Portal de Periódicos da CAPES

Formation of Functional Conduction Block During the Onset of Reentrant Ventricular Tachycardia

2016; Lippincott Williams & Wilkins; Volume: 9; Issue: 12 Linguagem: Inglês

10.1161/circep.116.004462

1941-3149

Edward J. Ciaccio, James Coromilas, Andrew L. Wit, Nicholas S. Peters, Hasan Garan,

Atrial Fibrillation Management and Outcomes

HomeCirculation: Arrhythmia and ElectrophysiologyVol. 9, No. 12Formation of Functional Conduction Block During the Onset of Reentrant Ventricular Tachycardia Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBFormation of Functional Conduction Block During the Onset of Reentrant Ventricular Tachycardia Edward J. Ciaccio, PhD, James Coromilas, MD, Andrew L. Wit, PhD, Nicholas S. Peters, MD, PhD and Hasan Garan, MD Edward J. CiaccioEdward J. Ciaccio From the Department of Medicine, Division of Cardiology (E.J.C., H.G.) and Department of Pharmacology (A.L.W.), Columbia University College of Physicians and Surgeons, New York, NY; Department of Medicine, Division of Cardiovascular Disease and Hypertension, Rutgers University New Brunswick, NJ (J.C.); and Department of Medicine, Cardiovascular Sciences, Imperial College London, United Kingdom (N.S.P.). , James CoromilasJames Coromilas From the Department of Medicine, Division of Cardiology (E.J.C., H.G.) and Department of Pharmacology (A.L.W.), Columbia University College of Physicians and Surgeons, New York, NY; Department of Medicine, Division of Cardiovascular Disease and Hypertension, Rutgers University New Brunswick, NJ (J.C.); and Department of Medicine, Cardiovascular Sciences, Imperial College London, United Kingdom (N.S.P.). , Andrew L. WitAndrew L. Wit From the Department of Medicine, Division of Cardiology (E.J.C., H.G.) and Department of Pharmacology (A.L.W.), Columbia University College of Physicians and Surgeons, New York, NY; Department of Medicine, Division of Cardiovascular Disease and Hypertension, Rutgers University New Brunswick, NJ (J.C.); and Department of Medicine, Cardiovascular Sciences, Imperial College London, United Kingdom (N.S.P.). , Nicholas S. PetersNicholas S. Peters From the Department of Medicine, Division of Cardiology (E.J.C., H.G.) and Department of Pharmacology (A.L.W.), Columbia University College of Physicians and Surgeons, New York, NY; Department of Medicine, Division of Cardiovascular Disease and Hypertension, Rutgers University New Brunswick, NJ (J.C.); and Department of Medicine, Cardiovascular Sciences, Imperial College London, United Kingdom (N.S.P.). and Hasan GaranHasan Garan From the Department of Medicine, Division of Cardiology (E.J.C., H.G.) and Department of Pharmacology (A.L.W.), Columbia University College of Physicians and Surgeons, New York, NY; Department of Medicine, Division of Cardiovascular Disease and Hypertension, Rutgers University New Brunswick, NJ (J.C.); and Department of Medicine, Cardiovascular Sciences, Imperial College London, United Kingdom (N.S.P.). Originally published22 Nov 2016https://doi.org/10.1161/CIRCEP.116.004462Circulation: Arrhythmia and Electrophysiology. 2016;9:e004462IntroductionReentrant ventricular tachycardia is an important clinical problem. It is caused by an electric circuit traveling in 1 or more loops that are constrained to the ventricular myocardium.1,2 The arrhythmia often originates after myocardial infarction, in the region near the infarcted area. This proximate region or infarct border zone (IBZ) contains surviving strands of myocardial fibers that continue to conduct electrical activity.1,2 Electrical conduction there is constrained by the infarcted region, which does not activate, by the geometry of the remaining viable myocardial tissue, and by the contour of the heart surface in proximity to the IBZ.Several mechanisms have been suggested to explain the formation of functional block leading to and maintaining reentry, all of which likely have a contribution. One possibility is that disparate gap junctional connecting properties suppress activation from one myocyte to the next. If, for example, there are less gap junctions between connecting cells than normal, it can result in an insufficient current density during activation and may lead to conduction block.3 Another mechanism that has been proposed is the presence of disparities in ion channel properties in the IBZ.4 Such regions would have differences in the functioning of sodium, potassium, and calcium current ion channel gates between the postinfarction myocardial cells, when compared with their normal properties. This would alter the depolarization and repolarization phases of the action potential, possibly preventing activation between cells or causing it to be slowed or weak. Similar to the proposed gap junction mechanism, if the ion channel properties are disparate across a transition zone, it may result in a lack of electrical connectivity. On one side of the transition zone, when activation occurs, ions are not transferred with sufficient density or rapidity to cells on the other side of the transition zone, preventing activation of the neighboring cells. Disparities in gap junction and ion channel properties may be important to the formation of functional conduction block leading to reentrant ventricular tachycardia circuits.A third possible mechanism has been put forth to explain conduction block leading to the onset and maintenance of reentrant ventricular tachycardia circuits.5–8 Source–sink mismatch can cause block of the electrical activation wavefront when it travels in certain directions with respect to the structural characteristics of the IBZ substrate. This mismatch would be expected to occur only when the wavefront orientation causes it to travel across a sudden transition from a smaller to a larger volume of viable myocardial substrate. In the next sections, the properties pertaining to this phenomenon are considered in detail, and the clinical ramifications will be described in the Discussion section, along with the contribution of molecular-level properties. By developing a biophysical model of reentry based on wavefront curvature and source–sink mismatch, the localization of the arrhythmogenic tissue for radiofrequency catheter ablation can potentially be improved.Functional Conduction Block at the Lateral Isthmus BoundariesThe presence of critically convex wavefront curvature at geometric transition points and its effects on the formation of functional electric conduction block at the onset of reentrant ventricular tachycardia is assessed in this section. The work was initially developed from electric activation mapping studies in postinfarction canine hearts and from analyses to determine IBZ geometry. By registrating activation maps with IBZ geometry, it was possible to determine the substrate characteristics leading to functional conduction block. Locations of block formation could then be predicted by formulating equations to describe alterations in the activation wavefront at geometric transitions.In order for reentry to occur, the activation wavefront must be impeded from propagating in certain directions. In canine as well as clinical postinfarction, most reentrant circuits have a double-loop pattern of activation and reside in the left ventricular myocardium,9,10 as illustrated in Figure 1. The electric impulse proceeds as a single wavefront through a constrained region known as the central common pathway or isthmus of the reentrant circuit. Here, lateral lines of functional conduction block during tachycardia, which are not present during sinus rhythm, bound the isthmus. The impulse then proceeds through the isthmus exit, whereupon it bifurcates into 2 distinct activation wavefronts, which circle around the bounding lines of conduction block through an area known as the outer circuit pathway. At the opposite end of the reentry isthmus, these 2 distinct wavefronts coalesce, at which point the single impulse reenters the constrained central common pathway. Since the area is constrained, it is ideally here, within the central common pathway, that the wavefront should be blocked using radiofrequency ablation energy to create a lesion. This would prevent recurrence of the arrhythmia if the lesion created successfully prevents conduction through the isthmus region.Download figureDownload PowerPointFigure 1. Diagram of a double-loop reentrant circuit with functional conduction block at the lateral isthmus boundaries.Another characteristic of this source of ventricular tachycardia in both canine experiments11,12 and clinical cases13 is that when multiple reentry morphologies arise, they often originate from a shared isthmus region (Figure 2). For each morphology, the orientation, shape, and length of the isthmus will differ, but they will each overlap the same general area. In the Figure 2 schematic representation, examples of 2 (Figure 2A) and 3 (Figure 2B) reentry morphologies are shown. Arrows denote propagation direction through the isthmus for each of the morphologies.Download figureDownload PowerPointFigure 2. Drawing that illustrates how 2 (A) or 3 (B) reentry morphologies can share a common location but have differing isthmus orientation. Shaded arrows show propagation direction for each morphology. The isthmus regions are noted by dark gray shading. Functional block at lateral boundaries is delineated as thick black lines. Functional block location may vary slightly between morphologies, because of differences in timing and orientation of the arriving activation wavefront and tachycardia cycle length.14The characteristics of reentrant ventricular tachycardia in the IBZ can be explained based on an impedance or source–sink mismatch at a transition region, resulting in functional block.5–7,15 In canine postinfarction, the reentry isthmus forms at the region where the IBZ is thinnest, that is, there are fewest surviving myocytes between infarct and heart surface.8–10 The thickness of this region, when reentry is inducible, has been measured to be on the order of 100 to 500 µm.8–10 When the reentrant circuit resides in the subepicardium or subendocardium, it is bounded by infarct at depth and by the actual heart surface. It, therefore, propagates approximately 2-dimensionally, along an xy plane, if the fact that the heart is actually a curved surface is neglected. The thickness of the IBZ can then be considered to be a third or z axis. In early experiments, it was shown that when the activation wavefront was constrained to traveling in the xy plane, at an area which then opened to a distal expansion, functional conduction block can occur (Figure 3).6 Two configurations are shown. In one diagram, the regions of conduction block, shown as thick black lines, are aligned edgewise, and in the other they are in parallel. In both cases, when the wavefront, whose direction of propagation is denoted by the arrow, exits a narrowed region to a distal expansion, block will occur if the narrowed region is on the order of 0.5 mm (500 µm) or less. This occurs because the available current from activation at the constrained region is insufficient to activate the greater volume of tissue in the distal direction. The wavefront itself changes in curvature at the transition (dotted lines denote the leading edge of the activation wavefront at several locations). When there is no change in geometry in the propagating direction, the activating wavefront is a straight line. However, when the wavefront passes through regions with differing geometry, it becomes curved. When it travels from a lesser to a greater volume of conducting tissue, it becomes convex in curvature, that is, the leading edge of the wavefront points outward at its center in the direction of travel, and it slows. This is shown in both schematics where there is a distal expansion in the propagation direction. If the wavefront becomes critically convex, because of a very large change from lesser to greater volume in the travel direction, conduction block will occur in the forward direction. Contrarily, at regions with distal constriction as in the schematic representation at right in Figure 3, the wavefront becomes concave and speeds up.Download figureDownload PowerPointFigure 3. Illustration of wavefront curvature (dotted lines) and functional block (double black lines) occurring at a distal expansion of the volume conductor, for 2 different configurations of lateral block (thick black lines).The occurrence of a critical wavefront convexity in the xy plane can be mathematically formulated as follows. The velocity of impulse conduction without curvature θo is dependent on the longitudinal resistance R of the conducting medium15:Download figure ((1)) The conduction velocity is θo when the wavefront propagates through a conducting medium lacking changes in volume along the path, that is, the leading edge of the wavefront remains a straight line. However, when the geometry of the conducting medium changes, the wavefront curves, with the conduction velocity given byDownload figure ((2)) where θc is the conduction velocity contribution due to wavefront curvature. In the IBZ, θc can be estimated as follows15,16:Download figure ((3)) where D, the diffusion coefficient, is the current flow due to the transmembrane potential gradient, with a value of 0.05 to 0.2 mm2/ms in ventricular myocardium.17 The term ρ is the degree of wavefront curvature and has units of mm−1. Thus,16Download figure ((4)) The degree of wavefront curvature ρ is related to the local radius of curvature r7,15:Download figure ((5)) HenceDownload figure ((6)) If there is no conduction across the lateral boundaries, so that no-flux conditions exist, the wavefront can be modeled as a circular arc5:Download figure ((7)) where β is the angle between the travel direction at the midline and the no-flux boundaries. The no-flux boundaries are infarcted tissue or the heart surface, present at either end of the wavefront. This configuration is illustrated in Figure 4. The chord length w and the travel direction of the wavefront are shown as dotted lines. The arc between the 2 thick lines represents the propagating wavefront. The thick lines are no-flux boundaries, either infarct or heart surface. On the basis of this configuration,Download figureDownload PowerPointFigure 4. The diagram shows how wavefront curvature can be represented as an arc of a circle, with a chord length w, and an angle β in the propagation direction with respect to no-flux boundaries. The no-flux boundaries at the wavefront ends, either composed of infarcted tissue or being the boundary at the heart surface, are delineated as thick black lines.Download figure ((8)) As the fraction on the right-hand side of the equation increases, the wavefront will slow, or block whenDownload figure ((9)) where the arrow denotes that the value of the left-hand term approaches that of the right-hand term. The fraction is maximized when β=90° (sin β=1), and w is minimized. If θo=0.4 mm/ms, a typical speed of propagation in the IBZ, D=0.1 mm2/ms, and β=90°, then w=0.5 mm (500 µm) so that θ=0. If β<90°, then w must be <500 µm for conduction block to occur.This configuration can be extended along the z axis (IBZ thickness axis). On the basis of geometric considerations and trigonometry, along the z axis8Download figure ((10)) where T is the thickness of the border zone, and c is the space step (distance between nodes used for calculation). Suppose that ΔTmax is the maximum change from thin-to-thick IBZ in the vector field, and further suppose thatDownload figure ((11)) Then from Equation 10,Download figure ((12)) Let the space step c be 1 mm; then this can more simply be written as follows:Download figure ((13)) Thus, from Equations 4 and 13,Download figure ((14)) Although IBZ geometry varies along the x, y, and z axes, it is approximated as uniform over small space steps on the order of c=1 mm (Equation 14). When the change in IBZ thickness ΔT is maximized from thin to thick, and the initial thickness T is minimized, then the quantity ΔT/T in Equation 14 is maximized, and the wavefront will be most likely to block. This would be expected to happen where the IBZ is thinnest, that is, at the location where the isthmus forms, because T is minimized there, and specifically at the lateral boundaries, where the change in ΔT in the thin-to-thick direction is maximized, as is actually observed.8–10 This phenomenon is illustrated as a thickness map in Figure 5A. The isthmus, which is the thinnest IBZ, is noted as a dark gray rectangle at center, shown with an average thickness of approximately 100 µm. Surrounding it laterally are areas of outer circuit pathway which are much thicker, greater than approximately 600 µm, thus rendering ΔT large. Block occurs as the wavefront travels through the isthmus at points along the lateral boundaries, as illustrated by short white arrows and double lines. However, propagation through the isthmus at the exit point is unimpeded, or there may be slower conduction but not block, because ΔT is of lesser magnitude per unit spatial distance (lighter gray area with gradient).Download figureDownload PowerPointFigure 5. The schematic representation illustrates how functional conduction block may occur all along the lateral isthmus boundaries (A), or there may be very slow and variable propagation across the lateral boundaries (dotted arrows, B), depending on the thickness transition there. A more uniform thickness gradient between the isthmus long axis at the centerline and the lateral edges increases wavelet speed in that direction (longer dotted arrows, B).Based on activation wavefront curvature, it is also possible to predict regions with more rapid conduction velocity in the IBZ during reentrant tachycardia.8 Suppose that the transition from thinnest border zone to thickest outer circuit pathway is a change Z, thusDownload figure ((15)) where dTi is the change in border zone thickness at space step i, and the total thickness change Z from thinnest to thickest border zone or vice versa is the sum of changes from space step 1 to n. Suppose the change Z transpires all in 1 space step j. This can be represented as follows:Download figure ((16a)) Download figure ((16b)) Very slow conduction or block will occur at j in the thin-to-thick direction due to the large thickness change Z there (large source–sink mismatch). In the thick-to-thin direction, a transient increase in conduction velocity θ will occur at space step j, however, θ=θo elsewhere along the path. Thus, it can be postulated that gradual changes in T along each space step according to Equation 15, rather than 1 sharp change at j as in Equation 16, will minimize travel time. To determine the exact relationship, from Equations 4 and 12,Download figure ((17)) The transit time (TT) for the activation wavefront to travel from one point in the circuit to another can be estimated from the inverse of the conduction velocity given by Equation 17Download figure ((18a)) Download figure ((18b)) for i=1 to n space steps. SinceDownload figure ((19)) and dTi is the change in IBZ thickness between space steps i and i+1Download figure ((20a)) Download figure ((20b)) LetDownload figure ((21a)) andDownload figure ((21b)) ThenDownload figure ((22)) To minimize the transit time from the thinnest IBZ (isthmus location) to the thicker tissue of the outer pathway or vice versa, the quotient rule can be used, that is,Download figure ((23)) Thus,Download figure ((24)) with the equation set to zero, which reduces toDownload figure ((25)) HenceDownload figure ((26)) which has an approximate solution ofDownload figure ((27)) for n space steps, and a total thickness change Z from isthmus to outer pathway or vice versa, regardless of whether the sign of dTi is positive or negative (wavefront convex or concave). Therefore, according to Equation 27, the most rapid path from thin-to-thick or thick-to-thin, would be one in which there is a constant, minimized thickness change ΔTi for all space steps i=1 to n. Although Equations 15 to 27 partially pertain to travel in the outer pathway where functional block does not occur, and it is not desirable to ablate there because it is not a constrained region, knowledge of the location of the fastest pathway may be important to determine how the circuit can interact with other potential circuit structures located elsewhere in the IBZ.Electrogram Fractionation at the Lateral Isthmus BoundariesAlthough the lateral isthmus boundaries can be treated, to a first approximation, as having a uniform and steep thickness transition all along each boundary (Figure 5A), this is not precisely correct. The change in thickness along each lateral boundary can be highly variable, because the thickness within the isthmus versus that in the outer pathway on the opposite side of the lateral boundaries can be quite variable when measured by histology or magnetic resonance imaging (MRI).8–10 If the change in thickness across the boundary causes critically convex wavefront curvature to occur along its entire length, then functional conduction block will form all along the boundary (Figure 5A). Suppose however that alterations in thickness at the transition are variable, and although they cause the activation wavefront to be convex, it does not become critically convex.18 Under such conditions, activation will proceed outward from the boundary without blocking, albeit very slowly, as is shown in Figure 5B. As for the uniform boundary in Figure 5A, the electrical impulse similarly travels through the isthmus of Figure 5B along the long axis. At the lateral boundaries, however, the thickness transitions are less steep and also variable when compared to Figure 5A. Thus, critically convex curvature is not attained, and the activation wavelets can propagate laterally. The lateral speed of propagation depends on the sharpness of the thickness transition there and is slower for steeper transitions (noted as smaller arrow length to connote lesser speed in Figure 5B). This lateral propagation does not break through to the outer pathway, because the wavefronts traveling along the circuit propagate and activate the outer pathway on either side of the isthmus rapidly, thus maintaining the functional integrity of the block lines at all segments along the lateral boundaries.When a bipolar electrode is positioned in proximity to the lateral boundaries with variable thickness transition during reentrant ventricular tachycardia, the difference in voltage at the bipoles will depend on the sum of all of the electric fields emanating from the disrupted, distinct activating wavelets, which are thereby independent voltage sources.18,19 This is given, to a first approximation, by the following equation18:Download figure ((28)) where φe is the extracellular voltage observed at an observation point P and time to, k1, and k2 are constants, x and y are the distances from source to recording electrode along the x and y axes in the coordinate system, and j is the grid square number, with the summation being for all grid squares that are activating at time to. If the electrode is bipolar, then the bipolar extracellular signal can be estimated as follows18:Download figure ((29)) where p and n represent the positive and negative electrodes, respectively. The bipolar extracellular signal can be calculated at each time epoch, typically every millisecond when the digital sampling rate is 1 kHz. Figure 6 illustrates how such wavefront discontinuities can cause electrogram fractionation. Figure 6A is the schematic representation of Figure 5B for reference, and Figure 6B shows the leading edge of several activation wavelets (tip of laterally directed arrows) in each case. According to Equation 29, the voltage observed at the bipolar electrode (example location shown as double circles in Figure 6A and 6B) will depend upon the location, length, and timing of each activating wavelet travelling laterally outward along the isthmus boundaries. Since the electric field diminishes as the distance from the source increases, a bipolar electrode in proximity to one of the lateral isthmus boundaries will be much less affected by the electric activity of distant wavelets. The bipolar electrode voltage will, however, be significantly affected by local electric activity from the activation wavelets in close proximity along the transition boundary. Because the wavelets have different distances from the electrode, varying timing and segment length, and different angle with respect to the bipole, the resulting bipolar electrogram, as described by Equation 29, will be fractionated, with each deflection caused by 1 or more of the independent wavelets of activation crossing the boundary. Larger deflections in the electrogram will occur when those segments of the boundary in closest proximity to the electrode, and those with longer segment length, activate nearby to one and then the other of the bipoles (Figure 6, labels 1 and 2). Immediately adjacent segments with differing speed of propagation will also cause relatively large deflections in the electrogram (Figure 6, label 3).Download figureDownload PowerPointFigure 6. The figure shows how activation wavefront discontinuities can form at the lateral boundaries because of the varying speed of the wavelet at each segment (A and B). The distinct wavelets are noted by dotted arrows in A and a few are shown by short lines in B. A bipolar recording electrode is depicted as a double circle in A and B. The bipolar electrogram recorded from this location is fractionated (C) because of the proximity of several discontinuous activation wavelets 1 to 3.Additional, lesser electrogram deflections will appear and overlap besides those shown in Figure 6C, caused by voltage sources at wavelets more distant from the bipole, and depending on their timing. Since the electric field diminishes as the distance increases from the source (activation wavefront) to the receiver (electrode), fractionation because of discontinuities in the laterally traveling wavelets can only be readily observed by a bipolar electrode positioned within approximately 1–2 cm of the lateral isthmus boundaries. Thus, this mechanism cannot be used to explain electrogram fractionation that might occur very distant from the isthmus location. However, in canine postinfarction, much of the observed electrogram fractionation occurs in proximity to the lateral lines of block during reentrant ventricular tachycardia.18 Furthermore, electrogram fractionation can also occur at the same region during sinus rhythm, if the wavefront travel direction also causes it to transition across zones of sharp gradient from thin-to-thick along the IBZ.18Unidirectional Block During Premature StimulationReentrant ventricular tachycardia is initiated when a premature stimulus, with a coupling interval of approximately 120–170 ms, is imparted to the IBZ at an appropriate location.1,2,20 This is illustrated in Figure 7, which is a schematized activation map with 20 ms isochronal spacing, and IBZ thickness is given by the grayscale bar. Approximate IBZ thicknesses along the z axis are shown. In Figure 7A, the location where the reentry isthmus will form is outlined by dotted black lines. The premature excitation wavefront originating from the stimulus site propagates outward and blocks along a boundary line in the thin-to-thick direction, which is a few centimeters distant (Figure 7A, white dashed line).21 The wavefront then bifurcates and travels around the block area. The distinct wavefronts coalesce on the opposite side of this boundary and proceed to travel across it in the opposite direction. If sufficient time for recovery of excitability has occurred, the activation wavefront reenters the previously excited region, resulting in onset of reentry (Figure 7B). Thus, this boundary is a unidirectional line of conduction block. As reentry is initiated (Figure 7B), the wavefront traveling within the isthmus will block in the lateral directions due to the steep thin-to-thick transitions there (short white arrows and double lines). In Figure 7B, there is also a thin-to-thick transition at the exit point. Block does not form there, however. This can be explained by the dependency for the ability of a particular degree of convex wavefront curvature to cause functional conduction block on the rate of stimulation.6,21 At the short coupling intervals of premature stimulation, 120–170 ms, functional block will occur even when the thin-to-thick transition is only moderately steep (Figure 7A, dashed white line). However, at the longer coupling interval of tachycardia, averaging 200 ms,22 a much steeper transition is necessary for block to occur (Figure 7B, dotted black lines). Thus, for this particular reentrant tachycardia circuit, with cycle length of 200 ms, block does not occur at the isthmus exit (time approximately 10–20 ms on the map of Figure 7B). The steepness of the spatial thickness transition may also be less at the isthmus exit, when compared with the entrance, further increasing the likelihood of successful propagation there at ventricular tachycardia cycle lengths averaging approximately 200 ms. The pattern of activation during premature stimulation, the necessity of the location of the stimulus site to be in proximity to the protoisthmus exit, the location of a moderately steep thin-to-thick transition at the protoisth