T-Wave Complexity: Diagnostic and Prognostic Significance

Luigi De Ambroggi

Dept. of Cardiology, Istituto Policlinico San Donato, University of Milan, Italy

Correspondence: L De Ambroggi, Istituto Policlinico San Donato, 20097 San Donato Milanese, Italy.
E-mail: luigi.deambroggi@unimi.it, phone/fax +39.02.52774520

1.  Introduction

The T wave morphology reflects the sequence of repolarization process in the ventricles, and experimental studies on wedge preparations have clearly demonstrated how variations of the “physiological” transmural dispersion of repolarization result in different patterns of T-U waves (Yan and Antzelevitch, 1998).

Many investigations have focused on the key role of ventricular repolarization abnormalities in the genesis of cardiac arrhythmias. Schematically, vulnerability to arrhythmias can arise from 2 conditions of repolarization process: 1) a state of heterogeneity of repolarization, i.e. a greater than normal dispersion of recovery times, and 2) a dynamic (beat to beat) variation of repolarization sequence. This last condition, which frequently occurs in ischemic situations, can be detected by different methods (e.g., analysis of T wave alternans, RR/QT relation variations). The first condition can be detected by analyzing a single beat, using the 12-lead ECG or multiple thoracic leads, that is body surface potential maps (BSPM).

2.  Body Surface Potential Maps

BSPM have 2 major advantages over the conventional 12 leads: 1. to explore the entire chest surface (in fact, BSPM is the only method that provides all the information on the cardiac electric field available at the body surface); 2. to be more sensitive in detecting local electrical events, such as local conduction disturbances or regional heterogeneities of ventricular recovery. On the other hand, the method of BSM is impractical for wide clinical use.

Different methods of analysis of BSPM have been used to study repolarization potentials.

QRST Integral Maps

Areas of QRST deflections mainly reflect the intrinsic repolarization properties and are largely independent of ventricular excitation sequence. A complex, multipolar pattern has been related, on the basis of experimental observations, to local heterogeneities of the ventricular recovery process and thus to cardiac states of vulnerability to arrhythmias (Urie et al., 1978). In our experience, a clear multipolar pattern is visible only in a small percentage of patients affected by ventricular arrhythmias. A multipolar distribution most likely reflects only gross regional inequalities of repolarization, and may not represent a marker sufficiently sensitive for minor disparities.

Eigenvector Analysis

In 1981 Lux et al. proposed a method by which each potential map could be represented as a weighted sum of a limited number of fundamental patterns (eigenvectors) common to both control subjects and patients. This method makes possible the detection and quantitation of nondipolar components not evident on visual inspection of the integral maps. In fact, the first 3 eigenvectors, displayed as eigenvector contour maps, generally show a smooth dipolar distribution with different locations of the peak values, whereas the eigenvectors beyond the 3rd have a multipolar distribution. Thus, the cumulative contribution of the eigenvectors beyond the 3rd to an individual map, expressed as percentage contribution of the total map content, has been considered the “non-dipolar content” of that map. According to this method, we calculated the non-dipolar content of QRST integral maps in different groups of patients (LQTS, old myocardial infarction with and without episodes of sustained ventricular tachycardia) and in a control group of healthy subjects (De Ambroggi et al., 1986; Bertoni et al., 1987). On average, the non-dipolar content was significantly lower in controls than in LQTS patients or in patients with MI and ventricular tachycardia.

Principal Component Analysis

By applying principal component analysis of all ST-T waves recorded, we proposed to compute the similarity index (ratio of first eigenvalue by the sum of all eigenvalues). The value of similarity index is inversely proportional to the variability of T wave morphologies and a low value is considered a marker of repolarization heterogeneity. In our experience, similarity index was significantly lower than normal in patients affected by congenital LQTS (De Ambroggi et al., 1991), in patients with arrhythmogenic right ventricular dysplasia and ventricular tachycardias (De Ambroggi et al., 1997) and in patients with myocardial infarction.

Other repolarization variables

In order to analyzing the instantaneous variations of repolarization potentials we considered two indices: early repolarization deviation index (ERDI) and late repolarization deviation index (LRDI). Visually, the pattern of potential maps is generally constant during normal repolarization, apart from changes in amplitude. The ERDI and LRDI are numerical indices which describe deviations from this behavior during repolarization, from the J point to the T peak and from the peak to the end of T wave, respectively (Corlan and De Ambroggi, 2000). We computed these indices in small series of patients with different cardiac diseases (ARVD, LVH due to aortic stenosis, myocardial infarction with and without arrhythmias), and in some groups significant differences from normal subjects were found.

An other way to explore repolarization complexity was recently proposed by Fuller et al. (2000). In an experimental model of a dog heart suspended in a torso shaped tank they demonstrated that the T wave width measured on the root mean square tracing of the surface ECGs is significantly correlated with the repolarization dispersion calculated directly from recovery times of the 64 epicardial electrograms.

3.  12-Lead ECG

In recent years various methods for quantification of repolarization heterogeneity from the standard 12-lead ECG have been proposed.

QT Dispersion (QTd)

The measurement of 12-lead QT interval dispersion was widely used as an index of repolarization heterogeneity mainly because of its simplicity, but it has several limitations. The major limitation is that this measure can not be related to the “true” spatial heterogeneity of repolarization, since each surface lead is influenced by the electrical activity of the entire heart. Moreover there are other well-known methodological limitations (e.g., accuracy of measurements, inter-intraobserver variability, number of leads used) which can partly explain the controversial results reported in the literature (Malik, 2000). In summary, whereas initial results coming from small retrospective studies seemed to prove the prognostic value of QTd as risk stratifier, more recent prospective trials did not confirm these data (Zabel et al.,1998). Actually, QTd can be considered a gross estimate of repolarization abnormalities.

Principal Component Analysis (PCA)

PCA has been applied also for analysis of the 12-lead ECG waveforms. Generally, as for BSPM, the method defines several independent components, that contain all the information of the T waves of the 12-lead ECG. Recently Okin et al. (2002) reported that an increased PCA ratio was an independent predictor of cardiovascular mortality in a large population of American Indians.

T Wave Morphology Descriptors

In order to identify more precise descriptors of the 12-lead T wave morphology a set of new variables has been proposed (Acar et al., 1999), that measure the spatial and temporal variations of T wave morphology, the difference of the mean wavefront direction between ventricular depolarization and repolarization, the non-dipolar component. These variables have the advantage to be not critically dependent on time domain measurements (as the identification of T wave end) and show good reproducibility.

The prognostic value of the total cosine R-to-T (TCRT), an estimate of the angle between repolarization and depolarization wavefront, and the T wave loop dispersion were found significantly associated with clinical events in 261 post-myocardial infarction patients (Zabel et al., 2000).

Recently, an other descriptor the T wave residuum (TWR) was found to predict mortality in 772 US veterans with cardiovascular diseases followed-up for 10.4 ± 3.8 years (Zabel et al., 2002). On univariate analysis, patients with T wave residua above the median value had significant worse survival compared to patients with values below the median. After entering into a stepwise backward Cox regression analysis clinical variables and T wave loop morphology variables, TWR maintains an independent prognostic value for prediction of mortality.

4.  Conclusion

Complexity of the surface T waves reflects complexity of the repolarization process in the heart, which in turn influence arrhythmogenesis.

Several variables can describe T wave complexity with different degree of accuracy. Nevertheless, the pathophysiologic meaning of each descriptor, that characterizes the complexity, has still to be clarified by experimental and modeling studies.

References

Acar B, Yi G, Hnatkova K, et al. Spatial, temporal and wavefront direction characteristics of 12-lead T wave morphology. Med Biol Eng Comput 37:574-84, 1999.

Bertoni T., Breghi ML., Marconi M., Bonifaccio G., De Ambroggi L. Usefulness of the integral maps to detect vulnerability to malignant arrhythmias in patients with old myocardial infarction. In: Schubert E. ed., Electrocardiology '87. Berlin, Akademie-Verlag, 1988, 247-250

De Ambroggi L., Bertoni T., Locati E., Stramba-Badiale M., Schwartz PJ. Mapping of body surface potentials in patients with the idiopathic long QT syndrome. Circulation 74:1334-1345, 1986

De Ambroggi L, Negroni MS, Monza E, Bertoni T, Schwartz PJ. Dispersion of ventricular repolarization in the long QT syndrome. Am J Cardiol 68:614-620, 1991.

De Ambroggi L, Aimè E, Ceriotti C, Rovida M, Negroni S. Mapping of ventricular repolarization potentials in patients with arrhythmogenic right ventricular dysplasia. Principal component analysis of the ST-T waves. Circulation 96: 4314-4318, 1997.

Corlan A.D, De Ambroggi L. New quantitative methods of ventricular repolarization analysis in patients with left ventricular hypertrophy. Ital Heart J 1: 542-548, 2000.

Fuller MS, Sandor G, Punske B, et al. Estimates of repolarization dispersion from electrocardiographic measurements. Circulation 102: 685-691, 2000.

Lux RL, Evans AK, Burgess MJ, Wyatt RF, Abildskov JA. Redundancy reduction for improved display and analysis of body surface potential maps. I. Spatial compression. Circ Res 49:186-196, 1981.

Malik M. QT dispersion: time for an obituary ? Eur Heart J 21: 955-57, 2000

Okin PM, Devereux RB, Fabsitz RR, et al. Principal component analysis of the T wave and prediction of cardiovascular mortality in american Indians. The strong heart study. Circulation 105: 714-719, 2002.

Urie PM, Burgess MJ, Lux RL, Wyatt RF, Abildskov JA. The electrocardiographic recognition of cardiac states at high risk of ventricular arrhythmias. Circ Res 42:350-358, 1978.

Yan GX, Antzelevitch C. Cellular basis for the normal T wave and the electrocardiographic manifestations of the long QT syndrome. Circulation 98:1928-1936, 1998.

Zabel M, Klingenheben T, Franz MR, Hohnloser S. Assessment of QT dispersion for prediction of mortality or arrhythmic events after myocardial infarction. Results of a prospective, long-term follow-up study. Circulation 97: 2543-50, 1998.

Zabel M, Acar,B, Klingenheben T, et al. Analysis of T wave morphology for risk stratification after myocardial infarction. Circulation 102: 1252-57, 2000.

Zabel M, Malik M, Hnatkova K, et al. Analysis of T wave morphology from the 12-lead electrocardiogram for prediction of long-term prognosis in male US veterans. Circulation 105: 1066-70, 2002.