Should Patients in Cardiogenic Shock Undergo Rescue Angioplasty after Failed Fibrinolysis? Comparison of Primary versus Rescue A

A strategy of early revascularization with intensive hemodynamic support saves lives in patients with cardiogenic shock in comparison to conventional medical management.⁶ The SHould we emergently revascularise Occluded Coronaries for cardiogenic shocK (SHOCK) trial demonstrated that in patients with AMI complicated by cardiogenic shock, early mechanical revascularization reduced 6- and 12-month mortality compared with initial medical stabilization, including intra-aortic balloon pump (IABP) and fibrinolytic therapy.⁷ Based on these findings, the American College of Cardiology and the American Heart Association listed early mechanical revascularization as a Class 1A recommendation for patients younger than 75 years with AMI in cardiogenic shock in their revised guidelines.⁸

Restoration of coronary blood flow is a major predictor of survival in patients with cardiogenic shock.⁹ Fibrinolytic therapy fails to restore thrombolysis in myocardial infarction (TIMI) grade 3 flow in up to 40% of patients at 90 minutes.¹⁰ Rescue angioplasty (rPCI) has been used to restore coronary flow after failed fibrinolysis in AMI.¹¹ As all the randomized trials on rPCI have excluded patients with shock, the benefit of rPCI in patients with cardiogenic shock in the context of failed fibrinolysis is not known.^11–13 We sought to identify differences in the clinical and angiographic characteristics, success rate of the intended procedure, and 30-day and 1-year mortality between patients in cardiogenic shock undergoing rPCI or primary angioplasty (PPCI).

Methods

Patient population. All patients undergoing PCI for cardiogenic shock at our institution between March 1994 and March 2005 were included. We adopted a strategy of emergency angiography and attempted PCI for cardiogenic shock complicating AMI in 1994. This service was offered to all surrounding district general hospitals serving our unit. The aim was to offer emergency revascularization within 12 hours of the onset of symptoms, although revascularization was attempted in some patients later than this because of continuing chest pain. The service was offered around the clock. A clinical diagnosis of cardiogenic shock was made if the patients fulfilled the following criteria (as defined by the British Cardiovascular Intervention Society): blood pressure < 100 mmHg; pulse > 100 beats/minute; and the patient was cool, clammy or requiring inotropes, intra-aortic balloon pump (IABP), or cardiopulmonary support to assist circulation. There was no requirement for measurement of cardiac filling pressures or cardiac output by invasive monitoring for the diagnosis to be made. Measurement of urine output was not mandatory. After clinical evaluation, echocardiography and occasionally left ventriculography were used to identify patients with cardiogenic shock caused by papillary muscle rupture, interventricular septal defect, or cardiac rupture before PCI, and these patients were not included in the analysis.

Interventional procedures. Coronary angiography was performed without routine left ventriculography. When Glycoprotein (GP) IIb/IIIa inhibitor use was likely, weight-adjusted heparin was used to achieve an activated clotting time (ACT) of 200–250 seconds (not greater than 250 seconds): for others, an ACT of 300 seconds was the target. The use of an IABP in the catheterization laboratory was left to the discretion of the attending cardiologist, with insertion of the device before, during or after PCI. The strategy involved percutaneous intervention to the culprit vessel if this could be identified, with intervention on other critical coronary stenoses only attempted in the absence of early hemodynamic improvement. The aim was to keep the procedure as short as possible and minimize the contrast load.

The use of adjunctive pharmacological treatment and other mechanical devices in the catheterization laboratory was left to the discretion of the attending cardiologist. Early unit guidelines were in place regarding the use of GP IIb/IIIa inhibitors following previous favorable experience with the selective use of these agents. The initial guidelines favored the bailout use of these agents rather than planned use, especially in the context of cardiogenic shock after full-dose thrombolytic, given the concern about bleeding risks. The indications for administration were: (1) the presence of intracoronary filling defects or haze after PCI, suggestive of thrombus; (2) slow flow after intracoronary stenting; (3) persistent dissection flaps; (4) suspected inlet or outlet dissection flaps after stenting; and (5) after long or multiple stenting, especially in the context of AMI. Subsequently, following additional trials of abciximab in primary PCI, abciximab has been used more liberally in patients, particularly those who have not been treated with full-dose fibrinolytic.

Our stenting strategy changed over time, according to availability of trial evidence to support the use of stenting in patients with AMI. Initially, stents were used for suboptimal balloon angioplasty results, but more recently, elective stenting has been the preferred strategy. If one or more coronary stents were deployed, ticlopidine 250 mg twice daily after an initial immediate dose of 500 mg or clopidogrel 75 mg once daily after an initial dose of 300 mg was administered (most recently, this has been increased to 600 mg). Patients were subsequently transferred to the coronary care unit or the cardiac intensive care unit if they were being ventilated. The use of thrombectomy devices in both arms was equal and rare.

Definitions and outcomes. The principal outcome of interest for the present study was all-cause mortality at 30 days and 1 year. We also assessed success or failure of the PCI procedures. A failed PCI procedure was defined as death during the procedure or need for urgent emergency (immediate) coronary artery bypass graft surgery (CABG) or TIMI flow < 3 in the infarct-related vessel. Multivessel coronary artery disease was defined if a stenosis > 50% was present in more than 1 main coronary vessel. The coronary lesions were classified according to the ACC/AHA classification: A and B1 were regarded as simple lesions, whereas B2 and C were considered as complex. For an acutely occluded vessel, when angiographic attributes of the lesion could not be evaluated, the lesion was classified as B1.

Statistical analysis. Data are expressed as mean value ± standard deviation (SD) for continuous variables, and as frequency with percentages for categorical variables. Differences between means of continuous variables were tested by the unpaired Student’s t-test as appropriate. When variables were not normally distributed, the nonparametric Mann-Whitney U-statistic tests assessed intergroup differences. Frequencies were compared using the chi-square (or Fisher’s exact test analysis when the expected value of cells was < 5). Cumulative survival curves for mortality were constructed according to the Kaplan-Meier method, and differences between the curves were tested for significance by the log-rank statistic. Thirty-day survival was also analyzed using Cox proportional hazards regression. The variables such as baseline characteristics, clinical management and angiographic findings with a Cox regression univariate p-value < 0.05 were grouped and entered into intermediate models. The covariates with significant p-values from this model were then combined to determine a final model with covariate p-values < 0.05. Data were analyzed with SPSS for Windows, version 13 (SPSS Inc., Chicago, Illinois).

Results

Of 171 patients with cardiogenic shock complicating AMI undergoing coronary intervention, 65 underwent PPCI and 59 patients underwent rPCI. The baseline characteristics are shown in Table 1.

The mean ages of the patients were 61.9 ± 11.9 years in the PPCI group and 62.8 ± 9.7 years in the rPCI group (p = 0.66). The other baseline demographics were similar, except patients in the rPCI group were more likely to be interhospital transfers (64% vs. 43%; p = 0.02) and had a longer door-to-balloon time (median 298 [IQR 395 to 180] minutes vs. 131 [IQR 215 to 90] minutes in the PPCI group; p < 0.01). Previous full-dose fibrinolytic with reteplase, tPA and tenecteplase had been administered in 17% of the rPCI patients, the others being treated with streptokinase.

Angiographic findings. The angiographic findings and procedural data are listed in Tables 2 and 3. An independent operator estimated all TIMI flows. Before the intervention, an occluded infarct-related artery (IRA) (TIMI 0/1 flow) was observed in 94% of the patients in the PPCI group and 88% in the rPCI group, and IRA TIMI 3 flow was 2% in the PPCI group and 8% in the rPCI group (p = 0.08).

There was no significant difference in the extent of coronary artery disease and the infarct-related artery. Single and multivessel coronary artery disease was observed in 18% and 82% in the PPCI group and 15% and 85% in the rPCI group, respectively (p = 0.64). The infarct-related artery was the left main in 5% of patients, the left anterior descending in 45%, the circumflex in 15%, the right coronary artery in 31% and a bypass graft in 4% of the patients. The infarct-related artery was more often in the left anterior descending artery in the rPCI group (53% vs. 38%), but this difference was not statistically significant (p = 0.12). The most frequent site of occlusion of the infarct-related artery was proximally in 63% of the patients (62% in PPCI, 64% in rPCI).

The intervention resulted in TIMI 3 flow in the IRA in 56% in the rPCI group and 74% in the PPCI group (p = 0.04). Technically successful PCI, defined by a patent vessel with TIMI 3 flow < 50% residual stenosis, no emergency CABG from the catheterization laboratory and successful transfer to the coronary care or cardiac intensive care unit was seen in 69% of the PPCI group and 54% in the rPCI group (p = 0.09). A stent was implanted in 62% of patients (63% in the PPCI group, 61% in the rPCI group; p = 0.81). The patients receiving stents had a higher TIMI flow grade 3 (rPCI: stent vs. balloon angioplasty was 75% vs. 26%, PPCI: stent versus balloon angioplasty was 83% vs. 58%, respectively). Mechanical support with an intra-aortic balloon pump was used in 90% patients, with similar use in rPCI (92%) and PPCI (89%) (p = 0.67). Overall, glycoprotein IIb/IIIa usage was 32%, but lower in the rPCI group compared with the PPCI group (20% vs. 42%; p < 0.01). The need for emergency bypass surgery was not different between both PCI strategies (2% in rPCI vs. 3% in PPCI).

PCI outcome. The total one-year survival rate was 40%. There was a higher 30-day survival rate in the PPCI group (57%) compared with the rPCI group 29% (p < 0.01), and similarly with 1-year survival rates (51% vs. 29%, p = 0.01), respectively (Figure 1). The impact of TIMI flow after PCI on mortality is given in Table 3. The patients with a postprocedure TIMI flow < 3 had a lower survival rate (18% in PPCI and 15% in the rPCI). In the patients with final TIMI flow 3, the 1-year survival rate was higher in the PPCI group (63% vs. 39%; p = 0.04). Patients who had successful PPCI and rPCI procedures had a higher survival rate at 1 year (67% and 41%) than those who had unsuccessful PCI (15%). Stenting was significantly associated with TIMI 3 flow and had lower mortality compared to balloon angioplasty (rPCI: stent vs. balloon angioplasty 61% vs. 87%, PPCI: stent vs. balloon angioplasty 46% vs. 54%). One-year mortality in patients > 70 years of age with cardiogenic shock undergoing rPCI was 100% (n = 15), and 70% with PPCI (n = 14).

Temporal trends. There was a decrease in 1-year mortality from 64% for the period 1994–2002, to 50% for the period 2003–2005 (rPCI: 72% to 67%, PPCI: 54% to 42%, for the respective periods). The use of stents increased from 50% to 92% (rPCI: 51% to 100%; PPCI: 49% to 88%) and rate of final TIMI flow grade 3 increased from 60% to 78% (rPCI: 51% to 75%; PPCI: 71% to 79%), for the periods 1994–2002 and 2003–2005, respectively. The use of GP IIb/IIIa inhibitors increased from 23% to 53% over the period, with significant increased use in PPCI procedures (27% to 67%) compared to rPCI procedures (19% vs. 25%).

Correlates of 30-day mortality (Table 4). Independent correlates of death at 30 days based on multivariate analysis are as follows: rescue angioplasty (p = 0.03), anterior myocardial infarction (p = 0.04), multivessel disease (p = 0.07), and final post-PCI TIMI flow grade < 3 (p < 0.01) (Figure 2). There was a univariate association between GP IIb/IIIa inhibitor use and improved survival, but this did not remain significant after controlling for other factors. The use of stents was not found to be a predictor of improved survival.

Discussion

In this retrospective observational study, we found that in the setting of cardiogenic shock, rPCI was associated with a higher 30-day and 1-year mortality than PPCI. Rescue angioplasty was found to be an independent predictor of 30-day mortality for patients undergoing PCI in the setting of cardiogenic shock.

The patients with cardiogenic shock complicating failure of fibrinolysis are a higher-risk group than those undergoing primary angioplasty for cardiogenic shock or angioplasty for cardiogenic shock complicating a reinfarction.¹⁴ In some patients, progression to an adverse outcome is not altered by attempted emergency revascularization. In fact, an adverse outcome may be determined well before emergency revascularization can even be contemplated. Patients undergoing rPCI may be sicker than those undergoing PPCI in this context because of the inherent delay in getting the patient into the cath lab, especially if the patient needs transportation from a district general hospital to the PCI center. In our cohort, significantly more patients in the rPCI group were transferred than in the PPCI group (64% vs. 43%; p = 0.02). It has been postulated that failed lysis is associated with high thrombus and plaque burden, which may result in a high rate of the no-reflow phenomenon, which itself can affect the outcome of the PCI procedure.

In our series, patients in shock undergoing rPCI had a particularly high 1-year mortality rate compared with those undergoing PPCI (71% vs. 49%; p < 0.01). Administration of a fibrinolytic before PCI did not affect survival in patients undergoing PCI for cardiogenic shock in the SHOCK trial registry, although the registry data as a whole do suggest that patients with cardiogenic shock should be considered for fibrinolytic treatment if there is no option for revascularization. However, it is unclear how many patients in the SHOCK registry or trial had cardiogenic shock accompanying a diagnosis of failed fibrinolysis. In the SHOCK trial, 56.3% of all randomly assigned patients, 49.3% of those in the revascularization arm and 49% of those receiving PCI had received a thrombolytic. Only 21% of those receiving PCI had TIMI 3 flow in the culprit vessel, so some patients did have cardiogenic shock complicating failed fibrinolysis.⁴

The mortality in our cohort was strongly related to the procedural success of PCI. There was an increased survival rate in both groups if PCI was successful, but an unsuccessful PCI was associated with a lower survival rate. A high mortality rate observed in patients with PCI failure (TIMI < 3 flow after the intervention) underscores the importance of TIMI 3 flow in this setting.¹⁵ In our cohort of patients, TIMI 3 flow rates in the infarct-related artery was 74% in the PPCI group and 56% in the rPCI group. This is certainly lower than the 90% TIMI 3 flow rates reported from PCI in patients with AMI without cardiogenic shock.¹⁶ In the SHOCK trial the success rate of PCI, defined as TIMI 3 or 2 flow and a residual stenosis < 50%, was 77% in the early invasive group. In comparison, our study had a high percentage of patients in the PPCI and rPCI groups with postprocedure TIMI flow grade 2 or 3 (88% in the PPCI group and 66% in the rPCI group, respectively). The increase in the time interval between symptom onset and start of PCI was associated with an increase in mortality, advocating a rapid invasive strategy in shock patients.

Even with the early interventional approach, mortality is high for patients in cardiogenic shock who are > 75 years of age. In the ALKK (Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte) registry, 63% of the patients > 75 years of age died.¹⁷ The number of patients > 75 years of age undergoing PPCI or rPCI in our cohort is small (n = 12), creating difficulties in interpretation of the results and in making recommendations for optimal treatment of these patients. In our cohort, 1-year mortality in patients > 75 years of age with cardiogenic shock undergoing rPCI was 100%, compared with 85% in the PPCI group. The decision to perform intervention in these elderly patients should be weighted on the biological age and the comorbidity of the individual patient.

Our stenting rates have increased from 32% to 89% over the years. The use of stents has been restricted in some patients, mainly because of failure to cross the culprit lesion, calcified vessels, small vessels and other difficult anatomy and, especially in our earlier experience, stents were not routinely used when results with balloon angioplasty were good. The overall use of stents in the rPCI and PPCI groups was 61% and 63%, respectively. The patients receiving stents had a higher final TIMI flow grade 3 and had lower mortality. This has been shown in the GRACE study, which showed a fourfold relative increase in hospital survival with stenting.¹⁸ Whether this reflects a specific advantage of stenting or simply defines a lower-risk cohort is unclear. The patients in cardiogenic shock undergoing rPCI without stents had higher mortality.¹⁹

GP IIb/IIIa inhibitor use has been shown to be associated with improved TIMI 3 flow and myocardial reperfusion rates in patients in cardiogenic shock undergoing PCI.^20–22 In our cohort, the GP IIb/IIIa inhibitor use in the rPCI group was lower (20%) compared with the PPCI group (42%). This infrequent use in the rPCI group is related to the prior administration of streptokinase as opposed to fibrin-specific agents. Bleeding risks in the setting of full-dose streptokinase are not insignificant.²³

Rescue angioplasty, anterior MI, multivessel disease and postprocedure TIMI flow grade < 3 in the context of cardiogenic shock were found to be independent predictors of 30-day mortality. This has implications for the transfer of acutely ill patients from a district general hospital to a regional cardiothoracic unit. Considering both logistic burden and the costs of emergency transfer for rescue angioplasty in patients with cardiogenic shock after fibrinolysis, proof of the benefit of rPCI from randomized trials is needed, but it may be extremely difficult to perform a randomized trial. It may be more appropriate to reduce the incidence of cardiogenic shock by earlier reperfusion strategies which will almost certainly reduce morbidity and mortality once shock occurs. This is probably best achieved by adopting a strategy of PPCI for all patients presenting with ST-elevation AMI. If this cannot be achieved for logistical reasons, then either a strategy of PPCI for high-risk patients (e.g., anterior infarction, the elderly and those with prior infarction), or early transfer of high-risk patients undergoing fibrinolysis for rPCI where applicable, should be considered.

Study limitations. This study may suffer from bias related to its retrospective analysis design. This is a single-center study and interventions were carried out by a single group of interventional cardiologists. We have been unable to collect the data on the onset of cardiogenic shock to intervention times, as these data were often not available from the referring hospital case notes.

We did not perform a full hemodynamic study, including measurement of filling pressures or calculation of cardiac output, before performing angiography and attempted revascularization. However, we are satisfied on clinical grounds that cardiogenic shock was present in each instance. Although patients were examined thoroughly prior to examination and selective echocardiography, and ventriculography was performed prior to a decision concerning angioplasty, we cannot be sure that some patients did not have acute mitral regurgitation, but all patients underwent echocardiography subsequently and no patient had mitral regurgitation as the dominant problem.

Another potential limitation could be the heterogeneity of the data collected, which reflects the continuous evolution of the PTCA techniques of revascularization over the 10-year period of the study.

Conclusions

In the setting of cardiogenic shock, rPCI patients had a lower final TIMI 3 flow and higher 1-year mortality than those undergoing PPCI. Even patients with a successful rPCI procedure had higher 1-year mortality than those with successful PPCI. Mortality was high for both groups if the PCI was unsuccessful. Rescue angioplasty was found to be an independent predictor of mortality at 30 days along with presentation with anterior MI, the presence of multivessel disease and postprocedure TIMI flow grade < 3. As cardiogenic shock remains a problem associated with adverse outcome, attempts to reduce its incidence are to be encouraged. In all patients presenting with AMI, early diagnosis, prompt initiation of reperfusion therapy, and early liaison with a revascularization unit at the first signs of hemodynamic instability are essential in the management of patients with cardiogenic shock. Rescue angioplasty in elderly patients (> 75 years of age) in cardiogenic shock may be a futile treatment.

References

1. Goldberg RJ, Samad NA, Yarzebski J, et al. Temporal trends in cardiogenic shock complicating acute myocardial infarction. N Engl J Med 1999;340:1162–1168.
2. Carnendran L, Abboud R, Sleeper LA, et al. Trends in cardiogenic shock: Report from the SHOCK Study. The SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK? Eur Heart J 2001;22:444–445.
3. Hasdai D, Topol EJ, Califf RM, et al. Cardiogenic shock complicating acute coronary syndromes. Lancet 2000;26;356:749–756.
4. Sanborn TA, Sleeper LA, Bates ER, et al. Impact of thrombolysis, intra-aortic balloon pump counterpulsation, and their combination in cardiogenic shock complicating acute myocardial infarction: A report from the SHOCK Trial Registry. SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK? J Am Coll Cardiol 2000;36:1123–1129.
5. Hasdai D, Holmes DR Jr, Topol EJ, et al. Frequency and clinical outcome of cardiogenic shock during acute myocardial infarction among patients receiving reteplase or alteplase. Results from GUSTO-III. Global Use of Strategies to Open Occluded Coronary Arteries. Eur Heart J 1999;20:128–135.
6. Moosvi AR, Khaja F, Villanueva L, et al. Early revascularization improves survival in cardiogenic shock complicating acute myocardial infarction. J Am Coll Cardiol 1992;19:907–914.
7. Hochman JS, Sleeper LA, Webb JG, et al. Early revascularization in acute myocardial infarction complicated by cardiogenic shock. SHOCK Investigators. SHould we emergently revascularize Occluded coronaries for cardiogenic shocK. N Engl J Med 1999;341:625–634.
8. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1999 Guidelines for the Management of patients with acute myocardial infarction). J Am Coll Cardiol 2004;44:E1–E211.
9. Webb JG, Lowe AM, Sanborn TA, et al. Percutaneous coronary intervention for cardiogenic shock in the SHOCK trial. J Am Coll Cardiol 2003;42:1380–1386.
10. The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both on coronary-artery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993;329:1615–1622.
11. Gershlick AH, Stephens-Lloyd A, Hughes S, et al, REACT Trial Investigators. Rescue angioplasty after failed thrombolytic therapy for acute myocardial infarction. N Engl J Med 2005;353:2758–2768.
12. Ellis SG, Da Silva ER, Heyndrickx G, et al. Randomized comparison of rescue angioplasty with conservative management of patients with early failure of thrombolysis for acute anterior myocardial infarction. Circulation 1994;90:2280–2284.
13. Sutton AG, Campbell PG, Graham R, et al. A randomized trial of rescue angioplasty versus a conservative approach for failed fibrinolysis in ST-segment elevation myocardial infarction: The Middlesbrough Early Revascularization to Limit INfarction (MERLIN) trial. J Am Coll Cardiol 2004;44:287–296.
14. Ross AM, Lundergan CF, Rohrbeck SC, et al. Rescue angioplasty after failed thrombolysis: Technical and clinical outcomes in a large thrombolysis trial. GUSTO-1 Angiographic Investigators. Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries. J Am Coll Cardiol 1998;31:1511–1517.
15. Hamada S, Nishiue T, Nakamura S, et al. TIMI frame count immediately after primary coronary angioplasty as a predictor of functional recovery in patients with TIMI 3 reperfused acute myocardial infarction. J Am Coll Cardiol 2001;38:666–671.
16. Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: A quantitative review of 23 randomised trials. Lancet 2003;361:13–20.
17. Zeymer U, Vogt A, Zahn R, et al, The Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte (ALKK). Predictors of in-hospital mortality in 1333 patients with acute myocardial infarction complicated by cardiogenic shock treated with primary percutaneous coronary intervention (PCI): Results of the primary PCI registry of the Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte (ALKK). Eur Heart J 2004;25:322–328.
18. Dauerman HL, Goldberg RJ, White K, et al. Revascularization, stenting, and outcomes of patients with acute myocardial infarction complicated by cardiogenic shock. Am J Cardiol 2002;90:838–842.
19. Antoniucci D, Valenti R, Migliorini A, et al. Abciximab therapy improves survival in patients with acute myocardial infarction complicated by early cardiogenic shock undergoing coronary artery stent implantation. Am J Cardiol 2002;90:353–357.
20. Chan AW, Chew DP, Bhatt DL, et al. Long-term mortality benefit with the combination of stents and abciximab for cardiogenic shock complicating acute myocardial infarction. Am J Cardiol 2002;89:132–136.
21. Huang R, Sacks J, Thai H, et al. Impact of stents and abciximab on survival from cardiogenic shock treated with percutaneous coronary intervention. Catheter Cardiovasc Interv 2005;65:25–33.
22. Zeymer U, Tebbe U, Weber M, et al, ALKK Study Group. Prospective evaluation of early abciximab and primary percutaneous intervention for patients with ST elevation myocardial infarction complicated by cardiogenic shock: Results of the REO-SHOCK trial. J Invasive Cardiol 2003;15:385–389.
23. Antman EM, Giugliano RP, Gibson CM, et al. Abciximab facilitates the rate and extent of thrombolysis: Results of the thrombolysis in myocardial infarction (TIMI) 14 trial. The TIMI 14 Investigators. Circulation 1999;99:2720–2732.