The influence of tissue plasminogen activator (t-PA) and heparin versus heparin alone on anatomic characteristic of patent infarct-related coronary arteries and the development of these angiographic descriptors in coronary arteries that remain patent during the hospital course was examined in 108 patients who participated in a placebo-controlled trial of recombinant tissue-type plasminogen activator in acute myocardial infarction. Coronary angiography was performed 18 [+ or -] 6 hours after treatment in 47 patients (group A) and at 10 days in 61 patients (group B). Quantitative coronary angiography of the infarct-related lesion was performed, and luminal irregularity was quantitated with an ulceration index. Of the 47 patients in group A, 7 (29%) treated with placebo had Thrombolysis in Myocardial Infarction grade 2 or 3 perfusion, whereas 18 (78%) treated with t-PA had grade 2 or 3 (p Angiographic studies demonstrated the capacity of thrombolytic therapy to reestablish patency of occluded, infarct-related coronary arteries in patients presenting with acute transmural myocardial infarction. However, reocclusion after initial successful thrombolysis remains a difficult clinical problem, with an incidence ranging between 15 and 30%.[1] It has been suggested that reocclusion occurs more frequently at the site of severe residual coronary stenosis[2]; however, this issue remains controversial.[3] It is also now known that after thrombolytic therapy has been completed and patency reestablished, considerable remodeling of the residual coronary stenosis may occur during the subsequent hospital course.[4] However, there is little information concerning the angiographically defined natural history of coronary vessels that reperfuse spontaneously in patients who are not treated with thrombolytic therapy. If significant changes were to occur in these patients, it would be important to account for them in evaluating future, angiographically based, mechanistic studies of the efficacy of thrombolytic therapy. New studies of this problem are unlikely to emerge given the current ethical imperative to administer thrombolytic therapy to such patients. We previously reported the results of a randomized, placebo-controlled, double-blind trial of tissue plasminogen activator (t-PA) versus placebo in heparinized patients with acute Q-ware myocardial infarction in which left ventricular global and regional function was the principal end point.[5] In the present report of those patients, we compare the influence of t-PA and heparin versus heparin alone on: (1) qualitative and quantitative anatomic characteristics of patent infarct-related coronary arteries soon after presentation of acute myocardial infarction, and (2) the development of these angiographic descriptors in coronary arteries that remain patent during the hospital course. METHODS The details of this protocol were reported previously.[5] Inclusion criteria for study entry were: age between 20 and 75 years; and typical ischemic chest pain lasting >30 minutes, with ST elevation [is greater than or equal to] 1 mm in [is greater than or equal to] 2 inferior leads or leads and aVL, or [is greater than or equal to] 1.5 mm in [is greater than or equal to] 2 anterior precordial leads. Patients were excluded if they presented >3.75 hours after the onset of chest pain. Patients were recruited from 10 collaborating Toronto hospitals and randomized to receive either double-chain t-PA (Burroughs Wellcome) or matching placebo. Study drug: The first 39 patients received the study drug intravenously at 0.4 MU/kg/hour for hour 1, 0.14 MU/kg/hour for hour 2, and 0.03 MU/kg/hour for the subsequent 8 hours; the final 79 patients received 0.4 MU/kg/hour for hour 1, 0.08 MU/kg/hour for hour 2, and 0.03 MU/kg/hour for the subsequent 4 hours. Ten percent of the first hour's dose was administered as an initial bolus injection. This change in infusion duration was made as a result of independent studies with this agent indicating that longer infusion times may increase hemorrhagic risk without providing improved patency.[6] All patients received heparin, concomitant with the study drug, commencing with a bolus injection of 4,000 IU, followed by an infusion of 1,000 IU/hour. Heparin administration was continued for 96 hours with the objective of maintaining the partial thromboplastin time between 1.5 and 2 times the control value. In all patients, 325 mg/day of enteric-coated aspirin was commenced orally before cessation of heparin therapy. Cardiac catheterization: Angiography was performed according to the schema outlined in Figure 1. Forty-seven patients initially admitted to 3 core institutions with facilities for cardiac catheterization underwent coronary and left ventricular angiography, according to protocol, 18 [+ or -] 6 hours after administration of study drug (group A). Because we also wanted to examine the development of both quantitative and qualitative angiographic features between the first and second angiograms obtained in group A patients, we performed this comparison in all patients who did not have 100% obstruction of the infarct-related coronary artery on the initial angiogram (i.e., eligible patients for this analysis were those with Thrombolysis in Myocardial Infarction [TIMI] perfusion grades 1 to 3). Repeat coronary angiography was performed in those patients on hospital day 10. Of 66 patients initially admitted to hospitals without facilities for catheterization (group B), 61 underwent coronary angiography only on day 10 after transfer to 1 of 3 randomly assigned core hospitals. One patient from each of groups A and B was excluded from angiographic analysis, because their angiograms were not available for interpretation at the core laboratory. Four patients in group B did not undergo angiography for other reasons (2 died and 2 refused). Angiographic assessment of the infarct-related artery: All coronary angiograms were independently reviewed at the core angiographic site without knowledge of treatment assignment or timing of angiography. The infarct-related coronary artery lesion was determined using the admission electrocardiogram and left ventricular angiogram, and by identification of characteristic infarct-lesion morphology.[7] Infarct-related perfusion was classified according to the criteria of the TIMI investigators; for purposes of analysis of reocclusion, patency in group A was defined as TIMI perfusion grade 2 or 3.[8] Quantitative coronary angiography was performed by obtaining 2 orthogonal projections of the infarct-related lesion. Each projection was quantitative analyzed using the Coronary Angiography Analysis System method of Reiber et al.[9] The lumen was identified by an automatic edge-detection algorithm.[10] The lesion outline was corrected for magnification and pincushion distortion, and the minimal lesion diameter, diameter of the adjacent normal segment and percent diameter stenosis were determined geometrically for each projection. The cross-sectional area could then be determined by the product of the diameters in the right and left anterior oblique projections.[10] In 11 angiograms, analysis could only be performed in 1 projection; in these cases, the lumen was assumed to be circular. The degree of luminal irregularity was quantitated by calculation of an ulceration index according to methods described previously.[7] The infarct-related lesion was divided into 10 equal segments. The diameter of the artery at its second most stenotic point was divided by the maximal intralesional diameter. This number, the ulceration index, is inversely proportional to the degree of luminal irregularity within the lesion and is independent of the maximal luminal stenosis. Care was taken to reproduce similar projection views in the repeat studies of patients with 2 angriograms. Statistical analysis: Demographic and clinical baseline characteristics of the 2 treatment groups were summarized previously.[5] Continuous data are presented as mean [+ or -] SE. Categoric variables were compared using chi-square. Normally distributed, continuous variables were compared using Student's t test (2-tailed), with paired or unpaired observations as appropriate. Non-normally distributed continuous variables were compared using Wilcoxon signed-rank test. RESULTS The distribution of TIMI perfusion grades in the 47 group A patients who underwent early catheterization is depicted in Figure 2 according to treatment assignment. Seven patients (29%) treated with placebo had TIMI grade 2 or 3 perfusion, whereas 18 (78%) treated with t-PA had grade 2 or 3 patency (p In Figure 3, the frequency of TIMI grade 2 or 3 perfusion in patients undergoing early catheterization is compared with that in those undergoing only late catheterization (group B). In contrast to the early patency data from group A (placebo 29% and t-PA 78%), group B showed a placebo patency of 59% (17 of 29) at day 10 as compared with a t-PA patency of 75% (24 of 32) (p = 0.10) (observed difference 16%; 95% confidence interval -8 to 40%). The development of minimal luminal area and ulcerative index for group A patients with patent arteries is shown in Table I; analogous data for patients in group B who were studied only on hospital day 10 are also shown. The minimal luminal area increased significantly in both placebo- and t-PA-treated patients between 18 hours and 10 days. Although this change on an individual basis was proportionately greater in patients treated with t-PA than with placebo (87 vs 41%), it did not achieve statistical significance. There was no difference in luminal characteristics among patients treated with t-PA when this analysis was performed according to the 2 different t-PA maintenance regiments. Although the luminal area of the culprit vessel tended to be larger on hospital day 10 in patients treated with t-PA than with placebo in both groups A and B, this difference also did not achieve statistical significance. There was no difference in reference diameters between patients treated with placebo and t-PA, nor in the anatomic distribution of culprit vessels. The ulcerative index was similar in placebo and t-PA subsets in groups A and B, and there was no change in the ulcerate index over time in either placebo or t-PA patients in group A. The luminal areas of the 2 patients in the placebo group and the 2 in the t-PA group who reoccluded after early study in group A were 0.53, 0.15, 0.37 and [1.07 mm.sup.2], respectively. [TABULAR DATA I OMITTED] DISCUSSION These data provide new insights into the natural history of the infarct-related coronary artery of patients presenting with acute myocardial infarction and ST-segment elevation. Although t-PA produced marked and rapid improvement in overall patency (as defined by TIMI grade 2 or 3 perfusion) early after infarction as compared with heparin alone, this difference was attenuated at 10 days because of spontaneous recanalization in the placebo group. Although there was no statistically significant difference in the area of the infarct vessel lumen either early or later in the hospital course in the 2 treatment groups of patients who had The data on ulceration index indicate no difference in infarct-vessel morphology irrespective of whether patients received t-PA or placebo. The data contrast with the recent report of Davies et al,[4] which suggested substantial improvement in a modified ulceration index in patients with myocardial infarction treated with streptokinase and maintained on heparin during their hospital course. Davies calculated an ulceration index using the maximal intralesional diameter divided by the minimal diameter, as opposed to use of the second most severe diameter within the lesion, as in the present report. Hence, the Davies index is closely linked to the degree of stenosis and would be expected to decrease with time because of lessening of the maximal lesion stenosis. Difference in thrombolytic agents, the timing of sequential angiography, and other protocol features may also explain the disparity between their findings and ours. The similarity to group A lumen areas in group B patients who were studied only at hospital discharge gives further support to the substantial remodeling and increase in lumen area, which occur in the culprit infarct-related artery after the presentation. Future studies of the efficacy of new thrombolytic agents should consider the substantial improvement we found in the luminal areas of patients receiving heparin alone. Acknowledgment: It is a pleasure to acknowledge the technical assistance of Lois Adams, and the secretarial assistance of Linda Gray. [1] White CW. Recurrent ischemic events after successful thrombolysis in acute myocardial infarction. The Achilles' heel of thrombolytic therapy. Circulation 1989;80:1482-1485. [2] Badger RS, Brown BG, Kennedy JW, Mathey D, Gallery CA, Bolson EL, Dodge HT. Usefulness of recanalization to luminal diameter of 0.6 millimeter or more with intracoronary streptokinase during acute myocardial infarction in predicting 'normal' perfusion status, continued arterial patency and survival at one year. Am J Cardiol 1987;59:519-522. [3] Ellis SG, Topol EJ, George BS, Kereiakes DJ, Debowey D, Sigmon KN, Pickel A, Lee KL, Califf RM. Recurrent ischemia without warning -- analysis of risk factors for in-hospital ischemic events following successful thrombolysis with intravenous tissue plasminogen activator. Circulation 1989;80:1159-1165. [4] Davies SW, Marchant B, Lyons JP, Timmis AD, Rothman MT, Layton CA, Balcon R. Coronary lesion morphology in acute myocardial infarction: demonstration of early remodeling after streptokinase treatment. J Am Coll Cardiol 1990; 16:1079-1986. [5] Armstrong PW, Baigrie RS, Daly PA, Haq A. Gent M, Robert RS, Freeman MR, Burns R, Liu P, Morgan CD. Tissue plasminogen activator. Toronto (TPAT) placebo-controlled randomized trial in acute myocardial infarction. J Am Coll Cardiol 1989;13:1469-1476. [6] Kalbfleish J, Thadani U, LittleJohn JK, Brown G, Magorien R, Kutcher M, Taylor G, Maddox WT, Campbell WB, Perry J, Spann JF, Vetrovec G. Kent RS, Armstrong PW, Evaluation of a prolonged infusion of recombinant tissue-type plasminogen activator (duteplase) in preventing reocclusion following successful thrombolysis in acute myocardial infarction. Am J Cardiol 1992;69:1120-1127. [7] Wilson RF, Holida MD, White CW. Quantitative angiographic morphology of coronary stenoses leading to myocardial infarction or unstable angina. Circulation 1986;73:286-293. [8] TIMI Study Group. The Thrombolysis in Myocardial Infarction (TIMI) trial: phase 1 findings. N Engl J Med 1985;312:923-296. [9] Reiber JHC, Serruys PW, Koodman CJ, Wijns W, Slager CH, Gerbrands, JJ, Schuurbiers JCH, Den Boer A, Hugenholtz P. Assessment of short-, medium- and long-term variations in arterial dimensions from computer-assisted quantitation of coronary cineangiograms. Circulation 1985;71:280-288. [10] Langer A, Wilson RF, Comparison of manual versus automated edge detection for determining degrees of luminal narrowing by quantitative coronary angiography. Am J Cardiol 1990;67:885-888. [11] Van Lierde J, DeGeest H, Verstraete M, Van de Werf F. Angiographic assessment of the infarct-related residual coronary stenosis after spontaneous or therapeutic thrombolysis. J Am Coll Cardiol 1990;16:1545-1549. [12] Morgan CD, Roberts RS, Haq A, Baigrie RS, Daly PA, Gent M, Armstrong PW. Results from the Tissue Plasminogen Activator: Toronto (TPAT) placebo-controlled trial. Coronary patency, infarct size and left ventricular function following thrombolytic therapy for acute myocardial infarction. J Am Coll Cardiol 1991; 17:1451-1457. Amin Haq, MD, Christopher D. Morgan, MD, Robert F. Wilson MD, Paul A. Daly, MD, Robert S. Baigrie, MD, Carl W. White, MD, Robert Roberts, MTech, Michael Gent, MSc, and Paul W. Armstrong, MD, and the Toronto Tissue Plasminogen Activator Trial Study Group From St. Michael's Hospital; the Western and General Divisions, The Toronto Hospital; Sunnybrook Health Science Centre; and University of Toronto, Toronto, Ontario, Canada; the Cardiovascular Division, Department of Medicine, University of Minnesota, Minneapolis, Minnesota; and the Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada. This study was supported by Burroughs Wellcome Co., Research Triangle Park, North Carolina.