Christopher N. Conley, MD

Resident Grand Rounds

February 15, 2000

Coronary artery disease is the leading cause of death and disability in the United States, approximately 1 million deaths annually. Estimated costs of loss of productivity and direct costs to the healthcare system are estimated at $287 billion¹. A number of therapeutic interventions have been studied in large trials, and many of them have become standard of care regarding the treatment of acute coronary syndromes. Among the most studied medical therapies are aspirin, anti-thrombotic agents such as heparin (and newer low molecular weight heparins), beta-blockers, and more recently thrombolytic agents.

Unstable angina is the acceleration of stable angina patterns or development of rest and nocturnal angina, and is thought to be due to rupture or fissure at the site of an atherosclerotic plaque, and subsequent platelet adherence and thrombosis. Unstable angina accounts for more than one million hospital admissions annually². Much basic research has been focused on the interaction of platelets with exposed subendothelial materials in efforts to target specific steps in the sequence of platelet adherence and subsequent thrombosis, and thus to specific steps that might be blocked to prevent the potential adverse outcomes of acute myocardial infarction, non-Q wave MI, or death. These efforts uncovered various glycoproteins that, once exposed, interact with platelets inducing binding of the platelets to the tissue, and conformational change in platelet structure to promote further platelet recruitment and thrombus generation.

Glycoprotein IIb/IIIa is one of a number of proteins that have been described in the literature, and platelet receptors for this protein have been characterized and studied extensively, as has the mechanism of binding of the receptor to the ligand. This protein is thought to be central in the mechanism of platelet adherence and activation of the thrombin-generating pathway leading to intermittent thrombosis resulting in unstable angina³. Efforts then turned toward specific ways of blocking this interaction, and several agents with this property have been developed and tested in a variety of clinical settings, including unstable angina.

The purpose of this discussion is to review the various agents studied in this clinical setting, and to examine the clinical trials that assess their efficacy.

• Cyclic heptapeptide derived from venom of the Southeastern pigmy snake
• Platelet function returns to 50% of normal within 4 hours of stopping infusion
• Evaluated as medical therapy for unstable angina, acute MI, and percutaneous intervention

• Studied as medical therapy for unstable angina, and as adjunctive therapy in patients undergoing percutaneous revascularization (IMPACT II), and acute MI (IMPACT- AMI)

• One of the newer agents, a non-peptide compound
• Studied as medical therapy for unstable angina, and acute MI (PARADIGM)

• Nonpeptide derivative of tyrosine
• Bleeding time returns to normal within 4 hours of stopping infusion
• Studied as medical therapy for unstable angina, and as adjunctive therapy in patients undergoing percutaneous revascularization (RESTORE)

• Fab fraction of the human-murine chimeric monoclonal antibody (murine Fab sequences attached by constant regions of human Ig (to decrease immunogenicity))

• Studied extensively in patients undergoing percutaneous revascularization in both stable and unstable angina, and in acute MI (eg. EPIC, EPISTENT, CAPTURE, EPILOG, RAPPORT, ADMIRAL, ERASER, GRAPE)

• No studies as medical therapy without revascularization, thus not discussed in this paper

7. Adverse Effects.

• No intracranial bleeding
• Higher incidence of hematomas in both low-dose (5.2%) and high-dose (6.6%) groups; aspirin (1.4%)
• Patients requiring transfusions: low-dose group (5.2%), high dose group (2.6%), aspirin (1.4%)
• Most bleeding was insignificant and consisted of ecchymoses and hematomas associated with IV lines

8. Comments.

• Small study, primarily used as a "pilot" or "dose-finding" study
• Not powered to detect differences in clinical events
• May provide some pathophysiological reason behind the greater mortality of MI in women, eg. basic research has reported that female platelets bind greater numbers of fibrinogen molecules than men

2. Inhibition of Platelet Glycoprotein IIb/IIIa with Eptifibitide in Patients with Acute Coronary Syndromes⁵

• The data failed to reach statistical significance at a number of endpoints during the study drug infusion, including incidence of death, acute MI, urgent intervention, or recurrent angina.

 

• Also failing to reach significance were a number of endpoints measured at 1 month. Among these are death, acute MI, frequency of intervention for ischemia, recurrent ischemia and events per patient.
• However, there was a difference in the combination endpoint of death/acute MI in the higher dose groups versus placebo at one month, with an incidence of death or MI at 1 month of 2.5% vs 8.1% (OR=0.29, 95% CI 0.09-0.94), NNT 18.

• If you treat 18 patients with one of the two higher dose lamifiban regimens for an average of 84 hours, you will prevent one death or acute MI.

• The lack of statistical significance was likely due to the small study size in comparison to other studies, as most of the comparisons of the treatment groups and placebo groups favored lamifiban, especially the higher doses (see attached Table 2)

7. Adverse Events.

• Significantly higher incidence of minor bleeding (defined as all bleeding that affected the patients' daily activities) in the study drug versus placebo, particularly at the higher doses, 11.7% and 17.1% in groups 4 and 5, respectively, versus 1.6% in the placebo group.

• If you treat seven patients with high dose lamifiban for an average of 84 hours you will cause one minor bleed.

• Major bleeding was higher as well, but to a lesser degree, and seemed to be associated with concomitant use of heparin. The difference was not statistically significant.
• There was no intracranial or life-threatening bleeding.

8. Comments.

• Primarily a dose finding study
• Sustained combined endpoint reduction at 1 month is significant.
• Using the NNT and NNH data above: If you treat 126 patients with unstable angina with high dose lamifiban (group 4 or 5 above) you will prevent 7 deaths or MIs and cause 18 minor bleeds.
• Lower incidence of NQWMI at admission (~14%) than other studies

1. Purpose. To evaluate the clinical efficacy of lamifiban in differing doses and with or without intravenous heparin compared to heparin alone
2. Design. Randomized, double blind, placebo controlled, two-by-two factorial
3. Patients.

• Chest pain within 12 hours of randomization
• Must have ECG changes consistent with ischemia
• Exclusion criteria were similar to other studies
• 2,282 patients enrolled
• Baseline characteristics of each study group were similar in regards to age, sex, weight, past medical history, blood pressure, Killip class at entry, and history of myocardial infarction and prior revascularization

• If you treat 19 patients with lamifiban and heparin for 72 hours you will prevent 1 death or MI within 6 months.

• Other treatment groups (low dose lamifiban without heparin, high dose lamifiban ± heparin) showed no statistically significant difference in either death, myocardial infarction, or both.

7. Adverse Events.

• Adverse events were more likely to occur in the high dose group compared to the placebo group.
• Rates of intermediate and major bleeding were 12.1% compared to 5.5% in the placebo group, P=0.002, NNH 15. The risk of major bleeding or need for transfusion was increased ³ 3 fold in the high dose group compared with placebo.

• In the treatment group that demonstrated composite benefit over 6 months (low dose lamifiban with heparin) rates of major bleeding, red cell transfusions, thrombocytopenia, and stroke were similar.

• Much higher rates of bleeding complications in the high dose ± heparin groups

8. Comments.

• Substantial reduction in combined endpoint at 6 month follow-up suggests promise for lamifiban and heparin combination as therapy for MI
• Seems to contradict with earlier (smaller) Canadian Lamifiban Study; larger trial to address this issue is underway
• Appears to be safe for use with heparin at the smaller dose

• Endpoints at 7 days were not significantly different either, specifically there was no difference in the composite endpoint, refractory ischemia, MI or death (combined), MI, or death. The rates of these outcomes were consistently lower in the tirofiban group, but the study lacked sufficient power to detect a difference
• Similarly, events at 30 days failed to reach statistical significance in all endpoints measured except one. Patients randomized to tirofiban had a lower rate of death than those in the placebo group, 2.3% vs. 3.6%, p=0.02 (ARR 1.3%, NNT 76.9). As in the 7-day data, the rates of the endpoints consistently favored tirofiban but lacked adequate power to show if a difference truly existed.

• Treating 77 people with tirofiban for 48 hours will prevent 1 death within the first month.
• See Survival Curve, Fig 2 from text

• Interestingly, as the patients were followed for the entire 30 days some differences emerged with regards to subsequent treatment.

• Among patients that received medical therapy alone (~62% of patients), tirofiban reduced both death (2.2% vs 4.1%, RR=0.53 (95% CI 0.32-0.89), NNT=53) and death or MI combined (3.6% vs 6.2%, RR=0.58 (95% CI 0.38-0.87), NNT=38).

• In English: If you treat 53 patients with tirofiban (without revascularization) for 48 hours, you will prevent one death at 30 days. Similarly, if you treat 38 patients with tirofiban (without revascularization) for 48 hours you will prevent one death or MI at 30 days.

• An even larger benefit was seen in the composite endpoint in those getting revascularized (~22% of patients in this study): 21.6% vs 27.3%, RR=0.72 (95% CI 0.53-0.98) NNT=18, reaching the composite endpoint.
• Note that the absolute event rate was higher in patients who underwent percutaneous revascularization. This treatment was not randomized and likely represents a subgroup of patients with more severe clinical presentations.
• See Figure 2(attached) for subgroup analysis

7. Adverse Events

• Major bleeding was uncommon in this trial and was no different in the two groups.
• Thrombocytopenia was more common in the tirofiban group, but with return to baseline platelet count within days and without clinical sequelae

8. Comments

• This study involved high risk patients as compared to others, with ~75% of patients with ECG changes and ~25% with MI at enrollment
• Diminishing benefit after 48 hours raises the question of whether a longer infusion would be more beneficial
• Demonstrates clear benefit in several subgroups, trend in most others. The only subgroup that suggested heparin may be better than tirofiban was the subgroup of patients without ECG changes
• Subgroup of revascularized patients received larger benefit, as seen in the PURSUIT trial
• Unlike PURSUIT demonstrated significant benefit in those not revascularized as well

1. Purpose. To evaluate the clinical efficacy of tirofiban with and without heparin compared with heparin alone in patients with unstable angina and non-Q-wave myocardial infarction
2. Design. Randomized, double blind, placebo controlled, multicenter
3. Patients.

• Entry criteria were prolonged anginal chest pain or repetitive angina at rest or with minimal exercise in the previous 12 hours prior to randomization.
• Patients must have new or transient ECG changes or enzyme elevation
• Exclusion criteria were persistent (>20 min) ST segment elevation, recent revascularization, identifiable cause of angina, and various bleeding predispositions as in previous trials.
• Overall, 1915 patients were randomized
• Baseline characteristics of the treatment groups were similar regarding age, sex, race, PMH, cardiac risk factors, ECG changes, and concomitant medical therapy

4. Treatment.

• Randomized to one of the following groups

1. Tirofiban bolus/infusion with heparin placebo bolus/infusion
2. Tirofiban and heparin combined
3. Heparin bolus/infusion with tirofiban placebo bolus/infusion

• Heparin and heparin placebo infusions were monitored and titrated by an unblinded and uninvolved physician
• Aspirin was given to all patients at enrollment and daily thereafter
• Antianginal therapy was directed by the individual physicians.
• Infusions were given for at least 48 hours (mean 71.3 ± 20 hours), and if PTCI performed:

1. the heparin infusion was changed to open-label, and
2. the tirofiban infusion was decreased and continued for an additional 12-24 hours

5. Endpoints.

• Primary endpoints: a composite of death, new MI, and refractory ischemia within 7 days or rehospitalization within 7 days, 30 days, or 6 months
• Secondary endpoints were the same composite endpoint as above but for 48 hours and 30 days, and death, MI, refractory ischemia, death and MI within 48 hours, 30 days, and 6 months

• A final secondary endpoint was the 30-day outcome of patients with refractory ischemia and percutaneous intervention during the initial hospitalization

• Major bleeding was defined as ¯ Hgb of ³ 4mg/dl, transfusion of ³ 2 units of red cells, intracranial or retroperitoneal hemorrhage, or need for surgical intervention. Thrombocytopenia was defined as a platelet count of <90,000/m l.

6. Results.

• Approximately 45% of patients in each group had an admitting diagnosis of non-Q-wave myocardial infarction
• Results consistently favored tirofiban plus heparin versus heparin alone in almost all endpoints at each time interval (See Table 3, attached), although the significance of many of these endpoints varied
• The following are tables of the most significant findings.
• Note that tirofiban-only group was stopped early (after 345 patients enrolled) due to excess mortality in this group

• Of the 16 deaths in this group, 1 was from sepsis, 1 from PE; all others were from cardiovascular causes, including 4 that were peri-intervention
• No identifiable cause of this increased mortality was found
• Contrasts directly with the PRISM study, where tirofiban + ASA was shown to have a significant mortality benefit over 30 days

• During the first 48 hours there was an early benefit shown for both myocardial infarction and the combined of myocardial infarction and death. Those reaching the predefined composite endpoint (death or MI or refractory ischemia) at 48 hours approached but failed to reach significance in favor of the tirofiban + heparin group (5.7% vs. 7.8%, p=0.08). Most of the benefit in the combined group appears to come from the decrease in MI.

• At seven days, the trend of decreasing ischemia in the tirofiban group became significant. The composite endpoint was significant at 7 days, without any difference in the incidence of death. This trend of statistical significance for the reduction in myocardial infarction and recurrent ischemia was continued throughout the trial, but without and difference in death or readmission for unstable angina.

• At 30 days there was a persistent reduction in refractory ischemia, MI, and death and MI endpoints

• Continued trend at the 6-month follow-up.

• No significant difference in incidence of death or MI individually overall in the 6 months of the study
• Subgroup analyses demonstrated a significant benefit of combination therapy in many subgroups, particularly in the aged (³ 65 years), male sex, whites, heavy (>85kg), nondiabetics, nonsmokers, known previous CAD, myocardial infarction at entry, ST segment elevation, and previous beta blocker and aspirin therapy.
• Among those with refractory ischemia, the use of tirofiban plus heparin reduced the percentage of patients taken for revascularization, 67.6% vs 89.4%, RR=0.76, 95% CI 0.59-0.97, NNT=4.6.
• Among those proceeding to revascularization, the tirofiban + heparin group had a lower incidence of the composite endpoint: 8.8% vs 15.3%, RR 0.55, 95%CI 0.32-0.94, NNT=15
• Endpoint frequencies among those medically managed (46% of total patients) were less convincing, with 95%CI of the relative risk crossing 1

7. Adverse Events.

• No intracranial bleeding or deaths from bleeding
• No significant increases in bleeding (either major or minor), need for transfusion, or thrombocytopenia. The absolute rates were somewhat higher in the tirofiban + heparin group, however.

8. Comments

• Combination therapy more likely to be beneficial in high risk patients undergoing percutaneous intervention
• Composite endpoint analysis clouds the picture of which events are actually reduced
• Significant early reduction in death or MI endpoint, which was persistent (though somewhat reduced) over the entire 6 months
• Difficult to explain the early mortality increase in the tirofiban-only group, especially with 1616 patients in the PRISM study showing mortality benefit

PURSUIT, Figure 2.

Subgroups and their Odds Ratios.

Results from the Canadian Lamifiban Study, Table 2.

Data that met statistical significance is summarized in the text. This table shows the consistent benefit of patients treated with lamifiban at varying doses, although many of the data points fell short of statistical significance.

Figure 3 from PRISM