Case study: 55-year-old man with thrombocytopenia

Patient's platelet count drops significantly after hospital day 14.

Issue: July 1, 2007

ByAnne Blaes, MD, MS

Editor's note: As part of its commitment to fellows in training, HemOnc Today continues to offer a series of important case presentations, written by fellows, followed by commentary from our editorial board. This month, Anne Blaes, MD, a fellow at the University of Minnesota and an editorial board member of Hem/Onc Today’s Fellow’s Affairs section offers a case and Gregory Vercellotti, MD, professor of medicine at the University of Minnesota and section editor of Hem/Onc Today’s Vascular Disorders section, provides the commentary.

A 55-year-old man presented with a scalp laceration and T3 and T12 compression fractures after falling down an escalator. He had a prolonged stay complicated by confusion, pain and methicillin-sensitive Staphylococcus aureus bacteremia.

On admission, the patient’s hemoglobin was 10 g/dL. His platelets were 200,000 k/mcL and had remained normal throughout the hospital stay. His white blood cell count was 13 g/dL. With the trauma, his hemoglobin drifted down to 5.6 g/dL. The patient was transfused with two units of blood on the third hospital day. His hemoglobin subsequently remained stable at 8 g/dL.

On hospital day 14, the patient’s platelet count dropped from 238,000 k/mcL to 6,000 k/mcL.

The patient’s medical history was significant for hepatitis C, hypertension and bilateral knee replacements requiring red cell transfusions, but no prior thrombocytopenia.

At the time of his thrombocytopenia, he was on the following medications: nafcillin for the preceding three days, quinine (Quinine, AR Holding), enoxaparin (Lovenox, Sanofi Aventis) for the preceding three days, heparin flushes, omeprazole, aspirin, furosemide, haloperidol, metoprolol and morphine. He had also received two doses of vancomycin four days prior to the drop in his platelets.

Anne Blaes, MD
Anne Blaes

He had a history of heavy alcohol use, but had been sober for two years. His history did not include tobacco or drug use.

On physical examination the patient was afebrile. His blood pressure was 160 mm Hg/80 mm Hg and his heart rate was 102 bpm. The patient appeared chronically ill and was wearing a body brace, which prevented an abdomen examination. He had mild scleral icterus. He also had some oozing from the gingiva as well as his scalp laceration. His lungs were clear and his heart was regular. He had no petechiae, bruising or lymphadenopathy.

His laboratory results at presentation were as follows:

Basic metabolic panel: normal.
Alanine transaminase: 25 U/L; aspartate transaminase: 45 U/L; total bilirubin: 3.4 U/L (direct, 3.1 U/L); alkaline phosphatase: 164 U/L; albumin: 2.2 g/dL.
White blood cell count: 18.4 k/mcL; hemoglobin: 10 g/dL; platelets: 6,000 k/mcL; mean corpuscular volume: 92 mcm³.
Prothrombin time international normalized ratio: 1.2.
Partial thromboplastin time: normal.
Fibrinogen: normal.
d-dimer: 7,900.
Heparin-induced thrombocytopenia (HIT) antibody: positive.

Peripheral smear revealed normochromic and normocytic anemia with increased red cell regeneration, mild leukocytosis with left-shift, mild absolute eosinophilia, hemophagocytosis and marked thrombocytopenia with large hypergranular platelets. No schistocytes were visualized.

Labs were drawn. Enoxaparin and quinine were discontinued, and nafcillin was switched to clindamycin. The patient was started on dexamethasone 40 mg orally daily for four days.

Additional laboratory results were as follows:

Retic count: 2.9%.
Haptoglobin: 150 mg/dL.
LDH-lactate dehydrogenase: 220 U/L.
Serotonin release assay: negative.
Coombs test: positive immunoglobulin G; confirmatory red cell eluate test: positive.
Anti-JK antibody: positive.

The patient’s clinical course deteriorated and he was intubated for acute respiratory distress syndrome. He had no signs of bleeding other than oozing from his scalp laceration. He received periodic platelet transfusions with no change in platelet count.

By day three he had minimal response to dexamethasone. He was started on IV immunoglobulin 0.4 g/kg for five days. He was also given IV methylprednisone (Medrol, Pfizer) 500 mg daily for three days.

The patient had no response to therapy and a bone marrow aspirate was performed on hospital day 26. Bone marrow aspirate revealed a mildly hypercellular bone marrow with mild granulocytic hyperplasia, mildly decreased iron stores, mild-moderate leukocytosis with toxic neutrophilia, marked thrombocytopenia and increased megakaryocytes.

The platelet antibody to human platelet alloantigen-5a (HPA–5a) was identified. The diagnosis of posttransfusion purpura was made.

The patient received transfusions with 5a negative platelets on three separate occasions with a platelet bump from 3,000 k/mcL to 21,000 k/mcL.

By hospital day 40 the patient’s platelet count gradually increased, returning to levels of 70,000 k/mcL.

Case Questions from Gregory M. Vercellotti, MD

Gregory M. Vercellotti, MD
Gregory M. Vercellotti

Gregory M. Vercellotti, MD, section editor for Hem/Onc Today’s Vascular Disorders section, said the case poses the following questions:

What is the best way to approach thrombocytopenia with such a precipitous drop in platelets?

This man presented a common dilemma for practicing physicians, abrupt decrease in circulating platelets in an ill patient. As we teach our medical students, one must approach this problem pathophysiologically: (1) production problems, (2) destruction and (3) sequestration. At the outset, careful review of the history and physical and laboratory findings are paramount. The review of the blood smear is the most effective activity in solving this conundrum.

This patient had no schistocytes, no megaloblastic features and decreased but large platelets that were not clumped. This suggests a destructive process with circulating young large platelets. He has a serious S. aureus infection and hepatitis C, no obvious nutritional disorders and the peripheral smear did not indicate myelodysplastic features such as pseudo–Pelger-Huët cells. The best test for assessing a platelet production issue is still the bone marrow aspirate and biopsy. In this patient, this was indicated.

Destruction can be consumptive or immunological. Hepatitis C infections can be associated with immune processes such as vasculitis, cryoglobulinemia or immune thrombocytopenic purpura. S. aureus can interact with platelets promoting early clearance. Drugs including nafcillin, vancomycin, quinine, heparin, furosemide and others could provide an immune/drug mediated destruction of platelets. He had a positive HIT antibody, but the timing and degree of thrombocytopenia is rarely seen in HIT. Compensated diffuse intravascular coagulation is possible with the increased d-Dimer. He did receive a blood transfusion and may have post transfusion purpura, but this is uncommon in men and the platelet count usually drops within five to 10 days after the transfusion. The conjugated bilirubin is elevated — possibly related to liver dysfunction — but an autoimmune hemolytic process with idiopathic thrombocytopenic purpura (Evans syndrome) might be considered. We do not know his spleen size on exam but worry about liver disease with portal hypertension with his alcohol history and low albumin.

What is the mechanism of posttransfusion purpura?

The patient clearly had a destructive process as he had severe thrombocytopenia in the face of increased megakaryocytes in the bone marrow. The Coombs test was positive along with anti-JK antibody. Could this be an Evans syndrome-like picture?

Corticosteroids had minimal benefit. It did suggest sensitization from JK–positive red blood cell transfusion. The positive Coombs test could be an autoantibody or an alloantibody. Posttransfusion purpura is a rare self-limiting thrombocytopenia occurring five to 10 days post transfusion of blood products in patients lacking a specific platelet antigen, usually HPA-1a. Eighty-five percent to 99% of individuals are HPA-1a positive. Most patients have a history of sensitization with prior transfusions or pregnancies. Ninety-nine percent of cases are in women. If rechallenged, alloantibodies that fix complement develop against the platelet-specific antigen that they are lacking, but are present on donor platelets. As a result, the transfused platelets and the patient’s own platelets are also destroyed through the adsorption of the antigen on their own platelets. Other possibilities include that the patient’s platelets adsorb immune complexes via platelet Fc receptors and are destroyed, or that an autoantibody develops in response to the foreign protein. The counts are usually <10,000 per mcL and resolve spontaneously in two weeks. IV immunoglobulin, corticosteroids, plasmapheresis and splenectomy may be helpful. Transfusion of antigen-specific platelets may be helpful if bleeding, but random donor platelets can cause severe allergic reactions.

In this patient, the platelets were negative for HPA–5a (located on glycoprotein 1a and present on 91% to 99% of individuals). Posttransfusion purpura has been reported in a few cases in HPA–5a negative patients. This patient had previous blood transfusions during knee replacements. Most red blood cell transfusions today are leuko-depleted, but still some platelet antigens may be transfused. It is not clear in this case what exact role posttransfusion purpura played in his thrombocytopenia as other confounding factors were in play. Regardless, clinicians faced with thrombocytopenic patients must fall back on their pathophysiology, reviewing mechanisms of cytopenias and of course looking at the blood smear.

Editor’s note: The important blood smear observation of “hematophagocytosis” noted before administered transfusions clearly suggests that the patient has autoimmune hemolysis and, thus, possibly a marked propensity to generate antibodies, including an antibody to PLA5, a generally weak antigen. – Harry S. Jacob, MD