to platelet serotonin release that exceeds the negative control, but does not exceed the cutoff for positivity. Indeterminate patterns come from positive serotonin release at both low and high heparin concentrations, suggesting that the platelet activation was not heparin-dependent and could be due to platelet antibodies directed against targets other than heparin/PF4 complexes. Possibilities include autoantibodies, HLA antibodies, alloantibodies directed against platelet-specific glycoprotein targets, or circulating immune complexes. Heat-induced aggregation of IgG in the serum preparation step can also cause indeterminate results.
Although the SRA remains the gold standard test, it is technically difficult to perform, usually requiring radioactivity, and requires a constant source of fresh donor platelets known to be reactive to HIT antibodies. Donor platelet preparation is critical and care must be taken not to induce platelet activation prior to performing the assay. Contamination of the platelet preparation with granulocytes or monocytes can cause false-negative results due to binding of the antibodies to white blood cell Fc receptors, making antibodies unavailable to interact with platelets. The inclusion of weak-positive, strong-positive, and negative controls confirms that the donor platelets can respond to even weak HIT sera. For these reasons, this type of testing is performed mainly at academic centers or reference labs and requires a several day turnaround time.
Similar to the SRA, other functional assays look for platelet activation in the presence of low heparin concentrations and inhibition of activation by excess heparin. Some labs offer heparin-induced platelet aggregation on donor platelet-rich plasma by light transmission aggregometry. The endpoint is platelet aggregation rather than platelet release of radio-labeled serotonin. This method has been shown to be less sensitive and less specific than the SRA.
Flow cytometric methods can detect platelet activation markers, such as P-selectin and annexin V, expressed by donor platelets in the presence of patient sera and heparin. Other flow cytometric methods detect platelet microparticles generated when heparin/PF4 antibodies bind and activate platelets through the FcγRIIa receptor in the presence of heparin. The microparticles are formed from platelet membranes and act as a substrate for thrombin generation. These flow cytometric methods, however, are commonly used in clinical labs.
Rapid immunoassays. Rapid immunoassays can detect heparin/PF4 antibodies in patient serum in an hour or less. In the U.S., one assay, a strictly qualitative particle immunofiltration assay, has been cleared by FDA. Outside the U.S., a gel agglutination assay is available that uses red polystyrene beads coated with heparin/PF4 complexes. To perform the test, the test card is incubated with the patient serum and then centrifuged. Particles cross-linked by antibodies remain at the top of the gel reaction chamber, resulting in a red color. Non-cross-linked particles sink to the bottom, resulting in a colorless reaction chamber. Results from the gel agglutination assay can be semi-quantitative if testing is performed on multiple dilutions of patient serum, and the dilution titer may be more predictive of antibody significance, similar to the increased significance of high OD values in the ELISA tests.
Because the characteristics of the rapid assays are similar to those of ELISA, the results of these tests do not appear to eliminate the need for functional testing in certain groups of patients. But rapid turnaround times are sometimes advantageous. For example, when physicians strongly suspect that a patient has HIT, they should discontinue heparin therapy before the results of central lab tests are available. These patients receive anticoagulation therapy in the form of a direct thrombin inhibitor. In some circumstances, physicians will transition patients with low clinical probability of HIT to a direct thrombin inhibitor.
HITting the Right Diagnosis
Clinical scoring systems are a useful tool to determine the pre-test probability of patients at risk for HIT. Low scores have been shown to have high negative-predictive value, but the positive predictive value of intermediate or high scores is limited. Patients that fall into the latter category require lab tests to make a more definitive diagnosis. Physicians will also sometimes order tests for patients with low clinical probability. Figure 2 demonstrates one possible approach to HIT testing, and Table 1 summarizes the clinical and laboratory features of HIT.
Immunoassays to detect heparin/PF4 antibodies are the usual first-line tests. As stated previously, a negative ELISA combined with low or moderate pre-test probability has good negative predictive value, making a diagnosis of HIT unlikely. Rare false-negative immunoassays have been reported in HIT patients with antibodies directed against antigens other than heparin/PF4, such as interleukin-8 or neutrophil-activating peptide-2. A strong-positive ELISA result in a high-probability patient is strongly suggestive of HIT.
Evaluation of Heparin-Induced Thrombocytopenia
Thrombocytopenia (percent platelet decrease, nadir platelet count) es
- Baseline platelet count necessary to accurately interpret decrease
- Patient may not develop absolute thrombocytopenia
- Platelets are an acute phase reactant
- Some patients have received platelet transfusions
Timing of platelet count fall
- Usually begins on day 5 to 10 after initiating heparin therapy
- May occur rapidly in patients with recent heparin exposure
- Rare delayed-onset cases
Thrombosis, skin necrosis at heparin injection sites, allergic reactions to heparin bolus
- Patient may have thrombosis due to underlying medical condition
Other causes for
- Other potential causes for thrombocytopenia are common in hospitalized patients (DIC, other drugs, etc.)
Immunoassays (ELISA, rapid assays)—Usual first-line test
- High sensitivity but poor specificity for HIT diagnosis
- Most detect IgG, IgM, or IgA antibodies but only IgG cause HIT
- Not all antibodies are capable of platelet activation
- High OD values more significant than low OD values, improving specificity
- Complex test with moderately long turnaround times
- Rapid assays not widely available
Functional platelet activation assays (SRA, others)—Confirmatory tests
- Greater specificity for HIT than immunoassays
- Technically demanding tests performed primarily at academic institutions/reference labs
- Long turnaround times
*Refer to Lo et al. J Thromb Haemost 2006;4:759–65 for a description of the 4 T’s clinical scoring system.
Abbreviations: HIT, heparin-induced thrombocytopenia; ELISA, enzyme-linked immunosorbent assay; OD, optical density; SRA, serotonin release assay; DIC, disseminated intravascular coagulation.
Between these two extremes are patients who need to be evaluated further. Functional assays are the second line of testing for these patients, which includes ELISA-positive patients who have only intermediate clinical probability and ELISA-negative or weak-positive patients in whom the clinical suspicion is high. Functional platelet activation assays such as the SRA are useful because of their greater specificity, allowing HIT to be either excluded or confirmed in many patients. If the SRA results are indeterminate, borderline, or don’t make clinical sense, however, clinical re-evaluation is necessary and testing may need to be repeated.
An important point to remember is that ELISA results will be positive in many patients who do not have clinical HIT, especially surgical patients without suggestive clinical features. Such results could be considered false positives for a HIT diagnosis because the antibodies detected in these patients do not activate platelets.
As for the 70-year old cardiothoracic patient described earlier in this review, the answer to whether he had HIT—a very common question physicians face—is not always straightforward. For this particular patient, the clinical pre-test probability was low. He had only a 20% drop in platelets from the baseline that occurred 2 days after beginning heparin. He also had no previous heparin exposures, no evidence of thrombosis, and there were other more plausible causes for his thrombocytopenia. Furthermore, his ELISA result was only weakly positive. Correlation of the clinical presentation, combined with knowledge of the proper interpretation of lab tests would exclude a HIT diagnosis for this patient.
Alberio L. Heparin-induced thrombocytopenia: some working hypotheses on pathogenesis, diagnostic strategies and treatment. Curr Opin Hematol 2008;15:456–64.
Greinacher A, Juhl D, Strobel U, Wessel A, et al. Heparin-induced thrombocytopenia: a prospective study on the incidence, platelet-activating capacity and clinical significance of antiplatelet factor 4/heparin antibodies of the IgG, IgM, and IgA classes. J Thromb Haemost 2007;5:1666–73.
Lo GK, Juhl D, Warkentin TE, Sigouin CS, 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;4:759–65.
Lo GK, Sigouin CS, Warkentin TE. What is the potential for overdiagnosis of heparin-induced thrombocytopenia? Am J Hematol 2007;82:1037–43.
Pouplard C, Gueret P, Fouassier M, Ternisien C, et al. Prospective evaluation of the '4Ts' score and particle gel immunoassay specific to heparin/PF4 for the diagnosis of heparin-induced thrombocytopenia. J Thromb Haemost 2007;5:1373-9.
Price EA, Hayward CP, Moffat KA, Moore JC, et al. Laboratory testing for heparin-induced thrombocytopenia is inconsistent in North America: a survey of North American specialized coagulation laboratories. Thromb Haemost 2007;98:1357–61.
Sheridan D, Carter C, Kelton JG. A diagnostic test for heparin-induced thrombocytopenia. Blood 1986;67:27–30.
Warkentin TE. Platelet count monitoring and laboratory testing for heparin-induced thrombocytopenia. Arch Pathol Lab Med 2002;126:1415–23.
Warkentin TE. New approaches to the diagnosis of heparin-induced thrombocytopenia. Chest 2005;127:35S–45S.
Warkentin TE, Sheppard JI, Moore JC, Sigouin CS, et al. Quantitative interpretation of optical density measurements using PF4-dependent enzyme-immunoassays. J Thromb Haemost 2008;6:1304–12.
Kristi J. Smock, MD, is assistant medical director of the Hemostasis/Thrombosis Laboratory at ARUP Laboratories, and assistant professor of pathology at the University of Utah, Salt Lake City.
George M. Rodgers, MD, PhD, is medical director of the Hemostasis/Thrombosis Laboratory at ARUP Laboratories, and professor of pathology and medicine at the University of Utah, Salt Lake City.