Comparison of Immediate Invasive, Delayed. Tissue-Type Plasminogen Activator. Invasive, and Conservative Strategies After. Clinical Investigation

Transcription

1 Clinical Investigation 1457 Comparison of Immediate Invasive, Delayed Invasive, and Conservative Strategies After Tissue-Type Plasminogen Activator Results of the Thrombolysis in Myocardial Infarction (TIMI) Phase TI-A Trial* William J. Rogers, MD, Donald S. Baim, MD, Joel M. Gore, MD, B. Greg Brown, MD, Robert Roberts, MD, David 0. Williams, MD, James H. Chesebro, MD, Joseph D. Babb, MD, Florence H. Sheehan, MD, Frans J.Th. Wackers, MD, Barry L. Zaret, MD, Thomas L. Robertson, MD, Eugene R. Passamani, MD, Richard Ross, MSc, Genell L. Knatterud, PhD, and Eugene Braunwald, MD, for the TIMI II-A Investigators* To assess the value and timing of percutaneous transluminal coronary angioplasty (PTCA) after thrombolytic therapy for acute myocardial infarction (AMI), 586 patients in the Thrombolysis in Myocardial Infarction Study Phase II-A were randomized among three treatment strategies, one using immediate coronary arteriography followed by PTCA if appropriate (immediate invasive strategy group, n=195), a second that deferred angiography and PTCA for hours (delayed invasive strategy group, n=194), and a third, more conservative, approach in which PTCA was used only if ischemia occurred spontaneously or at the time of predischarge exercise testing (conservative strategy group, n=197). Predischarge contrast left ventricular ejection fraction, the primary study end point, was similar among the patients in all three treatment groups and averaged 49.3%. The finding of a patent infarctrelated artery at the time of predischarge arteriography was equally common among the patients in the three groups (mean, 83.7%); however, the mean residual infarct artery stenosis was greater in the patients in the conservative strategy group (67.2%) as compared with the patients in the immediate invasive (50.6%) and the delayed invasive strategy groups (47.8%) (p<o.ool). Immediate invasive strategy led to a higher rate of coronary artery bypass graft surgery (CABG) after PTCA (7.7%) than did delayed invasive and conservative strategies (2.1% and 2.5%, respectively; p<0.01). Furthermore, among patients not undergoing CABG during the first 21 days, blood transfusion of more than 1 unit was used in 13.8% of the patients in the immediate invasive strategy group, 3.1% of the patients in the delayed invasive strategy group, and 2.0% of the patients in the conservative strategy group (p<0.001). At 1-year follow-up, the three treatment groups had similar cumulative rates of mortality (8.7%, pooled over all groups), fatal and nonfatal reinfarction (8.5%), combined death and reinfarction (14.5%), and CABG (17.2%), although the cumulative performance rate of PTCA remained higher in the invasive groups (immediate invasive strategy group, 75.8%; delayed invasive strategy group, 64.3%; and conservative strategy group, 23.9%; p<0.001). Thus, because conservative strategy achieves equally good short- and long-term outcome with less morbidity and a lower use of PTCA, it seems to be the preferred initial management strategy. (Circulation 1990;81: ) *A listing of principal and coinvestigators is given in the Supported by research contracts from the National Heart, Lung, "Appendix." A statement regarding financial interests of TIMI and Blood Institute, Bethesda, Maryland. principal investigators has been previously published (N Engi J Address for correspondence: William J. Rogers, MD, 334 Lyons Med 1989;320:618-footnote). Harrison Research Building, University of Alabama Medical Cen- From the Thrombolysis in Myocardial Infarction (TIMI) Coor- ter, Birmingham, AL dinating Center, Baltimore, Maryland. Received October 5, 1989; revision accepted January 10, 1990.

2 1458 Circulation Vol 81, No 5, May 1990 H igh-grade residual stenosis is common immediately after thrombolysis for acute myocardial infarction (AMI) and predicts subsequent reocclusion of the infarct-related artery.1 Based on earlier studies demonstrating that percutaneous transluminal coronary angioplasty (PTCA) is possible soon after starting thrombolytic therapy,2-6 it seemed reasonable to hypothesize that PTCA performed very early after symptoms of infarction might lower the probability of reocclusion, augment myocardial reperfusion, and aid in the salvage of ischemic myocardium. It was unclear, however, whether PTCA performed in the AMI setting (with altered clotting function and variable states of hemodynamic and arrhythmic instability) would be safer and more likely to preserve life and myocardium than a more conservative approach that reserved myocardial revascularization for only those patients who manifested recurrent ischemia. Accordingly, the National Heart, Lung, and Blood Institute Thrombolysis in Myocardial Infarction (TIMI) II-A trial was undertaken to compare two invasive strategies, one using immediate PTCA when feasible (immediate invasive strategy) and the other See p 1707 using PTCA delayed by hours (delayed invasive strategy), with a more conservative strategy of "watchful waiting" (conservative strategy) in patients with AMI treated with recombinant tissuetype plasminogen activator (rt-pa). A preliminary report of the results through 21 days of study entry for patients randomized to the immediate invasive and delayed invasive strategy groups has been published previously.7 The purpose of the present study was to provide a comprehensive description of the TIMI II-A findings, including the following topics not addressed in the preliminary report: 1) outcome of the patients allocated to the conservative strategy group, 2) quantitative ventriculographic and coronary arteriographic analyses, 3) resting and exercise radionuclide ventriculographic findings at the time of hospital discharge and at 6 weeks' follow-up, and 4) survival, interim events, and additional procedures performed during 1 year after entry. Additionally, whereas individual clinic assessments of infarct-vessel patency and PTCA success were presented in the preliminary report, core radiographic laboratory assessment of these parameters is provided in the present study. Methods Study Design A detailed description of the design of TIMI 1I-A has been previously published.7 The principal objective of TIMI IT-A was to establish the feasibility and optimal timing of PTCA in patients receiving rt-pa for AMI. The primary study end point was contrast left ventriculographic ejection fraction at the time of hospital discharge, although a number of prospectively defined secondary end points were also assessed, including the value and success of protocol PTCA, predischarge infarct-vessel patency, exercise test performance, need for nonprotocol procedures (including coronary arteriography and coronary revascularization), and frequency of complications (including death, nonfatal reinfarction, and need for transfusion). Patient Population Between April 11, 1986, and June 30, 1988, 3,534 patients were entered into TIMI II, of whom 586 were enrolled in the TIMI II-A substudy at seven participating clinical sites ("Appendix") (Figure 1). Investigators at these sites had special interest and documented proficiency in PTCA for AMI. Eligible for TIMI II were patients less than 76 years old with ischemic chest pain of at least 30 minutes' duration and ST elevation of at least 1.0 mv in at least two contiguous electrocardiographic leads who gave informed consent, who could have rt-pa administered within 4 hours of onset of AMI symptoms, and in whom none of the previously described exclusion criteria were present.7 Administration of rt-pa, Heparin, and Aspirin The rt-pa used in TIMI II was supplied by Genentech Inc. and prepared by the suspension culture technique (G11044). Study entry commenced when rt-pa infusion was initiated. In the first 195 patients, rt-pa was given as a dosage of 150 mg/6 hr; subsequently, because of an unacceptable frequency of hemorrhagic complications, the dosage was reduced to 100 mg/6 hr (391 patients),8,9 administered as an intravenous bolus of 6 mg and a constant infusion of 54 mg during the first hour, 20 mg during the second hour, and 5 mg during each of the next 4 hours. Before rt-pa administration, patients were prospectively classified according to myocardial infarction (MI) 4"risk status" as defined previously.7 Immediately after beginning rt-pa, a bolus of 5,000 USP units heparin was injected intravenously, followed by a 5-day constant infusion, initially at 1,000 units/hr but adjusted to maintain the activated partial thromboplastin time fold the upper limit of normal. On day 6, 10,000 units s.c. heparin every 12 hours was begun and continued until the predischarge exercise test. In the first 127 patients, aspirin was given at a dose of 81 mg daily for 5 days beginning on the day of rt-pa administration and at a dose of 325 mg daily thereafter; subsequently, aspirin was initiated on day 2 rather than day 1. Initial Treatment Assignments Unique randomization treatment allocation schedules were prepared by the TIMI Coordinating Center for each clinical center to ensure assignment of equal numbers of patients to the three treatment groups at each site. Assignment to the immediate invasive strategy, delayed invasive strategy, or conservative strategy group (Figure 2) was accomplished at each clinic using a security-protected

3 Rogers et al TIMI II-A Trial A Substudy (586) Immediate Invasive Stratgy M.~.. Delayed Inveasve Strategy (1681) 77 /3 _ ' Conservative Strategy (1658) Beta Blocker Substudy (1434) FIGURE 1. Diagram showing interrelation of TIMI II studies. TIMI II trial included 3,534 patients. There were three studies as follows: 1) TIMI II main study (large square, n=3,339) in which patients were randomized between delayed invasive strategy (n=1,681) and conservative strategy groups (n=1,658); 2) TIMIII-A substudy (hatched rectangle, n=586), overlapping with main study but including immediate invasive strategy treatment group; and 3) (-blocker substudy (n=1,434), a component of main study in which patients eligible for immediate intravenous 3-blocker were randomized (in a 2x2 factorial design) between immediate and delayed 3-blocker therapy as well as between delayed invasive strategy and conservative strategy groups. microcomputer system or, earlier in the study, sealed treatment allocation envelopes..management Strategies Coronary arteriography was to be performed within 120 minutes of rt-pa initiation in patients allocated to the immediate invasive strategy group and within hours of rt-pa initiation in patients allocated to the delayed invasive strategy group. PTCA of the infarct-related artery was to be attempted immediately after angiography if the coronary anatomy was suitable.7 In the immediate invasive strategy group, PTCA was performed in patients even if the infarctrelated artery was closed ("rescue angioplasty"), whereas in the delayed invasive strategy group, PTCA was limited to patients with open infarct-related arteries unless there was clinical evidence of continuing myocardial ischemia. Coronary anatomy unsuitable for PTCA of the infarct-related artery included one or more of the following: 1) residual stenosis of less than 60%, 2) lesion length greater than or equal to 20 mm, 3) lesion located at bifurcation, distal position, or beyond a tortuous proximal vessel, or 4) operator AMI Within 4 Hrs Onset IV I rt-pa Immediate Delayed Conser- Inasive nioasive valvel N = 195 H = 194 H = 197 Pre-Discharge Rest/Exercise RVG Coronary & LV Angios 6 Week Clinic Follow-up & Rest/Exercise RVG I 1 Yr Clinic Follow-up & Treadmill Exercise Test FIGURE 2. Diagram of TIMI II-A study design. AMI, acute myocardial infarction; IV; intravenous; rt-pa, recombinant tissue-type plasminogen activator; RVG, radionuclide ventriculogram; LV, left ventricular; Angios, angiograms. judgment that abrupt closure of involved vessel could cause catastrophic hemodynamic consequences. Standardized projections of the coronary arteries were recorded at baseline and again after PTCA. Full improvement after PTCA was defined as normal or improved flow in the infarct-related artery and both an absolute reduction in lumen stenosis by at least 20% and a final lumen stenosis of less than 60%. Partial improvement was defined as unchanged or improved flow with only one of the above changes in lumen stenosis. Patients randomized to the conservative strategy group underwent routine coronary care unit management.10 Coronary arteriography in this group was discouraged until immediately before hospital discharge (days 8-10 after entry) unless there was the recurrence of ischemia refractory to progressive stepwise medical therapy with nitrates, [3-blockers, and calcium antagonists. In all groups, coronary artery bypass graft surgery (CABG) was limited to patients with pressing clinical indications and coronary anatomy more suitable for CABG than PTCA (i.e., left main coronary artery disease or advanced multivessel disease in which closure of the infarct-related artery would cause severe hemodynamic compromise). Concomitant Care and Interim Events Except for PTCA assignment, patients in each of the three treatment groups were managed identically. On the initial presentation, all patients received sublingual nitroglycerin, prophylactic lidocaine, and morphine. Unless there were contraindications, nifedipine mg three times daily was administered for 96 hours. Unless contraindicated, metoprolol was administered after the predischarge contrast and radionuclide studies herein described, at 50 mg twice daily for 1 day, and then 100 mg twice daily for the next year. Guidelines for management of complications of myocardial infarction were specified within the TIMI Manual of Operations to ensure uniformity of care among the participating clinical units. Careful record was made of recurrence of chest pain, electrocardiographic changes, or cardiac enzyme rise. Using predetermined definitions,7 these events were classified as definite MI, recurrent ischemia, or "no

4 1460 Circulation Vol 81, No 5, May 1990 event" by an independent mortality and morbidity classification committee with no knowledge of treatment assignment. Predischarge Studies and Management To simultaneously assess left ventricular function and clinical response to submaximal exercise, patients in each of the three treatment groups underwent resting and bicycle-exercise radionuclide ventriculography on days 8-10 after entry. Equilibrium radionuclide ventriculography was performed at rest after in vivo red blood cell labeling with technetium 99m pertechnetate"l and then during 3-minute stages of 200 and 400 kilopond meters (kpm)/min workload, respectively. Predischarge exercise testing was restricted to two stages (400 kpm/min) or until one of the following end points was reached: 1) heart rate of 120 beats/min, 2) complex ventricular arrhythmias, 3) onset of angina pectoris, 4) decrease in blood pressure of more than 20 mm Hg, 5) exhaustion or marked dyspnea, or 6) ST segment depression of at least 2 mm (0.08 seconds after J point) as compared with the resting electrocardiogram (ECG). Radionuclide data from each clinical center were stored on magnetic tape or 8-in. floppy disks, forwarded to the TIMI Radionuclide Core Laboratory, and analyzed without knowledge of treatment assignment. After the radionuclide studies, coronary arteriograms in standardized projections and contrast left ventriculograms in the right anterior oblique projection were obtained. A grid of known dimensions was filmed at the level of the left ventricle for calibration. The contrast studies were analyzed by the Radiographic Core Laboratory without knowledge of treatment assignment and using previously described quantitative methodology After the discharge radionuclide and contrast studies, medical therapy was advised unless there was evidence of left main coronary artery stenosis of at least 70% or the presence of resting or exerciseinduced ischemia, in which case revascularization with PTCA or CABG was permitted if appropriate. Follow-up Studies at Six Weeks and One Year After Entry At 6 weeks (window, 6-12 weeks after entry), resting and exercise radionuclide ventriculography was repeated. Unlike the predischarge study, exercise was continued until a limiting symptom appeared or age-predicted (220 minus age in years) maximal heart rate was achieved. At 1 year after entry (window, months), a maximum symptom-limited Bruce treadmill exercise test15 was performed. At both follow-up visits, a detailed interview was conducted regarding interim events and procedures. Statistical Methodology It was determined during the planning phase that approximately 200 patients in each of the three treatment groups would be required to detect a three-unit difference in global ejection fraction (EF) as measured by contrast ventriculography with 85% power. Ap value of 0.05 was used for the assessment of this primary end point. The power to detect differences in subgroups comprising 50 or more patients was 85% for differences of 7.2 units, and the p value for this assessment was Except for the primary end point, p values in the range of were considered borderline, andp values of less than 0.01 were considered statistically significant. The statistical testing of differences in mean EF among the three treatment strategy groups was performed by using a one-way analysis of variance in the SAS statistical package program.16 Analysis of variance procedures were also used to test comparisons among the groups in the means of other continuous variables. Comparisons of proportions of clinical events among the three treatment groups was made with two-by-three contingency analyses, each yielding a x2 statistic with two degrees of freedom.16 Life-table rates were calculated by the product limit methods of the SAS LIFETEST software program. All reported p values were two sided and not adjusted for multiple testing. In all analyses in this study, patients were included in the group to which they were randomized regardless of whether coronary angiography and PTCA were actually performed within the time specified for that treatment group. The results presented in this study are based on all data processed in the TIMI Coordinating Center as of August 7, Results Baseline Characteristics Of the 586 patients enrolled in TIMI II-A, 195 were allocated to the immediate invasive strategy group, 194 were allocated to the delayed invasive strategy group, and 197 were allocated to the conservative strategy group (Table 1). The three randomized groups were similar with respect to age, race, and sex distribution. The majority (66.2%) of patients were classified as "not low risk," primarily because of the presence of anterior MI (52.0%), history of previous infarction (16.9%), or advanced age (13.0%). Despite randomization, slightly more patients in the immediate invasive strategy group were classified as "not low risk" (72.8%) as compared with the delayed invasive strategy (61.9%) and conservative strategy (64.0%) groups (p=0.05); however, the prevalence of the individual major risk factors (age,.70 years; previous MI; and anterior MI) were not significantly different among the three treatment groups. Intravenous rt-pa was initiated at a mean of 2.9 hours after onset of pain, at which time chest pain was still present in 86.7% of the patients. Performance of Coronary Arteriography and PTCA According to protocol design, the performance of coronary arteriography and PTCA differed significantly among the three groups. In the first group (immediate invasive strategy), coronary arteriography was performed within 2 hours of rt-pa initiation

5 Rogers et al TIMI II-A Trial 1461 TABLE 1. Baseline Characteristics Characteristics Age (mean yr) Race (white) Sex (male) Not low risk Age.70 yr Previous MI Anterior MI Rales.1/3 lung fields Hypotension and sinus tachycardia Atrial fibrillation or flutter Other risk factors History of angina History of CHF History of hypertension History of diabetes mellitus Ongoing chest pain at rt-pa initiation Time from onset of pain to study entry (mean hr) MIL, myocardial infarction; Immediate invasive (%) Delayed invasive (%) (n = 195) (n = 194) CHF, congestive heart failure. in 193 patients; however, PTCA was actually attempted in only 141 patients (Figure 3). Protocol PTCA in the immediate invasive strategy group resulted in improvement (as assessed by the TIMI Core Radiographic Laboratory) in 76.6% of the patients in whom it was attempted (Figure 4), including 73.2% of patients (82 of 112) with patent but X of PatIents - During Protocol Time Wlndow - Immediate Invasive Delayed Invasive (N = 195) (N = 194) l Conservative (%) (n= 197) stenotic infarct-related arteries and 89.7% of patients (26 of 29) with still-occluded arteries. Major complications occurred within 24 hours of PTCA in 8.5% of patients (Figure 4). In the second group (delayed invasive strategy), coronary arteriography was performed between 18 and 48 hours of rt-pa initiation in 175 patients; -WIthin 21- Days of Entry PTCA Not Don. (445) PTCA Done (565) Conservative (N = 197) PTCA Done -s-(18) - p FIGURE 3. Bar graph of performance ofpercutaneous transluminal coronary angioplasty (PTCA). Reasons for not performing PTCA during protocol time window included the finding of infarct-vessel stenosis of less than 60% (No Lesion); complex coronary anatomy unsuitable for PTCA (Unsuitable); total occlusion of infarct vessel in patients hours after entry without clinical evidence of ongoing ischemia (Closed); no catheterization performed during protocol time window (No Cath); and other reasons (Other). Shown for comparison is frequency of PTCA in conservative strategy group (n= 197) within 21 days of entry. Cumulative incidence of initial PTCA within 21 days ofentry was 75% in immediate invasive strategy group (n=195) and 63% in delayed invasive strategy group (n=194).

6 1462 Circulation Vol 81, No 5, May 1990 Immediate Invasive Strategy (N = 195) I PTCA < 2 hr (N = 141) An glographic imiprovement 77% Mua or Compil- within caltion hr of PTCA 2.1% Type -0. Dth MI CADG (N) (3) (4) (5) Delayed invasitv Strategy (N = 194) I PTCA at hr (N = 108) % 3.7% 1.9% Death M] CAIG (1) (4) (2) however, PTCA was actually attempted during this time window in 108 patients (Figure 3). PTCA resulted in improvement in 89.8% of the patients in whom it was attempted in this group, and major complications within 24 hours of PTCA occurred in 4.6% of the patients in whom it was attempted (Figure 4). Of the 19 patients who did not undergo coronary arteriography during the hour window, five had died before invasive evaluation and 12 Conservative Strategy (N = 197) "Early PTCA" 'Lat PTCA" (N = 14) (N = 21) Death Mi CABG (0) (1) (2) Death M) CABG (1) (2) (2) FIGURE 4. Diagram showing outcome ofpatients havingpercutaneous transluminal coronary angioplasty (PTCA). Angiographic improvement was more likely with delayed invasive strategy group as compared with immediate invasive strategy group (90% vs. 77%o, p=0.01), and major complications were fewer (8.5% vs. 4.6%, p=ns). Among patients in conservative strategy group, only a small number of l'~~~~~~~~~~~~~~~ patients underwent either "Early PTCA " (before scheduled predischarge coronary arteriogram) or "Late PTCA" (after predischarge coronary arteriogram); however, a trend toward more angiographic improvement (p=0.02) andfewer complications (p=ns) was also evident in "Late PTCA " group as compared with "Early PTCA" group. Note that complications categories are not mutually exclusive. PTCA, percutaneous transluminal coronary angioplasty; MI, myocardial infarction; CABG, coronary artery bypass graft surgery. other patients had undergone urgent catheterization before 18 hours, which led to PTCA in 10 patients (one of whom also had subsequent CABG). In the third group (conservative strategy), the intent was to perform coronary arteriography only at the time of hospital discharge (days 8-10 after entry) but not to perform PTCA unless prompted by either spontaneous or exercise-induced ischemia. Because of severe or prolonged recurrent ischemic pain by 1, Performance of PTCA % Having PTCA Immediate Invasive (N = 195) Delayed Invasive (N = 194) _.._... Conservative (N = 197) X Having CABG iu 80 [ 60 F 40 F Immediate Invasive (N = 195) Dolayed Invasive (N = 194) Conerative (N = 197) * Days Following Study Entry Days Following Study Entry FIGURE 5. Plottings of performance of initial coronary revascularization procedures during first 42 days. Performance of percutaneous transluminal coronary angioplasty (PTCA) was significantly lower in conservative strategy group (left panel) although at the end of 42 days, performance of coronary artery bypass graft surgery (CABG) was similar among the three treatment groups (right panel). 0 -

7 3, and 5 days after entry, early coronary arteriography was performed in a cumulative 6.1%, 8.6%, and 12.2%, respectively, of the patients and was followed by PTCA in 4.1%, 5.6%, and 6.6% of these respective patients (Figure 5, left panel). Altogether, 17.8% of the patients in the conservative strategy group underwent coronary angiography before the scheduled hospital discharge catheterization, which led to PTCA in 7.1% of patients (14) and CABG in 3.0% of patients (six). Of the 14 patients in the conservative strategy group who developed recurrent ischemia prompting early angiography followed by PTCA, PTCA resulted in improvement in 57.1%, including four of six patients who had an occluded infarct-related artery at the time of arteriography. During the 24 hours after PTCA, however, two patients (14.3%) in this group developed major complications (Figure 4). By contrast, in the 21 patients in the conservative strategy group who underwent PTCA after predischarge coronary arteriography but within 21 days of study entry, PTCA resulted in improvement in 90.4% of patients, including one of two patients who were found to have an occluded infarct-related artery. During the next 24 hours, major complications occurred in two patients (9.5%) (Figure 4). Rescue Angioplasty In the immediate invasive strategy group, 29 patients underwent acute dilation of totally occluded vessels. Of the 29 patients, 93.1% had patent infarctrelated arteries after PTCA, and the 21-day mortality rate was 6.9%. Predischarge catheterization, however, in 24 of these patients showed that 20.8% of these "rescue PTCA" vessels had reoccluded. In contrast, a total of 67 patients (35 in the delayed invasive strategy group and 32 in the conservative strategy group) had occluded infarct-related arteries at the time of initial catheterization. Although PTCA was attempted in 21 of these 67 patients with recurrent ischemia, the great majority (n=41) of these patients were discharged without PTCA or CABG, and their 21-day mortality rate was 7.3%. Performance of CABG Bypass surgery was performed in 67 of 586 patients (11.4%) in the TIMI 11-A trial within 21 days of study entry, including 24 patients after attempted PTCA and 43 patients referred directly to CABG based on anatomic findings (Figure 5, right panel). Although the use of CABG for anatomy unsuitable for PTCA was comparable among the three groups (17, 12, and 14 patients, respectively), there was significantly greater use of CABG after attempted PTCA in the immediate invasive strategy group (7.7% of patients) than in either the delayed invasive strategy group (2.1% of patients) or the conservative strategy group (2.5% of patients) (p=0.01). The results of primary bypass surgery for PTCAunfavorable anatomy were quite good, with a 21-day mortality rate of 2.3% (one of 43 patients) and a Rogers et al TIMI II-A Trial 1463 perioperative infarction rate of 0.0%. On the other hand, the results of CABG after attempted PTCA were less satisfactory, with a 21-day mortality rate of 16.7% (four of 24 patients) and a perioperative infarction rate of 25.0% (six of 24 patients). Patency of Infarct-Related Artery The identity of the infarct-related artery, as determined from the first coronary arteriogram obtained on each patient, was similar among the three treatment groups (Table 2). Protocol coronary arteriograms obtained at a mean of 1.4 and 32.8 hours after rt-pa initiation in the patients in the immediate invasive strategy and delayed invasive strategy groups, respectively, demonstrated patency (TIMI grade 2 or 3) of the infarct-related artery in 81.4% of the patients in the immediate invasive strategy group and 83.4% of the patients in the delayed invasive strategy group. Coronary arteriograms were obtained before hospital discharge (mean, 9.3 days after entry) in 80.5% of the TIMI IL-A patients and showed patency of the infarct-related artery in 81.1% of the patients in the immediate invasive strategy group, 84.0% of the patients in the delayed invasive strategy group, and 85.7% of the patients in the conservative strategy group (p=ns). In the 126 patients without data regarding predischarge patency, reasons included death before angiography (38 patients), recent CABG (29 patients), patient refusal (25 patients), and miscellaneous other reasons (34 patients). In an attempt to compensate for absence of predischarge coronary arteriography in some patients, those patients who underwent CABG (n = 29), those patients who had recurrent MI (n=4), or those patients who died before predischarge arteriography (n=38) were assumed to have occluded vessels, whereas the remaining patients (n=51) were assumed to have discharge patency unchanged from their earlier study (44 open and seven closed). With these imputations, patency could be estimated in 582 of the study patients (99.3%), with imputed predischarge patency rates of 71.8%, 72.7%, and 76.7% in the three treatment groups, respectively (p=ns). Coronary arteriograms suitable for quantitative analysis of the infarct-related artery were available in 99.7% of the patients in the PTCA groups who underwent early protocol coronary arteriography, in all of the patients who underwent protocol PTCA, and in 97.5% of patients who underwent protocol predischarge coronary arteriography (Table 2). On the initial arteriogram, the severity of the infarctartery stenosis was significantly greater in the patients in the immediate invasive strategy group than in the patients in the delayed invasive strategy group (79.9% vs. 75.2%,p=0.01), and the proportion of infarct arteries having stenosis of less than 60% was significantly lower (6.2% vs. 14.4%, p=0.01). Immediately after protocol PTCA, infarct-artery stenosis was reduced by an absolute 35.2% and 36.5%, respectively, in the two groups (p=0.65).

8 1464 Circulation Vol 81, No 5, May 1990 TABLE 2. Coronary Arteriographic Data Immediate invasive Delayed invasive Conservative (n= 195) (n =194) (n = 197) Arteriographic data n % n % n % p Protocol arteriography performed Early (2 hr or hr) (n) Predischarge (n) Time from onset of rt-pa to arteriography Early (hr) Predischarge (days) Infarct-related artery* t... LMCA LAD LCx RCA Normal Infarct artery found patent (grade 2 or 3) (%) Early protocol arteriogram Predischarge arteriogram Observed Imputed* Quantitative analysis of infarct-related artery Stenosis, mean All patients Early protocol arteriogram Predischarge arteriogram <0.001 Patients having protocol PTCA (n) Pre-PTCA <0.001 Post-PTCA <0.001 Predischarge arteriogram Patients not having protocol PTCA Early protocol arteriogram Predischarge arteriogram Stenosis <60% All patients Early protocol arteriogram Predischarge arteriogram <0.001 Patients having protocol PTCA Pre-PTCA Post-PTCA Predischarge arteriogram Patients not having protocol PTCA Early protocol arteriogram Predischarge arteriogram LMCA, left main coronary artery; LAD, left anterior descending coronary artery; LCx, left circumflex artery; RCA, right coronary artery; PTCA, percutaneous transluminal coronary angioplasty; MI, myocardial infarction; CABG, coronary artery bypass graft. *Infarct artery was determined from first catheterization patient received. tlncludes all catheterizations performed within 21 days of rt-pa treatment. *Closed includes TIMI grade 0 or 1 at hospital discharge catheterization or death, MI, CABG, or closed after early PTCA, or closed after protocol catheterization. Open includes TIMI grade 2 or 3 at hospital discharge catheterization or open after early PTCA or open after protocol catheterization. Conservative strategy >delayed invasive, immediate invasive. At the time of hospital discharge, highly significant group. Specifically, the mean stenosis was less (50.6% differences in the severity of the residual infarct- and 47.8% vs. 67.2%,p<0.001) and the percentage of vessel stenosis were observed between the patients in patients with a residual stenosis of less than 60% was the two PTCA groups and the conservative strategy greater (67.6% and 75.0% vs. 31.0%, p<0.001) for

9 Rogers et al TIMI II-A Trial 1465 TABLE 3. Predischarge Contrast Left Ventriculogram Data Immediate invasive Delayed invasive Conservative (n= 195) (n =194) (n =197) Left ventriculogram data n Mean n Mean n Mean p Ejection fraction (mean) Observed All patients Low-risk patients Not low-risk patients No previous MI Previous MI Anterior MI Not anterior MI Imputed* All patients Low-risk patients Not low-risk patients No previous MI Previous MI Anterior MI Not anterior MI End-diastolic volume index (mi/m2) End-systolic volume index (mu/m2) Stroke volume index (ml/m2) Cardiac index (I/min/m2) MI, myocardial infarction. *Imputations: 1) ejection fraction of zero assigned to patients who died before predischarge contrast ejection fraction could be determined; 2) contrast ejection fraction estimated from radionuclide ejection fraction, if available, as described previously.7 Total number of imputations made: Immediate invasive group, 35 patients, that is, 14 deaths and 21 estimated from radionuclide study; Delayed invasive group, 30 patients, that is, 10 deaths and 20 estimated from radionuclide study; Conservative group, 25 patients, that is, 13 deaths and 12 estimated from radionuclide study. the immediate invasive strategy and delayed invasive strategy groups versus the conservative strategy group, respectively. At the time of predischarge catheterization, total occlusion (TIMI grade 0 or 1) of the infarct vessels was found in 18.9% of the patients (28 of 148) in the immediate invasive strategy group (including 33.3% of patients [10 of 30] with initially occluded arteries and 14.9% of patients [18 of 121] with initially patent vessels). Total occlusions were found in 16.0% of patients (23 of 144) in the delayed invasive strategy group (including 12.5% of patients [one of eight] with initially occluded vessels who underwent PTCA, 100% of patients [19 of 19] with initially occluded vessels who did not undergo PTCA, and 3.2% of patients [three of 93] with initially patent infarct vessels who underwent PTCA). Contrast Left Ventriculography at Hospital Discharge Contrast left ventriculograms were performed before hospital discharge in 468 of the study patients (79.9%), and overall, EF averaged 49.3%. In the immediate invasive strategy, delayed invasive strategy, and conservative strategy groups, global EF of at least 50% was seen in 55.0%, 52.0%, and 43.6% of the patients (p=0.11), respectively, whereas EF of less than 30% was seen in 4.0%, 4.6%, and 7.3% (p=0.38), respectively. In patients without predischarge contrast EF, radionuclide EF data were used as previously described,7 and zeros were substituted for EFs of patients who died before EF assessment. These assumptions increased the number of available EF determinations to 558 (95.2%) and reduced the mean EF to 46.4%. Neither observed nor imputed contrast left ventriculographic EFs differed significantly among the three treatment groups (Figure 6, upper panel). Furthermore, end-systolic and end-diastolic volume indexes, cardiac index, and measurement of wall motion in the infarct and noninfarct zones, respectively, were similar among the three randomized groups (Table 3 and Figure 6, lower panel). Because of the slight imbalance in risk groups at baseline, EF was assessed separately in low-risk and not low-risk patients as well as in those patients with and without previous and anterior MIs, again showing no significant differences in predischarge contrast EF among the three treatment groups (Table 3). Radionuclide Ejection Fractions at Hospital Discharge and 6-Week Follow-up Paired rest and exercise radionuclide EFs were available at the time of hospital discharge in 399 of the patients (68.0%) and at the 6-week follow-up in

10 1466 Circulation Vol 81, No 5, May 1990 Percent Emm D1w _ Ins VA Cons SD/Chord 0 r (0.5) (1) (1.5) (2) (2.5) (3) N Infarct Site Non-Infarct Site 370 patients (63.1%). Reasons for missing paired radionuclide EF determinations at the time of hospital discharge included medical instability or orthopedic disability (43 patients), death less than 21 days after study entry (37 patients), recent CABG (20 patients), patient refusal (eight patients), technically unsatisfactory study (60 patients), and other reasons (19 patients); at the 6-week follow-up, reasons included death before 6 weeks (after study entry) (45 patients), medical instability or orthopedic disability (34 patients), recent CABG (seven patients), patient refusal (50 patients), technically unsatisfactory study (41 patients), and other reasons (39 patients). There was no difference in rest or exercise EF at either time among the three treatment groups (Table 4). Peak exercise stage was similar among the three groups and averaged 2.2 stages at the time of hospital discharge and 3.0 stages at 6 weeks. The difference between peak exercise and baseline EF was greater at the time of hospital discharge in the delayed invasive strategy group (mean, 4.2 EF units) as compared with the immediate invasive strategy (mean, 2.8 EF units) and the conservative strategy group (mean, 2.0 EF units) (p=0.03); at the 6-week follow-up, however, this trend was still present but no longer statistically significant. FIGURE 6. Bar graph showing contrast left ventriculography at hospital discharge. Upper panel: Ejection fraction, the primary study end point, was similar in the three treatment groups, both when ejection fraction was measured directly (Observed) and when zeros and radionuclide ejection fraction data were imputed for those patients in whom contrast ejection fraction was not measured (Imputed). Lower panel: Segmental wall motion abnormalities were of similar magnitude among the three treatment groups in both infarct site and noninfarct site. Imm, immediate; Del, delayed; Inv, invasive; Cons, conservative; solid bar, immediate invasive strategy group; double cross-hatched bar, delayed invasive strategy group; cross-hatched bar, conservative strategy group. Positive Exercise Tests at Hospital Discharge and 6-Week Follow-up Bicycle exercise tests were performed before hospital discharge in 461 patients (78.7%) and at 6-week follow-up in 416 patients (71.0%). Positive tests (i.e., chest pain, ST depression interpreted as ischemia by the TIMI clinic, or EF decrease.5 units) were observed in a similar proportion of patients within each of the three treatment groups (Table 5) although at the time of hospital discharge, there was a nonsignificant trend toward more positive tests in the conservative strategy group. This trend was no longer evident at the 6-week follow-up. The presence of a positive exercise test at the time of hospital discharge was significantly associated with residual infarct-vessel stenosis on the predischarge coronary arteriogram. Specifically, patients with positive rather than negative predischarge exercise tests had, on the average, fewer patent infarct arteries (77.1% vs. 86.4%, p=0.02), a higher mean infarct artery residual stenosis (61.6% vs. 52.8%, p =0.002), and fewer infarct arteries with stenosis of less than 60% (47.5% vs. 61.2%,p=0.01). Clinical Events The incidence of death, fatal and nonfatal reinfarction, and death or reinfarction was similar among

11 Rogers et al TIMI II-A Trial 1467 TABLE 4. Predischarge and Six-Week Follow-up Rest and Exercise Radionuclide Left Ventricular Ejection Fractions Immediate invasive Delayed invasive Conservative (n=195) (n =194) (n=197) Study data n Mean n Mean n Mean p Hospital discharge study Rest EF* Baseline exercise EFt Peak exercise EF Exercise EF-baseline EF Peak exercise stage EF change with exercise (n, %) Fall >5% No change Increase.5% Infarct zone segmental EF (rest) Computed t Adjusted week follow-up Rest EF* Baseline exercise EFt Peak exercise EF Exercise EF-rest EF Peak exercise stage EF change with exercise (n, %) Fall 25% No change Increase.5% Infarct zone segmental EF Computedt Adjusted EF, ejection fraction. *Rest EF was obtained on imaging table with patient supine. tbaseline exercise EF was obtained on exercise table with patient supine or semisupine and feet in stirrups of bicycle. tbased on average of regional wall motion in zone supplied by infarct-related artery. Adjusted for distribution of infarct-related coronary artery in each treatment group. TABLE 5. Bicycle Exercise Test Data at Hospital Discharge and 6-Week Follow-up Immediate invasive Delayed invasive Conservative (n=195) (n=194) (n=197) Test data n % n %S n % p Patients with exercise test performed Hospital discharge weeks Exercise test findings (% of all patients having exercise tests) Hospital discharge Chest pain ST depression* EF decrease.5 units Any of above, i.e., "positive test" weeks Chest pain ST depression* EF decrease.5 units Any of above, i.e., "positive test" EF, ejection fraction. *TIMI clinic interpretation of ischemic ST depression.

12 1468 Circulation Vol 81, No 5, May 1990 Cumulative Number of Patients with Major Procedures & Events Immediate Invasive Strategy (N = 195) Percent with PTCA CABG Death Re-MI 21 Days Weeks -_ Yeor -_ Delayed Invasive Strategy (N = 194) Percent with PTCA CABG Death Re-MI Conservative Strategy (N = 197) Percent with PTCA CABG Death Re-MI FIGURE 7. Diagram of cumulative number of patients with major procedures and events through 1 year after entry. Percentage of patients having (initial) percutaneous transluminal coronary angioplasty (PTCA) differed significantly (p<o.ooj) among the three strategy groups at each time interval. Coronary artery bypass graft surgery (CABG) frequency also differed (p <0.05) between groups at 21 days after entry but not at other time intervals. Death and reinfarction (Re- MI) occurred with equalffrequency among the three treatment groups at each time interval. the patients in the three treatment groups at 21 days, 42 days, and 1 year after entry (Figure 7 and Table 6). Death between 21 days and 1 year after entry in the patients in the immediate invasive strategy, delayed invasive strategy, and conservative strategy groups occurred in only 1.0%, 2.0%, and 2.6% of the patients, respectively. During the first 21 days after entry, CABG after PTCA was significantly more common in the patients in the immediate invasive strategy group than in the patients in the other two groups as were transfusions of more than 1 unit in patients not undergoing CABG (Table 6). At both the 6-week and l-year follow-ups, the patients in the three randomized treatment groups reported similar activity levels, angina class, and frequency of rehospitalization. The interim performance of CABG between the time of hospital discharge and 1 year after entry was similar among the patients in the three treatment groups and averaged 5.2% (Table 6). The cumulative performance of CABG throughout the first year after entry (including the initial hospitalization) was similar among the patients in the three groups and averaged 17.2% (Figure 7). The interim performance of PTCA between the time of hospital discharge and 1 year after entry was similar among the patients in the three groups and averaged 5.6% (Table 6). Between the time of entry to the study and 1-year follow-up, PTCA was performed in 75.8% of the patients in the immediate invasive strategy group, 64.3% of the patients in the delayed invasive strategy group, and 23.9% of the patients in the conservative strategy group (p<0.001). At the 1-year follow-up, revascularization (PTCA or CABG, or both) had been performed in 84.4% of the patients in the immediate invasive strategy group, 72.3% of the patients in the delayed invasive strategy group, and 38.6% of the patients in the conservative strategy group (p<0.001). Treadmill Exercise Test Findings at 1-Year Follow-up Treadmill exercise tests were performed in 397 patients (67.7%) at the 1-year follow-up visit. Exercise duration, hemodynamics, and frequency of ischemic response were similar among the patients in the three treatment groups (Table 7). Discussion The TIMI II-A trial demonstrates that patients receiving rt-pa therapy for AMI who are treated by an aggressive approach including either routine immediate (<2 hours) or delayed (18-48 hours) coronary arteriography followed by PTCA, if feasible, have no better global or segmental left ventricular function at the time of hospital discharge than patients managed by a more conservative approach in which coronary arteriography is delayed until hospital discharge and coronary revascularization is reserved for those patients with recurrent spontaneous or exercise-induced ischemia. Further, the conservatively managed patients have a similar proportion of patent infarct-related coronary arteries at the time of hospital discharge, similar rest and exercise EFs at the time of discharge and 6-week follow-up, and similar frequency of death, reinfarction, and rehospitalization during the subsequent year. Rationale for Routine PTCA Residual atherosclerotic plaque and thrombus compromise coronary perfusion and facilitate reocclusion after thrombolytic therapy for AMI, leading many to believe that prophylactic PTCA in this setting might improve myocardial function. Thus, when it was demonstrated that PTCA could be performed during and after thrombolytic therapy, the policy of routine PTCA after thrombolytic therapy quickly became established at many hospitals. The TIMI TI-A results reported herein and the overall TIMI II results recently published,10 however, have demonstrated that a conservative approach of watchful waiting without prophylactic PTCA produces results equivalent to the routine PTCA strategy. Why, then, has the promising concept of prophylactic PTCA in AMI not been proven valid?

13 Why Has Routine PTCA Not Been Proven Valuable? Unlikely eplanations PATIENT SELECTION. If TIMI II-A had selected very-low-risk patients for study, a good outcome might have been expected regardless of the treatment strategy used. But the TIMI II-A patients were not in a low-risk stratum as evidenced by the fact that two thirds of the patients who entered the trial were at higher-than-usual risk for AMI mortality because of their age of more than 70 years, history of previous MI, or presence of an anterior MI (Table 1). Patients were eligible for TIMI II regardless of the extent of left ventricular dysfunction if they were alert enough to provide written informed consent. Because rt-pa was available in most communities only through the TIMI II trial during the majority of the study enrollment period, there was considerable enthusiasm for referral of a wide spectrum of AMI patients to the trial, making a selection bias toward low-risk patients very unlikely. On the other hand, the inclusion and exclusion criteria for TIMI II were designed to select ideal candidates for thrombolytic therapy, namely, patients with recent onset of symptoms of AMI, definite clinical and electrocardiographic criteria for AMI, and low anticipated risk of hemorrhagic complications. Accordingly, TIMI II did not address the selection of reperfusion interventions in other groups of patients, for example, those patients more than 75 years old, those patients presenting late or with either unstable angina or non-q wave infarction, and those patients with contraindications to thrombolytic therapy. Although these groups of patients will require further study, their exclusion from TIMI II does not compromise the relevance of its findings to the care of the intended population. PTCA PERFORMANCE. If the patients randomized to the PTCA strategies had not received PTCA when indicated or if the results of PTCA had been poor, these facts might have explained why outcome in the invasive strategy groups failed to surpass that of the conservative strategy group. In the patients in the immediate invasive strategy and delayed invasive strategy groups, however, PTCA was performed within the protocol time window in 72% and 56% of the patients, respectively. Importantly, when PTCA was not performed, the reasons were consistent with good clinical judgment and common practice (generally, the lack of significant stenosis or the presence of disease better managed by CABG than by PTCA). Among patients receiving protocol PTCA, the procedure resulted in angiographic improvement (as assessed by the Radiographic Core Laboratory) in 77% of the patients in the immediate invasive strategy group and in 90% of the patients in the delayed invasive strategy group. (PTCA-related improvement was reported by the individual TIMI clinics as 84% in the patients in the immediate invasive strategy group and 93% in the patients in the delayed invasive strategy group.7) PTCA reduced the residual stenosis Rogers et al TIMI II-A Trial 1469 of the infarct artery by an average of 36% (absolute units) (Table 2); these PTCA results are excellent by present standards. ASSESSMENT OF LEFT VENTRICULAR FUNCTION. Left ventricular contrast EF at the time of hospital discharge was the primary end point of TIMI II-A. If this variable had been improperly or incompletely assessed, the interpretation of the study could have been biased. Direct measurement of contrast left ventricular EF was available in 80% of the study patients and averaged 49-50% in each of the three treatment groups (Table 3). In those patients without direct measurement of EF, we used radionuclide EF data (when available) and assigned an EF of zero to nonsurvivors; using these imputations, assessment of EF was possible in 95% of the study patients, but still there were no intergroup differences. In an attempt to identify subgroups of patients among the treatment groups in whom EF might have differed, we examined patients with previous MI, with anterior MI, and classified as not low risk for any reason. Further, we assessed parameters of left ventricular function that might be more sensitive than global EF (e.g., end-diastolic and end-systolic volume, regional wall motion in the infarct zone, and resting and exercise radionuclide EF at the time of hospital discharge and at the 6-week follow-up); however, we could find no significant difference in these indexes of ventricular function among the patients in the three randomized treatment groups. TIMING OF END-POINT ASSESSMENT. It might be argued that a hospital-discharge end point is too early to assess differences in outcome among the patients in the three treatment groups. Had the outcome of the conservative strategy been equivalent to the PTCA strategies during the initial hospitalization only to result in frequent relapses after hospital discharge, the advantages of this approach would be compromised. One-year follow-up data confirm, however, that not only is the mortality in the patients in the three groups similar at 1 year after entry (Table 6), but the risks of reinfarction alone or reinfarction and death are also similar (Figure 7 and Table 6). The incremental mortality (1.9%) between the time of hospital discharge and 1-year follow-up in the TIMI II-A population is extraordinarily low, especially in comparison with a previous large trial of thrombolytic therapy without systematic revascularization that showed an incremental mortality rate of 6.5%, between the time of discharge and 1-year follow-up. 17,18 The frequency of repeat hospitalization, as well as the frequency of interim cardiac catheterization, PTCA, and CABG, was similar among the patients in the treatment groups throughout the first year of follow-up. From entry until the 1-year follow-up, the proportion of patients having CABG did not differ in the three groups. Overall revascularization procedures through 1 year of follow-up remained substantially fewer in the conservative strategy group because of the lower use of PTCA.

14 1470 Circulation Vol 81, No 5, May 1990 v v v v 0 -o 0, 1-1 No0 n 0- X ON - 0,C \C N 00 _~ f C~r-: C r fi'ir _ _-" C 0 I" _. _ t w)or \00=- r10 00 e~n0c -~ldro 00 N nrcr- C14t 0o66 d _- 6 od ootr ir 0) C en 0 V) 00 It T* W- ~1- ON1 c CII -~I" e00 nr-,- ONN4,1r- r rlr (1-] = ON000 - ' " - t n - oo,-w, 1- -~ 1_ 0 1 0) 1- ON 11 0._;> a) CO.O 0) co 0. N -~ ~ 00 \ 00O e'nclc - Cl C~~~ON00 NNN- - V... rc. 0o6. 6 (i ---n t 1- - "Cl l nenn 45 cc r-rt-26 'It L(W 11 T "O N "Cl CAC 000CON 0neN00 c V) c - in = c C.~~~~~~~~~~~~~~~ 00 r cn 1- t l r. ' el l(1 r 00 ON N -cl 00 cl, ~i &4 ci- L-6 00N C'N N NL r) I. I-S t -VCN~NtŌN000-= -' 1r) ct m n - -o -- -, - 1 *> a0 Ci,a 0o', >.C Cu 4-0 0) F. *= C rm.t~~~~~~~~~~~~~~~~~~~~~~~~~~~~t Ct ed -ed - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~- - C $.0)0COC 0 $.i 4- cc 'S F5,0 C,c ccc A -ss sg. et < u8=uhouu3ze88z := v < u~~~~~~~~~~~~~~~~~~~~0

15 Rogers et al TIMI II-A Trial N N. OC o \Orfl r NO N- cc OO 0 )00 : : : : eno en N- - \ o v v v,c Ln 00 0 o en o 6o 6o vv v C) 00 - O - N cia c, In *_ en ri. ei&~.en O t N en ON O) en o -- N CO mn O o tr 6O 06 ON 0ON -- N en en ON4 - cn O 1 en0 -t t "t Ntn4V) en -4 N4 -.~ -_ N V) ONV-)t O NN en en t -Nen ttfln - en - -0 N n en - 0Nc t- N~ " Tt N. C -~ r- t,n.. `. CA,.... tn ne eno No tn en- N -- -N V-)'c V)UVr N-> m Ofl t en ONONO mm 4 CD o c o o ~\ONN<>b "RAi 0NONO -N en 0000ON)cr NN N ~ N enene NNN r NN" l. 00 ONON1~ t ONONONC tritri vtr do\oo enentcrc N ON NN N c 00 t-t t t ṯ I Ni ll10-1 \ en en en co 0 CO CO CO 3.t -^ CO- 0c) r.q <n COQ m 01) C) CO 0.0 cvnc.9 > o= COt: U CO U) 0. OC)Q C)O< U h. 0 0 C) 0 - O O~~~~~~~~~C * _ Qu _~~~~~~~C * lh,= * W~~ C) CO 0..c t U *. CO C 0 c 0CO 0 N p.t M. - '0 6n :>.0 C)U co C) m -~ U -C)~ p ' -_ 3 > t 3>, \O - 0o - P..

16 1472 Circulation Vol 81, No 5, May 1990 TABLE 7. Treadmill Exercise Test Findings at 1-Year Follow-up Immediate invasive (n = 195) Delayed invasive (n= 194) Conservative (n = 197) Exercise test data n % n % n % p Patients with test Exercise duration (mean min) Heart rate (beats/min) Baseline (mean) Peak exercise (mean) Systolic blood pressure (mm Hg) Baseline (mean) Peak exercise (mean) Diastolic blood pressure (mm Hg) Baseline (mean) Peak exercise (mean) Exercise-induced angina Exercise-induced flat or downsloping ST depression of >1.0 mv beyond baseline "Severe ischemic exercise response`* *Final exercise stage less than stage 3 and 1) exercise limiting angina, 2) J-point depression > 1.5 mm with horizontal or downsloping ST segments 80 msec after J point, 3) J-point elevation > 1 mm with horizontal or upsloping ST segments in a non-q wave lead, or 4) J-point depression.2.1 mm with upsloping ST segments but depressed <2 mm 80 msec after the J point. Likely explanations REVASCULARIZATION PERFORMED ONLY WHEN IT COUNTS" IN THE CONSERVATIVE STRATEGY GROUP. In TIMI I-A, PTCA was performed for clinically evident recurrent spontaneous ischemia in 9.1% of the patients in the conservative strategy group within the first 7 days after entry. Exercise testing suggested that the conservative strategy group was still more likely to have evidence of residual ischemic myocardium shortly before hospital discharge, a finding that generally predicts poorer 1-year survival.19 These exercise test findings corresponded to quantitative coronary arteriography at the time of hospital discharge, which confirmed that the infarct-related arteries in this subgroup of patients had more severe residual stenoses than those patients in the invasive strategy subgroups. The potentially deleterious consequences of these findings, however, were attenuated by the subsequent revascularization procedures that they prompted. PTCA, CABG, or both were performed after exercise testing in 14.4% of the patients in the conservative strategy group by 21 days, in 18.7% of the patients by 42 days, and in 27.3% of the patients by 1 year after entry. It is highly likely that targeting revascularization interventions to patients with spontaneous or exercise-induced ischemia minimized recurrent ischemic events in the conservative strategy group and rendered their outcome equivalent to the patients in the routine PTCA groups (in whom revascularization was performed less selectively). This finding is consistent with previous studies showing that limited exercise testing before the time of hospital discharge is capable of identifying groups of patients at high and low risk for recurrent ischemic events during the subsequent year Waiting to perform revascularization until this test is performed seems to carry no increased risk of mortality or reinfarction for the patient. Furthermore, patients in whom revascularization was not used according to conservative strategy might have enjoyed an equally favorable outcome because of a subcritical (and clinically unimportant) residual infarct-artery stenosis, a lack of remaining viable myocardium in the infarct zone, or freedom from the small but finite risks accompanying revascularization. FAILURE OF RESCUE PTCA TO AUGMENT OVERALL OUTCOME. Patency of the infarct-related artery is observed 90 minutes after initiation of intravenous rt-pa treatment in approximately 75% of patients.23 It has been proposed that routine immediate coronary arteriography would be useful in all patients receiving thrombolytic therapy to identify the 25% of patients with still-occluded infarct-related arteries and offer them rescue angioplasty.24 Although the TIMI 11-A trial was not designed specifically to test the benefit of rescue angioplasty, indirect information on this approach is available. Among the 29 patients in the immediate invasive strategy group who underwent acute dilation of totally occluded vessels, the primary success rate was 90% and the hospital mortality rate was 6.9%; however, the reocclusion rate of these rescue PTCA vessels was 20.8% at the time of predischarge angiography. In contrast, among the 67 patients with occluded infarct-related arteries on initial arteriography in the other two groups, most (61.2%) were discharged without PTCA or CABG and with a hospital mortality of 7.3%. Thus, from this assessment of patients in nonrandomized subgroups, rescue PTCA after rt-pa does not guarantee continued arterial patency at discharge, and the majority of patients with persistently occluded infarct-related arteries do well.

17 Clinical Implications The TIMI TI-A trial demonstrates that routine immediate PTCA in patients receiving rt-pa for AMI is associated with significant morbidity as compared with a less aggressive strategy of delaying the PTCA for hours. Routine immediate or delayed early catheterization followed by PTCA, however, is not required to preserve ventricular function in patients receiving rt-pa for AMI. Instead, patients can be safely managed by a conservative policy of watchful waiting, followed by predis- This analysis does not include patients in the delayed invasive strategy and conservative strategy groups who died before catheterization (unknown patency status), and thus, a randomized trial of rescue PTCA might still be considered. But proponents should remember that routine early catheterization to detect occlusion exposes all patients (occluded or patent) to certain risks not associated with conservative strategy, which might compromise any benefits of rescue PTCA and yield overall results worse than with conservative strategy. INCREASED COMPLICATIONS OF ROUTINE PTCA IN AMI. Goals of routine PTCA in AMI are to improve infarct-vessel caliber, lower rate of reocclusion, and improve ventricular function and patency rate at the time of discharge. In this study, however, infarct-related vessel patency was the same regardless of whether patients had been assigned to routine PTCA. Although early routine PTCA might have benefited some patients, these benefits might have been obscured by serious complications that occurred in other patients. In both invasive strategy treatment groups, infarct vessels that had been patent after PTCA were occluded at the time of hospital-discharge catheterization (17.6% in the patients in the immediate invasive strategy group and 4.3% in the patients in the delayed invasive strategy group), consistent with previous reports documenting significant reocclusion frequencies in patients undergoing PTCA in the setting of AMI.2,3,5,25 Patients undergoing immediate PTCA during thrombolytic therapy seem to be at a particularly increased risk for reocclusion, possibly secondary to intraplaque hemorrhage26 associated with balloon-induced trauma in this setting. Alternatively, balloon-induced trauma to plaque, vascular endothelium, and fresh thrombus can stimulate new thrombus formation (and reocclusion) by the exposure of thrombogenic substrates to circulating blood.27 Two other major trials of immediate PTCA in patients receiving rt-pa, similar to TIMI IT-A, have failed to demonstrate improvement in ventricular function as compared with less aggressive treatment strategies.23'28 One of these trials28 also observed an increased requirement for blood transfusions because of a higher hemorrhagic complication rate in the immediate PTCA group, similar to that noted in the TIMI IT-A immediate invasive strategy group (Table 6). Early coronary arteriography and PTCA in the TIMI TI-A invasive strategy treatment groups occasionally resulted in early CABG. CABG after PTCA was performed significantly more frequently in the patients in the immediate invasive strategy group (7.7%) than in the patients in the delayed invasive strategy group (2.1%) or the conservative strategy group (2.5%) (p=0.01); the outcome of CABG performed for failed PTCA was much less favorable than that of CABG performed as a primary revascularization mode for patients with complex coronary anatomy. The latter group of patients, of course, was a highly select population who had already survived Rogers et al TIMI II-A Trial 1473 the acute phase of AMI, a fact that in itself might have explained their lower surgical risk. Other considerations COMPARISON WITH LARGER TIMI II MAIN STUDY. Although sharing some patients in common with the TIMI II main study10 (Figure 1), the TIMI 1I-A trial differed importantly in the following several design features: 1) the presence of the immediate catheterization or PTCA group (the immediate invasive strategy group), 2) the use of left ventricular EF as the primary end point rather than death or reinfarction, and 3) the requirement that all patients undergo coronary arteriography and left ventricular angiography before hospital discharge. Despite these differences, the results of the TIMI II-A trial were remarkably consistent with those of the larger TIMI TI main study, showing no benefit with the use of the invasive management strategy. CONCOMITANT PHARMACOLOGICAL THERAPY. The role of ancillary pharmacological agents for patients receiving thrombolytic therapy for AMI is incompletely defined and the subject of ongoing clinical trials. Although the TIMI IT-A trial routinely used nifedipine and lidocaine, metaanalyses of the use of these agents for AMI in the prethrombolytic era have shown no benefit.29 These drugs, however, have not been adequately evaluated in the setting of thrombolytic treatment for AMI, and it is unknown whether nifedipine might be useful when PTCA is contemplated and whether lidocaine might be protective against ventricular arrhythmias occurring in the setting of reperfusion. Intravenous heparin, used routinely after rt-pa initiation in TIMI IT, has been demonstrated in two recent controlled trials30'31 to promote infarct vessel patency at 7-72 hours after rt-pa initiation. Aspirin reduces the incidence of acute coronary thrombosis in patients undergoing PTCA32,33 and also reduces mortality synergistically with intravenous streptokinase in patients with AMI.34 In TIMI TI-A, aspirin was begun on the day of rt-pa initiation in one third of the patients and on the subsequent day in the remainder of the patients. It is conceivable that if aspirin had been initiated on day 1 in all patients, the incidence of reocclusion in patients in the immediate invasive strategy group undergoing PTCA might have been lower but possibly at the risk of more hemorrhagic complications.

18 1474 Circulation Vol 81, No 5, May 1990 charge exercise testing and coronary arteriography to evaluate ischemia. The TIMI II-A experience suggests that during the first 7 days, coronary revascularization for clinically evident recurrent ischemia will be required in approximately 10% of these patients. Although the conservatively managed patients have more severe infarct-vessel stenoses at the time of predischarge catheterization and a higher likelihood of a positive exercise test (leading to coronary revascularization in another 15% of the patients during the next 2 weeks), the frequency of major events (i.e., death and reinfarction) over the subsequent year is similar to patients managed with routine PTCA. Because the conservative approach achieves equally good short- and long-term outcome with less morbidity and a lower use of PTCA, it seems to be the preferred management strategy. Appendix TIMI Phase II Participants Study Chairman. Eugene Braunwald, MD, Harvard University, Boston. Clinical Centers Phase Il-A Baylor College of Medicine, Houston, TX. Principal Investigator: Robert Roberts, MD. Coinvestigators: Phyllis Nelson, RN, Steven Minor, MD, Craig Pratt, MD, Albert Raizner, MD, William L. Winters, MD, Mario S. Verani, MD, John M. Lewis, MD, Jacques Heibig, MD, Neal Kleiman, MD, Mary Kay Vander- Molen, RN. Brown University, Providence, RI. Principal Investigator: David 0. Williams, MD. Coinvestigators: Thomas M. Drew, MD, Raymon S. Riley, MD, Harvey J. White, MD, David Shefcyk, MD, John Joelson, MD, Edward Thomas, MD, Barry Sharaf, MD, Frank Fedele, MD, Michael Nathanson, MD, George Mc- Kendall, MD, David Becker, MD, Diane L. Hardink, RN, Mark Macedo, RN, Gary Weeks, MD, Robert Mich, MD, Eric Berger, MD. Harvard University, Boston, MA. Principal Investigator: Donald S. Baim, MD. Coinvestigators: Daniel Diver, MD, Steven Herson, MD, John E. Markis, MD, Raymond G. McKay, MD, Beverly Lorell, MD, Cynthia (Brewer) Senerchia, RN, MS, Gail A. Carey, RN, Jeffrey Schweiger, RN. Mayo Foundation, Rochester, MN. Principal Investigator: James H. Chesebro, MD. Coinvestigators: Dennis R. Bresnahan, MD, Bernard J. Gersh, MD, Fletcher A. Miller, MD, Michael B. Mock, MD, Hugh C. Smith, MD, Robert Frye, MD, David Hayes, MD, Ian Clements, MD, William Freeman, MD, John Rumberger, MD, Ray Gibbons, MD, Rick Nishimura, MD, Richard Rodeheffer, MD, Roger Click, MD, Jae Oh, MD, Lawrence Sinak, MD, Diane Klees, LPN, Laurie Meyers, LPN, Ronald Vlietstra, MD, John Bresnahan, MD, David Holmes Jr., MD, Guy Reeder, MD. University of Massachusetts, Worcester, AL4. Principal Investigator: Joel Gore, MD. Coinvestigators: Joseph S. Alpert, MD, Joseph R. Benotti, MD, Jeffrey Leppo, MD, Ira S. Ockene, MD, James F. Rippe, MD, Bonnie H. Weiner, MD, James Dalen, MD, John M.J. Gaca, MD, Steven P. Ball, RN, Jeanne Corrao, RN, Cathleen Mahan, RN. Bridgeport Hospital, Bridgeport, CT. Principal Investigator: Joseph D. Babb, MD. Coinvestigators: Zosimo A. Adefuin, MD, Mitchell Driesman, MD, Jay Meizlish, MD, Donna Yasick, RN. University of Alabama, Birningham, AL. Principal Investigator: William J. Rogers, MD. Coinvestigators: Joaquin G. Arciniegas, MD, William A. Baxley, MD, Robert C. Bourge, MD, Thomas M. Bulle, MD, Terry B. Cooper, MD, Larry S. Dean, MD, Randall Hess, DO, William A.H. MacLean, MD, Silvio E. Papapietro, MD, Carlos Saenz, MD, Alfred W.H. Stanley, MD, Michael T. Simpson, MD, Karen Bynum, RN, Terri Eubanks, RN, Larry Maske, RN. Coordinating Center Maryland Medical Research Institute, Baltimore, MD. Principal Investigator: Genell L. Knatterud, PhD. Coinvestigators: Michael L. Terrin, MD, MPH, Sandra Forman, MA, Dorothy T. Harris, Richard Ross, MSc, Patricia C. Wilkins, BS, Margaret Bryant, PhD, Paul L. Canner, PhD, Margie Carroll, Joyce Depkin, BS, Judy Dotson, Cheryl Fiery, Moscoe Johnson, Cheryl Kelly, Peggy Noble, BS, Bruce Thompson, PhD, William R. Bell, MD, Leonard Scherlis, MD. Radiographic Core Laboratory University of Washington, Seattle, WA. Principal Investigator: Harold T. Dodge, MD. Coinvestigators: B. Greg Brown, MD, PhD, J. Ward Kennedy, MD, Florence H. Sheehan, MD, Brad Bisson, Edward Bolson, MS. Radionuclide Core Laboratory Yale University, New Haven, CT. Principal Investigator: Barry Zaret, MD. Coinvestigators: Frans Wackers, MD, David S. Kayden, MD, Kathleen Davis, RTNM, Robin Green, RTNM. Coagulation Core Laboratory University of Vermont, Burlington, VT. Principal Investigator: Kenneth Mann, PhD. Coinvestigators: David Stump, MD, Desire Collen, MD, Edwin Bovill, MD, Russell Tracy, PhD. ECG Core Laboratory for Qualifying ECGs George Washington University, Washington, DC. Principal Investigator: Allan M. Ross, MD. Coinvestigators: George B. Bren, MD, Alan G. Wasserman, MD. ECG Core Laboratory for Exercise ECGs St. Louis University, St. Louis, MO. Principal Investigator: Bernard R. Chaitman, MD. Coinvestigators: Robert D. Wiens, MD, Leslee Shaw, MS, Maryellen Haueisen, BS, Liwa T. Younis, MD, PhD.

19 coronary thrombolysis with recombinant tissue plasminogen activator: Importance of coronary angioplasty. J Am Coll Cardiol 1985;6: Erbel R, Pop T, Henrichs KJ, von Olshausen K, Schuster CJ, Rupprecht HJ, Steuernagel C, Meyer J: Percutaneous transluminal coronary angioplasty after thrombolytic therapy: A prospective controlled randomized trial. J Am Coll Cardiol 1986;8: Williams DO, Ruocco NA, Forman S, and the TIMI Investigators: Coronary angioplasty after recombinant tissue-type plasminogen activator in acute myocardial infarction: A report from the thrombolysis in myocardial infarction (TIMI) trial. J Am Coil Cardiol 1987;10:45B-50B 7. The TIMI Research Group: Immediate vs. delayed catheterization and angioplasty following thrombolytic therapy for acute myocardial infarction: TIMI II A results. JAMA 1988; 260: TIMI Operations Committee, Braunwald E, Knatterud GL, Passamani E, Robertson TL: Announcement of protocol change in thrombolysis in myocardial infarction trial (letter). J Am Coll Cardiol 1987;9: TIMI Operations Committee, Braunwald E, Knatterud GL, Passamani ER, Robertson TL, Solomon R: Update from the thrombolysis in myocardial infarction trial (letter). JAm Coll Cardiol 1987;10: The TIMI Study Group: Comparison of invasive and conservative strategies after treatment with intravenous tissue plasminogen activator in acute myocardial infarction: Results of the thrombolysis in myocardial infarction (TIMI) phase II trial. N EnglJ Med 1989;320: Callahan RF, Froelich JW, McKusick KA, Leppo J, Strauss HW: A modified method for the in vivo labeling of red blood cells with Tc-99m: Concise communication. J Nucl Med 1982; 23: Sandler H, Dodge HT: The use of single plane angiocardiograms for the calculation of left ventricular volume in man. Am Heart J 1968;75: Sheehan FH, Bolson EL, Dodge HT, Mathey DG, Schofer J, Woo HW: Advantages and applications of the centerline method for characterizing regional ventricular function. Circulation 1986;74: Brown BG, Gallery CA, Badger RS, Kennedy JW, Mathey D, Bolson EL, Dodge HT: Incomplete lysis of thrombus in the moderate underlying atherosclerotic lesion during intracoronary infusion of streptokinase for acute myocardial infarction: Quantitative angiographic observations. Circulation 1986;73: Bruce RA, Hornsten TR: Exercise stress testing in evaluation of patients with ischemic heart disease. Prog Cardiovasc Dis 1979;11: SAS Institute, Inc: SAS User's Guide: Statistics, Version 5 Edition. Cary, NC, SAS Institute Inc, 1985, pp Gruppo Italiano Per Lo Studio Della Streptochinasi Nell'Infarto Miocardico (GISSI): Effectiveness of intravenous throm- National Heart, Lung, and Blood Institute Program Office National Institutes of Health, Bethesda, MD. Principal Investigator: Eugene R. Passamani, MD. Coinvestigators: Thomas L. Robertson, MD, Gordon Lan, PhD, Rachel Solomon, MHS, George Sopko, MD. Pathology Core Laboratory National Institutes of Health, Clinical Center, Bethesda, MD. Principal Investigator: William C. Roberts, MD. Coinvestigator: Jay Kalan, MD. PTCA Quality Control Laboratory Brown University, Providence, Rl. Principal Investigator: David 0. Williams, MD. Coinvestigators: Ramon Riley, MD, Harvey White, MD, Barry Sharaf, MD, Frank Fedele, MD, Edward Thomas, MD, Thomas Drew, MD, John Joelson, MD, Diane Hardink, RN. Drug Distribution Center Cooperative Studies Program, VA Medical Research Service, Albuquerque, NM. Principal Investigator: Cindy Colling, RPh, MS. Coinvestigators: Claire Haakenson, RPh, MS, Mike Sather, RPh, MS. TIMI Phase II Committees Executive Committee. Eugene Braunwald, MD, Chairman; Bernard Chaitman, MD, James Chesebro, MD, Harold Dodge, MD, Genell Knatterud, PhD, Kenneth Mann, PhD, Hiltrud Mueller, MD, Eugene Passamani, MD, Robert Roberts, MD, William Rogers, MD; Burton Sobel, MD, David Stump, MD, David Williams, MD, Barry Zaret, MD. Hemorrhagic Event Review (HERC). James Chesebro, MD, Chairman; Andrew Berke, MD, Edwin Bovill, MD, Frederick Feit, MD, Joel Gore, MD, L. David Hillis, MD, Costas Lambrew, MD, Roy Leiboff, MD, John Markis, MD, Louis Offen, MD, Craig Pratt, MD, Scott Sharkey, MD, George Sopko, MD, Michael Terrin, MD. Mortality and Morbidity Classification Committee (MMCC). Myron Weisfeldt, MD, Chairman; William Baker, MD, Michael Cowley, MD, Kenneth Kent, MD, Edgar Lichstein, MD, Thomas Robertson, MD, Leonard Scherlis, MD, Michael Terrin, MD. Safety and Data Monitoring Committee. Francis Klocke, MD, Chairman; John Bailar, MD, Richard Conti, MD, David DeMets, PhD, Valentin Fuster, MD, Thomas Killip, MD, Harold Roberts, MD, Leroy Walters, PhD. Ex-officio Members: Eugene Braunwald, MD, Genell Knatterud, PhD, Eugene Passamani, MD, Thomas Robertson, MD. Steering Committee. Members of the Steering Committee are the Study Chairman and Principal Investigators from TIMI Clinical Centers, Core Laboratories, Coordinating Center, and NHLBI Program Office. Rogers et al TIMI II-A Trial 1475 References 1. Harrison DG, Ferguson DW, Collins SM, Skorton DJ, Ericksen EE, Kioschos JM, Marcus ML, White CW: Rethrombosis after reperfusion with streptokinase: Importance of geometry of residual lesions. Circulation 1984;69: Meyer J, Merx W, Schmitz H, Erbel R, Kiesslich T, Dorr R, Lambertz H, Bethge C, Krebs W, Bardos P, Minale C, Messmer J, Effert S: Percutaneous transluminal coronary angioplasty immediately after intracoronary streptolysis of transmural myocardial infarction. Circulation 1982;66: Papapietro SE, MacLean WA, Stanley AW Jr, Hess RG, Corley N, Arciniegas JG, Cooper TB: Percutaneous transluminal coronary angioplasty after intracoronary streptokinase in evolving acute myocardial infarction. Am J Cardiol 1985; 55: Topol EJ, Weiss JL, Brinker JA, Brin KP, Gottlieb SO, Becker LC, Bulkley BH, Chandra N, Flaherty JT, Gerstenblith G, Gottlieb SH, Guerci AD, Ouyang P, Llewellyn MP, Weisfeldt ML, Shapiro EP: Regional wall motion improvement after

20 1476 Circulation Vol 81, No 5, May 1990 bolytic treatment in acute myocardial infarction. Lancet 1986; 1: Gruppo Italiano Per Lo Studio Della Streptochinasi Nell'Infarto Miocardico (GISSI): Long-term effects of intravenous thrombolysis in acute myocardial infarction: Final report of the GISSI study. Lancet 1987;2: Corbett JR, Dehmer GJ, Lewis SE, Woodward W, Henderson E, Parkey RW, Blomqvist CG, Willerson JT: The prognostic value of submaximal exercise testing with radionuclide ventriculography before hospital discharge in patients with recent myocardial infarction. Circulation 1981;64: Theroux P, Waters DD, Halphen C, Debaisieux JC, Mizgala HF: Prognostic value of exercise testing soon after myocardial infarction. N Engl J Med 1979;301: Davidson DM, DeBusk RF: The prognostic significance of a single exercise test 3 weeks after uncomplicated myocardial infarction. Circulation 1980;61: Weld FM, Chu KL, Bigger JT, Rolnitzky LM: Risk stratification with low-level exercise testing 2 weeks after uncomplicated myocardial infarction. Circulation 1981;64: Topol EJ, Califf RM, George BS, Kereiakes DJ, Abbottsmith CW, Candela RJ, Lee KL, Pitt B, Stack RS, O'Neill WW, and the Thrombolysis and Angioplasty in Myocardial Infarction Study Group: A randomized trial of immediate versus delayed elective angioplasty after intravenous tissue plasminogen activator in acute myocardial infarction. N Engl J Med 1987; 317: Califf RM, Topol EJ, George BS, Boswick JM, Lee KL, Stump D, Dillon J, Abbottsmith C, Candela RJ, Kereiakes DJ, O'Neill WW, Stack RS, and the TAMI Study Group: Characteristics and outcome of patients in whom reperfusion with intravenous tissue-type plasminogen activator fails: Results of the thrombolysis and angioplasty in myocardial infarction (TAMI) I trial. Circulation 1988;77: Stack RS, O'Connor CM, Mark DB, Hinohara T, Phillips HR, Lee MM, Ramirez NM, O'Callaghan WG, Simonton CA, Carlson EB, Morris KG, Behar VS, Kong Y, Peter RH, Califf RM: Coronary perfusion during acute myocardial infarction with a combined therapy of coronary angioplasty and highdose intravenous streptokinase. Circulation 1988;77: Waller BF, Rothbaum DA, Pinkerton CA, Cowley MJ, Linnemeier TJ, Orr C, Irons M, Helmuth RA, Wills ER, Aust C: Status of the myocardium and infarct-related coronary artery in 19 necropsy patients with acute recanalization using pharmacologic (streptokinase, r-tissue plasminogen activator), mechanical (percutaneous transluminal coronary angioplasty) or combined types of reperfusion therapy. J Am Coll Cardiol 1987;9: Fuster V, Badimon L, Cohen M, Ambrose JA, Badimon JJ, Chesebro J: Insights into the pathogenesis of acute ischemic syndromes. Circulation 1988;77: Simoons ML, Arnold AER, Betriu A, DeBono DP, Col J, Dougherty FC, Von Essen R, Lambertz H, Lubsen J, Meier B, Michel PL, Raynaud P, Rutsch W, Sanz GA, Schmidt W, Serruys PW, Thery C, Uebis R, Vahanian A, Van de Werf F, Willems GM, Wood D, Verstraete M for the European Cooperative Study Group for Recombinant Tissue-Type Plasminogen Activator (rtpa): Thrombolysis with tissue plasminogen activator in acute myocardial infarction: No additional benefit from immediate percutaneous coronary angioplasty. Lancet 1988;1: Yusuf S, Wittes J, Friedman L: Overview of results of randomized clinical trials in heart disease: I. Treatments following myocardial infarction. JAMA 1988;260: Bleich SD, Nichols T, Schumacher R, Cooke D, Tate D, Steiner C, Brinkman D: The role of heparin following coronary thrombolysis with tissue plasminogen activator (t-pa) (abstract). Circulation 1989;80(suppl II):II Ross AM, Hsia J, Hamilton W, Chaitman B, Roberts R, Kleiman NS: Heparin vs. aspirin after recombinant tissue plasminogen activator therapy in myocardial infarction: A randomized trial (abstract). JAm Coll Cardiol 1990;15:64A 32. Barnathan ES, Schwartz JS, Taylor L, Laskey WK, Kleaveland JP, Kussmaul WG, Hirshfeld JW: Aspirin and dipyridamole in the prevention of acute coronary thrombosis complicating coronary angioplasty. Circulation 1987;76: Schwartz L, Bourassa MG, Lesperance J, Aldridge HE, Kazim F, Salvatori VA, Henderson M, Bonan R, David PR: Aspirin and dipyridamole in the prevention of restenosis after percutaneous transluminal coronary angioplasty. N Engl J Med 1988;318: ISIS-2 Collaborative Group: Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2. Lancet 1988; 2: KEY WORDS * percutaneous transluminal coronary angioplasty * thrombolysis * ejection fraction * clinical trials

21 Comparison of immediate invasive, delayed invasive, and conservative strategies after tissue-type plasminogen activator. Results of the Thrombolysis in Myocardial Infarction (TIMI) Phase II-A trial. W J Rogers, D S Baim, J M Gore, B G Brown, R Roberts, D O Williams, J H Chesebro, J D Babb, F H Sheehan and F J Wackers Circulation. 1990;81: doi: /01.CIR Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 1990 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: published Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Circulation is online at: