Relationship between QT dispersion and the incidence of early ventricular arrhythmias in patients with acute myocardial infarction

International Journal of Cardiology 62 (1997) 211–216

Relationship between QT dispersion and the incidence of early ventricular arrhythmias in patients with acute myocardial infarction ˇ ´ Teodora Zaninovic-Jurjevic, ´ ´ Ante Matana, Nikola Bradic´ Luka Zaputovic´ *, Zarko Mavric, Department of Internal Medicine, Division of Cardiology, Clinical Hospital Centre Rijeka, and University of Rijeka School of Medicine, Rijeka, Croatia Received 11 June 1997; received in revised form 3 September 1997; accepted 3 September 1997

Abstract There is controversy about the influence of QT dispersion on the incidence of early ventricular arrhythmias in patients with acute myocardial infarction (AMI). The QT and QTc dispersion (QTd, QTcd) between two groups of patients with AMI were compared: 39 patients with early sustained ventricular tachycardia or ventricular fibrillation (VT / VF) and 40 patients without such arrhythmias. QTd and QTcd were calculated from the admission and predischarge ECG, expressed as the difference between the maximum and minimum QT and QTc interval in 12 leads. The coefficient of variability was also calculated (VQT, VQTc). Groups did not differ significantly in age, incidence of previous infarction, Killip class, electrolyte status, infarct location, expected and final ECG infarct size, enzymatic infarct size, thrombolytic treatment and reperfusion rate, i.e., in variables that could influence the VT / VF occurrence. On admission, patients with VT / VF had significantly greater QTd (77623 vs 53627 ms, P,0.001) and QTcd (90629 vs 62628 ms, P,0.001); VQT and VQTc were also significantly higher. Although similar differences existed on predischarge ECG, they were smaller. The results indicate that QT dispersion varies during the illness, and that measurements of QT dispersion could be helpful in predicting serious ventricular arrhythmias.  1997 Elsevier Science Ireland Ltd. Keywords: QT dispersion; Ventricular arrhythmias; Acute myocardial infarction

1. Introduction Acute myocardial infarction (AMI) is the leading cause of death in developed countries [1]. Of all deaths due to AMI, more than 60% occur within the first few hours as a consequence of ventricular tachycardia and / or fibrillation (VT / VF) [2]. Therefore, it is of great clinical interest to procure simple and noninvasive measurements of myocardial electrical instability which could predict such fatal events. It is well known that the prolonged QT interval in

*Corresponding author. Tel.: 1385 51 218059; fax: 1385 51 218059.

patients with AMI reflects abnormal ventricular repolarization and is associated with higher risk of lifethreatening arrhythmias [3,4]. More recently, the dispersion of QT interval, defined as the interlead variability of QT interval in surface ECG, was found to be the measure of regional differences in ventricular repolarization important in the pathogenesis of ventricular arrhythmias [5–8]. Although several studies reported that the measurement of QT dispersion is of value in the prediction of serious ventricular arrhythmias in patients with AMI [9–12], some authors reached negative conclusions [13,14]. Because of this controversy [15], we tried to determine the relationship between the QT dispersion

0167-5273 / 97 / $17.00  1997 Elsevier Science Ireland Ltd. All rights reserved. PII S0167-5273( 97 )00226-X

212

L. Zaputovic´ et al. / International Journal of Cardiology 62 (1997) 211 – 216

and the incidence of early ventricular arrhythmias in patients with AMI.

2. Methods All patients with AMI admitted to our coronary care unit during 1996 were considered for inclusion in the study. Patients with atrial fibrillation or flutter, ventricular hypertrophy, preexcitation, intraventricular conduction abnormalities, permanent ventricular pacing, those treated with Class I or III antiarrhythmics and patients who died during hospitalization, were excluded. According to the VT / VF occurrence, two groups of patients were formed: 39 patients with early sustained VT or VF (within 24 h of the infarct onset), and 40 patients randomized (using randomnumber tables) for the control group among the patients without such arrhythmias. The electrocardiographic diagnosis of VT was established by the occurrence of a series of three or more consecutive, bizarrely shaped premature ventricular complexes whose duration exceeded 120 ms, with the ST-T vector pointing opposite to the major QRS deflection, and the rate over 120 beats min 21 . The sustained VT was defined as VT lasting over 30 s or requiring termination because of hemodynamic collapse [16]. The VF was recognized by the presence of irregular undulations of varying contour and amplitude, without possible distinction of QRS complexes, ST segments and T waves, and always accompanied by hemodynamic collapse [16]. The QT intervals were measured by two independent investigators in the 12-lead surface ECGs recorded at a paper speed of 50 mm s 21 , with a maximal resolution level of 10 ms. The investigators were blinded to all clinical data. The end of the T wave was defined as a point at which the line of the maximal downslope of the T wave crosses the baseline and, if the U wave was present, as the nadir between the T and the U wave. The QTc interval was calculated according to the Bazett’s formula [17]. The QT dispersion (QTd) and QTc dispersion (QTcd) were calculated from the admission and predischarge ECG, expressed in milliseconds as the difference between maximum and minimum QT and QTc interval in 12 leads. The coefficient of vari-

ability (V) for each variable (VQT, VQTc) was also calculated according to the formula: V5(standard deviation / arithmetic mean)3100. The arithmetic mean of two values obtained by the investigators for each ECG variable in the same patient was used for further statistical analysis. All clinical, laboratory and ECG variables that could affect the QT interval or VT / VF occurrence were followed and compared between the groups: age, gender, incidence of previous infarction, Killip class [18], electrolyte status (K, Ca, Mg), infarct location, expected and final ECG infarct size, enzymatic infarct size, thrombolytic treatment and reperfusion rate achieved by thrombolysis. The expected ECG infarct size was calculated from the admission ECG by previously developed formulas based on the initial ST segment elevation [19]. The final ECG infarct size was determined by the Selvester’s QRS score on the predischarge ECG [20]. The enzymatic infarct size was computed by the integration of the creatine kinase (CK) time-activity curve, using originally developed software based on a previously established mathematical model [21,22]. The expected and final ECG infarct size were expressed as a percentage of the left ventricular mass, and the enzymatic infarct size as a cumulative CK in blood in IU l 21 . Continuous variables were expressed as mean6SD. Differences between the groups were tested by Student’s t-test or Mann–Whitney’s test, as appropriate. Discrete variables were compared between the groups by the chi-square analysis or the Fisher’s exact test, as appropriate. Differences were considered significant at P,0.05; P values for two-tailed tests were used.

3. Results During 1996, 424 patients with AMI were admitted and treated in our coronary care unit. Among them, 48 patients suffered from early malignant ventricular arrhythmias (11.3%). There were 38 patients with sustained VT (9%), and ten patients with VF (2.3%). In four patients with VT, arrhythmia termination was achieved by intravenous lidocaine (10.5% success rate); others were successfully treated with DC cardioversion. Nine patients with early VT / VF were

L. Zaputovic´ et al. / International Journal of Cardiology 62 (1997) 211 – 216

excluded from the study: two with extensive AMI died during the hospitalization, others were excluded according to previously described exclusion criteria. Since 32 patients with sustained VT and seven patients with VF included in the study did not differ significantly in any analysed variable, their data were pooled for further analysis. Among 293 patients without VT / VF and any of the exclusion criteria 40 were randomized for the control group. The clinical characteristics of the patients with and without VT / VF are shown in Table 1. The two groups did not differ significantly in age, the incidence of previous infarction, Killip class on admission, frequency of thrombolytic treatment and reperfusion rate achieved by thrombolysis. The only significant difference was the greater proportion of male gender in the arrhythmia group (90 vs 68%, P50.03). There was no significant difference in the electrolyte status (K, Ca, Mg) on admission; all mean values were within the normal range (Table 2). No significant difference in the electrolyte status existed neither prior to hospital discharge. Comparing the two groups of patients, we found differences neither in the infarct location nor in the expected and final ECG infarct size. Although the enzymatic infarct size was greater in the arrhythmia group (37276942 vs 34356747 IU l 21 ), the difference was not significant (Table 2). Patients with AMI and early VT / VF had a significantly longer QT interval on admission ECG in comparison with the control group (366640 vs 347641 ms, P50.04) (Table 3). These patients had significant QT and QTc dispersion on admission

213

Table 2 Electrocardiographic and laboratory findings VT / VF N539

No arrhythmia N540

P

Electrolyte status on admission (mmol l 21 ) K 4.160.5 Ca 2.360.2 Mg 0.960.1

4.260.4 2.360.2 0.960.1

NS

Infarct location Anterior Inferior Non Q

19 (48%) 20 (50%) 1 (2%)

NS

ECG infarct size (% of LV mass) Expected 2267 Final 1766

2064 1665

NS NS

Enzymatic infarct size (cumulative CK in IU l 21 )

34356747

NS

21 (54%) 16 (41%) 2 (5%)

37276942

Table 3 223QT interval and QT dispersion on admission ECG VT / VF N539

No arrhythmia N540

P

QT interval QTc interval (ms)

366640 396629

347641 387623

0.04 NS

QT dispersion QTc dispersion (ms)

77623 90629

53627 62628

,0.001 ,0.001

Coefficient of variability (V) VQT interval 6.561.8 VQTc interval (%) 7.061.9

5.562.3 5.762.3

0.04 0.007

(.70 ms), significantly greater than the patients without arrhythmias (77623 vs 53627 ms, P,0.001, and 90629 vs 62628 ms, P,0.001, respectively) (Table 3). The variability of the QT and QTc interval (VQT and VQTc) was also significantly greater in the

Table 1 Clinical characteristics of the patients Arrhythmia (VT / VF) N539 (32 / 7)

No arrhythmia N540

Age (years) Male gender Previous infarction

59612 35 (90%) 4 (10%)

62611 27 (68%) 3 (8%)

Killip class on admission I II III IV

31 (80%) 6 (15%) 2 (5%) 0

34 (85%) 6 (15%) 0 0

Thrombolytic treatment Reperfusion

8 (21%) 5 / 8 (62%)

7 (18%) 4 / 7 (57%)

P NS 0.03 NS

NS

NS NS

L. Zaputovic´ et al. / International Journal of Cardiology 62 (1997) 211 – 216

214

Table 4 QT interval and QT dispersion on predischarge ECG VT / VF N539

No arrhythmia N540

P

QT interval QTc interval (ms)

384642 400627

370635 389632

NS NS

QT dispersion QTc dispersion (ms)

69626 79629

52625 62626

0.007 0.006

5.462.4 5.662.2

NS NS

Coefficient of variability (V) VQT interval 6.462.6 VQTc interval (%) 6.662.6

arrhythmia group (6.561.8 vs 5.562.3%, P50.04, and 7.061.9 vs 5.762.3%, P50.007, respectively). Although similar differences in the QT interval duration and dispersion between the two groups of patients existed on predischarge ECG, they were smaller (Table 4). The QTd and QTcd were still significantly greater in the arrhythmia group (69626 vs 52625 ms, P50.007, and 79629 vs 62626 ms, P50.006, respectively). For other ECG variables (i.e., QT, QTc and VQT, VQTc) the differences were no more significant (Table 4).

4. Discussion The QT interval does not exactly reflect the ventricular depolarization and repolarization time, because in all portions of the ventricles repolarization is not finished at the time when the end of the QT interval is recorded. These slight physiological differences in regional ventricular recovery time cannot be registered in the surface ECG and are not of clinical relevance. In patients with AMI, regional ischemia, combined with increased sympathetic activity, causes enlarged spacial and temporal dispersion of repolarization, which may be responsible for intraventricular reentry phenomena and VT / VF occurrence [23]. Such patients usually show a prolongation of the QT interval and its interlead variability in the surface ECG [23,24]. Therefore, it is logical and expected that the significant QT prolongation and dispersion should be associated with the greater incidence of VT / VF in patients with AMI, and that these variables should be a predictor of these arrhythmias. Why then were the obtained results in earlier studies contradictory?

Certainly, not only one explanation exists, although most could be related to the study design and methodology used. In this study the analysis was performed on a consecutive series of patients with AMI. Patients with early VF and those with early sustained VT were first analysed separately. Since there were no significant differences in QT interval and QT dispersion between these two sets of patients, and because there were only seven patients with VF, we considered it appropriate to pool their data for further analysis. Of all patients without VT / VF and any of the exclusion criteria 40 were randomized to form an adequate control group. Some authors have chosen patients with and without VT / VF by matching them in variables that could influence the arrhythmia occurrence [13]. Such a selection could have influenced their results. Since the precise determination of the end of T wave in atrial fibrillation or flutter is impossible, and because of the influence of the cyclelength variability on the QT interval, we excluded such patients from the study. All conditions with secondary changes of repolarization (i.e., ventricular hypertrophy, preexcitation, intraventricular conduction abnormalities and permanent ventricular pacing) were also excluded. The Selvester’s QRS scoring system for the ECG estimation of the infarct size was also not applicable in these patients [20]. Because of the possible drug influence on QT interval, three patients treated with Class I or III antiarrhythmics were also excluded. In spite of the many conditions that can make the QT analysis inadequate, only seven patients with VT / VF were excluded from the study for such a reason (18%). The methodology of QT interval measurements is also of great importance. It was found that QT dispersion determined in ECG tracings recorded at a paper speed of 50 or 100 mm s 21 was less variable than that obtained from recordings taken at a speed of 25 mm s 21 . The intraobserver and interobserver variability of QT dispersion measurements also exists [25,26]. We have, therefore, used the arithmetic mean of two values obtained by two independent investigators, blinded to all clinical data, for each ECG variable in the same patient, and a paper speed of 50 mm s 21 . The interobserver variability was also analysed, and was found to be low (within 5%). Comparing the two groups of patients with AMI

L. Zaputovic´ et al. / International Journal of Cardiology 62 (1997) 211 – 216

we noticed no significant differences in all variables reflecting the extent of jeopardized myocardium and / or left ventricular function. Therefore, it seems that only primary electrophysiological abnormalities of repolarization influenced the VT / VF occurrence. In addition to the direct effect of ischemia on the myocardium, it is possible that reperfusion after transient ischemia causes serious ventricular arrhythmias [27,28]. Such proarrhythmic effect of the reperfusion on jeopardized myocardium could be explained by the reperfusion injury. On the other hand, some studies described the beneficial effect of reperfusion on QT dispersion [29,30]. Because of the small number of patients treated and reperfused with thrombolytic therapy in both groups, we made no such analysis, neither was this the goal of our study. As there were no significant differences in thrombolytic treatment and reperfusion rate achieved by thrombolysis in the compared groups of patients (Table 1), it seems that thrombolytic therapy did not influence our results. A great proportion of patients were treated with beta-blocker on discharge. Although a greater proportion of beta-blocker therapy existed in the arrhythmia group, the difference was not significant (83 vs 76%, P5NS). The sex difference between the patients with VT / VF and the control group could be a potential confounding factor. Regarding the physiologically shorter QT interval in males than in females, the greater proportion of male gender could have shortened the mean QT interval in the arrhythmia group. On the contrary, this QT interval was longer on both admission and predischarge ECG, reflecting electrophysiological abnormalities in these patients. In spite of the sex influence on QT interval duration, it is not expected that sex could have such an impact on QT dispersion. It could be argued that abnormalities of QT dispersion might merely reflect a long QT interval in the patients with VT / VF. Although prolonged QT interval and QT dispersion are both associated with AMI, QT prolongation is the consequence of a longer ventricular repolarization time, while QT dispersion reflects the regional differences in repolarization of the ischemic myocardium. In this study, the mean QT and QTc interval were longer in patients with VT / VF on both admission and predischarge ECG (Tables 3 and 4), but the difference was significant on the low

215

level (P50.04) only for the QT interval on admission. On the other hand, the differences in QT dispersion were highly significant for all analysed values (Tables 3 and 4). Therefore, it seems that dispersion of the QT interval, rather than the QT interval duration, could be more predictive for such arrhythmias. The results of QT dispersion analysis in our study indicate that QT dispersion varies during the AMI. The measurements of QT dispersion in patients with AMI could be helpful in the prediction of serious ventricular arrhythmias.

References [1] American Heart Association: 1990 heart facts. Dallas: American Heart Association National Center, 1990;1. [2] Myerburg RJ, Castellanos A. Cardiac arrest and sudden cardiac death. In: Braunwald E, editor. Heart disease: A textbook of cardiovascular medicine. Philadelphia, USA: W.B. Saunders, 1997; 742–79. [3] Laakso M, Aberg A, Savola J, Pentikainen PJ, Pyorala K. Diseases and drugs causing prolongation of the QT interval. Am J Cardiol 1987;59:862–5. [4] Kupersmith J. Long QT syndrome. In: Singer I, Kupersmith J, editors. Clinical manual of electrophysiology. Baltimore, USA: Williams and Wilkins, 1993; 143–68. [5] Day CP, McComb JM, Campbell RWF. QT dispersion: an indication of arrhythmia risk in patients with long QT intervals. Br Heart J 1990;63:342–4. [6] Pye M, Quinn AC, Cobbe SM. QT interval dispersion: a noninvasive marker of susceptibility to arrhythmia in patients with sustained ventricular arrhythmias?. Br Heart J 1993;71:511–4. [7] Higham PD, Campbell RWF. QT dispersion. Br Heart J 1994;71:508–10. [8] Zabel M, Franz MR, Klingenheben T, Hohnloser SH. QT dispersion in the 12-lead surface ECG reflects inhomogeneity in the terminal part of ventricular repolarization (abstract). PACE 1995;18:821. [9] Hohnloser SH, Van de Loo A, Arendts W, Zabel M, Just H. QT dispersion in the surface ECG as a parameter of increased electrical vulnerability in acute ischemia. Z Kardiol 1993;82:678–82. [10] Xiang H. The relationship between increased QT dispersion of acute myocardial infarction and ventricular fibrillation. Chung Hua Hsin Hsueh Kuan Ping Tsa Chih 1993;21:282–3. ¨ [11] Aksoyek S, Batur MK, Kabakci MG et al. QT dispersion in survivors of acute myocardial infarction: implications for lifethreatening arrhythmias. (abstract). PACE 1995;18:1125. [12] Higham PD, Furniss SS, Campbell RWF. QT dispersion and components of the QT interval in ischemia and infarction. Br Heart J 1995;73:32–6. [13] Leitch J, Basta M, Dobson A. QT dispersion does not predict early ventricular fibrillation after acute myocardial infarction. PACE 1995;18:45–8. [14] Fiol M, Marrugat J, Bergada J, Guindo J, Bayes de Luna A. QT dispersion and ventricular fibrillation in acute myocardial infarction (letter). Lancet 1995;346:1424–5.

216

L. Zaputovic´ et al. / International Journal of Cardiology 62 (1997) 211 – 216

[15] Surawicz B. Will QT dispersion play a role in clinical decisionmaking?. J Cardiovasc Electrophysiol 1996;7:738–59. [16] Zipes, DP. Specific arrhythmias: diagnosis and treatment. In Braunwald E, editor. Heart disease: A textbook of cardiovascular medicine. Philadelphia, USA: WB Saunders, 1997; 675–87. [17] Bazett HC. An analysis of the time-relations of electrocardiograms. Heart 1920;7:353–70. [18] Killip T, Kimball JT. Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients. Am J Cardiol 1967;20:457–64. [19] Aldrich HR, Wagner NB, Boswick J. Use of initial ST-segment deviation for prediction of final electrocardiographic size of acute myocardial infarcts. Am J Cardiol 1988;61:749–53. [20] Hindman NB, Schocken DD, Widmann M et al. Evaluation of a QRS scoring system for estimating myocardial infarct size. V. Specificity and method of application of the complete system. Am J Cardiol 1985;55:1485–90. [21] Roberts R, Henry PD, Sobel BE. An improved basis for enzymatic estimation of infarct size. Circulation 1975;52:743–54. ˇ Matana A, Bradic´ N. Electrocardiographic [22] Zaputovic´ L, Mavric´ Z, and enzymatic estimation of acute myocardial infarct size in the evaluation of therapeutic success after thrombolytic treatment. Cro Med J 1994;35:32–6. [23] Surawicz B. Electrophysiologic substrate of torsade de pointes: dispersion of repolarization or early afterdepolarizations?. J Am Coll Cardiol 1989;14:172–84.

[24] Singh S, Evans SJ, Rathod S, Insel L, Blumberg S, Bodenheimer M. QT dispersion and inducibility of ventricular tachycardia during programmed electric stimulation (abstract). PACE 1995;18:821. [25] Van de Loo A, Arendts W, Hohnloser SH. Variability of QT dispersion measurements in the surface electrocardiogram in patients with acute myocardial infarction and in normal subjects. Am J Cardiol 1994;74:1113–8. [26] Glancy JM, Weston PJ, Bhullar HK, Garratt CJ, Woods KL, de Bono DP. Reproducibility and automatic measurement of QT dispersion. Eur Heart J 1996;17:1035–9. [27] Kaplinsky E, Ogawa S, Michelson EL, Dreifus LS. Instantaneous and delayed ventricular arrhythmias after reperfusion of acutely ischemic myocardium: evidence for multiple mechanisms. Circulation 1981;63:333–40. [28] Kimura S, Bassett AL, Saoudi NC, Cameron JS, Kozlovskis PL, Myerburg RJ. Cellular electrophysiological changes and arrhythmias during experimental ischemia and reperfusion in isolated cat ventricular myocardium. J Am Coll Cardiol 1986;7:833–42. [29] Van de Loo A, Klingenheben T, Zabel M, Demmler M, Just H, Hohnloser SH. Changes in QT dispersion after acute myocardial infarction: relation to success of thrombolysis and to the occurrence of ventricular fibrillation (abstract). PACE 1994;17:752. [30] Moreno F, Villanueva T, Karagounis LA, Anderson JL. Reduction in QT interval dispersion by successful thrombolytic therapy in acute myocardial infarction. TEAM-2 Study Investigators. Circulation 1994;90:94–100.