Author + information
- Received July 23, 1998
- Revision received January 25, 1999
- Accepted February 15, 1999
- Published online June 1, 1999.
- Stefan H Hohnloser, MD, FACCa,* (, )
- Thomas Klingenheben, MDa,
- Markus Zabel, MDa,
- Matthias Schöpperl, MDa and
- Oliver Mauß, MSca
- ↵*Reprint requests and correspondence: Dr. Stefan H. Hohnloser, Department of Medicine, Division of Cardiology, J.W. Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany
The purpose of this study was to determine the prevalence, characteristics and the predictive value of nonsustained ventricular tachycardia (VT) for subsequent death and arrhythmic events after acute myocardial infarction (AMI).
Nonsustained VT has been linked to an increased risk for sudden death in coronary patients. It is unknown whether this parameter can be used for selection of high-risk patients to receive an implantable defibrillator for primary prevention of sudden death in patients shortly after AMI.
In 325 consecutive infarct survivors, 24-h Holter monitoring was performed 10 ± 6 days after AMI. All patients underwent coronary angiography, determination of left ventricular function and assessment of heart rate variability (HRV). Mean follow-up was 30 ± 22 months.
There was a low prevalence (9%) of nonsustained VT shortly after AMI. Nonsustained VT together with depressed left ventricular ejection fraction (LVEF) was found in only 2.4% of patients. During follow-up, 25 patients reached one of the prospectively defined end points (primary composite end point of cardiac death, sustained VT or resuscitated ventricular fibrillation; secondary end point: arrhythmic events). Kaplan Meier event probability analyses revealed that only HRV, LVEF and status of the infarct-related artery were univariate predictors of death or arrhythmic events. The presence of nonsustained VT carried a relative risk of 2.6 for the primary study end point but was not a significant predictor if only arrhythmic events were considered. On multivariate analysis, only HRV, LVEF and the status of the infarct artery were found to be independently related to the primary study end point.
There is a low prevalence of nonsustained VT shortly after AMI. Only 2% to 3% of all infarct survivors treated according to contemporary guidelines demonstrate both depressed LVEF and nonsustained VT. The predictive value of nonsustained VT for subsequent mortality and arrhythmic events is inferior to that of impaired autonomic tone, LVEF or infarct-related artery patency. Accordingly, the use of nonsustained VT to select patients for primary implantable cardioverter/defibrillator prevention trials shortly after AMI appears to be limited.
In studies conducted before the widespread use of acute revascularization therapy in acute myocardial infarction (AMI), the presence of nonsustained ventricular tachycardia (VT) was found to be a marker of increased risk for subsequent all-cause and arrhythmic mortality (1–3). In the thrombolytic era, the prognostic significance of nonsustained VT is more controversial (4,5). For instance, the GISSI-2 investigators found nonsustained VT in only 6.8% of infarct survivors undergoing thrombolytic therapy, and its presence was not an independent predictor of subsequent all-cause or arrhythmic mortality (4). Preliminary data from more recent studies indicate an even lower incidence of nonsustained VT after AMI, even in the presence of depressed left ventricular function (6). On the other hand, the MADIT trial, so far the only primary prevention study demonstrating superiority of cardioverter/defibrillator implantation compared with conventional antiarrhythmic treatment in coronary patients, used Holter-documented asymptomatic nonsustained VT together with reduced left ventricular function as its primary selection criteria (7). However, because >75% of patients enrolled had had their infarct more than six months before study entry (7), it remains to be demonstrated whether the presence of nonsustained VT serves as a prognostic marker shortly after AMI.
Accordingly, the present study examined the incidence, the characteristics and the predictive value of nonsustained VT for cardiac death and arrhythmic events in a large sample of consecutive infarct survivors treated at a single institution according to contemporary therapeutic guidelines. The value of nonsustained VT as a predictor of subsequent risk was compared with that of other established risk stratifiers such as left ventricular function, patency of the infarct-related artery or markers of autonomic tone. Patients were prospectively followed for 30 ± 22 months.
Consecutive patients presenting with AMI were considered for entry in the present study. Patients were eligible for participation if the following inclusion criteria were met: 1) a confirmed diagnosis of AMI based on clinical presentation (typical chest pain for >30 min, unresponsive to nitrates; ST segment elevation in at least two limb leads of >0.1 mV or in at least two precordial leads of ≥0.2 mV; elevated CK and CKMB levels), 2) Holter monitoring was performed before hospital discharge. All patients gave informed consent before entry in the study.
After discharge, patients were seen in the arrhythmia outpatient clinic at 4, 8 and 12 months after AMI, and in 6-month intervals thereafter. All episodes of nonfatal arrhythmic events, reinfarction and revascularization procedures were carefully recorded. Information about deceased patients was obtained from family members, their general practitioners and from the hospitals to which they had been admitted. Particular attention was given to the circumstances of each death.
The primary end point of the study was prospectively defined as a composite end point of cardiac mortality, documented sustained VT and resuscitated ventricular fibrillation (VF). A secondary end point of the study was arrhythmic events (defined as sudden cardiac death, documented sustained VT and resuscitated VF). Sudden death was defined as instantaneous, unexpected death or death within 1 h of symptoms onset not related to circulatory failure. Sustained VT was defined as a documented tachycardia of ventricular origin at a rate of ≥100 bpm and lasting for >30 s or resulting in hemodynamic collapse.
Selective coronary angiography was performed by the Judkins technique in all patients 7 to 30 days after the infarct. Multiple orthogonal views were obtained and recorded at standardized angles on cineangiographic film. After catheterization, the cineangiograms were reviewed independently and without knowledge of other clinical parameters. Antegrade perfusion of the infarct artery was graded according to the classification system of the Thrombolysis in Myocardial Infarction (TIMI) trial (8). Patients with an initially occluded artery who were revascularized before hospital discharge were considered to have a patent vessel during the statistical analysis.
Before hospital discharge (10 ± 6 days after AMI) and in stable clinical conditions, patients underwent 24-h ambulatory monitoring (modified V2 and V5 leads) using two-channel bipolar Holter recorders (model 8500; Marquette, Milwaukee, Wisconsin). The tapes were subsequently analyzed by running a laser scanner (model 8000, Marquette) with its arrhythmia analysis program to identify and label each QRS complex.
The presence of nonsustained VT was defined according to criteria previously defined in a prospective trial evaluating the value of prophylactic implantation of a cardioverter/defibrillator in survivors of myocardial infarction (7). Accordingly, at least three consecutive premature ventricular beats at a rate of ≥120 beats/min were considered nonsustained VT. All respective episodes were printed out from the Holter recordings and verified by a single experienced investigator who was blinded as to the clinical information of the patient. Analysis of heart rate variability (HRV) was performed as previously reported (9,10). The reproducibility of HRV determination in our laboratory has been published previously (9). A prospectively defined cut-point of the HRV index standard deviation of normal-to-normal intervals (SDNN) ≤70 ms was used to define a high-risk group (11).
Determination of left ventricular function
Left ventricular ejection fraction (LVEF) was calculated from the 30° right anterior oblique ventriculogram using a standard area-length method or by echocardiography or by radionuclide ventriculography. All LVEF determinations were done in duplicate by two separate investigators blinded to the clinical data of the patient.
Continuous values are reported as mean ± standard deviation. All data were analyzed using the Statistical Package for the Social Sciences (12). Comparisons between patients with and without events during follow-up were performed by means of the unpaired Student ttest for normally distributed continuous variables (two-sided) or the chi-square test for categorical data. The independent correlation of various risk stratifiers to events during follow-up was determined by means of logistic regression analysis with the occurrence of events as the dependent variable. Sensitivity was defined as the percent of patients with a positive test result from all patients with an end point, specificity as the percent of patients with a negative test result from all patients without an end point and positive predictive value as the percent of patients with an end point from all patients with a positive test result. Kaplan Meier event probability curves (13)were computed, stratifying patient groups according to the presence or absence of prespecified risk markers. The cumulative probability of events of two patient groups was compared by means of a log-rank test. Significance was considered with a p value ≤0.05.
From a total of 361 consecutive patients surviving AMI, Holter recordings at the time of hospital discharge were obtained in 325 (90%). In 36 patients, long-term electrocardiographic recordings were not available because patients refused to undergo Holter monitoring (n = 11), patients were discharged before Holter monitoring (n = 14), or due to technical failures (n = 11). Thus, the present analysis is based on data of 325 patients with complete clinical and electrocardiographic data (Table 1). There were 257 men and 68 women with a mean age of 58 ± 11 years. LVEF averaged 49 ± 11%; 51 patients (16%) had a LVEF of ≤35%. Coronary angiography before discharge demonstrated an open infarct-related artery in 255 of 325 patients (78%). In the remaining 70 patients (22%), the infarct vessel was occluded. Patients were followed for a mean of 30 ± 22 months after the index infarct (median 25 months). The percentage of patients treated with beta-adrenergic blocking agents was 80% at hospital discharge, 67% at one year and 59% at the two-year follow-up; the respective numbers for aspirin treatment were 92%, 85% and 80%; and for ACE inhibitors, 41%, 40%, and 38%. During follow-up, antiarrhythmic drug therapy was instituted in 20 patients for atrial fibrillation or flutter (n = 15), or symptomatic ventricular premature beats (n = 5). There was only one cardiac death in a patient on antiarrhythmic drug therapy with sotalol.
Incidence of nonsustained VT after myocardial infarction
In 29 of 325 patients (9%), Holter monitoring revealed the presence of at least one episode of nonsustained VT. Eight of these patients (28%) had a LVEF of ≤35%. There were no significant differences in clinical characteristics between patients with or without nonsustained VT.
Characteristics of nonsustained VT after myocardial infarction
There was an average of 2.2 ± 2.4 episodes of nonsustained VT per patient (range 1–11). Episodes of nonsustained VT consisted of a mean of 6.0 ± 6.1 consecutive beats (range 3–34). There was only one patient with a VT episode of >20 beats; disregarding this episode, the average VT episode consisted of only 5.0 ± 2.9 beats. Figure 1shows the distribution of maximal VT length in all 29 patients. The average cycle length of nonsustained VT episodes was 415 ± 65 ms, corresponding to a rate of 145 beats/min. The average number of VT episodes, their average number of beats and cycle length were similar in patients with a LVEF ≤35% compared with VT-positive patients with preserved left ventricular function (Table 2). The prevalence of nonsustained VT was similar in patients with an occluded (6 of 70) compared with those with a patent (23 of 255; p = NS) infarct-related artery.
Events during follow-up
During follow-up, 15 patients died from cardiac death, which was classified as sudden cardiac death in 7 patients. Eight additional patients developed sustained monomorphic VT, and two patients were successfully resuscitated from documented VF. Thus, a total of 25 prospectively defined study end points were observed.
Prognostic value of clinical parameters
Clinical parameters including gender, age, infarct location, peak creatine kinase (CK) levels, thrombolysis or acute percutaneous transluminal coronary angioplasty (PTCA) and discharge medication were included in the statistical analysis. On univariate analysis, event-free survivors were significantly younger (57 ± 11 years) compared with patients dying or suffering arrhythmic events (64 ± 6 years; p = 0.007). Thrombolytic therapy or acute rescue PTCA was administered less frequently to patients with events during follow-up (10 of 25 patients, 40%) as compared with patients without events (184 of 300, 61%; p = 0.04).
Prognostic value of investigational procedures
Table 3summarizes the prognostic information derived from the various investigational tests performed. Five variables were found to yield prognostic information; depressed left ventricular function and impaired HRV were associated with the highest positive predictive accuracy and carried the highest relative risk for reaching the primary study end point. The corresponding Kaplan Meier curves for survival free of the primary study end point (composite of cardiac mortality, documented sustained VT, resuscitated VF) are shown in Figures 2 through 5. ⇓⇓⇓⇓
The same analysis was repeated for the prospectively defined secondary study end point, i.e., arrhythmic events. As demonstrated in Table 3, only depressed HRV, a permanently occluded infarct vessel and reduced LVEF remained significant predictors of arrhythmic events during follow-up. The presence of nonsustained VT yielded a relative risk of only 1.4 for subsequent arrhythmic events, which did not reach statistical significance (Fig. 5).
Multivariate analysis of risk stratifiers
A stepwise regression analysis incorporating clinical and investigational variables was performed to assess independent risk parameters. For the primary study end point, only SDNN (χ29.5; p = 0.002), LVEF (χ25.0; p = 0.026) and the status of the infarct-related artery (χ24.8; p = 0.029) were demonstrated to be independently correlated to outcome. With respect to the secondary study end point, only SDNN (χ26.9; p = 0.009) and the status of the infarct-related artery (χ25.2; p = 0.02) were independent predictors.
Main study findings
This large prospective long-term follow-up study reveals several important findings. There is a low (9%) incidence of Holter-documented nonsustained VT in postinfarction patients treated according to contemporary therapeutic guidelines, including a high frequency of revascularization. In only 2.4% of all infarct survivors, nonsustained VT is found in the presence of reduced left ventricular function. The average nonsustained VT episode comprises only five to six beats at a rate of around 145 beats/min. Most important, however, the predictive value of nonsustained VT for subsequent mortality is low. Actually, when only arrhythmic events are considered, the presence of nonsustained VT yields no prognostic value at all. Evidence of impaired cardiac autonomic tone, a patent infarct artery and reduced left ventricular function are more powerful risk predictors.
Prevalence and characteristics of nonsustained VT after myocardial infarction
Several studies conducted before the widespread use of acute revascularization therapy for AMI have attempted to determine the prevalence and the prognostic value of nonsustained VT (2,3,14,15). In these trials, the prevalence for this arrhythmia ranged between 12% (2)and as high as 21% (15). In most of these studies, the presence of spontaneous nonsustained VT was found to be a predictor of all-cause mortality in the first two years after the index infarction. There is a relative paucity of data regarding the prevalence of nonsustained VT in infarct survivors who are treated according to contemporary therapeutic guidelines. One of the largest studies stems from the Gruppoltaliano per lo Studio della Sopravivenza nell’ Infarto Miocardito (GISSI) investigators, who reported a low incidence of nonsustained VT (4). They performed Holter monitoring at the time of hospital discharge in 8,676 infarct survivors and found at least one run of nonsustained VT in only 6.8% of their patients (4), a figure that is very similar to the one reported in this study. The GISSI investigators concluded from their observations that the prevalence of nonsustained VT was much lower than the results of studies conducted in the prethrombolytic era, whereas the incidence of ventricular premature beats was comparable with earlier observations (4).
Only a few studies have described the characteristics of nonsustained VT detected early after AMI. Bigger et al. examined 820 infarct survivors and found that in the 12% of patients with nonsustained VT on Holter monitoring, 26 patients had a three-beat run as their longest episode (2). Nonsustained VT consisted of 4 to 10 beats in another 61% of individuals, whereas only 11% of patients demonstrated >10 beats (2). The average VT rate was 121 ± 34 beats/min. These findings are comparable with those reported here, where the average VT rate was around 145 beats/min. Most VT episodes consisted of a 3-beat run, with only 2 patients having episodes of >10 beats. These findings are similar to those reported by Denes et al., who analyzed the Holter recordings of patients enrolled in the Cardiac Arrhythmia Suppression Trial (CAST) (16). In that study, VT length averaged 4.8 ± 2.4 beats and most patients had VT runs at rates of 120 to 149 beats/min.
Nonsustained VT in patients with depressed left ventricular function
In the Multicenter Automatic Defibrillator Inplantation Trial (MADIT) study, the combination of nonsustained Holter-documented VT and a LVEF ≤35% was shown to be effective in screening coronary patients at high risk for arrhythmogenic death who subsequently benefited from prophylactic implantable cardioverter/defibrillator (ICD) implantation (7). Accordingly, another goal of the present study was to determine the prevalence of this combined selection criterion in patients shortly after AMI. As expected, the average left ventricular function was well preserved in our patient population due to strict adherence to modern concepts of infarct therapy. Only 51 of 325 patients (16%) had a significantly impaired left ventricular function, defined as a LVEF ≤35%. Nonsustained VT was present in only 8 of 51 patients, resulting in a prevalence of the combined risk stratifiers of only 2.4%. This number is in excellent concordance with a preliminary report comprising 1,625 infarct survivors (6). Depressed LVEF and nonsustained VT were found in only 3.2% of these patients (6). These numbers indicate that the selection criteria defined by the MADIT investigators (7)are prevalent in only a few patients in the immediate postinfarction period.
Prognostic implication of nonsustained VT
Nonsustained VT was considered to represent a clinically useful risk stratifier in patients surviving AMI in the prethrombolytic era (2,3,14). The results of the present study demonstrate that this seems to no longer be the case. Actually, nonsustained VT was inferior to measures of contractile function and of impaired autonomic tone. Moreover, when only arrhythmic events during the follow-up period were taken into consideration, nonsustained VT was no longer a statistically significant risk predictor. This is again in accordance with previous reports such as the GISSI-2 findings (4). In this study, nonsustained VT failed to be a statistically significant risk predictor, whereas frequent premature beats were confirmed as independent risk factors for total and sudden death during the first six months after the acute event (4). This finding may also indicate that the prognostic value of nonsustained VT is linked to other variables. Another recently published study in 471 infarct survivors failed to demonstrate prognostic value of both isolated premature ventricular beats and nonsustained VT concerning overall mortality after AMI (17).
Postinfarction risk stratification for all-cause mortality and for arrhythmogenic mortality in particular remains imperfect. The most likely reason for this is the low overall postinfarction mortality rate. The bulk of evidence, however, indicates that probably the most accurate way of approaching this problem is to take into account measures of left ventricular function and cardiac autonomic tone. Several studies utilizing heart rate variability have indicated this (18–22), and the present study confirms those findings. A recent large multicenter trial enrolling 1,284 infarct survivors treated according to modern therapeutic guidelines is of particular interest (23). In that study, two measures of cardiac autonomic tone were utilized, namely heart rate variability and baroreflex sensitivity (9,24,25). The study provided sound evidence that after myocardial infarction the analysis of vagal reflexes has significant prognostic value independently of left ventricular contractile function (23). As a matter of fact, the highest risk of cardiac death in that study was found for patients with both a reduction in LVEF and in autonomic tone, a finding that is in agreement with the results presented here.
A potential limitation of our study relates to the sample size: although data of 325 patients were analyzed, it is still conceivable that an even larger study would have yielded improved predictive accuracy of nonsustained VT. Another important issue relates to the comparison of the present results with previous findings. For instance, in the GISSI trial (4), nonsustained VT failed to be an independent predictor on multivariate analysis, but ventricular premature beats were predictive. It is possible that if multivariate analysis would have been performed omitting the variable ventricular premature beats, nonsustained VT would have become a significant predictor. A third problem relates to the timing of Holter monitoring after AMI. In a recent study, 112 patients who had nonsustained VT within 72 h of AMI were identified and compared with a matched control group (26). In-hospital and follow-up mortality were not significantly different between both groups. On multivariate analysis, however, the time from presentation to occurrence of nonsustained VT was a predictor of subsequent mortality, with the highest risk for patients who had the arrhythmia approximately 24 h after AMI. The present study did not evaluate the prognostic value of nonsustained VT within the first 72 h of AMI, but rather that of VT runs documented 10 ± 6 days thereafter. Studies in the prethrombolytic era, on the other hand, found the highest prevalence of nonsustained VT one month after AMI (27). Thus, there seems to be some uncertainty as to the most appropriate time point at which Holter recordings should be assessed for optimization of the predictive value of nonsustained VT.
Given the high efficacy of the ICD in preventing arrhythmogenic death (28), there appears to be the need of further primary prevention trials, particularly in postinfarction patients. Such trials will only be accepted by the scientific community and by clinicians taking care of infarct survivors if they demonstrate a reduction in all-cause mortality (29). This requirement makes risk stratification particularly demanding in this patient population. From the results of this as well as from other studies, it seems that the use of spontaneous nonsustained VT documented at the time of hospital discharge is of limited value. Only very few patients after AMI demonstrate both nonsustained VT and depressed LVEF. Thus, the criteria applied to the MADIT population (7)are only rarely observed shortly after AMI. It remains to be demonstrated, however, that other risk stratifiers may yield clinical benefit in this respect. For instance, the predictive value of the combination of depressed left ventricular function and impaired heart rate variability is currently being examined in a large interventional study (30). Only such prospective intervention trials will be able to answer these clinically important questions (29).
☆ This study was supported by the German Heart Foundation, Frankfurt, Germany.
- acute myocardial infarction
- cardiac arrythmia suppression trial
- creatine kinase
- heart rate variability
- implantable cardioverter/defibrillator
- left ventricular ejection fraction
- multicenter automatic defibrillator implantation trial
- percutaneous transluminal coronary angioplasty
- standard deviation of normal-to-normal intervals
- thrombolysis in myocardial infarction
- ventricular fibrillation
- ventricular tachycardia
- Received July 23, 1998.
- Revision received January 25, 1999.
- Accepted February 15, 1999.
- American College of Cardiology
- Anderson K.P.,
- DeCamilla J.,
- Moss A.J.
- Multicenter Post-Infarction Research Group,
- Bigger J.T.,
- Fleiss J.L.,
- Rolnitzky L.M.
- BHAT Study Group,
- Kostis J.B.,
- Byington R.,
- Friedman L.M.,
- et al.
- Maggioni A.P.,
- Zuanetti G.,
- Franzosi M.G.,
- et al.
- Hohnloser S.H.,
- Franck P.,
- Klingenheben T.,
- et al.
- ↵Andresen D, Steinbeck G, Brüggemann T, et al. Can the MADIT results be applied to myocardial infarction patients at hospital discharge? J Am Coll Cardiol 1998;31 Suppl A:308 (abstr).
- Hohnloser S.H.,
- Klingenheben T.,
- van de Loo A.,
- et al.
- GISSI-2 Investigators,
- Zuanetti G.,
- Neilson J.M.,
- Latini R.,
- et al.
- ↵SPSS-X User’s Guide. 3rd ed. Chicago: SPSS, 1988.
- Multicenter Post-Infarction Research Group,
- Bigger J.T. Jr.,
- Fleiss J.L.,
- Kleiger R.,
- et al.
- European Infarction Study Group,
- Andresen D.,
- Bethge K.P.,
- Boissel J.P.,
- et al.
- Denes P.,
- Gillis A.M.,
- Pawitan Y.,
- et al.
- Groupe détude du pronostic de lı́nfarctus du myocarde (GREPI),
- Touboul P.,
- Andre-Fouet X.,
- Leizorovicz A.,
- et al.
- Farrell T.G.,
- Bashir Y.,
- Cripps T.,
- et al.
- Bigger J.T.,
- Fleiss J.T.,
- Rolnitzky L.M.,
- Steinman R.C.
- Fei L.,
- Malik M.
- Task Force of The European Society of Cardiology and the North American Society for Pacing and Electrophysiology
- Schwartz P.J.,
- Vanoli E.,
- Stramba-Badiale M.,
- et al.
- Schwartz P.J.,
- La Rovere M.T.,
- Vanoli E.
- Cheema A.N.,
- Sheu K.,
- Parker M.,
- et al.
- Kleiger R.E.,
- Miller J.P.,
- Thanavaro S.,
- et al.
- ↵Camm AJ, Karam R, Pratt CM. The azimilide post-infarction survival evaluation (ALIVE) trial. Am J Cardiol 1998;81 Suppl 6A:35–9D.