Sustained Suppression of Ischemic Complications of Coronary Intervention by Platelet GP IIb/IIIa Blockade With Abciximab
One-Year Outcome in the EPILOG Trial
Background—Blockade of the platelet glycoprotein IIb/IIIa receptor with the monoclonal antibody fragment abciximab was shown in a placebo-controlled randomized trial to reduce the incidence of acute ischemic complications within 30 days among a broad spectrum of patients undergoing percutaneous coronary revascularization. The durability of clinical benefit in this setting has not been established.
Methods and Results—A total of 2792 patients enrolled in the Evaluation in PTCA to Improve Long-term Outcome with abciximab GP IIb/IIIa blockade (EPILOG) trial were followed with maintenance of double-blinding for 1 year. Patients had been assigned at the time of their index coronary interventional procedure to receive placebo with standard-dose, weight-adjusted heparin (100 U/kg initial bolus), abciximab with standard-dose, weight-adjusted heparin, or abciximab with low-dose, weight-adjusted heparin (70 U/kg initial bolus). The primary outcome was the composite of death, myocardial infarction, or urgent repeat revascularization by 30 days; this composite end point and its individual components were also assessed at 6 months and 1 year. Rates of any repeat revascularization (urgent or elective), target vessel revascularization, and a composite of death, myocardial infarction, or any repeat revascularization were also reported. Follow-up at 1 year was 99% complete for survival status and 97% complete for other end points. By 1 year, the incidence of the primary composite end point was 16.1% in the placebo group, 9.6% in the abciximab with low-dose heparin group (P<0.001), and 9.5% in the abciximab with standard-dose heparin group (P<0.001). Each of the components of this composite end point was reduced to a similar extent. Nonurgent or target vessel repeat revascularization rates were not significantly decreased by abciximab therapy. Mortality rates over 1 year increased with increasing levels of periprocedural creatine kinase MB fraction elevation.
Conclusions—Acute reductions in ischemic events after percutaneous coronary intervention by abciximab are sustained over follow-up to at least 1 year. Early periprocedural myocardial infarctions suppressed by this therapy are associated with long-term mortality rates.
Blockade of the platelet surface membrane glycoprotein IIb/IIIa complex with abciximab (c7E3 Fab, ReoPro, Centocor, Malvern, Pa), a human-murine chimeric antibody Fab fragment, has been demonstrated in 4 large-scale placebo-controlled trials to markedly reduce the incidence of acute ischemic complications in the setting of percutaneous coronary revascularization.1 2 3 4 In the EPILOG (Evaluation in PTCA to Improve Long-term Outcome with abciximab GP IIb/IIIa blockade) trial, the 30-day composite end point of death, myocardial infarction, or emergency repeat revascularization among patients undergoing elective or urgent coronary intervention was reduced by 56% relative to placebo by a bolus and 12-hour infusion of abciximab.2 By 6 months of follow-up, the reduction in acute ischemic end points appeared to have been maintained, although no differences were present among the treatment groups with respect to the need for elective target vessel revascularization procedures.
To determine whether the marked acute benefit of abciximab therapy observed in EPILOG was preserved over the long term, double-blind status was maintained and 1-year follow-up was performed among patients enrolled in the trial. Moreover, the relation between suppression of periprocedural myocardial infarction by GP IIb/IIIa blockade and late cardiovascular events was assessed.
The details of the EPILOG trial have been previously described.2 In brief, 2792 patients undergoing elective or urgent percutaneous coronary revascularization were enrolled at 69 investigational sites in the United States and Canada. Patients with acute myocardial infarction or unstable angina with associated ECG changes during the previous 24 hours were excluded. Other exclusion criteria included planned stent implantation or rotational atherectomy, percutaneous coronary intervention performed within the prior 3 months, or conditions that would be associated with excessive bleeding risk.2 The protocol was approved by the Institutional Review Board at each clinical site, and patients gave informed consent for participation in the trial and follow-up at 1 year.
All patients were treated with oral aspirin at least 2 hours before the index procedure and for at least 6 months thereafter. Randomization was done by means of a central telephone hotline in a double-blind fashion to 1 of 3 treatment groups: placebo with standard-dose, weight-adjusted heparin; abciximab with standard-dose, weight-adjusted heparin; or abciximab with low-dose, weight-adjusted heparin. Abciximab was administered as a 0.25 mg/kg bolus 10 to 60 minutes before balloon inflation or device activation, followed by a 0.125 μg/kg per minute (maximum 10 μg/min) infusion for 12 hours. The standard-dose weight-adjusted heparin regimen consisted of an initial heparin bolus before the interventional procedure of 100 U/kg (maximum 10 000 U), with additional weight-adjusted boluses according to an algorithm intended to achieve and maintain an activated clotting time ≥300 seconds over the duration of the procedure. The low-dose weight-adjusted heparin group received an initial bolus of 70 U/kg (maximum 7000 U), with additional boluses as necessary to achieve and maintain an activated clotting time ≥200 seconds. To preserve blinding of all investigators and personnel involved in patient management, a “heparin coordinator” at each clinical site performed all activated clotting time measurements and directed heparin administration. Postprocedural use of heparin was discouraged, and vascular sheaths were to be removed within 2 to 6 hours (during abciximab infusion). Specific guidelines or algorithms were provided for management of vascular access sites, uncontrolled bleeding, urgent coronary artery bypass surgery, thrombocytopenia, or blood transfusions. Stents were used in 13% of patients to maintain patency after manifest or threatened abrupt closure.
Study End Points
The primary efficacy end point was a composite of death from any cause, myocardial infarction, or severe myocardial ischemia requiring urgent surgical or repeat percutaneous coronary revascularization within 30 days of randomization. A secondary end point was the composite of death, myocardial infarction, or any repeat percutaneous or surgical revascularization within 6 months of randomization. Both of these end points, as well as their individual components, were assessed at 1-year follow-up. End point classifications at all time points were made by a clinical events committee, which was blinded to study group allocation.
An end point in-hospital myocardial infarction was defined by 1 of 2 criteria2 : (1) new significant Q waves in ≥2 contiguous ECG leads or (2) elevation in creatine kinase (CK) or its MB isoenzyme ≥3 times the upper limit of local normal, representing an increase of ≥50% over the previous nadir level, in 2 samples collected at different times. After hospital discharge, myocardial infarction was defined by the development of new significant Q waves or MB isoenzyme elevation to >2 times the upper limit of normal. The MB isoenzyme value was used unless unavailable, in which case total CK was used.
Data Collection and Statistical Analysis
All patients were contacted by the study sites by telephone or written questionnaire between 366 and 428 days after randomization for assessment of 1-year outcomes. Survival status and the occurrence of cardiac events, including myocardial infarction, revascularization, cardiac catheterization, and rehospitalization were ascertained. Relevant hospital records were obtained for source documentation of any rehospitalization, outpatient catheterization, or outpatient revascularization procedure; death certificates and autopsy reports, if available, were obtained for any patients who died during the follow-up period. Data were collected on case report forms by study coordinators at the clinical sites. Investigators and study coordinators remained blinded to treatment allocations of individual patients until 1-year follow-up, database entry, and clinical events committee review were finalized.
Differences among patients with regard to efficacy were examined according to an intention-to-treat analysis. To be considered complete, the 1-year follow-up was required to have occurred at least 335 days after randomization. Survival analysis methods were used at 1 year in the same fashion as for the 30-day and 6-month analyses.2 Pairwise comparisons between the 2 abciximab treatments and the placebo treatment were made with the use of the log-rank test, with event rates calculated by the Kaplan-Meier method. Two-sided probability values are reported. Hazard ratios and 95% CIs were calculated to assess consistency of treatment effect across subgroups.
For the purpose of assessing correlation of periprocedural or early myocardial infarction with long-term clinical outcome, late clinical event rates (death, myocardial reinfarction, or repeat revascularization) were calculated by the Kaplan-Meier method and analyzed as a function of the ratio of peak periprocedural CK-MB over the upper limit of normal. Two separate analyses were performed: the first correlated periprocedural CK-MB elevations within 24 hours of randomization with subsequent end point events occurring between 24 hours and 1 year (excluding events occurring within the first 24 hours), and the second correlated early CK-MB elevations within the primary efficacy end point period of 30 days of randomization with end point events occurring between 30 days and 1 year (excluding events within the first 30 days). Results are expressed as risk ratios and 95% CIs.
Primary 30-Day Efficacy Analysis and Completeness of Follow-Up
Enrollment took place between February 28 and December 14, 1995. The trial had been terminated early on the recommendation of the Data and Safety Monitoring Committee, after the first interim analysis on December 11, 1995, demonstrated an unexpectedly strong 30-day treatment effect among the first 1500 patients. The results of the primary efficacy end point analysis at 30 days have been previously reported in detail.2 In brief, the incidence of the composite end point of death, myocardial infarction, or urgent revascularization for severe myocardial ischemia at 30 days was 11.7% in the placebo group, 5.2% in the abciximab with low-dose heparin group (P<0.0001), and 5.4% in the abciximab with standard-dose heparin group (P<0.0001).
The 1-year follow-up database was locked in April 1997. Of the 2792 patients enrolled, survival status at 335 days was unknown in 27 patients (1.0%), for whom the median follow-up was 207 days. For assessment of events other than death, a total of 75 (2.7%) patients had fewer than 335 days of follow-up with no death, myocardial infarction, or repeat revascularization reported before their last follow-up. Rates of completeness of follow-up were similar among the 3 treatment groups.
Efficacy Analysis at 1 Year
The incidence of the primary composite end point of death, myocardial infarction, or urgent revascularization (the composite used as the primary efficacy end point at 30 days) was 16.1% in the placebo group, 9.6% in the abciximab with low-dose heparin group (hazard ratio 0.56; 95% CI, 0.43 to 0.73, P<0.001), and 9.5% in the abciximab with standard-dose heparin group (hazard ratio 0.57; 95% CI, 0.43 to 0.74, P<0.001) by 1-year follow-up (Figure 1⇓). Each of the components of this composite end point was reduced to a similar extent (Table 1⇓). The treatment effect of the combined abciximab groups relative to placebo with regard to this end point was homogeneous across patient groups defined by age, sex, body weight, or indication for revascularization (Figure 2⇓).
A composite end point of death, myocardial infarction, or any revascularization (urgent or elective) was also assessed. By 1 year, this end point had occurred in 32.4% of patients randomized to placebo, 29.4% of patients randomized to abciximab with low-dose heparin (P=0.093), and 29.2% of patients randomized to abciximab with standard-dose heparin (P=0.077). The types and circumstances of revascularization procedures by 1 year are detailed in Table 2⇓. Overall revascularization rates were not different among the 3 treatment groups; a significant treatment effect of abciximab was observed only for urgent revascularization.
Event rates for the primary composite and selected individual end points at 30 days, 6 months, and 1 year are illustrated in Table 3⇓. In the placebo group, the timing of acute ischemic complications was clustered early in the course after coronary intervention: 72% of the primary composite end point events had accrued during the first 30 days (54% within 24 hours), with an additional 19% occurring between 30 days and 6 months and 9% between 6 months and 1 year. Therapy with abciximab resulted in a marked reduction in the risk of complications during the first 30-day period, after which time incremental event rates were essentially equal among the 3 treatment arms. Thus for the primary composite efficacy end point, the treatment effect achieved by abciximab early (at 30 days) was maintained without attenuation throughout the 1-year follow-up period (Figures 1⇑ and 3⇓): the absolute reduction in events (number of events prevented per 100 patients treated) in the combined abciximab groups versus placebo was 6.40 at 30 days, 6.35 at 6 months, and 6.55 at 1 year. Similarly, the absolute treatment benefit for the individual components of this primary composite end point was preserved to 1 year (Figure 3⇓), although there was a trend toward increased abciximab treatment effect in preventing death over long-term follow-up (events prevented per 100 patients treated=0.45 at 30 days and 0.85 at 1 year). In contrast, rates of repeat target vessel revascularization converged somewhat after 30 days (Figure 3⇓ and Table 3⇓), with no incremental benefit out to 6 months and 1 year, indicating that abciximab had no effect on the incidence of “clinical restenosis.”
Periprocedural or Early CK-MB Elevation and 1-Year Outcome
The relation between early myocardial infarction and long-term mortality rates, reinfarction, and repeat revascularization was evaluated as a function of the magnitude of CK-MB elevation above the upper limit of normal. Table 4⇓ and Figures 4A⇓ and 5A⇓ correlate periprocedural myocardial infarctions occurring within the first 24 hours after randomization with subsequent events after 24 hours; Table 5⇓ and Figures 4B⇓ and 5B⇓ correlate early myocardial infarctions occurring within 30 days after randomization with subsequent ischemic events after 30 days. Of the 533 patients experiencing CK-MB elevations within the first 30 days and surviving beyond 30 days, 478 (89.7%) had such elevations occur within the periprocedural 24-hour period.
The risk of death by 1 year was significantly increased among patients who experienced CK-MB elevation during either the periprocedural (first 24 hours, Figures 4A⇑ and 5A⇑) or early follow-up (first 30 days, Figures 4B⇑ and 5B⇑) period. Although the magnitude of increased risk of death was proportional to the extent of CK-MB elevation above the upper limit of normal, there was a significant hazard associated even with mild rises in CK-MB. The largest periprocedural or early myocardial infarctions (CK-MB ≥10 times control) were predictive of an 8-fold increased risk in death after the first 24 hours and an approximately 3-fold increased risk of death after the first 30 days.
Periprocedural or early CK-MB elevation was also associated with the risk of subsequent myocardial reinfarction (Tables 4⇑ and 5⇑), although the hazard ratios tended to be lower than those for late death and the gradient of reinfarction risk with increasing periprocedural CK-MB levels was less distinct. There was no significant association between periprocedural or early CK-MB elevations and subsequent repeat revascularization procedures; a nonsignificant trend was observed, however, toward fewer revascularization procedures with increasing early CK-MB levels (Table 5⇑).
The acute efficacy of platelet GP IIb/IIIa receptor blockade with abciximab among patients undergoing percutaneous coronary revascularization has been unequivocally demonstrated among almost 9000 patients.1 2 3 4 In EPILOG, rates of important ischemic end points among a broad spectrum of patients undergoing intervention were reduced by >50% in the first 30 days after the procedure.2 The current analysis demonstrates that the acute efficacy of this agent is durable over the long term. An absolute treatment effect of 6.4 acute ischemic events (death, myocardial infarction, or urgent repeat revascularization) prevented per 100 patients treated with abciximab was observed early in their clinical course and was maintained without attenuation over 1-year follow-up. All patient subgroups appeared to derive similar long-term benefit. Nonurgent revascularization procedures, however, were not prevented by abciximab, and there were no differences by 1 year among the treatment groups with regard to the overall need for surgical or repeat percutaneous revascularization procedures.
The mechanism by which a relatively brief period of intense GP IIb/IIIa receptor blockade by abciximab leads to sustained clinical benefit can be inferred from the time course of ischemic complications in EPILOG. Ischemic events tended to take place early after the interventional procedure, with nearly three quarters of the events that would ultimately occur over 1 year in the placebo group clustered within the first 30 days. During this early postprocedural time period, abciximab exerted its protective effect, reducing the incidence of ischemic end points by 56%. After 30 days, ischemic events occurred relatively infrequently and at rates that were essentially equivalent in the placebo and abciximab treatment groups; incremental benefit of abciximab in preventing these later events or elective target vessel revascularization (the clinical sequelae of restenosis) was not observed. This lack of influence of abciximab on complications after the acute phase is not surprising; despite the high avidity of abciximab binding to the GP IIb/IIIa receptor, with receptor occupancy observed by flow cytometry for over 15 days,5 inhibition of platelet aggregation by this agent is measurable for only 72 to 96 hours after the end of infusion.5 6 Thus clinical benefit is derived from abciximab therapy because the period of maximal risk is relatively short after coronary intervention, during which time this agent exerts a potent inhibitory effect on coronary platelet thrombus formation. Long-term durability of the clinical benefit suggests that the injured coronary plaque is largely passivated after this early period, reducing the stimulus for new thrombus formation and a rebound or “catch-up” phenomenon of late ischemic events. Nevertheless, it is notable that event rates nearly doubled between 1 month and 1 year in the abciximab therapy groups (composite end point increasing from 5.3% to 9.6%), demonstrating some degree of continued vulnerability of the coronary plaque even after the periprocedural period of intense platelet inhibition; this finding suggests a potential role for chronic therapy with oral GP IIb/IIIa receptor blockers to suppress late adverse events in these patients.
The long-term preservation of clinical benefit in EPILOG is concordant with the sustained reduction of ischemic events over 3 years in the first trial of abciximab during coronary intervention, EPIC (Evaluation of c7E3 for Prevention of Ischemic Complications).7 In EPIC, therapy with abciximab was also associated with a reduction in the need for target vessel revascularization procedures (“clinical restenosis”), a finding that was not reproduced in EPILOG. Reasons for the disparate effects of abciximab on nonurgent revascularization rates in the EPIC and EPILOG trials are unknown. It may be relevant, however, that placebo group rates of late repeat revascularization were considerably lower in EPILOG than in EPIC (19.4% vs 29.4%, respectively, at 6 months and 26.0% vs 32.6%, respectively, at 1 year), perhaps related to underlying differences in patient populations or improvements in interventional practice or postprocedural care. This improvement in baseline outcome in EPILOG relative to EPIC may have diminished the potential for abciximab to further prevent late revascularization events among patients in the EPILOG trial.
Notably, death was the only clinical end point in EPILOG for which the treatment effect of abciximab trended toward incremental improvement over 1-year follow-up. The mortality rate difference between placebo and abciximab groups nearly doubled between 30 days and 1 year, from 0.45% to 0.85% (Figure 3⇑); this change in the magnitude of the treatment effect of abciximab did not reach statistical significance, perhaps because of the low frequency of death over the relatively short period of follow-up in this population of patients undergoing coronary intervention. The apparent augmentation in mortality benefit over the long term in EPILOG is consistent with that observed in the 3-year follow-up of the EPIC trial7 : although death rates in EPIC were identical at 30 days in the placebo and abciximab bolus and infusion groups, an absolute mortality rate difference of 1.8% was present at 3 years. When the mortality rates among treated patients at 1 year in EPILOG are pooled with those at 3 years in EPIC, the hazard ratio for death in the patients randomized to abciximab compared with placebo is 0.73 (P=0.049).
Given that any antiplatelet effect of the 12-hour periprocedural abciximab infusion persists for no longer than 2 to 3 weeks, the apparent increasing mortality benefit with this agent would not be expected to be due to a direct effect of abciximab on the vascular wall during the year after coronary intervention. Instead, it might be postulated that late mortality rate reduction is a downstream consequence of suppression of early ischemic events. The findings of this analysis support that hypothesis. The most frequent major ischemic complication in the first 30 days after revascularization in EPILOG was myocardial infarction, prospectively classified as Q-wave, large non–Q-wave (CK-MB >5 times the control value), or small non–Q-wave (CK-MB=3 to 5 times control). Treatment with abciximab markedly reduced the incidence of all myocardial infarctions by 30 days, with the predominant effect (in terms of absolute number of events prevented) on the large infarctions (Q-wave from 0.8% to 0.45%, large non–Q-wave from 5.6% to 2.3%, and small non–Q-wave from 1.9% to 1.1%). The 1-year analysis presented here demonstrates that the occurrence of periprocedural myocardial infarction significantly increased the risk for later death and that the magnitude of increased risk was proportional to the degree of early CK-MB enzyme elevation. Peak CK-MB levels that qualified as large non–Q-wave infarctions (CK-MB >5 times control), the predominant event inhibited by abciximab therapy in EPILOG, were associated with a 5-fold increase in the risk of death after the first 24 hours and a nearly 3-fold increase in late death after 30 days. Importantly, however, no safe threshold of enzyme elevation was apparent, with adverse late outcome observed even among patients with myocardial necrosis producing levels of CK-MB just over normal values.
The clinical importance of periprocedural non–Q-wave myocardial infarctions after percutaneous coronary revascularization has been controversial. Early small reports with limited or no long-term follow-up suggested that these events were not associated with clinical sequelae.8 Subsequently, however, virtually every study that has examined the impact of periprocedural enzyme release over an adequate follow-up period has demonstrated that patients who have myocardial infarction during or after coronary intervention are at significantly greater risk for late cardiac death than those who do not.7 9 10 11 12 13 14 Some studies have suggested that late consequences of periprocedural CK elevation are not fully manifest until after 1 year.7 9 10 12 In the 3-year follow-up report of the EPIC trial, for example, an association between periprocedural myocardial infarction and subsequent death was clearly observed, but survival curves for patients treated with placebo or abciximab did not begin to diverge until after the first year.7 The apparent divergence of mortality rates among treatment groups in EPILOG within the first year may reflect the fact that infarctions were suppressed to a greater extent in EPILOG than in EPIC (57% vs 40% relative risk reductions, respectively). Taken together, the findings over long-term follow-up of patients in the EPIC and EPILOG trials provide compelling evidence that CK-MB enzyme release after coronary intervention is a clinically relevant event and suggest that prevention of periprocedural myocardial necrosis may be linked to long-term reduction in cardiac mortality rates.
The association between early CK-MB elevations and late ischemic events other than death is less clear. In EPILOG, only an inconsistent trend was observed toward a greater risk of reinfarction with increasing levels of periprocedural CK-MB, a finding that suggests that the mechanism of late death in patients with periprocedural infarction is not mediated by recurrent infarction. An interesting inverse relation was noted between the occurrence and severity of early CK-MB elevations and subsequent revascularization procedures, however, in that patients with periprocedural infarction less frequently underwent repeat revascularization. It is probable that patients with CK-MB elevations during their index procedure had less viable myocardium in the distribution of their treated vessels, were therefore less likely to develop signs or symptoms of recurrent ischemia if restenosis occurred, and were thus less frequently treated by repeat revascularization. This type of reverse interaction between early ischemic events and late revascularization procedures must be considered in the evaluation of new therapies for ischemic heart disease, as interventions such as GP IIb/IIIa blockade that diminish the likelihood of early myocardial infarction may appear to increase the late risk of “restenosis,” whereas in fact only increasing the clinical perception of restenosis and the impetus for treatment.
Blockade of the platelet GP IIb/IIIa receptor with abciximab during urgent or elective coronary intervention in the EPILOG trial markedly diminished the risk of early ischemic complications, with sustained suppression of these events over at least 1-year follow-up. Incremental treatment effect in terms of reduction in the need for elective repeat revascularization procedures was not observed. Early clinical benefit imparted by this agent may be augmented over the long term, as the periprocedural myocardial infarctions that are prevented by therapy with abciximab are clearly correlated with subsequent late death.
This study was supported by Centocor, Inc (Malvern, Pa), and Eli Lilly and Co (Indianapolis, Ind).
↵1 A complete list of the principal investigators and study coordinators of the EPILOG (Evaluation in PTCA to Improve Long-term Outcome with abciximab GP IIb/IIIa blockade) Study Group can be found in N Engl J Med 1997;336:1689–1696.
- Received August 6, 1998.
- Revision received December 30, 1998.
- Accepted January 15, 1999.
- Copyright © 1999 by American Heart Association
Mascelli MA, Lance ET, Damaraju L, Wagner CL, Weisman HF, Jordan RE. Pharmacodynamic profile of short-term abciximab treatment demonstrates prolonged platelet inhibition with gradual recovery from GP IIb/IIIa receptor blockade. Circulation. 1998;97:1680–1688.
Tcheng JE, Ellis SG, George BS, Kereiakes DJ, Kleiman NS, Talley JD, Wang AL, Weisman HF, Califf RM, Topol EJ. Pharmacodynamics of chimeric glycoprotein IIb/IIIa integrin antiplatelet antibody Fab 7E3 in high-risk coronary angioplasty. Circulation. 1994;90:1757–1764.
Topol EJ, Ferguson JJ, Weisman HF, Tcheng JE, Ellis SG, Kleiman NS, Ivanhoe RJ, Wang AL, Miller DP, Anderson KM, Califf RM, for the EPIC Investigator Group. Long-term protection from myocardial ischemic events in a randomized trial of brief integrin β3 blockade with percutaneous coronary intervention. JAMA. 1997;278:479–484.
Abdelmeguid AE, Topol EJ, Whitlow PL, Sapp SK, Ellis SG. Significance of mild transient release of creatine kinase-MB fraction after percutaneous coronary intervention. Circulation. 1996;94:1528–1536.
Harrington RA, Lincoff AM, Califf RM, Holmes DR, Berdan LG, O’Hanesian MA, Keeler GP, Garratt KN, Ohman EM, Mark DB, Jacobs AK, Topol EJ, for the CAVEAT investigators. Characteristics and consequences of myocardial infarction after percutaneous coronary intervention: insights from the coronary angioplasty versus excisional atherectomy trial (CAVEAT). J Am Coll Cardiol. 1995;25:1693–1699.
Tardiff BE, Califf RM, Tcheng JE, Lincoff AM, Sigmon KN, Harrington RA, Mahaffey KW, Teirstein P, Heinsimer J, Topol EJ, for the IMPACT-II Investigators. Post-intervention cardiac enzyme elevations: prognostic significance in IMPACT II. J Am Coll Cardiol. 1996;27:83A. Abstract.
Redwood SR, Popma JJ, Kent KK, Pichard AD, Satler LF, Mintz GS, Hong MK, Bucher TA, Merritt AJ, Leon MB. Predictors of late mortality following ablative new-device angioplasty in native coronary arteries. J Am Coll Cardiol. 1996;27:167A. Abstract.