Is the apico-basal gradient larger than the transmural gradient?

Abstract

In this issue of the Journal of Cardiovascular Pharmacology Bauer and colleagues (1) describe experiments in which needle electrodes were used to investigate differences in effective refractory periods (ERP) both transmurally and along the baso-apical axis in the canine left ventricle. This work is an extension of previous work of the same group, which demonstrated absence of a functionally relevant transmural gradient in the canine left ventricle (2). In the present study the authors confirm the absence of such a transmural gradient both at baseline and after the administration of either dofetilide, which blocks the rapid component (IKr) of the delayed rectifier current or chromanol 293B, which blocks the slow component (IKs) of the delayed rectifier current. The novel information in this study concerns three issues. i) The authors describe an apico-basal difference in ERP at baseline: at 850 ms cycle length ERPs are longer at the apex than at the base. ii) Dofetilide increases this baso-apical difference, because it prolongs refractoriness more at the apex than at the base. iii) Chromanol 293B abolishes this difference. DISPERSION IN ACTION POTENTIAL DURATION OF REFRACTORY PERIODS Under normal conditions it seems permitted not to distinguish between dispersion in refractory periods or action potential duration, because the former is more or less a function of the latter. This relation is lost during acute ischemia when postrepolarization refractoriness develops (3). In heart failure this relationship between action potential duration and refractory periods is less established, simply because of lack of data. Dispersion of refractoriness is an important electrophysiologic parameter for reentry. Han and Moe (4) and Janse et al. (5) were the first to show that differences between longest and shortest refractory periods in the dog ventricle were in the order of 40 ms. In the atria, Allessie at al. (6) showed that a difference in refractoriness of about 15 msec is sufficient for the occurrence of unidirectional block followed by reentry. In terms of arrhythmogenesis it is not only important how large dispersion is, but also ` where ' it is, or, more precisely over which distance. Thus, in the study of Allessie et al. (6) the dispersion occurred over adjacent sites. A gross dispersion of 40 ms over the whole ventricle as in the dog (3,4) cannot easily be interpreted as an arrhythmogenic substrate when spatial information is lacking. In fact, an amount of dispersion in refractory periods (or action potential duration), which would exactly match the time needed for full activation of the heart, would render the ventricle less vulnerable for arrhythmias. At the same time such complete synchronization of the moment of repolarization over the ventricles would lead to disappearance of the T-wave from the EKG. Thus, regions of long and short refractory periods must be close together for (unidirectional) conduction block to occur. Another prerequisite is a sufficiently large size of the area of unidirectional block. If the area with long refractory period is small, no reentry will be possible. In normal ventricular myocardium, dispersion of refractory periods is not large enough to initiate reentry. Dispersion of refractory period can be increased by local cooling and values of 95-145 ms allow premature stimuli at the site of shortest refractoriness to induce arrhythmias in the canine left ventricle (7). Similar increases of the dispersion of refractory period are observed during acute ischemia (8). THREE TYPES OF REGIONAL DISPERSION IN THE VENTRICLE There is dispersion in refractory periods, action potential duration and in density of underlying membrane currents between the right ventricle and the left ventricle, between the base and the apex and transmurally between endocardium and epicardium with an intermediate M region. There are no studies that systematically have compared local electrograms (activation recovery intervals) with action potential duration and local refractory periods and with local membrane currents within one and the same study. It has been shown that the Ito current, responsible for phase 1 of the action potential (the early repolarization directly following the upstroke) and the IKs current are larger in the right ventricle of the dog than in the left ventricle (9). Secondly, in rabbit ventricle the density of IKs is larger at the base than at the apex and also larger than IKr both at the base and at the apex (10). Together the delayed rectifier currents were larger at the base than at the apex. The ratio of IKs and IKr is different as well, being 0.5 at the apex and 3 at the base. Finally, in the transmural direction the IKs current is smaller in the midmyocardial M region than at the endocardium and at the epicardium, at least in the dog (11), although the functional relevance is a debated issue (12–15). The absence of a transmural gradient in refractory periods in the present study (1) is another “brick in the wall” in the vivid debate on the significance of M-cells. In this brief editorial we will focus on the novel observations on apico-basal differences in refractory periods, which is -in contrast to the literature on the transmural gradients-a topic with very limited information from previous studies; for review see Burton and Cobb (16). APICO-BASAL DISPERSION IN ACTION POTENTIAL DURATION OR REFRACTORY PERIODS In the present study the refractory periods were longer at the apex than at the base. This is in agreement with data in the guinea pig where the apex-base difference is 25 ms and the transmural difference is 16 ms (17). Also in the rabbit heart, action potential duration is longer at the apex than at the base (10). In contrast, in pig hearts the action potentials are longer at the base than at the apex (18). The apico-basal disparity in refractory periods might be attributed to the difference in expression of delayed rectifier current. Thus, the group of Kodama and colleagues (10) showed that total delayed rectifier current is smaller at the apex than at the base, which would be in agreement with the longer apical refractoriness (1) and action potential duration in the rabbit left ventricle (10). However, for the components of the delayed rectifier these gradients are different (10), which complicates the picture at different heart rate and in response to antiarrhythmic agents. Interestingly, apico-basal differences in the expression of the HERG gene responsible for the IKr current, has also been demonstrated in the ferret ventricle (19). Viswanathan and Rudy investigated the effect of relative densities of IKr and IKs on action potential duration (20). This study revealed that cells with a low IK value, but a relatively high IKr content give rise to a large increase in action potential duration, when IKr is reduced. This is in line with the observations made by Bauer et al. (1). The increase in ERP dispersion that the authors observed after application of dofetilide is small compared to the increase in dispersion of action potential duration observed by Kodama's group (10). The latter group, however, carried out the measurements in isolated cells, and it is likely that coupling in the intact heart weakens the effect of dofetilide. In the simulation study of Viswanathan and Rudy (20), a reduction of IKs eliminates the action potential duration differences. This is similar to the effects of chromanol 293B, observed by Bauer et al. (1). They report the disappearance of the dispersion in refractoriness along the apico-basal axis in response to chromanol. This effect of IKs blockers shows species variability and depends also on heart rate (21). The deactivation of the current has to be slow enough to allow relevant amplitude of the current between two action potentials with a short coupling interval, either at high heart rate or after a closely coupled extra beat as in guinea pig ventricle (21). If the deactivation of IKs is fast enough as in rabbit ventricle, there is no relevant IKs current at high heart rate anyway, with or without blockade (21). ABSENCE OF A TRANSMURAL GRADIENT IN REFRACTORY PERIODS In this study Bauer et al. (1) -again- (2) report absence of transmural differences in the refractory period in contrast to previous data from the group of Antzelevitch (22). Ventricular myocardium of various species consists of three, electrophysiologically different, cell types: epicardial, endocardial and M-cells. The most convincing evidence for the existence of these cell types is obtained from canine ventricle (11,12,14,22), although such differences have been observed in cells isolated from the human right ventricle (23) and in wedge preparations from the human left ventricle as well (24). The M-cells, which are located between the epi- and endocardial cells, differ from them in that M-cells have longer action potential duration, especially at low heart rates. This is caused by a smaller, slowly activating delayed rectifier current IKs in M cells as compared to epi and endocardial cells (11). A detailed discussion about this topic is outside the topic of this editorial, but in the intact canine ventricle dispersion of action potential duration is less than 45 ms at cycle lengths of 2000 ms, in line with the results of Bauer et al. in the present study on refractoriness in the intact heart (1). In these hearts, a tight electrical coupling between cells is present, which may mask the intrinsic electrophysiological differences as has been demonstrated in several computer model studies (25–28). Taggart et al. (15) have measured repolarization in the left ventricle in human hearts during normoxia and ischemia. Even in compromised myocardium of patients with underlying coronary artery disease, transmural repolarization differences within the ventricular wall are absent even at cycle lengths as long as 1500 ms (15). This suggests that electrical uncoupling must be so profound to reveal the intrinsic dispersion in refractory periods or action potential duration that this would lead to arrhythmias anyway, because conduction velocity would fall to such low values that reentry would ensue. The question may arise why tight, but normal electrical coupling between cells prevents intrinsic differences in refractoriness to become evident in transmural direction but not in apico-basal direction. One may speculate that this is due to the fact that midmyocardial cells are sandwiched between epi and endocardial cells over a short distance, while the distance in the apico-basal direction is too large to annihilate the intrinsic differences in refractoriness present at apical and basal areas. METHODOLOGICAL CONSIDERATIONS AND LIMITATIONS The strength of the present study (1) is that the investigators performed direct measurements of the refractory period. Because this is a time-consuming method, most investigators apply indirect techniques, such as the recording of action potentials, monophasic action potentials or activation recovery intervals. The question arises how accurate these indirect measurements are in determining dispersion of the refractory period; they all neglect the role of postrepolarization refractoriness. In view of the present study it is a pity that the authors did not determine activation recovery intervals in addition to the ERPs, to provide information about the value of this indirect method. However, in previous work in the dog, the relationship between local activation recovery intervals and local action potential duration has been established (29). Recently, Chinushi et al. did the same for the correlation between local activation recovery intervals and local endocardial refractory periods in the human ventricle (30). Another inadequacy of the study is that no attempts were made to induce arrhythmias. Certainly, as the authors point out, the differences in refractoriness may have been too small to permit reentry, but differences in spread of activation could have been observed following premature stimulation.