Mechanisms for Absence of Inverse Relationship Between Coupling Intervals of Premature Impulses Initiating Reentrant Ventricular Tachycardia and Intervals Between Premature and First Tachycardia Impulses
Background During initiation of tachycardias by programmed stimulation (PES), an inverse relationship between the coupling interval of the premature impulse (V1V2) and the interval between the premature impulse and the first impulse of tachycardia (V2T1) has been proposed to be a specific indicator of reentry. However, an inverse relationship has not always been observed during initiation of clinical reentrant ventricular tachycardias (VTs).
Methods and Results Reentrant VT was initiated by PES in twelve 4-day-old infarcted dog hearts. The relationship between V1V2 and V2T1 was always direct. Mapping of the epicardial border zone (EBZ) indicated that initiation of VT was secondary to functional orthodromic block of V2, propagation of V2 around the line of block, and antidromic propagation through the original location of the block. In 7 dogs, the line of orthodromic block and the pathway of orthodromic propagation were similar for different V1V2 coupling intervals. Orthodromic conduction time around the line to its distal side was longer at shorter V1V2 intervals, but a shorter antidromic delay in the area of unidirectional block for shorter V1V2 intervals, possibly reflecting small changes in the conduction pathway involving deeper layers of the EBZ, resulted in shorter V2T1 intervals. In the other 5 dogs, the orthodromic conduction pathway of V2 around the line of block changed markedly, with a shorter pathway for shorter V1V2 intervals resulting in shorter V2T1 intervals.
Conclusions An inverse relationship between V1V2 and V2T1 is not a specific indicator of functional reentry.
Originally, the possibility that a VT could be initiated and stopped by programmed stimulation was considered to be an indication that the underlying mechanism was reentry and not abnormal impulse generation (automaticity).1 2 However, the later demonstration that triggered activity caused by delayed afterdepolarizations could also be initiated and stopped by programmed stimulation3 made it clear that additional criteria were needed to distinguish between the two types of arrhythmias. One additional criterion that has been suggested is an inverse relationship between the coupling interval of the stimulated premature impulse and the coupling interval between that premature impulse and the first impulse of tachycardia, based on the premise that the earlier the premature impulses, the longer the conduction time through the reentrant circuit because of propagation in increasingly refractory myocardium.4 Because that same relation has been shown in in vitro studies to be direct or flat for initiation of triggered activity caused by delayed afterdepolarizations,3 5 an inverse relationship, when it occurs in the study of a clinical tachycardia, seemed to be a good marker to differentiate reentrant from triggered arrhythmias. However, despite this prediction, an inverse relationship has been reported during initiation in only ≈40% of clinical VTs that all other evidence indicates are caused by reentry.4 6 7 The question of what mechanisms might underlie the lack of an inverse relationship during initiation of reentry has not been answered and was the focus of our study in a laboratory model, the infarcted canine heart.8 9 10 A preliminary report of these data has been published in abstract form.11
Myocardial infarction was produced in mongrel dogs weighing 30 to 40 kg by a two-stage ligation of the LAD.9 Four days after coronary occlusion, the dogs were anesthetized (sodium pentobarbital 20 to 30 mg/kg IV) and ventilated, and the chest was opened by a median sternotomy. An electrode array was sutured on the epicardial surface of the left ventricular anterior and lateral surfaces, including the thin layer of epicardium that survives over the infarcted region (EBZ).9 The opened chest was then covered with a sheet of plastic to retain heat and moisture. All care and use of animals conformed to the guidelines of the American Physiological Society and the American Association for Accreditation of Laboratory Animal Care.
Electrode Array and Recording Instrumentation
In some of the experiments, the electrode array consisted of 292 bipolar electrodes10 made of silver disks with a diameter of 1 mm and a distance between poles in the bipolar pair of 2 mm. The distance between bipolar pairs was 5 to 10 mm. In other experiments, the electrode array consisted of 312 bipolar electrode pairs also made of 1-mm-diameter silver disks with a distance between poles of 3.2 mm. The distance between bipolar pairs was 4.8 to 6.4 mm.
For the studies with the 292-electrode array, signals from 192 electrodes could be recorded simultaneously. The electrode leads, chosen with a switch box, were led into preamplifiers with automatic gain controlled by a microprocessor. The electrograms were filtered (10 to 1000 Hz), multiplexed, sampled (2000 samples per second), and digitized by an 8-bit analog-to-digital converter.9 For experiments with the 320-electrode array, signals from all electrodes were simultaneously bandpass filtered (15 to 500 Hz), sampled at 1000 samples per second, and digitized by a 16-bit analog-to-digital converter. All digitized signals were pulse-code modulated and stored on a wide-band tape recording system (Ampex).
Two ECG leads and arterial blood pressure were continuously monitored on an Electronics for Medicine DR-12 oscillographic recorder. For induction of VT, the heart was stimulated through bipolar electrodes on the noninfarcted right ventricle adjacent to the LAD with a basic train of stimuli (S1S1) at the longest cycle length that ensured capture (250 to 350 ms). Single premature stimuli (S2) were applied at progressively decreasing coupling intervals until they could not initiate a premature response. Basic and premature stimuli were 2 ms in duration and 2 to 4 times diastolic threshold. Sustained VT was defined as tachycardia lasting longer than 30 seconds. All sustained tachycardias were monomorphic, whereas nonsustained tachycardias (<30 seconds in duration) were either monomorphic or polymorphic.
Our methods for determining local activation times and drawing isochronal maps have been described.9 10 We defined conduction block as occurring where there were more than three interpolated isochrones between adjacent electrodes (>40 ms) with the wave front moving in opposite directions on different sides of the region designated as block. Because the bipolar electrodes were 5 to 10 mm apart, 40 ms between adjacent electrodes provides a conduction velocity of 10 to 20 cm/s, assuming conduction directly from one electrode to the next, which is possible in nonuniformly anisotropic myocardium.12 However, it is unlikely that slow conduction is occurring when adjacent wave fronts are moving in opposite directions. The probable exit route of the reentrant impulse from the EBZ to the ventricles was determined by locating the electrodes on the margin of the array that were excited just before the onset of the QRS, on the assumption that the impulse does not enter the ventricles through intramural connections at the center of the EBZ.13 This assumption may not always be correct; the detailed anatomy of the EBZ was not studied in these experiments. The relationship between V1V2 (coupling interval of ventricular activation by the last basic impulse, V1, and by the premature impulse, V2) and V2T1 (coupling interval between premature ventricular activation, V2, and activation by the first tachycardia impulse, T1) was measured on the ECG (time between the peaks of the R waves of the QRSs) and on epicardial electrograms adjacent to the site of stimulation (time between activations). The data were fitted by a linear regression model. To test the hypothesis that the slope of the regression line was different from zero, a significance level of P<.05 was selected (SigmaStat; Jandel Scientific Software). All data are expressed as mean±SD.
Characteristics of Initiation of VT
VT (unsustained and sustained) was induced in 34 hearts by single premature stimuli. Of this total number, 12 were included in this study that met the following criteria: (1) tachycardias were induced over a wide range of coupling intervals (window of initiation; Table⇓); (2) complete reentrant circuits were mapped in the EBZ; and (3) the first beat of the tachycardia had the same QRS morphology for all premature coupling intervals (see “Discussion”). In 8 of the 12 experiments, sustained monomorphic VT was induced at one or more coupling intervals within the window of initiation (Table⇓). Nonsustained monomorphic tachycardia with the same QRS morphology occurred at the longer coupling intervals. In 5 of the 8 experiments in which sustained tachycardia was initiated, the morphology of the QRS during the sustained tachycardia was the same as the QRS morphology of the first beat of the tachycardia, and in 3 experiments it was different (Table⇓). Nevertheless, the QRS morphology of the first nonstimulated impulse was always the same for each coupling interval. In the other 4 experiments, sustained VT could not be induced for any coupling interval in the window (Table⇓).
We did not find an inverse relationship between V1V2 and V2T1 in any of the 12 experiments when measured on the ECG or at electrograms near the stimulus site. In fact, in 11 of the 12 experiments the relationship was direct: as the V1V2 interval decreased, the V2T1 interval also decreased. In 1 experiment, the relationship measured on the ECG was flat (ie, V2T1 did not change over a range of V1V2), but it was direct when measured on the electrograms at the stimulus site. Fig 1A⇓ shows the initiation of a sustained monomorphic VT at the longest and shortest premature coupling intervals (V1V2) of 190 ms (top) and 163 ms (bottom), at a basic drive cycle (V1V1) of 280 ms. In this experiment, all tachycardias had the same QRS morphology as the first nonstimulated beat. To the right is the direct relationship measured on the ECG. Fig 1B⇓ shows the initiation of VT at premature coupling intervals of 205 ms (top) and 190 ms (bottom), at a basic cycle length of 300 ms in a different experiment in which only a V1V2 coupling interval of 190 ms resulted in sustained monomorphic tachycardia. The ECG morphology of the sustained tachycardia was different from the morphology of the initiating impulses (Fig 1B⇓, bottom). The QRS morphologies of the first tachycardia impulse at each coupling interval for nonsustained tachycardia were similar to each other and to the QRS of the first two beats of the sustained tachycardia. To the right is the direct relationship on the ECG.
From the activation maps, we found two different general patterns of initiation of VT that were responsible for the direct relationships, although there were some similar features between the two.
Conduction of Premature Impulses With Different Coupling Intervals Over Similar Reentrant Pathways (Pattern 1 in Table⇑)
In these hearts, conduction of the premature impulse that initiated reentry blocked in the orthodromic direction. Propagation of the premature impulse proceeded around the ends of the line of block and activated its distal side. Antidromic reexcitation of the proximal side of the line of block then completed the reentrant circuit and caused the onset of the tachycardia. The premature impulses initiated at different coupling intervals had similar activation patterns on a macroscopic level. As the coupling of the premature impulse decreased, orthodromic conduction time around the line of block increased, but antidromic conduction time decreased. The summation of the orthodromic and antidromic conduction decreased with decreasing coupling intervals. In Fig 2A⇓, the direct relationship for the seven experiments that had this characteristic is plotted. The regression lines have a positive slope (P<.05).
Activation maps from one of the experiments in this group are shown in Fig 3⇓ (ECG in Fig 1A⇑). Fig 3A⇓ shows the activation pattern of the last stimulated impulse, S1, initiated at the LAD margin (pulse symbol) (arrows show the direction of propagation). There is a small region of conduction disturbance at the basal margin (activation times 58 to 64), but the remainder of the EBZ is activated without block. The fingerlike pro- jection of the 10-ms isochrone into the EBZ suggests some intramural connections in this region. The electrograms in Fig 4A⇓, recorded from the sites circled in Fig 3A⇓, show impulse propagation during the basic drive (V1). Fig 3B⇓ shows propagation of the premature impulse (long V1V2 coupling interval of 190 ms) that initiated sustained VT. At isochrones 20 to 30 ms, the stimulated wave front blocked orthodromically, shown by the long heavy black line. The line of block as defined (see “Methods”) was characterized by a long delay between adjacent electrodes with wave fronts moving in opposite directions on either side of the line and was ≈35 to 40 mm long. The line was located within 10 to 20 mm of the region with intramural connections (possibly the margin with normal myocardium), as indicated by the relatively large area activated nearly simultaneously within isochrones 10 to 20. It is doubtful that such rapid activation of large regions can occur only by propagation over a narrow epicardial layer. During sinus rhythm, there was epicardial breakthrough in this region, supporting this interpretation. Fig 4A⇓ shows electrograms recorded across the line of block (V2). Electrogram c, with an activation time of 16 ms, was proximal to the line of block, whereas electrode d (asterisk) was at or just distal to the line. The propagating wave then proceeded around the edges of the line of block (isochrones 30 to 70 ms in Fig 3B⇓) and activated the distal side (isochrones 90 to 140 ms) antidromically. In Fig 4A⇓, electrodes e and d were activated at 146 and 156 ms by the premature wave front that reached the distal side of the line of block.
Fig 3C⇑ shows activation resulting in the first beat of the tachycardia. Activation in this time window begins at the asterisk (the 6-ms activation time that is circled is equal to the 156-ms circled activation time in Fig 3B⇑). The proximal side of the line of block was reactivated antidromically (isochrones 10 to 70 ms). The antidromic activation time on the epicardial surface in the region in which the original line of block was located (isochrones 10 to 50 ms) was prolonged, requiring 46 ms to activate a region of ≈4.8 mm. Although the apparent conduction velocity is ≈10 cm/s, a true conduction velocity cannot be calculated because the exact pathway of propagation on a microscale is not known. In fact, conduction velocity may not be so slow, because an altered pathway of activation involving deeper layers of myocardial cells might have been used (see “Discussion”). Electrograms recorded in this region (electrodes b and c in Fig 4A⇑ within the dashed rectangle) show double potentials: two deflections separated by an isoelectric segment. The delay in antidromic propagation corresponds to the delay between the two potentials at each of these recording sites. The relative sharpness of the two deflections in electrogram c might not be expected if the delay was a result of only very slow conduction on the epicardial surface and favors the interpretation of the involvement of deeper layers in the antidromic conduction pathway (see “Discussion”). Farther away, activation proceeded more rapidly (isochrones 50 to 90 ms in Fig 3C⇑) to the LAD margin of the electrode array (electrode a [T1] in Fig 4A⇑). The exit route to the ventricles causing the first beat of tachycardia was toward the LAD margin. The V2T1 interval at electrode a (Fig 4A⇑) was 214 ms. That the exit route was close to the stimulating electrode explains why the morphology of the first beat of the tachycardia is similar to the stimulated beats (Fig 1A⇑). After reaching the LAD margin, the propagating wave split (Fig 3C⇑) and proceeded counterclockwise to the apical margin (isochrones 80 to 130 at the left) and clockwise to the basal margin (isochrones 80 to 130 at the right). Later, the two waves coalesced around isochrone 190 to 200 (Fig 3C⇑) near the left lateral margin of the electrode array, and activation proceeded toward the LAD margin (isochrones 190 to 250) to complete the circuit (circled activation 249 and Fig 4A⇑, electrode e, T2). Fig 3D⇑ shows the activation map of the second beat of the tachycardia. Activation begins in the 10-ms isochrone (asterisk) where the map in Fig 3C⇑ ends and moves around two lines of block (heavy black lines) (isochrones 30 to 130 to the left and right). Activation time of the region in which the original line of block was located (isochrones 10 to 30 ms) was faster, with an apparent conduction velocity of ≈1 m/s. Because this velocity seems too high for ventricular myocardium, the rapidity of activation may result from a contribution to activation from deeper layers of surviving myocardial cells in this region. Note that in electrogram b (Fig 4A⇑, T2), the second deflection of the double potential is much smaller in amplitude and the delay between the two deflections is decreased, whereas in electrogram c (T2), a very small undulation in the isoelectric potential may be the remnant of the double potential. The pattern of activation shown in Fig 3D⇑ is similar to the pattern of activation that occurred later on during the sustained tachycardia.
Fig 5⇓ shows isochronal maps during initiation of VT in this same experiment by a shorter coupled premature impulse with a V1V2 of 166 ms (electrode a in Fig 4B⇑ and bottom of Fig 1A⇑). The sequence of activation during the basic drive was similar to the map in Fig 3A⇑ and is not shown. The pattern of propagation of the premature impulse (Fig 5A⇓) was similar to the premature impulse with the longer coupling interval (Fig 3B⇑). A long line of block formed during propagation in the orthodromic direction (heavy black line); propagation proceeded around the edges of the line of block and excited antidromically the distal side of the line at ≈179 to 219 ms. However, there were some quantitative differences. The line of block seemed to shift slightly (≈5 mm) toward the LAD margin, between electrodes b and c (Fig 4B⇑, V2) instead of between electrodes c and d as for the longer coupling interval (Fig 4A⇑, V2), and was ≈10 mm longer (50 to 55 mm). The time for propagation of the premature impulse from the LAD margin to the distal side of the line of block increased to 160 ms (difference between the activation times at circled electrodes 27 and 187 in Fig 5A⇓) at the shorter coupling interval from the 139 ms (difference between the activation times at circled electrodes 17 and 156 in Fig 3B⇑) at the longer premature coupling interval. The increase was the result of slower activation and the longer line of block. Fig 5B⇓ shows the pattern of antidromic propagation that caused the first beat of the tachycardia. The pattern was similar to the premature impulse with the longer coupling interval: the proximal side of the original line of block was antidromically activated (isochrones 0 to 10 ms, asterisk). However, the region of antidromic breakthrough was wider. The interval between the time of activation of the distal side (electrode c, Fig 4B⇑) and the reactivation of the proximal side (electrode a) during antidromic propagation was 37 ms for the shorter coupled premature impulse, in contrast to 75 ms for the longer coupled impulse (Fig 4A⇑). The shorter time to reactivate the region antidromic to the block may have resulted from a change in the antidromic pathway to involve less of the deeper myocardium (see “Discussion”). Note that a double potential is not prominent at electrode c, although there is a very small second slow deflection at electrode b (dashed rectangle in Fig 4B⇑) for this shorter coupling interval that corresponds to the more rapid antidromic activation. In addition, the slight shift in the location of the line of block may influence the occurrence of the double potential. Activation continued then following a pattern of figure-eight reentry, with the exit route located at the LAD margin. Therefore, despite the longer time required to activate the distal side of the line of block for the shorter coupling interval (Fig 4C⇑ orthodromic, solid circles), the resulting V2T1 interval at the electrode near the stimulating site (electrode a) was 197 ms for the shorter coupling interval and 214 ms for the longer coupling, indicating a direct relationship between V1V2 and V2T1 (Fig 4C⇑, solid triangles) because of the more rapid antidromic activation (Fig 4C⇑, solid squares). Activation of the second beat of the tachycardia and the subsequent tachycardia was similar to the activation pattern shown in Fig 3D⇑.
The patterns of activation for the other experiments in this group were almost identical to the example shown in Figs 3 through 5⇑⇑⇑. Double potentials occurred in the region of antidromic activation at long V1V2 intervals and diminished or disappeared with the decreased antidromic activation time that occurred when V1V2 decreased.
Conduction of Premature Impulses With Different Coupling Intervals Over Different Macroscopic Reentrant Pathways (Pattern 2 in Table⇑)
The activation patterns associated with initiation of tachycardia for the experiments in this group were more diverse. The stimulated premature impulses that initiated reentry still blocked, but activation of the distal side of the line of block occurred through different pathways for short and long coupling intervals, as shown by the activation maps. The pathway was shorter and took less time to activate the distal side of the line of block for the shorter coupling intervals than for the longer coupling intervals, accounting for the shorter V2T1 intervals at shorter V1V1 intervals. Fig 2B⇑ plots that relationship for all experiments. The regression lines have a positive slope (P<.05).
Fig 6⇓ shows the pattern of activation for one experiment in this group during initiation of a monomorphic nonsustained VT lasting 6 impulses by a premature stimulus delivered near the LAD margin (pulse symbol in A and B), with the longest V1V2 coupling interval (172 ms). Activation of the EBZ by the last impulse of the basic train (arrows) (cycle length 300 ms) is shown in Fig 6A⇓. Several of the electrograms recorded during the basic drive (recording sites are circled and labeled a through f on the map), when the reentrant circuit was located after premature stimulation, are shown in Fig 7A⇓ (V1). Fig 6B⇓ shows the premature impulse. Activation proceeded from the stimulus site (pulse symbol) for about 20 to 50 ms before blocking (horizontal heavy black line). Additional lines of block also formed toward the apex (heavy black lines). Electrograms recorded proximal to the line of block (a, b, f) and distal to the line of block (c, d, e) are shown in Fig 7A⇓ (V2), with their activation times indicated. Propagation occurred around both ends of the line of block, with the distal side activated after 147 to 178 ms. The electrograms recorded from the circled electrodes on the distal side of the line of block on the map (Fig 6B⇓) are shown in Fig 7A⇓ (e, d, c, with activation times of 147, 152, and 178 ms). The map in Fig 6C⇓ overlaps the one in Fig 6B⇓; it has a time window that starts (asterisk) at the 140-ms isochrone of Fig 6B⇓. It shows antidromic activation (isochrones 10 to 30). Unlike in Figs 3 through 5⇑⇑⇑, there was no long delay in activation by the antidromic impulse in the region in which the line of orthodromic block was located (isochrones 10 to 30 and curved black arrow) (electrogram recordings d, c, b, and a during T1 in Fig 7A⇓). The region in which the antidromic wave front crossed the right side of the line of block was also close to the exit route to the ventricles (at the basal margin) that caused the first impulse of the tachycardia. Conduction time from the stimulus site near electrode a (circled activation time 4 ms in Fig 6B⇓) to the right distal side of the line of block (electrode d; circled activation time 152 ms in Fig 6B⇓) was 148 ms. Conduction time from this region back to electrode a was 9 ms, accounting for the V2T1 coupling interval of 157 ms near the site of stimulation (Fig 7A⇓). A reentrant circuit then formed around the right end of the line (curved black arrows; isochrones 10 to 150 in Fig 6C⇓). Fig 6D⇓ shows the reentrant activation around the line of block to the right during the second tachycardia impulse (curved black arrows, isochrones 10 to 150). Block occurred in this circuit, terminating the tachycardia after 6 impulses (not shown).
Fig 8⇓ illustrates the activation patterns during initiation of a nonsustained VT lasting 7 impulses, with a premature impulse having the shortest V1V2 coupling interval of 148 ms in this same experiment. Propagation of the basic drive was the same as before (Fig 6A⇑). There is a long delay in epicardial activation by the premature impulse after isochrones 30 to 40 (Fig 8A⇓) in the same region as where the longest coupled premature impulse blocked (see Fig 6B⇑). However, there is evidence from the activation times that block of this early premature impulse did not occur, despite the large differences in activation times at adjacent electrodes where the heavy dashed line is located, because the wave front is moving in the same direction on both sides of this line. The region distal to the center of the line is activated between 111 and 126 ms in an orthodromic direction (open arrows and shaded area; see electrodes c, d, e, and f during V2 in Fig 7B⇑). Because electrograms recorded here did not show fractionation (see Fig 7B⇑, electrodes b, c, and d), the slow activation may have involved a change in pathway through deeper surviving myocardial cell layers rather than slow conduction on the epicardial surface. From this shaded area there is orthodromic activation (isochrones 120 to 140), so that wave fronts on opposite sides of the region of delay are moving in the same direction. There is collision of this orthodromic wave front (open arrows) with the wave front propagating around the right side of the line of block (curved black arrows). Therefore, the true line of block was shorter (solid heavy black line). The map in Fig 8B⇓ overlaps the one in Fig 8A⇓; time 117 ms in Fig 8A⇓ corresponds to time 10 ms in Fig 8B⇓. The small region activated between 0 and 10 ms at the asterisk indicates the premature wave front moving in the orthodromic direction (straight black arrow); the region at the end of the right line of block activated between 0 and 10 ms (asterisk) indicates the premature wave front coming around the right end of the line of block (curved arrows). The merging of the two wave fronts occurs in the shaded region between the 30-ms isochrones. Because this large (shaded) region is activated within 6 ms, there may also be some contribution to epicardial activation from deeper cell layers. From this area, reexcitation of the proximal side of the line of block occurred (curved open arrows). Activation continued to the right around the remnants of the line of block in a reentrant pattern (isochrones 40 to 120 and curved black arrows). Conduction time from the stimulus site near electrode a (circled activation time 4 ms in Fig 8A⇓) to the right distal side of the line of block where the impulse crossed the line in the antidromic direction (electrodes c and d; circled activation times 117 to 118 ms in Fig 8A⇓) was 113 to 114 ms (compare with 148 ms for the longer coupled premature impulse in Fig 6⇑). Conduction time from this region back to electrode a was 10 ms (compare with 9 ms for the longer coupled premature impulse). Therefore, the shorter orthodromic conduction time to the distal side of the line of block accounted for the shorter V2T1 coupling interval of 124 ms near the site of stimulation (Fig 7B⇑). The corresponding value for the longer coupled premature impulse was 157 ms (Fig 7A⇑), which indicates a direct relationship between V1V2 and V2T1.
The variability of the pattern of propagation of premature impulses in this group is illustrated in another experiment in Fig 9⇓ in which nonsustained VT was induced. Activation during the basic drive is not shown. Propagation of a long coupled (170-ms) premature impulse (Fig 9A⇓) created a long “horseshoe”-shaped line (heavy black lines) that was not present during basic drive. Propagation moved around the extension of this line of block at the right to activate its distal side (isochrones 40 to 130; circled activation times 132 and 138), although there also may have been some orthodromic activation, as indicated by activation at one site distal to the line at 110 ms (open arrows), before arrival of the wave front moving around the right end of the line. The distal side of the rest of this complex line of block was activated by wave fronts moving around both its right and left ends (curved arrows, isochrones 100 to 240). Fig 9B⇓ illustrates the initiation of the first nonstimulated beat. The time window of the map overlaps Fig 9A⇓; the 10-ms isochrone in Fig 9B⇓ is equal to the 110-ms isochrone in Fig 9A⇓. The wave front at the right moves antidromically to reexcite the proximal side of the line of block after 80 ms at the asterisk (activation time 82). Despite reexcitation of the proximal side at the basal margin, the exit route to the rest of the ventricles was closer to the LAD margin of the electrode array (near circled activation times 110). The onset of the QRS in the ECG (not shown) occurred at time 110 ms. Excitation continued to rotate around the line of block at the right (isochrones 90 to 150, curved arrows). The orthodromic conduction time from time 10 ms (near the stimulus site) to time 138 ms (distal to the right edge of the line of block) in Fig 9A⇓ was 128 ms. The antidromic conduction time from time 38 in Fig 9B⇓ (which is the same as time 138 in Fig 9A⇓) to time 110 (near the stimulus site) was 72 ms. Therefore, at this premature coupling interval (170 ms), V2T1 was 200 ms. Fig 9C⇓ shows the pattern of propagation of a premature impulse with a short coupling interval of 140 ms. The location, shape, and size of the line of block were similar to the longer coupled premature impulse (Fig 9A⇓). Activation occurred around the rightmost branch of the line of block (isochrones 40 to 110) to activate its distal side between 149 and 153 ms (circled). However, in contrast to the long coupled premature impulse, the distal side of the line of block closer to the LAD margin was activated slightly earlier (asterisk, 141 ms circled) by a wave front that moved to the left around the line of block. Isochrone 10 in Fig 9D⇓ is the same as isochrone 120 in 9C and shows the initiation of the nonstimulated response. Reactivation of the proximal side of the line of block started close to the LAD margin of the electrode array (asterisk and circled activation times 21 and 18). The exit route was at the LAD margin of the electrode array (the onset of the QRS of the ECG occurred at time 35 ms), similar to the exit route for the longer coupled premature impulse (Fig 9B⇓). However, the time of antidromic activation of the proximal side to the exit point (at the LAD margin of the electrode array) was shorter for the shorter coupling (139 ms) than for the longer coupling interval (200 ms). As a consequence, the relationship between V1V2 and V2T1 was direct.
It has been proposed (and sometimes shown) that during the initiation of reentrant tachycardia by premature impulses, the relationship between the coupling interval of the premature impulse to the previous basic impulse and its coupling interval to the first impulse of tachycardia should be inverse because of progressively slower conduction of the premature impulse in the reentrant circuit.4 6 7 14 Therefore, this characteristic might distinguish reentrant arrhythmias from triggered activity caused by delayed afterdepolarizations in which the relationship is direct or flat.3 5 However, an inverse relationship has been reported for only ≈40% of clinical cases of VT4 6 7 that all other lines of evidence strongly suggest are reentrant. Mechanisms that might be responsible for the lack of an inverse relationship during initiation of reentry had not been previously determined in clinical or laboratory studies.
Reentrant VT can be induced by programmed electrical stimulation in a canine model of 4-day-old healing myocardial infarction.8 9 The reentrant circuits causing tachycardia often occur in a narrow epicardial rim of surviving muscle, the EBZ, and propagation in the circuits can be mapped.9 15 16 We investigated whether an inverse relationship occurred during initiation of VT in this model and did not find one. In fact, the relationship was always direct; as the coupling interval of the premature impulse to the basic impulse decreased, so did its coupling to the first impulse of the tachycardia (Fig 2⇑). By necessity, these measurements could only be made in a subset of the total number of hearts with tachycardia, those in which tachycardia could be induced by single premature impulses over a wide range of coupling intervals and complete reentrant circuits could be mapped. In addition, we chose only those examples in which the QRS complex of the first tachycardia beat was the same for all coupling intervals, as has usually been done when inverse relationships are sought during the initiation of clinical VT. This characteristic suggests that the initiating series of premature impulses might have traveled through the same reentrant circuit, although this supposition turned out to be incorrect (see below).
On a macroscopic level that could be resolved by the spatial resolution of the electrode array, there were two different patterns of activation that resulted in the direct relationship. In the first pattern (Figs 3 through 5⇑⇑⇑), premature impulses with different coupling intervals to the basic impulse blocked in the orthodromic direction in a similar region of the EBZ. The premature wave fronts conducted around the ends of the line of block to activate its distal side also similarly for different coupling intervals. The premature wave fronts then propagated antidromically past the region in which orthodromic block had occurred to reexcite myocardium on the proximal side before exiting the circuit to cause the first impulse of tachycardia. This pattern of initiation has been described in previous publications.9 15 16 17 Orthodromic conduction block has been attributed to prolonged refractory periods of the EBZ17 but also may involve its anisotropic properties.9 Orthodromic conduction time to the distal side of the line of block prolonged with increasing prematurity because of slowing of activation, as if the premature impulse were conducting in progressively more refractory myocardium, and because of lengthening of the line of block. What was not expected was the characteristics of antidromic conduction of the premature impulse to complete its transit through the reentrant circuit. At relatively long premature coupling intervals, antidromic activation took a long time, whereas at shorter coupling intervals, this time decreased. There are several possible interpretations for this observation. Because a large number of isochrones needed to be interpolated in this region of antidromic activation at long coupling intervals (Fig 3C⇑), antidromic conduction block may have occurred in the most superficial epicardial cell layers where the recording electrodes were located, because of insufficient recovery time from the orthodromic activation. However, because antidromic activation always continued away from this region in the same direction in which the impulse was propagating when it entered it, to complete the reentrant circuit, it seems likely that the impulse was able to bypass this region of block in the epicardial layers through alternative pathways involving deeper layers or tracts of surviving myocardial cells. This possibility is favored by the location of the line of block toward the margin of the EBZ, where its thickness increases substantially and where there might be intramural connections.18 19 Intramural reentry has previously been shown to occur at the margins of the EBZ.20 The electrograms recorded just proximal to this region of antidromic activation showed a double deflection separated by an isoelectric interval proportional to the activation delay (Fig 4⇑). The first deflection in the double potential most likely resulted from activation of the myocardium on the distal side of the line of block, because the time of its occurrence followed closely the activation of other electrodes in the conduction pathway of the premature impulse. The second deflection in the double potential occurred nearly simultaneously with activation by the antidromic impulse at electrodes on the proximal side of the line of block. The isoelectric interval between deflections occurred during the delay between activation on the distal side and the proximal side, during which the premature wave front may have been propagating through deeper myocardium out of the recording field of the surface electrodes. Early coupled premature impulses, because of increased orthodromic delay before they reached the distal side of the line of block, in fact arrived at the distal side slightly later than the long coupled premature impulses, allowing the myocardium at the line of block more time to recover. In addition, a shorter action potential duration resulting from orthodromic propagation up to the line of block for shorter coupled premature impulses21 may facilitate more rapid recovery of excitability. Because of the increased time for recovery at the line of block, antidromic activation through this region was much more rapid and might involve more superficial epicardial pathways. The more rapid antidromic activation of the region in which the line of block was located was accompanied by the diminution or disappearance of the double potentials (Fig 4B⇑). An alternative explanation for the slow antidromic activation of long coupled premature impulses is that there was very slow conduction through the region of block that did not involve alternative pathways, because the myocardium in this region had only barely recovered excitability by the time the premature impulse reached the distal side. If the refractory period in this region is longer than elsewhere in the border zone,17 such delayed recovery would be expected. However, one would also expect to see low-amplitude, fractionated electrograms rather than double potentials with an isoelectric segment, if such slow conduction occurred,22 unless early premature excitation of the distal side of the line resulted in very slow antidromic propagation of nonregenerative graded response,23 in which case no activity might be detectable between the two deflections.
A second type of activation pattern associated with a direct relationship during initiation of reentry was a macroscopic change in the activation pattern of the premature impulse detectable with the resolution of the mapping electrode array, even though the exit route and the QRS morphology remained unaltered. In some of these experiments, there was a decrease in the time for orthodromic activation of the distal side of the line of block with decreasing coupling intervals of the premature impulse (Figs 6 through 8⇑⇑⇑). The decrease in time was a result of a decrease in the length of the pathway traveled by the premature impulse to the distal side of the line of block. Premature impulses with shorter coupling intervals propagated orthodromically through a small segment of the large region over which longer coupled premature impulses blocked (Fig 8A⇑). Even though the activation time of this region was long, the distal side of the line of block was still activated earlier than by the longer coupled premature wave fronts that reached the distal side only by propagating around the ends of the line of block. Here again, it is likely that deeper layers of surviving myocardial cells may have provided alternative pathways of conduction for the shorter coupled premature impulses to the distal side of a line of superficial epicardial block and that orthodromic conduction velocity may have been faster than it appeared on the epicardial surface. Alternatively, more slowly propagating early premature impulses might provide additional time for recovery of critical sites within the zone of block, allowing orthodromic propagation through the line of “block” (compare Figs 6B⇑ and 8A⇑). However, fractionated electrograms, which would be expected to occur if this were the case (see above), were not observed (Fig 7⇑). As a result of the altered orthodromic propagation for the shorter coupling intervals, the size of the reentrant circuit appears to be smaller than for the longer coupling intervals (Fig 8B⇑). In other experiments, such as the one shown in Fig 9⇑, orthodromic conduction of early and late premature impulses occurred around the ends of the line of block, but antidromic propagation to the same exit site occurred at different regions of the block line. A shorter distance between the site of antidromic activation and the exit site for earlier premature impulses accounted for the direct relationship of V1V2 to V2T1 (Fig 9⇑).
The absence of an inverse relationship in this experimental model does not mean that the concept of the association of an inverse relationship with the initiation of reentry is incorrect. This animal model may not be an accurate representation of clinical events, and the results may apply only to anisotropic reentry. However, the results do show some possible and previously unrecognized causes for the absence of an inverse relationship. An inverse relationship would most likely exist if the reentrant circuit was a fixed anatomic pathway and the region in which orthodromic block of premature impulses occurs is fully recovered by the time the distal side of the line of block is activated over the full range of premature impulses that initiate reentry. In this situation, increased conduction slowing in any part of the reentrant circuit that is expected to accompany increased prematurity would be manifested as an increase in transit time of the premature impulse around the entire circuit and an increased interval between the initiation of the premature impulse and the first impulse of the tachycardia. There would be no cause for decreased antidromic conduction delay with increased prematurity, because no alternative conduction pathways would be available. Therefore, the occurrence of an inverse relationship in some cases of clinical VT and its absence in others might indicate a different mechanism for reentry, with anatomic reentry favoring an inverse relationship and functional or anisotropic reentry not favoring an inverse relationship. In addition, the oc- currence of a direct relationship in clinical studies on VT is not sufficient, by itself, to discount a reentrant mechanism.
Selected Abbreviations and Acronyms
|EBZ||=||epicardial border zone|
|LAD||=||left anterior descending coronary artery|
This study was supported by grant HL-31393 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md.
- Received April 7, 1997.
- Revision received June 5, 1997.
- Accepted June 19, 1997.
- Copyright © 1997 by American Heart Association
Wellens HJJ. Electrical Stimulation of the Heart in the Study and Treatment of Tachycardias. Amsterdam, The Netherlands: University of Amsterdam; 1971. Doctoral thesis. Leiden, The Netherlands: HE Stenfert Kroese NV; 1971.
Wellens HJJ. Value and limitations of programmed electrical stimulation of the heart in the study and treatment of tachycardias. Circulation. 1978;57:845-853.
Wit AL, Rosen MR. Afterdepolarizations and triggered activity: distinction from automaticity as an arrhythmogenic mechanism. In: Fozzard HA, Haber E, Jennings RB, Katz AM, Morgan HE, eds. The Heart and Cardiovascular System: Scientific Foundations. New York, NY: Raven Press; 1992:2113-2163.
Josephson ME, Marchlinski FE, Buxton AE, Waxman HL, Doherty JU, Kienzle MG, Falcone R. Electrophysiologic basis for sustained ventricular tachycardia: role of reentry. In: Josephson ME, Wellens HJJ, eds. Tachycardias: Mechanism, Diagnosis, Treatment. Philadelphia, Pa: Lea & Febiger; 1984:305-323.
Johnson NJ, Rosen MR. The distinction between triggered activity and other cardiac arrhythmias. In: Brugada P, Wellens HJJ, eds. Cardiac Arrhythmias: Where To Go From Here. Mount Kisco, NY: Futura Publishing Co; 1987:129-145.
Josephson ME, Horowitz LN, Farshidi A, Kastor JA. Recurrent sustained ventricular tachycardia, I: mechanisms. Circulation. 1978;57:431-440.
Wellens HJJ, Duren DR, Lie KI. Observations on mechanisms of ventricular tachycardia in man. Circulation. 1976;54:237-244.
El-Sherif N, Hope RR, Scherlag BJ, Lazzara R. Reentrant ventricular arrhythmias in the late myocardial infarction period, II: patterns of initiation and termination of reentry. Circulation. 1977;55:702-719.
Dillon SM, Allessie MA, Ursell PC, Wit AL. Influence of anisotropic tissue structure on reentrant circuits in the subepicardial border zone of subacute canine infarcts. Circ Res. 1988;63:182-206.
Coromilas J, Saltman AE, Waldecker B, Dillon SM, Wit AL. Electrophysiological effects of flecainidine on anisotropic conduction and reentry in infarcted canine hearts. Circulation. 1995;91:2245-2263.
Cabo C, Deruyter B, Coromilas J, Wit AL. Mechanisms for absence of inverse relationship between coupling intervals of premature impulses initiating ventricular tachycardia and intervals between premature and first tachycardia impulses. Pacing Clin Electrophysiol. 1996;19:728. Abstract.
Spach MS, Dolber PC. Relating extracellular potentials and their derivatives to anisotropic propagation at a microscopic level in human cardiac muscle: evidence for electrical uncoupling of side-to-side fiber connections with increasing age. Circ Res. 1986;58:356-371.
Bigger JT Jr, Goldreyer BN. The mechanism of supraventricular tachycardia. Circulation. 1970;42:673-688.
El-Sherif N, Smith RA, Evans K. Canine ventricular arrhythmias in the late myocardial infarction period, VIII: epicardial mapping of reentrant circuits. Circ Res. 1981;49:255-265.
Cardinal R, Savard P, Carson DL, Perry J-B, Pag P. Mapping of ventricular tachycardia induced by programmed stimulation in canine preparations of myocardial infarction. Circulation. 1984;70:136-148.
Gough WB, Mehra R, Restivo M, Zeiler RH, El-Sherif N. Reentrant ventricular arrhythmias in the late myocardial infarction period in the dog, XIII: correlation of activation and refractory maps. Circ Res. 1985;57:432-442.
Wit AL, Allessie MA, Bonke FIM, Lammers W, Smeets J, Fenoglio JJ Jr. Electrophysiologic mapping to determine the mechanism of experimental ventricular tachycardia initiated by premature impulses: experimental approach and initial results demonstrating reentrant excitation. Am J Cardiol. 1982;49:166-185.
Mehra R, Zeiler RH, Gough WB, El-Sherif N. Reentrant ventricular arrhythmias in the late myocardial infarction period, IX: electrophysiologic-anatomic correlation of reentrant circuits. Circulation. 1983;67:11-24.
Kramer JB, Saffitz JE, Witkowski FX, Corr PB. Intramural reentry as a mechanism of ventricular tachycardia during evolving canine myocardial infarction. Circ Res. 1985;56:736-754.
Elharrar V, Surawicz B. Cycle length effect on restitution of action potential duration in dog cardiac fibers. Am J Physiol. 1983;244:H782-H792.
Gardner PI, Ursell PC, Fenoglio JJ Jr, Wit AL. Electrophysiologic and anatomic basis for fractionated electrograms recorded from healed myocardial infarcts. Circulation. 1985;72:596-611.
Kao CY, Hoffman BF. Graded and decremental in heart muscle fibers. Am J Physiol. 1958;194:187-196.