- Email: [email protected]
Multivariate associates of QT dispersion in patients with acute myocardial infarction: Primacy of patency status of the infarct-related artery
Multivariate associates of QT dispersion in patients with acute myocardial infarction: Primacy of patency status of the infarct-related artery Labros A. Karagounis, MD, Jeffrey L. Anderson, MD, Fidela L. Moreno, MD, and Sherman G. Sorensen, MD, for the TEAM-3 Investigators Salt Lake City, Utah
Background QT dispersion (QTd; QT interval maximum minus minimum) has been shown to reflect regional variations in ventricular repolarization and is increased in patients with life-threatening ventricular arrhythmias. Methods To determine correlates of QTd in patients who had had myocardial infarction (MI), 207 patients (158 men, aged 57 ± 11 years) with acute MI who were treated with alteplase or anistreplase within 2.7 ± 0.9 hours of symptom onset were studied. Angiograms at a median of 27 hours after thrombolysis showed reperfusion (Thrombolysis in Myocardial Infarction grade ≥2) in 184 (88%) patients. QT was measured in 10 ± 2 leads on discharge electrocardiograms with a computerized analysis program interfaced with a digitizer. Associations of QTd with 24 variables related to patient characteristics, acute MI, angiography, interventions, and radionuclide ventriculography were evaluated by univariate and multivariate regression.
Results Univariate associations with QTd (p ≤ 0.10) were Thrombolysis in Myocardial Infarction flow grade 0/1 versus 2/3 (QTd = 75 ± 33 msec vs 53 ± 22 msec, p < 0.0001), minimal luminal diameter (p = 0.007), left ventricular ejection fraction at discharge (p = 0.007), reinfarction (p = 0.01), number of leads with ST elevation (p = 0.05), end-systolic volume at discharge (p = 0.04), time to peak creatine kinase (p = 0.06), and YST elevation (p = 0.10). Independent associates of QTd were Thrombolysis in Myocardial Infarction grade 0/1 versus 2/3 (p < 0.0001), reinfarction (p = 0.005), and ejection fraction (p = 0.02). Conclusions Successful thrombolysis is associated with less QTd in patients after acute MI. Our results support the hypothesis that QTd after MI depends on reperfusion status, reinfarction, and left ventricular function. Reduction in QTd may be an additional mechanism by which the benefit of thrombolytic therapy is realized. (Am Heart J 1998;135:1027-35.)
A substantial body of evidence supports the notion that prolongation of ventricular repolarization as measured by the QT interval of standard 12-lead electrocardiography (ECG) is a predictor for life-threatening ventricular arrhythmias in patients with ischemic and nonischemic heart disease.1-6 More recently, a new noninvasive index measuring ventricular repolarization disparity, QT dispersion (QTd), has been introduced, defined as the maximum interlead variation of QT on the 12-lead ECG and linked to malignant ventricular tachyarrhythmias. Because QTd is a noninvasive and easily obtained parameter, a number of studies have been undertaken to determine its clinical From the University of Utah School of Medicine, LDS Hospital. Submitted Jan. 21, 1997; accepted Dec. 10, 1997. Reprint requests: Labros A. Karagounis, MD, University of Utah Medical School, Division of Cardiology, 50 North Medical Dr., Salt Lake City, UT 84132. Copyright © 1998 by Mosby, Inc. 0002-8703/98/$5.00 + 0 4/1/88293
utility for identifying patients at high risk for ventricular arrhythmias.7-17 Previous studies have established that in-hospital and long-term benefits of thrombolysis are closely related to early reestablishment and maintenance of coronary blood flow. These improved outcomes appear to be mediated by preservation of the mechanical properties of the left ventricle and favorable effects on electrophysiologic parameters.18-21 Because QTd may be an index of arrhythmia predisposition, it is important to identify the factors that influence QTd in patients with a recent acute myocardial infarction, a high-risk group for subsequent arrhythmic events. In an earlier study our group demonstrated that successful reperfusion leads to reduced QTd in patients who have acute myocardial infarction and are treated with thrombolytic therapy.22 This study was undertaken in a distinct study population to (1) expand and confirm our previous observations, (2) identify the independent associates of QTd,
American Heart Journal June 1998
1028 Karagounis et al.
and (3) specifically, to evaluate the association of QTd with left ventricular function in patients treated with thrombolytic therapy early after acute myocardial infarction.
Methods Patient selection The study population included 325 patients with acute myocardial infarction consecutively recruited to a multicenter, double-blind, randomized study comparing alteplase (Activase) and anistreplase (Eminase) (TEAM-3, third trial of Thrombolysis with Eminase in Acute Myocardial Infarction). Inclusion and exclusion patient criteria and other details of the study have been previously published.23 In brief, patients were enrolled if they were ≤76 years old, had typical chest pain lasting for ≥30 minutes but not >4 hours that did not respond to nitroglycerin, and had ST-segment elevation compatible with acute myocardial infarction. Patients were excluded for the following reasons: cardiogenic shock, previous coronary bypass surgery, recent coronary angioplasty, contraindications to thrombolysis, recent administration of streptokinase or anistreplase, and absence of informed consent. In addition to the original study exclusions, patients were also excluded from this study if they were not in sinus rhythm or had bundle branch block or any other intraventricular conduction abnormality on the electrocardiogram selected for analysis in the study. All patients received thrombolytic therapy within 4 hours of symptom onset. Aspirin and intravenous heparin were routinely given.
Coronary angiography Coronary perfusion status was determined by angiography performed 1 day (18 to 48 hours) after the initiation of lytic therapy. Multiple angiographic views were obtained including the optimal view for maximizing the degree of stenosis of the infarct-related artery and the orthogonal view. A single experienced observer assessed coronary patency by reading all angiograms in a blind fashion at a central laboratory (LDS Hospital). Anterograde perfusion of the infarct-related artery was graded according to the classification system of the Thrombolysis in Myocardial Infarction (TIMI) trial24 as follows: grade 0, no anterograde perfusion; grade 1, minimal perfusion, penetration of the obstructed lumen with negligible distal flow; grade 2, partial perfusion: coronary bed perfuses distal to the obstruction but at a delayed rate of filling and clearing; and grade 3, complete perfusion: coronary bed perfuses distal to the obstruction with a normal rate of filling and clearance. Interobserver variability was previously shown to be small in our laboratory for the determination of patency (TIMI grade 2 or 3 perfusion) versus occlusion (grade 0 or 1 perfusion).25 The degree of residual stenosis of the infarct-related artery was quantified with a validated technique modified from Brown et al.26,27
Radionuclide ventriculography Convalescent radionuclide ventriculography was performed at 10 days (7 to 10 days after dosing or hospital discharge, whichever came first). Gated blood pool scans were performed with standard protocols28,29 on the same equipment for each individual patient on all studies at each center. The standardized procedure enabled centralized reading of ejection fractions and end-systolic and end-diastolic volumes at the study’s core radionuclide laboratory. Individual studies were copied onto floppy computer disks by the study centers and sent to the core radionuclide laboratory, where blinded assessment was made with previously standardized and validated techniques.
ECG Standard 12-lead ECG was performed on the tenth day of hospitalization or at hospital discharge, whichever came first. At the time of their convalescent ECG, patients were not taking class I or class III antiarrhythmic drugs, and their electrolyte status was generally normal. All standard 12-lead electrocardiograms were recorded at 25 mm/sec paper speed and performed immediately before hospital discharge. All electrocardiograms were examined retrospectively by a single observer who was unaware of the patients’ drug and coronary patency status. Measurements were performed with a commercially available computer program (Configurable Measurement System, Salt Lake City, Utah) interfaced with a Calcomp 9000 digitizer. Intraobserver, interobserver, and interstudy variability with this measurement system was shown to be small in a previous study.22 QT interval was measured from the onset of the QRS complex to the end of the T wave, defined as its return to the TP isoelectric baseline. The QT was measured to the nadir of the curve between the T and U waves when U waves were present. Whenever possible, the average measurement of three complexes for each lead was taken. If the end of the Twave could not be reliably determined, or when the T wave was isoelectric or of very low amplitude, the QT measurement was not made in that lead, and it was excluded from analysis. A lower limit of five or more technically adequate leads per electrocardiogram was required for inclusion in this study. The duration of the QRS was also measured and averaged in analyzable lead. JT was calculated after the QT was subtracted from the corresponding QRS duration. QTd was defined as the difference between the maximum and minimum QT interval measurements among all of the measured (12) leads on the standard electrocardiogram. QTc (or heart rate-corrected QT interval) was calculated according to Bazett’s formula30,31 as follows: QTc = QT/square root of the R-R interval. QTc, JT, and JTc dispersion were calculated in a manner similar to QTd.
Statistics Results are expressed as mean ± SD unless stated otherwise. Significant univariate associations of QTd with the fol-
American Heart Journal Volume 135, Number 6, Part 1
Karagounis et al. 1029
Table I. Patient characteristics and variables of interest n = 207 57 ± 11 24 81 ± 16 34 42 14 14
Demographics Age (yr, mean ± SD) Sex (% female) Weight (kg, mean ± SD) History of angina (%) History of hypertension (%) History of myocardial infarction (%) History of diabetes mellitus (%) Acute myocardial infarction variables Time lapsed from symptom onset to therapy (hr, mean ± SD) Anterior MI vs other (%) Sum of ST-segment elevation (mV, mean ± SD) No. of leads with ST elevation (mean ± SD) Alteplase (vs anistreplase) (%) Reinfarction (%) Peak CK (IU/L, mean ± SD) Time to peak CK (hours, mean ± SD) Angiographic data TIMI 2, 3 (vs 0, 1) (%) TIMI 3 (vs 0, 1 ,2) (%) Minimal lumen diameter (mm, mean ± SD) Baseline LVEF (%) No. of diseased arteries (%) One-vessel Two-vessel Three-vessel Interventions PTCA (%) CABG (%) Radionuclide findings at discharge Discharge EF (%) 53 ± 12 Discharge diastolic volume (ml, mean ± SD) Discharge systolic volume (ml, mean ± SD)
2.7 ± 0.9 37 1.1 ± 0.8 4.0 ± 1.6 51 3 2315 ± 1885 11.6 ± 6.4 88 75 0.98 ± 0.52 54 ± 12.3 48 34 15 30 12 163 ± 33 78 ± 31
MI, Myocardial infarction; CK, creatine kinase; LVEF, left ventricular ejection fraction; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting.
lowing 24 variables were first identified: patient characteristics (age, sex, weight, history of angina, history of myocardial infarction, history of hypertension, history of diabetes mellitus, history of congestive heart failure), acute myocardial infarction variables (location, time from symptom onset to therapy, sum of ST elevation, number of leads with ST elevation, thrombolytic drug, reinfarction, peak creatine kinase, and time to peak), angiographic findings at 1 day (TIMI grade flow 0/1 vs 2/3, minimal lumen diameter, number of diseased arteries), interventions (angioplasty, bypass surgery) and radionuclide data at discharge (global left ventricular ejection fraction, end-diastolic and end-systolic volumes). To determine the multivariate independent associates between QTd and the initially identified univariate predictors, stepwise multivariate regression modeling was performed. A p value of ≤0.1 was required to allow a factor to enter the model, and p of ≥0.15 was required to subsequently remove a factor. (We chose a p ≤ 0.1 because we wished to identify all potentially important determinants of QTd.32) Similar analyses were performed used as dependent variables QTc, JT, and JTc dispersion, each in separate regression models. Analyses were
performed with SPSS 6.1 (SPSS Inc., Chicago, Ill.) statistical software for Macintosh (Apple Computer, Cupertino, Calif.). A p value <0.05 was considered significant.
Results Patient characteristics Of the 325 patients enrolled in TEAM-3, 118 patients were excluded because of the following reasons: no discharge electrocardiograms (n = 29), poor quality electrocardiograms (n = 12), unavailable records (n = 3), no angiographic scoring done (n = 20), presence of bundle branch block (n = 24), grafted vessels (n = 11), atrial fibrillation (n = 3), or no discharge radionuclide left ventriculogram (n = 16). The remaining 207 patients satisfied the entry criteria for this study, including 158 men and 49 women with a mean age of 57.1 ± 11 years. Patients were given alteplase (n = 106) or anistreplase (n = 101) 2.7 ± 0.9 hours from symptom onset. Table I summarizes other patient characteristics and variables of interest.
American Heart Journal June 1998
1030 Karagounis et al.
Table II. Comparisons of QTd by perfusion grade TIMI perfusion grade
No. of patients QTd (msec) QTcd (msec)
0
1
2
3
p Value*
16 74 ± 25 85 ± 28
8 74 ± 18 80 ± 16
27 57 ± 30 66 ± 34
156 50 ± 20 57 ± 23
≤0.0001 ≤0.0001
*Analysis of variance.
Table III. Comparison of QTd by perfusion grade TIMI perfusion grade
No. of patients QTd (msec) QTcd (msec)
TIMI perfusion grade
0, 1
2, 3
p Value
0, 1, 2
3
p Value*
24 74 ± 23 84 ± 24
183 51 ± 22 59 ± 25
≤0.0001 ≤0.0001
51 65 ± 28 74 ± 31
156 50 ± 20 57 ± 23
0.0001 ≤0.0001
*Analysis of variance.
Angiographic and left ventricular function data All 207 patients had coronary angiography done at 32.3 ± 29.8 hours after thrombolytic therapy. The infarct-related lesion was identified within the right circumflex coronary artery in 129 (62%) patients and the left anterior descending coronary artery in 76 (38%) patients. Two patients had an undefined infarct-related artery and were assigned a TIMI perfusion grade of 3. A total of 156 (75.4%), 27 (13.0%), 8 (3.9%), and 16 (7.7%) patients had TIMI perfusion grades 3, 2, 1, and 0, respectively. The total patency rate (defined as TIMI grades 2 and 3) was high (88%) in these 207 patients and was not significantly different between alteplase and anistreplase (p = 0.75). Radionuclide ventriculography was performed 7.7 ± 2.9 days after thrombolytic therapy.
ECG data ECG was performed at 8 ± 4 days from symptom onset. All 12 leads were measurable in 39 (18.8%) patients. In the remaining 168 (81%) patients the QT interval was measurable in 9.6 ± 1.5 leads. The mean QT interval averaged over all 12 leads was 395 ± 37 msec (range 302 to 481 msec). The mean QTc interval was 430 ± 29 msec (range 341 to 533 msec). The maximum QT interval was 421 ± 39 msec (range 315 to 517 msec) and occurred in one of the precordial leads in 117 (57%) patients and in one of the limb leads in 90 (44%) patients. The maximum QTc was 459 ± 33
msec (range 360 to 585 msec). The minimum QT interval was 367 ± 40 msec (range 261 to 460 msec) and occurred in one of the precordial leads in 102 (49%) patients and in one of the limb leads in 105 (51%). The minimum QTc was 397 ± 32 msec (range 298 to 509 msec). The average QTd for this study group was 54 ± 23 msec (range 14 to 152 msec), and the average QTc dispersion was 62 ± 26 msec (range 15 to 173 msec).
QTd and perfusion status Table II shows QT and QTc dispersion by perfusion grade in all patients. Significant differences were seen in QT and QTc dispersion among the TIMI perfusion groups (p ≤ 0.0001). There was also a gradual decrease in QT and QTc dispersion with improving (increasing) TIMI perfusion grades. These results did not change when the patients treated with alteplase and anistreplase were analyzed separately. There was a trend to less mean QT and QTc dispersion in patients treated with alteplase than in those treated with anistreplase, but statistical significance was not reached. Table III shows the correlation of QT and QTc dispersion with perfusion grade in all patients when comparing TIMI grades 0/1 versus 2/3 and TIMI grades 0/1/2 versus 3. Significant differences were seen in QT and QTc dispersion between TIMI grades 0, 1 versus 2/3, and between TIMI grades 0/1/2 and 3 (p ≤ 0.0001 for both comparisons). These results did not change when
American Heart Journal Volume 135, Number 6, Part 1
Karagounis et al. 1031
Table IV. Univariate associations of QTd Variable Patient characteristics Age (yr) Sex Weight History of angina History of hypertension History of myocardial infarction History of diabetes mellitus Acute myocardial infarction variables Time lapsed from symptom onset to therapy Anterior MI vs other ∑ST-segment elevation (mm) No. of leads with ST elevation Alteplase vs anistreplase Reinfarction Peak CK Time to peak CK Angiographic data TIMI 0, 1 vs 2, 3 TIMI 0, 1, 2 vs 3 Minimal lumen diameter (mm) Infarct-related artery: LAD vs other No. of diseased arteries Interventions PTCA CABG Radionuclide data at discharge Discharge EF Discharge end-diastolic volume Discharge end-systolic volume
Coefficient ± SEM
Correlation
F
p Value
0.14 ± 0.15 0.06 ± 3.73 –0.08 ± 0.10 4.27 ± 3.33 0.32 ± 3.22 –0.81 ± 4.64 3.35 ± 4.57
0.06 0.00 0.05 0.09 0.01 0.01 0.05
0.85 0.00 0.60 1.64 0.01 0.03 0.54
0.36 0.99 0.44 0.20 0.92 0.86 0.46
–1.08 ± 1.85 4.57 ± 3.28 3.26 ± 1.97 1.94 ± 0.97 –4.48 ± 3.16 21.98 ± 8.87 0.53 ± 0.01 0.47 ± .25
0.04 0.10 0.12 0.14 0.09 0.18 0.05 0.13
0.34 1.94 2.75 4.02 2.00 6.14 0.53 3.55
0.56 0.17 0.10 0.05 0.16 0.01 0.47 0.06
22.66 ± 4.70 14.62 ± 3.54 –8.24 ± 3.00 5.26 ± 3.27 3.20 ± 2.09
0.32 0.28 0.19 0.11 0.11
23.21 17.03 7.55 2.60 2.36
–1.03 5.23 ± 4.86
0.02 0.07
0.09 1.16
0.77 0.28
–0.37 ± 0.14 0.62 ± 0.05 0.11 ± 0.05
0.19 0.09 0.15
7.48 1.72 4.51
0.007 0.19 0.04
0.0000 0.0001 0.007 0.11 0.13
F, F statistic; EF, left ventricular ejection fraction; MI, myocardial infarction; CK, creatine kinase; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting.
the patients treated with anistreplase and streptokinase were analyzed separately.
Univariate and multivariate associates of QTd Table IV shows the univariate and Table V the multivariate associations with QTd. Variables that met the significance level of p ≤ 0.1 (∑ST elevation, number of leads with ST elevation, reinfarction, time to peak creatine kinase, TIMI grade flow 0/1 vs 2/3, minimal lumen diameter, reinfarction, left ventricular ejection fraction at discharge, and end-systolic volume) were further subjected to multivariate modeling to determine their independent association with QT dispersion. The most significant multivariate associate was found to be the TIMI grade (p ≤ 0.0001) followed by reinfarction (p = 0.005) and left ventricular ejection fraction (p = 0.02). Similarly, in separate models where QTc, JT, and JTc dispersion were used as the dependent variable, TIMI grade also emerged as the strongest associated factor (p ≤ 0.0001 in all three models). The significance of the association
was not diminished when TIMI grade was dichotomized into grades 0/1/2 versus 3 (Table VI): p ≤ 0.0001 for QT, QTc, JT, or JTc dispersion in separate models.
Discussion Summary of study findings In this study we sought to determine the significant associates of dispersion in ventricular repolarization (the difference between maximum and minimum QT interval measurements in any of the 12 leads on a standard electrocardiogram) in patients who had acute myocardial infarction and were treated with thrombolytics. We analyzed 24 variables describing patient clinical characteristics, myocardial infarction characteristics, and angiographic and ventriculographic parameters. We found by multivariate regression modeling that the perfusion status of the infarct-related artery, TIMI grade 0/1 vs 2/3 flow (0/1 = 75 msec, 2/3 = 53 msec, p < 0.0001), was the most important associate of
American Heart Journal June 1998
1032 Karagounis et al.
Table V. Multivariate associations of QTd by TIMI 0, 1 vs 2, 3 Regression model parameters
Variable
Coefficient ± SEM
p Value
R = 0.40 R2 = 0.16 F = 12.68, p = 0.0000
TIMI grade 0, 1 vs 2,3 Reinfarction Discharge EF
–22.16 ± 4.62 24.76 ± 8.75 –0.31 ± 0.13
0.0000 0.005 0.02
Variable
Coefficient ± SEM
p Value
TIMI grade 0, 1, 2 vs 3 Reinfarction Discharge EF
–13.43 ± 3.53 23.16 ± 8.91 –0.31 ± 0.13
0.0002 0.01 0.03
F, F statistic of regression model; R, multiple correlation coefficient; R2, coefficient of variation.
Table VI. Multivariate associations of QTd by TIMI 0, 1 vs 2, 3 Regression model parameters R = 0.35 R2 = 0.12 F = 9.61, p = 0.0000 Abbreviations as in Table V.
QTd, followed by whether reinfarction (yes = 74 msec, no = 51 msec, p = 0.005) occurred during hospitalization and by left ventricular ejection fraction at discharge (p = 0.02). These findings suggest that successful and sustained restoration of blood flow leads to reduced dispersion of repolarization.
Comparison with previous work Our group first reported the association of TIMI grade flow and QTd in an earlier study of streptokinase and anistreplase of acute myocardial infarction.22 In that study patients with TIMI grade 2/3 had lesser QTd on their postmyocardial infarction 12-lead ECG than did those with grade 0/1 flow (54 ± 20 vs 94 ± 29 msec, p = 0.0001). This association was independent of age, sex, drug assignment, infarct site, infarctrelated vessel, and number of measurable leads. However, left ventricular function was not measured, and the degree of independence of QTd and cardiac dysfunction could not be ensured. This study validates the association of TIMI flow grade and QTd and confirms its independence from patient and infarct characteristics. It is important that this study adds to the earlier study in showing that perfusion grade continues to be a strong predictor of convalesce QTd even after accounting for differences in left ventricular function (ejection fraction, end-systolic and end-diastolic volumes). This study also differs in assessing patency at “plateau” (1 day), after the ongoing processes of early reperfusion and reocclusion largely have reached equilibrium.
Mechanisms of thrombolytic therapy benefits The main goal of thrombolytic therapy in patients with acute myocardial infarction is the establishment and maintenance of coronary patency to improve left ventricular function and decrease mortality. The use of thrombolytic agents has been documented to reduce the mortality rate after infarction; the mechanisms by which these benefits are mediated appear to include improvements both in mechanical and electrical function.18-21,33 In this study of patients who had infarction we have shown that successful thrombolysis is associated with less dispersion in QT interval measurements. There also was an association of left ventricular function with QTd of lesser magnitude, suggesting that the effects of successful reperfusion on QTd are independent of those leading to the preservation of left ventricular function. Also, patients with reinfarction representing late failure of reperfusion had a greater dispersion than those without. Thus by establishing sustained patency, thrombolytic therapy may prevent or attenuate the development of an abnormal electrophysiologic substrate after myocardial infarction. Our results thus suggest that reduction in QTd may be an additional mechanism of the benefit of thrombolytic therapy. It should be pointed out that the predictive value for QTd of our multivariate model was of modest strength, because it was measured by the coefficient of variation (Table V, R2 = 0.16). The inability to achieve a better prediction of QTd indicates that important additional unknown factor(s) affect ventricular repolarization (such as autonomic influences), or the accuracy of our
American Heart Journal Volume 135, Number 6, Part 1
method of measuring QTd may be limited (e.g., by significant “noise” contaminating the true signal).
Dispersion of repolarization and arrhythmogenesis Dispersion of refractoriness is a measure of nonhomogeneous recovery of excitability in a given mass of cardiac tissue and is determined by the combined effects of the differences in activation times and action potential durations. A great deal of experimental work has substantiated the association between dispersion of repolarization and malignant ventricular tachyarrhythmias. Han and Moe34 were the first to demonstrate (in an animal model) the relation between dispersion of refractoriness and a decrease in fibrillation threshold under diverse experimental conditions. Kuo et al.7 proposed three possible mechanisms (based on their own experiments) by which increased dispersion predisposed to arrhythmias: (1) focal reexcitation, (2) reentry facilitated by a conduction from an area with a shorter refractory period to an area with a longer refractory period, and (3) reentry facilitated by conduction from an area with a longer refractory period to an area with a shorter refractory period. In the studies of Restivo et al.35 of canine myocardial infarction, dispersion of effective refractory period in a normal heart averaged 30 msec and was dramatically increased in the infarcted heart to 160 msec. The reentry waveform was initiated by a premature complex that originated from the region of shortest refractoriness and was conducted towards regions with longer refractory periods. Vassallo et al.36 measured the dispersion of excitability in the left ventricle in three groups of patients. In seven individuals with normal hearts and no arrhythmias, the average dispersion of total recovery time was 52 msec. In six patients with previous myocardial infarction who had sustained ventricular tachycardia, the mean dispersion of total recovery time was 90 msec. In the five patients with long QT syndrome and previous cardiac arrest, dispersion was the longest, averaging 114 msec. It was concluded that nonuniform recovery of excitability in the left ventricle is caused by dispersion of refractoriness or activation time and depends on the pathophysiogic substrate. The link between QT variation across the 12 leads and regional differences of ventricular repolarization was provided by Cowan et al.,37 who were able to demonstrate that regional differences in ventricular recovery time were reflected by epicardial monophasic action potentials recordings during open heart surgery. More recently, Cao et al.38 found an excellent correla-
Karagounis et al. 1033
tion between surface QTd and monophasic action potential dispersion in rat hearts. Increased QTd was observed in patients with acute myocardial infarction who had ventricular tachyarrhythmias compared with those with a benign course.39,40 Increased dispersion also was noted in patients with ventricular tachyarrhythmias compared with control (normal) groups. Pye et al.16 found that in patients with sustained ventricular tachycardia, QTd was significantly longer (77 msec) than in those without (38 msec). Perkiomaki et al.41 showed a greater QTd in patients with cardiac arrest or monomorphic ventricular tachycardia (104 ± 41 msec) compared with that in healthy individuals (38 ± 14 msec) or that in patients with previous myocardial infarction but without a history of ventricular tachyarrhythmias (65 ± 31 msec). In multivariate analysis, among clinical and angiographic variables QTd was the strongest independent factor and most effectively separated the patient groups with and without susceptibility to ventricular tachyarrhythmias (p < 0.001). QTd is also increased in patients with long QT syndromes at risk of ventricular arrhythmias,8 patients with hypertrophic cardiomyopathy,12 and patients with chronic heart failure.13 To definitively establish the predictive value of QTd for mortality after myocardial infarction, a sizable prospective trial is required.
Limitations A major limitation of this study is that the QT interval is not measurable in every lead and is difficult to measure with precision, as recognized by previous investigators.42,43 However, the inability to measure QT interval in some leads will tend, if anything, to result in an underestimation rather than an overestimation of QTd. To ensure reliable QT measurements we applied a consistent method, and we used a reliable computerized system. The reader of the QT intervals was blinded to patient clinical characteristics and outcomes. Similarly, grading of TIMI perfusion was carried out without knowledge of patients’ clinical status or ECG findings. Manual measurement of QTd is a laborious process. Given the potential application of QTd as a noninvasive predictor of life-threatening ventricular tachyarrhythmias, computerized measurement of ventricular repolarization based on objective standardized criteria would add to the current capability. Finally, long-term outcomes of patients were not available in our study population to verify the predic-
1034 Karagounis et al.
tive value of QTd for clinical events (the study was too small for mortality assessment, even if available).
Conclusions In conclusion, the results of our study show that successful thrombolysis is the primary associate of QTd in patients after acute myocardial infarction. Our data support the hypothesis that QTd after myocardial infarction depends on reperfusion status and to a lesser degree on left ventricular function. Reduction in QTd may be an additional mechanism of benefit of successful thrombolytic therapy. Because homogeneity of recovery time is thought to protect against arrhythmias and dispersion of recovery time is thought to be arrhythmogenic, the reduction in QTd seen in our patients may be interpreted as a mechanism of the benefit of thrombolytic therapy.
References 1. Schwartz PJ, Wolf S. QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 1978;57:1074-7. 2. Ahnve S, Helmers C, Lundman T, Rehnqvist N, Sjogren A. QTc intervals in acute myocardial infarction: first-year prognostic implications. Clin Cardiol 1980;3:303-8. 3. Wheelan K, Mukharji J, Rude RE, et al. Sudden death and its relation to QT-interval prolongation after acute myocardial infarction: two-year follow-up. Am J Cardiol 1986;57:745-50. 4. Puddu PE, Bourassa MG. Prediction of sudden death from QTc interval prolongation in patients with chronic ischemic heart disease. J Electrocardiol 1986;19:203-12. 5. Algra A, Tijssen JGP, Roelandt JRTC, Pool J, Leibsen J. QTc prolongation measured by standard 12 lead electrocardiography is an independent risk factor for sudden death due to cardiac arrest. Circulation 1991;83:1888-94. 6. Schouten EG, Dekker JM, Meppelink P, et al. QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation 1991;84:1516. 7. Kuo CS, Munakata K, Reddy CP, Surawicz B. Characteristics and possible mechanism of ventricular arrhythmia dependent on the dispersion of action potential durations. Circulation 1983;67:1356-67. 8. Day CP, McComb JM, Campbell RFW. QT dispersion: an indicator of arrhythmia risk in patients with long QT intervals. Br Heart J 1990;63:342-4. 9. Linker NJ, Colonna P, Kekwick CA, Till J, Camm AJ, Ward DE. Assessment of QT dispersion in symptomatic patients with congenital long QT syndromes. Am J Cardiol 1992;69:634-8. 10. Hii JT, Wyse DG, Gillis AM, Duff HJ, Solylo MA, Mitchell LB. Precordial QT interval dispersion as a marker of torsade de pointes. Disparate effects of class Ia antiarrhythmic drugs and amiodarone. Circulation 1992;86:1376-82. 11. Miorelli M, Buja G, Melacini P, Fasoli G, Nava A. QT-interval variability in hypertrophic cardiomyopathy patients with cardiac arrest. Int J Cardiol 1994;45:121-7. 12. Barr CS, Naas A, Freeman M, Lang CC, Struthers AD. QT dispersion and sudden unexpected death in chronic heart failure. Lancet 1994; 343:327-9.
American Heart Journal June 1998
13. Buja G, Miorelli M, Turrini P, Melacini P, Nava A. Comparison of QT dispersion in hypertrophic cardiomyopathy between patients with and without ventricular arrhythmias and sudden death. Am J Cardiol 1993; 72:973-6. 14. Higham PD, Furniss SS, Campbell RW. QT dispersion and components of the QT interval in ischaemia and infarction. Br Heart J 1995; 73:32-6. 15. Jordaens L, Missault L, Pelleman G, Duprez D, De Backer G, Clement DL. Comparison of athletes with life-threatening ventricular arrhythmias with two groups of healthy athletes and a group of normal control subjects. Am J Cardiol 1994;74:1124-8. 16. Pye M, Quinn AC, Cobbe SM. QT interval dispersion: a non-invasive marker of susceptibility to arrhythmia in patients with sustained ventricular arrhythmias? Br Heart J 1994;71:511-4. 17. Zareba W, Moss AJ, le Cessie S. Dispersion of ventricular repolarization and arrhythmic cardiac death in coronary artery disease. Am J Cardiol 1994;74:550-3. 18. Sager PT, Perlmutter RA, Rosenfeld LE, McPherson CA, Wackers FTh, Batsford WP. Electrophysiologic effects of thrombolytic therapy in patients with transmural anterior myocardial infartion complicated by left ventricular aneurysm formation. J Am Coll Cardiol 1988;12:19-24. 19. Zaret BL, Wackers FJ, Terrin M, et al. Does left ventricular ejection fraction following thrombolytic therapy have the same prognostic impact described in the prethrombolytic era? Results of the TIMI II Trial. J Am Coll Cardiol 1991;17:214A. 20. Gang ES, Lew AS, Hong M, Wang FZ, Liebert CA, Peter T. Decreased incidence of ventricular late potentials after successful thrombolytic therapy for acute myocardial infarction. N Engl J Med 1989;321:712-6. 21. Vatterrott PJ, Hammil SC, Bailey KR, Wiltgen CM, Gersh BJ. Late potentials on signal-averaged electrocardiograms and patency of the infarct-related artery in survivors of acute myocardial infarction. J Am Coll Cardiol 1991;17:330-7. 22. Moreno FL, Villanuela MT, Karagouis LA, Anderson JL. Reduction in QT interval dispersion by successful thrombolytic therapy in acute myocardial infarction. Circulation 1994;90:94-100. 23. Anderson JL, Becker LC, Sorensen S, et al. Anistreplase versus alteplase in acute myocardial infarction: comparative effects on left ventricular function, 1 day patency, and morbidity: the TEAM-3. J Am Coll Cardiol 1992;20:753-66. 24. Chesebro JH, Knatterud G, Roberts R, et al. Thrombolysis in myocardial infarction (TIMI) trial, phase I: a comparison between intravenous plasminogen activator and intravenous streptokinase. Circulation 1987; 76:142-54. 25. Hackworthy RA, Sorensen SG, Fitzpatrick PB, et al. Dependence of assessment of coronary artery reperfusion on angiographic criteria and interobserver variability. Am J Cardiol 1988;62:538-42. 26. Moreno FLl, Stoops KL, Hackworthy RA, Menlove RL, van Bree R, Anderson JL. Quantification of rate of coronary artery disease progression by a new method of angiographic analysis. Am Heart J 1991; 121:1062-70. 27. Brown BG, Bolson E, Frimer M, Dodge HT. Quantitative coronary arteriography. 1977;55:329-37. 28. Pavel DG, Zimmer AM, Patterson VN. In-vivo labeling of red blood cells with 99mTc: a new approach to blood pool visualization. Nucl Med 1977;18:305-8. 29. Callahan RJ, 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;23:315-8.
American Heart Journal Volume 135, Number 6, Part 1
30. Bazett H. An analysis of the time relationships of the heart. Heart 1920;7:353-70. 31. Ahnve S. Correction of the QT interval for heart rate: review of different formulas and the use of Bazett’s formula in myocardial infarction. Am Heart J 1985;109:568-74. 32. Kleinbaum DG, Kupper LL, Muller KE. Selecting the best regression equation. Applied regression analysis and other multivariate methods. 2nd ed. Boston: PWS-Kent; 1988. p. 314-40. The Duxbury series in statistics and decision sciences. 33. Califf RM, Topol EJ, Gersh BJ. From myocardial salvage to patient salvage in acute myocardial infarction: the role of reperfusion therapy. J Am Coll Card 1989;14:1382-8. 34. Han J, Moe GK. Nonuniform recovery of excitability in ventricular muscle. 1964;14:44-60. 35. Restivo M, Gough WB, El-Sherif N. Reentrant ventricular rhythms in the late myocardial infarction period: prevention of reentry by dual stimulation during basic rhythm. Circulation 1988;77:429. 36. Vassallo JA, Cassidy DM, Kindwall KE, Marchlinski FE, Josephson ME. Nonuniform recovery of excitability in the left ventricle. Circulation 1988;78:1365-72.
Karagounis et al. 1035
37. Cowan JC, Hilton CJ, Griffiths CJ, et al. Sequence of epicardial repolarization and configuration of T wave. Br Heart J 1988;60:424-33. 38. Cao W, Ranade V, Jaffrani N, Khosla S, Somberg JC. Relationship between QT dispersion and action potential duration. J of Investigative Medicine 1996;43:430A. 39. Higham PD, Furniss SS, Campell RWF. QT dispersion in ischaemia and infarction: a marker of primary ventricular fibrillation? (Abstract). Eur Heart J 1991;12:87. 40. Potratz J, Djonlagic H, Mentzel H. Prognostic significance of QT dispersion in patients with acute myocardial infarction. (Abstract) Eur Heart J 1993;14:254. 41. Perkiomaki JS, Koistinen MJ, Yli Mayry S, Huikuri HV. Dispersion of QT interval in patients with and without susceptibility to ventricular tachyarrhythmias after previous myocardial infarction. J Am Coll Cardiol 1995;26:174-9. 42. Lepeschkin E, Surawicz B. The measurement of the QT interval of the electrocardiogram. Circulation 1952;6:378-88. 43. Ahnve S. Errors in the visual determination of corrected QT (QTc) interval during myocardial infarction. J Am Coll Cardiol 1985; 5:699-702.
Background QT dispersion (QTd; QT interval maximum minus minimum) has been shown to reflect regional variations in ventricular repolarization and is increased in patients with life-threatening ventricular arrhythmias. Methods To determine correlates of QTd in patients who had had myocardial infarction (MI), 207 patients (158 men, aged 57 ± 11 years) with acute MI who were treated with alteplase or anistreplase within 2.7 ± 0.9 hours of symptom onset were studied. Angiograms at a median of 27 hours after thrombolysis showed reperfusion (Thrombolysis in Myocardial Infarction grade ≥2) in 184 (88%) patients. QT was measured in 10 ± 2 leads on discharge electrocardiograms with a computerized analysis program interfaced with a digitizer. Associations of QTd with 24 variables related to patient characteristics, acute MI, angiography, interventions, and radionuclide ventriculography were evaluated by univariate and multivariate regression.
Results Univariate associations with QTd (p ≤ 0.10) were Thrombolysis in Myocardial Infarction flow grade 0/1 versus 2/3 (QTd = 75 ± 33 msec vs 53 ± 22 msec, p < 0.0001), minimal luminal diameter (p = 0.007), left ventricular ejection fraction at discharge (p = 0.007), reinfarction (p = 0.01), number of leads with ST elevation (p = 0.05), end-systolic volume at discharge (p = 0.04), time to peak creatine kinase (p = 0.06), and YST elevation (p = 0.10). Independent associates of QTd were Thrombolysis in Myocardial Infarction grade 0/1 versus 2/3 (p < 0.0001), reinfarction (p = 0.005), and ejection fraction (p = 0.02). Conclusions Successful thrombolysis is associated with less QTd in patients after acute MI. Our results support the hypothesis that QTd after MI depends on reperfusion status, reinfarction, and left ventricular function. Reduction in QTd may be an additional mechanism by which the benefit of thrombolytic therapy is realized. (Am Heart J 1998;135:1027-35.)
A substantial body of evidence supports the notion that prolongation of ventricular repolarization as measured by the QT interval of standard 12-lead electrocardiography (ECG) is a predictor for life-threatening ventricular arrhythmias in patients with ischemic and nonischemic heart disease.1-6 More recently, a new noninvasive index measuring ventricular repolarization disparity, QT dispersion (QTd), has been introduced, defined as the maximum interlead variation of QT on the 12-lead ECG and linked to malignant ventricular tachyarrhythmias. Because QTd is a noninvasive and easily obtained parameter, a number of studies have been undertaken to determine its clinical From the University of Utah School of Medicine, LDS Hospital. Submitted Jan. 21, 1997; accepted Dec. 10, 1997. Reprint requests: Labros A. Karagounis, MD, University of Utah Medical School, Division of Cardiology, 50 North Medical Dr., Salt Lake City, UT 84132. Copyright © 1998 by Mosby, Inc. 0002-8703/98/$5.00 + 0 4/1/88293
utility for identifying patients at high risk for ventricular arrhythmias.7-17 Previous studies have established that in-hospital and long-term benefits of thrombolysis are closely related to early reestablishment and maintenance of coronary blood flow. These improved outcomes appear to be mediated by preservation of the mechanical properties of the left ventricle and favorable effects on electrophysiologic parameters.18-21 Because QTd may be an index of arrhythmia predisposition, it is important to identify the factors that influence QTd in patients with a recent acute myocardial infarction, a high-risk group for subsequent arrhythmic events. In an earlier study our group demonstrated that successful reperfusion leads to reduced QTd in patients who have acute myocardial infarction and are treated with thrombolytic therapy.22 This study was undertaken in a distinct study population to (1) expand and confirm our previous observations, (2) identify the independent associates of QTd,
American Heart Journal June 1998
1028 Karagounis et al.
and (3) specifically, to evaluate the association of QTd with left ventricular function in patients treated with thrombolytic therapy early after acute myocardial infarction.
Methods Patient selection The study population included 325 patients with acute myocardial infarction consecutively recruited to a multicenter, double-blind, randomized study comparing alteplase (Activase) and anistreplase (Eminase) (TEAM-3, third trial of Thrombolysis with Eminase in Acute Myocardial Infarction). Inclusion and exclusion patient criteria and other details of the study have been previously published.23 In brief, patients were enrolled if they were ≤76 years old, had typical chest pain lasting for ≥30 minutes but not >4 hours that did not respond to nitroglycerin, and had ST-segment elevation compatible with acute myocardial infarction. Patients were excluded for the following reasons: cardiogenic shock, previous coronary bypass surgery, recent coronary angioplasty, contraindications to thrombolysis, recent administration of streptokinase or anistreplase, and absence of informed consent. In addition to the original study exclusions, patients were also excluded from this study if they were not in sinus rhythm or had bundle branch block or any other intraventricular conduction abnormality on the electrocardiogram selected for analysis in the study. All patients received thrombolytic therapy within 4 hours of symptom onset. Aspirin and intravenous heparin were routinely given.
Coronary angiography Coronary perfusion status was determined by angiography performed 1 day (18 to 48 hours) after the initiation of lytic therapy. Multiple angiographic views were obtained including the optimal view for maximizing the degree of stenosis of the infarct-related artery and the orthogonal view. A single experienced observer assessed coronary patency by reading all angiograms in a blind fashion at a central laboratory (LDS Hospital). Anterograde perfusion of the infarct-related artery was graded according to the classification system of the Thrombolysis in Myocardial Infarction (TIMI) trial24 as follows: grade 0, no anterograde perfusion; grade 1, minimal perfusion, penetration of the obstructed lumen with negligible distal flow; grade 2, partial perfusion: coronary bed perfuses distal to the obstruction but at a delayed rate of filling and clearing; and grade 3, complete perfusion: coronary bed perfuses distal to the obstruction with a normal rate of filling and clearance. Interobserver variability was previously shown to be small in our laboratory for the determination of patency (TIMI grade 2 or 3 perfusion) versus occlusion (grade 0 or 1 perfusion).25 The degree of residual stenosis of the infarct-related artery was quantified with a validated technique modified from Brown et al.26,27
Radionuclide ventriculography Convalescent radionuclide ventriculography was performed at 10 days (7 to 10 days after dosing or hospital discharge, whichever came first). Gated blood pool scans were performed with standard protocols28,29 on the same equipment for each individual patient on all studies at each center. The standardized procedure enabled centralized reading of ejection fractions and end-systolic and end-diastolic volumes at the study’s core radionuclide laboratory. Individual studies were copied onto floppy computer disks by the study centers and sent to the core radionuclide laboratory, where blinded assessment was made with previously standardized and validated techniques.
ECG Standard 12-lead ECG was performed on the tenth day of hospitalization or at hospital discharge, whichever came first. At the time of their convalescent ECG, patients were not taking class I or class III antiarrhythmic drugs, and their electrolyte status was generally normal. All standard 12-lead electrocardiograms were recorded at 25 mm/sec paper speed and performed immediately before hospital discharge. All electrocardiograms were examined retrospectively by a single observer who was unaware of the patients’ drug and coronary patency status. Measurements were performed with a commercially available computer program (Configurable Measurement System, Salt Lake City, Utah) interfaced with a Calcomp 9000 digitizer. Intraobserver, interobserver, and interstudy variability with this measurement system was shown to be small in a previous study.22 QT interval was measured from the onset of the QRS complex to the end of the T wave, defined as its return to the TP isoelectric baseline. The QT was measured to the nadir of the curve between the T and U waves when U waves were present. Whenever possible, the average measurement of three complexes for each lead was taken. If the end of the Twave could not be reliably determined, or when the T wave was isoelectric or of very low amplitude, the QT measurement was not made in that lead, and it was excluded from analysis. A lower limit of five or more technically adequate leads per electrocardiogram was required for inclusion in this study. The duration of the QRS was also measured and averaged in analyzable lead. JT was calculated after the QT was subtracted from the corresponding QRS duration. QTd was defined as the difference between the maximum and minimum QT interval measurements among all of the measured (12) leads on the standard electrocardiogram. QTc (or heart rate-corrected QT interval) was calculated according to Bazett’s formula30,31 as follows: QTc = QT/square root of the R-R interval. QTc, JT, and JTc dispersion were calculated in a manner similar to QTd.
Statistics Results are expressed as mean ± SD unless stated otherwise. Significant univariate associations of QTd with the fol-
American Heart Journal Volume 135, Number 6, Part 1
Karagounis et al. 1029
Table I. Patient characteristics and variables of interest n = 207 57 ± 11 24 81 ± 16 34 42 14 14
Demographics Age (yr, mean ± SD) Sex (% female) Weight (kg, mean ± SD) History of angina (%) History of hypertension (%) History of myocardial infarction (%) History of diabetes mellitus (%) Acute myocardial infarction variables Time lapsed from symptom onset to therapy (hr, mean ± SD) Anterior MI vs other (%) Sum of ST-segment elevation (mV, mean ± SD) No. of leads with ST elevation (mean ± SD) Alteplase (vs anistreplase) (%) Reinfarction (%) Peak CK (IU/L, mean ± SD) Time to peak CK (hours, mean ± SD) Angiographic data TIMI 2, 3 (vs 0, 1) (%) TIMI 3 (vs 0, 1 ,2) (%) Minimal lumen diameter (mm, mean ± SD) Baseline LVEF (%) No. of diseased arteries (%) One-vessel Two-vessel Three-vessel Interventions PTCA (%) CABG (%) Radionuclide findings at discharge Discharge EF (%) 53 ± 12 Discharge diastolic volume (ml, mean ± SD) Discharge systolic volume (ml, mean ± SD)
2.7 ± 0.9 37 1.1 ± 0.8 4.0 ± 1.6 51 3 2315 ± 1885 11.6 ± 6.4 88 75 0.98 ± 0.52 54 ± 12.3 48 34 15 30 12 163 ± 33 78 ± 31
MI, Myocardial infarction; CK, creatine kinase; LVEF, left ventricular ejection fraction; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting.
lowing 24 variables were first identified: patient characteristics (age, sex, weight, history of angina, history of myocardial infarction, history of hypertension, history of diabetes mellitus, history of congestive heart failure), acute myocardial infarction variables (location, time from symptom onset to therapy, sum of ST elevation, number of leads with ST elevation, thrombolytic drug, reinfarction, peak creatine kinase, and time to peak), angiographic findings at 1 day (TIMI grade flow 0/1 vs 2/3, minimal lumen diameter, number of diseased arteries), interventions (angioplasty, bypass surgery) and radionuclide data at discharge (global left ventricular ejection fraction, end-diastolic and end-systolic volumes). To determine the multivariate independent associates between QTd and the initially identified univariate predictors, stepwise multivariate regression modeling was performed. A p value of ≤0.1 was required to allow a factor to enter the model, and p of ≥0.15 was required to subsequently remove a factor. (We chose a p ≤ 0.1 because we wished to identify all potentially important determinants of QTd.32) Similar analyses were performed used as dependent variables QTc, JT, and JTc dispersion, each in separate regression models. Analyses were
performed with SPSS 6.1 (SPSS Inc., Chicago, Ill.) statistical software for Macintosh (Apple Computer, Cupertino, Calif.). A p value <0.05 was considered significant.
Results Patient characteristics Of the 325 patients enrolled in TEAM-3, 118 patients were excluded because of the following reasons: no discharge electrocardiograms (n = 29), poor quality electrocardiograms (n = 12), unavailable records (n = 3), no angiographic scoring done (n = 20), presence of bundle branch block (n = 24), grafted vessels (n = 11), atrial fibrillation (n = 3), or no discharge radionuclide left ventriculogram (n = 16). The remaining 207 patients satisfied the entry criteria for this study, including 158 men and 49 women with a mean age of 57.1 ± 11 years. Patients were given alteplase (n = 106) or anistreplase (n = 101) 2.7 ± 0.9 hours from symptom onset. Table I summarizes other patient characteristics and variables of interest.
American Heart Journal June 1998
1030 Karagounis et al.
Table II. Comparisons of QTd by perfusion grade TIMI perfusion grade
No. of patients QTd (msec) QTcd (msec)
0
1
2
3
p Value*
16 74 ± 25 85 ± 28
8 74 ± 18 80 ± 16
27 57 ± 30 66 ± 34
156 50 ± 20 57 ± 23
≤0.0001 ≤0.0001
*Analysis of variance.
Table III. Comparison of QTd by perfusion grade TIMI perfusion grade
No. of patients QTd (msec) QTcd (msec)
TIMI perfusion grade
0, 1
2, 3
p Value
0, 1, 2
3
p Value*
24 74 ± 23 84 ± 24
183 51 ± 22 59 ± 25
≤0.0001 ≤0.0001
51 65 ± 28 74 ± 31
156 50 ± 20 57 ± 23
0.0001 ≤0.0001
*Analysis of variance.
Angiographic and left ventricular function data All 207 patients had coronary angiography done at 32.3 ± 29.8 hours after thrombolytic therapy. The infarct-related lesion was identified within the right circumflex coronary artery in 129 (62%) patients and the left anterior descending coronary artery in 76 (38%) patients. Two patients had an undefined infarct-related artery and were assigned a TIMI perfusion grade of 3. A total of 156 (75.4%), 27 (13.0%), 8 (3.9%), and 16 (7.7%) patients had TIMI perfusion grades 3, 2, 1, and 0, respectively. The total patency rate (defined as TIMI grades 2 and 3) was high (88%) in these 207 patients and was not significantly different between alteplase and anistreplase (p = 0.75). Radionuclide ventriculography was performed 7.7 ± 2.9 days after thrombolytic therapy.
ECG data ECG was performed at 8 ± 4 days from symptom onset. All 12 leads were measurable in 39 (18.8%) patients. In the remaining 168 (81%) patients the QT interval was measurable in 9.6 ± 1.5 leads. The mean QT interval averaged over all 12 leads was 395 ± 37 msec (range 302 to 481 msec). The mean QTc interval was 430 ± 29 msec (range 341 to 533 msec). The maximum QT interval was 421 ± 39 msec (range 315 to 517 msec) and occurred in one of the precordial leads in 117 (57%) patients and in one of the limb leads in 90 (44%) patients. The maximum QTc was 459 ± 33
msec (range 360 to 585 msec). The minimum QT interval was 367 ± 40 msec (range 261 to 460 msec) and occurred in one of the precordial leads in 102 (49%) patients and in one of the limb leads in 105 (51%). The minimum QTc was 397 ± 32 msec (range 298 to 509 msec). The average QTd for this study group was 54 ± 23 msec (range 14 to 152 msec), and the average QTc dispersion was 62 ± 26 msec (range 15 to 173 msec).
QTd and perfusion status Table II shows QT and QTc dispersion by perfusion grade in all patients. Significant differences were seen in QT and QTc dispersion among the TIMI perfusion groups (p ≤ 0.0001). There was also a gradual decrease in QT and QTc dispersion with improving (increasing) TIMI perfusion grades. These results did not change when the patients treated with alteplase and anistreplase were analyzed separately. There was a trend to less mean QT and QTc dispersion in patients treated with alteplase than in those treated with anistreplase, but statistical significance was not reached. Table III shows the correlation of QT and QTc dispersion with perfusion grade in all patients when comparing TIMI grades 0/1 versus 2/3 and TIMI grades 0/1/2 versus 3. Significant differences were seen in QT and QTc dispersion between TIMI grades 0, 1 versus 2/3, and between TIMI grades 0/1/2 and 3 (p ≤ 0.0001 for both comparisons). These results did not change when
American Heart Journal Volume 135, Number 6, Part 1
Karagounis et al. 1031
Table IV. Univariate associations of QTd Variable Patient characteristics Age (yr) Sex Weight History of angina History of hypertension History of myocardial infarction History of diabetes mellitus Acute myocardial infarction variables Time lapsed from symptom onset to therapy Anterior MI vs other ∑ST-segment elevation (mm) No. of leads with ST elevation Alteplase vs anistreplase Reinfarction Peak CK Time to peak CK Angiographic data TIMI 0, 1 vs 2, 3 TIMI 0, 1, 2 vs 3 Minimal lumen diameter (mm) Infarct-related artery: LAD vs other No. of diseased arteries Interventions PTCA CABG Radionuclide data at discharge Discharge EF Discharge end-diastolic volume Discharge end-systolic volume
Coefficient ± SEM
Correlation
F
p Value
0.14 ± 0.15 0.06 ± 3.73 –0.08 ± 0.10 4.27 ± 3.33 0.32 ± 3.22 –0.81 ± 4.64 3.35 ± 4.57
0.06 0.00 0.05 0.09 0.01 0.01 0.05
0.85 0.00 0.60 1.64 0.01 0.03 0.54
0.36 0.99 0.44 0.20 0.92 0.86 0.46
–1.08 ± 1.85 4.57 ± 3.28 3.26 ± 1.97 1.94 ± 0.97 –4.48 ± 3.16 21.98 ± 8.87 0.53 ± 0.01 0.47 ± .25
0.04 0.10 0.12 0.14 0.09 0.18 0.05 0.13
0.34 1.94 2.75 4.02 2.00 6.14 0.53 3.55
0.56 0.17 0.10 0.05 0.16 0.01 0.47 0.06
22.66 ± 4.70 14.62 ± 3.54 –8.24 ± 3.00 5.26 ± 3.27 3.20 ± 2.09
0.32 0.28 0.19 0.11 0.11
23.21 17.03 7.55 2.60 2.36
–1.03 5.23 ± 4.86
0.02 0.07
0.09 1.16
0.77 0.28
–0.37 ± 0.14 0.62 ± 0.05 0.11 ± 0.05
0.19 0.09 0.15
7.48 1.72 4.51
0.007 0.19 0.04
0.0000 0.0001 0.007 0.11 0.13
F, F statistic; EF, left ventricular ejection fraction; MI, myocardial infarction; CK, creatine kinase; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting.
the patients treated with anistreplase and streptokinase were analyzed separately.
Univariate and multivariate associates of QTd Table IV shows the univariate and Table V the multivariate associations with QTd. Variables that met the significance level of p ≤ 0.1 (∑ST elevation, number of leads with ST elevation, reinfarction, time to peak creatine kinase, TIMI grade flow 0/1 vs 2/3, minimal lumen diameter, reinfarction, left ventricular ejection fraction at discharge, and end-systolic volume) were further subjected to multivariate modeling to determine their independent association with QT dispersion. The most significant multivariate associate was found to be the TIMI grade (p ≤ 0.0001) followed by reinfarction (p = 0.005) and left ventricular ejection fraction (p = 0.02). Similarly, in separate models where QTc, JT, and JTc dispersion were used as the dependent variable, TIMI grade also emerged as the strongest associated factor (p ≤ 0.0001 in all three models). The significance of the association
was not diminished when TIMI grade was dichotomized into grades 0/1/2 versus 3 (Table VI): p ≤ 0.0001 for QT, QTc, JT, or JTc dispersion in separate models.
Discussion Summary of study findings In this study we sought to determine the significant associates of dispersion in ventricular repolarization (the difference between maximum and minimum QT interval measurements in any of the 12 leads on a standard electrocardiogram) in patients who had acute myocardial infarction and were treated with thrombolytics. We analyzed 24 variables describing patient clinical characteristics, myocardial infarction characteristics, and angiographic and ventriculographic parameters. We found by multivariate regression modeling that the perfusion status of the infarct-related artery, TIMI grade 0/1 vs 2/3 flow (0/1 = 75 msec, 2/3 = 53 msec, p < 0.0001), was the most important associate of
American Heart Journal June 1998
1032 Karagounis et al.
Table V. Multivariate associations of QTd by TIMI 0, 1 vs 2, 3 Regression model parameters
Variable
Coefficient ± SEM
p Value
R = 0.40 R2 = 0.16 F = 12.68, p = 0.0000
TIMI grade 0, 1 vs 2,3 Reinfarction Discharge EF
–22.16 ± 4.62 24.76 ± 8.75 –0.31 ± 0.13
0.0000 0.005 0.02
Variable
Coefficient ± SEM
p Value
TIMI grade 0, 1, 2 vs 3 Reinfarction Discharge EF
–13.43 ± 3.53 23.16 ± 8.91 –0.31 ± 0.13
0.0002 0.01 0.03
F, F statistic of regression model; R, multiple correlation coefficient; R2, coefficient of variation.
Table VI. Multivariate associations of QTd by TIMI 0, 1 vs 2, 3 Regression model parameters R = 0.35 R2 = 0.12 F = 9.61, p = 0.0000 Abbreviations as in Table V.
QTd, followed by whether reinfarction (yes = 74 msec, no = 51 msec, p = 0.005) occurred during hospitalization and by left ventricular ejection fraction at discharge (p = 0.02). These findings suggest that successful and sustained restoration of blood flow leads to reduced dispersion of repolarization.
Comparison with previous work Our group first reported the association of TIMI grade flow and QTd in an earlier study of streptokinase and anistreplase of acute myocardial infarction.22 In that study patients with TIMI grade 2/3 had lesser QTd on their postmyocardial infarction 12-lead ECG than did those with grade 0/1 flow (54 ± 20 vs 94 ± 29 msec, p = 0.0001). This association was independent of age, sex, drug assignment, infarct site, infarctrelated vessel, and number of measurable leads. However, left ventricular function was not measured, and the degree of independence of QTd and cardiac dysfunction could not be ensured. This study validates the association of TIMI flow grade and QTd and confirms its independence from patient and infarct characteristics. It is important that this study adds to the earlier study in showing that perfusion grade continues to be a strong predictor of convalesce QTd even after accounting for differences in left ventricular function (ejection fraction, end-systolic and end-diastolic volumes). This study also differs in assessing patency at “plateau” (1 day), after the ongoing processes of early reperfusion and reocclusion largely have reached equilibrium.
Mechanisms of thrombolytic therapy benefits The main goal of thrombolytic therapy in patients with acute myocardial infarction is the establishment and maintenance of coronary patency to improve left ventricular function and decrease mortality. The use of thrombolytic agents has been documented to reduce the mortality rate after infarction; the mechanisms by which these benefits are mediated appear to include improvements both in mechanical and electrical function.18-21,33 In this study of patients who had infarction we have shown that successful thrombolysis is associated with less dispersion in QT interval measurements. There also was an association of left ventricular function with QTd of lesser magnitude, suggesting that the effects of successful reperfusion on QTd are independent of those leading to the preservation of left ventricular function. Also, patients with reinfarction representing late failure of reperfusion had a greater dispersion than those without. Thus by establishing sustained patency, thrombolytic therapy may prevent or attenuate the development of an abnormal electrophysiologic substrate after myocardial infarction. Our results thus suggest that reduction in QTd may be an additional mechanism of the benefit of thrombolytic therapy. It should be pointed out that the predictive value for QTd of our multivariate model was of modest strength, because it was measured by the coefficient of variation (Table V, R2 = 0.16). The inability to achieve a better prediction of QTd indicates that important additional unknown factor(s) affect ventricular repolarization (such as autonomic influences), or the accuracy of our
American Heart Journal Volume 135, Number 6, Part 1
method of measuring QTd may be limited (e.g., by significant “noise” contaminating the true signal).
Dispersion of repolarization and arrhythmogenesis Dispersion of refractoriness is a measure of nonhomogeneous recovery of excitability in a given mass of cardiac tissue and is determined by the combined effects of the differences in activation times and action potential durations. A great deal of experimental work has substantiated the association between dispersion of repolarization and malignant ventricular tachyarrhythmias. Han and Moe34 were the first to demonstrate (in an animal model) the relation between dispersion of refractoriness and a decrease in fibrillation threshold under diverse experimental conditions. Kuo et al.7 proposed three possible mechanisms (based on their own experiments) by which increased dispersion predisposed to arrhythmias: (1) focal reexcitation, (2) reentry facilitated by a conduction from an area with a shorter refractory period to an area with a longer refractory period, and (3) reentry facilitated by conduction from an area with a longer refractory period to an area with a shorter refractory period. In the studies of Restivo et al.35 of canine myocardial infarction, dispersion of effective refractory period in a normal heart averaged 30 msec and was dramatically increased in the infarcted heart to 160 msec. The reentry waveform was initiated by a premature complex that originated from the region of shortest refractoriness and was conducted towards regions with longer refractory periods. Vassallo et al.36 measured the dispersion of excitability in the left ventricle in three groups of patients. In seven individuals with normal hearts and no arrhythmias, the average dispersion of total recovery time was 52 msec. In six patients with previous myocardial infarction who had sustained ventricular tachycardia, the mean dispersion of total recovery time was 90 msec. In the five patients with long QT syndrome and previous cardiac arrest, dispersion was the longest, averaging 114 msec. It was concluded that nonuniform recovery of excitability in the left ventricle is caused by dispersion of refractoriness or activation time and depends on the pathophysiogic substrate. The link between QT variation across the 12 leads and regional differences of ventricular repolarization was provided by Cowan et al.,37 who were able to demonstrate that regional differences in ventricular recovery time were reflected by epicardial monophasic action potentials recordings during open heart surgery. More recently, Cao et al.38 found an excellent correla-
Karagounis et al. 1033
tion between surface QTd and monophasic action potential dispersion in rat hearts. Increased QTd was observed in patients with acute myocardial infarction who had ventricular tachyarrhythmias compared with those with a benign course.39,40 Increased dispersion also was noted in patients with ventricular tachyarrhythmias compared with control (normal) groups. Pye et al.16 found that in patients with sustained ventricular tachycardia, QTd was significantly longer (77 msec) than in those without (38 msec). Perkiomaki et al.41 showed a greater QTd in patients with cardiac arrest or monomorphic ventricular tachycardia (104 ± 41 msec) compared with that in healthy individuals (38 ± 14 msec) or that in patients with previous myocardial infarction but without a history of ventricular tachyarrhythmias (65 ± 31 msec). In multivariate analysis, among clinical and angiographic variables QTd was the strongest independent factor and most effectively separated the patient groups with and without susceptibility to ventricular tachyarrhythmias (p < 0.001). QTd is also increased in patients with long QT syndromes at risk of ventricular arrhythmias,8 patients with hypertrophic cardiomyopathy,12 and patients with chronic heart failure.13 To definitively establish the predictive value of QTd for mortality after myocardial infarction, a sizable prospective trial is required.
Limitations A major limitation of this study is that the QT interval is not measurable in every lead and is difficult to measure with precision, as recognized by previous investigators.42,43 However, the inability to measure QT interval in some leads will tend, if anything, to result in an underestimation rather than an overestimation of QTd. To ensure reliable QT measurements we applied a consistent method, and we used a reliable computerized system. The reader of the QT intervals was blinded to patient clinical characteristics and outcomes. Similarly, grading of TIMI perfusion was carried out without knowledge of patients’ clinical status or ECG findings. Manual measurement of QTd is a laborious process. Given the potential application of QTd as a noninvasive predictor of life-threatening ventricular tachyarrhythmias, computerized measurement of ventricular repolarization based on objective standardized criteria would add to the current capability. Finally, long-term outcomes of patients were not available in our study population to verify the predic-
1034 Karagounis et al.
tive value of QTd for clinical events (the study was too small for mortality assessment, even if available).
Conclusions In conclusion, the results of our study show that successful thrombolysis is the primary associate of QTd in patients after acute myocardial infarction. Our data support the hypothesis that QTd after myocardial infarction depends on reperfusion status and to a lesser degree on left ventricular function. Reduction in QTd may be an additional mechanism of benefit of successful thrombolytic therapy. Because homogeneity of recovery time is thought to protect against arrhythmias and dispersion of recovery time is thought to be arrhythmogenic, the reduction in QTd seen in our patients may be interpreted as a mechanism of the benefit of thrombolytic therapy.
References 1. Schwartz PJ, Wolf S. QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 1978;57:1074-7. 2. Ahnve S, Helmers C, Lundman T, Rehnqvist N, Sjogren A. QTc intervals in acute myocardial infarction: first-year prognostic implications. Clin Cardiol 1980;3:303-8. 3. Wheelan K, Mukharji J, Rude RE, et al. Sudden death and its relation to QT-interval prolongation after acute myocardial infarction: two-year follow-up. Am J Cardiol 1986;57:745-50. 4. Puddu PE, Bourassa MG. Prediction of sudden death from QTc interval prolongation in patients with chronic ischemic heart disease. J Electrocardiol 1986;19:203-12. 5. Algra A, Tijssen JGP, Roelandt JRTC, Pool J, Leibsen J. QTc prolongation measured by standard 12 lead electrocardiography is an independent risk factor for sudden death due to cardiac arrest. Circulation 1991;83:1888-94. 6. Schouten EG, Dekker JM, Meppelink P, et al. QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation 1991;84:1516. 7. Kuo CS, Munakata K, Reddy CP, Surawicz B. Characteristics and possible mechanism of ventricular arrhythmia dependent on the dispersion of action potential durations. Circulation 1983;67:1356-67. 8. Day CP, McComb JM, Campbell RFW. QT dispersion: an indicator of arrhythmia risk in patients with long QT intervals. Br Heart J 1990;63:342-4. 9. Linker NJ, Colonna P, Kekwick CA, Till J, Camm AJ, Ward DE. Assessment of QT dispersion in symptomatic patients with congenital long QT syndromes. Am J Cardiol 1992;69:634-8. 10. Hii JT, Wyse DG, Gillis AM, Duff HJ, Solylo MA, Mitchell LB. Precordial QT interval dispersion as a marker of torsade de pointes. Disparate effects of class Ia antiarrhythmic drugs and amiodarone. Circulation 1992;86:1376-82. 11. Miorelli M, Buja G, Melacini P, Fasoli G, Nava A. QT-interval variability in hypertrophic cardiomyopathy patients with cardiac arrest. Int J Cardiol 1994;45:121-7. 12. Barr CS, Naas A, Freeman M, Lang CC, Struthers AD. QT dispersion and sudden unexpected death in chronic heart failure. Lancet 1994; 343:327-9.
American Heart Journal June 1998
13. Buja G, Miorelli M, Turrini P, Melacini P, Nava A. Comparison of QT dispersion in hypertrophic cardiomyopathy between patients with and without ventricular arrhythmias and sudden death. Am J Cardiol 1993; 72:973-6. 14. Higham PD, Furniss SS, Campbell RW. QT dispersion and components of the QT interval in ischaemia and infarction. Br Heart J 1995; 73:32-6. 15. Jordaens L, Missault L, Pelleman G, Duprez D, De Backer G, Clement DL. Comparison of athletes with life-threatening ventricular arrhythmias with two groups of healthy athletes and a group of normal control subjects. Am J Cardiol 1994;74:1124-8. 16. Pye M, Quinn AC, Cobbe SM. QT interval dispersion: a non-invasive marker of susceptibility to arrhythmia in patients with sustained ventricular arrhythmias? Br Heart J 1994;71:511-4. 17. Zareba W, Moss AJ, le Cessie S. Dispersion of ventricular repolarization and arrhythmic cardiac death in coronary artery disease. Am J Cardiol 1994;74:550-3. 18. Sager PT, Perlmutter RA, Rosenfeld LE, McPherson CA, Wackers FTh, Batsford WP. Electrophysiologic effects of thrombolytic therapy in patients with transmural anterior myocardial infartion complicated by left ventricular aneurysm formation. J Am Coll Cardiol 1988;12:19-24. 19. Zaret BL, Wackers FJ, Terrin M, et al. Does left ventricular ejection fraction following thrombolytic therapy have the same prognostic impact described in the prethrombolytic era? Results of the TIMI II Trial. J Am Coll Cardiol 1991;17:214A. 20. Gang ES, Lew AS, Hong M, Wang FZ, Liebert CA, Peter T. Decreased incidence of ventricular late potentials after successful thrombolytic therapy for acute myocardial infarction. N Engl J Med 1989;321:712-6. 21. Vatterrott PJ, Hammil SC, Bailey KR, Wiltgen CM, Gersh BJ. Late potentials on signal-averaged electrocardiograms and patency of the infarct-related artery in survivors of acute myocardial infarction. J Am Coll Cardiol 1991;17:330-7. 22. Moreno FL, Villanuela MT, Karagouis LA, Anderson JL. Reduction in QT interval dispersion by successful thrombolytic therapy in acute myocardial infarction. Circulation 1994;90:94-100. 23. Anderson JL, Becker LC, Sorensen S, et al. Anistreplase versus alteplase in acute myocardial infarction: comparative effects on left ventricular function, 1 day patency, and morbidity: the TEAM-3. J Am Coll Cardiol 1992;20:753-66. 24. Chesebro JH, Knatterud G, Roberts R, et al. Thrombolysis in myocardial infarction (TIMI) trial, phase I: a comparison between intravenous plasminogen activator and intravenous streptokinase. Circulation 1987; 76:142-54. 25. Hackworthy RA, Sorensen SG, Fitzpatrick PB, et al. Dependence of assessment of coronary artery reperfusion on angiographic criteria and interobserver variability. Am J Cardiol 1988;62:538-42. 26. Moreno FLl, Stoops KL, Hackworthy RA, Menlove RL, van Bree R, Anderson JL. Quantification of rate of coronary artery disease progression by a new method of angiographic analysis. Am Heart J 1991; 121:1062-70. 27. Brown BG, Bolson E, Frimer M, Dodge HT. Quantitative coronary arteriography. 1977;55:329-37. 28. Pavel DG, Zimmer AM, Patterson VN. In-vivo labeling of red blood cells with 99mTc: a new approach to blood pool visualization. Nucl Med 1977;18:305-8. 29. Callahan RJ, 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;23:315-8.
American Heart Journal Volume 135, Number 6, Part 1
30. Bazett H. An analysis of the time relationships of the heart. Heart 1920;7:353-70. 31. Ahnve S. Correction of the QT interval for heart rate: review of different formulas and the use of Bazett’s formula in myocardial infarction. Am Heart J 1985;109:568-74. 32. Kleinbaum DG, Kupper LL, Muller KE. Selecting the best regression equation. Applied regression analysis and other multivariate methods. 2nd ed. Boston: PWS-Kent; 1988. p. 314-40. The Duxbury series in statistics and decision sciences. 33. Califf RM, Topol EJ, Gersh BJ. From myocardial salvage to patient salvage in acute myocardial infarction: the role of reperfusion therapy. J Am Coll Card 1989;14:1382-8. 34. Han J, Moe GK. Nonuniform recovery of excitability in ventricular muscle. 1964;14:44-60. 35. Restivo M, Gough WB, El-Sherif N. Reentrant ventricular rhythms in the late myocardial infarction period: prevention of reentry by dual stimulation during basic rhythm. Circulation 1988;77:429. 36. Vassallo JA, Cassidy DM, Kindwall KE, Marchlinski FE, Josephson ME. Nonuniform recovery of excitability in the left ventricle. Circulation 1988;78:1365-72.
Karagounis et al. 1035
37. Cowan JC, Hilton CJ, Griffiths CJ, et al. Sequence of epicardial repolarization and configuration of T wave. Br Heart J 1988;60:424-33. 38. Cao W, Ranade V, Jaffrani N, Khosla S, Somberg JC. Relationship between QT dispersion and action potential duration. J of Investigative Medicine 1996;43:430A. 39. Higham PD, Furniss SS, Campell RWF. QT dispersion in ischaemia and infarction: a marker of primary ventricular fibrillation? (Abstract). Eur Heart J 1991;12:87. 40. Potratz J, Djonlagic H, Mentzel H. Prognostic significance of QT dispersion in patients with acute myocardial infarction. (Abstract) Eur Heart J 1993;14:254. 41. Perkiomaki JS, Koistinen MJ, Yli Mayry S, Huikuri HV. Dispersion of QT interval in patients with and without susceptibility to ventricular tachyarrhythmias after previous myocardial infarction. J Am Coll Cardiol 1995;26:174-9. 42. Lepeschkin E, Surawicz B. The measurement of the QT interval of the electrocardiogram. Circulation 1952;6:378-88. 43. Ahnve S. Errors in the visual determination of corrected QT (QTc) interval during myocardial infarction. J Am Coll Cardiol 1985; 5:699-702.