Incidence and Clinical Course of Thrombotic Thrombocytopenic Purpura Due to Ticlopidine Following Coronary Stenting FREE

Thrombotic thrombocytopenic purpura (TTP) is a rare and often fatal disorder with an estimated incidence of 3.7 cases per 1 million people (0.0004%).¹ Laboratory findings of profound thrombocytopenia and microangiopathic hemolytic anemia and clinical findings of fluctuating mental status, mild renal dysfunction, and fever characterize TTP. The central pathogenic feature of this disorder is widely disseminated platelet aggregates, with prominent concentrations within the microcirculation of the pancreas, adrenal glands, heart, brain, and kidneys.² A number of precipitating causes of TTP have been identified, including malignancies, immunological disorders, pregnancy, toxic substances, infectious agents, and several drugs.³ One of these drugs is the antiplatelet agent ticlopidine hydrochloride, which was first associated with TTP in 1991 in a report of 4 patients.⁴ More recently, a review of 60 cases of TTP related to ticlopidine therapy raised concerns that the occurrence of this disorder may not be as rare as previously thought.⁵

Ticlopidine in combination with aspirin has been the mainstay of antithrombotic therapy following coronary stenting for the prevention of stent thrombosis since 1995. Because of the safety and antithrombotic efficacy afforded by this combination of antiplatelet agents compared with previous regimens, the use of intracoronary stenting has been steadily increasing.⁶ It is estimated that more than 1 million patients worldwide had a percutaneous coronary intervention in 1998, with 50% to 60% of these cases involving a coronary stent.⁷ Both the total number of interventions and the percentage of patients undergoing coronary interventions who receive stents are expected to increase during the ensuing years. Currently, nearly all patients receiving a coronary stent receive 2 to 4 weeks of dual antiplatelet therapy following the procedure.

Because of the large population of cardiac patients at risk, we sought to determine, through a multicenter, retrospective observational trial, the true incidence and clinical course of TTP associated with short-term exposure to ticlopidine following coronary stenting.

The predefined primary study cohort used to determine the incidence of ticlopidine-associated TTP after stenting was composed of all patients undergoing a percutaneous coronary intervention with a stent at 63 centers within the United States and Canada that made up the Evaluation of Platelet IIb/IIIa Inhibitor for Stenting (EPISTENT) study group.⁸ All centers had on-site coronary intervention laboratories. Each institution was asked to determine the total number of patients undergoing a percutaneous coronary intervention involving a stent at their institution during the preceding year and identify the number of cases of TTP after stenting treated at their institution during that same period. Cases of TTP were identified through case records and confirmed by telephone follow-up to each site's plasmapheresis center. For each case of TTP, a case report form was filled out through record review that included preprocedural medical history and medications, procedural treatment data, postprocedural medications and monitoring, and clinical, laboratory, and treatment data after TTP was diagnosed.

Additional cases of ticlopidine-associated TTP following stenting were collected to better define the clinical presentation and course of the disorder. These cases were identified through questionnaires with approximately 200 institutions not making up the primary study centers as well as cases from the primary centers that were not within the 1-year period or were noncoronary interventions. Again, clinical data were obtained through chart review for all of these patients. No patients reported herein have previously been described in other case series of ticlopidine-associated TTP.

Incidence rates and 95% confidence intervals (CIs) were calculated using the TRUE EPISTAT statistical software package (Epistat Services, Richardson, Tex). The observed rated of TTP following stenting was compared with the expected rate using the Poisson probability distribution.

During a 1-year period spanning 1996 to 1997, 43,322 patients were treated with coronary stents at the 63 centers. Nine cases of TTP (0.02%) following coronary stenting were identified at 6 of these centers during the study period. In this cohort, the risk of developing TTP following coronary stenting was 1 case per 4814 patients treated, with a 95% CI of 1 case per 2533 patients treated (0.04%) to 1 case per 10,541 patients treated (0.009%). The observed rate in our cohort greatly exceeded the expected incidence of 3.7 cases per 1 million patients in the general population (P<.001 by Poisson confidence limits). When compared with the rate of TTP in the general population, the risk of developing TTP due to ticlopidine use following coronary stenting was 50-fold higher (standardized morbidity ratio, 56.14; 95% CI, 25.67-106.56).

An additional 10 patients were identified from a population that was not part of the primary cohort—9 patients from other hospitals and 1 patient with a renal stent from 1 of the primary centers. Therefore, data were available for a total of 19 patients who developed TTP following treatment with ticlopidine after intravascular stenting (Table 1). The mean age of patients with TTP was 61 years and 31% were women. No patients had previously been exposed to ticlopidine remote to the procedure. Four of the 19 patients began ticlopidine therapy prior to their stent procedure. All 18 patients whose intraprocedural medications were documented received heparin and 3 patients also received adjunctive abciximab. No patient underwent atherectomy. The mean total length of stents placed was 26 mm, with a range of 9 to 60 mm. Seventy-two percent of stents were Palmaz-Schatz stents (Cordis, Johnson & Johnson Co, Warren, NJ). The intervention procedure was considered successful in all patients. Following the procedure, all patients received ticlopidine hydrochloride, 250 mg twice a day, and daily aspirin. Two patients were also discharged receiving warfarin. Ticlopidine was prescribed for 2 weeks in 2 patients, 4 weeks in 16 patients, and 8 weeks in 1 patient, although the total duration of therapy was 2 weeks or less in 4 patients (21%), 2 to 4 weeks in 14 patients (74%), and 4 weeks or more in 1 patient (5%). Three patients had completed their prescribed ticlopidine treatment regimen prior to the diagnosis of TTP and for the remaining 16 patients, ticlopidine therapy was discontinued after the recognition of signs and symptoms that were eventually shown to be TTP. The average duration of ticlopidine therapy was 22 days, with a range of 5 to 60 days. The diagnosis of TTP was made, on average, 25 days after starting ticlopidine, with a range of 8 to 67 days. Among the 3 patients who had completed their intended ticlopidine course, TTP was diagnosed 3 to 19 days after completion. One of these patients (patient 13 in the tables) had a normal screening complete blood cell count on day 14, his final day of therapy, and was admitted 3 days later with new onset of seizures.

Symptoms and hematological data are presented in Table 2. Seventy-three percent of patients had central nervous system symptoms, 47% had some level of renal dysfunction, and 32% had recurrent cardiac symptoms. For 2 patients, the presenting symptom of TTP was chest pain. The mean nadir platelet count was 11.4 × 10⁹/L (range, 3-48 × 10⁹/L), and the mean nadir hemoglobin level was 90 g/L (range, 59-126 g/L). Sixty-eight percent of patients were treated with plasmapheresis and 68% received glucocorticoids. The overall mortality rate was 21% (4/19), but in those patients who were not treated with plasmapheresis, the mortality rate was 67% (4/6) compared with none of the 13 who did receive plasmapheresis. The average length of hospital stay for surviving patients was 27 days (range, 6-107 days).

Thrombotic thrombocytopenic purpura as a complication of ticlopidine therapy was first described in a report of 4 patients in 1991.⁴ After this initial report, only sporadic descriptions of ticlopidine and TTP appeared in the literature.^9- 12 In 1998, Bennet et al⁵ compiled all reported cases of TTP associated with ticlopidine from a number of sources. The 60 patients identified by this group were gathered during a period of nearly a decade and throughout the world, so incidence could not be defined. Also, although their series included 12 patients who received ticlopidine after a stent procedure, the heterogeneity of the cases reported and the methods used to acquire these data precluded an accurate detailed description of the procedural characteristics and clinical course of the stented patients. However, their report was the first to suggest that TTP due to ticlopidine therapy may be more common than previously thought, and led to a change in the package insert for Ticlid (Roche Laboratories Inc brand of ticlopidine, Nutley, NJ) stating that the incidence of ticlopidine-associated TTP may be as high as 1 case in every 2000 to 4000 patients treated. The fact that 80% of their patients developed TTP less than 4 weeks after the start of therapy and 15% within 2 weeks suggested that the nearly half-million patients in the United States who annually undergo coronary stenting followed by ticlopidine might represent a population at significant risk for TTP.

We determined the incidence of TTP associated with ticlopidine following stenting by analyzing the population of patients who received a stent at 63 centers across the United States and Canada during 1 year. Among the nearly 45,000 patients receiving coronary stents at these centers, 9 definitive cases of TTP were identified. Based on the 95% CI of our incidence (1 case per 2533 to 1 case per 10,541 patients treated), approximately 50 to 200 patients would be expected to develop this life-threatening complication annually in North America because of ticlopidine treatment following coronary stenting.

Previous reports of TTP following cardiovascular surgery have suggested that endothelial injury alone may be the inciting event for the development of TTP.¹³ This would imply that the stent procedure itself may possibly lead to TTP in susceptible patients. However, this appears unlikely because TTP following stenting has never been reported in patients treated with an intravascular stent without concomitant ticlopidine use, despite almost no ticlopidine use after stenting between 1987 and 1994. Also, the fact that ticlopidine therapy alone has been well associated with the development of TTP makes its use after stenting almost assuredly the underlying cause of TTP in stented patients.

It is likely that our results may underestimate the actual incidence of TTP due to ticlopidine therapy following stenting. Because some patients' presenting symptoms are cardiac in origin, it is possible that TTP may manifest as an out-of-hospital death that is erroneously ascribed to subacute stent thrombosis, without TTP recognized as the underlying cause. Also, because each patient was not prospectively followed up for the development of TTP, it is possible that patients may have developed TTP and presented to other facilities without the knowledge of the cardiologist, hematologist, or their associated plasmapheresis center.

Our findings among all 19 patients with TTP confirm that early recognition and the initiation of proper treatment for these patients is crucial for minimizing mortality. Patients receiving plasmapheresis had 0% mortality compared with 67% in those not treated with plasmapheresis. The primary reason plasmapheresis was not initiated was because of a delay in the diagnosis. However, even with prompt therapy, the diagnosis of TTP was associated with substantial morbidity and prolonged hospitalization—a mean of 25 days in our study.

No concomitant medications or patient or procedural characteristics were found to predict the development of TTP. Unlike the female predominance generally described with TTP, 69% of our patients were men, which is more consistent with the demographics of patients undergoing stenting. The majority of our patients were prescribed 4 weeks of ticlopidine therapy following stenting, with the majority of patients developing TTP within the third week of therapy. The presenting symptoms were central nervous system complaints in two thirds of our patients and recurrent cardiac symptoms in 2 patients.

Two of the most concerning findings among our patients were the development of TTP with less than 2 weeks of therapy in 4 patients and the delay in onset of TTP of up to 19 days after the completion of a patient's course of ticlopidine therapy. Because the percentage of all patients who had stents and were treated with 2 weeks vs 4 weeks of ticlopidine is not known, it is impossible to say whether the smaller number of patients in our cohort treated with only 2 weeks of ticlopidine therapy reflects a protective effect of the shorter therapy or only mirrors a smaller percentage of all patients treated with only 2 weeks of ticlopidine therapy. What is clear is that 2 weeks of therapy does not completely eliminate the risk of TTP. Also, the onset of symptoms as early as 5 days after starting ticlopidine therapy and as long as 19 days after completing treatment suggests the need for more frequent and longer routine monitoring.

The cause of TTP is not fully understood, but ultralarge von Willebrand factor multimers, not normally present in plasma, have been described in patients with TTP.¹⁴ These abnormal multimers can bind to platelets and induce intravascular aggregation. Two recent reports have described the presence of an inhibitory antibody against von Willebrand factor–cleaving protease in patients during the acute phase of TTP.^15- 16 A preliminary report of a patient with ticlopidine-induced TTP has also demonstrated the presence of an inhibitor of von Willebrand factor–cleaving protease.¹⁷ This suggests that ticlopidine or a metabolite may either directly inhibit or elicit an antibody that inhibits von Willebrand factor–cleaving protease in susceptible patients.

Clopidogrel, like ticlopidine, is a thienopyridine derivative that selectively inhibits the adenosine diphosphate (ADP) receptor, thus blocking ADP-induced platelet aggregation and ADP-mediated amplification of platelet activation by other agonists.¹⁸ It is structurally identical to ticlopidine except for the addition of a carboxymethyl side group. However, unlike ticlopidine, there have been no reports of blood dyscrasias associated with clopidogrel. Results from the CAPRIE (Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events) Trial, which included more than 19,000 patients, showed no excess in neutropenia compared with aspirin-treated patients.¹⁹ Importantly, no cases of TTP were reported in this trial and none have surfaced after broad clopidogrel administration since its release in March 1998. The fact that clopidogrel's metabolites are different from all of the metabolites of ticlopidine, each possessing either the initial carboxymethyl ester group or the carboxylic acid group (resulting from hydrolysis of the carboxymethyl ester group) (written communication, J. Marc Herbert, PhD, Sanofi Recherche, October 1998), likely explains the lack of TTP and other adverse effects found with clopidogrel.

Thrombotic thrombocytopenic purpura due to ticlopidine therapy following coronary artery stenting is unusual, but, at 0.02%, it occurs more commonly than previously thought. Current routine monitoring of blood counts is unlikely to unmask TTP prior to clinical presentation. Educating all patients treated with ticlopidine about the signs and symptoms of TTP might allow for a more rapid diagnosis and treatment, which are critical for decreasing mortality. Although limiting therapy to 2 weeks after stenting is thought by many to decrease the risk of adverse effects from ticlopidine, it does not prevent the development of TTP. Clopidogrel may offer a safer alternative to ticlopidine as part of a dual antiplatelet regimen following stenting. However, prospective trials, which are currently under way, are necessary to definitively determine the safety and efficacy of this agent in patients undergoing coronary stenting.