Association of Left Ventricular Remodeling and Nonuniform Electrical Recovery Expressed by Nondipolar QRST Integral Map Patterns in Survivors of a First Anterior Myocardial Infarction
Background Progressive left ventricular dilatation after myocardial infarction is associated with a high mortality rate, the majority of which is arrhythmogenic in origin. The underlying mechanism of this relation remains unknown. It has been suggested, however, that left ventricular dilatation is accompanied by changes in repolarization characteristics that may facilitate the occurrence of life-threatening ventricular arrhythmias.
Methods and Results We examined 62-lead body surface QRST integral maps during sinus rhythm in 78 patients at 349±141 days after thrombolysis for a first anterior myocardial infarction. Visual map analysis was directed at discriminating dipolar (uniform repolarization) from nondipolar (nonuniform repolarization) patterns. In addition, the nondipolar content of each map was assessed quantitatively with the use of eigenvector analysis. Nondipolar map patterns were present in almost one third of the patients (32%). Left ventricular end-systolic and end-diastolic volumes were assessed echocardiographically before discharge and after 3 and 12 months with the use of the modified biplane Simpson rule. The increase in left ventricular end-systolic volume 1 year after myocardial infarction was more pronounced in patients with nondipolar QRST integral map patterns (14.47±14.10 versus 4.22±8.44 mL/m2, P=.017). In patients with an increase in end-systolic volume of more than 16 mL/m2 (upper quartile), the prevalence of nondipolar maps was 89% compared with 29% in patients with dilatation of less than 16 mL/m2. In addition, the nondipolar content of maps in patients in the upper quartile was significantly increased compared with the lower quartiles (49±14% versus 37±12%, P=.013). Logistic regression analysis revealed that an end-systolic volume of more than 42 mL/m2 after 1 year contributed independently to the appearance of nondipolar maps. Patients with high-grade ventricular arrhythmias showed a higher nondipolar content (49±17% versus 39±10%, P=.013). QTc dispersion did not discriminate between patients with and those without high-grade ventricular arrhythmias. Also, the association between left ventricular remodeling and nondipolar map patterns was confirmed prospectively in an additional group of 15 patients.
Conclusions Nondipolar map patterns are present in 32% of patients after thrombolysis for a first anterior myocardial infarction and are associated with increased left ventricular dilatation. These data support the hypothesis that left ventricular dilatation after myocardial infarction leads to changes in repolarization characteristics that may facilitate the occurrence of life-threatening ventricular arrhythmias.
Left ventricular remodeling after myocardial infarction is characterized by expansion of the infarcted area and dilatation and hypertrophy of the noninfarcted myocardium.1 This process is accompanied by an overall increase in left ventricular volume. An increase in left ventricular end-systolic volume after myocardial infarction has been identified as an important risk factor for sudden cardiac death.2 At present, the exact mechanism of this association is unclear. It has been reported that increased loading conditions and dilatation of the heart occur in parallel with changes in repolarization characteristics3 4 5 6 and the occurrence of early afterdepolarizations.4 6 These altered electrophysiological conditions may predispose to the occurrence of ventricular arrhythmias.7 8 9
Local repolarization properties during sinus rhythm can be assessed noninvasively with the use of multiple-lead ECG (body surface mapping). The QRST deflection area or isointegral has been identified both in theory10 11 12 and in experiments with animals13 as a marker of changes in local repolarization properties, regardless of the site of origin of ventricular activation.14 A high incidence of nondipolar QRST integral map patterns has been reported in patients with documented sustained ventricular tachycardia or ventricular fibrillation.15 16 17 18 In contrast, healthy subjects usually demonstrate a dipolar QRST integral map pattern.19 20 It has been suggested that nondipolar QRST integral map patterns reflect nonuniformity of local recovery properties and can be used to identify patients who are at risk of developing life-threatening ventricular arrhythmias.16 21 22
In the present study, we hypothesized that left ventricular remodeling after myocardial infarction leads to changes in local repolarization properties that can be detected noninvasively as nondipolar QRST integral map patterns and may express a potentially arrhythmogenic state.
The present study was part of the Captopril and Thrombolysis Study (CATS), in which the effect of captopril treatment, started during thrombolysis, was evaluated in patients with a first anterior myocardial infarction.23 Main end points included left ventricular remodeling, neurohumoral activation, and ventricular arrhythmias. In CATS, 298 patients from 12 hospitals in the Netherlands were included. Selection criteria included a typical history of chest pain consistent with acute myocardial infarction, onset of symptoms no more than 6 hours before admission, and ECG criteria for acute anterior myocardial infarction including at least 1-mm ST-segment elevation in leads I and aVL and/or 2-mm ST-segment elevation in two or more precordial leads of the 12-lead ECG, consistent with anterolateral, anteroseptal, and/or anterior myocardial infarction. Patients had to be eligible for thrombolytic therapy. Exclusion criteria included the presence of a prior myocardial infarction, left bundle-branch block, and severe heart failure (Killip class III or IV).
In 78 of the 298 patients, 62-lead body surface mapping was performed during sinus rhythm. Baseline characteristics of all CATS patients and of the patients who underwent a body surface mapping recording are listed in Table 1⇓. The prevalence of nondipolar map patterns in the chronic phase of myocardial infarction was determined in all 78 patients who underwent body surface mapping. The relation between nondipolarity of QRST integral maps and remodeling was analyzed in a subset of 50 patients in whom left ventricular volume assessment at 1 year was available. In addition, the association between nondipolarity and the occurrence of ventricular arrhythmias was investigated in 50 patients with Holter monitoring performed at 1 year.
Body Surface Mapping
No patients had angina pectoris and all were in sinus rhythm at the time of investigation. Patients with right bundle-branch block or second- or third-degree atrioventricular block were excluded. None of the patients had left bundle-branch block. The time interval between body surface mapping and myocardial infarction ranged from 5 to 28 months (349±141 days).
Body surface mapping was performed with the use of a portable microcomputer-based mapping system that includes a 62-lead electrode array (Fig 1⇓). This array also includes the six standard precordial leads (V1 through V6). The procedure of recording ECG signals has been described previously.24 25 In summary, 62 unipolar lead recordings were made with Wilson’s central terminal as reference. Single-beat ECG tracings were stored on floppy disk after analog-digital conversion with a sampling frequency of 500 Hz. The noise of this system was estimated to be 0.03 mV. The stored data were then transferred to an Amiga 500 microcomputer (Commodore-Amiga Inc) for further processing. Offset differences, linear baseline drifting, or both were corrected with a computer algorithm. ECG signals of unacceptable quality were replaced by values calculated from surrounding leads. The average number of leads that was rejected per measurement was 0.5±0.7.
Isopotential maps were generated at a 2-millisecond interval during the QRST complex. The onset of the QRS complex was defined as the instant in time when the earliest activation exceeded 0.05 mV; offset of the T wave was set at the point in time when the maximum positive or negative voltage dropped below 0.05 mV. U waves were not included in the QRST interval. Then, an integral map of the complete QRST interval was constructed,19 and a hard copy of this map was produced. QRST integral maps were evaluated visually by two observers who were unaware of left ventricular dimensions. Maps were defined as nondipolar if three or more extrema were present. An extreme was considered present if an area of equal polarity included recordings of at least two unipolar leads. Maps with so-called pseudopods16 without additional extrema were considered to be dipolar. If differences existed between the evaluations of the observers, maps were reevaluated and classified after discussion until consent was reached.
In addition, a supportive quantitative analysis was carried out. Singular value decomposition was applied to the QRST intervals of all 78 patients with the use of matlab software (The MathWorks Inc) installed on a Sun Sparc Station 2 computer (Sun Microsystems Inc). The first 12 eigenvectors of the entire set were derived, and the QRST integral map of each patient was expressed in terms of these eigenvectors as described by Lux et al.26 These authors demonstrated that by using the first 12 eigenvectors of the complete data set, each individual map could be reconstructed without a significant loss of information, yielding a 16:1 data reduction. In addition, the first 3 eigenvectors of their set revealed dipolar patterns, whereas eigenvectors 4 through 12 showed nondipolar patterns. Therefore, the nondipolar content of each map was defined as the contribution of eigenvectors 4 through 12 relative to the contribution of eigenvectors 1 through 12. In the first 20 patients, two consecutive measurements were carried out to assess the reproducibility of this method.
According to the CATS protocol, two-dimensional echocardiography was performed within 24 hours after admission, after 3 days, before discharge, and at 3 and 12 months after myocardial infarction. Left ventricular end-systolic and end-diastolic volumes were calculated from a two- and four-chamber view with the use of the modified biplane Simpson rule.27 The ejection fraction was calculated from these volume measurements. Measurements were made off-line from end-diastolic and end-systolic still frames with the use of a dataview microsonics cardiac analysis system (Nova Microsonics). Left ventricular volumes were indexed for body surface area. Furthermore, regional wall motion abnormalities were evaluated with the wall motion score recommended by the American Society of Echocardiography.27 In this scoring system, the left ventricle is divided into 16 segments, with each segment scored as 1 for normokinesia, 2 for hypokinesia, 3 for akinesia, 4 for dyskinesia, and 5 for an aneurysmal segment. A wall motion score index (WMSI) was computed as the sum of the scores of all segments divided by the number of segments evaluated. Twelve evaluable segments were considered a minimum to reliably assess WMSI.
The occurrence of ventricular arrhythmias was assessed with the use of two-channel 24-hour ambulatory Holter monitoring (Reynolds Medical Tracker recorder) 12 months after myocardial infarction. Analysis was performed with a Reynolds Medical pathfinder 3 analysis system. Tapes were analyzed for the presence of uniform and multiform premature ventricular beats, pairs of premature ventricular beats, and ventricular tachycardia. Pairs were defined as a repetition of two ventricular beats with a maximum interval of 0.6 second. Ventricular tachycardia was defined as a repetition of three or more ventricular beats with a rate exceeding 100 beats per minute.
High-grade ventricular arrhythmias were defined as Lown classes 4A and 4B (paired ventricular premature beats and ventricular tachycardia) because patients with repetitive forms of ventricular premature beats are at especially high risk of sudden cardiac death.28 Life-threatening ventricular arrhythmias, as mentioned in the text, were defined as primary ventricular fibrillation, ventricular tachycardia degenerating into ventricular fibrillation, or ventricular tachycardia leading to hemodynamic collapse.
Enzymatic infarct size was determined by analysis of the cumulative α-hydroxy-butyrate dehydrogenase (α-HBDH) washout curve, calculated from α-HBDH samples twice daily during the first 5 days, as described by van der Laarse et al.29
The presence of a Q-wave myocardial infarction was investigated by analysis of the 12-lead ECG. QT intervals were measured from the onset of the QRS to the end of the T wave, defined as a return to the T-P baseline. When U waves were present, the QT interval was measured to the nadir of the curve between the T and U waves. When the end of the T wave could not be reliably identified, that lead was not included in subsequent analysis. QT and QTc dispersions were determined in all patients in the body surface mapping substudy and were defined as the difference between the maximum and minimum QT intervals measured on the 12-lead ECG. Bazett’s formula was used to correct for heart rate: QTc=QT/√RR (RR interval measured in seconds).
Coronary angiography was performed when considered necessary by the individual investigator. All available angiograms were studied for the presence or absence of an occluded infarct-related artery. The severity of coronary artery disease was expressed as one-, two-, or three-vessel disease.
Data are presented as mean±SD. Differences were evaluated using paired or unpaired Student’s t test. Logistic regression analysis was used to identify variables independently predicting the occurrence of nondipolar map patterns and high-grade ventricular arrhythmias, respectively.30 A two-sided value of P<.05 was considered significant unless stated otherwise. Calculations were made with an IBM PS/2 personal computer and spss/pc+ version 4.0 software.
Body Surface Mapping
Visual analysis revealed nondipolar maps in 25 of 78 patients (32%) and dipolar maps in the remaining 53 of 78 patients (68%). Examples of typical dipolar and nondipolar QRST integral map patterns are given in Fig 2⇓. In Fig 2A⇓, a dipolar QRST integral map of a patient with a limited myocardial infarction and no left ventricular dilatation is displayed. Fig 2B⇓ features a nondipolar QRST integral map of a patient with a large myocardial infarction and a nearly twofold increase in end-systolic volume after 1 year. Body surface maps were recorded at 200 and 221 days after myocardial infarction, respectively.
Quantitative assessment of nondipolarity was performed using singular value decomposition. The first 12 eigenvectors of the complete data set and their singular values are represented in Fig 3⇓. These 12 eigenvectors include the QRST complexes of all 78 patients. This computed set of 12 eigenvectors was used to calculate the nondipolar content of the QRST interval of each patient as the contribution of eigenvectors 4 through 12 relative to eigenvectors 1 through 12. In the first 20 patients, the nondipolar content was assessed in two consecutive measurements. It was found that nondipolarity between measurements correlated well (r=.89, P<.001). The mean difference between measurements was 1±7% (range, 0% to 12%).
The nondipolar content was 36±10% of the maps that were identified as dipolar and 52±13% of the maps identified as nondipolar. The difference between these groups was statistically significant (P<.001). In contrast, QTc dispersion on the standard 12-lead ECG did not differ between patients with dipolar and those with nondipolar maps (102±46 versus 97±36 milliseconds), and there was no correlation between QTc dispersion on the 12-lead ECG and nondipolar content (r=.06, P=.632). (Additional prospective data are presented in “Appendix 1.”)
Echocardiographic data of patients with and without nondipolar maps are summarized in Table 2⇓. Left ventricular dimensions were significantly different between patients with and those without nondipolar maps. End-systolic volume indexed for body surface area (LVESVI) was already higher in the nondipolar group at day 1. This difference persisted until months 3 and 12. The absolute increase in end-systolic volume was more pronounced in patients with nondipolar maps. End-diastolic volume index (LVEDVI) was also larger in the nondipolar group from day 1 on, and here, also, more dilatation appeared to be present. However, the difference in increase in end-diastolic volume between the two groups was not statistically significant.
A lower ejection fraction and higher WMSI were found in patients with nondipolar map patterns. Ejection fraction remained stable in the first 12 months after myocardial infarction. The WMSI showed a small improvement during the first year of follow-up in patients with dipolar maps, whereas this index remained unchanged in those with nondipolar map patterns.
In Fig 4⇓, the percentage of patients with nondipolar maps is given for the four quartiles of left ventricular dilatation 1 year after myocardial infarction (see Fig 4⇓ legend). Especially in the upper quartile (increase in end-systolic volume of more than 16 mL/m2), the prevalence of nondipolar maps was high (eight of nine patients, 89%), in parallel with a higher nondipolar content in these patients (37±12% in the lower three quartiles versus 49±14% in the upper quartile, P=.013). In contrast, there was no difference in QTc dispersion on the 12-lead ECG between these groups (90±41 versus 105±37 milliseconds).
In Table 3⇓, variables other than left ventricular dilatation are considered that might explain the presence of a nondipolar map pattern. Patients with nondipolar map patterns showed a larger enzymatic infarct size. This was not in parallel with more Q-wave infarctions or a higher percentage of patients with an occluded infarct-related artery. However, multivessel disease was more prominent in patients with nondipolar maps. The number of interventions, time interval between infarction and body surface mapping, heart rate, and medication were comparable between groups. A logistic regression analysis was performed to assess whether left ventricular remodeling contributed independently to the presence of nondipolar maps. End-systolic volume at 1 year was selected as the parameter of remodeling. First, enzymatic infarct size (α-HBDH of more than 1089 U/L, the median value in this subset) was entered into the model. After this, four levels of end-systolic volume (quartiles) were added. An end-systolic volume of more than 42 mL/m2 (upper quartile) was found to contribute independently to the appearance of nondipolar maps (Table 4⇓). Although patients with an end-systolic volume of more than 42 mL/m2 in the subgroup with an infarct size of more than 1089 U/L were randomized to receive captopril more frequently than were patients with an end-systolic volume of less than 42 mL/m2 (55% versus 33%), patients with a large end-systolic volume more frequently required open-label captopril treatment (36% versus 0%), had more Q-wave infarctions (91% versus 67%), more often had multivessel coronary artery disease (50% versus 38%), and had an occluded infarct-related artery in a higher proportion of patients (50% versus 25%). None of these differences were statistically significant, possibly due to the small number of patients in these subgroups. (Additional prospective data are given in “Appendix 1.”)
Incidence of Ventricular Arrhythmias Detected During Holter Monitoring After 1 Year
Holter monitoring after 1 year showed that patients with nondipolar maps demonstrated ventricular arrhythmias more frequently (Table 5⇓). However, these differences were not statistically significant. The overall incidence of ventricular tachycardia was low. Only one patient had nonsustained ventricular tachycardia, and this patient showed a dipolar map pattern. In patients with left ventricular dilatation of more than 16 mL/m2, ventricular arrhythmias were relatively frequent. All four of six patients with high-grade ventricular arrhythmias (Lown classes 4A and 4B) in this subgroup had nondipolar map patterns.
The nondipolar content of maps in 14 patients with high-grade ventricular arrhythmias was 49±10% versus 39±17% in 41 patients without these arrhythmias (P=.013).
There was no significant difference in QTc dispersion on the 12-lead ECG between patients with and those without high-grade ventricular arrhythmias (116±44 versus 101±38 milliseconds).
In Table 6⇓, a number of possible determinants of high-grade ventricular arrhythmias are listed. A logistic regression analysis (forward-stepwise model) showed that change in end-systolic volume after 1 year was the only independent predictor of these arrhythmias.
Medication and Nondipolar Map Patterns
There were no significant differences in concomitant medication between patients with dipolar and those with nondipolar map patterns (Table 7⇓). However, patients with dipolar map patterns tended to be randomized to receive captopril more frequently, and a high percentage of patients were receiving β-blockers. Apparently, patients with nondipolar map patterns more frequently required open-label captopril treatment for congestive heart failure.
Left ventricular dysfunction is an important determinant of life-threatening ventricular arrhythmias after myocardial infarction.2 31 In recent years, experimental evidence has become available indicating that left ventricular dilatation is associated with increased dispersion of repolarization.3 4 5 6 These studies suggest that left ventricular remodeling after myocardial infarction, a process that is characterized by left ventricular dilatation, may also lead to changes in repolarization characteristics. Increased dispersion of refractoriness in dilated ventricles may lower the threshold for the occurrence of life-threatening ventricular arrhythmias.32 33
In the present study, body surface mapping was used to detect changes in repolarization characteristics accompanying left ventricular dilatation after myocardial infarction. Increased dispersion of repolarization, represented by nondipolar map patterns, appeared to be more frequent in patients with left ventricular dilatation, especially when dilatation exceeded 16 mL/m2 (89% prevalence of nondipolar maps). This was supported by a higher nondipolar content of QRST integral map patterns in these patients. In addition, when corrected for infarct size, end-systolic volume after 1 year still contributed independently to the appearance of nondipolar QRST integral map patterns. The present data indicate that left ventricular remodeling after myocardial infarction is accompanied by altered repolarization characteristics, which may facilitate the occurrence of life-threatening ventricular arrhythmias.
Changes in Left Ventricular Dimensions and Occurrence of Ventricular Arrhythmias
The prognostic significance of left ventricular dimensions was demonstrated by White et al,2 who found that end-systolic volume was the most powerful predictor of death after myocardial infarction. Because death was found to be sudden in 70% of cases in this study, left ventricular dilatation was believed to be associated with an increased risk for fatal ventricular arrhythmias. In two recent experimental studies,34 35 the relation between left ventricular volume changes and the occurrence of ventricular arrhythmias was investigated. Hansen et al35 observed in their isolated canine heart model that a sudden increase in left ventricular volume applied during diastole induced frequent ventricular premature beats and occasional nonsustained ventricular tachycardia. In a similar study, Franz et al34 investigated the arrhythmogenic effect of slow versus rapid increases in left ventricular volumes in the isolated Langendorff perfused rabbit heart. Both forms of volume change induced an increase in ventricular arrhythmias, but especially rapid volume pulses caused frequent ectopic ventricular excitation.
These studies confirm that an increase in left ventricular volume is accompanied by frequent ventricular arrhythmias. In the present study, high-grade ventricular arrhythmias were a relatively rare observation (Table 5⇑). However, paired ventricular premature beats were especially frequent in patients with dilatation of more than 16 mL/m2 in the presence of nondipolar map patterns.
Changes in Repolarization Characteristics
Changes in repolarization characteristics accompanying increased loading conditions or an increase in left ventricular volume have been described in both experiments3 in animals and clinical studies.4 5 6 Lerman et al3 demonstrated in dogs that an increase in the preload of the heart significantly shortens the refractory period. Obstruction of the right ventricular outflow tract during balloon valvuloplasty4 and performance of the Valsalva maneuver6 have also been found to reduce the refractory period. In patients being weaned from extracorporeal life support after bypass surgery, Taggart et al5 observed a shortening of the refractory period when the left ventricle was filled to normal dimensions. Because these changes in refractory period were not uniformly distributed over the ventricle, dispersion in refractoriness was considered to be increased.
No long-term studies are available showing the relation between left ventricular dilatation and changes in repolarization characteristics. All of the previous studies describe an experimental situation in which left ventricular volume was increased for a short period of time. The present study describes repolarization characteristics after chronic left ventricular dilatation.
Early afterdepolarizations were observed during increased loading conditions in two of the previously mentioned clinical studies.4 6 This phenomenon was detected during balloon valvuloplasty for pulmonary valve stenosis4 and during a Valsalva maneuver carried out during routine coronary angiography,6 both by measurement of monophasic action potentials. Early afterdepolarizations are oscillations in the membrane potential that occur during the plateau of the repolarization phase that may induce single or multiple action potentials when they reach threshold level9 (triggered activity).
In the above-mentioned studies, early afterdepolarizations were found in a minority of cases. The exact role of early afterdepolarizations in the cause of ventricular arrhythmias in patients with left ventricular dysfunction remains unclear.36
Influence of Thrombolytic Therapy
Arrhythmogenic vulnerability is decreased after thrombolytic therapy.37 Previous studies have shown that successful thrombolysis leads to a reduced incidence of late potentials, a marker of the presence of an arrhythmogenic substrate that may cause reentrant ventricular tachyarrhythmias.38 In addition, it is known that a patent infarct-related artery is associated with less left ventricular dilatation during follow-up.39 In the present study, left ventricular dilatation is associated with increased dispersion of refractoriness. Therefore, a relation between patency of the infarct-related artery and the presence of nondipolar map patterns may be expected. The finding that nondipolar map patterns were relatively more frequent in patients with an occluded infarct-related artery is consistent with this hypothesis. However, a significant difference in nondipolarity could not be detected, possibly due to a small patient cohort.
If this trend could be confirmed, it would imply that thrombolytic therapy not only reduces the risk of the development of an anatomic substrate after myocardial infarction but also is capable of reducing dispersion of refractoriness, another important arrhythmogenic factor.
Noninvasive Assessment of Repolarization Characteristics by Body Surface Mapping
Wilson et al10 suggested that the deflection area of the QRST complex (QRST integral or ventricular gradient) was determined by differences in the duration of repolarization. Subsequent theoretical models confirmed this relation.11 12 Abildskov et al13 provided direct evidence for this relation in an experimental study in dogs. Changes in refractory period induced by local warming were closely associated with similar changes in QRST integrals measured at the same epicardial site. This association was found to be independent of the site of pacing (five sites). The authors concluded that QRST integrals represent primary repolarization characteristics, which are independent of the site of ventricular activation, as opposed to secondary repolarization characteristics, which are dependent on the activation site as well as on a disturbed activation sequence due to bundle-branch block or prior infarction.
In humans, QRST integrals have been studied extensively with the use of body surface mapping.15 16 17 18 19 20 It has been demonstrated in healthy subjects that QRST integral maps obtained during sinus rhythm demonstrate a homogeneous dipolar distribution over the thorax.19 20 However, maps of patients with documented ventricular arrhythmias show inhomogeneous nondipolar patterns in a majority of cases (63% to 71%).15 16 17 18 21 This association suggests that there is an increased vulnerability for ventricular arrhythmias if QRST integrals are inhomogeneously distributed over the body surface. The underlying mechanism has not been clarified in humans. However, based on the previously mentioned experiments in animals, it may be deduced that regional differences in QRST integrals on the body surface reflect local, potentially arrhythmogenic differences in repolarization properties in the heart.
In the present study, nondipolar maps were found in 32% of all patients. A similar proportion has been found in other postinfarction studies,16 despite the application of different criteria for the classification of QRST integral maps into dipolar and nondipolar patterns. In contrast to previously mentioned studies, nondipolarity of QRST integral maps did not provide an index of serious ventricular arrhythmias in the present study, as no patient died suddenly or had a clinically relevant ventricular arrhythmia.
Nondipolar QRST Integral Map Patterns and Left Ventricular Dilatation
We hypothesized in the present study that left ventricular dilatation leads to altered repolarization properties that can be detected with body surface QRST integrals. The incidence of nondipolar QRST integral map patterns was high in patients with left ventricular dilatation: eight of nine patients with an increase in LVESVI of more than 16 mL/m2 (upper quartile) had nondipolar patterns. This relation was further confirmed by the finding of a significant increase in nondipolar content in patients with left ventricular dilatation of more than 16 mL/m2 (Fig 4⇑). These data indicate that remodeling is frequently accompanied by changes in repolarization properties that represent an underlying electrophysiological condition that may facilitate the occurrence of ventricular arrhythmias. It should be noted that these changes could not be detected with QTc dispersion on the standard 12-lead ECG.
Until now, studies of the association between left ventricular dimension and dispersion of repolarization were carried out in noninfarcted ventricles. Therefore, when the effect of remodeling on repolarization characteristics is studied, infarct size should be taken into account. In the present study, it is shown that end-systolic volume after 1 year significantly contributes to the appearance of nondipolar map patterns, regardless of enzymatically determined infarct size.
Nondipolar Content of QRST Integral Map Patterns and Ventricular Arrhythmias
In the present study, the nondipolar content of QRST integral maps was significantly higher in patients with high-grade ventricular arrhythmias during Holter monitoring. However, logistic regression analysis revealed that the change in end-systolic volume after 1 year was the only independent predictor of these arrhythmias. The nondipolar content of the QRST integrals did not contribute independently. This is probably due to the fact that in the present study, left ventricular dilatation and nondipolarity of QRST integral maps were closely related. The finding that left ventricular dilatation is a stronger predictor of ventricular arrhythmias than nondipolarity of QRST integral maps may indicate that in addition to dispersion of refractoriness, other mechanisms, such as slow conduction,40 myocardial stretch,35 or early afterdepolarizations,6 are responsible for the arrhythmogenic effects of left ventricular dilatation.
Body surface mapping was performed within a time window of 5 to 28 months (cross-sectional) after myocardial infarction had occurred and was correlated with echocardiograms obtained at days 1 and 3, before discharge, and after 3 and 12 months. QRST integral map patterns obtained when echocardiographic recordings were made might have shown different results. In addition, if the development of dispersion of repolarization is related to ventricular remodeling, a time-dependent effect on the development of nondipolar QRST integral maps may be expected. In the present study, there was no difference in the prevalence of nondipolar maps before or after 12 months of follow-up (Table 3⇑). Furthermore, there was no relation between duration of follow-up before body surface mapping and nondipolar content of the map. Therefore, a time-dependent effect could not be demonstrated in the present study.
However, “Appendix 1” provides a description of 15 new patients who underwent serial body surface mapping and echocardiography after myocardial infarction. Data from these patients show that changes in nondipolarity do occur during follow-up after myocardial infarction. In the present study, it is demonstrated that a change from a dipolar pattern to a nondipolar pattern occurs in parallel with an increase in end-systolic volume, and a conversion from a nondipolar pattern to a dipolar pattern is accompanied by a decrease in end-systolic volume. Thus, despite the occurrence of time-dependent changes in map patterns, the relation of nondipolarity in map pattern to increased end-systolic volume remains intact.
In the retrospective study, there was no direct correlation between the development of life-threatening ventricular arrhythmias and the presence or absence of nondipolar map patterns. This finding is not surprising, since in the total CATS population only 6 of 298 patients (2%) died suddenly during the first year after myocardial infarction. In addition, the present study investigated patients who survived for at least 5 months after myocardial infarction and were well enough to visit the outpatient clinic to undergo body surface mapping. Therefore, a group of relatively well patients was selected. This may also explain why no serious ventricular arrhythmias were documented and why the relation with nondipolar map patterns could not be established. However, the nondipolar content of maps in patients with high-grade ventricular arrhythmias during Holter monitoring was significantly higher than in patients without these arrhythmias, underlining the association between ventricular arrhythmias and dispersion of refractoriness.
Even though the present study demonstrates the relation between left ventricular remodeling and nondipolarity of QRST integral maps, these data do not allow prediction of repolarization characteristics on the basis of an enlarged end-systolic volume. It does imply that when risk is assessed in postinfarction patients, an enlarged end-systolic volume should be considered a potent risk factor, with increased dispersion of refractoriness as a possible underlying arrhythmogenic mechanism.
The present study describes a correlation between left ventricular remodeling and nondipolar QRST integral map patterns after myocardial infarction. These findings do not allow inference about causality or prognosis. However, they do provide a step toward the understanding of the relation between left ventricular dilatation and life-threatening ventricular arrhythmias.
The present study underlines that left ventricular remodeling may prove, through altered repolarization properties, an important risk factor for the occurrence of life-threatening ventricular arrhythmias after myocardial infarction.
To investigate the presence or absence of time-dependent changes in nondipolarity of QRST integral maps after acute myocardial infarction, 15 other patients underwent serial body surface mapping 8±5 and 640±223 days after myocardial infarction. Left ventricular dimensions were assessed with the use of echocardiography before hospital discharge and 12 months after myocardial infarction. Measurements were performed as described in “Methods.” Infarct location was assessed by standard 12-lead ECG criteria. Of all patients, 6 had an anterior, 8 had an inferior, and 1 had a lateral myocardial infarction. Only 1 patient had a previous myocardial infarction. All patients received thrombolytic therapy after hospital admission. Coronary angiography was performed before hospital discharge. In 12 of 15 patients (80%), the infarct-related artery was patent.
In Table 8⇓, it is shown that there was no significant increase in nondipolar content during follow-up. End-systolic and end-diastolic volumes increased significantly, but the ejection fraction remained relatively unchanged.
At baseline, 4 of 15 patients (27%) had nondipolar maps. Of these patients, 2 also had a nondipolar map at follow-up, and 2 featured a change of a nondipolar into a dipolar map. Two other patients had dipolar maps at baseline and developed nondipolar maps at follow-up.
The 2 patients who demonstrated conversion from dipolar to nondipolar maps showed an increase in end-systolic volume of 7 and 5 mL/m2, resulting in an LVESVI of 39 and 30 mL/m2, respectively. The nondipolar content increased with 26% and 3% to 58% and 37%, respectively. The 2 patients for whom maps changed from nondipolar to dipolar showed a decrease in end-systolic volume of 5 and 2 mL/m2, resulting in an LVESVI of 18 and 26 mL/m2, respectively. The nondipolar content of the maps in these patients decreased from 14% and 9% to 43% and 35%, respectively.
At baseline, there were no significant differences in infarct size, end-systolic volume, end-diastolic volume, or ejection fraction between patients with dipolar and those with nondipolar map patterns. However, patients who showed nondipolar map patterns at follow-up had a significantly larger end-systolic volume after 1 year (32.85±5.31 versus 22.02±5.37 mL/m2, P=.004). The increase in end-systolic volume was also more pronounced in patients with nondipolar maps, although this difference was not significant (8.79±5.30 versus 2.05±6.12, P=.074). Multiple regression analysis, with nondipolar content at follow-up as a continuous dependent variable, revealed that the nondipolar content at baseline and the end-systolic volume after 1 year were the only independent predictors of the nondipolar content at follow-up, accounting for almost 60% of the variation of this variable (R2=.59).
We conclude that the data from this small set of prospectively evaluated patients support the hypothesis that left ventricular remodeling and nondipolar content as a measure of dispersion of refractoriness are clearly related.
Principal Investigators and Participating Centers
P.H. van der Burgh, MD, Medisch Spectrum Twente, Enschede; R. Enthoven, MD, PhD, St Sophia Ziekenhuis, Zwolle; J.P.M. Hamer, MD, PhD, Academisch Ziekenhuis, Groningen; O. Kamp, MD, PhD, Academisch Ziekenhuis VU, Amsterdam; J.H. Kingma, MD, PhD, St Antonius Ziekenhuis, Nieuwegein; D.J.A. Lok, MD, St Geertruiden Ziekenhuis, Deventer; C.H. Peels, MD, Catharina Ziekenhuis, Eindhoven; and R.P. Wielenga, MD, St Ignatius Ziekenhuis, Breda, Netherlands.
The authors thank Drs A.C. Linnenbank and M. Potse for support concerning the eigenvector analysis and editorial suggestions; J.G.P. Tijssen, PhD, for valuable statistical advice; and W.A.M. Dolman and A.J. Dijkstra for technical assistance in recording the body surface maps.
↵1 List of principal investigators and participating centers is provided in “Appendix 2.”
- Received December 28, 1994.
- Accepted January 31, 1995.
- Copyright © 1995 by American Heart Association
Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial infarction: experimental observations and clinical implications. Circulation. 1990;81:1161-1172.
White HD, Norris RM, Brown MA, Brandt PWT, Whitlock RML, Wildt CJ. Left ventricular end-systolic volume as the major determinant of survival after recovery from myocardial infarction. Circulation. 1987;76:44-51.
Lerman BB, Burkhoff D, Yue DT, Franz MR, Sagawa K. Mechanoelectrical feedback: independent role of preload and contractility in modulation of canine ventricular excitability. J Clin Invest. 1985;76:1843-1850.
Levine JH, Guarnieri T, Kadish AH, White RI, Calkins H, Kan JS. Changes in myocardial repolarization in patients undergoing balloon valvuloplasty for congenital pulmonary stenosis: evidence for contraction-excitation feedback in humans. Circulation. 1988;77:70-77.
Taggart P, Sutton PM, Treasure T, Lab M, O’Brien W, Runnalls M, Swanton RH, Emanuel RW. Monophasic action potentials at discontinuation of cardiopulmonary bypass: evidence for contraction-excitation feedback in man. Circulation. 1988;77:1266-1275.
Taggart P, Sutton P, John R, Lab M, Swanton H. Monophasic action potential recordings during acute changes in ventricular loading induced by the Valsalva manoeuvre. Br Heart J. 1992;67:221-229.
Han J, Moe GK. Nonuniform recovery of excitability in ventricular muscle. Circ Res. 1964;14:44-60.
Allessie MA, Bonke FM, Schopman FJG. Circus movement in rabbit atrial muscle as a mechanism of tachycardia, II: the role of nonuniform recovery of excitability in the occurrence of unidirectional block, as studied with multiple microelectrodes. Circ Res. 1976;39:168-177.
Binah O, Rosen MR. Mechanisms of ventricular arrhythmias. Circulation. 1992;85(suppl I):I-25-I-31.
Abildskov JA, Evans AK, Lux RL, Burgess MJ. Ventricular recovery properties and QRST deflection area in cardiac electrograms. Am J Physiol. 1980;239:H227-H231.
SippensGroenewegen A, Muilwijk SLC, Kingma JH, van Hemel NM, Hauer RNW, Janse MJ, Dunning AJ. Incidence of nonuniform ventricular recovery in patients with postinfarction ventricular arrhythmias: significance of infarct location and type of spontaneously occurring arrhythmia. Eur Heart J. 1991;12(II-IV):109. Abstract.
Gardner MJ, Montague TJ, Armstrong CS, Horacek BM, Smith ER. Vulnerability to ventricular arrhythmia: assessment by mapping of body surface potential. Circulation. 1986;73:684-692.
Mitchell LB, Hubley-Kozey CL, Smith ER, Wyse DG, Duff HJ, Gillis AM, Horacek BM. Electrocardiographic body surface mapping in patients with ventricular tachycardia: assessment of utility in the identification of effective pharmacological therapy. Circulation. 1992;86:383-393.
Montague TJ, Smith ER, Cameron DA, Rauraharju PM, Klassen GA, Felmington CS, Horacek BM. Isointegral analysis of body surface maps: surface distribution and temporal variability in normal subjects. Circulation. 1981;63:1166-1172.
Abildskov JA, Green LS, Lux RL. Detection of disparate ventricular repolarisation by means of the body surface electrocardiogram. In: Cardiac Electrophysiology and Arrhythmias. Orlando, Fla: Grune & Stratton; 1985:495-499.
Yasumura S, Kubota I, Ikeda K, Tsuiki K, Yasui S. Using body surface mapping to detect vulnerability to ventricular arrhythmias in patients with coronary artery disease. J Electrocardiol. 1987;20:114-120.
Abildskov JA, Green LS. The recognition of arrhythmia vulnerability by body surface electrocardiographic mapping. Circulation. 1987;75(suppl III):III-79-III-85.
Kingma JH, van Gilst WH, Peels KH, Dambrink J-HE, Verheugt FWA, Wielenga RP. Acute intervention with captopril during thrombolysis in patients with first anterior myocardial infarction. Eur Heart J. 1994;15:898-907.
Reek EJ, Grimbergen CA, van Oosterom A. A low-cost 64-channel microcomputer-based data acquisition system for bedside registration of body surface maps. In: d’Alche P, ed. Advances in Electrocardiology. Caen, France: Centre de Publications de l’Universite de Caen; 1985:112-114.
SippensGroenewegen A, Spekhorst H, van Hemel NM, Kingma JH, Hauer RNW, Janse MJ, Dunning AJ. Body surface mapping of ectopic left and right ventricular activation: QRS spectrum in patients without structural heart disease. Circulation. 1990;82:879-896.
Lux RL, Evans AK, Burgess MJ, Wyatt RF, Abildskov JA. Redundancy reduction for improved display and analysis of body surface potential maps. Circ Res. 1981;49:186-196.
Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH, Tajik AJ. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-dimensional Echocardiograms. J Am Soc Echocardiogr. 1989;2:358-367.
Bigger JT Jr, Weld FM, Rolnitzky LM. Which postinfarction ventricular arrhythmias should be treated? Am Heart J. 1982;103(pt 2):660-666.
van der Laarse A, Kerkhof PL, Vermeer F, Serruys PW, Hermens WT, Verheugt FW, Bar FW, Krauss XH, van der Wall EE, Simoons ML. Relation between infarct size and left ventricular performance assessed in patients with first acute myocardial infarction randomized to intracoronary thrombolytic therapy or to conventional treatment. Am J Cardiol. 1988;61:1-7.
Altman DG. Relation between several variables. In: Practical Statistics for Medical Research. London, UK: Chapman & Hall; 1990:325-364.
Kuo CS, Atarashi H, Reddy CP, Surawicz B. Dispersion of ventricular repolarization and arrhythmia: study of two consecutive ventricular premature complexes. Circulation. 1985;72:370-376.
Kuo CS, Reddy CP, Munakata K, Surawicz B. Mechanism of ventricular arrhythmias caused by increased dispersion of repolarization. Eur Heart J. 1985;6(suppl D):63-70.
Franz MR, Cima R, Wang D, Profitt D, Kurtz R. Electrophysiological effects of myocardial stretch and mechanical determinants of stretch-activated arrhythmias. Circulation. 1992;86:968-978.
Hansen DE, Craig CS, Hondeghem LM. Stretch-induced arrhythmias in the isolated canine ventricle. Circulation. 1990;81:1094-1105.
Taggart P, Sutton P, Lab M. Interaction between ventricular loading and repolarisation: relevance to arrhythmogenesis. Br Heart J. 1992;67:213-215.
Kersschot IE, Brugada P, Ramentol M, Zehender M, Waldecker B, Stevenson WG, Geibel A, De Zwaan C, Wellens HJ. Effects of early reperfusion in acute myocardial infarction on arrhythmias induced by programmed stimulation: a prospective, randomized study. J Am Coll Cardiol. 1986;7:1234-1242.
Dambrink J-HE, Tobé TJM, Peels CH, van Gilst WH, Kingma JH, on behalf of the CATS Investigators Group. Relation between left ventricular dilatation and QRS duration after a first anterior myocardial infarction. Eur Heart J. 1994;15S:494. Abstract.
Taccardi B. Distribution of heart potentials on the thoracic surface of normal human subjects. Circ Res. 1963;12:341-352.