Author + information
- Received August 23, 2005
- Revision received November 9, 2005
- Accepted November 16, 2005
- Published online April 4, 2006.
- Carlo Pappone, MD, PhD⁎ (, )
- Gabriele Vicedomini, MD,
- Francesco Manguso, MD, PhD,
- Filippo Gugliotta, BE,
- Patrizio Mazzone, MD,
- Simone Gulletta, MD,
- Nicoleta Sora, MD,
- Simone Sala, MD,
- Alessandra Marzi, MD,
- Giuseppe Augello, MD,
- Laura Livolsi, MD,
- Andreina Santagostino, MD and
- Vincenzo Santinelli, MD ()
- ↵⁎Reprint requests and correspondence:
Dr. Carlo Pappone, Department of Arrhythmology, University Hospital San Raffaele, Via Olgettina, 60, 20132 Milan, Italy
Objectives We assessed feasibility of magnetic catheter guidance in patients with atrial fibrillation (AF) undergoing circumferential pulmonary vein ablation (CPVA).
Background No data are available on feasibility of remote navigation for AF ablation.
Methods Forty patients underwent CPVA for symptomatic AF using the NIOBE II remote magnetic system (Stereotaxis Inc., St. Louis, Missouri). Ablation was performed with a 4-mm tip, magnetic catheter (65°C, maximum 50 W, 15 s). The catheter tip was guided by a uniform magnetic field (0.08-T), and a motor drive (Cardiodrive unit, Stereotaxis Inc.). Left atrium maps were created using an integrated CARTO RMT system (Stereotaxis Inc.). End point of ablation was voltage abatement >90% of bipolar electrogram amplitude.
Results Remote ablation was successful in 38 of 40 patients without complications. The median mapping and ablation time was 152.5 min (range, 90 to 380 min) but was much longer in the first 12 patients (192.5 min vs. 148 min; p = 0.012). Median ablation time was 49.5 min (range, 17 to 154 min), but it was much shorter in the last 28 patients than in the first 12 patients (49 min vs. 70 min; p = 0.021). Patients receiving remote ablation had longer procedure times than control patients (p < 0.001) with similar mapping time but shorter ablation time on right-sided pulmonary veins. Many more mapping points regardless of their location were collected remotely (p < 0.001).
Conclusions Remote magnetic navigation for AF ablation is safe and feasible with a short learning curve. Although all procedures were performed by a highly experienced operator, remote AF ablation can be performed even by less experienced operators.
Catheter ablation techniques for atrial fibrillation (AF) are evolving with targets for radiofrequency (RF) ablation of AF increasingly being selected based on anatomic considerations. Such anatomic ablation techniques require precise catheter localization and stable contact during ablation. Even experienced electrophysiologists occasionally encounter difficulties in maintaining stable catheter contact, especially in some regions of complex anatomy where catheter stability is indeed crucial (1–5). Limitations of manually deflected conventional ablation catheters in performing complex catheter maneuvers may also contribute to these challenges. A recent feasibility study in animals suggests that continuous stability of ablation catheters can be obtained by the combination of the external magnetic field and flexibility of the distal portion of a magnetic catheter (6). Based on experimental and clinical data (6,7), we evaluated safety and feasibility of such navigation system in patients undergoing AF ablation. Here, we report the first human experience on magnetic remote navigation in 40 patients undergoing circumferential pulmonary vein ablation (CPVA) for AF.
This study included 40 patients (60% male patients; median age, 57 years, range, 28 to 75 years; Table 1)undergoing CPVA for drug-refractory symptomatic paroxysmal (62.5%) or permanent AF (median duration, 46.5 months; range, 12 to 286) with the use of the remote magnetic navigation NIOBE II system (Stereotaxis Inc., St. Louis, Missouri). Patients had no contraindications for magnetic navigation such as implanted cardioverter defibrillators, pacemakers, or vascular clips. Twenty-eight patients, matched for gender, age, history of AF, duration of AF, previous antiarrhythmic drugs tried, associated disease, and all clinical characteristics, were selected as a procedural control group from patients who underwent standard CPVA ablation between September 2004 and June 2005. Ablation procedure and selection criteria were similar for both groups. Patients were studied in a fasting state, and before the procedure gave their written permission after informed consent was obtained.
Remote magnetic navigation system
The remote magnetic navigation system consists of two focused-field permanent magnets of a neodymium-iron-boron compound that are computer-controlled and located on either side of the body, as previously reported (7). The magnetic field creates a 360° omnidirectional rotation of the device by uniform magnetic field (0.08-T) within an approximately spherical navigation volume of 20-cm diameter (NaviSphere) inside the patient’s chest. The magnetic navigation system is integrated with a modified C-arm single-plane digital imaging system (Artis, Siemens, Malvern, Pennsylvania). The combination of rotation, translation, and tilt movements of the magnets adjusts the magnetic field to any desired orientation in the spherical 20-cm diameter navigation volume. The maximum X-ray imaging angles are limited to approximately 30° left anterior oblique (LAO) and right anterior oblique (RAO). The operator is positioned in a separate control room, at a distance from the X-ray beam and the patient’s body.
A 4-mm flexible catheter (NaviStar-RMT, Biosense Webster, Diamond Bar, California) incorporating a small permanent magnet in the tip and additional magnets in the distal portion of the device can be deflected in the desired direction and steered by the magnetic navigation system (Fig. 1).The catheter is advanced and retracted by a mechanical device (Cardiodrive, Stereotaxis Inc.). All magnetic field vectors can be stored and, if necessary, reapplied while the magnetic catheter is navigated automatically (Fig. 2).The video workstation-based User Interface (Navigant, Stereotaxis Inc.) used together with the permanent magnets and the Cardiodrive unit (Stereotaxis Inc.) permits accurate orientation changes of the catheter by 1° increments and advancement or retraction by 1-mm steps (Fig. 3).Additionally, X-ray image data can be transferred from the X-ray system to the user interface of the magnetic navigation system to provide an anatomical reference.
CARTO RMT integration
The magnetic navigation system is integrated with a newly developed electroanatomical mapping system, the CARTO RMT mapping system (Stereotaxis Inc.) (Fig. 4).The CARTO RMT system sends real-time catheter tip location and orientation data to the magnetic navigation system. It also sends target locations, groups of points, and anatomical surface information from the electroanatomical map to the magnetic navigation system. The real-time catheter location information can be displayed on the Navigant reference X-ray images, enabling continuous real-time monitoring of the catheter tip position (Fig. 3) even without acquiring a fresh X-ray image.
Magnetic navigation procedure
After transseptal puncture, the magnetic catheter was advanced to the target positions in the left atrium and guided by using the X-ray system and user interface monitors and the catheter advancer (Cardiodrive) system. Directional catheter navigation was performed with a graphical user interface. A desired magnetic field vector was drawn and displayed on orthogonal fluoroscopic views, as shown in Figure 3. A control computer then calculates the appropriate orientation of each of the magnets and moves them to produce the desired magnetic field vector. Manipulation of the magnetic field vector usually requires a few seconds to redirect the field vector in a new direction with time proportional to the magnitude of the directional change. Both real-time and saved images are used to determine the appropriate orientation of the device (Fig. 3). When applying the field, magnetic field-induced torque orients the device in the appropriate direction. The magnetic field can be controlled in three dimensions from NaviSphere (Fig. 3). The magnetic field can be automatically applied by double-clicking on a desired orientation of NaviSphere. The Navigant user interface also provides a set of preset field directions that enable easier access to specific anatomical landmarks, such as the pulmonary veins (PVs) (Fig. 3) or mitral valve (Fig. 4). When a preset is applied, it can steer the catheter near the approximate region indicated. Based on our AF ablation strategy, navigation targets included all PVs, the ridge between the left superior PV and the left atrial appendage, the mitral valve annulus (between the inferior aspect of the left-sided encircling ablation line and the mitral annulus), the mitral isthmus line, the left posterior wall, and the left atrial roof. Navigation success was judged by a combination of electrogram analysis and analysis of fluoroscopic images (Figs. 3and 4). Stable magnetic contact was based on fluoroscopic image, minimum variation, good quality on electrogram recording, and CARTO stability criteria.
Remote mapping and ablation protocol
First, the transseptal sheath was positioned just proximal to the fossa ovalis after a magnetic catheter was placed into the left atrium (Fig. 2). After synchronizing with respiratory and cardiac cycles, such as inspiration and end diastolic period, a pair of best matched RAO/LAO images were transferred and kept into Navigant screen as background references for orientation and navigation (Fig. 3). Because two distinct planes of X-ray data are available, spatial locations in two dimensions can be localized by marking on corresponding points in the X-ray images and using epipolar geometry. The user marks the support or base of the catheter (the distal portion of the sheath) on the pair of X-ray images. This provides Navigant with the data needed to compute field orientations corresponding to particular targets. Next, a PV location was selected, either as a target by marking on suitable locations in the reference X-ray images, or by selecting a preset magnetic field vector based on selected study protocol from the list on Navigant (Figs. 3and 4). Throughout, points were simultaneously acquired for mapping (Fig. 5).The NaviStar-RMT (Biosense Webster) magnetic catheter was used for both mapping and ablation.
Circumferential PV ablation was performed with a target temperature of 65°C, a power limit of 50 W. Ablation lines were performed by sequentially navigating to contiguous points with a single 15-s application of RF current to achieve >90% reduction in the bipolar electrogram amplitude (Fig. 5), and/or peak-to-peak bipolar electrogram amplitude <0.1 mV inside the line. The mitral isthmus ablation line was extended to near the region of the left inferior pulmonary vein (LIPV), taking care to ensure, using impedance and anatomical map information, to stay outside the PV ostium (Fig. 5). The lesion line consisting of contiguous point-by-point locations of RF power delivery can be resent to and recorded on the fluoroscopic image. Potential vagal target sites are identified during the procedure as previously reported (8). Briefly, vagal reflexes are considered sinus bradycardia (<40 beats/min) or asystole, atrioventricular block, or hypotension occurring within a few seconds of the onset of RF application. If a reflex is elicited, RF energy is delivered until such reflexes are abolished, or for up to 30 s. The end point for ablation at these sites is termination of the reflex, followed by sinus tachycardia or AF. Failure to reproduce the reflexes with repeat RF is considered confirmation of local vagal denervation. A single ablation location was considered as failure if each of the selected targets was not completed within 60 min. Patients underwent transthoracic echocardiography before and after the procedure to assess potential complications.
Continuous variables are expressed as median and interquartile range, with range values, and compared by use of the Mann-Whitney Utest. For categorical variables the chi-square test was performed, unless the exact test was required for frequency tables when more than 20% of the expected values were <5. Statistical tests were performed with SPSS for Windows (version 13.0.1, SPSS Inc., Chicago, Illinois).
Remote catheter mapping
First three cases
Retrograde aortic approach was attempted only in the first patient in whom the magnetic catheter could not be passed through the aortic valve and therefore could not reach navigation targets within the left atrium. In the first three patients, the initial sheath position was unstable causing repeat dragging of both sheath and magnetic catheter into the right atrium (two patients) or even the right ventricle (one patient). In these cases, the transseptal sheath was pulled back in the inferior vena cava orifice while the catheter wire was left in the left atrium. With the first three cases, many attempts and time-consuming manipulations of the magnetic field were required to refine precise and stable catheter position.
The transseptal sheath was positioned just proximal to the fossa ovalis to allow the greatest movement of the magnetic wire catheter. Before mapping, the position of the magnetic catheter was also controlled by manual advancement or retraction of the catheter through the vascular sheath. Sheath insertion and positioning of diagnostic catheters, including the magnetic catheter, required a median of 7.9 min (range, 5 to 12 min). Median fluoroscopy exposure was 32.3 min (range, 21.8 to 81.4 min). After crossing the atrial septum and positioning the transseptal sheath, the physician left the interventional room to perform mapping and ablation from the control room. Remote magnetic navigation and ablation to all targeted sites was successfully achieved in 38 of 40 patients without complications.
Navigation to the PV ostia was guided by real-time biplanar fluoroscopic images, but venography was not performed to identify the location of PV ostia. We accurately navigated all PVs (Fig. 3), and then, by advancing magnetic catheter using the Cardiodrive unit (Stereotaxis Inc.), we acquired a median of 274 points (range, 139 to 382 points) for accurate maps of the left atrium (Table 1, Fig. 5). Minimal orientation changes of the catheter up to 5° were used for acquiring points around venoatrial junction region and up to 15° for the remaining left atrial reconstruction. At the beginning of the learning curve and for the first 10 patients, the catheter orientation was frequently adjusted using the keyboard rotation and deflection keys to reach mapping targets. In all cases, the catheter was retracted and advanced to access all PVs by using this sequence when feasible: left superior pulmonary vein (LSPV), LIPV, right superior PV, and finally right inferior PV. Afterwards, the mitral valve annulus and the left atrial appendage were accessed by selecting different field directions on NaviSphere. Finally, the magnetic catheter was navigated in sequence to the posterior wall, the roof, the septal wall, and the anterior wall.
Remote catheter ablation
In the first three patients, both mapping and ablation were extremely time consuming and in two of them RF applications were made manually with standard catheters. In the remaining 38 patients, the magnetic catheter successfully reached all targets. We started the encircling lesions around the left-sided PVs, which were completed by repeatedly moving and adjusting the magnetic catheter. The right-sided PVs were then encircled. During this process, the NaviSphere display orientation was adjusted to match that of the CARTO map to facilitate adjustment of the catheter directly from the NaviSphere. No interference of magnetic field with the surface 12-lead electrocardiogram (ECG) was observed in any patient. At the end of the procedure and at discharge, all patients were in stable sinus rhythm without antiarrhythmic drugs.
Overall, the median procedure time including mapping and ablation was 152.5 min (range, 90 to 380 min). In the last 28 patients, this time shortened to 148 min (range, 90 to 209 min) (Table 1). The median ablation time was 49.5 min (range, 17 to 154 min) for encircling all PVs. The mitral isthmus line and posterior lines were performed in a median of 12.5 min each. To evaluate the impact of the magnetic navigation learning curve for mapping and ablation, we compared procedural times for the first 12 patients and the last 28 patients. The median procedure time, including both mapping and ablation times, was 192.5 min (range, 92 to 380 min) in the first 12 patients and 148 min (range, 90 to 209 min) in the last 28 patients (p = 0.012) while ablation time significantly shortened in the last 28 patients (Table 1, Fig. 6).In the first 12 patients, a median of 34.5 min (range, 22.4 to 81.4 min) of fluoroscopy was used for mapping and ablation while it shortened to 30.3 min (range, 21.8 to 60.4 min) in the last 28 patients (p = 0.065) (Table 1).
Comparison between remote CPVA and control patients
Comparison of clinical and procedure parameters of the last 28 patients undergoing remote navigation and their control patients is shown in Table 1. By definition, cases and control patients were similar in clinical characteristics. Overall, procedure time was longer in remote navigation than in control group with similar mapping time but shorter ablation time (Tables 1 and 2).⇓Many more CARTO mapping points were collected in the remote group than in control group (Table 1). Vagal denervation, defined as elicitation and abolition of vagal reflexes during the procedure, did not differ in both groups (39.3% in remote ablation vs. 28.6% in standard approach; p = 0.397). Remote ablation time of right-sided PVs was shorter than standard ablation but similar for the remaining targets (Table 2). Atrial fibrillation was converted to sinus rhythm in two and three patients during standard CPVA and remote CPVA, respectively. At the end of the procedure, electrical cardioversion was successful in all patients in AF, regardless of the procedure.
No adverse effects occurred during navigation and ablation of the soft magnetic catheter despite very complex and broad movements inside the left atrium.
This is the first human study of a magnetically remote-guided catheter with the use of a new catheter advancer system for AF ablation. Our results indicate that remote magnetic navigation can safely maneuver magnetic soft catheters within the left atrium and around the PV ostia for successful ablation with shorter RF time probably due to the stability of catheter contact, particularly for the right-sided PVs. The integration of a stable magnetic catheter with a newly developed electroanatomical mapping system is very useful to reconstruct an accurate electroanatomical map by acquiring many more points than is possible manually for successful ablation even of challenging areas within the left atrium. Appropriate transseptal sheath positioning is important for assuring catheter stability. The remote magnetic guidance system in patients undergoing CPVA appears to be safe, feasible, and, more importantly, quick and easy to learn, reducing in all cases fluoroscopic exposure time for the operator.
Remote magnetic navigation and ablation
Precise target localization and catheter stability are prerequisite for adequate RF applications minimizing risks of potential complications. Stiff manually deflectable catheters with unidirectional or bidirectional deflection radius have several inherent limitations as stable wall contact may be difficult to achieve particularly in regions of complex cardiac anatomy. Usually, challenging sites within the LA include the anterior aspect of right-sided PVs, the narrow ridge that separates the left atrial appendage from the left PVs, and/or the mitral isthmus where movement of the valve greatly limits catheter stability. Incomplete lesions in such areas, particularly at mitral isthmus, facilitate development of multiple gaps, which may in the majority of the cases lead to post-ablation incessant macro-re-entrant atrial tachycardia (2). Despite aggressive protocols, a high rate of conduction recurrence across electrically disconnected segments late after ablation with standard, manually deflectable catheters has been reported to occur in up to 80% of the PVs four to five months after ablation (5). In the current study, all targeted sites regardless of their anatomic position were successfully ablated remotely, and abatement of all atrial potentials was reached rapidly in all targeted sites, usually disappearing within 10 s. This enhanced flexibility and compliance of the catheter could also compensate for cardiac and respiratory motion. These findings confirm previous studies in animals that demonstrated lesion block after RF applications by magnetic catheter even on trabeculated myocardium, which is the most difficult area to achieve linear conduction block (6). Vagal denervation was also obtained similarly to standard CPVA (8).
Catheter stability and sheath positioning
In the current study, to achieve optimal catheter stability associated with the greatest freedom of movement for navigation even in critical areas, initial positioning of the sheath was made just proximal to the fossa ovalis for mapping and ablation of left-sided PVs or lower in the inferior vena cava orifice for the right-sided PVs. Sometimes, the sheath may pull back during navigation, together with the soft magnetic catheter. When this inadvertently occurs, it is possible to rapidly regain access to the transseptal puncture simply by applying a stored magnetic field corresponding to entry into the left atrium. In this study, navigation to the right-sided PV region required much larger catheter deflections with shorter ablation times as compared with standard CPVA. These were more easily obtained positioning the sheath into the inferior vena cava orifice. Conversely, positioning of the sheath proximal to the fossa ovalis was required for better remote navigation and ablation of left-sided PVs. The same positioning was used for successful ablation of the ridge between the left atrial appendage and the LSPV. Reapplication of a previously applied magnetic field vector was useful to repeatedly navigate to previously visited targets decreasing ablation time in challenging sites.
CARTO RMT mapping integration system
In the current study, more mapping points were obtained with the NIOBE system (Stereotaxis Inc.) than with manual catheter manipulation because it is easier to obtain multiple points with this system assuring in all cases a homogeneous mapping density for successful ablation even of particularly challenging regions. The CARTO RMT (Stereotaxis Inc.) system was also able to send real-time catheter tip location and orientation data to the magnetic navigation system, which include target locations, groups of points, and anatomical surface information from the electroanatomical map to the magnetic navigation system. Also, magnetic field orientations corresponding to specific map points were stored on the magnetic navigation system and were re-applied if desired to repeatedly and accurately return to previously visited locations on the map.
Comparison of remote ablation with standard, manually deflectable catheters
Circumferential PV ablation, which is one of the two current predominant approaches for AF ablation, is a complex procedure requiring a rather extensive learning curve. This relatively long learning curve could limit the wide application of the procedure in the clinical practice (9). The current study is the first report in the literature prospectively documenting a single center experience of using remote magnetic navigation for AF ablation. Although an expert operator with extensive experience with standard, manually deflectable catheters performed all procedures, early results suggest that remote navigation is indeed a simple, safe, and useful system for AF ablation not requiring a substantial learning curve as the end point was successfully reached in almost all patients (38 of 40). In the first few patients, the procedure duration and fluoroscopy exposure times were excessively long as they were an expression of the underlying learning curve. This was mainly due to the need to confirm catheter positioning and stability visually during both mapping and RF applications. Both mapping and ablation were performed from control room due to remote navigation nature, reducing in all cases fluoroscopic exposure time for the operator. Although the manual approach may well be operator-dependent, the remote approach is not solely dependent on a single operator, but is most dependent on a well-trained team. This could explain why the overall procedure time was longer in the remote group than in the control group, while mapping time or PV denervation in both approaches remained similar. However, we believe that procedure times will improve as more experience is gained with it in the future. Ablation time of circumferential lesions around right-sided PVs was shorter, remotely suggesting that with this approach there are no specific challenging sites avoiding unnecessary RF energy applications.
There are no data on remote magnetic catheter navigation for AF ablation in humans. Our feasibility data confirm a recent study with this method in 42 patients with atrioventricular nodal re-entrant tachycardia where no complications were reported (7). This system was successful in performing completely controlled mapping and slow pathway modulation in 27 patients or ablation in 15 patients with a mean overall procedure time of 145 min calculated from puncture to sheath extraction.
No acute complications occurred during magnetic navigation and ablation confirming previous experimental and clinical studies on safety (6,7,10). In animals, attempts to intentionally perforate the heart with the magnetic catheter did not result in significant endocardial injury indicating a very low risk of cardiac injury (6). Our experience confirms these experimental observations as complex tip movements of the soft catheter within the left atrium did not result in any complications. However, further larger studies are required to evaluate whether remote navigation is associated with less complications than standard CPVA (9,11). In our initial experience, echocardiography performed before and after the procedure did not show abnormalities. Although previous animal studies reported interfering signal components in the presence of the magnetic field on surface ECG (1), we did not observe any interference with both surface ECG and intracardiac electrograms.
Mapping and navigation are displayed on separated screens, which may cause longer procedure times. Therefore, it seems to be reasonable to provide a single screen to improve remote procedure times.
The present and the future
At present, our experience is limited to a relatively small number of patients, and there are incomplete data on long-term follow-up. The learning curve with this navigation system appears to be short, but our experience is limited just to 40 cases performed by a highly skilled operator. Previous experience with magnetic navigation in the left atrium among 13 patients with left-sided accessory pathways also indicated that PV ostia stability may be problematic just at the beginning of the learning curve (10). The results of this study have important clinical implications as magnetic soft catheters can easily be navigated precisely and safely in the left atrium even in challenging sites.
The authors express their appreciation to Leonardo Mandile, RN, for his technical assistance and Rosa Michela for her continuous secretarial assistance.
This study was supported by a grant from Johnson & Johnson. This work received technical support from Stereotaxis Inc.
- Abbreviations and Acronyms
- atrial fibrillation
- circumferential pulmonary vein ablation
- left anterior oblique
- left inferior pulmonary vein
- left superior pulmonary vein
- pulmonary vein
- right anterior oblique
- Received August 23, 2005.
- Revision received November 9, 2005.
- Accepted November 16, 2005.
- American College of Cardiology Foundation
- Pappone C.,
- Rosanio S.,
- Oreto G.,
- et al.
- Pappone C.,
- Manguso F.,
- Vicedomini G.,
- et al.
- Ernst S.,
- Ouyang F.,
- Lober F.,
- Antz M.,
- Kuck K.H.
- Cappato R.,
- Negroni S.,
- Pecora D.,
- et al.
- Faddis M.N.,
- Blume W.,
- Finney J.,
- et al.
- Ernst S.,
- Ouyang F.,
- Linder C.,
- et al.
- Pappone C.,
- Santinelli V.,
- Manguso F.,
- et al.
- Faddis M.N.,
- Chen J.,
- Osborn J.,
- Talcott M.,
- Cain M.E.,
- Lindsay B.D.
- Pappone C.,
- Oral H.,
- Santinelli V.,
- et al.