RSS Feeds
Thrombotic Thrombocytopenic Purpura Associated with Ticlopidine: A Review of 60 Cases
- Charles L. Bennett, MD, PhD;
- Peter D. Weinberg, BS;
- Karine Rozenberg-Ben-Dror, PharmD;
- Paul R. Yarnold, PhD;
- Hau C. Kwaan, MD, PhD; and
- David Green, MD, PhD
- From Northwestern University and Chicago Veterans Affairs HealthCare System-Lakeside Division, Chicago, Illinois. Acknowledgments: The authors thank Drs. Diane Wysowski and Janos Bacsanyi of the U.S. Food and Drug Administration for providing information on retrieval of reports and for review of the manuscript; Drs. Dave Aboulafia, Harvey Einhorn, James Grossman, Robert Hecht, Chitra Raman, Tariq Mahmood, Peter Townley, Michael Voralia, and Walter Kahr, as well as Hal Lieberman, Nurit Begani, and Dr. Josh Levy and the staff of HemaCare Corp., Sherman Oaks, California, for contributing case histories and helpful comments; Dr. James George, for helpful comments; and our index patient, for her assistance and interest. Requests for Reprints: Charles Bennett, MD, PhD, Veterans Affairs Medical Sciences Building, #205, 400 East Ontario, Chicago, IL 60611. Current Author Addresses: Dr. Bennett and Mr. Weinberg: Veterans Affairs Medical Sciences Building, #205, 400 East Ontario, Chicago, IL 60611.
Abstract
Background: Thrombotic thrombocytopenic purpura, a life-threatening multisystem disease, has been infrequently associated with use of ticlopidine, a platelet anti-aggregating agent.
Purpose: To review 60 cases of ticlopidine-associated thrombotic thrombocytopenic purpura.
Data Sources: Medical records, published case reports, and case reports submitted to the U.S. Food and Drug Administration.
Study Selection: Instances of ticlopidine-associated thrombotic thrombocytopenic purpura were identified.
Data Synthesis: Ticlopidine had been prescribed for less than 1 month in 80% of the patients, and normal platelet counts had been found within 2 weeks of the onset of thrombotic thrombocytopenic purpura in most patients. Mortality rates were higher among patients who were not treated with plasmapheresis than among those who underwent plasmapheresis (50% compared with 24%; P < 0.05).
Conclusions: Ticlopidine use may be associated with the development of thrombotic thrombocytopenic purpura, usually within 1 month of initiation of therapy. The onset of ticlopidine-associated thrombotic thrombocytopenic purpura is difficult to predict, despite close monitoring of platelet counts.
Thrombotic thrombocytopenic purpura is a life-threatening multisystem disease characterized by thrombocytopenia, microangiopathic hemolytic anemia, neurologic changes, progressive renal failure, and fever [1, 2]. Its frequency is estimated to be only 3.7 cases per year per 1 million persons, with mortality rates ranging from 10% to 20% [3-5]. Although the cause of thrombotic thrombocytopenic purpura is unknown, many drugs, including penicillin, antineoplastic chemotherapy agents, and oral contraceptives, have been associated with the syndrome [6].
Ticlopidine, an antiplatelet agent, has rarely been associated with thrombotic thrombocytopenic purpura [6-14]. From the time when ticlopidine was first marketed (in October 1991) through March 1995, the U.S. Food and Drug Administration received reports of 25 ticlopidine users who developed thrombotic thrombocytopenic purpura [7]. Ticlopidine is used for treatment of intermittent claudication; for prevention of thrombotic strokes; and, more recently, for patients with cardiac stents. Early clinical trials identified neutropenia with bone marrow aplasia as the most common serious side effect of ticlopidine (with an incidence of 2.4% in clinical trials), but instances of ticlopidine-associated thrombotic thrombocytopenic purpura have been infrequently described [8-14]. We describe clinical characteristics, time course until the onset of disease, response to treatment, and outcomes in 60 patients with ticlopidine-associated thrombotic thrombocytopenic purpura.
Methods
Cases of ticlopidine-associated thrombotic thrombocytopenic purpura were identified from five sources: 1) reports obtained from MedWatch, the U.S. Food and Drug Administration's spontaneous reporting system [n = 36]; 2) previously published case reports (n = 11) [8-14]; 3) case summaries given to us after an informal survey of 20 hematologists on ticlopidine-associated thrombotic thrombocytopenic purpura [n = 6]; 4) cases reported to us from physicians affiliated with a large therapeutic plasmapheresis program (HemaCare, Inc., Sherman Oaks, California [n = 4]); and 5) medical records of patients with thrombotic thrombocytopenic purpura who were seen by physicians at Northwestern Memorial Hospital, Chicago, Illinois (n = 3). The case definition for ticlopidine-associated thrombotic thrombocytopenic purpura included use of ticlopidine and development of thrombotic thrombocytopenic purpura as measured by the scoring system proposed by Rose and Eldor [5]. The scoring system comprises four categories: platelet count (x109/L), hemoglobin level (g/L), serum creatinine level (micromol/L), and neurologic changes. Each category is assigned a score of 0, 1, or 2 on the basis of set criteria; a total combined category score of 4 or more indicates a diagnosis of thrombotic thrombocytopenic purpura.
Data Synthesis
We identified 60 cases of ticlopidine-associated thrombotic thrombocytopenic purpura (Table 1). Almost two thirds of the patients were older than 60 years of age, half were male, and 72% had received ticlopidine for stroke prevention. Normal platelet counts within 2 weeks of onset of thrombotic thrombocytopenic purpura were documented in two thirds of patients for whom platelet counts before disease onset were available. Ticlopidine therapy was stopped in all patients upon diagnosis of thrombotic thrombocytopenic purpura. The most common concomitant medications were aspirin (22%), metoprolol (12%), diltiazem (15%), hydrochlorothiazide (7%), and nitroglycerin (7%). In almost all cases, ticlopidine was the only new drug prescribed in the month preceding the onset of thrombotic thrombocytopenic purpura. Ticlopidine had been prescribed for less than 2 weeks for 15% of patients and less than 1 month for 80% of patients.
In all patients, thrombotic thrombocytopenic purpura was manifested by thrombocytopenia, anemia, neurologic changes, or renal dysfunction: Platelet counts were less than 20 × 109/L in two thirds of patients; hemoglobin levels were less than 90 g/L in 28% of patients; neurologic changes, including focal deficits, convulsions, or coma, were seen in 75% of patients; and renal insufficiency with a serum creatinine level greater than 221 µmol/L was seen in 25% of patients. Plasmapheresis was performed in 63% of patients; the rate was significantly higher among patients who were younger than 60 years of age (81% for persons <60 years of age and 54% for persons ≥ 60 years of age; P = 0.028).
Overall, the survival rate for patients with thrombotic thrombocytopenic purpura was 67%. Although no patient has had relapse, none have been rechallenged with ticlopidine. Plasmapheresis was the most important predictor of death: The mortality rate was 24% among patients who underwent plasmapheresis and 50% among patients who did not undergo this procedure (P = 0.049).
Discussion
We described the clinical characteristics of 60 patients with ticlopidine-associated thrombotic thrombocytopenic purpura. This condition is severe: Among patients in our study, the mortality rate was 33%. The disease affected both men and women and occurred among older persons who had frequently received the drug for prevention of stroke and among younger patients who usually had received the drug after cardiac stent placement. The individual case histories of patients in our study are similar to the few cases of ticlopidine-associated thrombotic thrombocytopenic purpura cases described elsewhere [8-14]. The clinical presentation of thrombocytopenia, anemia, neurologic changes, renal dysfunction, responsiveness to treatment, and overall mortality for ticlopidine-associated thrombotic thrombocytopenic purpura that we found are similar overall to those for patients with thrombotic thrombocytopenic purpura in general [3-5].
Undergoing plasma exchange is important for patients with thrombotic thrombocytopenic purpura. In a randomized trial, Rock and colleagues [3] reported a mortality rate of 22% among persons who underwent plasma exchange compared with a mortality rate of 83% among patients who underwent plasma infusion [3]. Similarly, we found a mortality rate of 24% among patients who underwent plasma exchange compared with a mortality rate of 50% among patients who did not undergo this procedure.
Our description of cases of ticlopidine-associated thrombotic thrombocytopenic purpura is limited because we used a clinical definition of the syndrome. Most of our patients had a score of 4 or more according to the scoring system of Rose and Eldor; four patients had a score of 3 and a diagnosis of thrombotic thrombocytopenic purpura made by an attending hematologist [5]. Patients were not directly seen by us, and patient information was provided by case reports (rather than hospital records) that often did not include complete records of laboratory values and clinical findings. These factors prevented us from obtaining complete clinical information on all patients.
The mechanism by which ticlopidine induces the clinical manifestations of thrombotic thrombocytopenic purpura is unclear. Ticlopidine, a thienopyridine compound, alters platelet function by inhibiting the binding of adenosine 5′ diphosphate to its adenylyl cyclase-coupled receptor site [15, 16]. The drug decreases clottable fibrinogen but not fibrinogen antigen [17] and decreases fibronectin [18]. A metabolite of ticlopidine rather than the parent compound is thought to be responsible for these actions [19]. It is paradoxical that a drug that inhibits platelet function is incriminated in a disorder that is manifested by platelet thrombi. This suggests that in patients who develop thrombotic thrombocytopenic purpura, a metabolite of ticlopidine with an activity very different from that of the parent compound may produce this disorder.
Most physicians are aware that neutropenia, hepatic cholestasis, and thrombocytopenia are adverse effects of ticlopidine [6, 20]. One review of an estimated 10 million patient-years of ticlopidine treatment [7] identified 645 cases of aplastic anemia, bone marrow suppression, pancytopenia, and agranulocytosis, of which 102 (16%) were coded as fatal. Because hematologic side effects of ticlopidine almost always occur within the first 3 months of therapy, the U.S. Food and Drug Administration issued advisory warnings when ticlopidine was approved, advising physicians to perform complete blood counts every 2 weeks for 12 weeks. However, compliance with this recommendation is unlikely to assist in early detection of thrombotic thrombocytopenic purpura. For most patients in our study, platelet counts were within normal limits 2 weeks before the onset of thrombotic thrombocytopenic purpura.
Our study raises concern about the potential for underdiagnosis and subsequent undertreatment of ticlopidine-associated thrombotic thrombocytopenic purpura. After cardiac stent placement or after saphenous vein grafting for treatment of peripheral vascular disease, patients usually receive ticlopidine for 2 to 4 weeks. Given this short period, many cardiologists do not monitor complete blood counts; consequently, thrombocytopenia associated with thrombotic thrombocytopenic purpura may go unrecognized. In our study, 12 patients developed thrombotic thrombocytopenic purpura after receiving ticlopidine for 3 weeks or less after placement of cardiac stents; 3 of these patients died. For older patients who receive ticlopidine to prevent stroke, the onset of thrombotic thrombocytopenic purpura may be insidious, with symptoms of altered neurologic function that mimic a stroke. Physicians may attribute these changes to cerebrovascular disease and may not institute plasmapheresis in a timely manner. In our study, only 60% of patients with thrombotic thrombocytopenic purpura who received ticlopidine for stroke prevention were treated with plasmapheresis. In some patients who survived, plasmapheresis was withheld because of good patient response to fresh frozen plasma, platelets, or steroids. In several of the patients who died, thrombotic thrombocytopenic purpura was diagnosed only in retrospect.
Ticlopidine is being used more frequently for prevention of stroke and clot formation after cardiac stent placement and as a treatment for intermittent claudication. The large number of cases of ticlopidine-associated thrombotic thrombocytopenic purpura that we have reviewed suggests that caution should be taken in prescribing this compound. Patients selected for ticlopidine therapy should be fully informed about the serious hematologic toxicity, including neutropenia and thrombotic thrombocytopenic purpura, that can develop within the first few weeks of therapy.
Drs. Rozenberg-Ben-Dror and Kwaan: Chicago Veterans Affairs HealthCare System-Lakeside Division, 333 East Huron, Chicago, IL 60611.
Dr. Yarnold: Time Life Building, #300, 303 East Ohio, Chicago, IL 60611.
Dr. Green: Rehabilitation Institute of Chicago, #1407, 345 East Superior, Chicago, IL 60611.
- Copyright ©2004 by the American College of Physicians
