Platelets are cell fragments released from megakaryocytes that circulate in an inactive state in the blood for 8 to 10 days. One third of the platelets are sequestered in the spleen and released back into the circulation in response to epinephrine.
Platelet activation, the first step in hemostasis, occurs in a three-stage process. Platelets adhere to exposed subendothelial surfaces of injured vessel walls (mediated by von Willebrand factor [vWF]). Platelets then change shape and release adenosine diphosphate and triphosphate, serotonin, fibrinogen, vWF, and factor V. Clopidogrel irreversibly binds to the P2Y12 adenosine diphosphate receptor, blocking platelet activation and aggregation. During shape change, glycoprotein IIb/IIIa receptors are exposed, bind fibrinogen, and initiate reversible platelet-platelet aggregation. The intravenous antiplatelet agents abciximab, eptifibatide, and tirofiban inhibit fibrinogen binding during this stage. In the final stage, arachidonic acid is released and converted to thromboxane A2, promoting irreversible platelet aggregation and resulting in a platelet plug (aspirin acts by irreversibly acetylating thromboxane A2). Finally, platelets provide the phospholipid membrane necessary to complete the coagulation cascade (see Bleeding Disorders).
Thrombocytopenia must be evaluated according to the platelet count and the clinical setting. Platelet counts greater than 100,000/µL (100 × 109/L) do not increase bleeding risk. Counts between 50,000 and 100,000/µL (50-100 × 109/L) do not require treatment and are adequate for surgical procedures, although 100,000/µL (100 × 109/L) is desirable for neurosurgery. Platelet counts between 30,000 and 50,000/µL (30-50 × 109/L) must be corrected before surgical intervention, and counts less than 10,000/µL (10 × 109/L) are associated with risk of spontaneous bleeding.
The first step of the evaluation is to review the peripheral blood smear and confirm the count. Pseudothrombocytopenia occurs when a patient has antibodies to ethylenediaminetetraacetic acid (EDTA), causing platelets to clump together in vitro. An accurate count can be obtained from blood drawn in citrate or heparin instead of EDTA. Inaccurate platelet counts may also occur if the platelets are exceptionally large (complete blood count machine may count them as erythrocytes) or if erythrocyte fragments (schistocytes) are counted by the machine as if they were platelets (leading to a higher than actual platelet count).
When thrombocytopenia is confirmed, symptoms are assessed, with attention paid to mucocutaneous bleeding (epistaxis, gum bleeding, menorrhagia, hematuria, melena, or hematochezia) and easy bruising. A thorough review of medications and supplements is required, especially in relation to the timing of onset of the thrombocytopenia. Drug and alcohol exposure must be documented and quantified, and dietary restrictions noted (for example, vegans may be deficient in vitamin B12). The patient's medical history may show related disorders, such as HIV infection, hepatitis C infection, or thyroid diseases (hyperthyroidism and hypothyroidism may be associated with thrombocytopenia).
The physical examination may reveal blood blisters in the mouth (“wet purpura”), petechiae or ecchymoses on the skin, or splenomegaly.
Disorders associated with bone marrow infiltration (myelofibrosis, metastatic tumors, granulomatous diseases) can decrease platelet production, as can nutritional deficiencies (vitamin B12 or folate) or abnormalities in stem cell maturation (aplastic anemia and myelodysplasia). Other cytopenias often accompany the low platelet count.
Splenomegaly is a common medical problem that causes thrombocytopenia without platelet destruction through increased sequestration. The thrombocytopenia is usually associated with anemia and leukopenia.
Unlike sequestration, non-immune platelet destruction is caused by abnormal platelet aggregation occurring in the setting of disseminated intravascular coagulation or a microangiopathic hemolytic anemia (MAHA) such as thrombotic thrombocytopenic purpura (TTP) or hemolytic uremic syndrome (HUS) (see following sections).
Patients with TTP present with MAHA and thrombocytopenia. Acquired TTP is caused by a deficiency in the metalloprotease ADAMTS13, which is responsible for cleaving high-molecular-weight vWF multimers. These multimers accumulate, generating platelet-rich thrombi in small vessels. These thrombi consume platelets and shear erythrocytes, fragmenting them into schistocytes (see Erythrocyte Disorders for figure) and may cause end-organ injury. Autoantibodies to ADAMTS13 are responsible for the deficiency in most patients, but hereditary TTP has been reported in some. Drugs such as ticlopidine, quinine, cyclosporine, gemcitabine, and vascular endothelial growth factor inhibitors (such as bevacizumab) can cause TTP. Drug abuse with oxymorphone, 3,4-methylenedioxymethamphetamine (“ecstasy”), and cocaine has also been reported to cause TTP.
Acquired TTP presents with MAHA (evidenced by elevated lactate dehydrogenase and decreased haptoglobin levels, schistocytes on peripheral blood smear, and a negative direct antiglobulin [Coombs] test result) and thrombocytopenia (with normal coagulation study results). An ADAMTS13 level less than 10% supports the diagnosis; an ADAMTS13 level greater than 50% should suggest an alternate diagnosis. The clinical picture is described in Table 19. All patients have MAHA and thrombocytopenia, but only 5% of patients present with all the aforementioned signs. The mortality rate associated with TTP is high, so treatment must begin immediately without waiting for ADAMTS13 test results. Treatment is successful in 85% of patients.
Initial treatment involves therapeutic plasma exchange to remove the high-molecular-weight vWF multimers and replace the deficient ADAMTS13 (plasma infusion, although not a definitive treatment, can be started if therapeutic plasma exchange is delayed). Glucocorticoids are added to decrease autoantibody production. Remission is defined by a normalization of the platelet count and is usually seen after 7 to 10 plasma exchange treatments. Rituximab may be added for patients with refractory disease.
HUS is characterized by thrombocytopenia, MAHA, and acute kidney injury. Classic HUS presents after an acute diarrheal illness caused by enterohemorrhagic Escherichia coli O157:H7 (although other strains have also been implicated). Diagnosis is confirmed by stool culture and polymerase chain reaction for Shiga toxin, free Shiga toxin, or O157:H7 antigen testing. Management is supportive (fluids and transfusions as needed). Antibiotics do not alter the course of the disease. Classic HUS has a 4% associated mortality rate, and long-term sequelae include hypertension and mild kidney disease. Atypical HUS presents without diarrhea and may be caused by complement dysregulation or may occur secondary to drugs (see drugs listed in the TTP section previously), systemic lupus erythematosus, or other infection (HIV, Streptococcus pneumoniae). In severe cases of HUS or when the distinction from TTP is unclear, treatment with plasma exchange should begin immediately, with eculizumab (a monoclonal antibody that binds to complement C5 and prevents generation of the membrane attack unit, C5b-9) added if atypical HUS is suspected. Drug-induced HUS is managed supportively with drug avoidance.
Peripheral destruction of platelets with decreased production is a feature of immune thrombocytopenic purpura (ITP) and is caused by autoantibodies directed against glycoproteins on the platelet surface membrane. Because only 6% of patients with platelet counts between 100,000 and 150,000/µL (100-150 × 109/L) develop persistent thrombocytopenia, many guidelines suggest the diagnosis of ITP be limited to patients with platelet counts less than 100,000/µL (100 × 109/L). ITP can occur alone, can be triggered by medications, or can be associated with other disorders, such as systemic lupus erythematosus, chronic lymphocytic leukemia, HIV, hepatitis C, or Helicobacter pylori infection.
ITP is characterized by the duration of the thrombocytopenia. Acute ITP lasts fewer than 3 months, persistent ITP lasts 3 to 12 months, and chronic ITP lasts longer than 12 months. Many patients are asymptomatic until the platelet count decreases to less than 10,000/µL (10 × 109/L). Petechiae and ecchymoses without lymphadenopathy or splenomegaly are notable physical signs. Laboratory findings are limited to a low platelet count. Antiplatelet antibody testing is not sensitive or specific. Testing for HIV and hepatitis C should be performed. Bone marrow evaluation is not necessary unless clinical findings suggest an alternate diagnosis. Platelet autoantibody testing is not recommended. Although platelet autoantibody assays are highly specific in identifying patients with ITP, they are not sensitive in ruling out the disorder, and results do not correlate with clinical outcomes.
In adults with newly diagnosed ITP and a platelet count less than 30,000/µL (30 × 109/L) who are asymptomatic or have minor mucocutaneous bleeding, treatment with glucocorticoids is recommended. Asymptomatic patients or those with minor mucocutaneous bleeding with a platelet count greater than 20,000/µL (20 × 109/L) may be managed as outpatients. Recommended initial therapy includes a short course (<6 weeks) of prednisone or dexamethasone. The response to intravenous immune globulin (IVIG) is faster and may be indicated in patients with more severe thrombocytopenia and life-threatening bleeding. Complications from glucocorticoids may include mood disorders, insomnia, fluid retention, hyperglycemia, and hypertension; complications from IVIG may include infusion reactions (headache, chills, anaphylaxis), kidney disease, and thrombosis. Those patients who relapse or have ITP refractory to these treatments require second-line treatments, such as splenectomy, rituximab, or thrombopoietin receptor agonists. Thrombopoietin receptor agonists, such as eltrombopag or avatrombopag (each given orally) and romiplostim (given subcutaneously weekly), may be beneficial in patients with a target platelet count of 50,000/µL (50 × 109/L), but they must be taken continuously to prevent relapse. Splenectomy may be considered in patients unresponsive to or intolerant of second-line therapy but should be delayed for 1 year after diagnosis because of the possibility of delayed remission.
Because patients with chronic ITP may be asymptomatic, treatment decisions must balance the risk of bleeding against treatment-related toxicities.
Type I heparin-induced thrombocytopenia (HIT) refers to a non−immune-mediated decrease in platelets occurring within the first few days of exposure to heparin. No intervention is needed because type I is clinically benign. Type II HIT is an immune-mediated thrombocytopenia occurring 5 to 10 days after exposure. It is caused by antibodies against platelet factor 4 (complexed to heparin). Associated thrombosis, which includes typical deep venous thrombosis, pulmonary embolism, or more unusual acute arterial occlusions, can be life threatening. The risk is highest with unfractionated heparin compared with low-molecular-weight heparin. Most patients with HIT develop platelet counts less than 150,000/µL (150 × 109/L), but severe thrombocytopenia is unusual. In patients with high baseline platelet counts, a decrease of greater than 50%, even if still within the “normal” range, warrants concern. Skin necrosis at the site of heparin injection or progressive thromboembolic events in patients receiving heparin are signs of possible HIT, even if the platelet count remains normal. The 4T scoring system (Table 20) is useful for diagnosis. A low probability score reliably excludes the diagnosis; a high score should prompt empiric therapy pending confirmatory tests.
Testing for heparin-associated antibodies is necessary to confirm the diagnosis. Testing for HIT involves a screening test and a confirmatory test. The enzyme-linked immunosorbent assay for platelet factor 4 antibodies is a very sensitive screening test (a negative test result rules out the diagnosis) but is not specific. The diagnosis must be confirmed with a functional test such as the serotonin release assay or the heparin-induced platelet aggregation assay (Figure 27). Classic platelet aggregation assays provide a more detailed evaluation of platelet function, but they are labor intensive and not used in initial screening.
Heparin must be discontinued immediately. An alternate anticoagulant should then be started (unless the patient is actively bleeding). Argatroban, bivalirudin, danaparoid, fondaparinux, or a non–vitamin K antagonist oral anticoagulant (such as rivaroxaban) are alternative options. Warfarin should not be the initial alternate anticoagulant because antithrombotic efficacy requires 3 to 5 days of therapy, during which time declining levels of protein C increase the thrombotic risk. Transitioning to warfarin is safe after thrombocytopenia resolves. Anticoagulation should be continued for 2 to 3 months in patients with HIT without documented thromboembolic events or 3 to 6 months for patients with HIT with associated thrombosis. Patients should be instructed to avoid heparin for life.
Patients presenting with mucocutaneous bleeding and normal complete blood count, prothrombin time, and activated partial thromboplastin time should be evaluated for platelet dysfunction, which can result from rare hereditary disorders or acquired disorders (Table 21). Von Willebrand disease is a much more common inherited cause of coagulation factor deficiency leading to platelet dysfunction (see Bleeding Disorders). After platelet dysfunction is identified, treatment depends on the clinical situation. Acute bleeding can be managed with platelet transfusions, but complete reversal of antiplatelet agent effects may be detrimental if the therapy is lifesaving (for example, patients with recent cardiac stents). Minor dental procedures or menorrhagia can be managed with antifibrinolytic agents (tranexamic acid or ε-aminocaproic acid).
The decision to discontinue antiplatelet medication in patients with acute bleeding is complicated, influenced by the interplay between the severity of bleeding and the risk of thrombotic vascular complications. In patients with acute gastrointestinal bleeding, evidence suggests (1) aspirin for primary prevention of atherosclerotic cardiovascular disease should be discontinued, perhaps permanently; (2) aspirin for secondary prevention need not be discontinued, but if aspirin is stopped, it should be resumed promptly when hemostasis is achieved; (3) dual antiplatelet therapy, with ongoing cardiology evaluation ensuring its appropriateness, need not be routinely discontinued; (4) the P2Y12 receptor antagonist should be stopped and aspirin continued for more serious bleeding in patients taking dual antiplatelet therapy, and it should be restarted promptly after achieving adequate hemostasis.
Platelet function can be assessed by a variety of laboratory tests. The Platelet Function Analyzer-100 is useful as a screening device, replacing the traditional bleeding time. The test is run on a citrated blood sample exposed to collagen and either epinephrine or adenosine diphosphate. It is a sensitive screen for von Willebrand disease and will detect abnormalities caused by antiplatelet medications. The Verifynow® test assesses for continued antiplatelet activity of aspirin or clopidogrel, but guidelines do not recommend using it to guide therapy. Thromboelastography and rotational thromboelastometry are bedside devices that analyze platelet function, coagulation, and clot stability and may be useful as screening tests in patients with unexplained bleeding.