J.A. Abildskov, MD

Long QT Syndrome

Possible mechanisms in the long QT syndrome and in torsade de pointes have been described in previous sections. Prolongation of the relative refractory period prolonged the QT interval and increased vulnerability to fibrillation in the automata model. During that period propagation was slow and allowed reentry during a wider window in which fibrillation followed single responses or after fewer responses to train stimuli. Prolongation of the absolute refractory period also prolonged the QT interval but decreased vulnerability to fibrillation in the model. During that period there were fewer responses to train stimuli per unit time and a narrower window in which single stimuli initiated fibrillation. ECG waveform effects of prolonged absolute and relative refractory periods differed. Prolongation of the absolute refractory period prolonged both onset and termination of T waves while prolongation of the relative refractory period prolonged only T wave termination.

Independent propagation of excitation in one path followed by reentry of the first was demonstrated in the model and proposed as the mechanism of torsade de pointes. The association of torsade de pointes and long QT interval was attributed to conduction defects and/or disparate recovery of excitability. Subsequent observations have provided additional details concerning the mechanism of torsade de pointes and its relation to the long QT syndrome. Layered distribution of recovery properties and slow propagation during the relative refractory period were shown to provide suitable conditions for torsade de pointes. The normal layered distribution of endocardial to epicardial recovery properties provides multiple potential paths with adjacent longer and shorter recovery times and the varied location and different recovery times of the paths are compatible with varied location and times of premature excitation to initiate torsade de pointes. Prolongation of the relative refractory period provides the conduction delay required for reentrant excitation.

In the model, the endocardial to epicardial gradient of recovery properties was simulated by 5 paths of 5 units thickness each. Initiation of torsade de pointes in layers with longer recovery (endocardium) required longer duration or more marked degree of slow conduction as well as later premature excitation. After normal excitation originating in the endocardium serial reentry in those layers remained confined there and other layers with briefer recovery were activated without reentry. In contrast, the torsade de pointes process of serial reentry initiated in “epicardial” layers spread to other layers.

In the model, the endocardial to epicardial gradient of recovery properties was simulated by 5 paths of 5 units thickness each. Initiation of torsade de pointes in layers with longer recovery (endocardium) required longer duration or more marked degree of slow conduction as well as later premature excitation. After normal excitation originating in the endocardium serial reentry in those layers remained confined there and other layers with briefer recovery were activated without reentry. In contrast, the torsade de pointes process of serial reentry initiated in “epicardial” layers spread to other layers.

In summary of findings concerning prolonged QT intervals: Prolongation of the absolute RP represented by action potential plateau resulted in decreased vulnerability, QT prolongation and early T wave onset. Prolongation of the relative RP resulted in increased vulnerability, and normal time of T wave onset. The T amplitude was decreased since it reflected action potential downstrokes rather than plateaus. Waveform abnormalities were approximately the same in early and late portions of the T wave since prolongation of the relative refractory period was present in both endocardial and epicardial regions. This was evidenced by symmetry of initial and terminal T waveform and by a biphed deflection in curves of the difference between control records and those from matrices with prolonged relative refractory period. The combination of prolonged relative refractory period and layered distribution of recovery properties provided suitable conditions for torsade de pointes.

Local Fibrillation

The wavelet hypothesis of fibrillation proposed a mechanism of widespread reentry but did not exclude local mechanisms. The automata model used to illustrate and investigate the hypothesis is also applicable to the demonstration and study of local reentry and findings elucidate the necessary conditions and manifestations.

Fibrillation due to local reentry has been investigated in a matrix of 625 units with local areas of 25 to 100 units having different properties than others. Fibrillation was initiated by train stimulation and local origin demonstrated by cessation after isolation of the responsible region by an inexcitable boundary or by ablation of the region.

Local fibrillation required different conditions in the responsible region (L) and the remainder of the matrix (M). These included a) high K value range b) lower K value mean c) conduction delay or d) inexcitable units including ones with geometry simulating the pulmonary veins. It was also necessary that matrix K values had sufficiently high mean or low range to prevent self-sustained fibrillation outside the responsible region. Particular conditions in M and L regions in various combinations were factors in local fibrillation. Representative examples were L regions of 100 units (K 3-4) with lower mean K values or higher range (K 3-6) than the M region (K 4-5). Train stimulation of the L region resulted in fibrillation that ceased in both regions when they were separated by an inexcitable boundary. Conduction delay (0-2 T.S.) in an L region of 100 units resulted in fibrillation that ceased in the M but not L region but ceased in both with an L region of 25 units. These findings were due to particular sets of K values and not general for local regions with lower mean, higher range or delayed conduction. In each case it was necessary that conditions in L regions allowed self-sustained fibrillation while those in M regions did not.

Inexcitable units arranged to simulate the pulmonary veins and adjacent left atrium allowed local fibrillation even with similar K values in L and M regions. The geometry provided paths between and around the simulated veins. Together with nonuniform recovery these patterns exhibited slow propagation or local block resulting in self sustained reentrant excitation.

ECG, Waveform and Nonuniform Local Lesions

As noted in a previous section, effects of local lesions on ECG waveform were due to boundaries between lesions and matrix when these were in different states. Endocardial lesions with prolonged recovery altered the terminal T wave with the same polarity as the QRS. Endocardial lesions with brief recovery altered initial to mid portions of T waves with polarity opposite to QRS. Epicardial lesions with prolonged recovery altered T waves during their entire duration with polarity opposite to that of the QRS and epicardial lesions with brief recovery altered the initial portion of the T wave with polarity the same as QRS.

Nonuniform recovery within localized lesions resulted in combinations of the above effects. For example, endocardial lesions with both reduced and prolonged recovery altered both initial and terminal portions of T waves with initial effects of the same and late effects of opposite polarity to the QRS.

Conduction Defects and Vulnerability to Single Stimuli

Effects of nonuniform conduction defects on vulnerability to train stimuli have been published (J. Cardiovasc. Electrophysiol. 1992; 3:48-55) and described earlier in this review. Uniform defects also increased vulnerability to train stimulation and both uniform and nonuniform defects increased vulnerability to single stimuli. As with nonuniform defects, effects depended on recovery properties of the matrix. Effects were most marked when fibrillation threshold was high since conduction defect effects occurred with each response. Conduction defects also increased vulnerability to single stimuli in terms of duration of the period in which fibrillation was initiated. Examples are shown in the table below.

Termination of Atrial Fibrillation

Conditions that terminated simulated atrial fibrillation were investigated. Fibrillation was initiated by train stimulation of 9 units for 200 time steps (TS) in matrices with varied properties. At 400 TS K values were altered and matrix behavior noted until 800 TS or until fibrillation ceased.

The increase of mean K values required to terminate fibrillation was related to vulnerability. Matrices with greater vulnerability due to high range a low mean K values, conduction defects or prolonged relative refractory period required greater increase of K values to terminate fibrillation.

In specific examples, fibrillation in a matrix with K values 2-7 (range 5) required an increased to 5-10 while one with K values 3.5-5.5 (range 2) required only an increase to 4.5-6.5. In the matrix with higher K value range fibrillation was initiated by single stimuli at 39 to 56 TS and FT to train stimulation was 53.3 (21.6). In the matrix with lower K range fibrillation to single stimuli occurred only at 38 TS and Ft was 104.1 (40.7).

Fibrillation in a matrix with K values 2-6 (mean 3) required an increase to 5-7 to terminate while a matrix of 3-5 (mean 3) required an increase to only 4-6. In the matrix with the lower mean fibrillation was initiated by stimuli at 23-31 TS and FT was 30.6 (7.8). In the matrix with higher mean, fibrillation was initiated by a single stimulus only at TS 33 and FT was 66.0 (21.4). Termination of fibrillation in matrices with conduction defects or prolongation of the relative refractory period also required grater increase of mean K. Data concerning termination of atrial fibrillation are summarized the in the following table:

Termination of fibrillation required a greater increase of mean K value when combined with conduction delay as occurs with quinidine administration. In an example, fibrillation in a matrix with K value of 3 to 5 ceased when values were increased to 4-6 but required an increase of 5-7 when combined with a conduction defect of 1.

Summary:Fibrillation that was difficult to start was easy to stop. When a particular set of conditions was required to initiate fibrillation only that set required change to terminate fibrillation. Fibrillation that was easily started was difficult to stop. Multiple combinations of conditions allowed fibrillation, and termination required change of all.

Conduction Defects and ECG Waveform

Conduction defects had effects on both QRS and ST-T deflections. Effects on the QRS were due to delayed activation of the defect and altered activation sequence outside the defect. Effects on the ST-T deflection were the result of delayed recovery in the defect and altered recovery sequence external to the defect. In matrices with the ventricular pattern of an endo-epicardial gradient of long to short recovery duration and endocardial origin of excitation, endocardial defects altered both early and late portions of the QRS and late portions of the T wave. Epicardial defects altered late portions of the QRS and early portions of the ST-T deflection. Polarity of the alterations was determined by differences in the state of defects and the adjacent matrix. During the QRS and early portion of the T wave the still excitable defect was adjacent to inexcitable matrix. During later portions of the T wave the excited and still refractory defect was adjacent to recovered matrix. These conditions resulted in deflections of opposite polarity superimposed on the QRST complex due to excitation and recovery of the entire matrix. The delayed activation of defects also altered the pattern of excitation and subsequent recovery outside the defects with additional effects on ECG waveform these effects were the result of delayed activation so waveform effects had the same polarity as those directly due to the conduction defect.

The duration and magnitude of waveform effects of conduction defects were determined by the duration of delay and recovery in defects in relation to the state of adjacent matrix and the degree of altered activation and recovery external to the defects. Plots of the difference between waveforms with and without conduction defects showed deflections of equal area and opposite polarity during the QRS and T waves. Findings provide an example of the “ventricular gradient” concept of Wilson that QRST area is independent of activation sequence.

Multiple Factors and Vulnerability

The individual effects of refractory period mean and range, conductions defects and relative refractory period on vulnerability to fibrillation have been described in previous sections. Combinations of the factors modified their effects as descried in this section.

The major mechanism in the effect of multiple factors on fibrillation threshold to train stimuli was the number of responses required to initiate fibrillation. Mean refractory period duration determined the number of responses to train stimulation in a particular period and refractory period range and relative refractory period duration determined nonuniformity of propagation o each response. Since effects of refractory period range or relative refractory period duration occurred per response, their effects were greater with larger number of responses. In one example, FT in a matrix with mean K value of 4.5 was 116 with a K range of 2 and 85 with a range of 3 (change 31). With a lower mean K value of 4, FT was 77 with a range of 2 and 62 with a range of 3 (change 15). In another example, FT in a matrix with mean K value of 4.5 was reduced by 37 by a conduction defect of 2 but only by 12 when mean K value was 4. Duration of the relative RP was also shown to have a greater effect on fibrillation threshold when that threshold was high. The medical significance of these findings is evidence that when fibrillation threshold is already low the addition of factors that further increase vulnerability has relatively little effect.

Increased mean refractory period reduced vulnerability to single stimuli in terms of later and shorter period for initiation of fibrillation. The addition of increased refractory period range, relative refractory period duration or conduction defects then had less effect than with lower refractory periods. An example was a matrix with mean K value of 3.5 and range of 3 (K2-5) in which fibrillation was initiated by stimuli of 28 to a44 time steps. With increased range to 21, (K 1.5-5.5) fibrillation was initiated by stimuli during 30 time steps at 19-49. With higher mean K value of 4 however, increased range to 4 (K 2-6) resulted in a fibrillation period of only 18 time steps from 34 5o 52. Increased relative refractory period duration and conduction defects also had less marked effects in combinations with higher mean K value.

Nonuniform Local Lesions

Waveform effects of nonuniform recovery in localized lesions were determined by plotting the difference of records with and without lesions. Endocardial lesions with refractory period range greater than that in the adjacent matrix increased T wave duration and had biphasic effects on T waveform. T wave onset was determined by the shortest recovery in the lesion and normal early activation of the endocardium. Termination of the T wave was determined by latest recovery in the lesion. Those conditions resulted in an initial negative and later positive deflection superimposed on the normal T wave.

Epicardial lesions with greater RP range than that in the adjacent matrix resulted in earlier T wave onset and an initial positive deflection superimposed on the T wave. A later negative deflection was the result of longer recovery in the lesion than the adjacent matrix.

Endocardial lesions with increased mean recovery duration resulted in a positive deflection in the difference curve. Onset of that deflection was determined by the shortest recovery and termination by the longest recovery in the lesion.

Epicardial lesions with increased mean recovery duration resulted in a negative deflection in the difference curve. Onset of that deflection was later than the T wave onset. Termination was determined by the latest recovery in the lesion plus activation time.

Endocardial-Epicardial Recovery and Waveform

Waveform effects of varied mean and range of recovery duration were largely opposite in endocardial and epicardial regions. Increased K value mean in the endocardial region increased the difference of endo-epi recovery times and increased the amplitude of mid and terminal portions of the T wave with prolongation of the wave. The difference curve showed a positive deflection extremely beyond the control T wave duration. Increased mean K value in the epicardial region decreased the difference of endo-epi recovery times and the amplitude of early portions of the T wave. The T wave onset was delayed and the difference curve showed a negative deflection.

Increased K value range in either endo- or epicardial regions had biphasic effects on T waveform. In the case of the endocardium effects occurred in the mid and late portions of the T wave. In mid T wave lower K values reduced the endo-epi difference and reduced T amplitude. In late T wave higher K values increased T amplitude and prolonged T wave duration. Increased range in the epicardial region resulted in earlier T wave onset, increased early T wave and decreased later T wave amplitude.

Increased mean in both endocardial and epicardial regions decreased early and increased late T wave amplitude. Epicardial increase reduced the difference of endo-epi recovery times during early portions of the T wave. Endocardial increase resulted in greater difference of endo-epi recovery times during later portions of the T wave. Total effects were similar to the combined effects of increased mean in epicardial and endocardial effects individually.

Increased range in both endocardial and epicardial regions increased early, decreased mid and increased late T wave amplitude. Earlier onset of epicardlal recovery resulted in earlier T wave onset and increased early T wave amplitude due to earlier epicardial recovery. Mid portions of T wave amplitude were decreased due to later recovery of endocardial portions of the matrix at that time. The increased range in the endocardium resulted in increased late T amplitude and later T wave termination.

Graphic T Wave Interpretation

Findings suggest that curves of the difference between normal T waves and those being evaluated may be useful for the recognition and definition of repolarization abnormalities. Clinical application would require definition of representative normal waveforms for particular leads but findings with the model provide guides for clinical studies. Findings employed a matrix with an endocardial to epicardial gradient of long to short recovery duration designated as “normal”. QRST waveform after stimulation at the endocardium was calculated and subtracted from waveforms calculated from a variety of other matrices. These included varied mean an/or range of recovery durations in endocardial and epicardial regions, relative refractory period durations and localized abnormalities of activation or recovery. The curves resulting from the subtraction represented the effects of these conditions on the matrix designated as normal.

Regional recovery.In curves of normal T waves subtracted from test waves, early recovery in the epicardial region resulted in positive deflections and early recovery in the endocardial region in negative deflections. Later than normal recovery in those regions had opposite polarity effects. The times of onset and termination of deflections were related to the duration and distribution of recovery in matrices. Increased duration in the endocardium resulted in late positive deflections ending after the control T wave. Increased range of recovery duration resulted in a late biphasic deflection in the difference curve. The initial portion was negative due to briefer recovery in the endo- than epicardial region. The terminal T wave portion ended later than the normal T due to the longer recovery portion of the increased range. Increased mean in the epicardium resulted in a negative deflection in the difference curve due to later than normal recovery in portions of that region. Increased range in the epicardial region resulted in an initial positive and later negative deflection due respectively to earlier and later recovery than normal.

Localized lesions consisting of K values differing from the neighboring matrix had effects similar to those of regional K values. For example, endocardial lesions with increased K values resulted in late positive deflections in the difference curve and endocardial lesions with increased range of K values resulted in late biphasic deflections of negative-positive polarity.

Prolongation of the relative refractory period resulted in biphasic difference curves Increased duration of recovery in the epicardial region reduced the early difference of recovery durations in epicardium and endocardium resulting in the initial negative deflection. Increased duration of recovery in the epicardial region resulted in the later positive deflection. Increased range of recovery durations in either endocardium or epicardium also resulted in biphasic difference corves. That in the epicardium resulted in a plus-minus deflection and that that the endocardium began with T wave onset and ended after control T wave termination.

The most likely clinical utility of subtraction curves is detection of “abnormalities” confined to mid portions of T waves. Altered recovery durations that did not include minimum and/or maxima affected T waveform without change of T wave or QT duration. Prolongation in the epicardium reduced the difference of endo-epi recovery durations and resulting T wave amplitude. Prolongation in the endocardium increased the endo-epi difference and increased T wave amplitude. Prolongation in both epi- and endocardium resulted in initial reduction of endo-epi difference and reduced early T amplitude and later increased difference and increased late T amplitude.

Inexcitable Units

Inexcitable units had effects on both ECG waveform and vulnerability to fibrillation. Mechanisms of waveform effects were loss of contribution of the areas to waveform and altered activation and recovery outside the areas. Magnitude of the effects depended on size and location of the areas and characteristics of the associated matrix. Curves of the difference between a designated “normal” record and those from matrices with localized inexcitable regions showed QRS and T wave features due to the region. Inexcitable units in the epicardial region reduced the dimensions of the epicardial portion of the epicardial-endocardial boundary and reduced early T wave amplitude. Inexcitable units in the endocardial region altered both initial and late T wave portions.

Scattered inexcitable units.As noted in a previous section the usual effect of randomly scattered units on vulnerability was increased fibrillation threshold due to blocking potential reentrant paths. The finding may be applicable to effects of fibrous tissue as well as those of inexcitable sites. Such tissue has been shown to participate in excitation but with fewer sites of propagation than muscle. Scattered inexcitable units in the model reduce the number of excitable neighbors and the finding of decreased vulnerability suggests excitable fibrous tissue may have a protective role in arrhythmogenesis.

Multiple Factors and Waveform

Changes of recovery in endocardial and epicardial regions had largely opposite effects on T waveform. Increased duration in the endocardium increased the difference from epicardium and increased T wave amplitude. Increased duration in the epicardium decreased the difference and reduced T wave amplitude. Increased range in the endocardium resulted in decreased followed by increased T wave amplitude and increased slope of the T wave downstroke. Increased range in the epicardium resulted in increased then decreased T amplitude and increased slope of the T wave upstroke. Waveform features due to epicardial conditions occurred in early portions of the T wave and those due to endocardial conditions in later T wave portions.

Various combinations of the T wave feature occurred with varied conditions in endocardial and epicardial regions. In an example with decreased mean and increased range in both endo- and epicardial regions: T wave onset was early due to brief recovery duration in the epicardium and T termination early due to brief recovery in endocardium. Increased slope of initial and terminal T wave portions reflected increased range in epicardial and endocardial regions. Completion of epicardial recovery was represented by the time of maximal T amplitude at the end of the upstroke. Onset of endocardial recovery was represented by the time of maximal T amplitude preceding the downstroke. Equal or overlapping times of epicardial recovery completion and endocardial recovery onset resulted in a T wave peak between the times of endocardial completion and epicardial onset of recovery.

« Prev (The Automata Model of Arrhythmias and the CVRTI - Part II)Next (Part IV) »