What's Wrong With This Picture: Woman With Sudden Drop in Platelet Count
On her third day of hospitalization for acute gallstone pancreatitis, a 49-year-old woman’s platelet count fell to 113,000/μL from 216,000/μL the previous day.
History. The patient has a history of stable atherosclerotic coronary disease, tobacco abuse, and migraine headaches. Her gallstone pancreatitis has been treated with bowel rest and analgesics, with gradual relief of abdominal pain and improvement in bowel function. At 48 hours after admission, she exhibited none of the Ranson criteria for pancreatitis mortality prediction. She has no bleeding diatheses and has not experienced significant postoperative bleeding. Her current medications include morphine sulfate, 2 mg every 2 hours for pain, as needed; metoprolol, 25 mg twice daily; aspirin, 325 mg daily; and heparin, 5000 units subcutaneously 3 times daily.
Examination. The patient appears to be in no acute distress. Systolic blood pressure is 140 mm Hg; heart rate is 70 beats per minute. There is no gingival bleeding, petechiae, or bruising. Some tenderness is noted over the midepigastrium, but there is no abdominal or flank bruising.
Laboratory studies. White blood cell (WBC) count is 6300/μL with a normal differential; hemoglobin level is 13.4 g/dL; neither is significantly changed from the 2 previous days’ values. The patient’s peripheral blood smear is shown.
What do the clinical history and peripheral smear suggest is the cause of the patient’s sudden drop in platelet count?
A. Heparin-induced thrombocytopenia
B. Disseminated intravascular coagulation as a result of pancreatitis
C. Immune thrombocytopenic purpura
D. Pseudothrombocytopenia as a result of platelet satellitism
Answer on next page
,
WHAT’S WRONG:
The peripheral blood smear reveals platelet satellitism, D; this artifactual phenomenon occurs only in vitro and has no clinical significance. Satellitism is seen in 0.09% to 0.2% of both healthy and hospitalized patients when their blood is collected in tubes that contain the anticoagulant EDTA.
EDTA removes calcium from the glycoprotein IIb/IIIa complex, exposing platelet epitopes that react with antibodies in the patient’s serum; this results in clumping. In patients whose blood samples show platelet satellitism, antibodies also adhere to the neutrophil Fc3 receptor. The platelets then bind to neutrophils in a rosette formation (Figure 1).1
Automated platelet counters typically define platelets as all particles between 2 and 20 fL. If the machine cannot detect individual differences when platelets are clumped or are arranged in rosettes, the automated platelet count will be falsely depressed. In addition, the WBC distribution scattergram may show an unusual spike that reflects the larger neutrophil-platelet complexes, as seen in this patient (Figure 2).
A CAVEAT
Misdiagnosis of the cause of thrombocytopenia can lead to serious problems, including surgical delays, withheld treatment, and even unnecessary corticosteroid therapy and splenectomy.1 The first step in the evaluation of a falling platelet count is manual confirmation of the count to rule out artifactual pseudothrombocytopenia.
OUTCOME OF THIS CASE
After the suspected automated counter error was confirmed by evaluation of the peripheral blood smear, the complete blood cell count was repeated. The specimen was collected in citrate, a weaker chelator of calcium. The platelet count was 230,000/μL, and no clumping was evident. A new scattergram performed on the same day showed a typical distribution of WBC sizes.
The remainder of the patient’s hospital stay was uneventful. She was discharged with surgical follow-up for a planned cholecystectomy.
REFERENCE:
1. George JN, Rizvi MA. Thrombocytopenia. In: Beutler E, Lichtmen MA, Collier BS, et al, eds. William’s Hematology. 6th ed. New York: McGraw-Hill; 2001:1495-1537.