Non-Invasive Evaluation o Ventricular Repolarisation

Philippe Coumel, M.D, F.E.S.C.
Hôpital Lariboisière, 2, Rue Ambroise-Paré, 75010- Paris, France.

Conventional electrocardiography has always dealt with ventricular repolarization in terms of morphology rather than of duration and variability. It is easy for the cardiologist to diagnose at a glance abnormalities suggesting ischemia or left ventricular hypertrophy. However, evaluating the QT duration is not part of routine ECG lecture, and even more so for variability. The cardiologist only knows of QT dynamicity through the formula first proposed by Bazett (QTc = QT x _RR [in sec]) but in practice he rarely uses it. There was a long-lasting controversy in the literature about the poor prognosis significance of a prolonged QTc interval in heart disease. In fact QTc actually is the marker of a poor prognosis only when observed in the absence of congestive heart failure. In itself this notion shows how much the significance of QT duration depends on the context in which it is observed. In any case this information is important because it is the only one available for the clinician on what is going on at the cellular level in term of action  potential duration and behavior.

DYNAMICITY OF VENTRICULAR REPOLARIZATION

A traditionally admitted notion is that QT duration variations have an exponential pattern when plotted against RR intervals, as expressed by Bazett formula. In fact this formula reflects the QT/RR relationships established from comparing individual ECG records obtained in comparable conditions (i.e. at rest during daytime). This should be reconsidered. The QT duration physiologically depends on the heart rate and on the autonomic nervous system, but obviously these two variables are by no means independent. One can easily verify that QT interval shortens in two different conditions: heart rate acceleration obtained by pacing, or exercise in a patient whose heart frequency is dependent on a fixed rate pacemaker. Such approaches have always found linear relationships between QT and RR values, at variance from what is obtained by exercising an individual in sinus rhythm.

The large number of correction formulae proposed after Bazett are strictly related to the conditions of provoking the QT modification, and in fact they only have in common not to form reliable tools for exploring QT dynamicity. Two resting individuals with heart rates of, say, 70 and 90 beats per minute are not equivalent, in terms of vago-sympathetic balance, to a third individual whose heart rate goes from 70 to 90 on the occasion of exercise. All formulae have in common not to propose a linear relationship between QT and RR, which most probably results from the combination of at least two (or more) linear relations with different slopes operating at the same time.

Thus, studying QT dynamicity meets important difficulties. Not only QT measurement is not easy to make, but it is difficult to select pure information concerning the rate on the one hand, the autonomic influences on the other hand. The Holter technique gives the opportunity of a new approach because of the possibilities offered by computerized selection of data. Measuring exactly QT duration is indeed difficult, but once determined, the measurement algorithm can reliably evaluate differences of QT intervals with a precision of the order of one millisecond or so. Holter recordings are technically imperfect, but this problem disappears when the measurement is performed on templates of averaged QRS-T complexes. In fact the problem is to collect the data in such a way that the information contained is homogeneous.

Relation between QT and RR intervals at Holter can be studied on a beat-by-beat or on an hourly basis, but in any case the correlation is always linear with a Pearson coefficient of the order of 0.90. However, it is very easy to show that QT duration at identical rates is shorter at daytime and at night by about 20 to 30 ms. Thus, one can easily realize that no correction formula will ever apply to comparing QT duration at daytime and at night at the same heart rate.

One the condition to collect the QT data in segments of recording including a stable heart rate (because QT changes consecutive to heart rate variations are not instantaneous and take about one minute) on can easily explore the QT rate-dependence at daytime and at night (figure) and verify that the slope of the regression line is physiologically steeper at daytime compared to night. The basic value of the QT/RR slope at night on the one hand, and its modification at daytime compared to night seem to be the important characteristics of ventricular repolarization when one explores the diseased hearts. Normal values of the slopes have been determined in our laboratory. They slightly vary with age and sex. The following values should be considered as normal in men and women : 0.154 and 0.195, respectively  [95% Confidence Interval 0.028-0.053] for the day period, 0.133 and 0.170 at night [95% CI 0.022-0.051]

QT DYNAMICITY IN HEART DISEASES

We have recently explored QT dynamicity by the preceding method in various conditions and the results can be briefly summarised.

Occurrence of arrhythmic events in long-term follow up of infarcted patients: 2 cohorts of patients with an old myocardial infarction were compared, and 30 patients with ventricular tachyarrhythmias during the follow-up were matched with patients without such events but no different in terms of age, sex, NYHA functional class and left ventricular ejection fraction. The cohort of patients with ventricular tachyarrhythmia and/or cardiac arrest at the outcome significanly differed from controls by a steeper QT/RR slope at night, and a reduced difference between daytime and nighttime slopes. The significance was more marked in terms of QT dynamicity than of heart rate variability in these patients.

QT dynamicity and heart failure: patients with and without left ventricular hypertrophy, with and without heart failure were compared to controls, and the QT/RR slopes as well as the day-to-night differences allow to differenciate the groups according to the following trends : the more severe the disease, the steeper the night slope, the less marked the day-to-night difference of the slopes, and the more reduced the difference of the QT duration between day and night at the same heart rate.

QT dynamicity and the long QT syndrome: now that the possibility is offered to characterize for sure patients who are carriers or non carriers of the various genetical forms of the long QT syndrome, we had the opportunity to study a large cohort of families totalizing more than 100 individuals, and to characterize the QT dynamicity according to the presence or the absence of the disease, and to the presence or the absence of symptoms in carriers. The long QT syndrome in its  KvLQT1 form is characterized by a steeper QT/RR slope at night in carriers compared to non-carriers. More importantly, symptomatic carriers not only have a steeper night slope compared to non symptomatic carriers, but the former have a daytime slope that is less steep than the night slope : they have a paradoxical pattern of this fundamental behavior. Recalling that the circadian changes of the QT/RR slope reflect the autonomic influences on the basic phenomenon of rate-dependence, it is clear that this inverted behavior of QT dynamicity certainly has to do closely with the fundamental mechanism of the ventricular arrhythmias in this syndrome.  In conclusion, studying dynamicity of ventricular repolarization as a function of heart rate and autonomic influences considered separately is a very promising diagnostic and prognostic approach of heart disease, and the fundamental reason for this situation is that it provides the best information we can obtain from surface ECG on cellular electrophysiology.