Ventricular repolarization in diabetes mellitus: still much to learn.

Abstract

More than a decade has passed since the revival of the concept of QT interval interlead variability [1]. The difference between maximum and minimum QT intervals was called QT dispersion and proposed as an index of spatial inhomogeneity of ventricular repolarization, a presumed marker of arrhythmia risk. Since then, a large amount of scientific work has been reported and some recent reviews published on this subject [2, 3]. Nevertheless, QT interval dispersion has not fulfilled many of its initial promises. As pointed out by Surawicz [4] in 1996, four mandatory pre-conditions must be accomplished before QT dispersion clinical utility be established; and these problems have not been solved yet. The measurement methods of QT interval duration have not been standardized,making it difficult to compare results from different investigations. Although normal values have been proposed (less than 50–60 ms), there is a large overlap range between normal and abnormal values and, in the abnormal range, between those prone or not to serious ventricular arrhythmias. The question about its sensitivity and specificity has also not been solved,although they seem comparable to those of other known electrocardiographic risk predictors, such as T-wave alternans, RR variability or signal-averaged ECG [5]. And finally, there is still debate about what QT interval dispersion really means [6]. We no longer believe in the original concept that it reflects the spatial dispersion of ventricular recovery times, a measure of regional heterogeneity of ventricular repolarization [1], as supported by some early reports [7, 8]. More recent investigations [9, 10] demonstrated that increased QT dispersion probably signifies a global abnormal ventricular repolarization pattern, reflected by an uncommon projection of a more complex vetorcardiographic T-wave loop morphology. The main drawback of QT interval dispersion measurement is its recognized poor reproducibility [11], which lies basically in two aspects. First, the difficulty of determining the T-wave end, particularly in ECGs with abnormal T-wave patterns. Second, the fact that QT dispersion is a relatively small value compared to the QT interval duration. So a relatively small error in QT interval measurement amplifies the error in QT dispersion calculation.Thus,studies with a small number of ECG evaluations are the least probable of concluding or elucidating clinical or physiological relationships with repolarization abnormalities, because most of their demonstrated differences could rest simply on measurement errors. That is also the reason why no repolarization parameter should be solely used as risk stratifier or cardiovascular event predictor. In summary, we think that eletrocardiographic ventricular repolarization parameters in general, and QT dispersion in particular,should be considered crude and imprecise measurements of global ventricular repolarization abnormalities, which by no means signifies that they have no value at all.As ventricular repolarization is of considerable importance in several conditions, such as in diabetes mellitus, we think that it is plausible to quantify its abnormalities, with the presumption that increasing values are probably associated with the severity of repolarization abnormality. Although some new descriptors of altered ventricular repolarization patterns have been investigated, such as T-wave axis or principal component analysis of T-wave vector loop, at present QT interval duration and dispersion remain the C R 0 53 C. Cardoso, MD, PhD · G. Salles, MD Internal Medicine Department Clementino Fraga Filho University Hospital Federal University of Rio de Janeiro Rio de Janeiro, Brazil