Radiofrequency catheter ablation of atrial fibrillation: Electrical modification suggesting transmurality is faster achieved with remote magnetic catheter in comparison with contact force use

Background: Remote magnetic navigation (RMN) and contact force (CF) sensing catheters are available technologies for radiofrequency (RF) catheter ablation of atrial fibrillation (AF). Our purpose was to compare time to electrogram (EGM) modification suggesting transmural lesions between RMN and CF‐guided AF ablation. Methods and results: A total of 1,008 RF applications were analyzed in 21 patients undergoing RMN (n = 11) or CF‐guided ablation (n = 10) for paroxysmal AF. All procedures were performed in sinus rhythm during general anesthesia. Time to EGM modification was measured until transmurality criteria were fulfilled: (1) complete disappearance of R if initial QR morphology; (2) diminution > 75% of R if initial QRS morphology; (3) complete disappearance of R’ of initial RSR’ morphology. Impedance drop as well as force time integral (FTI) were also assessed for each application. Mean CF at the beginning of each RF application in the CF group was 11 ± 2 g and mean FTI per application was 488 ± 163 gs. Time to EGM modification was significantly shorter in the RMN group (4.52 ± 0.1 seconds vs. 5.6 ± 0.09 seconds; P < 0.00001). There was no significant difference between other procedural parameters. Conclusion: Remote magnetic AF ablation is associated with faster EGM modification suggesting transmurality than optimized CF and FTI‐guided catheter ablation.

atrial fibrillation; contact force; remote magnetic navigation; transmurality

The procedural success of radiofrequency (RF) paroxysmal atrial fibrillation (AF) ablation is associated with complete and perennial pulmonary veins (PVs) isolation by adjacent transmural lesion creation. After AF ablation, PV reconnections are mainly the fact of non‐transmural lesion creating gaps of conduction in the lines, responsible for clinical recurrences.[1] Transmurality obtention is multifactorial, and may be predicted by electrograms (EGMs) modification.[2]

There is now overwhelming proof that electrode–tissue contact force (CF) during RF application is a major determinant of RF lesions size.[3] Using CF has been shown to decrease ablation, procedural time, and fluoroscopy time, but also to reduce clinical recurrences.[4] , [5] , [6] , [7] A force time integral (FTI) between 392 and 404 gs and impedance drop at 8 or 10 ohms have been correlated transmurality obtention.[8] , [9] , [10] Remote magnetic navigation (RMN) is an alternative technology for RF ablation.[11] , [12] Comparatives studies between RMN and conventional (without CF) AF ablation showed a noninferiority efficiency in terms of clinical recurrences, but a better profile of safety and an undeniable comfort for the operator.[13]

The purpose of our study was to compare transmurality obtention (based upon EGM modification) between RMN and optimized CF‐guided AF ablation. We also analyzed the time to obtain EGM modification (from the beginning of RF energy), according to different parameters: the RF delivery sites, the impedance drop, and the FTI in CF group.

Twenty‐one consecutive patients undergoing RF catheter ablation for symptomatic drug‐refractory paroxysmal AF, at the Princess Grace Hospital (Monaco), between June 2013 and June 2014, were included in this study from a population of 50. Only patients who were in sinus rhythm the day of the procedure and performed under general anesthesia were selected for this study.[14] Exclusion criteria were hyperthyroidism, left atrial thrombus, decompensated heart failure, stroke, myocardial infarction, or gastrointestinal bleeding within 4 weeks prior to the intervention, and life expectancy < 6 months. The study was approved by the institutional ethical local committee on human research. Written informed consent for the procedure was obtained for all patients.

Procedures were performed with vitamin K antagonist continuation with a target INR of 2.0–3.0. All catheters were advanced via the femoral vein under ultrasound guidance.[15] Intravenous heparin was administered immediately after the insertion of the femoral sheaths, and continuously given to maintain an activated clotting time of 300–350 seconds throughout the procedure. A 6F steerable decapolar catheter (Irvine Bio Inc., St. Jude Medical, Inc., St. Paul, MN, USA) was positioned in the coronary sinus. After left atrial thrombus exclusion and transesophageal‐guided double transseptal puncture, a circular mapping catheter within a sheath was advanced into the left atrium (Lasso and Preface, Biosense Webster, Diamond Bar, CA, USA). In the RMN group, a 3.5‐mm open‐irrigated, magnetic ablation catheter (Navistar Thermocool RMT, Biosense Webster, Diamond Bar) was then advanced most of the time through a SL1 sheath into the left atrium (St. Jude Medical, Inc.), or within a steerable robotic sheath (V‐Cas Deflect, Stereotaxis).[16] In the CF group, an irrigated CF sensing catheter (Thermocool SmartTouch, Biosense Webster) was introduced either via an SL1 sheath into the left atrium, or within a steerable sheath (Agilis, St. Jude Medical, Inc.). Circumferential PV ablation was performed using a three‐dimensional mapping system (Carto 3, Biosense Webster) in conjunction with the integrated CT image of the left atrium and real‐time fluoroscopy. The RF generator (Stockert, Biosense Webster) was set to temperature‐controlled RF delivery with a maximum temperature of 42 °C and a nominal power limit of 40 W (flow 30 mL/minute) and 30 W (flow 17 mL/ minute) at the posterior left atrial wall. RF current was applied until local bipolar EGM modification became suggestive of transmural lesion (see below), and until the appearance of a transmural lesion as revealed by the Visitag™ software in both groups (with criteria of FTI > 400 gs in the CF group, and impedance drop > 8 ohms in the RMN group) (Fig. [NaN] ).[17] Each ablation was performed in a point‐by‐point manner aiming a force > 10 g. An esophageal probe was inserted for luminal esophageal temperature monitoring and its height was constantly adjusted under fluoroscopy in order to be as close as possible to the position of the catheter tip.

Electrical isolation of all PVs defined as entrance conduction block was proven by the circular mapping catheter. Adenosine was also injected to search any PV dormant conduction at the end of the PVI. Procedures were all performed by three experienced electrophysiologists (N.S., D.‐G.L., and S.‐S.B.).

EGMs were acquired on a digital electrophysiological recording system (Prucka Engineering, Inc., Houston, TX), with a sampling rate of 1 kHz. Bipolar signals were recorded with a 30–500 Hz filtering and a 50 Hz notch pass to avoid artifacts. Morphological analysis and measurements of the bipolar EGM on the distal poles of the ablation catheter were performed after the procedure. Amplification and speed of display (100–200 mm/s) were optimized. Poor‐quality EGMs, mainly due to artifacts, were excluded from the final analysis. The first two major positive deflections were defined as R and R’ and the first and second major negative deflections were defined as Q and S, respectively. After the procedure, for each RF pulse delivery (Visitag™), the corresponding EGMs were analyzed offline, comparing the EGM morphology immediately before the RF is applied, until EGM modification (suggesting transmurality) during ongoing RF delivery. Their predominant morphology was classified into one of the following patterns: QS, QR, QRS, RS, or RSR’. Depending on initial distal bipolar EGM morphology, transmurality was defined by the occurrence during RF and the persistence after RF of one of the following EGM criteria: (1) complete disappearance of the positivity when there was initial QR morphology, (2) diminution > 75% of the positivity when there was initial QRS morphology, or (3) complete disappearance of the R’ positivity when there was initial RSR’ morphology (Fig. [NaN] ). Peak‐to‐peak amplitude, positive and negative amplitude, and EGM morphology before RF at the obtention of transmurality criteria and at the end of RF delivery were also recorded. Time to EGM modification suggesting transmurality, from the beginning of each RF application, was measured in both groups, in an offline analysis after the procedure using the Odyssey™ video recording system (Stereotaxis).

All patients were monitored in the hospital for at least 12 hours. On the day after the procedure, a 12‐lead surface ECG was acquired to confirm normal sinus rhythm. Previous antiarrhythmic drugs were usually pursued for at least 1 month after ablation. Oral anticoagulation was continued after the procedure. After hospital discharge, all patients were followed in an outpatient clinic every 3 months for at least 1 year. At each visit, subjects were questioned for symptoms and documented arrhythmia recurrences, and current medication was assessed, including a 24‐hour Holter recording. A documented AF or atrial tachycardia (AT) episode lasting longer than 30 seconds outside a blanking period of 3 months after the index procedure was considered as recurrent AT/AF.

The statistical analysis was made with GraphPad Prism 5 (San Diego, CA, USA). Numerical variables are expressed as mean ± SD. Comparison of continuous variables utilized an unpaired Student's t‐test. Categorical data were compared by the χ² test. A two‐tailed P value < 0.05 was considered statistically significant.

Twenty‐one patients (11 in RMN group and 10 in CF group) were included. Patients’ clinical and anatomical characteristics were comparable between both groups, except for the presence of stroke, congestive heart failure, and coronary artery disease (Table [NaN] ). The number of isolated ipsilateral circles after first pass RF delivery was 15/22 in the RMN group versus 14/20 in the CF group (P = 0.89). A carina (inter‐PV) line was then performed in case of nonisolated circles allowing to acutely achieve bidirectional block at the PV‐antrum level in all patients in both groups. The number of reconnected veins after the 30‐minute waiting time needing additional RF delivery was 5/44 in the RMN group versus 3/40 in the CF group (P = 0.55).

Patients characteristics and procedural parameters

1 *RMN: remote magnetic navigation group.

2 +CF: contact force group.

A total of 1,117 point‐by‐point left atrial RF application were performed (554 in RMN group and 563 in CF group). Sixty‐eight RF applications were excluded because of very fragmented or low‐amplitude potentials, and as a consequence, EGM analysis could not be reliably performed. We also excluded from the final analysis 24 sites with QS or RS initial morphology suggesting preexisting scar. In addition, 17 RF applications were not included due to impedance rise revealing a catheter instability. The final analysis was performed on 1,008 RF applications (respectively, 506 in the RMN group and 502 in the CF group) (Fig. [NaN] ).

Initial EGM morphology at the beginning of each RF application was, respectively, in RMN and CF group: QRS (50.5%; 53%), RSR’ (35.8%; 36.1%), and QR (15%; 10.9%) (Fig. [NaN] ). The distribution of EGM morphologies before ablation was comparable between the two groups, suggesting a similar orientation of the catheters (either RMN or CF) and tissue contact (Figs. [NaN] , [NaN] ).

Mean CF at the beginning of RF delivery was 11 ± 2 g. Regardless of initial EGM morphology, at the time to obtaining transmural lesion (Fig. [NaN] ), peak‐to‐peak EGM amplitude reduction was similar in both groups (45% in RMN and 43% in CF group). Time to EGM modification suggesting transmurality was significantly shorter in the RMN group (4.52 ± 0.1 seconds vs. 5.6 ± 0.09 seconds; P < 0.00001); this difference was also significant for each of the RF application sites (Fig. [NaN] A).

There was no significant difference about impedance drop (13.59 ± 6.3 ohms in RMN group vs. 13.02 ± 5.9 ohms in CF group, P = 0.13). In both groups, the impedance drop tended to be higher in the left PV, when compared to the right PV, but without reaching statistical significance: 13.8 ± 6.7 versus 13.3 ± 5.8 for the right PV in the RMN group (P = 0.46); 13.2 ± 5.7 versus 12.7 ± 6.08 for the right PV in the CF group (P = 0.32) (Fig. [NaN] B).

Mean FTI was 488 ± 163 gs. FTI in left PV tended to be more important (500.9 ± 151 gs) versus 476 ± 173 gs in the right PV (P = 0.09) (Fig. [NaN] C).

Impedance drop was poorly but significantly correlated with FTI in CF group (r = 0.29; P < 0.0001) (Fig. [NaN] A) and with time to EGM modification suggesting transmurality (r = −0.07; P = 0.02) (Fig. [NaN] B). No correlation was found between FTI and time to EGM modification.

In the right posteroinferior PV, there was a lower FTI and impedance drop; also, the time to EGM modification was longer.

After a mean follow‐up of 21 ± 2 months, 2 patients experienced a recurrence in the CF group (respectively, 91 and 270 days after the index procedure), versus only 1 in the RMN group 380 days after the initial procedure (P = 0.047). The 3 patients all benefited from a repeat procedure consisting in a successful PV reisolation.

To the best or our knowledge, this is the first study to investigate the time to reach the electrical transmurality criteria based on bipolar EGM modifications during RF application for AF ablation. Our study shows that the time to obtain EGM modification as a surrogate for electrical transmurality is shorter with RMN technology. Despite similar power programming, RF application efficiency was obtained faster with RMN system in comparison with an open‐irrigated CF‐sensing catheter with optimized CF (mean CF 11 ± 2 g at the beginning of RF delivery and with mean FTI of 488 ± 163 gs). That could be explained by a better catheter stability with magnetic navigation. This one second difference for EGM modification between the groups cannot be extrapolated into clinical practice, and the study was not designed to compare clinical efficacy between the two technologies based upon EGM analysis. EGM modification during RF application is one electrical criterion serving as a surrogate for transmurality, among several others, but for this electrical transmurality to be perennial, the operator must prolong the RF application, and FTI (and eventually FTI with power integration called FTPI) cutoff is another criterion that can be used, as well as impedance drop.

A similar distribution of EGM morphologies between the two groups was found before RF applications, suggesting a similar orientation of the catheters on the tissue with both technologies.[18] Our results are in line with a recent comparative study between RMN and CF‐guided AF ablation.[19] Natale et al. emphasized the importance of stable CF (variation < 20%) in reducing clinical recurrences at 1 year.[20]

Impedance drop was relatively similar (13.59 ± 6.3 ohms in RMN group and 13.02 ± 5.9 in CF group; P = 0.13). The impedance drop was measured between the time just before RF application and the time to achieve EGM transmurality criteria (impedance drop was thus measured at mean 4.52 and 5.6 seconds, respectively, in RMN and CF group). These results are in agreement with Anter et al. who reported that an impedance drop > 8 ohms or > 5% was associated with a decrease of premature venous reconnection rate and clinical recurrences at 6 months. Reichlin et al. proposed a 10‐ohm value of impedance drop with Smartouch® catheter and an 8 ohm value with Tacticath® catheter as an indicator of good contact.[10] In our study, a good correlation between impedance drop and FTI was found in CF group (r = 0.29; P < 0.0001), as reported by Ullah et al. and Makimoto et al. (mean CF and impedance drop).[21] , [22]

The small number of patients and lack of randomization are the most important limitations of this study. This is nevertheless compensated by the high number of RF applications analyzed in each group, yielding highly significant main results. Moreover, clinical recurrences investigation was not our aim but rather the investigation of RF application efficiency according to EGM criteria of transmurality.

In our study, EGMs were analyzed only in bipolar mode. Though, EGM transmurality criteria in unipolar mode have been reported to be also reliable markers: complete disappearance of negativity on distal unipolar EGM after RF application allowed to predict transmurality with sensibility (Se) and specificity (Sp) of 100% in Otomo et al. study. In this same study, bipolar EGM analysis predicted transmurality with Se and Sp of 100% if RSR’ and QR initial morphology and with 85% of Se and 95% of Sp if QRS initial morphology.[2] However, distal unipolar EGM during RF application presented a background noise making difficult a reliable analysis of EGM modification during RF delivery in our laboratory. Noteworthily, unipolar modification, when present, during RF delivery is obtained later in comparison with bipolar EGM modification (15 ± 1 seconds) in a previous study.[23]

During RF delivery, EGM modification (used as a surrogate for electrical transmurality) was faster obtained with remote magnetic AF ablation in comparison to optimized CF‐guided catheter ablation. EGM morphology analyses in both groups before RF delivery confirmed a similar orientation of both catheters (RMN and CF) on the tissue.

Graph: Examples of annotation algorithm in the contact force group using the force time integral values (left image) and in the remote magnetic group using impedance drop cutoff

Graph: image%5ft/jce13222-fig-0001-t.gif

Graph: Distal bipolar electrogram criteria of transmurality after radiofrequency, depending on initial morphology (horizontal arrows). Examples of recorded electrograms before and after radiofrequency, with criteria of transmurality met [Color figure can be viewed at ]

Graph: image%5ft/jce13222-fig-0002-t.gif

Graph: Flowchart of the total number of radiofrequency deliveries analyzed, in both groups (remote magnetic navigation or contact force sensing catheters), and kept for final analysis after exclusion

Graph: image%5ft/jce13222-fig-0003-t.gif

Graph: (A) Catheter orientation relative to the endocardial surface (depending on the respective contact of the distal and immediately proximal electrodes) impacts the morphology of the bipolar electrograms. (B) Distribution of the electrograms morphologies before ablation was similar between the groups [Color figure can be viewed at ]

Graph: image%5ft/jce13222-fig-0004-t.gif

Graph: Example of “beat‐to‐beat” electrogram modification suggesting transmurality, during ablation, with a remote magnetic catheter, with obtention of a QS morphology after 1,200 milliseconds of radiofrequency delivery [Color figure can be viewed at ]

Graph: image%5ft/jce13222-fig-0005-t.gif

Graph: (A) Mean time (in seconds) for achieving electrogram modification suggesting transmurality in both groups, according to the different locations around the pulmonary veins. (B) Mean impedance drop in both groups according to the different locations around the pulmonary veins. (C) Mean force time integral (FTI) in the contact force group only obtained according to the different locations around the pulmonary veins [Color figure can be viewed at ]

Graph: image%5ft/jce13222-fig-0006-t.gif

Graph: (A) Correlation between impedance drop and force time integral (FTI) in the contact force group only. (B) Correlation between impedance drop and time to electrogram modification suggesting transmurality (seconds) in both groups [Color figure can be viewed at ]

Graph: image%5ft/jce13222-fig-0007-t.gif