Chief problem: Development of thrombocytopenia on day 6 of hospitalization.
History: A 46-year-old woman with a history of chronic obstructive pulmonary disease (COPD) was admitted to the ICU with the diagnosis of acute-on-chronic, hypoxemic, hypercapnic respiratory failure and health care-associated pneumonia. Initial interventions included mechanical ventilation for four days and intravenous antibiotics. Heparin, 5000 U/8 h, was given for deep venous thrombosis (DVT) prophylaxis. Complete blood counts were obtained every three days. Admission platelet count was 224 × 109 cells/L, but by day 3 it had dropped to 175 × 109 cells/L, and on day 6, platelet count was noted to be 96 × 109 cells/L.
The patient's medical history was significant for hypertension, chronic hypoxemic respiratory failure, and COPD. She had a 50-pack-year smoking history but reported no use of alcohol or illicit drugs. She had no known drug allergies. Hospital medications included the following: vancomycin, 1500 mg/12 h; levofloxacin, 750 mg/d; cefepime, 2 g/12 h; pantoprazole, 40 mg/d; enalapril, 10 mg/d; ipratropium/albuterol, 0.5 mg/2.5 mg/3 mL nebule every 4 hours; and heparin, 5000 U/8 h.
Physical examination: Vital signs were as follows: blood pressure, 125/75 mm Hg; heart rate, 88 beats/min; temperature, 98.7° F; respiratory rate, 18 breaths/min; and oxygen saturation (SpO2) 92% on patient's baseline supplemental oxygen at 2 L/min. Physical exam was largely unremarkable except for crackles at bilateral lung bases. There were no signs or symptoms suggestive of arterial or venous thrombosis.
Labs: Hemoglobin level was 13.3 g/dL, hematocrit was 40%, and white blood cell count was 13.3 × 109 cells/L. Blood and sputum cultures obtained on admission showed no growth.
Once progressive thrombocytopenia was identified, multiple laboratory studies were ordered. Peripheral smear revealed thrombocytopenia with no schistocytes. Partial thromboplastin time, prothrombin time, international normalized ratio (INR), lactate dehydrogenase level, haptoglobin level, and complete metabolic panel were within normal limits. A presumptive diagnosis of heparin-induced thrombocytopenia (HIT) was made based on a “4 T's” score of 5 (see Table). Blood was sent to a reference laboratory for anti-PF4/heparin ELISA. Heparin was discontinued, and the patient was placed on sequential compression devices for ongoing DVT prophylaxis.
Two days later, the platelet count had dropped to 43 × 109 cells/L, and the anti-PF4/heparin ELISA result was positive, with an optical density (OD) of 2.926. The patient was started on argatroban, and a serotonin release assay was sent to a reference lab for confirmatory testing. Ultimately, the serotonin release assay result was positive, and the patient was maintained on argatroban for one week until her platelet count increased to 104 × 109 cells/L, at which time she was placed on oral apixaban, 5 mg twice daily, and was discharged to complete an additional three weeks of therapy.
Four days later, the patient returned to the hospital with acute-on-chronic hypoxemic respiratory failure. A CT angiogram of the chest revealed bilateral pulmonary emboli.
Q: Was our patient diagnosed correctly?
A: Yes. The 4 T's score indicated that the patient had an intermediate probability of HIT. An anti-PF4/heparin ELISA was obtained and returned as strongly positive. Subsequently, a functional assay (serotonin release assay) was obtained to confirm the diagnosis and was positive.
Q: Was our patient managed and discharged appropriately?
A: No. While the 4 T's score was used appropriately to determine the probability of HIT, the patient was managed suboptimally for the following reasons:
- The patient should have been started on argatroban as soon as it was determined that clinical probability for HIT was intermediate or higher (4 T's score ≥4) (1, 2).
- Argatroban should have been continued until the platelet count was 150 × 109 cells/L or higher (1, 3).
- The patient should have been transitioned from argatroban to warfarin prior to discharge, with a minimum of five days' overlap of the two drugs (1, 2, 3). Currently, none of the direct-acting oral anticoagulants (DOACs) available in the United States (edoxaban, dabigatran, apixaban, and rivaroxaban) are FDA-approved for the treatment of HIT.
- Bilateral lower-extremity compression ultrasonography could have been performed to rule out occult DVT, since the presence of a DVT would influence treatment duration of warfarin at discharge and change the diagnosis to heparin-induced thrombocytopenia and thrombosis (HITT) (1).
Heparin-induced thrombocytopenia has a prevalence of 0.1% to 5% in patients receiving heparin (4, 5). It is an adverse drug reaction that occurs when heparin-dependent IgG antibodies bind to heparin/platelet factor 4 complexes to activate platelets. The result is a hypercoagulable state that leads to thrombocytopenia (HIT) or thrombocytopenia and thrombosis (HITT) (4).
HIT usually develops after more than five days of heparin therapy (5), with a range of four to 15 days (4, 6) in patients without heparin exposure in the preceding 100 days. Approximately 60% of patients with HIT experience a decline in platelets five to 10 days after exposure (5). The risk of HIT is 10 times higher with unfractionated heparin (UFH) than with low-molecular-weight heparin (LMWH) (3, 5, 6). HIT is 1.5 to 2 times more likely to occur in women than in men (4) and is more likely in surgical or trauma patients (1% to 5%) than in medical or intensive care patients (<1%) (5).
A diagnosis of HIT should be considered when a patient develops new thrombocytopenia (platelet count <150 × 109 cells/L) or experiences a proportional decline in platelets of more than 30% to 50% after heparin exposure. The most common complication of HIT is venous thrombosis such as DVT or pulmonary embolism, but arterial thrombotic events such as myocardial infarction, thrombotic stroke, and limb ischemia are all possible (3).
Clinical scoring systems to predict the likelihood of HIT include the HIT Expert Probability (HEP) score and the 4T's scoring system (5). Both the 4 T's score and the HEP score consider the magnitude and timing of the decrease in platelet count, the presence of thrombosis, and other potential causes for the observed thrombocytopenia (4), but differ in how specific clinical features are scored.
The HEP score includes eight clinical features and has a sensitivity of 100% and a specificity of 60% (5, 6, 7). Limitations of this scoring system include its complexity and the need for validation in larger prospective studies (5). The 4T's scoring system, as the name implies, includes four clinical categories. In a meta-analysis, the negative predictive value of a low 4 T's score was 99.8%, while the positive predictive value of an intermediate or a high 4 T's score was 14% or 64%, respectively (8).
Laboratory work-up and management of the patient with suspected HIT hinge on the clinical probability of the diagnosis. For a patient with a low clinical probability (4 T's score of ≤3), no further laboratory work-up is necessary and heparin therapy can be continued. For a patient with an intermediate to high clinical probability of HIT (4 T's score ≥4), heparin should be discontinued, an alternate anticoagulant should be initiated, and an anti-PF4/heparin ELISA should be obtained (1). The anti-PF4/heparin ELISA has a sensitivity of 99% (5), so if the anti-PF4/heparin ELISA result is negative, HIT is unlikely and heparin may be reintroduced.
If the anti-PF4/heparin ELISA is weakly positive (OD of 0.4 to 0.99) and the patient has an intermediate clinical probability (4 T's score of 4 to 5), HIT is also unlikely and heparin may be reinitiated. However, if the anti-PF4/heparin ELISA is weakly positive (OD of 0.4 to 0.99) and the patient has a high clinical probability (4 T's score 6 to 8), a functional assay (serotonin release assay or heparin-induced platelet activation assay) should be obtained. Last, in patients with intermediate or high clinical probability (4 T's score ≥4), if the anti-PF4/heparin ELISA is moderately or strongly positive (OD ≥1.00), a functional assay should be obtained. These functional assays offer both high sensitivity (90% to 98%) and specificity (90% to 95%). If these assays are positive, HIT is likely, and if they are negative, HIT is unlikely (1).
Treatment of HIT begins with stopping all heparin products and starting an alternate anticoagulant. The American College of Chest Physicians' 2012 guidelines “Treatment and Prevention of Heparin-Induced Thrombocytopenia” recommend anticoagulation with lepirudin, argatroban, or danaparoid (3). Of these agents, both lepirudin and danaparoid have been discontinued in the United States, thus the direct thrombin inhibitor argatroban is the only FDA-approved agent currently available to treat HIT. Alternative anticoagulation should be maintained until the platelet count increases to at least 150 × 109 cells/L, at which time a vitamin K antagonist (VKA) can be initiated at dosages of 5 mg or less (1, 3). It is recommended that VKA and alternate anticoagulant therapy overlap for a minimum of five days and until the INR is within the target range consistently before the patient is transitioned to VKA monotherapy (1, 2, 3). Patients with HIT should receive therapeutic anticoagulation for four weeks (2, 3). Bilateral lower-extremity ultrasound to rule out occult DVT may be considered (1), since the presence of a thrombus would change the diagnosis from HIT to HITT and change the treatment duration to three months (1, 2).
Interventions that are not currently recommended for the treatment of HIT include fondaparinux, DOACs, and prophylactic platelet transfusion. The 2012 guidelines did not recommend fondaparinux because the highest level of evidence at that time was limited to case series (3). Although none of the DOACs are approved for HIT, there are some published reports of their use (9, 10, 11, 12). Likewise, prophylactic platelet transfusions are not routinely recommended, but platelet transfusion may be employed in the event of bleeding or during a procedure with a high risk of bleeding (2, 3).
Several potential errors are highlighted in this case, including delay and early discontinuation of argatroban therapy. As a result of this suboptimal management, the patient developed a thrombotic complication, highlighting the importance of accurate recognition and treatment of heparin-induced thrombocytopenia.