Author + information
- Received August 22, 1997
- Revision received February 19, 1998
- Accepted March 5, 1998
- Published online June 1, 1998.
- Allan M Ross, MD, FACCa,
- Conor F Lundergan, MDa,
- Steven C Rohrbeck, MD, FACCa,
- Deneane H Boyle, MPHa,
- Marcel van den Brand, MD∗,
- Christopher H Buller, MD, FACC†,
- David R Holmes Jr, MD, FACC‡,
- Jonathan S Reiner, MD, FACCa,*,
- for the GUSTO-1 Angiographic Investigatorsa
- ↵*Address for correspondence: Jonathan S. Reiner, Division of Cardiology, George Washington University Medical Center, 2150 Pennsylvania Avenue NW, Washington, D.C. 20037
Objectives. We sought to assess the angiographic outcome, complication rates and clinical features of percutaneous transluminal coronary angioplasty (PTCA) after failed thrombolysis for acute myocardial infarction.
Background. “Rescue angioplasty” refers to mechanical reopening of an occluded infarct-related artery (IRA) after failed intravenous thrombolysis. Although the procedure is commonly performed, data describing its technical and clinical outcome are sparse. Early reports suggested that rescue PTCA is less often successful and produces more complications than primary PTCA. Other reports have described beneficial effects of successful rescue PTCA but adverse outcomes when PTCA is unsuccessful.
Methods. Using data from the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO-1) angiographic substudy, we compared clinical and angiographic outcomes of 198 patients selected for a rescue PTCA attempt with those of 266 patients with failed thrombolysis but managed conservatively and, for reference, with those of 1,058 patients with successful thrombolysis.
Results. Patients offered rescue PTCA had more impaired left ventricular function than those in whom closed vessels were managed conservatively. Rescue successfully opened 88.4% of closed arteries, with 68% attaining Thrombolysis in Myocardial Infarction (TIMI) grade 3 flow. The interventions did not increase catheterization laboratory or postprocedural complication rates. Multivariate analysis identified severe heart failure to be a determinant of a failed rescue attempt. Successful rescue PTCA resulted in superior left ventricular function and 30-day mortality outcomes, comparable to outcomes in patients with closed IRAs managed conservatively, but less favorable than in patients in whom thrombolytic therapy was initially successful. The mortality rate after a failed rescue attempt was 30.4%; however, five of the seven patients who died after failed rescue PTCA were in cardiogenic shock before the procedure.
Conclusions. Rescue PTCA tends to be selected for patients with clinical predictors of a poor outcome. It is effective in restoring patency. Patients who die after a failed rescue attempt are often already in extremis before the angioplasty attempt.
“Rescue angioplasty” refers to mechanical reopening of an occluded infarct-related artery (IRA) after failed intravenous thrombolysis. Although this procedure is common, data describing its technical and clinical outcomes are sparse. Early reports have suggested that percutaneous transluminal coronary angioplasty (PTCA) of the IRA is less often technically successful and more often associated with complications when performed early after intravenous thrombolytic therapy than when performed de novo (i.e., primary PTCA) (1–3). Support for the efficacy of rescue PTCA has been provided by Ellis et al. (4)in a small and highly selected group. This report describes the angiographic outcome, complication rates and clinical features regarding PTCA after failed thrombolysis in 198 patients treated in the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO-1) angiographic trial (5). Uniquely, the present study is prospective, compares results according to randomized thrombolytic assignment and includes protocol-driven follow-up angiography.
The performance of and outcome from rescue PTCA was a prespecified end point of the GUSTO-1 angiographic study of four thrombolytic regimens in patients with acute myocardial infarction (MI). The study was conducted in 76 centers in North America, Europe and Australia participating in the GUSTO-1 mortality trial. The entry criteria have been previously described (5,6); briefly, the study enrolled patients with ST segment elevation and chest pain duration <6 h. The study was approved by each site’s Institutional Review Board, and all patients gave written informed consent.
Treatment and angiographic randomization
The therapeutic and angiographic randomization strategies have been previously described (5,6). Patients received one of four treatment regimens: streptokinase (SK) (Kabbekinase, Kabi Vitrum, Sweden) with intravenous heparin (sodium heparin, Sanofi, Paris), adjusted to maintain an activated partial thromboplastin time between 60 and 80 s; SK with subcutaneous heparin, 12,500 U twice daily, beginning 4 h after thrombolytic initiation; tissue-type plasminogen activator (t-PA) (alteplase, Genentech) given in an accelerated dosing regimen with intravenous heparin; and a combination arm consisting of reduced-dose SK and t-PA with intravenous heparin. At the time of thrombolytic treatment randomization, patients were also randomized to one of four times for initial angiography after thrombolytic initiation. The times were 90 min, 180 min, 24 h and 5 to 7 days. All patients enrolled in the 90-min group were also scheduled for follow-up angiography 5 to 7 days later.
Coronary angiography and core angiographic laboratory procedures
This report concerns patients who were enrolled in the angiographic substudy and who were assigned to angiography 90 or 180 min after thrombolytic therapy. Rescue PTCA was defined as attempted guide wire crossing and balloon dilation of a closed IRA (Thrombolysis In Myocardial Infarction trial [TIMI] flow grade 0 or 1) undertaken within 6 h from the start of thrombolytic therapy. By protocol, rescue PTCA was permitted but not required when an occluded IRA was encountered. Therefore, the patients in this report underwent mechanical reperfusion at the operator’s discretion and not by random assignment. Rescue PTCA success was defined as attainment of TIMI flow grade 2 or 3 in the infarct-related vessel.
Details of core laboratory procedures have been previously described (5). Briefly, angiographic films were interpreted by an experienced angiographer (C.F.L., J.S.R. or A.M.R.) who had no knowledge of the treatment, angiographic randomization and clinical outcome. The first injection of the IRA was used to determine the TIMI flow grade (7). The final angiographic sequence filmed after rescue PTCA was used to determine the TIMI flow grade resulting from the intervention.
Two orthogonal views of the IRA were digitally acquired and quantitatively analyzed. Quantitative coronary analysis was performed, after manual centerline placement, using a first- and second-derivative edge-detection algorithm (ARTREK, University of Michigan). Lesion variables obtained included percent diameter stenosis and minimal lumen diameter. For determination of percent diameter stenosis, a “normal” reference segment was identified proximal to the stenosis. When two views were technically adequate, the more severe measured stenosis of the two views was used for this analysis.
End-systolic and end-diastolic ventriculographic silhouettes, defined by the core laboratory angiographer, were digitized and stored. The analysis was limited to images derived from a 30° right anterior oblique ventriculogram. Ejection fraction was calculated by software (VENTREK, University of Michigan) using the area–length method (8). Regional wall motion, defined as the mean excursion of the most abnormal 50% of chords, was calculated using the method described by Bolson et al. (9). Additional characterization of wall motion included determining the number of consecutive chords in the infarct zone >2 SD below the norm.
Because clinical outcomes of the two SK groups in the mortality trial were similar (6)and because patients assigned to SK with subcutaneous heparin received intravenous heparin in the angiographic substudy to support the early diagnostic and particularly interventional procedures, both SK groups were considered comparable and are combined in this analysis.
Results are expressed as the mean value ± SD unless otherwise noted. The chi-square test was used to compare categoric variables. Pairwise comparisons of more than two proportions were analyzed using the Bonferroni adjustment. Continuous end points of only two groups were compared by using the unpaired Student ttest; all p values are two-tailed. Continuous end points of more than two groups were compared by using analysis of variance and the Tukey posterior test. A p value ≤0.05 was considered significant. Statistical analysis was performed utilizing SAS software (SAS Institute, Inc.).
Of the 1,613 patients randomized to undergo protocol angiography at 90 or 180 min after initiation of thrombolytic therapy, 1,522 patients (94%) underwent the procedure within 6 h. Of this group, 464 patients (30%) had closed infarct-related vessels (TIMI flow grade 0 or 1); of these, 198 (43%) had attempted rescue PTCA.
Table 1displays the incidence of failed thrombolysis and rescue PTCA for the treatment assignments. Rescue PTCA was attempted when a closed IRA was encountered in a similar proportion across all treatment assignments.
With the exceptions of the incidence of diabetes, which was more common in patients in whom rescue PTCA was attempted, and of a strong trend for angioplasty to be offered to patients with hypotension, the baseline demographic characteristics were similar between patients selected for a rescue PTCA attempt and those with closed vessels managed by noninterventional methods (Table 2).
Table 3lists angiographic and ventriculographic characteristics of the two groups of patients with failed thrombolysis (those in whom rescue PTCA was attempted and those managed conservatively). By all measures of global or regional systolic function, patients with more impaired left ventricular function were those more frequently offered rescue PTCA.
Table 4details the technical results of rescue according to thrombolytic treatment assignment. Rescue successfully reopened 175 (88.4%) of 198 closed arteries (conversion to TIMI flow grades 2 or 3), with 41 (20.7%) of 198 becoming TIMI flow grade 2 and 134 (67.7%) of 198 becoming TIMI flow grade 3. We observed no thrombolytic drug-specific differences in patency outcomes after rescue PTCA. The quantitative angiographic indices of minimal lumen diameter and percent diameter stenosis after rescue PTCA were also similar across treatment groups. Patients who had successful rescue PTCA left the catheterization laboratory with a larger lumen in the infarct-related vessel compared with those who had thrombolytic success (minimal lumen diameter 1.40 ± 0.44 vs. 0.82 ± 0.47, p = 0.0001; percent diameter stenosis 46.2 ± 13.7 vs. 66.3 ± 14.1, p = 0.0001).
For patients with successful rescue PTCA assigned to initial angiography at 90 min, protocol-directed follow-up angiography was performed in 116 (66.3%) of 175 at a mean of 5.9 ± 2.4 days after enrollment. Of this group, 13 (11.2%) of 116 were found to have reoccluded infarct-related vessels. Three of the 29 vessels (10.3%) that were TIMI flow grade 2 after rescue PTCA became reoccluded, compared with 10 (11.5%) of 87 vessels that were TIMI flow grade 3 (p = 0.87). For comparison, patients who had successful thrombolysis and who underwent protocol-driven follow-up angiography had a reocclusion rate of 5.6%.
Table 5compares the baseline and angiographic characteristics of patients undergoing attempted rescue PTCA, stratified by whether the procedure was ultimately successful or unsuccessful. Multivariable logistic regression analysis found only the presence of severe heart failure (Killip class >2) to be a determinant of a failed rescue PTCA procedure (odds ratio 0.146, 95% confidence interval 0.035 to 0.631).
The rates of procedure-related complications for rescued and nonrescued patients with thrombolytic failure are shown in Table 6. For reference, this table also shows complication rates for patients with successful thrombolysis undergoing only diagnostic catheterization at 90 or 180 min after thrombolytic initiation. Overall, the incidence of complications related to rescue PTCA was low and similar to that of patients not undergoing rescue PTCA.
Table 7shows measures of left ventricular function obtained in patients undergoing scheduled follow-up ventriculography at 5 to 7 days. Successful rescue PTCA resulted in superior convalescent left ventricular function outcomes, comparable to patients with closed IRAs managed conservatively, but less favorable than in patients in whom thrombolytic therapy was initially successful.
Survival data at 30 days for the rescue group and for the entire 90-min and 180-min angiographic cohort are displayed in Figure 1. Patients with successful thrombolysis had the lowest 30-day mortality rate, followed by those with successful rescue PTCA and those in whom no rescue was attempted. Patients with failed rescue attempts had the highest death rate. Of significance is that five of the seven patients who died after rescue failure were noted to have been in cardiogenic shock at the start of the catheterization laboratory procedure (Table 8). The overall impact of shock on mortality in this series is displayed in Table 9.
Intraaortic balloon pumps were used frequently in patients in cardiogenic shock (11 of 20) undergoing rescue PTCA. Nine of these patients died nonetheless; of the nine other patients in shock not treated with a balloon pump, eight died.
In no rescue failure fatality was there evidence to implicate that the procedure was a technical contributing factor in the death.
This study describes the largest prospective and prespecified series to date of rescue PTCA procedures. It uniquely incorporates systematic, protocol-driven, early and convalescent angiographic and clinical outcome measures. In addition, it provides results stratified by randomly assigned thrombolytic treatment regimens.
Success of rescue PTCA
The findings provide strong evidence that rescue PTCA in patients who have failed to reperfuse after intravenous thrombolytic therapy for acute MI is very effective in restoring patency and that success is not influenced by the specific thrombolytic agent used. In an earlier study of 78 patients undergoing rescue PTCA, Ellis et al. (4)reported an angioplasty success rate of 92%. The Cohort of Rescue Angioplasty in Myocardial Infarction (CORAMI) study group reported a success rate of 90% (10). Recently, the TIMI group reported a success of 90% from 58 rescue attempts in the TIMI-4 trial (11). In the present study, PTCA restored TIMI flow grade 2 or 3 in 88% of patients. Our data, as well as those of the earlier, small studies, contradict the notion that rescue PTCA will not be as technically successful as in those referred for PTCA without antecedent lytic drugs (“primary PTCA”). A review of the published data reveals an aggregate success rate of 94% from five small primary PTCA studies (12–16). In the large Global Use of Strategies To Open occluded arteries in acute coronary syndromes (GUSTO IIb) angiographic substudy (17), the success rate for primary PTCA was 93% with a TIMI flow grade 3 outcome in 73%. In our rescue PTCA group, the aggregate rate of TIMI flow grade 3 was 68%. It is important to emphasize that the previous studies used differing definitions of success and only three of these studies (10,12,17)used the attainment of TIMI flow grade 2 or 3 in the definition of primary PTCA success.
Complications of rescue PTCA
Our findings also help dispel the concept that noncoronary complication rates after rescue PTCA are high. The consensus that PTCA in a thrombolytic milieu is a dangerous combination is derived principally from two trials: TIMI IIA (1)and Streptokinase Angioplasty Myocardial Infarction trial (SAMI) (2), both of which reported significantly higher rates of bleeding and emergency coronary artery bypass graft surgery in patients undergoing PTCA in the perithrombolytic period. In SAMI, 122 patients were randomized to direct PTCA or 1.5 million U of SK followed by angioplasty of the infarct-related vessel. The trial found no significant differences in PTCA success (98% for SK plus angioplasty vs. 93% for angioplasty alone). The group given thrombolytic therapy before PTCA did, however, demonstrate a substantially increased incidence of emergency bypass surgery (10% vs. 2%, p = 0.03) and bleeding complications requiring transfusion (39% vs. 8%). In contrast, in the present study, the rate of emergency bypass surgery in patients undergoing rescue PTCA was only 1%, and the incidence of major bleeding was 8.6%, equivalent to the rate in patients undergoing protocol-driven diagnostic angiography alone.
We found that patients with thrombolytic success fared better in terms of survival and convalescent ventricular function than those whose infarct-related vessel patency was achieved only after rescue PTCA. This observation is consistent with the advantage in time to patency in those patients with infarct-related vessels already patent before arrival at the angiographic laboratory compared with those who required a catheterization laboratory rescue to establish IRA reperfusion. Our data suggest that PTCA was usually selected for patients with clinical predictors of a poor outcome (significantly worse global and regional ventricular function). Successful rescue PTCA appears to yield improved convalescent left ventricular function, which becomes equivalent to that in patients not offered rescue PTCA (who had significantly better left ventricular indexes at the time of the early angiography).
Outcome after failed rescue PTCA
An intriguing finding in this and previous reports is the very poor outcome in the rescue-attempted-but-failed cohort. In our group of 23 such patients, 7 (30.4%) died. A report from the Thrombolysis and Angioplasty in Myocardial Infarction (TAMI) trials (18)found a similarly high mortality rate of 39.1% for patients with failed rescue PTCA. Gorfinkel et al. (19)recently reported, in a series of 125 patients, a mortality rate of 37.5% for those with failed rescue PTCA. Although it has been postulated that various components of the angioplasty procedure itself (negative inotropic effect of contrast, blood loss, intraluminal trauma) may contribute to the observed high mortality in this group of patients (18), our series indicates that the patients who die are already in extremis before the angioplasty attempt. In the present study, cardiogenic shock was already established before arrival in the catheterization laboratory in five of the seven patients who died after a failed rescue PTCA procedure. Intraaortic balloon pumping, often employed in such patients (20), did not provide evidence of benefit in this cohort.
These observations, although providing a powerful data base on the technical aspects of rescue PTCA, are only inferentially informative regarding the clinical utility of the procedure, because discovery of failed thrombolysis in this series was followed by an operator-determined clinical decision (not a randomization) to leave the infarct-related vessel closed or to attempt mechanical reopening. Baseline imbalances, particularly in ventricular function, prohibit reaching firm conclusions from the comparison of rescued versus nonrescued patient groups, but suggest that PTCA is selected for patients with clinical predictors of a poor outcome and is usually successful in restoring coronary blood flow.
☆ This study was funded by a combined grant from Bayer (New York, New York), CIBA-Corning (Medfield, Massachusetts), Genentech (South San Francisco, California), ICI Pharmaceuticals (Wilmington, Delaware) and Sanofi Pharmaceuticals (Paris, France).
- Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries
- infarct-related artery
- myocardial infarction
- percutaneous transluminal coronary angioplasty
- Thrombolysis in Myocardial Infarction trial
- tissue-type plasminogen activator
- Received August 22, 1997.
- Revision received February 19, 1998.
- Accepted March 5, 1998.
- by the American College of Cardiology
- O’Neill W.W,
- Weintraub R,
- Grines C.L,
- et al.
- Ellis S.G,
- da Silva E.R,
- Heyndrickx G,
- et al.
- Chesebro J.H,
- Knatterud G,
- Roberts R,
- et al.
- 9.↵Bolson EL, Kliman S, Sheehan F, Dodge HT, editors. Left ventricular segmental wall motion—a new method using local direction information. In: Computers in Cardiology. IEEE Trans Biomed Eng 1981:245–8.
- 12.↵Grines CL, Browne KF, Marco J, et al., for the Primary Angioplasty in Myocardial Infarction Study Group. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993;328:673–9.
- The Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes (GUSTO IIb) Angiographic Substudy Investigators
- Abbottsmith C.W,
- Topol E.J,
- George B.S,
- et al.
- Holmes D.R,
- Bates E.R,
- Kleiman N.S,
- et al.