Author + information
- Received July 26, 2000
- Revision received December 13, 2000
- Accepted January 5, 2001
- Published online April 1, 2001.
- Domenico Scrutinio, MD∗,* (, )
- Claudio Cimminiello, MD†,
- Ettore Marubini, PhD‡,
- Maria Vittoria Pitzalis, MD§,
- Matteo Di Biase, MD∥,
- Paolo Rizzon, MD§,
- the STAMI Group
- ↵*Reprint requests and correspondence: Dr. Domenico Scrutinio, Division of Cardiology, “S. Maugeri” Foundation, IRCCS, Institute of Rehabilitation, 70020 Cassano Murge, Bari, Italy
We sought to compare the efficacy of aspirin and ticlopidine in survivors of acute myocardial infarction (AMI) treated with thrombolysis.
The role of ticlopidine in secondary prevention after AMI has not yet been explored.
Of 4,696 patients with AMI treated with thrombolysis who were screened, 261 died in the hospital (5.6%) and 1,470 were enrolled in this randomized, double-blind, multicenter trial and allocated to treatment with either aspirin (160 mg/day) or ticlopidine (500 mg/day). The most frequent reasons for exclusion were refusal to give informed consent, planned myocardial revascularization, risk of noncompliance with study procedures, need for anticoagulant therapy and contraindications to the study treatments. The primary end point was the first occurrence of any of the following events during the six-month follow-up: fatal and nonfatal AMI, fatal and nonfatal stroke, angina with objective evidence of myocardial ischemia, vascular death or death due to any other cause.
The primary end point was recorded in 59 (8.0%) of the 736 aspirin-treated and 59 (8.0%) of the 734 ticlopidine-treated patients (p = 0.966). Vascular death was the first event in five patients taking aspirin and in six patients taking ticlopidine (0.7% vs. 0.8%; p = NS); nonfatal AMI in 18 and 8 (2.4% vs. 1.1%; p = 0.049); nonfatal stroke in 3 and 4 (0.4% vs. 0.5%; p = NS); and angina in 33 and 40 (4.5% vs. 5.4%; p = NS), respectively. The frequency of adverse reactions was not significantly different between the two groups.
No difference was found between the ticlopidine and aspirin groups in the rate of the primary combined end point of death, recurrent AMI, stroke and angina.
Both aspirin and ticlopidine have proved to be beneficial in patients with atherosclerotic disease, with relative risk reductions of 27% and 33%, respectively, in terms of the composite end point of vascular death, acute myocardial infarction (AMI) and stroke (1). In coronary artery disease (CAD), the beneficial effect of aspirin extends to a wide range of clinical conditions, including suspected AMI, postinfarction, stable and unstable angina and coronary artery bypass graft surgery (2). Ticlopidine has been shown to be beneficial in unstable angina (3), bypass surgery (4)and, in combination with aspirin, in reducing the risk of subacute stent thrombosis (5).
In current clinical practice, aspirin is the standard antiplatelet therapy in secondary prevention after MI (6). However, aspirin is a weak platelet inhibitor and is not protective against recurrent ischemic events in a sizable number of patients with CAD (1,7–9). Progression of CAD or mechanisms overcoming the platelet inhibitor effect of aspirin may be responsible for such recurrences (8,10). Adenosine diphosphate (ADP) inhibitors are currently used as an alternative to aspirin, but the role of ticlopidine in secondary prevention after AMI has not yet been explored. The present study compares the efficacy and safety of aspirin and ticlopidine in survivors of AMI treated with thrombolysis.
This was a randomized, double-blind, multicenter trial managed by a Coordinating and Data Management center, as well as Steering, Safety and Validation Committees. All events and adverse reactions were reviewed by the Validation and Safety Committees in blinded manner. The study protocol was approved by local Ethical Committees, and all patients gave written, informed consent.
Eligible patients were men and women >18 years old who had been discharged from the hospital after a MI treated with thrombolytic therapy, with the onset of MI no more than 30 days before randomization. Acute MI was diagnosed by the presence of at least two of the following: typical chest pain lasting >30 min; an increase in creatine kinase, aspartate transaminase or lactate dehydrogenase levels to at least twice the upper limit of normal; ST segment elevation ≥0.1 mV in any limb lead on the electrocardiogram (ECG) on admission to the coronary care unit or ≥0.2 mV in at least two contiguous precordial leads, or both; or the presence of new Q-waves lasting ≥40 ms in at least two contiguous ECG leads, or a new dominant R-wave in lead V1(R-wave ≥1 mm ≥S in lead V1) on the prerandomization ECG. Exclusion criteria were contraindications to the study drugs (e.g., hematopoietic or hemostatic disorders, a history of thrombocytopenia, leukopenia or peptic ulcer); the need for oral anticoagulant therapy; scheduled major or bypass surgery; severe comorbidity likely to limit the patient’s life expectancy; uncontrolled hypertension; geographic or other factors making study participation impractical; or participation in other concomitant trials. Each center recorded all consecutively screened patients who had thrombolysis in a log book, and checked the inclusion and exclusion criteria and reported their in-hospital outcomes.
The initial assessment included a medical history, general physical and cardiovascular examination and recording details of the qualifying AMI. Laboratory tests were carried out at baseline to rule out excluding conditions. The clinical follow-up visits took place after 45 days, three months and at the end of the six-month follow-up period. During these visits, information was collected on outcome end points, adverse events and concomitant medications, and the study drug for the next period was dispensed. Blood cell counts were repeated every two weeks during the first three months, then at each scheduled visit.
Compliance was assessed by counting the number of tablets returned at each follow-up visit, which was recorded in the case report form. Patients were given indistinguishable blister packs containing either ticlopidine (250-mg tablets) or plain aspirin (80-mg tablets), and they were asked to take one tablet twice a day after meals. The use of anticoagulants or additional antiplatelet agents was forbidden, and patients were instructed to avoid aspirin-containing products. If surgery was scheduled, the study drug was discontinued 10 days before and then resumed. All reasons for early, temporary or permanent discontinuation were recorded. The patients were to be followed up for six months, regardless of whether or not they were still taking the study drug.
End points and outcome events
The primary end point was the first occurrence of any of the following during the six-month follow-up period: fatal and nonfatal MI, fatal and nonfatal stroke, angina with objective evidence of myocardial ischemia, vascular death or death due to any other cause. Nonfatal MI was diagnosed by the presence of at least two of the following: typical chest pain lasting >30 min; increases in cardiac enzyme levels to at least twice the upper limit of normal; new and persistent ST segment elevations in any ECG leads; or new pathologic Q-waves, as described earlier. Nonfatal ischemic stroke was defined as a suddenly occurring focal neurologic deficit lasting >24 h whose ischemic origin was confirmed by a cerebral computed tomographic scan or magnetic resonance imaging.
Angina could be considered as an outcome event, provided it was documented by ECG changes during or just after the chest pain and/or Holter, ergometric or myocardial scintigraphy abnormalities demonstrating myocardial ischemia. Fatal reinfarction or ischemic stroke was diagnosed if death occurred within 28 days of the acute event. Vascular death, including sudden death, was any death clearly not due to nonvascular causes. Thrombocytopenia was defined as serious when the platelet count was <80,000/mm3, and neutropenia was diagnosed when the neutrophil count was <450/mm3.
It was estimated that ∼2,500 patients had to be treated and followed up for six months to detect a 5% absolute difference in the frequency of patients free of any of the events considered in the primary end point (alpha = 0.05, two-sided; power 0.85).
When the study was planned, thrombolysis was expected in 70% of patients with MI (11), and a frequency of 75% in event-free patients was assumed as the baseline reference value (12–15). Therefore, a recruitment period of 12 months was decided. During the screening process, it became evident that the actual proportion of thrombolysed patients was ∼40%, and consequently the recruitment period was extended to 22 months, but it was still not long enough to recruit the planned number. Furthermore, during the study, the overall frequency of event-free patients (monitored in blinded manner) was always ∼90%. After recruitment of 1,470 patients, the power calculation was revised, using the observed event-free frequency of 90% as baseline, with 5% as the absolute difference between groups (alpha = 0.05, two-sided test); a power of 95% was reached.
Intention-to-treat analysis was done by evaluating all the enrolled patients and all the validated outcome events. Primary analysis was based on the first occurrence of an event in the outcome cluster of death, MI, ischemic stroke and angina. Event-free time analysis was done using the log-rank two-sided test, and event-free survival curves were plotted using the Kaplan-Meier method. Time was calculated as the time that a patient participated in the trial, measured in days from the date of randomization to the date of the occurrence of the primary end point, or from the date of randomization to sixth-month follow-up of patients who were event-free. According to the protocol, all possible efforts were made to obtain complete follow-up information up to the end of the trial for patients who discontinued early. The following censoring procedure was adopted: “withdrawal event-free” patients at the end of the six-month trial; and “lost to follow-up” patients at the last available follow-up visit.
When a patient is exposed to several causes of heart failure, the occurrence of each can be analyzed by making allowance for the “competing risk” of the event considered (16). As allowed in the protocol, a further analysis was planned to investigate the possible different effects of treatments on each type of primary event. The crude cumulative incidences were calculated for three types of failure: nonfatal MI, angina (added with the few cases of nonfatal stroke) and death for any reason. Gray’s nonparametric method (17,18)was used to compare the crude cumulative incidences of each type of failure.
Safety assessments were based on the proportions of patients experiencing one or more episodes of a specific drug-related adverse event, and comparisons were made by using the chi-square test. All adverse events (serious and nonserious) were reported, as grouped by body system according to World Health Organization Adverse Reaction Terminology. A patient was counted only once in the total for a body system, regardless of the number of individual events reported within that system.
Between March 1995 and January 1997, 4,696 patients with AMI treated with thrombolysis were screened for inclusion in the trial. Of these, 261 patients (5.6%) died in the hospital and 1,470 were randomized to receive aspirin (n = 736) or ticlopidine (n = 734). The most frequent reasons for exclusion were refusal to give informed consent (25%), planned myocardial revascularization (14%), risk of noncompliance with study procedures (14%), need for anticoagulant therapy (5.3%) and contraindications to the study treatments (5.2%). The baseline characteristics of the 1,470 patients enrolled are listed in Table 1, and those at randomization in Table 2. In both groups, 96% of the patients met all the three diagnostic criteria for qualifying AMI. The early-phase treatments are reported in Table 3.
During the six-month follow-up, the primary end point (defined as the first occurrence of death, nonfatal MI or stroke or angina) occurred in 118 patients: 59 in the ticlopidine group and 59 in the aspirin group. The frequencies of the first events are presented in Table 4, and the cumulative incidence curves are shown in Figure 1. Five of these 59 patients in the aspirin group and six in the ticlopidine group experienced more than one event. In the aspirin group, one patient died late (>28 days) after MI; one died late after MI and documented angina; one had MI after angina; one presented with angina after MI; and one patient had angina recorded three times. In the ticlopidine group, one patient died late after MI; one died late after documented angina; and four presented with angina twice.
To investigate whether the treatments had different effects on the individual type of event, the crude cumulative incidence was estimated for each (Fig. 2). The curves for fatal events (Fig. 2A)were similar over time (p = 0.558). The crude cumulative incidence curves for nonfatal MI (Fig. 2B)started to diverge at day 20 and maintained this difference throughout the follow-up, showing more effective protection with ticlopidine (p = 0.049). The curves for angina and nonfatal stroke (Fig. 2C)diverged after the first month of treatment, but tended to converge toward the end of the follow-up period (p = 0.342). The rate of revascularization during follow-up was 13% in both groups.
Of the 1,470 randomized patients, 333 (22.6%) permanently discontinued the study drug before the end of the scheduled six-month follow-up period: 152 in the aspirin group and 181 in the ticlopidine group (p = 0.07). Of these patients, 271 were observed until the end of the trial; 30 (4.1%) in the aspirin group and 32 (4.4%) in the ticlopidine group were lost to follow-up. Thirty-five patients in the aspirin group and 53 in the ticlopidine group stopped the treatment because of adverse events; 26 in the aspirin group and 21 in the ticlopidine group withdrew because of the occurrence of an end point.
Of the 1,470 randomized patients, 162 (11%) reported at least one drug-related adverse event: 72 (9.8%) in the aspirin group and 90 (12.3%) in the ticlopidine group (p = 0.129). Serious adverse events occurred in 11 patients (1.5%) treated with aspirin (one of whom presented with two serious reactions: chest pain and dyspepsia) and 18 (2.4%) treated with ticlopidine (p = 0.196). The most frequent or meaningful adverse events (classified by organ system) are reported in Table 5.
For gastrointestinal symptoms, the following frequencies were found in the aspirin group and ticlopidine group, respectively: abdominal pain (3.4% and 3.3%), dyspepsia (2.0% and 1.9%), diarrhea (0.4% and 1.4%), nausea (0.3% and 0.5%) and melena in two patients treated with aspirin (0.3%). Eight adverse events in the aspirin group were serious (1.1%): seven required hospital admission (abdominal pain [n = 2], dyspepsia [n = 2], melena [n = 1], pancreatitis [n = 1] and hemorrhagic gastric ulcer [n = 1]); the eighth event was defined as medically serious nausea. In the ticlopidine group, 11 adverse events (1.5%) were serious and required hospital admission (abdominal pain [n = 3], dyspepsia [n = 4], diarrhea [n = 1], gastritis [n = 1], duodenal ulcer [n = 1]and gastric ulcer [n = 1]).
Platelet, bleeding and clotting disorders developed in seven patients treated with aspirin: thrombocytopenia (n = 3, one serious), gingival bleeding (n = 2), epistaxis (n = 1) and hematoma (n = 1). In the ticlopidine group, the complication included epistaxis (n = 3), thrombocytopenia (n = 4, one serious) and hematoma (n = 1). Aspirin was discontinued in four patients (bleeding [n = 1], hematoma [n = 1] and thrombocytopenia [n = 2]), and ticlopidine was discontinued in five patients (thrombocytopenia [n = 3] and epistaxis [n = 2]). There were no cases of cerebral bleeding.
White blood cell and reticuloendothelial system disorders occurred in one patient taking aspirin and in four patients taking ticlopidine. Two of those in the ticlopidine group had serious cases (one case of leukopenia and one of granulocytopenia), but both patients recovered without sequelae. One case of serious anemia occurred in the aspirin group, and one case of serious pancytopenia occurred in the ticlopidine group. In addition, four other serious adverse events were reported: one rash in the aspirin group and one case of rash, one case of elevated hepatic enzymes and one case of pulmonary infiltration in the ticlopidine group.
To the best of our knowledge, this is the first study comparing the effects of ticlopidine and aspirin on recurrent events after thrombolysis in patients with AMI. The dose of 160 mg/day of aspirin was chosen on the basis of the Second International Study of Infarct Survival (ISIS-2) results (19). This choice is consistent with the American College of Cardiology/American Heart Association (6)recommendation that a dose of 160 to 325 mg should be given on day 1 of AMI and continued indefinitely on a daily basis thereafter. At 160 mg/day, aspirin has a rapid clinical antithrombotic effect caused by immediate and near-total inhibition of thromboxane A2production (6); furthermore, the likelihood of side effects is low.
The main finding of the trial was that the incidence of the primary combined end point of vascular death, nonfatal MI or stroke or angina was no different between the two groups. However, the data also suggest that the risk of reinfarction after thrombolysis may be lowered further by ticlopidine (p = 0.049). It is worth noting that the incidence of nonfatal reinfarction in the aspirin-treated patients (2.4%) in the present study was very close to that in the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico-2 (GISSI-2) trial (2.5%) (20), despite a substantially greater use of aspirin (100% vs. 77%), beta-blockers (46% vs. 25%), angiotensin-converting enzyme inhibitors (53% vs. 10%) and revascularization procedures (13% vs. 6%) in our six-month follow-up study versus GISSI-2, respectively. Our results are consistent with the findings of the Studio della Ticlopidina nell’Angina Instabile trial (STAI) (3), in which the benefit of ticlopidine was entirely due to the lower incidence of fatal or nonfatal AMI. They are also in line with the more recent Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events (CAPRIE) trial (21), which compared clopidogrel with aspirin in patients at risk of ischemic events. In that study, there was a relative risk reduction of 8.7% (p = 0.043) for the primary outcome cluster of MI, ischemic stroke and vascular death and 19.2% for the outcome of MI alone, in favor of clopidogrel.
Reinfarction after thrombolysis
Most reinfarctions occurring late after thrombolysis (>1 month) result from reocclusion of the infarct-related artery (22–24). White et al. (25)found that reocclusion occurred between four weeks and one year in 25% of the patients who had successful thrombolysis and was clinically associated with reinfarction in almost half of the cases. The severity of residual stenosis proved to be a strong independent predictor of reocclusion in the year after successful thrombolysis (25–27).
In patients with severe residual stenosis, the rheology of blood flow in the vessel is altered, leading to increased shear stress and rethrombosis (28). Experimental studies have shown the limited efficiency of aspirin in inhibiting platelet-mediated thrombosis under high shear conditions (29–31). In agreement with these experimental findings, angiographic studies have shown only a limited benefit of aspirin in preventing late reocclusion of severe infarct-related artery stenoses after thrombolysis (25,27). Aspirin appears to protect against reocclusion of the less severe lesions, but these have a lower reocclusion rate anyway (25,27).
Under the high shear rate situations characteristic of severe stenosis, ADP plays an important role in mediating platelet aggregation (7,32,33), and ADP inhibitors have shear-independent antithrombotic effects in an experimental setting (30,31,34). It has not been determined whether this translates into a greater benefit of ADP inhibitors in preventing late reocclusion of the infarct-related artery, and its deleterious consequences (26), after thrombolysis. The lower reinfarction rate found in the present study is promising.
The incidence of drug-related adverse events did not differ significantly between the two groups. The most frequent were gastrointestinal disorders, which occurred in 7.3% of patients in the aspirin group and 8.7% of patients in the ticlopidine group. There were few serious side effects in either group; serious reversible neutropenia was observed in only two patients, both of whom were taking ticlopidine (0.3%). This rate is lower than that previously reported, and several reasons are possible: laboratory tests were done at baseline to rule out contraindications to the study drugs; ticlopidine was permanently discontinued in two additional patients presenting with mild neutropenia, which otherwise might have become severe; and, at present, all physicians are aware of the potential adverse effects associated with ticlopidine, and consequently, they pay careful attention to hematologic monitoring to detect any such effect promptly.
Overall, we found no difference in the frequency of the primary end point between the two treatment arms. However, there was a significantly lower rate of nonfatal reinfarction in the ticlopidine group (p = 0.049). Although the analysis aimed at assessing the effects of treatment on each type of failure combined in the primary end point had been planned by protocol, this finding must be viewed with some caution. The trial was not powered to detect a difference in the reinfarction rate between the two groups, and the number of reinfarctions was low. Furthermore, the mechanisms underlying rethrombosis are multifactorial (29).
Despite an apparent trend toward a lower reinfarction rate with ticlopidine, the frequency of the primary combined end point of death, reinfarction, stroke and angina was no different between the two treatment groups. These data indicate that an ADP inhibitor, such as ticlopidine, is at least as effective as aspirin in preventing recurrent events in post-MI patients treated with thrombolysis. (appendix)
STAMI Study Organization: Steering Committee:P. Rizzon (Chairman), D. Scrutinio, C. Cimminiello, E. Marubini, G. Rudelli, G. Feruglio;1Validation Committee:D. Scrutinio, C. Cimminiello, C. Pasotti, A. Carolei; Outcome Evaluation Committee:E. Marubini, G. Serio; Safety Committee:G. G. Nenci, M. Guslandi; Data Management:A. Passannante; Statistical Analysis:S. Geraci.
For a complete list of Collaborating Clinical centres, please see the April issue of JACCat http://www.cardiosource.com..
- adenosine diphosphate
- acute myocardial infarction
- coronary artery disease
- Received July 26, 2000.
- Revision received December 13, 2000.
- Accepted January 5, 2001.
- American College of Cardiology
- Antiplatelet Trialists’ Collaboration
- Studio della Ticlopidina nell’Angina Instabile Group,
- Balsano F,
- Rizzon P,
- Violi F,
- et al.
- Ryan T.J,
- Anderson J.L,
- Antman E.M,
- et al.
- Topol E.J
- Garcia-Dorado D,
- Théroux P,
- Tornos P,
- et al.
- Cipollone F,
- Patrignani P,
- Greco P,
- et al.
- ISAM Study Group,
- Schroder R,
- Neuhaus K.L,
- Leizorovicz A,
- et al.
- Ad Hoc Working Group of the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico (GISSI)-2 Data Base,
- Volpi A,
- De Vita C,
- Franzosi M.G,
- et al.
- TIMI Investigators,
- Williams D.O,
- Braunwald E,
- Knatterud G,
- et al.
- Kalbfleisch J.D,
- Prentice R.L
- Marubini E,
- Valsecchi M.G
- De Vita C,
- Franzosi M.G,
- Geraci E,
- et al.
- Simoons M.L,
- Serruys P.W,
- van den Brand M,
- et al.
- Rivers J.T,
- White H.D,
- Cross D.B,
- et al.
- White H.D,
- French J.K,
- Haner A.W,
- et al.
- Verheught F.W.A,
- Meijer A,
- Lagrand W.K,
- et al.
- Veen G,
- Meyer A,
- Verheugt A,
- et al.
- Hanson S.R,
- Sakariassen K.S
- Maalej N,
- Folts D
- Makkar R.R,
- Eigler N.L,
- Kaul S,
- et al.
- Stein B,
- Fuster V,
- Israel D.H,
- et al.
- Yao S.K,
- Ober J.C,
- McNatt J,
- et al.