Vanishing Platelets

Abstract

Glycoprotein IIb/IIIa inhibitors are established treatment for patients who develop acute coronary syndromes. Thrombocytopenia is known to occur following the administration of various drugs, including heparin and glycoprotein IIb/IIIa inhibitors. In the case of glycoprotein IIb/IIIa inhibitors, the mechanism is thought to be drug-dependent antibodies. In most cases, the thrombocytopenia is mild or moderate in severity. Severe thrombocytopenia (platelet count, <50 × 10⁹/L) is distinctly rare.

Herein, we report a case of tirofiban-induced thrombocytopenia in which the overall platelet count dropped precipitously to <1 × 10⁹/L within 12 hours of administration; recovery was relatively prolonged, possibly owing to concomitant renal insufficiency. The severity and the rapidity of onset emphasize the need to routinely check platelet counts early after tirofiban administration, in order to prevent sequelae.

Glycoprotein IIb/IIIa inhibitors (GPIs) are now established treatment for patients who develop acute coronary syndromes before, during, and after percutaneous coronary intervention (PCI). Thrombocytopenia is known to occur after the administration of various drugs, including heparin and GPIs.¹ In the case of GPIs, the mechanism is thought to be drug-dependent antibodies.^1–3 In most cases, the thrombocytopenia is mild or moderate in severity, and a severe (<50 × 10⁹/L) decline in platelet count is distinctly rare—usually under 2% in clinical trials.^4–7

We report a case of tirofiban-induced thrombocytopenia in which overall platelet count dropped precipitously to <1 × 10⁹/L within 12 hours, followed by the patient's relatively prolonged recovery, probably due to concomitant renal insufficiency. The rapidity of onset and the lack of bleeding manifestations emphasize the need to routinely check platelet counts early after tirofiban administration, in order to prevent sequelae.

Case Report

In March 2007, a 50-year-old diabetic, hypertensive man presented at our hospital within 2 hours of the onset of an acute anterior myocardial infarction. He had no history of bleeding diathesis, hematologic disorder, or heparin exposure. He was taken to the cardiac catheterization laboratory for primary PCI after pretreatment with 300 mg each of aspirin and clopidogrel, 5,000 IU of intravenous unfractionated heparin, and a weight-adjusted high-dose bolus of intravenous tirofiban (20 μg/kg over 3 min), followed by a tirofiban infusion at 0.15 μg/kg/min.

Coronary angiography revealed a totally occluded left anterior descending coronary artery and moderate disease in the circumflex and right coronary arteries. Primary PCI resulted in the implantation of a 2.25 × 24-mm Apollo paclitaxel-eluting stent (InTek Technology; Baar, Switzerland). An excellent angiographic result was obtained.

Initial laboratory investigations (the results of which became available after the emergent PCI procedure) showed a hemoglobin level of 13 g/dL, a total leukocyte count of 14 × 10⁹/L, a platelet count of 246 × 10⁹/L, and a serum creatinine level of 2.4 mg/dL. The patient was continued on a regimen of aspirin and clopidogrel, and the tirofiban infusion was reduced in consideration of his renal dysfunction. On the next morning (approximately 12 hours after the PCI), the patient's platelet count on routine testing was noted to be 8 × 10⁹/L, declining to <1 × 10⁹/L 1 hour later, upon repeat testing. Review of the peripheral smear of an ethylenediaminetetraacetic-acid blood sample confirmed the profound lack of platelets with no clumping (Fig. 1). All antiplatelet drugs including tirofiban were immediately discontinued, and the patient was treated with platelet transfusions in an effort to maintain a count of >10 × 10⁹/L. The course of the patient's platelet count and concomitant serum creatinine level over the next 7 days is shown in Figure 2. There was almost no change in counts for the first 48 hours and a very slight rise on days 3 and 4, with improvement beginning after day 4 and counts exceeding 100 × 10⁹/L on day 6 (Fig. 3). During this time, the patient was supported with platelet transfusions. Clopidogrel and aspirin were resumed when the platelet count exceeded 25 × 10⁹/L. Throughout this period, the patient remained hemodynamically stable and did not show any evidence of bleeding.

Fig. 1 Blood smear from the patient at 12 hours after initiation of tirofiban treatment shows a virtual absence of platelets (Leishman stain, orig. ×40).

Fig. 2 Graph shows time course of platelet counts (dots) and serum creatinine levels (squares) from day of admission onward.

Fig. 3 Blood smear from the same patient on day 6 shows improvement in platelet count (Leishman stain, orig. ×40)

Discussion

Intravenous GPIs are widely used for the prevention of thrombotic complications in patients who have acute coronary syndromes, particularly in the setting of PCI. Three intravenous GPIs (abciximab, tirofiban, and eptifibatide) are currently available for clinical use. Abciximab is a monoclonal antibody directed against the GP IIb/IIIa receptor, while tirofiban and eptifibatide are high-affinity small-molecule inhibitors of the same. There is a clear association between GPI use and thrombocytopenia.^8–10 In the vast majority of cases, thrombocytopenia is of little consequence; but occasionally, severe and potentially fatal declines in platelet count can be observed.¹¹ The reported incidence of severe thrombocytopenia (platelet count, <50 × 10⁹/L) is 0.4% to 1.6% with abciximab, 0.2% with eptifibatide, and 0.2% to 0.5% with tirofiban.⁶ More profound degrees of thrombocytopenia have been reported with readministration of GPIs in patients who have a history of thrombocytopenia with these drugs.¹² Five patterns of GPI--induced thrombocytopenia have been identified: acute severe thrombocytopenia (platelet count, <10 × 10⁹/L) within 12 hours of 1st exposure, acute thrombocytopenia within 12 hours of 2nd exposure, delayed thrombocytopenia (5–7 days after treatment), anaphylaxis after 1st or 2nd exposure, and pseudothrombocytopenia.¹ The first 4 have been reported with both abciximab and the small-molecule inhibitors, while pseudothrombocytopenia has been seen only with abciximab.⁶ The main mechanism of true GPI-induced thrombocytopenia appears to be drug--dependent antibodies,^2,13 some of which may be naturally occurring¹ (which leads to thrombocytopenia on 1st administration of the drug). Various specialized assays exist to identify these antibodies, but they are difficult to obtain and the results are difficult to validate.⁶

The treatment of GP IIb/IIIa-induced thrombocy-to-penia is largely supportive in nature, because the problem is usually transitory and self-limited. Support includes platelet transfusions to maintain the platelet count >20 × 10⁹/L.¹⁴ However, in prolonged or severe cases, intravenous immunoglobulin G and corticosteroids^6,15 have been used with success.

The case presented here raises several highly unusual issues. First, the thrombocytopenia was very acute (under 12 hr) and extremely profound—there were almost no platelets seen and the laboratory reported the lowest possible count (<1 × 10⁹/L). Although rapid-onset thrombocytopenia has been reported earlier with tirofiban,^16,17 to the best of our knowledge, a platelet-count drop of this magnitude has not been reported with any of the GPIs. Second, the onset of recovery of the platelet count took approximately 4 days—much longer than the half-life of tirofiban.¹⁸ One may speculate that delayed clearance of the drug due to contrast--induced nephropathy contributed to slow recovery via the continued induction of antibodies. Tirofiban usually has a fairly rapid clearance in the absence of renal insufficiency.¹⁸ Third, the patient showed no signs of bleeding despite an extremely low platelet count. Although this in itself may not be unusual, severe thrombocytopenia associated with GPIs confers a 7-fold increased risk of severe bleeding;¹⁹ and we believe that at such a low platelet count, measures that included the cessation of antiplatelet drugs and the transfusion of platelets circumvented this bleeding risk. This emphasizes the need for routine testing in all patients who receive GPIs.

Although GPI-induced thrombocytopenia is the most likely explanation of the decline in platelet count in our patient, another potential cause is heparin. Heparin-induced thrombocytopenia (HIT) type I tends to occur early in the course of treatment and is usually mild and asymptomatic.²⁰ More profound and severe is HIT type II, which is immunologic in origin and occurs approximately 5 days after initiation of treatment, unless there has been prior exposure to heparin.²⁰ Although we did not test for heparin-dependent antibodies, we believe that the acute profound thrombocytopenia seen in our patient was very unlikely to be heparin induced, because this patient had no prior exposure to heparin, nor did the time course fit. Pseudothrombocytopenia is another possible cause of the low platelet count in our patient, but that was effectively ruled out by a review of the peripheral blood smear (which failed to show clumped platelets).

In summary, we present a case of acute, profound, and life-threatening thrombocytopenia as a result of tirofiban administration. It was identified early upon routine testing and necessitated platelet transfusions along with the cessation of antiplatelet therapy. The relatively prolonged recovery probably reflects delayed clearance of the drug owing to the contrast-induced renal failure that ensued. We report what is, to the best of our knowledge, the lowest platelet count ever in response to tirofiban; and we consider ourselves somewhat lucky to have identified the abnormality on the basis of routine laboratory data. Quite clearly, the platelet-count estimation should have taken place much earlier. We believe that this case emphasizes the need to identify unexpected drops in platelet counts well before bleeding ensues. All patients who receive any GPI therapy should undergo routine early testing of counts.

Acknowledgment

The authors acknowledge the help of Dr. Usman Shaikh, consultant hematologist, Aga Khan University Hospital, for his valuable input on the case and most particularly for his assistance in procuring and interpreting the hematologic slides.