Effects of thrombolysis and atenolol or metoprolol on the signal-averaged electrocardiogram after acute myocardial infarction

Effects of Thrombolysis and Atenolol or Metopro/o/ on the Signal-Averaged Electrocardiogram After Acute Myocardial Infarction Pietro Santarelli, Mp! Gaetano A. Lanza, MD, Francesco Biscione, MD, Andrea Natale, MD, Giancarlo Corslni, MD, Carmine Riccio, MD, Eraldo Occhetta, MD, Paolo Rossi, MD, Maurizio Gronda, MD, Johannes Makmur, MD, Marco Zanetta, MD, Umberto Parravicini, MD, and Salvatore Toscano, MD

Late potentials (LPs) detected on the signat-avep aged (SA) electrocardiogram (ECG) predict arrhythmic events after acute myocardial infarction (AMI). The effect of thrombolysis on the incidence of LPs after AMI is controversial and its impact on subsequent arrhythmic events is not known. More over, the effects of p blockers on the SAECG have not been studied. Six hundred eighteen patients with AMI were studied; thrombolysis was given to 228 (37%). In comparison with patients treated conventionally, those receiving thrombolysis were significantly younger and more frequently male, had higher peak values of creatine kinase, a lower prevalence of no-wave AMI, and a higher incidence of ventricular fibrillation in the acute phase, and more frequently received p blockers. An SAECG obtained 6 to 6 days after AMI showed LPs in 24% of patients receiving and in 25% not re ceiving thrombolysis (p q NS). On admission, iu travenous 9 blockers were administered to 110 patients (16%); those receiving (3 blockers were younger, had lower peak values of creatine kinase and more frequently received thrombolysis. LPs were less frequently found in patients treated than in those not treated with p blockers (15 vs 27%; p = 0.007); however, this effect was found only in those with an ejection fraction 240%. lmlependent predictors of LPs by mutivariate analysis were an ejection fraction *40% (p = 0.007), ventricular fibrillation in the acute phase (p q 0.02), and absence of P-blocking therapy (p = 0.03). During a mean followup of 12 f 7 months, there were 39 cardiac deaths (6%), 13 of which were sudden (2%), and 9 sustained ventricular tachycardias. Thrombolysis significantly reduced cardiac mortality but not the occurrence of arrhythmic events (3.9 vs 7.7%), whereas p blockers reduced cardiac mortality (1.6 vs 7.3%, p = From the Institute of Cardiology, Catholic University, Rome, and the Late Potentials Italian Study (LAPIS) sites (see Appendix), Italy. Manuscript received December 3 1, 1992; revised manuscript received April 9, 1993, and accepted April 12. Address for reprints: Pietro Santarelli, MD, Istituto di Cardiologia, Universita’ Cattolica de1 S. Cuore, L. go A. Gemelli, 8, 00168 Roma, Italy.

0.03) and reduced the incidence of arrhythmic events by 50% (p = NS). In conclusion, thrombolytic therapy does not reduce the incidence of LPs after AMI and its beneficial effects on prognosis are not related to a reduction of anhythmic events. Conversely, p blockers administered in the acute phase of the infarction, followed by chronic oral therapy, significantly reduce the incidence of LPs; this effect may explain, at least in part, the reduced incidence of arrhythmic events during folJow-up. (AmJ Cardiol1993;72:525-931)

T

he most frequent mechanism leading to sudden death after acute myocardial infarction (AMI) is considered to be ventricular tachycardia or fibrillation. It was shown that 70 to 90% of patients with episodesof ventricular tachycardia have low-amplitude, high-frequency signals, called late potentials (LPs) on the signal-averagedelectrocardiogram (SAEZCG).t-3In patients recovering from AMl, the presenceof LPs is a strong predictor of subsequent arrhythmic events.47 Thus, any intervention that reducesthe incidence of LPs may be expected to improve the outcome of these patients. Prognosis after AMI can be improved by thrombolytic therapys9 and p blockers.lO,lt The effects of thrombolysis on the SAEZCGare controversial,12-1sand those of p blockers are unknown. In addition, there are no studies that have systematically explored the relation between reduction of arrhythmic events by thrombolysis and p blockers, and the effects of these treatments on the incidence of LPs. In 1987, we initiated a prospective, multicenter study to determine the prognostic significance of LPs after AMI. Data from that study were retrospectively analyzed to: (1) assessthe effects of thrombolytic therapy and /3-blocking agentson the incidence of LPs after AMI, and (2) determine whether the improved survival observedafter thrombolysis and therapy with P-blocking agents could be attributed to a reduced incidence of LPs. METHODS Study patients: BetweenJuly 1987and March 1990,

702 patients admitted with the diagnosis of AMI to the THROMBOLYSIS,p BLOCKERSAND SIGNAL-AVERAGEDECG 525

TABLE I Main Clinical Data of Patients in the Study Group No. of patients Age (years) Age 160 years Peak creatine kinase (mU/ml) Site of infarction Anterior Inferior Non-Q-wave Previous infarction Killip 1 2 Ventricularfibrillation (acute phase) Sustained ventricular tachycardia (acute phase) p blockers Ejection fraction (%) Ejection fraction < 40%

618 61 t 12 342 (55%) 1,558 * 1,262 249 305 64 80

(40%) (49%) (11%) (13%)

514 104 32 21 110 45 117

(83%) (17%) (5%) (3%) (18%) 2 9 (20%)

coronary care units of the 6 institutions participating in the Late Potentials Italian Study (LAPIS) were recmited. Patients were included in the study if they had: (1) a conlirmed diagnosis of AMI on the basis of clinical (typical chest pain lasting >30 minutes), electrocardiographic (ST elevation Zl mV in the peripheral leads,and 22 mV in the precordial leads) and creatine kinase changes;and (2) an SAECG recorded before discharge. Patients were excluded from the study if: (1) they had cardiogenic shock, atrial fibrillation or pacemaker rhythm at the time of the SAECG; (2) they needed antiarrhythmic therapy at the time of the SAECG; or (3) the noise level of the SAECG was 20.5 p,,V.All patients gave informed consent. Fifty-four patients (8%) were lost to follow-up and were excluded from analysis. For the objectivesof the present study,30 patients (5%) who presentedin Killip class3 or 4 on admissionto the coronary care unit were also excluded, because they had clear contraindications to acute B-blocking treatment. Thus, the study population comprised 618 patients with a mean age of 61 + 12 years. The relevant clinical data of the study group included in the analysis are presented in Table I.

Treatment: All patients admitted to the coronary care unit within 6 hours from the onset of symptoms and with no contraindications to thrombolysis were administered intravenous streptokinase(1,500,OOO U in 60 minutes) or recombinant tissue-type plasminogen activator (a bolus of 10 mg, followed by 50 mg in 60 minutes, and another 40 mg in 120 minutes). In the absenceof specific contraindications, B-blocking agents were also given at the discretion of the attending physician; atenolol or metoprolol (5 mg in 10 minutes) was administered and then repeated after 10 minutes if heart rate was 260 beats/minor systolic blood pressure was 300 mm Hg, or both. After intravenous administration, 1 oral dose of 100 mgldie of atenolol, or 100 mg twice daily of metoprolol was given. Any other therapy, either .acute or chronic, was not standardized and was at the discretion of the attending physician. Therapeutic decisions were not obtained on the basis of the results of the SAECG. Signal-averaged electrocardiogram: All patients had an SAECG obtained 6 to 8 days from AMI. The SAECG was obtained with an Arrhythmia Research Technology-l@ or 1200 System,with bidirectional Butterworth filtering (25 to 250 Hz), as previously described.3The following quantitative SAECG variables of the filtered QRS were evaluated: (1) total duration, (2) root-mean-squarevoltage of the last 40 ms (RMS-40), and (3) duration of the low-amplitude signals (~40 p,V) in the terminal portion (LAS-40). Between 200 and 400 QRS complexes were averaged for each recording to reach a noise level <0.5 p\! LPs were consideredto be present when 22 of the following criteria were fulfilled: (1) total QRS duration >114 ms, (2) RMS-40 ~25 kV, and (3) LAS-40 >32 ms. In this study, 44 patients (7%) with bundle branch block patterns on the standardECG were also included (30 with right and 14 with left bundle branch block). In these patients, the SAECG was analyzed using different filters (40 to 250 Hz) and criteria, according to Buckingham et ali9 In this group, criteria for the presenceof

TABLE II Main Clinical Data of Patients With and Without Late Potentials on the Signal-Averaged Electrocardiogram Late Potentials (%I No. of patients Age (years) Age 260 years Sex (men) Peak creatine kinase (mu/ml) Site of infarction Anterior Inferior Non-Q-wave Previous infarction Killip 1 2 Ventricular fibrillation (acute phase) Sustained ventricular tachycardia (acute phase) p bjockers Thrombolysis Ejection fraction (%) Ejection fraction ~40%

526

THE AMERICAN JOURNAL OF CARDIOLOGY

VOLUME 72

152 61 87 126 1,886

124.5) ? 11 (57) (83) + 1,505

No Late Potentials (%) 466 60 255 368 1,421

(75.5) 2 12 (55) (79) lr 1,098

58 85 9 30

(38) (56) (6) (20)

191 220 55 50

(41) (47) (12) (11)

129 23 14 8 16 54 42 43

(85) (15) (9) (5) (10) (35) + 8 (30)

385 81 18 13 94 174 47 74

(83) (17) (4) (3) (20) (37) k 9 (17)

SEPTEMBER 1,1993

p Value NS NS NS 0

0.054 0.004 NS 0.01 NS 0.007 NS 0 0.0007

TABLE III

Main Clinical

Data of Patients

Treated

and Not Treated

Thrombolysis No. of patients Age (years) Age 2 60 years Sex (men) Peak creatine kinase (mu/ml) Site of infarction Anterior Inferior Non-Q-wave Previous infarction Killip

228 59? 110 193 1,748 t

1

2 Acute ventricular fibrillation (acute phase) Sustained ventricular tachycardia (acute phase) p blockers Ejection fraction (%) Ejection fraction <40%

TABLE IV Results of the Signal-Averaged Treated

(%I

with Thrombolysis

No Thrombolysis

(37) 11 (48) (86) 1,314

390 (63) 62 k 12 232 (59) 301 (77) 1,410 2 1,155

104 118 6 27

(45) (52) (3) (12)

145 187 58 53

(37) (48) (15) (14)

202 26 17 10 54 46 36

(89) (11) (8) (4) (24) -c 9 (16)

312 78 15 11 56 45 81

(80) (20) (4) (3) (14) 2 9 (22)

Electrocardiogram

in Patients

Treated

(%)

p Value

0.007 0.007 0.025 0.001

0.006 0.05 0.11 0.004 NS 0.072

and Not

with Thrombolvsis* Thrombolysis (%) (n = 228)

Late potentials Filtered QRS (ms) Root mean square-last 40 ms (FV) Low amplitude signals <4OpV (ms) Ejixtion fraction 240% Late potentials Ejection fraction <40% Late potentials

54 108 29 49

No Thrombolysis (%) (n = 390)

Total (%)

(24) & 19 t 14 f 37

98 (25) 109 + 19 3Ok 19 53 t 67

152 (25) 109 2 19 302 17 57 + 57

40 (22)

62 (22)

102 (22)

13 (36)

30 (38)

43 (37)

*No difference is significant.

-

LPs (22) were as follows: (1) total QRS duration 245 ms, (2) RMS 117 pV, and (3) LAS-40 245 ms. Echocdim A 2-dimensionalechocardiographic examination was performed at each center by experiencedcardiologists before dischargeof patients. Ejection fraction was calculated in the apical 4-chamber view by the single-plane ellipse area-lengthmethod. The average of 3 measurementswas considered as the ejection fraction for each patient. A technically good echocardiographic examination was obtainedin 587 patients (95%). Followup: Patientswere followed up every 3 months at the clinics, by telephone and through their private physician. In the caseof death,,detailed information was obtained from the patient’s relatives, the attending physician or, in the case of death in a hospital not participating in the study, clinical records to determine the cause. Furthermore, clinical records were examined in the case of admission to the hospital for any event. Definitions: Sudden death was defined as iristantaneous, unexpected death, or death occurring within 1 hour of the onset of symptoms or during sleep. Sustained ventricular tachycardia was detied as a ventricular rhythm at a rate 2120 beats/min, lasting 230 seconds or leading to hemodynamic compromise. Arrhythmic events were considered the occurrence of either sudden death or sustainedventricular tachycardia. Statistics: Analysis of variance was used for comparing continuous variables, whereas discrete data were

i

compared by &i-square test. Bonferroni’s correction was used for multiple comparisons.Mtiltivariate analysis was performed using stepwise logistic regression to evaluate independent predictors of LPs; variables were inserted in the model when the p value was ~0.1 by univariate analysis. For purposesof analysis, patients were dichotomized with regard to ejection fraction (2 and <40%), peak creatine kinase median value (> and I 1,168 mu/ml) and age (> and 160 years). To reduce possible bias when patients were assigned to specific treatment, a further analysis was performed by comparing retrospectively matched patients, receiving and not receiving thrombolysis (n = 205 and 203, respectively) and P-blocking therapy (n = 92 and 341, respectively), with similar age, gender,site of AMI, previous AMI, peak creatine kinase and ejection fraction. When not otherwise indicated, values are reported as mean f. SD. A p value ~0.05 was consideredsignificant. RESULTS Incidence

of late potentials: LPs were recorded in 152 of 618 patients (24.5%) (Table II). Age, gender and Killip class were similar in patients with and without LPs. However, patients with LPs liad higher mean values of peak creatine kinase, and a higher incidence of ventricular fibrillation in the acute phase (~48 hours), previous AMI, and lower mean ejection fraction (42 + THROMBOLYSIS,

p BLOCKERS

AND SIGNAL-AVERAGED

ECG

527

fABlE

V Main Clinical Data of Patients Treated and Not Treated with p-Blocking Agents

No. of patients Age (years) Age 2 60 years Sex (men) Peakcreatine kinase (mU/ml) Site of infarction Anterior Inferior Non-Q-wave Previous infarction Killip

p Blockers (%)

No p Blockers (%I

p Value

110 (18) f 11

508 (82) 62 k 12 292 (57) 400 (79) 1,621 ” 1,297

0.003 0.02 N3

57.5 50 94 1,137

42 59 9 12

1

(45) (85) 2 704 (38) (54) (8)

207 246 55 68

(11)

95 (86) 15 (14) 2 (2)

2 Ventricular fibrillation (acute phase) Sustained ventricular tachycardia (acute phase) Thrombolysis Ejection fraction (%) Ejection fraction <40%

54 (49) 47 k 9 14 (13)

TABLE VI Results of the Sign,al-Averaged Electrocardiogram Treated with p-Blocking Agents

Late potentials Filtdred QRS (ms) Root mean square-last 40 ms (PV) L6w amplitude signals <40$/ (ms) Ejection fraction 240% Late potentials Ejection fraction < 40% Late potentials

16 106 27 52

10191 (111 5/14

(13)

NS

419 (82)

NS

(36)

0.08

0.11 0.004 0.15 0.062

in Patients Treated and Not

No f3 Blockers (%)

(14) f 17 f 9 2 39

NS

(11)

89 (17) 30 (6! 20 (4) 174 (34) 45 + 9 103 (21)

1 (1)

8 Blockers (%I

(41) (48f

0

Total (%)

p Value 0.007

136 (27)

152 (25)

109 z?I19

108 ‘- 19

0.14

30 5 18 52 + 61

302 17 52 f 58

0.045 NS

92/379

(24)

102/470

(22)

0.006

38/103

(37)

431117

(37)

NS

TABLE VII Clinical Everits in Pat/en& Treatkd and Not Treated with Thrombolysis Thrombolysis (%) Fojlbw-up (mos) Total death Cardiac death Sudden death Sustained ventricular tachycardia Arrhythmic events

12 10 9 4 4 8

+ 6 (4.4) (3.9)

(1.8) (1.8) (3.5)

8% vs 47 + 9%; p
ThE AMERICANJOURNALOF CARDIOLOGY VOLUME72

No Thrombolysis (%I 12” 7 37 (9.5) 30 (7.7) 9 (2.3) 5 (1.3) 14 (3.6)

Total 12 47 39 13 9; 22

f 7 (7.6) (6.31 (2.1) (1.4) (3.5)

p Value NS 0.02 0.06 NS NS NS

Effeds d thron&olysi& on iate potentials (Table IV): LPs were found ih 54 patients (24%) treated with

thrombolysis and ih 98 (25%) treated conventionally, with a nonsigniticant reduction of 6% (95% contidence limits -33%, +22%). Similarly, no difference was found in mean values of the duration of the filtered QRS and LAS-40, and the amplitude of RMS-40. Thrombolytic therapy had no effect on the incidence of LPs also in the subgroups of patients with ejection fractions 2 and ~40% (Table IV), in those with anterior and inferior AMI, and in those aged > and 560 years. C-~Ofpati~Wltil~-beta blockers (Table V): Atenolol or metoprolol was administered to 110of 618 patients (18%). Patientstreatedwith p-blocking agents were younger, had lower mean values of peak creatine kinase and more frequently received thrombolytic agents.The mean ejection fraction was not signiiicantly different between the 2 groups, but the incidence of an ejection fraction ~40% tended to be lower

SEPTEMBER1, 1993

TABLE

VIII

Clinical Events in Patients Treated and Not Treated with p-Blocking Agents

Follow-up (mos) Total death Cardiac death Sudden death Sustained ventricular tachycardia Arrhythmic events

p Blockers (%I

No p Blockers (%)

Total

p Value

12 f 5 2 (1.8)

12 * 7 45 (8.9)

12 F 7 47 (7.6)

NS 0.012

2 (1.8)

37 (7.3)

39 (6.3)

0.033

2 (1.8)

11 (2.2)

13 (2.1)

0 (0) 2 (1.8)

9 (1.8) 20 (3.9)

9 (1.4) 22 (3.5)

NS 0.16 0.28

in the group treated with P-blocking agents(13 vs 21%; Furthermore, P-blocking agents reduced both total and cardiac death (Table VIII). Moreover, arrhythmic p = 0.06). Effects of beta blackers on late potentials (Table events were decreased>50% (1.8 vs 3.9%), although VI): LPs were recorded in 16 patients (14.5%) treated this result did not reach statistical significance.However, and in 136 (27%) not treated with P-blocking agents, a sudden death did not appear to be influenced by psign&ant reduction of 46% (95% contidence limits blocking agents,whereasno ‘patient on P-blocking ther-13%, -79%; p = 0.007). Beta-blocking agents did not apy had sustainedventricular tachycardia during followsignificantly modify the duration of the filtered QRS and up. The 2 events (both sudden death) recorded in LAS-40, but reduced the mean value of RMS-40. patients treatedwith P-blocking agentsoccurred in those Analysis of subgroups, however, showed that p- without LPs. Thus, no patient treated with P-blocking blocking agents reduced LPs only in the group of pa- agents and with LPs had events during follow-up. tients with an ejection fraction 240% (Table VI); in this subgroup, LPs were present in 11%of those treated and DISCUSSION in 24% not treated (reduction of 55%; p = 0.006); conThrombolysis and signal-averaged electrocardi~ versely,in those with an ejection fraction <40%, the inci- gram: There are severalpotential mechanismsby which dence of LPs was 36 and 37%, respectively (p = NS). thrombolysis may decreaseLPs and electrical instabiliThe effectsof l3 blockers on the occurrenceof LPs were ty after AMI, including: (1) a complete salvageof myonot different in patients treated with and without throm- cardium; (2) a substantial limitation of necrosis, with bolysis (11 and 18%, respectively; p = NS). reduction of the “border zone” of the infarct from Multivariate analysis: Stepwise logistic regression which LPs occur; and (3) a beneficial impact on scar analysis showedthat the most powerful predictor of LPs healing, infarct expansion, ventricular remodeling and in patients with AMI was an ejection fraction <40% dilation.2°,21 Several studies have reported a significant reduction (chi-square 7.35; p = 0.007), followed by ventricular fibrillation in the tirst 48 hours (&i-square 5.22; p = 0.02). of LPs in patients with AMI treated by thrombolyThe third strongestpredictor of LPs was the absenceof sis.12-18However, these results are contradictory and P-blocking therapy (chi-square 4.63; p = 0.03). Other should be interpreted with caution. Gang et all2 reportindependentpredictors of LPs were peak creatine kinase ed incidences of LPs of 5% in patients treated and 23% in those not treated with thrombolysis (p = 0.01) on the (p ==O~~Xlll history of AMI (p = 0.03). mtroqx&ive matching: Becausepatients SAECG recorded within 48 hours of admission; howhad not been randomized to either thrombolysis or p ever, the SAEXG obtained at discharge did not show blockers, we selected 2 groups for each type of treat- signiIicant differences in LPs between the 2 groups (5 ment (thrombolytic and p blockade) with similar mean and 18%, respectively). Chew et all4 found incidences age, type of AMI, peak creatine kinase, previous AMI of LPs of 16% in patients treated and 43% in those not and ejection fraction to reduce bias. Again, the incidence treated with streptokinase(p = 0.003), using a 40 Hz filof LPs was similar in the 2 retrospectively matched ter, but the incidence was not significantly different using groups of patients with (52 of 205; 25%) and without a 25 Hz filter. In another study, Zimmermann et al,15 (55 of 203; 27%) thrombolysis, whereas it differed sig- using high-resolution beat-to-beatanalysis,reported incinificantly between those receiving and not receiving p dencesof LPs of 10 and 24% in patients receiving and blockers (13 and 23%, respectively; p = 0.03). not receiving thrombolysis, respectively; however, there Followup: During a mean follow up of 12 & 7 was no difference when LPs were evaluated by the months, there were 47 deaths (total mortality 8%), with SAEXG. Moreover, other studies did not observe any 39 cardiac deaths (6%), of which 13 (2%) were sudden. significant difference in the incidence of LPs between In addition, 9 patients (1.4%) had 21 episode of docu- patients treated and not treated with thrombolytic making the effects of thrombolysis on LPs mented sustainedventricular tachycardia.Thus, in all, 22 agents,6J7*18 controversial. arrhythmic events (3.5%) were recorded. The results of the present large multicenter study are Thrombolytic therapy reduced total and cardiac mortality, whereas it did not influence the occurrence of in agreementwith those latter studies, showing no sigarrhythmic events (3.5 and 3.6% in patients with and nificant effect of thrombolytic therapy per se on LPs. There are several possible explanations for the conwithout thrombolysis, respectively) (Table VII). The effects of thrombolysis were similar in patients both with tradictory results reported on the effectsof thrombolysis and without LPs. on the SAECG, including: (1) different techniques,times THROMBOLYSIS, p BLOCKERS AND SIGNAL-AVERAGED ECG 529

of recording, and deli&ions of abnormality; (2) different times of administration and types of thrombolytic agents; (3) different degreesof efficacy of thrombolysis (however, thrombolytic therapy appearedto be effective in the present study, becausewe observed,as expected, a lower mortality in treated patients); and (4) different clinical characteristicsof patients treated with and without thrombolysis (however, this possibility is unlikely in our study, becausethrombolysis did not reduce LPs in any comparablesubgroup, such as anterior and inferior AMI, low and high ejection fraction, and so forth). Furthermore, it should be considered that thrombolysis could have no apparent effect on the incidence of LPs, becausepatients with unsuccessfulreperfusion are included, whereas a significant effect would be evident only in those with a documented patent infarct artery12~14J5~17,22-25; this possibility cannot be ruled out in this study, becausecoronary angiography was not part of the protocol. However, effective thrombolysis also has the potential of increasing the incidence of LPs after AMI; heterogeneousinfarcts have been observed after coronary occlusion and reperfusion in animality and humans,27with secondary hemorrhage and patchy fibrosis, which could facilitate the occurrence of LPs. Beta Mockers and the signakveraged electdiagram: The effects of @blockingagentson the SAECG

were not investigated previously, although Farrell et al6 recently reported a lower, although not significant, incidenceof LPs in patients with AMI who received in-hospital p-blocking therapy. The present data show that l3 blockers signilicantly decreasethe incidence of LPs in AMI. Although patients treated with l3 blockers showed some differences indicating a lower risk than that of those not treated, thus suggesting a selection bias, l3 blockers maintained their independent effect on LPs when included in the multivariate analysis. This effect was confirmed in the retrospective analysis of 2 wellmatched groups of patients treated with and without l3 blockers. In the setting of acute ischemia, p blockers have been shown to induce a favorable redistribution of blood flow to ischemic subendocardialregions, to reduce infarct size and improve regional myocardial function,28-30 all factors that have potential beneficial effects on the electrophysiologic substrateand thus on the occurrence of LPs. Recently, Grines et a130showed a signilicant improvement in infarct zone wall motion in patients treated with intravenous p blockers after reperfusion therapy.In agreementwith that linding, l3 blockers were more effective in reducing the incidence of LPs when associated with thrombolytic therapy in the present study. Thus, we can speculatethat the lower incidence of LPs that we observed in patients treated with p blockers may reflect a reduced extent of the infarcted area or peri-infarct& ischemia, or a decreased infarct area dyssynergy, or a combination. The observation that l3 blockers decreasedthe incidence of LPs only in patients with an ejection fraction >40% further confirms that their effect was related to a successfullimitation of infarct extension, with a possible reduced risk of border zones constituted by vital and fibrotic areas. 530

THE AMERICANJOURNALOF CARDIOLOGY VOLUME72

Follow-up: Some studieshave suggestedthat the enhanced survival associated with thrombolytic therapy may also depend on a reduced incidence of LPs and, consequently,arrhythmic events.6J2In the present study, thrombolysis reduced mortality but had no effect on the incidence of arrhythmic events.Becausethe occurrence of LPs was not affectedby thrombolysis, we suggestthat the improvement of prognosis after thrombolytic therapy is not due to a reduction of electrical instability as indicated by the SAECG. Several studies have demonstratedthat l3 blockers prevent sudden death and recurrent AMI,loJ1 but the mechanism of this beneficial effect is not clear. Beta blockers significantly decreasedcardiac death and the incidence of arrhythmic events by >50%; it is tempting to speculatethat the reduction of mortality by l3 blockers is, at least in part, secondaryto an increasedelectrical stability, as indicated by a lower incidence of LPs in the present study, and that this effect is potentiated by thrombolytic treatment also. Study limitations: These lindings are the results of a retrospective analysis of a prospective study. Because patients were not randomized to separategroups with and without thrombolytic and P-blocking therapy, the results should be interpreted with caution. Possible bias in the selection of patients for specilic treatment cannot be excluded, and it is probable that patients with more depressedleft ventricular function may have been excluded from p-blocking treatment. However, the results of multivariate analysis and retrospectivematching both indicate a beneficial effect of @blocking therapy on the SAECG and may thus have important implications for future researchin this field. Acknowledgment: We are grateful to Attilio Maser-i, MD, for helpful suggestions in reviewing the manuscript.

APPENDIX Participants of the late Potentials Italian Study (LAPIS): Istituto di Cardiologia, Universita’ Cattolica,

Roma, Italy (Pietro Santarelli, MD, Gaetano A. Lanza, MD, Francesco Biscione, MD, and Andrea Natale, MD); Divisione di Cardiologia, OspedaleCivile, Caserta, Italy (Giancarlo Corsini, MD, and Carmine Riccio, MD); Divisione di Cardiologia, OspedaleMaggiore della Car&a’, Novara, Italy (Eraldo Occhetta, MD, and Paolo Rossi, MD); Divisione di Cardiologia, Ospedale Poveri Infermi, Borgosesia, Italy (Maurizio Gronda, MD, and Johannes Makmur, MD); Divisione di Cardiologia, Ospedale SS. Trinita’, Borgomanero, Italy (Marco Zanetta, MD, and Umberto Parravicini, MD); and Divisione di Cardiologia, OspedaleS. Filippo Neri, Roma, Italy (SalvatoreToscano,MD). 1. Simon MB. Identification of patients with ventdcula~ tachycmlia after myocadia.l infarction from signals in the terminal QRS complex. Circulation 1981; 64~235-242. 2. Breithwdt G, Borggrefe M. Recent advances in the identification of patients at risk of ventricular tachymhythmias: role of ventricular late potentials. Circulation 1987;75:1091-1097. 3. Denes P, Santaelli P, Hauser RG, Uretz EF. Quantitative analysis of the high frequency components of the terminal portion of the body surface QRS in nomml

SEPTEMBER1, 1993

subjects and in patients with ventricular tachycardia. Circulation 1983;67: 1129-l 138. 4. Kuchar DL, Thorbum CW, Sammel NL. Prediction of serious arhythmic events after myocardial infarction: signal averaged electrocardiogram, Halter monitoring and radionuclide venticulogmphy. J Am CoN Cardiol 1987;9:531-538. 5. Games JA, Winters SL, Martinson M, Machac J, Stewart D, Targonski A. The prognostic significance of quantitative signal-averaged variables relative to clinical variables, site of myocardial infarction, ejection fraction and ventricular premature beats: a prospective study. J Am Coll Cardiol 1989;13:377-384. 6. Famell TG, Bashir Y, Cripps T, Malik M, Poloniecki .I, Bennett ED, Ward DE, Camm AJ. Risk stratification for srrhythmic events in postiofarction patients based on heart rate variability, ambulatory electrocardiographic variables and the signal averaged electmcardiogrsm. J Am CoN Cardiol 1991;18:687-697. 7. Steinberg JS, Regan A, Sciacca RR, Bigger JT, Fleiss JL. Predicting arrhythmic events after acute myocardial infarction using the signal averaged electmcardiogram. Am .I Cardiol 1992;69:13-7.1. 8. Gmppo Italian0 per lo Studio delIa Streptochioasi nell’ Infarto miocardico (GISSI). Long termeffects of intravenous thrombolysis in acute myocaxlial infarction: fmal report of the GISSI study. Lancet 1987;ii:871-874. 9. ISIS-2 (Second International Study of Infarct Sun&l) Collaborative Group. Random&d trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-Z. Lancer 1988;2: 349-360. 10. ISIS-I (Fist International Study of Infarct Survival) Collaborative Group. Randomized trial of intravenous atenolol among 16027 cases of suspected acute myocardial infarction: ISIS-I. Lancet 1981;2:823-827. 11. MIAMI Trial Research Group. Metoprolol In Acute My-dial Infarction (MIAMI): a randomized placebo-controlled international trial. Eur Heart J 1985;6: 199-226. 12. Gang ES, Lew AS, Hong MM, Wang FZ, Siebert CA, Peter T. Decreased incidence of ventricular late potentials after successful thrombolytic therapy for acute myocardial infarction. N Engl .I Med 1989;321:712-716, 13. El& M, Lear J, Hod H, Rot&o Z, Truman S, Kaplinsky E, Abboud S. Effect of thrombolysis on the evolution of late potentials within 10 days of infarction. Br Heart J 1990;63:273-276. 14. Chew E-W, Morton P, Murtagh JG, Scott ME, O’Keef DB. Intravenous streptokinase for acute myocardial infarction reduces the occurrence of ventricular late potentials. Br Heart J 199&64:5-S. 15. Zimmennaon MM, Adamec R, Ciaroni S. Reduction in the frequency of ventricular late potentials after acute myocardial infarction by early thrombolytic therapy. Am J Cardiol 1991;67:697-703. 16. Lear J, Hold H, Rotstein Z, Truman S, Gansky S, Goldbowt U, Abboud S, Kaplinsky E, Eldar M. Effects of thrombolysis on the 1Zlexl signal-averaged ECG in the early postinfarction period. Am Heart J 1990; 120:495-502. 17. Twitto G, Risa AL, Zanchi E, Prati PL. The signal averaged electrx&iograrn and ventricular arrhythmias after thrombolysis for acute myocardial infarction. JAm Call Cardiol 1990; 15: 127@1276.

18. de Chillou C, Sadoul N, Briancon S, Etieone A. Factors detetminiig the occwrence of late potentials on the signal averaged electrocardiogram after a first myocardial infarction: a multivariate analysis. J Am Coil Cardiol 1991;18:1638-1642. 19. Buckingham TA, Thessen CC, Stevens LL, Redd RM, Kennedy HL. Effect of conduction defects on the signal averaged electrocardiographic determination of late potentials. Am J Cardiol 1988;61: 1265-1271. 20. Morgan CD, Roberts RS, Haq A, Baigrie RS, Daly PA, Gent M, Armstrong PW, for the TPAT Study Group. Coronary patency, infarct size, and left venhicular function after thrombolytic therapy for acute myocardial: results from the Tissue Plasminogen Activator Toronto (TPAT) placebo-controlled trial. JAm Co11Cardial 1991;17:1451-1457. 21. Lavie CJ, O’Keefe JH Jr, Chesebro JH, Clements lP, Gibbons RJ. Prevention of late ventricular dilatation after acute myocxdial infarction by successfol thrombolytic reperfosion. Am J Cardiol 1990;66:31-36. 22. Vatterott PJ, Hammill SC, Bailey KR, Wiltgen CM, Gersh BJ. Late potentials on signal averaged electxxardiogram and patency of the infarct-related ztery in survivors of acute myocardial infarction. J Am Co11 Cardioi 1991; 17:33&337. 23. Lange RA, Cigarroa RG, Wells PJ, Kremers MS, Hillis LD. Influence of anterograde flow in the infarct axtery on the incidence of late potentials after acute myocardial infarction. Am J Cardiol 1990;65:554-558. 24. Tranchesi B, Verstraete M, Van de Wezf F, de Abuquerque CP, Caramelli B, Gebara OC, Per&a WI, Moffa P, Bellotti G, Pileggi F. Usefulness of high-frequency analysis of signal-averaged surface elec~ocardiograms in acute myocardial infarction before and after coronary thrombolysis for assessing coronary reperfw sion. Am J Cardiol 1990;66: 11961198. 25. Aguirre FV, Kern MJ, Hsia J, Serota H, Janosik D, Greenwalt T, Ross AM, Chaitman BR. Importance of myocardial infarction artery patency on the prevalence of ventricular arrhythmia and late potentials after thrombolysis in acute myocardial infarction. Am J Cardiol 1991;68:1411%1416. 26. Smith GT, Soeter JR, Haston HH, McNamara JJ. Coronary reperfusion in primates. Serial electrocardiographic and histologic assessment. J Clin Invest 1974;54: 142k-1427. 27. Fujiwara H, Onodera T, Tanaka M, Fujiwara T, Wu DJ, Kawai C, Hamashima Y. A clinicopathologic study of patients with hemorrhagic myocardial infarction treated with selective coronary tbrombolysis with urokinase. Circulation 1986;73: 749-757. 28. Vatner SF, Baig H, Manders TW, Ochs H, Pagani M. Effects of pmpraoolol on regional myocardial function, electrocardiogram and blood flow in conscious dogs with myocardial infarction. J Clin Invest 1977;60:353-360. 29. Steingart RM, Matthews R, Gambino A, Kantrowitz N, Katz S. Effects of intravenous metoprolol on global and regional left ventricular function after coronary arterial reperfusion in acute myocaxiial infarction. Am J Cardiol 1989;63:767-771. 30. Grines CL, Booth DC, Nissen SE, Gorley JC, Bennet KA, DeMaria EN. Acute effects of pare&ml beta-blockade on regional ventricular function of infarct and noninfarct zones after repetfosion therapy in humans. J Am CON Cardiol 1991;17: 13X2-1387.

THROMBOLYSIS,p BLOCKERSAND SIGNAL-AVERAGEDECG 531