Volume 14 – Issue 5 – May 2014 Posted on: 4/23/14 0 Comments 5033 reads
David E. Krummen1, MD; Vijay Swarup2, MD; James Daubert3, MD; John Hummel4, MD; Gery Tomassoni5, MD; John M. Miller6, MD; Sanjiv M. Narayan1,7, MD PhD; 1University of California Medical Center, San Diego, California; 2Arizona Heart Hospital, Phoenix, Arizona; 3Duke University, Durham, North Carolina; 4Ohio State University, Columbus, Ohio; 5Central Baptist Hospital, Lexington, Kentucky; 6Indiana University, Indianapolis, Indiana; 7University of California Medical Center, Los Angeles, California
There is growing interest in the substrates that sustain atrial fibrillation (AF) after it has been triggered. Several groups now report that the electrophysiological substrate for AF includes rotational circuits or focal beats, whose elimination can improve long-term AF freedom compared to trigger ablation alone. However, a major source of confusion is why traditional mapping in AF does not identify rotational activity.
The answer is physiological: AF rotors differ fundamentally from macro-reentry around an obstacle. In macro-reentry, the primary mechanism (and ablation target) is the rotational circuit around the inert obstacle. Conversely, in an AF rotor, the primary mechanism is its core (phase singularity) that precesses (‘wobbles’) in small areas with emanating spiral arms that disorganize and fuse with surrounding activation (fibrillatory conduction). Simple activation or phase mapping are not ideal for this physiology, and so approaches using repolarization and conduction dynamics to reduce noise (e.g., Focal Impulse and Rotor Mapping, FIRM) have been developed to track AF rotors. FIRM-guided ablation targets the primary mechanism (rotor core or focal source origin) rather than emanating spiral arms. The endpoint is to ablate the 2-3 rotors present per AF patient, reconfirming elimination by repeat FIRM mapping. In independent laboratories, FIRM-guided ablation improves long-term freedom from paroxysmal and persistent AF versus trigger ablation alone. Multicenter randomized clinical trials are underway.
Keywords: Atrial Fibrillation, Spiral Wave, Rotor, Fibrillatory Conduction, Paroxysmal, Human, Ablation, Focal Source
There is increasing focus on ablating the substrates that maintain atrial fibrillation (AF) after it has been triggered.1 Structural substrate includes atrial fibrosis suggested by low voltage2 or gadolinium-enhanced magnetic resonance imaging.3 Electrophysiologically, several groups now report regions of rotational activation as drivers for AF. Targeting of rotors by Focal Impulse and Rotor Modulation (FIRM) has been shown to improve the long-term success of AF ablation compared to trigger ablation alone in multicenter studies.4,5 However, while FIRM and other approaches target rotors that lie in spatially reproducible regions for tens of minutes6-8 or even days,9,10 some methods reveal transient or no rotational activity.11-13
This brief review discusses some of the historical challenges in detecting AF rotors, then discusses the clinical method and pitfalls of FIRM-guided ablation, and summarizes long-term results.
WHY A ROTOR IS NOT MACRO-REENTRY
Electrophysiologists at all levels of training are familiar with macro-reentry, in which a stable circuit rotates around an anatomical obstacle. It is thus natural to extend this concept to an AF rotor, as revealed by two decades of intense investigation by Jalife and others14 and recently detected in patients by FIRM mapping (Figure 1).
Despite the superficial similarities, however, an AF rotor is quite different from macro-reentry. In macro-reentry, the rotating circuit is the primary mechanism around an ‘inert’ obstacle that can be mapped (using point-by-point mapping15 or even FIRM16). In AF, the reverse is true with the rotor core being the principle mechanism and the emanating spiral waves rapidly disorganizing via wavebreak and collision with the milieu14 — this fibrillatory conduction is a principal component of the localized source hypothesis for AF. Accordingly, while stable rotational activation in macro-reentry is easily traced by mapping activation sequence, rotors in AF have for decades been obscure using these approaches.17 The factors enabling rotor formation likely reflect changes in cellular ion channel function and cytosolic calcium dynamics, changes in cell-to-cell conduction and in the structural matrix consistent with remodeling.14
Figure 1 indicates a left atrial AF rotor in a 63-year-old gentleman. The isochronal (contour) map in Figure 1A indicates a clockwise spiral wave (arrow) for one cycle around the rotor (labeled). As indicated, the spiral wave collides and fuses with fibrillatory waves (arrows and double lines) during each cycle, producing variability (contributing to the disorganized appearance of AF). Thus, while electrograms at electrodes E3 to F3 show rotation for this cycle (Figure 1B), this may vary for each subsequent cycle. In addition, the rotor core precesses (‘wobbles’) over time, as shown by Davidenko et al in 199218 and quantified in patients by FIRM,9,19 that ‘drags’ spiral waves such that rotation may also precess away from E3 to F3. Accordingly, AF rotor videos (see online supplement in the published report of this case20) are used to guide clinical ablation, rather than the single snapshots used in illustrations.
Figure 2 provides more detail on rotor precession, in an 81-year-old gentleman with persistent AF mapped using FIRM. At first, one may expect the clockwise rotor in the inferior left atrium to exhibit consistent rotational activity on electrodes 1-8. However, Figure 2B shows the precession locus of the rotor core over 900 ms (≈4 cycles), that in general occupies areas of 2-3 cm2.19 This is sufficient, independent of fibrillatory conduction, to obscure rotor detection and cause irregular electrograms in experimental studies.21 Figure 2C details precession in one single AF cycle (190 ms), in which the rotor core starts at point α then passes through point β to γ . This locus falls outside the electrode sites 1-8, so that to track rotation one would have to know this trajectory in advance then ‘drag’ the array across it within 190 ms. As the core traverses this trajectory, emanating spiral waves are subject to wavebreak from repolarization and conduction dynamics,22 collision with external fibrillatory waves, Doppler effects18 and other structural factors. FIRM can identify AF rotors in this complex milieu by combining phase mapping with noise-reduction algorithms including rate-response of atrial (left or right) repolarization22-26 and conduction26,27 dynamics, and oscillations in AF rate.28
As mechanistic proof-of-concept that FIRM-mapped rotors are principal AF mechanisms in patients, the multicenter PRECISE study showed that elimination of FIRM-mapped rotors alone (without PV isolation) was able to eliminate paroxysmal AF on follow-up of patients undergoing ablation for the first time.29 Electrograms on the ablation catheter at FIRM-guided sites vary in their characteristics, and are often high amplitude and not predicted by fractionation.19 If signals are poor, basket repositioning should be attempted (see below).
FOCAL IMPULSE AND ROTOR MAPPING (FIRM): TECHNICAL APPROACH AND LIMITATIONS
Figure 3 summarizes the workflow for FIRM mapping and ablation. For macro-reentry, ablation would aim to bisect circus movement (the primary mechanism). In AF, therapy also targets the primary mechanism — but in this case, it is the rotor core. In the CONFIRM trial, 2.1 ± 1.0 sources were observed per patient, with typical total FIRM ablation time of 5-10 minutes per source, or 15-20 minutes total.4
FIRM uses wide-area mapping of both atria in AF,30 using commercially available basket catheters (Constellation™, Boston Scientific; or FIRMap™, Topera) to provide contact (i.e., not virtual) AF electrograms. AF recordings are exported for analysis to a computational system (RhythmView™, Topera) using the above algorithms to create diagnostic movies.30 Movies are interpreted by each operator to identify electrical rotors or focal sources in spatially reproducible locations, that have been observed in all paroxysmal AF and nearly all persistent AF patients (98% in CONFIRM,4 100% in recent external series6).
A practical limitation of FIRM-guided mapping and ablation is poor contact of the catheter with the atrial walls, as we have detailed previously.30 Figure 4A illustrates good basket coverage, producing high-confidence FIRM maps. Basket positioning is used to minimize electrodes overlying the superior and inferior mitral annulus (left atrium), or triscupid annulus (right atrium), that register ventricular (QRS) artifact. Moreover, AF rotors cannot lie in the valve orifice or at the atrial insertion of the valve where they will self-terminate due to proximity to a nonconducting boundary.14
Figures 4B and 4C show poor basket contact with the atrium, that produces low confidence FIRM maps. Such FIRM mapping is suboptimal, based upon clinical cases when the atria were larger than the largest available basket in CONFIRM4 and in external series by Shivkumar et al6 and Miller et al.5 Good results have still been reported when the rotor/focal source happens to lie at a site where the basket makes contact.6 In these cases, many electrodes away from this site may not be in contact, and basket repositioning is needed to map the entire chamber. Future improvements in basket sizing/conformity, particularly in the septal left atrium, may help to map AF in atria with very large diameters (>6 cm).
FIRM-GUIDED ABLATION: CLINICAL ENDPOINT AND LONG-TERM RESULTS
FIRM-guided ablation targets all rotors/sources in both atria. FIRM mapping is first performed in the right atrium, with the endpoint of elimination of all rotors/focal sources using immediate FIRM remapping. This currently takes <1 minute per map, whereas in early trials, remapping was not possible.4 Once all right atrial rotors/focal sources are eliminated, FIRM is repeated in the left atrium, again with the endpoint of eliminating all rotors and focal sources. Other ablation is then performed per operator judgment, including PV isolation in the CONFIRM trial4 and independent centers which
Rotor elimination as the primary endpoint of FIRM is supported by preliminary results of FIRM-only ablation in the multicenter PRECISE trial,29 and on-treatment analysis from the CONFIRM trial (Figure 5A). In this analysis (Figure 5A), patients in whom ablation lesions passed through rotors and focal sources, directly (FIRM-guided patients) or coincidentally (via anatomical lines/PVI), experienced most of the benefit of ablation (80% long-term AF freedom) while patients whose lesions missed all sources experienced little benefit (<20% freedom from AF).
During rotor elimination AF may terminate, predominantly to sinus rhythm. This was shown in the CONFIRM trial (Narayan et al4), in the first external series by Shivkumar at UCLA (Figure 5B) and others.6 In early cases, FIRM analysis took too long for practical remapping and so <10 minutes of ablation was applied to each source. Now that remapping is possible, FIRM ablation can stop once rotors/focal sources have been eliminated. Recent data shows that rotor elimination may be more predictive of long-term outcome than AF termination. Studies are needed to define the physiology of AF termination, particularly if it occurs after extensive ablation and/or to atrial tachycardia,31 and if this is equivalent to termination to sinus rhythm (Figure 4).
The CONFIRM trial (CONventional ablation with or without Focal Impulse and Rotor Modulation) was a prospective case cohort study4 enrolling 92 patients at 107 consecutive AF ablation procedures, of whom approximately two-thirds had nonparoxysmal AF. Single-procedure AF elimination in CONFIRM was higher for FIRM-guided than conventional FIRM-blinded cases (82.4% versus 44.9%;p<0.001) after 273 days (median; IQR 132-681). Very long-term follow-up of FIRM-guided ablation at three years in the CONFIRM trial are being collected and are promising.
Success may vary between centers predominantly due to patient selection, with lower success in redo (versus first-ablation) cases and in patients with atria larger than the largest basket size. Nevertheless, a substudy of the CONFIRM trial showed that FIRM-guided ablation retained its advantage over conventional ablation even in patients with typically unfavorable demographic factors such as sleep apnea.32
ROTOR MAPPING: FUTURE DIRECTIONS
FIRM-guided ablation has now been validated in many independent clinical laboratories in the U.S.5,6and Europe.33 Many groups are developing independent systems for rotor detection that should be compared in clinical outcome trials. Future improvements in basket design should help improve the technical performance of FIRM mapping. Clinically, the primary question is to confirm how FIRM-guided substrate ablation compares to trigger ablation in multicenter randomized trials. Such trials are underway. Mechanistically, studies are required to define how AF rotor cores or focal source origins precess, and spiral arms disorganize, in relation to the structural and electrophysiological milieu.
Rotors and focal sources are electrophysiological substrate sustaining AF in many patients, whose ablation can improve success over trigger-based ablation alone as shown now in laboratories in the U.S. and Europe. FIRM incorporates physiological algorithms (rate response of action potential duration and conduction) to track precessing rotor core(s) while simultaneously analyzing complex decay in emanating spiral arms from fibrillatory conduction. Ongoing clinical trials and mechanistic studies will further clarify the dynamics between AF rotors, fibrillatory dynamics and clinical outcomes.
David E. Krummen has received consulting fees from Insilicomed and research grants from the American Heart Association and NIH. He has also received fellowship support from Medtronic, St. Jude Medical, Biosense Webster, Boston Scientific and BIOTRONIK.
Vijay Swarup has received consulting fees/honoraria from Biosense Webster and research grants from Biosense Webster, Medtronic, Boston Scientific, St. Jude Medical and BIOTRONIK.
James Daubert has received consulting fees/honoraria from Medtronic, St. Jude Medical, Boston Scientific, Sorin Group and CardioFocus. He has received research grants from Boston Scientific, Biosense Webster, Medtronic and Gilead Sciences. He has also received fellowship support from Medtronic, Boston Scientific, BIOTRONIK, St. Jude Medical, Biosense Webster and Bard Electrophysiology.
John Hummel reports consulting fees/honoraria from Medtronic and research grants from St. Jude Medical and Boston Scientific.
Gery Tomassoni reports consulting fees/honoraria from Topera, Stereotaxis, Biosense Webster, St. Jude Medical, Boston Scientific, Pfizer, and AtriCure. He also serves as CMO of Stereotaxis.
John M. Miller reports consulting fees/honoraria from Biosense Webster, BIOTRONIK and Medtronic, and fellowship support from Medtronic, Boston Scientific, BIOTRONIK and Biosense Webster.
Sanjiv M. Narayan is supported by NIH (HL83359, HL103800). Dr. Narayan is co-author of intellectual property owned by the University of California Regents and licensed to Topera, Inc. Topera does not sponsor any research, including that presented here. Dr. Narayan holds equity in Topera, and reports having received honoraria from Medtronic, St. Jude Medical, BIOTRONIK and Boston Scientific. He has received consulting fees from the American College of Cardiology Foundation and Topera, and royalty income from UpToDate. His Institution has received fellowship support from Medtronic, St. Jude Medical, Biosense Webster, Boston Scientific and BIOTRONIK.