Relationship Between Sinus Rhythm Activation and the Reentrant Ventricular Tachycardia Isthmus
Background In canine hearts with inducible reentry, the isthmus tends to form along an axis from the area of last to first activity during sinus rhythm. It was hypothesized that this phenomenon could be quantified to predict reentry and the isthmus location.
Methods and Results An in situ canine model of reentrant ventricular tachycardia occurring in the epicardial border zone was used in 54 experiments (25 canine hearts in which primarily long monomorphic runs of figure-8 reentry were inducible, 11 with short monomorphic or polymorphic runs, and 18 lacking inducible reentry). From the sinus rhythm activation map for each experiment, the linear regression coefficient and slope were calculated for the activation times along each of 8 rays extending from the area of last activation. The slope of the regression line for the ray with greatest regression coefficient (called the primary axis) was used to predict whether or not reentry would be inducible (correct prediction in 48 of 54 experiments). For all 36 experiments with reentry, isthmus location and shape were then estimated on the basis of site-to-site differences in sinus rhythm electrogram duration. For long and short runs of reentry, estimated isthmus location and shape partially overlapped the actual isthmus (mean overlap of 71.3% and 43.6%, respectively). On average for all reentry experiments, a linear ablation lesion positioned across the estimated isthmus would have spanned 78.2% of the actual isthmus width.
Conclusions Parameters of sinus rhythm activation provide key information for prediction of reentry inducibility and isthmus location and shape.
Received February 2, 2001; revision received March 30, 2001; accepted April 10, 2001.
Treatment of reentrant ventricular tachycardia by catheter ablation methods is hampered by the difficulty in localizing the circuit, particularly when the circuit structure is complex, the tachycardia is short-lived, or when reentry is not inducible during electrophysiological study.1 If measurements of sinus rhythm activation could be used to accurately localize reentry circuit features, it could potentially greatly improve the cure rate under these circumstances. A number of clinical and experimental studies to determine the usefulness of sinus rhythm parameters for targeting reentry circuits have been reported. The time of latest depolarization during sinus rhythm has been partially correlated to the location of the reentry isthmus; however, the relationship is inexact.2,3 At the border zone, both normal and abnormal (low-amplitude, fractionated, or wide-deflection) electrograms are present2–5; these abnormal electrograms, however, can be present both within and away from the reentry circuit location and are therefore not a specific predictor of its position in the border zone. Therefore, methods for detection and measurement of abnormal sinus rhythm activation characteristics are not currently sufficient for targeting reentry circuits for catheter ablation, although the presence of abnormality suggests the proximity of an arrhythmogenic substrate.
When a reentrant circuit can be induced in the infarct border zone by programmed electrical stimulation in a canine model,6 the area where the isthmus forms has at least 2 conspicuous substrate properties: (1) it is the thinnest surviving cell layer of any area of the border zone,6,7 and (2) there is disarray of gap-junctional intercellular connections that extends the full thickness from the infarct to the surface of the heart.8 Because these substrate properties persist regardless of rhythm type, they may affect electrical conduction at the isthmus area during sinus rhythm. The hypothesis that these phenomena could be quantified and used to predict reentry inducibility and isthmus location and shape, when it occurs, was tested in this study.
A myocardial infarct was created by ligation of the left anterior descending coronary artery (LAD) in experiments in 54 canine hearts, and attempts to induce reentry in dogs anesthetized with sodium pentobarbital were made 4 to 5 days later by premature electrical stimulation.9 Bipolar electrograms were recorded from 196 to 312 sites in the epicardial border zone of the anterior left ventricle for 25 experiments with predominantly long runs of monomorphic reentry (≥10 beats, mean 181.9 beats), 11 experiments with short monomorphic or polymorphic runs (<10 beats, mean 4.5 beats), and 18 experiments in which reentry was not inducible. Programmed stimulation from the LAD or lateral, base, or center region of the ventricle proceeded with 10 S1 stimuli followed by a single premature stimulus. The premature coupling intervals were successively shortened on subsequent stimulus trains until reentry was induced. For consistency between experiments, the multielectrode array was placed on the heart with the same edge always positioned along the LAD margin. For simplicity, the ventricular area where recording sites in the multielectrode array were located was considered to encompass the entire infarct border zone.
Activation maps9 were created from data obtained from the border zone during sinus rhythm, pacing, and reentry, when it occurred. For each experiment, the sinus rhythm map was constructed from an arbitrary cycle at the beginning of the experiment before programmed stimulation, and pace maps were constructed from cycles of the pace train that led to onset of reentry. Reentry maps were constructed from an early cycle of ventricular tachycardia after stabilization of the circuit (long-run experiments) or for all cycles (short-run experiments). Inspection of sinus rhythm activation maps in canine hearts in which reentry was inducible suggested that the isthmus entrance and exit, respectively, tended to form along an axis from the area of last to first activity during sinus rhythm. Moreover, the activating wavefront during sinus rhythm was observed to advance in parallel to this axis, with uniform conduction velocity, and in the opposite direction to activation within the isthmus during reentry. To quantify this phenomenon, the last 10-ms interval during which ≥5 contiguous sites activated was ascertained. The xy center of this region was computed as the mean value of the site locations in the x and y directions, referenced to an arbitrary fiduciary point on the computerized electrode grid. The linear regression of activation times was computed along 8 rays originating from the geometric center of this last-activating region (45° ray separation with orientation such that 2 of the rays were precisely vertical on the grid). The activation times at 4 sites along each ray (0.8-cm spacing between sites), plus the center site itself (5 sites in all), were used for each regression (rays not entirely on the grid were excluded from analysis). The ray with highest r2 value was called the primary axis. The regression line slope along the primary axis (called the activation gradient, ▿) and the r2 value (called the activation uniformity) were plotted for all experiments as a scatterplot. From the scatterplots, the best linear thresholds to classify experiments in which reentry could versus could not be induced were determined manually for the activation uniformity parameter alone and for the activation gradient-uniformity parameters in tandem. In polymorphic experiments, 2 distinct last-activation regions sometimes occurred during sinus rhythm, and in such cases, primary axis parameters from each were computed and used for classification.
The electrogram duration, defined as that contiguous series of electrogram deflections with no isoelectric segment of >5 ms duration, encompassing the time of local activation at the recording site during 1 cardiac cycle, was also measured for all electrogram recordings obtained during the same cycle as that used to construct the sinus rhythm activation map. The starting and ending points, respectively, were considered to be the beginning and ending times at which contiguous electrogram deflections rose above the isoelectric level to an amplitude >10% of the maximum electrogram peak. Electrogram duration for all sites was mapped by use of the same computerized electrode grid that was used for activation mapping. The location of the xy center of last activation during sinus rhythm and the location of reentry arcs of block determined from the reentry activation map were superimposed on the electrogram duration map computerized grid. Separate means of electrogram duration were computed for (1) sites residing along the primary axis that were used for its regression equation, (2) all sites residing within the area where the isthmus formed, and (3) all sites in the border zone (including those within the isthmus). The mean electrogram duration along the primary axis was plotted versus the activation gradient along the primary axis for all experiments, and the resulting scatterplots were also used to classify experiments in which reentry could versus could not be induced as described above for the activation gradient-uniformity scatterplot.
Locations at which the difference in sinus rhythm electrogram duration between any 2 adjacent sites (horizontal, vertical, or diagonal direction) was ≥15 ms were marked on the computerized electrogram duration map grid. Selected marks were then connected to form the border of a contiguous region by use of the following algorithm implemented on a digital PC-type computer: (1) the region must encompass the xy center of last activation and the initial 1 cm of the primary axis extending from it; (2) marks were connected so as to (a) minimize the maximum distance between connections, followed by (b) minimize the mean distance between connections; and (3) the inscribed region must have surface area ≥2.0 cm2 (the approximate minimum isthmus surface area that was observed in any experiment). The contiguous region so formed (the estimated isthmus) was compared with the actual location and shape of the reentry isthmus (delineated by connecting the computerized grid locations of block line end points that were superimposed from the reentry activation map), and the mean±SEM was computed from all experiments. The direction designated by the primary axis was considered to be an approximation of the direction of activation through the actual isthmus during reentry. A straight line, called the estimated line for ablation, was then drawn perpendicular to the primary axis from one edge of the estimated isthmus to the other on the computerized grid. The location of the estimated line for ablation was chosen so as to bisect the estimated isthmus into halves with equal areas. The percentage of the width of the actual reentry isthmus that the estimated line for ablation spanned was then computed and tabulated.
The significance of quantitative variables was determined by computerized statistical procedures (SigmaStat, Jandel Scientific) as follows. For comparison of mean activation gradients and mean electrogram durations, the difference in means (t test) was calculated. For comparison of isthmus locations, first the xy center of the estimated isthmus was taken as the point along the primary axis 1 cm from the origin. Then the actual reentry isthmus location on the computerized grid was approximated as the mean xy location of the 4 end points of the 2 arcs of block that bounded the isthmus. For polymorphic experiments with multiple isthmus locations, the xy center of the actual isthmus closest to the estimate was used for statistical comparison. The linear regression of estimated versus actual xy centers was then calculated for all reentry experiments. A linear regression was also computed for percent overlap of isthmuses versus heart rate.
Figure 1 shows activation maps for sinus rhythm (A), premature stimulation (S2) from the center of the border zone (B), and reentry (C) and the electrogram duration map (D) for a canine experiment in which only long runs of monomorphic reentry with a single morphology were inducible. Wavefront propagation direction through the isthmus during reentry (C) is oriented in parallel but opposite to propagation in the same region during sinus rhythm (A). During sinus rhythm (A), the 5 or more last sites to activate within a 10-ms interval have activation times between 60 and 69 ms. Nearest to the xy center of last activation (+) is a site that activates at time 91 ms. The locations used to determine the linear regression, which included this site, are denoted by their activation times, and the rays are numbered from 1 to 8. The accompanying table shows activation gradient and uniformity for each ray. The ray with greatest activation uniformity (the primary axis) is ray 5 (r2=0.97). The primary axis has the lowest activation gradient (▿=0.41 m/s) and is approximately parallel to the isthmus long axis. The block lines that form during premature stimulation and during reentry partially align between areas of large disparity in sinus rhythm electrogram duration (for simplicity, only reentry arcs of block are superimposed on the electrogram duration map). On the basis of the isthmus estimation algorithm, boundary points of the estimated isthmus are given by cross-hatched circles (D). This area partially overlaps the actual reentry isthmus whose boundaries are formed by the superimposed arcs of block. Examples of electrograms with differing electrogram duration are shown (inset, D); within most of the reentry isthmus region, electrogram duration was relatively short. Other long-run experiments had properties similar to those shown in Figure 1. Along the primary axis for all long-run experiments, mean activation uniformity and gradient were r2=0.97±0.01 and ▿=0.67±0.04 m/s, respectively. The mean sinus rhythm electrogram duration for sites residing within the isthmus area for all experiments was 24.2±0.4 ms (mean of 18.4±2.2 sites per isthmus), which was significantly lower (P<0.001) than for the border zone as a whole (34.1±0.7 ms). For most long-run experiments, including that of Figure 1, wavefront orientation during sinus rhythm was approximately parallel to the primary axis; therefore, the activation uniformity and gradient along the primary axis were proportional to conduction uniformity and velocity, respectively, along the same axis.
Figure 2 shows maps from an experiment in which only short runs of 3 to 8 beats of monomorphic reentry could be induced. The sinus rhythm activation map (A) shows the region where the isthmus forms (shaded). The primary axis (ray 1 with r2=0.96) approximately aligned with the isthmus long axis and extended from late- to early-depolarizing regions during sinus rhythm (upward vertical direction originating at the larger 50-ms isochrone). For the reentry episode of Figure 2, activation maps of reentry beats 1 to 2 were similar (second beat is shown in B). The reentry arcs of block partially align at edges between areas with large disparity in electrogram duration (D). Mostly short-duration sinus rhythm electrograms are present at the reentry isthmus location. On the third reentry cycle, the left arc suddenly shifted inward (dotted in C) to align with a different edge of large disparity in electrogram duration (D). On the next (termination) cycle, the activating wavefront blocked at the narrowest width of the reentry isthmus (not shown). The boundary points of the estimated isthmus are shown (D, crosshatched circles), and as in Figure 1, they partially overlap the actual isthmus location. Other monomorphic short-run experiments had properties similar to those shown in Figure 2. Along the primary axis for all short-run experiments, mean activation uniformity and gradient were r2=0.94±0.01 and ▿=0.79±0.12 m/s, respectively. Also for all short-run experiments, mean sinus rhythm electrogram duration at the isthmus location was 22.7±0.7 ms (mean of 9.9±2.0 sites per isthmus), which was significantly lower (P<0.05) than for the border zone as a whole (28.8±0.6 ms). The mean sinus rhythm electrogram duration throughout the border zone was significantly less in short-run than in long-run experiments (P<0.05).
For experiments lacking reentry, activation uniformity and gradient along the primary axis were r2=0.93±0.02 and ▿=1.22±0.08 m/s, respectively, and the mean electrogram duration throughout the border zone was 31.3±0.5 ms. In most experiments lacking reentry, the wavefront propagation direction during sinus rhythm did not align with the primary axis.
Figure 3A shows a scatterplot of activation uniformity and gradient along the primary axis during sinus rhythm for each experiment. Shown are the best threshold to classify experiments by activation gradient alone (dotted line) and for activation gradient uniformity in tandem (dashed line). In 24 of 25 experiments with long runs of reentry (solid circles) and 9 of 13 primary axes present in 11 short-run experiments (solid triangles), each threshold predicted that reentry could be induced. In 15 of 18 experiments lacking reentry (open circles), the activation gradient threshold alone predicted that reentry could not be induced; prediction improved to 17 of 18 when the activation gradient-uniformity threshold was used. The difference in means in the activation gradient parameter between each of the 3 groups was significant (P<0.001). Figure 3B shows a scatterplot of the mean electrogram duration versus activation gradient computed along the primary axis for each experiment. For comparison, the best activation gradient threshold is shown (dotted line; same as in A). The best threshold for electrogram duration/activation gradient in tandem (dashed line) can be used to correctly classify experiments into those with or without inducible reentry with the same accuracy as the activation gradient-uniformity threshold of Figure 3A. In Figure 3B, the points representing experiments with short runs of reentry (solid triangles) tend to form a curvilinear boundary separating points representing experiments with long runs of reentry versus no reentry.
Figure 4 shows the estimated reentry isthmus (regions with grid lines), the estimated wavefront direction through it (arrows), the estimated best line for ablation (dashed lines), and the actual location of reentry arcs of block (thick curved lines) for each experiment with long runs of reentry. A through Y are ordered from shortest to longest reentry cycle length. Figure 4W shows estimates for the Figure 1 experiment (boundary points denoted in 1D). In 2 experiments (O and Y), 2 reentry morphologies occurred, and arcs of block are shown for each. For all long-run experiments, the estimated reentry isthmus surface area mostly overlapped with actual isthmus location (mean overlap 71.3±3.2%), which was independent of heart rate (P=0.25). The xy centers of the estimated and actual reentry isthmuses were linearly correlated (x, r2=0.77; y, r2=0.60; P<0.001). Also, the estimated best line for ablation extended across most of the width of the actual reentry isthmus (mean 88.2%).
Figure 5 shows the estimated isthmus parameters for experiments with short runs of reentry (polymorphic in A through E and monomorphic in F through K; separately ordered on the basis of cycle length). In 3 polymorphic experiments (A through C), only a single late-activating region was detected in the sinus rhythm activation map, although there were isthmuses at multiple locations during reentry. In D and E, 2 estimated isthmuses are shown because there were 2 late-activating regions and therefore 2 primary axes during sinus rhythm. The estimates for the experiment of Figure 2 are shown in Figure 5F. For all short-run experiments, the estimated reentry isthmus surface area partially overlapped actual isthmus location (mean overlap 58.6±9.0% for monomorphic experiments, 25.7±6.3% for polymorphic, and 43.6±7.5% overall), which was independent of heart rate (P=0.45). The xy centers of the estimated and actual reentry isthmuses were linearly correlated (x, r2=0.78; y, r2=0.81; P<0.001). Also, the estimated best line for ablation extended across more than half the width of the actual reentry isthmus (mean 55.4%).
Electrical Properties at the Isthmus Location
The results of this study suggest that the area over which the primary axis extends during sinus rhythm has special distinguishing electrical characteristics for experiments in which reentry was inducible versus those lacking inducibility. In experiments with monomorphic reentry, the primary axis often overlapped the actual isthmus location, which is that area of the border zone with thinnest layer of surviving myocytes6,7 and having full-thickness gap-junctional disarray.8 Disarray of gap-junctional intercellular connections is also present at the isthmus formation area in reentrant ventricular tachycardia in humans.10 Uniformity of gap-junctional disarray throughout the region8 may have been responsible for the uniform activation gradient and therefore conduction velocity uniformity (because the activating wavefront tended to propagate in parallel to the primary axis). Sinus rhythm electrogram duration tended to be short within the isthmus formation area and longer just outside it, resulting in large differences in electrogram duration at isthmus edges that were used to draw boundary points. Electrical activation at depth is often asynchronous with surface activation11; therefore, reduction of electrical activity at depth, due to thinness of the layer, may have acted to shorten electrogram duration within the isthmus region.
In many experiments, conduction was impeded at border areas between regions with a large disparity in electrogram duration, both during premature stimulation and during reentry (Figures 1B through 1D and 2B through 2D). Such border areas, or discontinuities, between regions with differing electrical properties are marked by the presence of increased effective axial resistivity.12 Under normal conditions, depolarizing current is sufficiently coupled across such discontinuities to maintain propagation; however, current available for activation is reduced during premature stimulation and reentry, causing slow conduction or block.12 An arbitrary threshold of 15-ms difference in electrogram duration was used to mark areas where arcs of block would form. Disparity in sinus rhythm electrogram duration, however, was not always large along the entire length where block lines actually formed (Figures 1D and 2D). This may have resulted from diffraction effects in which wavefront curvature, as it traverses a small aperture between impassable obstacles, increases beyond a critical value so that propagation ceases.13 Boundary points between regions with large disparity in electrogram duration would act as nearly impassable obstacles because of their high effective axial resistivity, whereas block would also occur along connecting segments between them, although disparity in electrogram duration is smaller, with presumably lowered effective axial resistivity, as a result of the aperture effect.
Clinical Significance of the Study
There is abundant evidence that ventricular tachycardia in human patients is often caused by reentrant excitation.1,2,5,14 There are several similarities between clinical observations and the measurements of electrical activation during sinus rhythm in dog hearts with and without inducible reentry. Clinical studies suggest that the extent of abnormal activation and number of fractionated sinus rhythm electrograms tend to be greater in patients with sustained reentry than with unsustained reentry,14 in accord with the result of this study that mean sinus rhythm electrogram duration throughout the border zone was significantly greater for experiments with long runs of reentry versus short runs of reentry (P<0.05). Also, clinical findings suggest that disrupted and delayed endocardial activation14,15 and prolonged, fractionated electrograms during sinus rhythm14 can distinguish patients with reentrant ventricular tachycardia from those with normal ventricles and those of prior infarction without reentry. In the present study, although the isthmus area tended to have short sinus rhythm electrogram duration, areas adjacent to it within the reentry circuit area often had much longer electrogram duration (Figures 1 and 2⇑), consistent with clinical findings. In addition, the last-activating region of the border zone during sinus rhythm tended to reside in proximity to the reentry isthmus both in clinical studies2,3 and in the present study. These similarities suggest that it may be possible to apply the methodology described here to targeting of clinical ablation sites, for example, using a noncontact mapping system.3 Differences in infarct age, the intracellular matrix, border zone location, and action potential characteristics, however, may cause clinical data to vary significantly from canine heart data used in the present study6 and therefore necessitate modification of the quantitative techniques.
Imprecision in activation mapping due to limited spatial resolution and/or ambiguous time of local activation at any given recording site16 will affect both the activation gradient measurements and localization of arcs of block. Use of a different threshold for electrogram duration measurements could alter the precise locations of boundary points. Both multiple deflections (fractionation) and a single wide deflection were considered indicative of abnormal cell presence and wavefront impediment; anatomic and histological correlation to support this hypothesis, however, was not performed in this series of experiments, which is an important limitation of this study. The results described here for functional reentrant circuits in a canine model may not be fully applicable to reentrant ventricular tachycardia occurring in humans, in whom anatomic arcs of block can occur more frequently.5
This study was supported by an Established Investigator Award from the American Heart Association to Dr Ciaccio, a Whitaker Foundation Research Award, and NIH NHLBI grants HL-31393 and HL-30557. The authors thank Drs Andrew Wit and James Coromilas for their participation in some of the experiments.
Stevenson WG, Friedman PL, Kocovic DZ, et al. Radiofrequency catheter ablation of ventricular tachycardia after myocardial infarction. Circulation. 1998; 98: 308–314.
Gardner PI, Ursell PC, Fenoglio JJ Jr, et al. Electrophysiologic and anatomic basis for fractionated electrograms recorded from healed myocardial infarcts. Circulation. 1985; 72: 596–611.
Wit AL, Janse MJ. Basic mechanisms of arrhythmias.In: Wit AL, Janse MJ, eds. The Ventricular Arrhythmias of Ischemia and Infarction. Armonck, NY: Futura; 1993: 1–160.
Scherlag BJ, Brachman J, Kabell G, et al. Sustained ventricular tachycardia: common functional properties of different anatomical substrates.In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology and Arrhythmias. Orlando, Fla: Grune and Stratton; 1985: 379–387.
Peters NS, Coromilas J, Severs NJ, et al. Disturbed connexin43 gap junction distribution correlates with the location of reentrant circuits in the epicardial border zone of healing canine infarcts that cause ventricular tachycardia. Circulation. 1997; 95: 988–996.
Dillon SM, Allessie MA, Ursell PC, et al. Influences of anisotropic tissue structure on reentrant circuits in the epicardial border zone of subacute canine infarcts. Circ Res. 1988; 63: 182–206.
Miller JM, Tyson GS, Hargrove WC III, et al. Arrhythmias/pacing/surgical correction: effect of subendocardial resection on sinus rhythm endocardial electrogram abnormalities. Circulation. 1995; 91: 2385–2391.
Spach MS, Miller WT III, Dolber PC, et al. The functional role of structural complexities in the propagation of depolarization in the atrium of the dog. Circ Res. 1982; 50: 175–191.
Cabo C, Pertsov AM, Baxter WT, et al. Wave-front curvature as a cause of slow conduction and block in isolated cardiac muscle. Circ Res. 1994; 75: 1014–1028.
Pogwizd SM, Corr PB. Reentrant and nonreentrant mechanisms contribute to arrhythmogenesis during early myocardial ischemia: results using three-dimensional mapping. Circ Res. 1987; 61: 352–371.