- Email: [email protected]
The effect of electrode density on the interpretation of atrial activation patterns in epicardial mapping of human persistent atrial fibrillation
1 2 3 4 5Q3 6 Q4 7 8 9 Q1 10Q5 11 12 13 14 15 16 17 18 19 20 21 Q6 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
The effect of electrode density on the interpretation of atrial activation patterns in epicardial mapping of human persistent atrial fibrillation Tomos E. Walters, PhD,*† Geoffrey Lee, PhD,*† Steven Spence, XX,* Jonathan M. Kalman, PhD*† From the *Department of Cardiology, The Royal Melbourne Hospital, Melbourne, Australia, and † Department of Medicine, University of Melbourne, Melbourne, Australia. BACKGROUND Mechanisms sustaining human persistent atrial fibrillation (AF) remain debated, with significant differences between high-density epicardial and global endocardial mapping studies. A key difference is the density of recording electrodes. OBJECTIVE We aimed to determine the differences in the prevalence of different atrial activation patterns, and specifically in the prevalence of rotational activations, with varying densities of bipolar electrodes.
P o .001). Simple broad wavefront activations became more prevalent (20% ⫾ 8% to 54% ⫾ 8%; P o .05) and complex patterns became less prevalent (48% ⫾ 8% to 9% ⫾ 8%; P o .05) with reducing density. The prevalence of rotational activity declined with bipole density, from median 5.0% (range 0.9%–12.1%) to 0% (range 0%–1.5%) (P ¼ .03). The largest change occurred between interbipole spacings of 5.0 3.5 and 5.0 7.1 mm.
METHODS Epicardial mapping was performed in 10 patients undergoing cardiac surgery, with bipolar electrograms recorded using a triangular plaque (6.75 cm2 area; 117 bipoles; 2.5-mm interbipole spacing) applied to the left atrial posterior wall or right atrial free wall. Dynamic wavefront mapping based on the timing of atrial electrograms was applied to 2 discrete 10-second AF segments. The spacing between bipolar electrode locations was increased from 2.5 3.5 mm in the horizontal and oblique directions to 5.0 3.5, 5.0 7.1, and 7.5 10.6 mm, with wavefront mapping repeated at each density.
CONCLUSION Apparent activation patterns in persistent AF vary significantly with electrode density. Low density underestimates the prevalence of complex and rotational patterns. The largest difference occurs between an interbipole spacing of 5.0 3.5 and a spacing of 5.0 7.1 mm. This may have important implications for mapping technology design.
RESULTS As density reduced, there was a significant change in relative proportions of the various activation patterns (F¼3.69;
(Heart Rhythm 2016;0:0–6) rights reserved.
Introduction The mechanisms by which human atrial fibrillation (AF) is maintained continue to be debated,1,2 with particular focus on the prevalence of rotational activation and its mechanistic significance as a focal driver of fibrillation. A range of highdensity epicardial mapping studies,3–8 with an interelectrode spacing of 1.2–2.5 mm and the use of activation mapping based on the timing of discrete bipolar or unipolar atrial electrograms (EGMs), have reported no more than infrequent and transient rotational activity. Conversely, studies of human AF using Dr Walters is supported by a Postgraduate Research Scholarship from the National Health and Medical Research Council and the National Heart Foundation of Australia. Dr Lee is supported by an Early Career Fellowship from the National Health and Medical Research Council of Australia. Dr Kalman is supported by a Practitioner Fellowship from the National Health and Medical Research Council of Australia. Address reprint requests and correspondence: Dr Jonathan M. Kalman, Department of Cardiology, The Royal Melbourne Hospital, 500 Parnassus Avenue, Melbourne 94158, Australia. E-mail address: [email protected].
1547-5271/$-see front matter B 2016 Heart Rhythm Society. All rights reserved.
KEYWORDS Atrial fibrillation; Mapping; Bipolar electrode; Density; Rotor I
2016 Heart Rhythm Society. All
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
9,10
lower-density endocardial mapping baskets, with 64 unipolar electrodes aiming to cover the entire atrial endocardial surface and subsequent phase analysis of the unipolar atrial EGMs, have suggested that human AF may be driven by a small number of rotors with a high degree of temporal stability.11 A key difference between these 2 groups of mapping studies is the density of recording electrodes, and it remains unclear what electrode density is required to accurately discern atrial activation patterns and to identify the presence of any rotors. In this epicardial mapping study of human persistent AF, we therefore aimed to determine the differences in the distribution of atrial activation patterns and in the prevalence of rotational activations with progressive reduction in the density of recording electrodes.
Methods Ten patients with long-standing persistent AF undergoing a first elective cardiac surgical procedure were studied (Table 1). Participants were undergoing coronary artery http://dx.doi.org/10.1016/j.hrthm.2016.01.030
T1
2 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 Q7 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 P 119 R 120 I 121 N 122 T 123 & W 124 E 125 B 126 4 127 C 128 / 129 F 130 P O 131
Table 1
Heart Rhythm, Vol 0, No 0, Month 2016 Participant characteristics
Characteristic
Value
Age (y) Sex: male AF duration (y) BMI (kg/m2) Hypertension Smoking Diabetes Vascular disease Aspirin Warfarin RAS inhibitor β-Blocker Calcium channel blocker Digoxin Class I antiarrhythmic medication Sotalol Amiodarone LA area (cm2) LVEF (%)
69 ⫾ 17 6 (60) 4.5 (2–6) 29.6 ⫾ 4.3 9 (90) 3 (30) 5 (50) 7 (70) 3 (30) 9 (90) 5 (50) 0 6 (60) 6 (60) 0 2 (20) 0 32.3 ⫾ 5.4 52.3 ⫾ 5.8
Values are presented as mean ⫾ SD, as n (%), or as median (interquartile range). AF ¼ atrial fibrillation; BMI ¼ body mass index; LA ¼ left atrium; LVEF ¼ left ventricular ejection fraction; RAS ¼ renin-angiotensin system.
bypass graft surgery (CABGS; n ¼ 3), aortic valve replacement (AVR; n ¼ 2), mitral valve replacement (MVR; n ¼ 2), CABGS/AVR (n ¼ 2), and CABGS/
MVR/AVR (n ¼ 1). Antiarrhythmic medications were discontinued Z5 half-lives before surgery. All participants gave written informed consent, with the protocol approved by the Melbourne Health Human Research and Ethics Committee.
Epicardial mapping protocol High-density atrial epicardial mapping was performed after median sternotomy and pericardial division, before cardioplegia and cardiopulmonary bypass. Mapping involved a custom-made triangular plaque (UniServices, Auckland, New Zealand) used in previous epicardial AF mapping studies.6,7 This plaque includes 128 silver-plated copper electrodes 0.7 mm in diameter (117 bipole pairs), with spacing between the location of bipole pairs of 2.5 mm in the horizontal and vertical directions and of 3.5 mm in the oblique direction, and a total mapping area of 6.75 cm2 (Figure 1). The plaque was positioned by the operating surgeon, and bipolar signals were recorded. Nine recordings from the posterior left atrial wall were made, with 4 from the right atrial free wall. Bipolar EGMs, with a sampling frequency of 1000 Hz and a band-pass filter of 0.05–400 Hz, were recorded using the UnEmap mapping system (UniServices). After the recording was visually scanned, 2 discrete 10-second AF segments with high-quality signals were analyzed.
Figure 1 High-density epicardial mapping plaque used in this study (left). Bipole density schemas 1–4 (right), with the locations of active bipolar recording sites shown in white. The distances between active recording sites in schema 1 are 2.5 mm in the horizontal or vertical direction and 3.5 mm in the oblique direction; in schema 2, the distances are 5.0 and 3.5 mm, respectively; in schema 3, the distances are 5.0 and 7.1 mm, respectively; and in schema 4, the distances are 7.5 and 10.6 mm, respectively.
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 F1 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
Walters et al 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245
Effect of Electrode Density in AF Mapping
Signal analysis UnEmap recordings were exported into customized software (Cardiac ElectroPhysiology Analysis System, Cuoretech, Sydney, Australia) for further analysis. An automated algorithm was used to identify discrete atrial EGMs by peak negative dV/dt events in the voltage signal. Consistent with previous reports6,7 a noise threshold of 0.1 mV, a maximum width criterion of 10 ms, and a refractory period of 50 ms were applied to avoid detection of broad far-field activations and multiple detections of the same activation complex. All automated annotations were visually verified and manually adjusted as required.
AF wavefront mapping Within the Cardiac ElectroPhysiology Analysis System, the series of activation times for each bipole referenced to the start of the 10-second AF segment and the Cartesian coordinates precisely locating each bipole were used to construct dynamic wavefront maps on the basis of the timing of each bipolar atrial activation. The sequences of activation times detected in the voltage signal at each bipole location over the 10-second segment were converted to cycling 0 to 2π radian signal phase values at equal time increments of 5 ms. Phase 0 was set at the instant of each successive activation detection. A parabolic phase template with a nominal duration of 100 ms was then applied to govern the phase variation from 0 to 2π radians over that interval. Finally, a gray color scale (0 white, 2π black) was applied to the resulting phase values, which, in combination with the parabolic template, was found to be most effective for clear visualization of activation patterns.
3 the 10-second time segment, initially using EGMs from all 117 bipole locations on the mapping plaque (Figure 1, schema 1), with spacing between bipole locations of 2.5 mm in the horizontal direction and 3.5 mm in the oblique direction (2.5 3.5 mm) as in previous reports.6,7 Analysis of dynamic wavefront maps was then repeated with exclusion of 58, 86, and 102 bipole locations (Figure 1). The resulting bipole location spacing was 5.0 3.5 mm (schema 2), 5.0 7.1 mm (schema 3), and 7.5 10.6 mm (schema 4). The spacing of the 2 electrodes comprising each included bipole was held constant throughout, with linear interpolation applied by the mapping software to demonstrate activation between the included bipole locations. All wavefront maps were de-identified and analyzed without knowledge of the map patterns for the differing electrode densities.
Statistics
Consistent with previous studies,3,4,6,7 activation patterns were classified as disorganized activity, rotational activations, focal activations, and linear wavefronts. Broad linear wavefronts activated the entire width of the plaque while narrow linear wavefronts had a breadth of r7 bipoles. Linear wavefronts were subclassified by the direction of propagation onto the triangular plaque (side 1, 2, or 3). The simultaneous presence within the dynamic wavefront maps of multiple broad or linear wavefronts was considered to represent complex activation. Rotational activation was defined as a wavefront rotating Z3601 within the mapped area and governing regional activation. Focal activations were represented by activation emanating from within the mapped region and subsequently spreading radially a distance of 43 bipoles. Disorganized activity failed to fulfill criteria for the patterns detailed above. It was composed of early activation inside the mapping area that failed to propagate 43 bipoles or activations that occurred as isolated beats dissociated from activation at adjacent bipole sites.
Statistical analysis was performed using commercially available software (GenStat 16th Edition, VSN International Ltd, Hempstead, United Kingdom, and STATA version 12.1, StataCorp, College Station, TX). Normality was tested using the Shapiro-Wilk method. Continuous variables are expressed as mean ⫾ SD if normally distributed and as median (interquartile range [IQR]) otherwise, and categorical variables are expressed as number of patients (percentage). Statistical significance was assessed at the .05 level. The overall effect of varying bipole density was examined using a linear mixed model, with the patient undergoing mapping included as a random term and the activation pattern type, bipole density, and the interaction of pattern type and density included as fixed terms. An overall F statistic for this model, including the interaction of pattern type and density, was computed, and the approximate least significant difference (method at the .05 level was used to estimate the significance of differences within the model. Six activation pattern types were included, with disorganized activity, focal activation, and rotational activation included individually, but narrow wavefront activation from each of the 3 sides of the mapping plaque grouped together, broad wavefront activation from each of the 3 sides of the mapping plaque grouped together, and the presence of multiple narrow wavefronts, multiple broad wavefronts, and simultaneous narrow and broad wavefronts grouped together as complex activation patterns. The linear model fulfilled requisite criteria, with normal distribution of residuals and homoscedasticity in the variance of residuals. Activation patterns were also analyzed individually. With marked heterogeneity between patients in the frequency of each activation pattern, as previously demonstrated,6 the nonparametric Friedman test for repeated measures was used to assess the significance of differences in the activation pattern frequency identified with each of the 4 electrode location densities.
Electrode density reduction
Results
The dynamic AF wavefront maps were analyzed to record the sequence and number of each activation pattern type over
In the initial analysis applying all 117 bipole locations (schema 1), disorganized activity accounted for 1% (IQR
Activation pattern classification
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
4 303 304 305 306 307 308 309 310 311 312 F2 313 314 315 316 F3 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 T2 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
0%–14%) of activations, rotational activity accounted for 5.0% (IQR 0.9%–12.1%), focal activations for 1.8% (IQR 0%–4.2%), single narrow linear wavefronts for 3% (IQR 0%–8.1%), multiple narrow linear wavefronts for 26.4% (IQR 12.3%–46.4%), single broad linear wavefronts for 11.4% (IQR 5.8%–31.4%), multiple broad linear wavefronts for 7.4% (IQR 2.3%–17.0%), and simultaneous broad linear and narrow linear wavefronts for 6.4% (IQR 4.2%–12.2%). Overall complex patterns accounted for 53.1% (IQR 41.5%– 56.8%) (Figure 2). The linear model revealed the change in the estimated proportions of total activations represented by each activation pattern as electrode density was reduced to be highly significant (F¼3.69; P o .001; Figure 3). Within this model, single broad wavefront activations became significantly more prevalent as a percentage of all observed activation patterns (20% ⫾ 8% with density schema 1, 25% ⫾ 8% with density schema 2, 51% ⫾ 8% with density schema 3, and 54% ⫾ 8% with density schema 4; P o .05), with the largest increase when stepping down in density from 5.0 3.5 to 5.0 7.1 mm. Conversely, complex activation patterns became significantly less prevalent (48% ⫾ 8% with density schema 1, 44% ⫾ 8% with density schema 2, 19% ⫾ 8% with density schema 3, and 9% ⫾ 8% with density schema 4; P o .05), again with a marked fall in prevalence when stepping down in bipole density from 5.0 3.5 to 5.0 7.1 mm. Analyzing each activation pattern in isolation across all mapped regions, there was a significant reduction in the representation of rotational activations and complex activation patterns with reduction in bipole location density (Table 2) and a significant increase only in the prevalence of single broad linear wavefront activation. For each of these activation patterns, there was again a particularly marked
Heart Rhythm, Vol 0, No 0, Month 2016 change in prevalence with a step-down in electrode density from 5.0 3.5 to 5.0 7.1 mm. There was no systematic change in the prevalence of disorganized activity, of focal activations, or of single linear wavefronts as a result of varying bipole location density (Table 2). Specifically focusing on rotational activations in the 11 of 13 mapped regions in which rotational activation patterns were observed, there was a significant reduction in the frequency of rotational activity as bipole density declined, from a median prevalence of 7.4% (IQR 4.3%–15.1%) with schema 1 to a median prevalence of 0% (IQR 0%–1.5%) with the application of the least dense schema (P ¼ .04). Again there was a marked step-down in the observed prevalence of rotational activity between a bipole location density of 5.0 3.5 mm and a density of 5.0 7.1 mm (6.9% [IQR 6.1%–13.2%] to 2.5% [IQR 0%–4.1%]).
Discussion The key finding of this study is that the atrial activation patterns observed in human persistent AF when applying wavefront mapping on the basis of the timing of discrete bipolar EGMs are dependent on the density of the recording electrodes. In particular, low density of bipolar electrode locations overstates the prevalence of simple broad linear wavefront activation, understates the prevalence of complex activation patterns, and understates the prevalence of rotational activations. The largest difference occurs with a stepdown from an interbipole spacing of 5.0 3.5 mm to a spacing of 5.0 7.1 mm. Multiple groups have used high-density epicardial mapping plaques to study atrial activation patterns in human AF.3–8 Such mapping is spatially limited but uses mapping systems with small interelectrode distances of 1.2–2.5 mm to
Figure 2 Distribution of activation pattern through the application of the most dense schema of bipolar electrode locations (2.5 3.5 mm). A: Single narrow linear wavefronts are grouped together and single linear broad wavefronts are grouped together regardless of the direction of activation of the mapped region. B: Multiple narrow linear wavefronts, multiple broad linear wavefronts, and the simultaneous presence of narrow linear and broad linear wavefronts are grouped together as complex activations.
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
Walters et al 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473
P R I N T & W E B 4 C / F P O
Effect of Electrode Density in AF Mapping
5 with 64 unipolar electrodes mounted on 8 splines.9,10 The cardinal difference in the output of these studies has been the finding of a high prevalence of temporally stable rotors, the mechanistic significance of which has been indicated by catheter ablation studies that have demonstrated enhanced medium-term efficacy when these rotors are targeted, in comparison to a standard approach of pulmonary vein isolation alone.10,13,14 It may be that this relates to the critical differences in methodology, which are the markedly lower electrode density and the application of phase mapping based on Hilbert transformation of the voltage signal to derive a phase representation of the voltage amplitude over time. On the basis of theoretical consideration of action potential duration and conduction velocity restitution curves, which suggests a minimal wavelength of reentry of 4–5 cm in human atria, Narayan and coworkers15,16 have argued that while the spatial resolution of a multielectrode basket catheter might not be expected to resolve the rotor core at the tip of a spiral wave, it is sufficient to resolve the rotating arms of the spiral wave emanating from this core.12 In a computational model of a temporally stable rotor with a wavelength of 5 cm and a coherent field spanning the entire field of view, the distance between locations at which virtual EGMs were sampled was progressively increased, moving from a spatial resolution of 2.5 mm to one of 12.5 mm. While at the lower resolution the rotor core could not be mapped and there was some distortion in the appearance of rotation, a rotational pattern could be discerned.12 Using another computational model, however, with a wavelength between successive spiral arms of 3.2 cm and a coherent domain of 6 cm beyond which there was degeneration into fibrillatory conduction, there was marked deterioration in the ability to discern rotational activity when moving from a spatial resolution of 6 mm to one of 12.5 mm. This is consistent with the observation in this study of progressive insensitivity to the presence of rotational activity between an interelectrode spacing of 2.5 3.5 mm and a spacing of 7.5 10.6 mm. The phase analysis of activation during analysis of AF was initially reported by Chen et al17 in high-resolution optical mapping studies of sheep atria. Initially, a lag-return method was used to determine the phase of cyclical optical
474 475 476 477 478 479 480 481 482 483 484 485 486 487 Figure 3 Output of the linear mixed model evaluating the prevalence of 488 each activation pattern type as a percentage of total activations with the 489 application of each of the 4 bipole densities. The frequency of activation 490 patterns remains little changed with the application of schemas 1 (2.5 3.5 491 mm) and 2 (5.0 3.5 mm). Single broad wavefront activations are significantly more prevalent with the application of the less dense schemas 3 and 4, while the 492 frequency of complex activation patterns is significantly lower with the 493 application of these sparse electrode arrangements (see details in text). 494 495 record unipolar3–5,8 or bipolar6–8 EGMs. A common feature 496 to all such studies is that analysis of atrial activation patterns 497 has been based on accurate identification of the timing of 498 discrete atrial EGMs. Taken together, these studies have 499 suggested that AF is maintained by multiple wavelet 500 propagation, with dissociated wavefronts in the remodeled 501 atrium separated by regions of fibrosis and poor tissue 502 coupling.4 They have suggested that dissociation between 503 epicardial and endocardial layers allows “endo-epi” wave504 front breakthrough that may aid perpetuation of fibrillation.5 505 While high-frequency foci have been described as a potential 506 sustaining source of fibrillation,8 rotational activity has been 507 relatively infrequent and transient.6,7 An inherent limitation 508 to such studies is that the mechanistic significance of such 509 patterns has not been tested by ablation. 510 An alternative approach to mapping human AF has 511 involved the application of more global but lower-density 512 recording systems, in particular multispline basket catheters. 513 These catheters, with an interelectrode distance of o4 mm at 514 the poles and o1 cm at the equator12 when deployed 515 optimally, aim to map the entire atrial endocardial surface 516 517 Table 2 Percentage of all atrial activations represented by each pattern, with the 4 different electrode density schemas sequentially applied 518 519 to the 10-s atrial fibrillation segments from each of the 13 mapped regions* 520 Schema 1 Schema 2 Schema 3 Schema 4 521 Activation pattern (2.5 3.5 mm) (5.0 3.5 mm) (5.0 7.1 mm) (7.5 10.6 mm) P 522 Disorganized 1.0 (0–14.0) 2.8 (0–14.5) 7.2 (0–11.9) 15.7 (2.5–21.0) .80 523 Rotational 5.0 (0.9–12.1) 6.1 (1.0–10.5) 0.9 (0–3.7) 0 (0–1.5) .03 Q8524 Focal 1.8 (0–4.2) 2.0 (0–3.6) 2.6 (1.0–9.3) 6.1 (1.7–10.0) .80 525 Single narrow linear wavefront 3.0 (0–8.1) 4.1 (0–8.3) 5.3 (0–12.5) 2.8 (0–18.9) 1.00 Single broad linear wavefront 11.4 (5.8–31.4) 22.8 (9.5–36.3) 48.2 (40.1–72.4) 54.5 (37.4–80.6) .0003 526 Complex 53.1 (41.5–56.8) 52.0 (35.6–53.1) 20.2 (11.8–25.8) 7.2 (1.0–12.1) .02 527 528 With decreasing density of bipole locations, a marked fall in the prevalence of complex and rotational activation was observed, offset by a significant 529 increase in the prevalence of simple broad linear wavefront activation. * The median and interquartile range are calculated from the percentages determined for each mapped region. 530
6 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
action potentials, and this was subsequently refined through the use of the Hilbert transform to derive the signal phase from the voltage-time series.18,19 Application of this methodology to the analysis of markedly lower-density extracellular EGMs may exacerbate the effect of low-density mapping on observed atrial activation patterns, as the interpolation of the Hilbert-based phase at dispersed electrodes, which may by chance be activated sequentially by unrelated wavefronts, carries an inherent bias toward the demonstration of rotational activity in the area between the electrodes.20 The spatial resolution of schema 1 in this study, as in previous epicardial mapping studies, was sufficient to demonstrate a high prevalence of complex activation patterns that often included the simultaneous presence of multiple discrete narrow wavefronts. At the resolution of schema 4, complex patterns were much less frequently discerned, with broad linear wavefront becoming dominant. It is possible that the application of other mapping algorithms to these low-density recordings might have demonstrated more rotational activity.
Study limitations This study has examined only the effect of reducing electrode density on one particular wavefront mapping system on the basis of accurate determination of the timing of discrete atrial EGMs. Likewise, we have studied only the results of highdensity but spatially limited epicardial mapping and have also not examined more global endocardial recordings.
Conclusion Atrial activation patterns observed in epicardial mapping of human persistent AF vary significantly with the density of the recording bipolar electrodes. Low density understates the prevalence of complex and rotational activation patterns. The largest change occurs with a step-down in spacing between bipole locations from 5.0 3.5 to 5.0 7.1 mm. This may have important implications for the design of mapping technologies.
References 1. Allessie M, De Groot N. CrossTalk opposing view: rotors have not been demonstrated to be the drivers of atrial fibrillation. J Physiol 2014;592: 3167–3170. 2. Narayan SM, Jalife J. CrossTalk proposal: rotors have been demonstrated to drive human atrial fibrillation. J Physiol 2014;592:3163–3166. 3. Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. Highdensity mapping of electrically induced atrial fibrillation in humans. Circulation 1994;89:1665–1680.
Heart Rhythm, Vol 0, No 0, Month 2016 4. Allessie MA, de Groot NMS, Houben RPM, Schotten U, Boersma E, Smeets JL, Crijns HJ. Electropathological substrate of long-standing persistent atrial fibrillation in patients with structural heart disease: longitudinal dissociation. Circ Arrhythm Electrophysiol 2010;3:606–615. 5. de Groot NMS, Houben RPM, Smeets JL, Boersma E, Schotten U, Schalij MJ, Crijns H, Allessie MA. Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough. Circulation 2010;122:1674–1682. 6. Lee G, Kumar S, Teh A, et al. Epicardial wave mapping in human long-lasting persistent atrial fibrillation: transient rotational circuits, complex wavefronts, and disorganized activity. Eur Heart J 2014;35:86–97. 7. Walters TE, Lee G, Morris G, et al. Temporal stability of rotors and atrial activation patterns in persistent human atrial fibrillation. JACC Clin Electrophysiol 2015;1:14–24. 8. Lee S, Sahadevan J, Khrestian CM, Cakulev I, Markowitz A, Waldo AL. Simultaneous bi-atrial high density (510-512 electrodes) epicardial mapping of persistent and long-standing persistent atrial fibrillation in patients: new insights into the mechanism of its maintenance. Circulation 2015;132:2108–2117. 9. Narayan SM, Krummen DE, Rappel W-J. Clinical mapping approach to diagnose electrical rotors and focal impulse sources for human atrial fibrillation. J Cardiovasc Electrophysiol 2012;23:447–454. 10. Narayan SM, Krummen DE, Shivkumar K, Clopton P, Rappel W-J, Miller JM. Treatment of atrial fibrillation by the ablation of localized sources: CONFIRM (Conventional Ablation for Atrial Fibrillation With or Without Focal Impulse and Rotor Modulation) trial. J Am Coll Cardiol 2012;60:628–636. 11. Narayan SM, Shivkumar K, Krummen DE, Miller JM, Rappel W-J. Panoramic electrophysiological mapping but not electrogram morphology identifies stable sources for human atrial fibrillation: stable atrial fibrillation rotors and focal sources relate poorly to fractionated electrograms. Circ Arrhythm Electrophysiol 2013;6:58–67. 12. Rappel W-J, Narayan SM. Theoretical considerations for mapping activation in human cardiac fibrillation. Chaos 2013;23:023113. 13. Narayan SM, Baykaner T, Clopton P, Schricker A, Lalani GG, Krummen DE, Shivkumar K, Miller JM. Ablation of rotor and focal sources reduces late recurrence of atrial fibrillation compared with trigger ablation alone: extended follow-up of the CONFIRM trial (Conventional Ablation for Atrial Fibrillation With or Without Focal Impulse and Rotor Modulation). J Am Coll Cardiol 2014;63:1761–1768. 14. Miller JM, Kowal RC, Swarup V, et al. Initial independent outcomes from focal impulse and rotor modulation ablation for atrial fibrillation: multicenter FIRM registry. J Cardiovasc Electrophysiol 2014;25:921–929. 15. Lalani GG, Schricker A, Gibson M, Rostamian A, Krummen DE, Narayan SM. Atrial conduction slows immediately before the onset of human atrial fibrillation: a bi-atrial contact mapping study of transitions to atrial fibrillation. J Am Coll Cardiol 2012;59:595–606. 16. Narayan SM, Kazi D, Krummen DE, Rappel W-J. Repolarization and activation restitution near human pulmonary veins and atrial fibrillation initiation: a mechanism for the initiation of atrial fibrillation by premature beats. J Am Coll Cardiol 2008;52:1222–1230. 17. Chen J, Mandapati R, Berenfeld O, Skanes AC, Gray RA, Jalife J. Dynamics of wavelets and their role in atrial fibrillation in the isolated sheep heart. Cardiovasc Res 2000;48:220–232. 18. Bray M-A, Wikswo JP. Considerations in phase plane analysis for non-stationary reentrant cardiac behavior. Phys Rev E 2002;65:051902. 19. Umapathy K, Nair K, Masse S, Krishnan S, Rogers J, Nash MP, Nanthakumar K. Phase mapping of cardiac fibrillation. Circ Arrhythm Electrophysiol 2010;3: 105–114. 20. Berenfeld O, Oral H. The quest for rotors in atrial fibrillation: different nets catch different fishes. Heart Rhythm 2012;9:1440–1441.
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621
The effect of electrode density on the interpretation of atrial activation patterns in epicardial mapping of human persistent atrial fibrillation Tomos E. Walters, PhD,*† Geoffrey Lee, PhD,*† Steven Spence, XX,* Jonathan M. Kalman, PhD*† From the *Department of Cardiology, The Royal Melbourne Hospital, Melbourne, Australia, and † Department of Medicine, University of Melbourne, Melbourne, Australia. BACKGROUND Mechanisms sustaining human persistent atrial fibrillation (AF) remain debated, with significant differences between high-density epicardial and global endocardial mapping studies. A key difference is the density of recording electrodes. OBJECTIVE We aimed to determine the differences in the prevalence of different atrial activation patterns, and specifically in the prevalence of rotational activations, with varying densities of bipolar electrodes.
P o .001). Simple broad wavefront activations became more prevalent (20% ⫾ 8% to 54% ⫾ 8%; P o .05) and complex patterns became less prevalent (48% ⫾ 8% to 9% ⫾ 8%; P o .05) with reducing density. The prevalence of rotational activity declined with bipole density, from median 5.0% (range 0.9%–12.1%) to 0% (range 0%–1.5%) (P ¼ .03). The largest change occurred between interbipole spacings of 5.0 3.5 and 5.0 7.1 mm.
METHODS Epicardial mapping was performed in 10 patients undergoing cardiac surgery, with bipolar electrograms recorded using a triangular plaque (6.75 cm2 area; 117 bipoles; 2.5-mm interbipole spacing) applied to the left atrial posterior wall or right atrial free wall. Dynamic wavefront mapping based on the timing of atrial electrograms was applied to 2 discrete 10-second AF segments. The spacing between bipolar electrode locations was increased from 2.5 3.5 mm in the horizontal and oblique directions to 5.0 3.5, 5.0 7.1, and 7.5 10.6 mm, with wavefront mapping repeated at each density.
CONCLUSION Apparent activation patterns in persistent AF vary significantly with electrode density. Low density underestimates the prevalence of complex and rotational patterns. The largest difference occurs between an interbipole spacing of 5.0 3.5 and a spacing of 5.0 7.1 mm. This may have important implications for mapping technology design.
RESULTS As density reduced, there was a significant change in relative proportions of the various activation patterns (F¼3.69;
(Heart Rhythm 2016;0:0–6) rights reserved.
Introduction The mechanisms by which human atrial fibrillation (AF) is maintained continue to be debated,1,2 with particular focus on the prevalence of rotational activation and its mechanistic significance as a focal driver of fibrillation. A range of highdensity epicardial mapping studies,3–8 with an interelectrode spacing of 1.2–2.5 mm and the use of activation mapping based on the timing of discrete bipolar or unipolar atrial electrograms (EGMs), have reported no more than infrequent and transient rotational activity. Conversely, studies of human AF using Dr Walters is supported by a Postgraduate Research Scholarship from the National Health and Medical Research Council and the National Heart Foundation of Australia. Dr Lee is supported by an Early Career Fellowship from the National Health and Medical Research Council of Australia. Dr Kalman is supported by a Practitioner Fellowship from the National Health and Medical Research Council of Australia. Address reprint requests and correspondence: Dr Jonathan M. Kalman, Department of Cardiology, The Royal Melbourne Hospital, 500 Parnassus Avenue, Melbourne 94158, Australia. E-mail address: [email protected].
1547-5271/$-see front matter B 2016 Heart Rhythm Society. All rights reserved.
KEYWORDS Atrial fibrillation; Mapping; Bipolar electrode; Density; Rotor I
2016 Heart Rhythm Society. All
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
9,10
lower-density endocardial mapping baskets, with 64 unipolar electrodes aiming to cover the entire atrial endocardial surface and subsequent phase analysis of the unipolar atrial EGMs, have suggested that human AF may be driven by a small number of rotors with a high degree of temporal stability.11 A key difference between these 2 groups of mapping studies is the density of recording electrodes, and it remains unclear what electrode density is required to accurately discern atrial activation patterns and to identify the presence of any rotors. In this epicardial mapping study of human persistent AF, we therefore aimed to determine the differences in the distribution of atrial activation patterns and in the prevalence of rotational activations with progressive reduction in the density of recording electrodes.
Methods Ten patients with long-standing persistent AF undergoing a first elective cardiac surgical procedure were studied (Table 1). Participants were undergoing coronary artery http://dx.doi.org/10.1016/j.hrthm.2016.01.030
T1
2 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 Q7 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 P 119 R 120 I 121 N 122 T 123 & W 124 E 125 B 126 4 127 C 128 / 129 F 130 P O 131
Table 1
Heart Rhythm, Vol 0, No 0, Month 2016 Participant characteristics
Characteristic
Value
Age (y) Sex: male AF duration (y) BMI (kg/m2) Hypertension Smoking Diabetes Vascular disease Aspirin Warfarin RAS inhibitor β-Blocker Calcium channel blocker Digoxin Class I antiarrhythmic medication Sotalol Amiodarone LA area (cm2) LVEF (%)
69 ⫾ 17 6 (60) 4.5 (2–6) 29.6 ⫾ 4.3 9 (90) 3 (30) 5 (50) 7 (70) 3 (30) 9 (90) 5 (50) 0 6 (60) 6 (60) 0 2 (20) 0 32.3 ⫾ 5.4 52.3 ⫾ 5.8
Values are presented as mean ⫾ SD, as n (%), or as median (interquartile range). AF ¼ atrial fibrillation; BMI ¼ body mass index; LA ¼ left atrium; LVEF ¼ left ventricular ejection fraction; RAS ¼ renin-angiotensin system.
bypass graft surgery (CABGS; n ¼ 3), aortic valve replacement (AVR; n ¼ 2), mitral valve replacement (MVR; n ¼ 2), CABGS/AVR (n ¼ 2), and CABGS/
MVR/AVR (n ¼ 1). Antiarrhythmic medications were discontinued Z5 half-lives before surgery. All participants gave written informed consent, with the protocol approved by the Melbourne Health Human Research and Ethics Committee.
Epicardial mapping protocol High-density atrial epicardial mapping was performed after median sternotomy and pericardial division, before cardioplegia and cardiopulmonary bypass. Mapping involved a custom-made triangular plaque (UniServices, Auckland, New Zealand) used in previous epicardial AF mapping studies.6,7 This plaque includes 128 silver-plated copper electrodes 0.7 mm in diameter (117 bipole pairs), with spacing between the location of bipole pairs of 2.5 mm in the horizontal and vertical directions and of 3.5 mm in the oblique direction, and a total mapping area of 6.75 cm2 (Figure 1). The plaque was positioned by the operating surgeon, and bipolar signals were recorded. Nine recordings from the posterior left atrial wall were made, with 4 from the right atrial free wall. Bipolar EGMs, with a sampling frequency of 1000 Hz and a band-pass filter of 0.05–400 Hz, were recorded using the UnEmap mapping system (UniServices). After the recording was visually scanned, 2 discrete 10-second AF segments with high-quality signals were analyzed.
Figure 1 High-density epicardial mapping plaque used in this study (left). Bipole density schemas 1–4 (right), with the locations of active bipolar recording sites shown in white. The distances between active recording sites in schema 1 are 2.5 mm in the horizontal or vertical direction and 3.5 mm in the oblique direction; in schema 2, the distances are 5.0 and 3.5 mm, respectively; in schema 3, the distances are 5.0 and 7.1 mm, respectively; and in schema 4, the distances are 7.5 and 10.6 mm, respectively.
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 F1 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
Walters et al 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245
Effect of Electrode Density in AF Mapping
Signal analysis UnEmap recordings were exported into customized software (Cardiac ElectroPhysiology Analysis System, Cuoretech, Sydney, Australia) for further analysis. An automated algorithm was used to identify discrete atrial EGMs by peak negative dV/dt events in the voltage signal. Consistent with previous reports6,7 a noise threshold of 0.1 mV, a maximum width criterion of 10 ms, and a refractory period of 50 ms were applied to avoid detection of broad far-field activations and multiple detections of the same activation complex. All automated annotations were visually verified and manually adjusted as required.
AF wavefront mapping Within the Cardiac ElectroPhysiology Analysis System, the series of activation times for each bipole referenced to the start of the 10-second AF segment and the Cartesian coordinates precisely locating each bipole were used to construct dynamic wavefront maps on the basis of the timing of each bipolar atrial activation. The sequences of activation times detected in the voltage signal at each bipole location over the 10-second segment were converted to cycling 0 to 2π radian signal phase values at equal time increments of 5 ms. Phase 0 was set at the instant of each successive activation detection. A parabolic phase template with a nominal duration of 100 ms was then applied to govern the phase variation from 0 to 2π radians over that interval. Finally, a gray color scale (0 white, 2π black) was applied to the resulting phase values, which, in combination with the parabolic template, was found to be most effective for clear visualization of activation patterns.
3 the 10-second time segment, initially using EGMs from all 117 bipole locations on the mapping plaque (Figure 1, schema 1), with spacing between bipole locations of 2.5 mm in the horizontal direction and 3.5 mm in the oblique direction (2.5 3.5 mm) as in previous reports.6,7 Analysis of dynamic wavefront maps was then repeated with exclusion of 58, 86, and 102 bipole locations (Figure 1). The resulting bipole location spacing was 5.0 3.5 mm (schema 2), 5.0 7.1 mm (schema 3), and 7.5 10.6 mm (schema 4). The spacing of the 2 electrodes comprising each included bipole was held constant throughout, with linear interpolation applied by the mapping software to demonstrate activation between the included bipole locations. All wavefront maps were de-identified and analyzed without knowledge of the map patterns for the differing electrode densities.
Statistics
Consistent with previous studies,3,4,6,7 activation patterns were classified as disorganized activity, rotational activations, focal activations, and linear wavefronts. Broad linear wavefronts activated the entire width of the plaque while narrow linear wavefronts had a breadth of r7 bipoles. Linear wavefronts were subclassified by the direction of propagation onto the triangular plaque (side 1, 2, or 3). The simultaneous presence within the dynamic wavefront maps of multiple broad or linear wavefronts was considered to represent complex activation. Rotational activation was defined as a wavefront rotating Z3601 within the mapped area and governing regional activation. Focal activations were represented by activation emanating from within the mapped region and subsequently spreading radially a distance of 43 bipoles. Disorganized activity failed to fulfill criteria for the patterns detailed above. It was composed of early activation inside the mapping area that failed to propagate 43 bipoles or activations that occurred as isolated beats dissociated from activation at adjacent bipole sites.
Statistical analysis was performed using commercially available software (GenStat 16th Edition, VSN International Ltd, Hempstead, United Kingdom, and STATA version 12.1, StataCorp, College Station, TX). Normality was tested using the Shapiro-Wilk method. Continuous variables are expressed as mean ⫾ SD if normally distributed and as median (interquartile range [IQR]) otherwise, and categorical variables are expressed as number of patients (percentage). Statistical significance was assessed at the .05 level. The overall effect of varying bipole density was examined using a linear mixed model, with the patient undergoing mapping included as a random term and the activation pattern type, bipole density, and the interaction of pattern type and density included as fixed terms. An overall F statistic for this model, including the interaction of pattern type and density, was computed, and the approximate least significant difference (method at the .05 level was used to estimate the significance of differences within the model. Six activation pattern types were included, with disorganized activity, focal activation, and rotational activation included individually, but narrow wavefront activation from each of the 3 sides of the mapping plaque grouped together, broad wavefront activation from each of the 3 sides of the mapping plaque grouped together, and the presence of multiple narrow wavefronts, multiple broad wavefronts, and simultaneous narrow and broad wavefronts grouped together as complex activation patterns. The linear model fulfilled requisite criteria, with normal distribution of residuals and homoscedasticity in the variance of residuals. Activation patterns were also analyzed individually. With marked heterogeneity between patients in the frequency of each activation pattern, as previously demonstrated,6 the nonparametric Friedman test for repeated measures was used to assess the significance of differences in the activation pattern frequency identified with each of the 4 electrode location densities.
Electrode density reduction
Results
The dynamic AF wavefront maps were analyzed to record the sequence and number of each activation pattern type over
In the initial analysis applying all 117 bipole locations (schema 1), disorganized activity accounted for 1% (IQR
Activation pattern classification
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
4 303 304 305 306 307 308 309 310 311 312 F2 313 314 315 316 F3 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 T2 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
0%–14%) of activations, rotational activity accounted for 5.0% (IQR 0.9%–12.1%), focal activations for 1.8% (IQR 0%–4.2%), single narrow linear wavefronts for 3% (IQR 0%–8.1%), multiple narrow linear wavefronts for 26.4% (IQR 12.3%–46.4%), single broad linear wavefronts for 11.4% (IQR 5.8%–31.4%), multiple broad linear wavefronts for 7.4% (IQR 2.3%–17.0%), and simultaneous broad linear and narrow linear wavefronts for 6.4% (IQR 4.2%–12.2%). Overall complex patterns accounted for 53.1% (IQR 41.5%– 56.8%) (Figure 2). The linear model revealed the change in the estimated proportions of total activations represented by each activation pattern as electrode density was reduced to be highly significant (F¼3.69; P o .001; Figure 3). Within this model, single broad wavefront activations became significantly more prevalent as a percentage of all observed activation patterns (20% ⫾ 8% with density schema 1, 25% ⫾ 8% with density schema 2, 51% ⫾ 8% with density schema 3, and 54% ⫾ 8% with density schema 4; P o .05), with the largest increase when stepping down in density from 5.0 3.5 to 5.0 7.1 mm. Conversely, complex activation patterns became significantly less prevalent (48% ⫾ 8% with density schema 1, 44% ⫾ 8% with density schema 2, 19% ⫾ 8% with density schema 3, and 9% ⫾ 8% with density schema 4; P o .05), again with a marked fall in prevalence when stepping down in bipole density from 5.0 3.5 to 5.0 7.1 mm. Analyzing each activation pattern in isolation across all mapped regions, there was a significant reduction in the representation of rotational activations and complex activation patterns with reduction in bipole location density (Table 2) and a significant increase only in the prevalence of single broad linear wavefront activation. For each of these activation patterns, there was again a particularly marked
Heart Rhythm, Vol 0, No 0, Month 2016 change in prevalence with a step-down in electrode density from 5.0 3.5 to 5.0 7.1 mm. There was no systematic change in the prevalence of disorganized activity, of focal activations, or of single linear wavefronts as a result of varying bipole location density (Table 2). Specifically focusing on rotational activations in the 11 of 13 mapped regions in which rotational activation patterns were observed, there was a significant reduction in the frequency of rotational activity as bipole density declined, from a median prevalence of 7.4% (IQR 4.3%–15.1%) with schema 1 to a median prevalence of 0% (IQR 0%–1.5%) with the application of the least dense schema (P ¼ .04). Again there was a marked step-down in the observed prevalence of rotational activity between a bipole location density of 5.0 3.5 mm and a density of 5.0 7.1 mm (6.9% [IQR 6.1%–13.2%] to 2.5% [IQR 0%–4.1%]).
Discussion The key finding of this study is that the atrial activation patterns observed in human persistent AF when applying wavefront mapping on the basis of the timing of discrete bipolar EGMs are dependent on the density of the recording electrodes. In particular, low density of bipolar electrode locations overstates the prevalence of simple broad linear wavefront activation, understates the prevalence of complex activation patterns, and understates the prevalence of rotational activations. The largest difference occurs with a stepdown from an interbipole spacing of 5.0 3.5 mm to a spacing of 5.0 7.1 mm. Multiple groups have used high-density epicardial mapping plaques to study atrial activation patterns in human AF.3–8 Such mapping is spatially limited but uses mapping systems with small interelectrode distances of 1.2–2.5 mm to
Figure 2 Distribution of activation pattern through the application of the most dense schema of bipolar electrode locations (2.5 3.5 mm). A: Single narrow linear wavefronts are grouped together and single linear broad wavefronts are grouped together regardless of the direction of activation of the mapped region. B: Multiple narrow linear wavefronts, multiple broad linear wavefronts, and the simultaneous presence of narrow linear and broad linear wavefronts are grouped together as complex activations.
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
Walters et al 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473
P R I N T & W E B 4 C / F P O
Effect of Electrode Density in AF Mapping
5 with 64 unipolar electrodes mounted on 8 splines.9,10 The cardinal difference in the output of these studies has been the finding of a high prevalence of temporally stable rotors, the mechanistic significance of which has been indicated by catheter ablation studies that have demonstrated enhanced medium-term efficacy when these rotors are targeted, in comparison to a standard approach of pulmonary vein isolation alone.10,13,14 It may be that this relates to the critical differences in methodology, which are the markedly lower electrode density and the application of phase mapping based on Hilbert transformation of the voltage signal to derive a phase representation of the voltage amplitude over time. On the basis of theoretical consideration of action potential duration and conduction velocity restitution curves, which suggests a minimal wavelength of reentry of 4–5 cm in human atria, Narayan and coworkers15,16 have argued that while the spatial resolution of a multielectrode basket catheter might not be expected to resolve the rotor core at the tip of a spiral wave, it is sufficient to resolve the rotating arms of the spiral wave emanating from this core.12 In a computational model of a temporally stable rotor with a wavelength of 5 cm and a coherent field spanning the entire field of view, the distance between locations at which virtual EGMs were sampled was progressively increased, moving from a spatial resolution of 2.5 mm to one of 12.5 mm. While at the lower resolution the rotor core could not be mapped and there was some distortion in the appearance of rotation, a rotational pattern could be discerned.12 Using another computational model, however, with a wavelength between successive spiral arms of 3.2 cm and a coherent domain of 6 cm beyond which there was degeneration into fibrillatory conduction, there was marked deterioration in the ability to discern rotational activity when moving from a spatial resolution of 6 mm to one of 12.5 mm. This is consistent with the observation in this study of progressive insensitivity to the presence of rotational activity between an interelectrode spacing of 2.5 3.5 mm and a spacing of 7.5 10.6 mm. The phase analysis of activation during analysis of AF was initially reported by Chen et al17 in high-resolution optical mapping studies of sheep atria. Initially, a lag-return method was used to determine the phase of cyclical optical
474 475 476 477 478 479 480 481 482 483 484 485 486 487 Figure 3 Output of the linear mixed model evaluating the prevalence of 488 each activation pattern type as a percentage of total activations with the 489 application of each of the 4 bipole densities. The frequency of activation 490 patterns remains little changed with the application of schemas 1 (2.5 3.5 491 mm) and 2 (5.0 3.5 mm). Single broad wavefront activations are significantly more prevalent with the application of the less dense schemas 3 and 4, while the 492 frequency of complex activation patterns is significantly lower with the 493 application of these sparse electrode arrangements (see details in text). 494 495 record unipolar3–5,8 or bipolar6–8 EGMs. A common feature 496 to all such studies is that analysis of atrial activation patterns 497 has been based on accurate identification of the timing of 498 discrete atrial EGMs. Taken together, these studies have 499 suggested that AF is maintained by multiple wavelet 500 propagation, with dissociated wavefronts in the remodeled 501 atrium separated by regions of fibrosis and poor tissue 502 coupling.4 They have suggested that dissociation between 503 epicardial and endocardial layers allows “endo-epi” wave504 front breakthrough that may aid perpetuation of fibrillation.5 505 While high-frequency foci have been described as a potential 506 sustaining source of fibrillation,8 rotational activity has been 507 relatively infrequent and transient.6,7 An inherent limitation 508 to such studies is that the mechanistic significance of such 509 patterns has not been tested by ablation. 510 An alternative approach to mapping human AF has 511 involved the application of more global but lower-density 512 recording systems, in particular multispline basket catheters. 513 These catheters, with an interelectrode distance of o4 mm at 514 the poles and o1 cm at the equator12 when deployed 515 optimally, aim to map the entire atrial endocardial surface 516 517 Table 2 Percentage of all atrial activations represented by each pattern, with the 4 different electrode density schemas sequentially applied 518 519 to the 10-s atrial fibrillation segments from each of the 13 mapped regions* 520 Schema 1 Schema 2 Schema 3 Schema 4 521 Activation pattern (2.5 3.5 mm) (5.0 3.5 mm) (5.0 7.1 mm) (7.5 10.6 mm) P 522 Disorganized 1.0 (0–14.0) 2.8 (0–14.5) 7.2 (0–11.9) 15.7 (2.5–21.0) .80 523 Rotational 5.0 (0.9–12.1) 6.1 (1.0–10.5) 0.9 (0–3.7) 0 (0–1.5) .03 Q8524 Focal 1.8 (0–4.2) 2.0 (0–3.6) 2.6 (1.0–9.3) 6.1 (1.7–10.0) .80 525 Single narrow linear wavefront 3.0 (0–8.1) 4.1 (0–8.3) 5.3 (0–12.5) 2.8 (0–18.9) 1.00 Single broad linear wavefront 11.4 (5.8–31.4) 22.8 (9.5–36.3) 48.2 (40.1–72.4) 54.5 (37.4–80.6) .0003 526 Complex 53.1 (41.5–56.8) 52.0 (35.6–53.1) 20.2 (11.8–25.8) 7.2 (1.0–12.1) .02 527 528 With decreasing density of bipole locations, a marked fall in the prevalence of complex and rotational activation was observed, offset by a significant 529 increase in the prevalence of simple broad linear wavefront activation. * The median and interquartile range are calculated from the percentages determined for each mapped region. 530
6 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
action potentials, and this was subsequently refined through the use of the Hilbert transform to derive the signal phase from the voltage-time series.18,19 Application of this methodology to the analysis of markedly lower-density extracellular EGMs may exacerbate the effect of low-density mapping on observed atrial activation patterns, as the interpolation of the Hilbert-based phase at dispersed electrodes, which may by chance be activated sequentially by unrelated wavefronts, carries an inherent bias toward the demonstration of rotational activity in the area between the electrodes.20 The spatial resolution of schema 1 in this study, as in previous epicardial mapping studies, was sufficient to demonstrate a high prevalence of complex activation patterns that often included the simultaneous presence of multiple discrete narrow wavefronts. At the resolution of schema 4, complex patterns were much less frequently discerned, with broad linear wavefront becoming dominant. It is possible that the application of other mapping algorithms to these low-density recordings might have demonstrated more rotational activity.
Study limitations This study has examined only the effect of reducing electrode density on one particular wavefront mapping system on the basis of accurate determination of the timing of discrete atrial EGMs. Likewise, we have studied only the results of highdensity but spatially limited epicardial mapping and have also not examined more global endocardial recordings.
Conclusion Atrial activation patterns observed in epicardial mapping of human persistent AF vary significantly with the density of the recording bipolar electrodes. Low density understates the prevalence of complex and rotational activation patterns. The largest change occurs with a step-down in spacing between bipole locations from 5.0 3.5 to 5.0 7.1 mm. This may have important implications for the design of mapping technologies.
References 1. Allessie M, De Groot N. CrossTalk opposing view: rotors have not been demonstrated to be the drivers of atrial fibrillation. J Physiol 2014;592: 3167–3170. 2. Narayan SM, Jalife J. CrossTalk proposal: rotors have been demonstrated to drive human atrial fibrillation. J Physiol 2014;592:3163–3166. 3. Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. Highdensity mapping of electrically induced atrial fibrillation in humans. Circulation 1994;89:1665–1680.
Heart Rhythm, Vol 0, No 0, Month 2016 4. Allessie MA, de Groot NMS, Houben RPM, Schotten U, Boersma E, Smeets JL, Crijns HJ. Electropathological substrate of long-standing persistent atrial fibrillation in patients with structural heart disease: longitudinal dissociation. Circ Arrhythm Electrophysiol 2010;3:606–615. 5. de Groot NMS, Houben RPM, Smeets JL, Boersma E, Schotten U, Schalij MJ, Crijns H, Allessie MA. Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough. Circulation 2010;122:1674–1682. 6. Lee G, Kumar S, Teh A, et al. Epicardial wave mapping in human long-lasting persistent atrial fibrillation: transient rotational circuits, complex wavefronts, and disorganized activity. Eur Heart J 2014;35:86–97. 7. Walters TE, Lee G, Morris G, et al. Temporal stability of rotors and atrial activation patterns in persistent human atrial fibrillation. JACC Clin Electrophysiol 2015;1:14–24. 8. Lee S, Sahadevan J, Khrestian CM, Cakulev I, Markowitz A, Waldo AL. Simultaneous bi-atrial high density (510-512 electrodes) epicardial mapping of persistent and long-standing persistent atrial fibrillation in patients: new insights into the mechanism of its maintenance. Circulation 2015;132:2108–2117. 9. Narayan SM, Krummen DE, Rappel W-J. Clinical mapping approach to diagnose electrical rotors and focal impulse sources for human atrial fibrillation. J Cardiovasc Electrophysiol 2012;23:447–454. 10. Narayan SM, Krummen DE, Shivkumar K, Clopton P, Rappel W-J, Miller JM. Treatment of atrial fibrillation by the ablation of localized sources: CONFIRM (Conventional Ablation for Atrial Fibrillation With or Without Focal Impulse and Rotor Modulation) trial. J Am Coll Cardiol 2012;60:628–636. 11. Narayan SM, Shivkumar K, Krummen DE, Miller JM, Rappel W-J. Panoramic electrophysiological mapping but not electrogram morphology identifies stable sources for human atrial fibrillation: stable atrial fibrillation rotors and focal sources relate poorly to fractionated electrograms. Circ Arrhythm Electrophysiol 2013;6:58–67. 12. Rappel W-J, Narayan SM. Theoretical considerations for mapping activation in human cardiac fibrillation. Chaos 2013;23:023113. 13. Narayan SM, Baykaner T, Clopton P, Schricker A, Lalani GG, Krummen DE, Shivkumar K, Miller JM. Ablation of rotor and focal sources reduces late recurrence of atrial fibrillation compared with trigger ablation alone: extended follow-up of the CONFIRM trial (Conventional Ablation for Atrial Fibrillation With or Without Focal Impulse and Rotor Modulation). J Am Coll Cardiol 2014;63:1761–1768. 14. Miller JM, Kowal RC, Swarup V, et al. Initial independent outcomes from focal impulse and rotor modulation ablation for atrial fibrillation: multicenter FIRM registry. J Cardiovasc Electrophysiol 2014;25:921–929. 15. Lalani GG, Schricker A, Gibson M, Rostamian A, Krummen DE, Narayan SM. Atrial conduction slows immediately before the onset of human atrial fibrillation: a bi-atrial contact mapping study of transitions to atrial fibrillation. J Am Coll Cardiol 2012;59:595–606. 16. Narayan SM, Kazi D, Krummen DE, Rappel W-J. Repolarization and activation restitution near human pulmonary veins and atrial fibrillation initiation: a mechanism for the initiation of atrial fibrillation by premature beats. J Am Coll Cardiol 2008;52:1222–1230. 17. Chen J, Mandapati R, Berenfeld O, Skanes AC, Gray RA, Jalife J. Dynamics of wavelets and their role in atrial fibrillation in the isolated sheep heart. Cardiovasc Res 2000;48:220–232. 18. Bray M-A, Wikswo JP. Considerations in phase plane analysis for non-stationary reentrant cardiac behavior. Phys Rev E 2002;65:051902. 19. Umapathy K, Nair K, Masse S, Krishnan S, Rogers J, Nash MP, Nanthakumar K. Phase mapping of cardiac fibrillation. Circ Arrhythm Electrophysiol 2010;3: 105–114. 20. Berenfeld O, Oral H. The quest for rotors in atrial fibrillation: different nets catch different fishes. Heart Rhythm 2012;9:1440–1441.
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621