View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

58 JACC Vol . 14, No . I July 18 58-64

Late Thrombolytic Therapy Preserves Left Ventricular Function in Patients With Collateralized Total Coronary Occlusion Primary End Point Findings of the Second Mount Sinai- University Reperfusion Trial

K. PETER RENTROP, MD, FACC,* FREDERICK FELT, MD, FACC,t WARREN SHERMAN, MD, FACC,* PETER STECY, MD, FACC,t SUSAN HOSAT, MS,* MARC COHEN, MD, FACC,* MARIANO REY, MD, FACC,t JOHN AMBROSE, MD, FACC,* MARK NACHAMIE, MD,t WILLIAM SCHWARTZ, MD, FACC,* WILLIAM COLE, MD,t ROBERT PERDONCIN, MD,* JOHN C . THORNTON, PHD New York, New York

The change in left ventricular ejection fraction from prein- icant improvement over those without collateral flow in the tervention to predischarge was prospectively assessed in streptokinase (5 .4 ± 2.5%) and streptokinase-nitroglycerin 33 patients with acute . Within 12 h (10 .6 ± 2 .7 %) arms, but not in the nitroglycerin arm . Time of symptom onset (mean 6.3 ± 2 .7 h), patients were to treatment did not influence the change in ejection randomly assigned to a double-blind intracoronary infusion fraction . In patients with initial subtotal occlusion, throm- of streptokinase, nitroglycerin, both streptokinase and ni- bolytic therapy was of no short-term benefit because ejec- troglycerin or conventional therapy without acute cardiac tion fraction increased by 6% in all three intervention catheterization . Treatment effects were also assessed in arms. prospectively defined angiographic subsets . These findings indicate that relatively late thrombolytic There was a significant interaction between streptoki- therapy results in significant myocardial salvage in those nase and nitroglycerin (p < 0 .01), resulting in an increase patients with collateralized total coronary occlusion . This in ejection fraction of 3 . percentage units in the combined benefit is potentiated by concomitant nitroglycerin therapy . treatment arm (p < 0 .001) . Patients with collateral flow to (J Am Coll Cardiol 18 ;1458-64) a totally obstructed infarct-related artery showed a signif-

Thrombolytic therapy is the treatment of choice for patients in experimental animals (4,5), it is still undefined in human with acute myocardial infarction who present within 3 h of patients . Several studies (6-8) suggest that the efficacy of pain onset (1-3) . Whereas the time limit for benefit from thrombolytic therapy, as assessed by the change in left interventions in evolving infarction has been clearly defined ventricular function, is dependent on initial angiographic status . Furthermore, it has been suggested (,10) that the benefit of thrombolytic therapy can be enhanced by adjuvant From the *Departments of Medicine, Mount Sinai School of Medicine and tNew York University School of Medicine, New York, New York . This study therapy . Therefore, we performed a prospective randomized was supported by the Cardiac Diseases Branch, Division of Heart and double-blind trial to assess the effects of intracoronary Vascular Diseases, National Heart, Lung, and Blood Institute, National streptokinase, intracoronary nitroglycerin or their combina- Institutes of Health, Bethesda, Maryland (Grant R01HL28843) . It was pre- sented in part at the 35th Annual Scientific Session of the American College of tion, administered within 12 h of the onset of chest pain . The Cardiology, March 186, Atlanta, Georgia and the 5th Scientific Session of main end point was change in ejection fraction in angio- the American Heart Association, November 186, Dallas, Texas . The strep- tokinase and the placebo were donated by Kabi Vitrum, Stockholm, Sweden . graphic subsets defined prospectively on the basis of results Manuscript received October 17, 188 ; revised manuscript received of the Registry of the European Society of Cardiology (6) January 18, 18, accepted February 21, 18 . and the First Mount Sinai-New York University Reperfusion Address for reprints K . Peter Rentrop, MD, St . Vincent's Hospital and Medical Center of New York, 153 West 11th Street . New York, New York Trial (). Patients were enrolled in three 10011 . hospitals and followed up for a minimum of 2 years .

©18 by the American College of Cardiology 0735-107/8/$3 .50

JACC Vol . 14, No . I RENTROP ET AL . 5 July 18 58-64 LATE THROMBOLYSIS AND COLLATERAL FLOW

Methods laboratory in a side by side fashion ; the core laboratory was unaware which study represented the pretreatment or end Study design and patients . A two by two factorial design point nuclear ventriculogram . Change in ejection fraction was selected to detect any interaction between intracoronary was expressed in ejection fraction percentage units . Repro- infusions of streptokinase and nitroglycerin . We prospec- ducibility of the core laboratory's analyses was assessed by tively assessed the influence of baseline angiographic status recycling 10% of the raw data . Intraclass correlation coeffi- in the patients in the intervention arms of the study as cients (13) in the recycled studies yielded reliabilities of 0 .6 follows total versus subtotal occlusion of the infarct-related for both preintervention and end point ejection fraction . artery, and presence or absence of collateral flow to the Coronary angiography . The infarct-related coronary ar- infarct vessel (6) . The study protocol was approved by the tery was classified as totally obstructed if there was com- committees on human research in the participating institu- plete cessation of anterograde flow, or as patent if there was tions . any degree of anterograde filling beyond the culprit lesion . All patients with suspected acute myocardial infarction Collateral vessels were determined to be present if any were screened in the participating hospitals . Patients <78 segment of the infarct-related artery filled in other than a years of age were eligible if they presented within 12 h of the continuous anterograde fashion . The extent of collateral onset of myocardial ischemic pain of ?30 min duration and filling was not further graded . demonstrated either a new ST segment elevation or depres- Electrocardiograms . The sum of R wave height and the sion over 0 .1 mV persisting 0 .08 s after the J point or number of pathologic Q waves were assessed from precor- pathologic Q waves in two contiguous leads of the screening dial leads V, to V6 in patients who had anterior wall electrocardiogram (ECG) . Exclusion criteria have been pub- infarction, and leads II, III and aVF in patients who had lished elsewhere () . inferior wall infarction . Treatment arms . After signing informed consent, pa- Statistical analysis . This trial was analyzed by intention tients were prospectively stratified according to duration of to treat . Continuous variables were expressed as mean pain ; Group A <2 h, Group B 2 to 12 h . They were values ± SD, except where SEM was used . Differences in randomly assigned to one of four treatment arms 1) intra- the distribution of baseline variables among the four treat- coronary infusion of streptokinase, 2,000 U/min ; 2) intracor- ment arms were compared using analysis of variance for onary infusion of nitroglycerin, 0 .01 mg/min ; 3) combined continuous variables and chi-square analysis for categorical infusion of streptokinase, 2,000 U/min and nitroglycerin, variables . 0.01 mg/min ; or 4) a control group receiving conventional The null hypothesis that the average change in ejection therapy without acute catheterization . A radionuclide ven- fraction was the same for the four treatment arms was tested triculogram at rest and a standard 12 lead ECG were by analysis of covariance . All changes in ejection fraction obtained before . In the three inter- reported are adjusted changes derived from the analysis of vention arms, aortic pressures were recorded before dye covariance (14) . The following were considered as possible injection . Intracoronary drug infusions were double-blind covariates age, preintervention ejection fraction, infarct and were maintained for a minimum of 75 min . If recanali- location, prior myocardial infarction, preintervention use of zation occurred, the intracoronary infusion was continued beta-blockers and duration of pain prior to infusion . In for at least an additional 30 min to a maximum of 120 min . addition, the number of diseased vessels was considered as Concomitant therapy . All patients underwent full antico- a possible covariate in the intervention groups . agulant therapy with intravenous heparin followed by war- In the mortality analysis, age, preintervention heart rate farin for 3 months . Nitrates, calcium channel antagonists and and ejection fraction were used as covariates . beta-adrenergic blockers were not administered routinely in the first 72 h . Percutaneous transluminal coronary angio- plasty or coronary artery bypass surgery was performed in Results 5% of the study patients before end point data acquisition . Patient profile. Of the 3,62 patients with suspected End point radionuclide ventriculograms and a 12 lead ECG myocardial infarction who were screened, 551 met the were obtained on day 10 to 14 . eligibility criteria . One hundred fifty-eight eligible patients Technical aspects of data acquisition and evaluation . were not enrolled, primarily because they refused to partic- Radionuclide ventriculography . Red blood cells labeled with ipate (n = 78) or because a complete study team was not 30 mCi of technetium-m were injected using a modified in available (n = 54) . Thus, the study group comprised 33 vivo technique (11) . Ejection fraction was calculated from patients ; 7 were assigned to the streptokinase arm, to the the "best septal" left anterior oblique view by a semiauto- nitroglycerin arm, 6 to the streptokinase-nitroglycerin arm mated various region of interest count method, applying a and 101 to the control arm . The intracoronary infusion was background correction (12) . The preintervention and end incomplete as a result of death or profound hemodynamic point studies of each patient were analyzed by the core instability in 18 of the 22 patients in the intervention groups .

60 RENTROP ET AL . JACC Vol . 14, No . I LATE THROMBOLYSIS AND COLLATERAL FLOW July 18 58-64

Table 1 . Baseline Variables and Duration and Dosage of Acute response to intracoronary interventions . The acute Intracoronary Therapy in the 33 Study Patients recanalization rate of totally occluded infarct-related arteries Variables was 60% (40 of 67) in the streptokinase arm, 63% (3 of 62) in the streptokinase-nitroglycerin arm (p = NS) and 8% (5 of Age (yr) 57 ± 10 Men (%) 77 65) in the nitroglycerin arm (p < 0 .01) . Systolic blood Prior myocardial infarction (%) 15 pressure decreased during the study drug infusion in the Current infarct location (%) nitroglycerin arm by 12 .1 ± 20 .2 mm Hg, in the streptoki- 47 Anterior nase-nitroglycerin arm by 11 .4 ± 1 .8 mm Hg and in the Inferior 47 streptokinase arm by 4 .8 ± 15 .6 mm Hg . The decrease was Lateral 6 greater in those patients treated with nitroglycerin alone or Infarct-related coronary artery (%) Left anterior descending 4 in combination with streptokinase than in patients treated Left circumflex 16 with streptokinase alone (p < 0 .05) . Right 35 Primary end point data . Paired gated blood pool scans 1 Left main suitable for analysis were available in 272 patients . The main Extent of * (%) reasons for missing data were insufficient quality of at least Single vessel 48 one of the studies (n = 45) or death before end point Double vessel 36 Triple vessel 16 measurements (n = 34) . There were no significant differ- Time to infusion* (h) 6 .3 ± 2 .7 ences in the baseline variables listed in Table 1 between the Duration of infusion (min) 85 .3 ± 24 .5 272 patients with paired gated blood pool studies and the 121 173' ± 53' Total dose of streptokinase* (U) patients with incomplete ejection fraction data, except for Total dose of nitroglycerin* (mg) 0 .85 ± 0 .22 gender distribution . The prevalence of female gender was Intraaortic systolic pressure* (mm Hg) 133 ± 2 Ejection fraction (%) 46.6 ± 14 .8 significantly greater among patients with incomplete ejection fraction data (30% [36 of 121] versus 20% [54 of 272], *Values obtained only in the three intervention arms . ± = mean respectively) (p < 0.03) . The missing data points were standard deviation . evenly distributed among the treatment arms . Baseline ejec- tion fraction values were not significantly different (Table 2) . There were no statistically significant differences in any There was a significant correlation of preintervention of the baseline variables among the four treatment arms ejection fraction with change in ejection fraction . None of (Table 1). The time interval from the onset of pain to the the other baseline variables considered as covariates added beginning of intracoronary infusion ranged from 1 .8 to 13 .6 significantly to the prediction of change in ejection fraction, h. The prevalence of initial subtotal obstruction of the including time, whether considered as the interval between infarct-related artery in patients who underwent interven- pain onset and infusion or by group (A [<2 h] versus B [2 to tional angiography was 27% (73 of 272) ; collaterals vessels to 12 h]) . the infarct-related artery were seen before intervention in In the model used to assess the significance of the 33% of patients (66 of 1) with complete obstruction of the changes in ejection fraction, all changes in ejection fraction infarct-related vessel and in 11% (8 of 72) of those with were adjusted for the preintervention value (Table 2) (14) . subtotal obstruction (p < 0 .001) . Myocardial infarction was There was a significant positive streptokinase-nitroglycerin confirmed by an increase in serum creatine kinase to an interaction (p = 0 .01) (Fig . 1) . The only significant improve- abnormal level in 1% of the patients (358 of 33) . ment in ejection fraction from preintervention to end point

Table 2 . Ejection Fraction Before Intervention and at Study End Point and Change in Ejection Fraction From Preintervention to End Point Study Adjusted for the Preintervention Value by Treatment Arm*

Adjusted Preintervention End Point Change

Arm No. Mean SD Mean SD Mean SD p Value

SK 6 43 .0 11 .8 44 .4 12 .6 0 .7 1 .2 0 .5 SK-NTG 64 48 .5 15 .3 51 . 16 .2 3 . 1 .2 <0 .001 NTG 72 47 .2 15 .5 45 .3 14 .5 -1 .7 1 .1 0.1 Control 67 46 .4 15 . 47 .5 16 .3 1 .1 1 .2 0.3

*The change in ejection fraction is expressed in percentage units . Change in ejection fraction was adjusted for preintervention value (14) . NTG = intracoronary nitroglycerin ; SK = intracoronary streptokinase ; SK-NTG = intracoronary streptokinase plus nitroglycerin .

JACC Vol . 14, No . I RENTROP ET AL . 61 July 1858-64 LATE THROMBOLYSIS AND COLLATERAL FLOW

SK - NTG

NTG

SK No SK TREATMENT ARM SK SK - NTG NTG

Figure 1 . Change (A) in ejection fraction (expressed as the change Figure 2 . Change (A) in ejection fraction (expressed as the change in percentage units) from preintervention to end point study ad- in percentage units) from preintervention to end point study by justed for preintervention ejection fraction in the four treatment treatment arm in patients with total coronary occlusion of the arms . C = control (no intracoronary therapy) ; NTG = intracoronary infarct-related artery before therapy . Closed bars = collateral chan- nitroglycerin ; SK = intracoronary streptokinase ; SK-NTG = intra- nels present ; striped bars = no collateral channels . Abbreviations as coronary streptokinase plus nitroglycerin . in Figure 1 .

occurred in the streptokinase-nitroglycerin arm (p < 0 .001) arm versus nitroglycerin arm, 13 .6 ± 3 .1 percentage units (Table 2) . The only significant difference in the change in (p < 0 .001) . ejection fraction among treatment arms was between the Noncollateralized total coronary occlusion was associ- streptokinase-nitroglycerin arm and the nitroglycerin arm ated with a significant decrease in ejection fraction in both (p < 0 .001) . the streptokinase and nitroglycerin arms, but not in the In patients assigned to the intervention groups, treatment streptokinase-nitroglycerin arm . effects were assessed by three pretreatment angiographic Within each treatment arm, the effect of collateral status subsets 1) total occlusion of the infarct-related artery with on the change in ejection fraction was evaluated in patients observed collateral flow (n = 52) ; 2) total occlusion of the with total coronary occlusion (Fig. 2) . In the streptokinase- infarct-related artery without collateral flow (n = 0) ; and 3) nitroglycerin arm, the subset with collateral flow showed a subtotal obstruction of the infarct-related artery without net ejection fraction improvement by 10 .6 ± 2 .7 percentage collateral flow (n = 48) . The group of patients with collater- units over the subset without collateral flow (p < 0 .001) alized subtotal obstruction (n = 8) was too small for analysis . (Table 3) . In the streptokinase arm, ejection fraction im- Among patients with collateralized total coronary occlu- proved by 5 .4 ± 2 .5 percentage units in patients with sion, only those in the streptokinase-nitroglycerin arm collateral flow over those without (p = 0 .03) . In the nitro- showed a significant improvement in ejection fraction of glycerin arm, there was no significant collateral effect . .8 ± 2 .1 percentage units (p < 0 .001) (Table 3, Fig . 2) . In Subtotal obstruction of the infarct-related vessel was this angiographic subset, the differences in the change in associated with a significant improvement in ejection frac- ejection fraction between the streptokinase-nitroglycerin tion of about 6% in all three intervention arms ; there were no arm and the other intervention arms were significant ; strep- significant differences among treatment arms (Table 3) . tokinase-nitroglycerin arm versus streptokinase arm, 8 .1 ± Secondary end point data . Twelve lead electrocardio- 2 .8 percentage units (p < 0 .005) ; streptokinase-nitroglycerin grams . Patients with anterior wall infarction in the streptoki-

Table 3. Change in Ejection Fraction (in percentage units) From Preintervention to End Point Study by Treatment Arm and Angiographic Subset

Infarct SK-NTG SK NTG Artery Collateral Obstruction Vessel No. Mean SEM No. Mean SEM No. Mean SEM

Total Yes 18 .8 2.1 t 20 1 .7 1. 14 -3 .8 2.3 Total No 25 -0.8 1 .7 30 -3.7 1 .6* 35 -5.0 1 .5# Subtotal No 17 6.0 2.1t 16 6.6 2.2t 15 5.7 2.2t Ejection fraction change is significantly different from zero with the following p values *<0.05, t<0 .01, t<0 .001 . Abbreviations as in Table 2.

62 RENTROP ET AL . JACC Vol . 14, No . I LATE THROMBOLYSIS AND COLLATERAL FLOW July 1858-64

Table 4. Electrocardiographic Change From Preintervention to unlikely to influence the change in ejection fraction . The End Point Study in Patients With Anterior Wall Infarction by intracoronary rather than the intravenous route of streptoki- Treatment Arm nase infusion was selected because of its higher recanaliza- Preintervention Change tion rates and its requisite angiographic documentation of

Arm n Mean SD Mean SEM baseline coronary anatomy and recanalization (17) . Recom- binant tissue-type plasminogen activator (rt-PA) was not A . Sum of R Wave Amplitude, Leads V, to V (mm) 6 available at the time of our study ; however, we achieved SK 40 28 21 -8 . 2 .1 , remarkably similar recanalization rates (18) . The requisite SK-NTG 33 2 18 -4 .2 2 .3 preintervention angiography in the present trial provides NTG 36 30 21 -10 . 2 .2t insights into the pathophysiology of acute myocardial infarc- Control 3 24 1 -10 .0 2 .1 t tion, which can no longer be obtained . Furthermore, our B . Number of Pathologic Q Waves . Leads V, to V 6 (mm) findings lead to hypotheses about the therapeutic potential of SK 40 1 .6 1 .6 1 .0 0 .3 second generation thrombolytic agents administered >3 h SK-NTG 33 2 .0 1.7 0.4 0 .3 after symptom onset . NTG 37 1 . 2 .1 0. 0 .3* Streptokinase-nitroglycerin interaction . The positive in- Control 40 2 .4 1 . 1 .0 0 .3t teraction between streptokinase and nitroglycerin that re- Change is significantly different from zero with the following p values sulted in significant improvement in ejection fraction in *<0 .001, t<0 .0001 . Abbreviations as in Table 2 . patients assigned to combined therapy is a major new finding of this study . The ECG data support this finding salvage of myocardium as evidenced by greater preservation of R nase-nitroglycerin arm showed beneficial treatment effects waves and less development of Q waves occurred only in because they experienced neither the significant loss of R those patients with anterior wall infarction who were as- wave sum nor the increase in the number of Q waves from signed to combined therapy . baseline to end point studies that occurred in the other There are at least two possible mechanisms for the treatment arms (Table 4) . With inferior wall infarction, no beneficial effect of the addition of nitroglycerin to throm- treatment effects were found ; all treatment arms demon- bolytic therapy . Although the dose of nitroglycerin infused strated a significant decrease in R wave sum and increase in into the infarct-related artery was low, there was a systemic the number of pathologic Q waves . effect, with a significant decrease in systolic blood pressure Mortality. Mortality did not differ significantly among and thus, in myocardial oxygen demand. Furthermore, it has treatment arms at end point study (streptokinase, 3 of 7 ; been established that nitroglycerin can increase collateral streptokinase-nitroglycerin, 3 of 6 ; nitroglycerin, 5 of ; flow (1-21) . Such hemodynamic and collateral flow effects control, 5 of 101) or at 2 years after enrollment (streptokinase, could have decreased the rate of myocardial necrosis, result- 12 of 6 ; streptokinase-nitroglycerin, 16 of 5 ; nitroglycerin, ing in greater myocardial salvage when reperfusion was 14 of 7; control, 1 of ; 6 patients lost to follow-up study) . effected by thrombolytic therapy . The % (15 of 158 patients) in-hospital mortality rate for Angiographic subsets. Our results indicate that treatment patients who were eligible but not randomized did not differ effects are highly dependent on prospectively defined angio- significantly from that observed in randomized patients . graphic baseline variables . In patients with noncollateralized total coronary occlusion, there was a lack of improvement in ejection fraction regardless of treatment assignment, indicat- Discussion ing that in this angiographic subset, meaningful myocardial Trial design . In this trial, therapy was delayed l to 2 h to salvage is not possible if thrombolytic therapy is initiated gather nuclear and angiographic baseline data . Conse- >3 h after coronary closure. This finding parallels data from quently, the number of patients treated within <3 h of the animal model (4,5) in which reperfusion >3 h after abrupt symptom onset is too small to analyze the effects of early coronary closure did not result in significant myocardial thrombolytic therapy . Other investigators (7) utilized early salvage. postintervention left ventriculograms in lieu of preinterven- Patients with collateralized total coronary occlusion had tion studies . However, we previously demonstrated (15) that the greatest potential for benefit . There was a significant improvement in left ventricular function can occur immedi- increase of .8 percentage units in ejection fraction with ately after reperfusion . Lack of paired ejection fraction data assignment to the streptokinase-nitroglycerin arm, suggest- in almost one-third of the randomized patients, which is ing that collateral flow extends the "time window" for common in trials of comparable design (16), could have thrombolytic therapy, particularly when nitroglycerin is ad- influenced the observed changes in ejection fraction . How- ministered concomitantly . The dramatic difference between ever, patients with complete and incomplete ejection frac- the observed increase in ejection fraction in the streptoki- tion data differed only in gender distribution, a difference nase-nitroglycerin arm and the decrease of 3 .8 percentage

JACC Vol . 14, No . 1 RENTROP ET AL . 63 July 18 58-64 LATE THROMBOLYSIS AND COLLATERAL FLOW

units in the nitroglycerin arm indicates that the favorable Appendix effects of nitroglycerin on the rate of necrosis are not of The Second Mount Sinai-New York University Reperfusion Trial in- value unless timely reperfusion is achieved . cluded the following participants The improvement in left ventricular ejection fraction seen Steering Committee K. Peter Rentrop, MD (Chairman), Frederick Feit, in patients with collateral flow and in those with initial MD (Secretary), Warren Sherman, MD, Arthur Fox, MD, Richard Gorlin, MD, Kenneth McKusick, MD, Arlie Cameron, MD, Stanley Goldsmith . MD, subtotal obstruction indicates functional recovery of stunned Mariano Rey, MD . myocardium in both angiographic subsets . On the basis of Clinical Units, Principal Investigators Mount Sinai Medical Center, New retrospective studies, it has been suggested (6,8) that resid- York, New York K . Peter Rentrop, MD (Principal Investigator), Marc Cohen, MD. John Ambrose, MD, Heinrich Blanke, MD, Barry Cohen, MD, Mark ual flow through a subtotal lesion extends the time window Milner, MD, Jeffrey Cowen, MD, Rohit Arora, MD, Stephen Winters, MD, for myocardial salvage by thrombolytic therapy in a similar Robert Phillips, MD, Robert Reichstein, MD, Robert Levine, MD, Rao way as does residual flow through collateral channels . How- Betina, MD, Susan Hosat, MS ; Bellevue Hospital, New York, New York Frederick Feit, MD (Principal Investigator), Peter Stecy, MD, Mariano Rey, ever, improvement in left ventricular function has also been MD, Mark Nachamie, MD, William Cole, MD, Frank Politzer, MD, Jeffrey observed in patients with subtotal obstruction who did not Kramer, MD, Mark Ehrich, MD, Gary Friedman, MD, Stephan Siegel, MD, Frank Prior, MD, Michael Attubato, MD, James Slater, MD, Aaron Gindea, receive thrombolytic therapy, albeit in a noncontrolled study MD, Sharon Napoli, RN, Susan Harty, RN, Bernardin, RN ; City (7) . Our study confirms this finding and shows that in this Hospital, Elmhurst, New York Warren Sherman, MD (Principal Investiga- angiographic subset, recovery of function is not augmented tor), Richard Schneider, MD (former Principal Investigator), Robert Perdon- cin, MD, William Schwartz, MD, Francis Lane, MD, Hee Lee, MD, R . Patel, by thrombolytic therapy . This finding is compatible with the MD, Naomi Hill, RN . hypothesis that patients with initial subtotal obstruction Data Coordinating Center John Thornton, PhD (Principal Investigator) . experienced spontaneous reperfusion before baseline angi- Ford Calhoun . PhD, Harry Smith Jr, PhD, Richard Fagerstrom, PhD, Jim Holt, MS, Lainy Berkowitz-Rudolph, RN, Jean Fisher, PhD, Mark Van ography, and that at the time of intervention, coronary flow Buskirk, MS, Gail McAvay, MS . at rest was no longer critically impaired by either fibrin- Radionuclide Ventriculography Core Laboratory Harvard University Kenneth McKusick, MD (Prin- containing thrombus or spasm (22) . School of Medicine, Boston, Massachusetts cipal Investigator), Tsunehiro Yasuda, MD . Conclusions . Although experimental (4,5) and some clin- Angiography Core Laboratory St . Luke's Hospital, New York, New ical data (1) suggest that the time limit for myocardial salvage York Arlie Cameron, MD (Principal Investigator), Harvey Kemp, MD . Electrocardiography Core Laboratory New York University Hospital by reperfusion is <6 h after coronary occlusion, improve- School of Medicine, New York, New York Mariano Rey, MD (Principal ment in ejection fraction has been demonstrated in patients Investigator) . treated up to 18 h after pain onset in observational studies Thallium Scintigraphy Core Laboratory Mount Sinai and New York University Schools of Medicine, New York, New York Stanley Goldsmith, (6,8,23) . Two large randomized studies (24,25) showed a MD (Principal Investigator), Joseph Sanger . MD, Stephen Horowitz, MD, survival benefit from thrombolytic therapy initiated up to Mary Farrell . 24 h after symptom onset . The mechanism for a benefit from Hematology Core Laboratory Mount Sinai Medical Center, New York, New York Jacob kand, MD (Principal Investigator) . late thrombolytic therapy has remained unclear. Our data Safety and Data Monitoring Committee William B . Hood, Jr, MD indicate that the presence of collateral flow to a totally (Chairman), Lewis C . Becker, MD, Paul Canner, PhD, Jay H . Cohn, MD . Max Halperin, PhD, Thomas Killip 111, MD, Victor J . Marder, MD, Elliot occluded coronary vessel slows the progression of myoc, , r- Rapaport, MD, Harmon Smith, PhD. Thomas L . Robertson, MD (Ex Officio dial necrosis . Because benefit from thrombolytic therapy aid Member) . not decrease between 3 and 12 h after symptom onset, myocardial necrosis may be incomplete for at least 12 h in patients with collateral flow . References Collateralized total coronary occlusion was seen in about I . Gruppo Italiano per lo Studio delta Streptochinasi nell' lnfarto Mio- 25% of the patients in both our trials (), whereas almost half cardico (GISSI) . Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction . Lancet 186 ;1 37-402 . of the patients had noncollateralized total occlusion . Thus, . Improved survival the mortality benefit of late thrombolytic therapy would be 2 . Simoons ML, van der Brand M, DeZwaan C, et al after early thrombolysis in acute myocardial infarction . Lancet 185 expected to be smaller than that of early treatment . Demon- 2 578-81 . stration of this mortality benefit would require a sample size 3 . Passamani E, Hodges M, Herman M, et al . The Thrombolysis in much larger than that in the present trial (25) . Myocardial Infarction (TIMI) phase II pilot study tissue plasminogen activator followed by percutaneous transluminal coronary angioplasty . J Augmentation of ejection fraction benefit by intracoro- Am Coll Cardiol 187 ;1051B-64B . nary nitroglycerin indicates that further evaluation of adju- 4 . Reimer KA, Lowe JE, Rasmussen MM, Jennings RB . The wavefront vant therapy with agents that decrease the rate of myocardial phenomenon of ischemic cell death . 1 . Myocardial infarct size vs duration necrosis is warranted . of coronary occlusion in dogs . Circulation 177 ;56 786-4. 5 . Costantini C, Corday E, Lang TW, et al . Revascularization after 3 hours of coronary arterial occlusion effects on regional cardiac metabolic We are indebted to Thomas Chalmers, MD for suggesting the two by two function and infarct size. Am J Cardiol 175 ;36368-84 . factorial trial design, as well as to the members of the Safety and Data . Changes in left ventricular ejection Monitoring Committee for their guidance in the conduct of the trial and their 6 . Rentrop P, Smith H, Painter L, Holt J fraction after intracoronary thrombolytic therapy . Circulation 183 ; review of the manuscript . 68(suppl 1) 1-55-60 .

04 RENTROP ET AL . JACC Vol . 14, No . 1 LATE THROMBOLYSIS AND COLLATERAL FLOW July 18 58-64

7 . Leiboff RH, Katz RJ, Wasserman AG . et al . A randomized, angiograph- 17 . Rentrop P . Thrombolytic therapy in patients with acute myocardial ically controlled trial of intracoronary streptokinase in acute myocardial infarction . Circulation 185 ;71 627-31 . infarction . Am J Cardiol 184 ;53 404-7 . 18 . Chesebro JH, Knatterud G, Roberts R, et al . Thrombolysis in Myocardial 8 . Rogers WJ, Hood WP Jr, Mantle JA, et al . Return of left ventricular Infarction (TIMI) trial, phase 1 a comparison between intravenous tissue function after reperfusion in patients with myocardial infarction ; impor- plasminogen activator and intravenous streptokinase . Circulation 187 ; tance of subtotal stenoses or intact collaterals . Circulation 184 76 142-54 . 6338-4. 1 . Fam WM, McGregor M . Effect of coronary vasodilator drugs on retro- . Rentrop P. Feit F, Blanke H, et al . Effects of intracoronary streptokinase grade flow in areas of chronic myocardial ischemia . Circ Res 164 and intracoronary nitroglycerin infusion on coronary angiographic pat- 15 355-65 . terns and mortality in patients with acute myocardial infarction . N Engl J 20. Goldstein RE, Stinson EB, Scherer JL, Seningen RP, Grehl TM, Epstein Med 184 311 1457-63 . SE . Intraoperative coronary collateral function in patients with coronary 10 . Hammerman H, Kloner RA, Briggs LL, Braunwald E . Enhancement of occlusive disease nitroglycerin responsiveness and angiographic corre- salvage of reperfused myocardium by early beta-adrenergic blockade lations . Circulation 174 ;428-308 . (timolol) . J Am Coll Cardiol 184 ;3 1438-43 . 21 . Cohen MV . Downey JM, Sonnenblick EH, Kirk ES . The effects of 11 . Callahan RJ, Froelich JW, McKusick KA, Leppo J, Strauss HW, A nitroglycerin on coronary collaterals and myocardial contractility . J Clin modified method for the in vivo labeling of red blood cells with tc-m Invest 173 52 2836-47 . concise communication . J Nucl Med 182 ;23 315-8 . 22 . Maseri A, L'Abbate A, Baroldi G, et al . Coronary vasospasm as a 12 . Burow RD, Strauss HW, Singleton R, et al . Analysis of left ventricular possible cause of myocardial infarction a conclusion derived from the function from multiple gated acquisition cardiac blood pool imaging study of "preinfarction" . N Engl J Med 178 ;21271-7 . comparison to contrast angiography . Circulation 177 561024-8. 23 . Smalling RW, Fuentes F, Matthews MW, et al . Sustained improvement in 13 . Shrout PE, Fleiss JL . Intraclass correlations uses in assessing rater left ventricular function and mortality by intracoronary streptokinase reliability . Psychol Bull 17 ;86 420-8 . administration during evolving myocardial infarction . Circulation 183 ; 68 131-8 . 14 . Ostle D, Mensing RW . Statistics in Research, 3rd ed . Ames, IA Iowa 24 . Yusuf S, Collins R, Peto R, et al . Intravenous and intracoronary fibrin- State University Press, 175 452-83 . olytic therapy in acute myocardial infarction overview of results on 15 . Rentrop P, Blanke H, Karsch KR. Effects of nonsurgical coronary mortality, reinfarction and side-effects from 33 randomized controlled reperfusion on the left ventricle in human subjects compared with trials . Eur Heart J 185 ;6556-85 . conventional treatment . Am J Cardiol 182 ;4 1-7 . 25 . ISIS-2 (Second International Study of Infarct Survival) Collaborative 16 . Milis Study Group . Hyaluronidase therapy for acute myocardial infarc- Group . Randomized trial of intravenous streptokinase, oral aspirin, both, tion results of a randomized, blinded, multicenter trial . Am J Cardiol or neither among 17,187 cases of suspected acute myocardial infarction 186 ;57 1236-43 . ISIS-2 . Lancet 188 2 34-60 .