Author + information
- Received October 28, 1998
- Revision received June 8, 1999
- Accepted July 19, 1999
- Published online November 1, 1999.
- C.Michael Gibson, MS, MD, FACC∗,*,
- Sabina Murphy, MPH∗,
- Ian B.A. Menown, MD, MRCP†,
- Rafael F. Sequeira, MD, FACC‡,
- Robert Greene, MD, FACC§,
- Frans Van de Werf, MD∥,
- Marc J. Schweiger, MD, FACC¶,
- Magdi Ghali, MD, FACC#,
- Martin J. Frey, MD, FACC∗∗,
- Kathryn A. Ryan, BS∗,
- Susan J. Marble, RN, MS∗,
- Robert P. Giugliano, MD, SM††,
- Elliott M. Antman, MD, FACC††,
- Christopher P. Cannon, MD, FACC††,
- Eugene Braunwald, MD, FACC††,
- for the TIMI Study Group
- ↵*Reprint requests and correspondence: C. Michael Gibson, Associate Chief of Cardiology, University of San Francisco, 3333 California Street, Suite 430, San Francisco, California 94118
This study evaluated the determinants of coronary blood flow following thrombolytic administration in a large cohort of patients.
Tighter residual stenoses following thrombolysis have been associated with slower coronary blood flow, but the independent contribution of other variables to delayed flow has not been fully explored.
The univariate and multivariate correlates of coronary blood flow at 90 min after thrombolytic administration were examined in a total of 2,195 patients from the Thrombolysis in Myocardial Infarction (TIMI) 4, 10A, 10B and 14 trials. The cineframes needed for dye to first reach distal landmarks (corrected TIMI frame count, CTFC) were counted as an index of coronary blood flow.
The following were validated as univariate predictors of delayed 90-min flow in two cohorts of patients: a greater percent diameter stenosis (p < 0.0001 for both cohorts), a decreased minimum lumen diameter (p = 0.0003, p = 0.0008), a greater percent of the culprit artery distal to the stenosis (p = 0.03, p = 0.02) and the presence of any of the following: delayed achievement of patency (i.e., between 60 and 90 min) (p < 0.0001 for both cohorts), a culprit location in the left coronary circulation (left anterior descending or circumflex) (p = 0.02, p < 0.0001), pulsatile flow (i.e., reversal of flow in systole, a marker of heightened microvascular resistance, p = 0.0003, p < 0.0001) and thrombus (p = 0.002, p = 0.03). Despite a minimal 16.4% residual stenosis following stent placement, the mean post-stent CTFC (25.8 ± 17.2, n = 181) remained significantly slower than normal (21.0 ± 3.1, n = 78, p = 0.02), and likewise 34% of patients did not achieve a CTFC within normal limits (i.e., <28 frames, the upper limit of the 95th percent confidence interval previously reported for normal flow). Those patients who failed to achieve normal CTFCs following stent placement had a higher mortality than did those patients who achieved normal flow (6/62 or 9.7% vs. 1/118 or 0.8%, p = 0.003).
Lumen geometry is not the sole determinant of coronary blood flow at 90 min following thrombolytic administration. Other variables such as the location of the culprit artery, the duration of patency, a pulsatile flow pattern and thrombus are also related to slower flow. Despite a minimal 16% residual stenosis, one-third of the patients treated with adjunctive stenting still have a persistent flow delay following thrombolysis, which carries a poor prognosis.
Coronary blood flow is a major determinant of clinical outcomes in patients with acute myocardial infarction (AMI) (1–7). To guide therapy better, further insight into the determinants of coronary blood flow is needed. The conventional Thrombolysis in Myocardial Infarction (TIMI) flow grade classification scheme (1)has been a valuable tool to compare angiographic and clinical outcomes following thrombolysis (2–7). This classification scheme is limited, however, by interobserver variability, its categorical nature, its limited statistical power, and the fact that nonculprit flow (used to gauge TIMI flow grade 3) is abnormally slow early in the course of AMI (8). Use of the Doppler velocity wire in a very large number of patients is limited by the need to instrument the artery in patients who may not require intervention.
Recently we described a new index of coronary blood flow called the corrected TIMI frame count (CTFC) in which the number of frames required for dye to first reach standardized distal landmarks is counted (8). Although tighter residual stenoses following thrombolysis have been associated with slower flow (9,10), the independent contribution of other angiographic, hemodynamic and clinical findings to delayed flow has not been fully evaluated. The goal of this study was to identify and validate the independent predictors of coronary blood flow following thrombolytic administration in over 2,000 patients, the largest series described to date by a single reader.
Corrected TIMI frame count data was pooled from the TIMI 4, 10A, 10B, and 14 trials. The TIMI 4 trial was a randomized double-blind comparison of three thrombolytic regimens (11): antistreplase (Eminase) versus front-loaded recombinant tissue plasminogen activator (rt-PA) (Activase or alteplase) versus combination therapy in 416 patients. The TIMI 10A trial was a nonrandomized, open-label, dose-escalation study of eight ascending doses of TNK (a mutant of recombinant tissue plasminogen activator) (5, 7.5, 10, 15, 20, 30, 40, 50 mg IV over 5 s) in 113 patients (12). TIMI 10B was an 880-patient randomized trial comparing 30, 40 and 50 mg of TNK versus front-loaded rt-PA (13). Data from 812 patients enrolled in the ongoing TIMI 14 trial comparing abciximab versus abciximab + a low-dose thrombolytic were also included (14). Angiography was performed at 60, 75 and 90 min following thrombolytic administration (8,11–13). Nitroglycerin (IV or sublingual) was administered every 15 min if the systolic blood pressure exceeded 110 mm Hg (8,11–13). Hemodynamics were evaluated at 90 min following thrombolysis in patients enrolled in the TIMI 14 trial. Few patients had a Fick cardiac output measured; therefore, the ventriculographic cardiac output was also estimated as stroke volume × heart rate at 90 min. The TIMI studies were approved by each participating center’s Institutional Review Board, and the trial was conducted according to the principles of the Declaration of Helsinki.
Angiographic analysis methods
The TIMI flow grade was assessed as previously defined at the TIMI Angiographic Core Laboratory (1). To evaluate objectively coronary flow as a continuous quantitative variable, the number of cineframes required for contrast to first reach standardized distal coronary landmarks in the infarct-related artery (the TIMI frame count) was measured using a frame counter on a cineviewer (8). Data presented here have been converted to and are reported using the most common cinefilming speed in the U.S.: 30 frames/s (15). All flow data was assessed by a single observer (CMG). A pulsatile flow pattern was defined as intermittent cessation of antegrade flow or frank flow reversal during systole. The optimal single-plane projection was selected that identified the stenosis in its greatest severity with minimal foreshortening or overlapping of branches, and end-diastolic frames were chosen for quantitative angiographic analysis using a previously described and validated automated edge-detection algorithm (16).
All analyses were performed using Stata version 5.0 (17). All continuous variable values are reported as the mean ± standard deviation. Analysis of variance (ANOVA), with a Bonferroni correction for multiple corrections or multiple linear regression, was utilized for the analysis of continuous variables. A p value <0.10 was required for retention in step-up linear regression models. The nonparametric Wilcoxon rank-sum test was used when the data were not normally distributed or when data were imputed to an occluded vessel. When appropriate, the chi-square test or logistic regression was used for analysis of categorical variables when appropriate. Patient identification numbers were randomized into two random cohorts, and the univariate and multivariate correlates were validated independently in the two cohorts.
Relationship of TIMI flow grades 2 and 3 to angiographic and hemodynamic variables
To gain further insight into the correlates of the conventional TIMI flow grades 2 and 3, their relationships to a variety of angiographic and hemodynamic variables were examined. The CTFC of patients with TIMI flow grade 2 was more than twice that of patients with TIMI flow grade 3 (Table 1), indicating a flow of less than half. Although no difference existed in the normal reference segment diameter between the two flow grades, the minimum lumen diameter of patients with TIMI flow grade 2 was smaller and the percent diameter stenosis was correspondingly tighter than in patients with TIMI flow grade 3 (Table 1). Patients with TIMI flow grade 2 more often received collaterals; they also had a higher proportion of left anterior descending artery (LAD) lesions, a greater incidence of both thrombus and pulsatile flow, a greater percent of the culprit artery distal to the stenosis compared with TIMI flow grade 3 but lower ejection fractions (Table 1). Although no difference was seen between TIMI flow grades 2 and 3 in the time from the onset of symptoms to treatment, a greater proportion of patients with TIMI flow grade 2 at 90 min were not open by 60 min (i.e., the duration of patency was <30 min, p < 0.0001) (Table 1). A lower systolic blood pressure was associated with arteries with TIMI flow grade 2, but no differences existed in diastolic pressures between flow grades (Table 1). Although the heart rate did not differ between the two flow grades, the ventriculographic cardiac output was lower in patients with TIMI flow grade 2 (Table 1). Patients with TIMI flow grade 2 tended to have higher left ventricular end-diastolic pressures and lower double products (90-min heart rate × 90-min blood pressure) (Table 1).
The following factors were validated as univariate predictors of TIMI flow grade 2 in two random cohorts of patients: smaller minimum lumen diameters, tighter percent diameter stenoses, failure to achieve patency early by 60 min, a pulsatile flow pattern, a greater percent of the culprit artery distal to the stenosis, the presence of a left-system culprit lesion and the 90-min ejection fraction. No difference existed between the two cohorts in height (TIMI flow grade 2 = 170.6 ± 9.4 cm [n = 256] vs. TIMI flow grade 3 = 171.1 ± 9.5 cm [n = 643], p = NS), weight (79.3 kg for both, n = 973), body surface area (3.76 m2for both, n = 899), age (TIMI flow grade 2 = 58.9 ± 11.8 years [n = 308], TIMI flow grade 3 = 58.7 ± 11.4 years [n = 744], p = NS) or gender (rate of TIMI flow grade 3 55.9% for men [n = 803] and 55.2% for women [n = 249]). There was a tendency for current smokers to have a higher rate of TIMI flow grade 3 (54.2% [n = 177] vs. 44.3% [n = 223; p = 0.054]).
Response of TIMI flow grades 2 and 3 at 90 min to adjunctive PTCA of the residual stenosis
To gain further insight into the relative contribution of the epicardial stenosis to the various TIMI flow grades, the response to adjunctive (for TIMI flow grades 2 and 3) and rescue percutaneous transluminal coronary angioplasty (PTCA) (for TIMI flow grades 0 and 1) was examined (Fig. 1, Table 2). Although both successful adjunctive and rescue PTCA largely normalized discrepancies in pre-PTCA CTFCs among the TIMI flow grades at 90 min, it is notable that neither adjunctive nor rescue PTCA restored a normal CTFC to TIMI flow grades 0/1 (p = 0.0005) or grade 2 (p = 0.0003) following PTCA (i.e., mean post-PTCA CTFCs were higher than a CTFC of 21.0 ± 3.1 [n = 78], the previously reported normal CTFC observed in the absence of acute MI ). In a paired analysis, adjunctive PTCA of the culprit artery restored a culprit CTFC (30.5 ± 23.1) that was the same as that in the nonculprit artery at 90 min (30.5 ± 13.6, paired n = 168, p = NS). However, both post-PTCA culprit and nonculprit flows were 45% slower than normal flow in the absence of AMI (21.0 ± 3.1 [n = 78] as previously defined , p < 0.0005). Although patients with TIMI flow grade 2 sustained a highly significant 47.0 ± 40.3 frame improvement in the CTFC following adjunctive PTCA (p < 0.0001), patients with TIMI flow grade 3 sustained a significant but smaller improvement in their CTFC (27.8 frames pre-PTCA to 23.1 frames post-PTCA, p = 0.02).
Placement of an intracoronary stent would be expected not only to relieve the residual stenosis but also to minimize the potential for intraluminal obstruction due to thrombus and dissection planes and also to relieve laminar flow disturbances (Fig. 1). In those patients in which both post-PTCA and post-stent frame counts were available, the CTFC improved by 4.1 ± 15.1 frames (from 26.6 ± 18.6 to 22.5 ± 11.8, p = 0.03, n = 63) following intracoronary stent placement in a paired analysis. Likewise, the rate of TIMI flow grade 3 improved from 80% post-PTCA to 88% following stent placement.
Despite a minimal 16.4 ± 16.1% residual stenosis following stent placement, the mean post-stent CTFC remained elevated, however, at 25.8 ± 17.2 (n = 181) frames (p = 0.02 vs. the previously defined normal value of 21.0 ± 3.1 for normal flow ), and 34% of patients did not achieve a CTFC within normal limits (i.e., <28 frames, the upper limit of the 95th percent confidence interval [CI] previously reported for normal flow). Those patients who failed to achieve a CTFC within normal limits following stent placement had a higher mortality than did those patients who achieved normal flow (1/118 or 0.8% vs. 6/62 or 9.7%, p = 0.003).
Relationship of the 90-min-corrected TIMI frame count to angiographic and hemodynamic variables
The relationship of a quantitative index of flow, the CTFC, to angiographic and hemodynamic variables was also evaluated. Tighter 90-min percent diameter stenoses and minimum lumen diameters were both univariate correlates of slower 90-min CTFCs (p < 0.001 for both) as was a longer length of the lesion (p = 0.006) (Table 3). An increased percent of the artery distal to the culprit stenosis (i.e., a more proximal lesion) was also associated with slower 90-min CTFCs (p = 0.001).
The presence of the following categorical variables was related to slower 90-min flow: failure to achieve patency (TIMI flow grades 2 or 3) by 60 min after thrombolytic administration (p < 0.0001), the presence of a pulsatile flow pattern (p < 0.0001), the location of the culprit artery in the left system (LAD or circumflex) (p < 0.0001), the presence of thrombus (p = 0.0002), and the presence of collaterals (p = 0.0003) (Table 3). The following variables were validated as predictors of slower CTFCs in the two patients cohorts: the 90-min percent stenosis (p < 0.0001 for both); the 90-min minimum lumen diameter (p = 0.0003 and p = 0.0008); the percent of the culprit artery distal to the stenosis (p = 0.03 and p = 0.02); the duration of patency (p < 0.0001 for both); the presence of pulsatile flow (p = 0.0003 and p < 0.0001); left-system infarct location (p = 0.02 and p < 0.0001); and the presence of thrombus (p = 0.002 and p = 0.03). There was a tendency for the following variables to be valid in the two separate random cohorts: the 90-min ejection fraction (p = 0.06 and p = 0.003); the presence of collaterals (p = 0.0002 and p = 0.17); and the ventriculographic cardiac output (p = 0.13 and p = 0.06). There were only modest trends for the CTFC to be related to weight (p = 0.10, n = 873), body surface area (p = 0.08, n = 803) and current smoking status (36.3 ± 15.7 [n = 98] vs. 41.4 ± 22.0 [n = 123], p = 0.056), none of which were valid in two separate cohorts.
Differences in outcomes among the three epicardial culprit arteries
The CTFC was slowest for LAD and circumflex culprit locations, and fastest for the right coronary artery (RCA) (p < 0.0001 vs. CTFC of left system) (Table 4). Likewise, the rate of TIMI flow grade 3 was lower in arteries supplying the left ventricle than in those supplying the right ventricle (p < 0.001). Minimum lumen diameters were smaller in the left system, but so were the normal reference segment diameters; consequently, percent diameter stenosis (which is independent of arterial diameter) did not differ between left and right systems. Left-system lesions were located significantly more proximally than in the right system (Fig. 2, p < 0.0001). Following PTCA, the CTFC was slower for left-system culprit lesions compared with RCA lesions (32.7 ± 26.3 [n = 117] vs. 25.5 ± 17.7 [n = 246], p = 0.002), despite the fact that there was no difference in percent diameter stenosis post-PTCA (41.7 ± 22.4%, n = 255 vs. 42.0 ± 20.0%, n = 292, p = NS).
Multivariable model of the 90-min CTFC
The final multivariable model for the 90-min CTFC was as follows: 90-min CTFC = 33.1 frames + 5.0 frames if a left-system lesion (p = 0.002) − 16.9 frames if open at 60 min (p < 0.001) + 0.21 frames × percent stenosis (p < 0.001) + 8.5 frames if pulsatile flow was present (p < 0.001) (overall model n = 672, r = 0.33). With the exception of left-system location, each independent multivariable remained a significant correlate of the 90-min CTFC when the model was tested in two independent, random cohorts.
In a model that was restricted to those patients with hemodynamic data available, systolic blood pressure (p = 0.001) and left ventricular end-diastolic pressure (p = 0.005) and ventriculographic cardiac output (p = 0.078) were also independent predictors of the 90-min CTFC in a model that corrected for percent stenosis, left-system location (p = 0.012) and pulsatile blood flow (p = 0.003). The model that included hemodynamic data had the highest r2of any model (r2= 0.50, r = 0.71, overall model p < 0.0001, n = 60).
Determinants of coronary blood flow at 90 min following thrombolytic administration
Previous studies have identified the presence of a tighter residual stenosis as a potential cause of TIMI flow grade 2 at 90 min following thrombolytic administration (9,10). The present study confirms the observation that increased residual stenosis severity is associated with TIMI flow grade 2 and higher TIMI frame counts. In addition to lumen geometry, however, the following were additional predictors of slower flow at 90 min: a shorter duration of patency (i.e., arteries closed at 60 min that opened by 90 min had slower 90-min frame counts than did those arteries that were open at 60 min); infarct location (LAD or left-system infarcts); more proximal infarcts with a greater percentage of the vessel distal to the stenosis; thrombus; and the presence of a pulsatile flow pattern (systolic flow reversal, a marker of heightened microvascular tone, but which may also be observed in other conditions such as hypertrophy). In those patients in whom hemodynamic data were available from TIMI 14, lower systolic arterial blood pressure, lower cardiac output and higher left ventricular end-diastolic pressures were also independently associated with slower flow. Current smoking was the only clinical variable that tended to be associated with more rapid flow, but smokers more often had RCA infarcts, and the infarct location rather than smoking status might account for the more rapid flow. The slower flow observed in arteries that opened more recently may reflect the fact that the transition from epicardial patency to myocardial reperfusion (or transient no/slow reflow) might also require more time with an increasing duration of ischemia.
The relief of the residual stenosis by conventional PTCA and the scaffolding provided by intracoronary stent placement present unique opportunities to examine the potential role of the residual stenosis and intraluminal obstruction to flow delays following thrombolysis. Both rescue and adjunctive PTCA of the residual stenosis at 90 to 120 min largely normalized discrepancies in pre-PTCA frame counts between the TIMI flow grades. In a paired analysis, adjunctive and rescue angioplasty restored flow in culprit vessels at a rate nearly identical to that of nonculprit arteries in the setting of AMI (30.5 frames each). These observations should not, however, be interpreted as demonstrating that PTCA completely restores “normal” flow. It is important to note that post-PTCA CTFCs and nonculprit CTFCs of 30 frames are both abnormally slowed, and they are approximately 45% slower (p < 0.001) than the CTFC of 21 frames previously reported in patients without AMI and normal flow (8). It is notable that the flow in nonculprit arteries was slower in those patients with TIMI flow grade 2 in the culprit artery than in those patients with an occluded culprit artery. The underlying cause of this association is unclear, but it may represent the impact of reperfusion injury (which had not occurred in the occluded arteries) or a local reflex that is activated in arteries that achieve patency.
Despite only a 16.4% residual diameter stenosis and the relief of intraluminal obstruction that would be anticipated following stent placement, flow was persistently slowed to nearly 26 frames (vs. 21 frames without myocardial infarction), and likewise 34% of stented vessels had abnormal flow with a CTFC ≥28 (the 95th percentile of the previously defined upper limit of normal). This persistent abnormality in flow is unlikely to be due to either the residual stenosis or intraluminal obstruction and may represent either the contribution of downstream microvascular resistance, or the slowing of flow due to hemodynamic compromise, a reflex or reduced oxygen consumption in infarcted or ischemic tissue. Whatever the mechanism mediating this persistent flow delay, those patients with persistently abnormal flow following relief of the stenosis by stenting did have a higher mortality.
Differences between the left and right coronary systems following successful thrombolysis
The CTFCs were slower and the rates of TIMI flow grade 3 were lower for the left system than the right system. Of note, the length of the culprit artery distal to left-system lesions was longer, and a greater percentage of the vessel lay distal to the culprit stenosis in left-system lesions (Fig. 2). It could be speculated that slower flow in the left system may be related to more extensive necrosis and increased myocardial edema as a result of the large myocardial mass subtended by the longer arteries. Perhaps another cause of slower flow in the left system is extravascular compression of intramyocardial arterioles during left ventricular systole (18–20). Coronary blood flow in the RCA is relatively independent of the phase of the cardiac cycle (21). The relatively thicker wall, the increased wall thickening during systolic contraction and the higher intracavitary pressures of the left ventricle may all produce higher intramyocardial pressure (22,23)than is observed in the thinner-walled right ventricle, which is also subjected to lower filling pressures.
Moreover, myocardial edema secondary to injury may further increase these systolic compressive forces acting upon arterioles in the left ventricular system. Indeed, the incidence of systolic flow reversal in this study was highest in LADs. Given these complex interrelationships among flow, infarct artery location and myocardium at risk, analyses comparing the clinical, enzymatic, ventriculographic or electrocardiographic outcomes of the various TIMI flow grades should correct for the fact that LAD and left-system locations are both associated with a higher incidence of TIMI flow grade 2 (8).
The TIMI frame counting technique was introduced when the TIMI 4 trial was one-third completed. Following the prospective utilization of the CTFC method in the TIMI 10A and 10B trials, the CTFC was evaluable in 97.3% of open arteries at 90 min (739/760). Hemodynamic data from the TIMI 4, 10A and 10B trials were not available, and assessment of the underlying effects of hemodynamic variables on culprit artery flow was based upon data from the TIMI 14 trial alone. Given the multiple variables that were analyzed in the present study, it is possible that some spurious results may have emerged. However, many of the p values in this study are below the 0.01 level, indicating that they are unlikely to have arisen by chance. The majority of the univariate correlates were validated in two random cohorts of the dataset, and the multivariable model of 90-min flow was validated in both random cohorts.
Recently we have demonstrated that when the force of injection is varied between the 10th and 90th percentiles for human injection rates, the CTFC may change by approximately two frames (24). Other investigators have recently reported that the TIMI frame counting method is very reproducible, with a coefficient of correlation of >0.95 between observers and differences between observers of <2 frames (25). French et al. (26–28)have likewise reported mean differences between observers of 0.75 frames. Although the TIMI frame count is a measure of time, it does not account for vessel length or volume, and it is therefore only an index of coronary velocity and flow. More direct measurements of flow determination through ancillary methods, such as positron emission tomography scanning, Doppler, or magnetic resonance imaging flow, are needed to verify these observations.
Aside from protocol-mandated nitrate administration, this study did not control for drug effect, and it is possible that other drugs (beta-blockers, calcium channel blockers, adenosine, sedatives) may have also modulated coronary flow globally. Preliminary pharmacologic data from Gregorini et al. (29)indicate that PTCA in AMI improves culprit flow from 47 to 38 frames, and stenting transiently improves the culprit CTFC to 23 frames. However, 15 min later the culprit artery CTFC slowed to 38 frames (29). Infusion of the alpha-adrenergic blocker phentolamine reaccelerated the CTFC to 17 frames (29). There was an association between culprit and nonculprit flows, and the investigators concluded that neural mechanisms may be responsible for the simultaneous slowing in both culprit and nonculprit vessels. Similar persistent slowing of culprit flow following stent placement has been reported by the PAMI (Primary Angioplasty in Myocardial Infarction) investigators, who have shown that patients with TIMI flow grade 3 before the stent placement have a CTFC of 26 following stent placement, and those patients with TIMI flow grade 2 before the stenting have a post-stent CTFC of 36 (30).
Lumen geometry is not the sole determinant of coronary blood flow at 90 min following thrombolysis. Other predictors of slowed 90-min flow include left coronary artery location, the late achievement of patency after 60 min, the presence of a pulsatile flow pattern, thrombus, a greater extent of myocardium subtended by the culprit artery and hemodynamic perturbations. Also, PTCA of the residual stenosis improves flow, but flow is persistently delayed to approximate that in the nonculprit artery. Following relief of the stenosis by stenting, approximately one-third of patients treated with thrombolysis still have a persistent flow delay of ≥28 frames. A persistent blood flow abnormality following relief of the stenosis by stenting carries a poor prognosis.
☆ Supported in part by a grant from Smith Kline Beecham, Philadelphia, Pennsylvania (TIMI 4); Genentech, South San Francisco, California (TIMI 10); Centocor and Eli Lilly Inc., Malvern, Pennsylvania, and Indianapolis, Indiana (TIMI 14).
- acute myocardial infarction
- analysis of variance
- corrected Thrombolysis in Myocardial Infarction frame count. (Number of frames required for dye to reach a standardized distal landmark)
- left anterior descending artery
- Primary Angioplasty in Myocardial Infarction
- percutaneous transluminal coronary angioplasty
- right coronary artery
- recombinant tissue plasminogen activator
- Thrombolysis in Myocardial Infarction
- a mutant of recombinant tissue plasminogen activator
- Received October 28, 1998.
- Revision received June 8, 1999.
- Accepted July 19, 1999.
- American College of Cardiology
- TIMI Study Group
- Vogt A.,
- Von Essen R.,
- Tebbe U.,
- et al.
- Karagounis L.,
- Sorensen S.G.,
- Menlove R.I.,
- et al.
- TEAM-3 Investigators,
- Anderson J.L.,
- Karagounis L.A.,
- Becker L.C.,
- et al.
- Gibson C.M.,
- Cannon C.P.,
- Daley W.L.,
- et al.
- GUSTO 1 Angiographic Investigators,
- Reiner J.S.,
- Lundergan C.F.,
- Fung A.,
- et al.
- TIMI 4 Investigators,
- Cannon C.P.,
- McCabe C.H.,
- Diver D.J.,
- et al.
- TIMI 10A Investigators,
- Cannon C.P.,
- McCabe C.H.,
- Gibson C.M.,
- et al.
- TIMI 10B Investigators,
- Cannon C.P.,
- Gibson C.M.,
- McCabe C.H.,
- et al.
- Thrombolysis in Myocardial Infarction (TIMI) 14 Investigators,
- Antman E.M.,
- Giugliano R.P.,
- Gibson C.M.,
- et al.
- Dotani I.,
- Dodge J.T.,
- Goel M.,
- et al.
- ↵StataCorp. Stata Statistical Software: Release 5.0. College Station, Texas. Stata Corporation, 1997.
- Bellamy R.F.
- Khouri E.M.,
- Gregg D.E.,
- Lowensohn H.S.
- Lowensohn H.S.,
- Khouri E.M.,
- Gregg D.E.,
- et al.
- Gallagher K.P.,
- Osakada G.,
- Matsuzaki M.,
- et al.
- Spaan J.A.E.,
- Breulas N.P.,
- Laird J.D.
- Ivanc T.B.,
- Crowe T.D.,
- Balazs E.M.,
- et al.
- HERO 1 Investigators,
- French J.K.,
- Straznicky I.T.,
- Webber B.J.,
- et al.
- Gregorini L.,
- Marco J.,
- Kozakova M.,
- et al.
- Abizaid A.S.,
- Popma J.J.,
- Grines C.L.,
- et al.