| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2002;73:1160-1168
© 2002 The Society of Thoracic Surgeons
a Departments of Cardiovascular Surgery and Electrophysiology, Sinai Samaritan and St. Luke’s Medical Centers, Milwaukee, Wisconsin, USA
Accepted for publication November 20, 2001.
* Address reprint requests to Dr Kress, Midwest Heart Surgery Institute, 2901 West Kinnickinnic River Parkway, Suite 511, Milwaukee, WI 53215, USA
e-mail: dkress{at}execpc.com
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Endocardial radiofrequency ablation was performed on 12 dogs with chronic atrial fibrillation. Lesions to isolate pulmonary veins in pairs, the left atrial appendage, and connecting lesions between these structures were administered in a randomized approach.
Results. Twelve dogs were in chronic atrial fibrillation for 31 ± 21 days before ablation. Atrial fibrillation was successfully ablated and rendered noninducible in all 12 dogs. All treatment failures observed with less than the full lesion pattern became a success when the remaining lesions were given.
Conclusions. Atrial fibrillation ablation using this left atrial lesion pattern is highly successful in this model. This approach may have significant utility as a concomitant procedure for patients with atrial fibrillation undergoing mitral valve procedures.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Several investigators subsequently altered the Maze 3 lesion pattern for the following reasons: (1) delayed onset of left atrial depolarization and disruption of blood supply to some atrial segments [2, 3]; (2) fluid retention caused in part by right atrial (RA) appendage excision and subsequent reduction in secretion of atrial natriuretic peptide [3]; and (3) electrical isolation of the left atrial (LA) region between pulmonary vein orifices, which can comprise as much as 33% of left atrial mass [3, 4]. Recent modifications to the Maze 3 have adopted alternative energy sources to accomplish linear lesions more quickly without incisions [5–7] and have adopted lesion patterns that are applied to the left atrium only [8–10]. Lesion patterns applied only to the right atrium using radiofrequency (RF) catheters appear to be less efficacious overall [11, 12] though they have been applied successfully primarily during surgery in conjunction with right-sided congenital repairs [13].
The purposes of this study were: (1) to characterize the appropriateness of a canine chronic atrial fibrillation (AF) model as a basis of comparison to chronic AF in patients; (2) to determine whether a specific left atrial linear ablation lesion pattern is effective in terminating chronic atrial fibrillation in this model and rendering it noninducible; and (3) to ascertain whether subsets of this lesion pattern are equally effective.
The lesion pattern consists of left atrial appendage isolation, bilateral pulmonary vein isolation, and connecting lesions between the right and left pulmonary vein lesions and to the left atrial appendage. Connecting lesions to the mitral annulus, right atrial lesions, septal lesions, and coronary sinus lesions were purposely left out of the model for theoretical and safety considerations, as discussed below.
|
Material and methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Chronic atrial fibrillation model
This model has been previously described [14]. The carotid artery and jugular vein were isolated by cutdown. Under fluoroscopy, a Swan-Ganz catheter was placed in the pulmonary artery through the jugular vein for measurement of the pulmonary wedge pressure. A transesophageal echo probe was placed for control measurement of the regurgitant ratio of fill of the mitral valve (diameter of the regurgitant jet/diameter of the mitral valve). A biopsy forceps catheter (Cordis 502-302L) was placed in the left ventricle by means of an introducer (Daig Fast-Cath 406181, 30 cm) in the carotid artery to cut or damage the mitral valve chordae to produce mitral valve regurgitation. Pulmonary wedge pressure and regurgitant ratio measurements were then made. Signs of significant mitral valve regurgitation included an increase in the pulmonary wedge pressure (
5 mm Hg) with V wave dominance and a regurgitant ratio of more than 0.25. These changes were usually immediate and permanent.
To induce AF, a permanent pacemaker lead (Medtronic, Inc, Minneapolis, MN) was placed in the RA appendage of each dog and attached to a modified permanent pacemaker (Medtronic Thera, model 8960) to allow for sustained high-rate pacing. The pacemaker was initially programmed at a rate of 30 beats/min and minimum output voltage. At 3 to 5 days after the initial implant, the pacemaker’s output voltage was increased to twice the diastolic capture threshold and the rate programmed to 400 beats/min. The ECGs were monitored at least three times per week. After AF was initiated, the pacer output was reduced to minimum output, and regular ECG monitoring continued. High-rate pacing was resumed if AF terminated spontaneously. After AF was sustained for at least 3 weeks (Table 1), endocardial RF ablation was performed.
|
The chest was opened by means of a sternotomy, and the heart was suspended in a pericardial cradle. Data were monitored continuously on a BERS oscilloscopic recorder (Bloom Associates, Denver, CO) and recorded as needed on photographic paper at a speed of 50 mm/s and 200 mm/s. Pacing and recording electrodes were attached to the left and right atrial appendages, and a roving probe was used for pacing and recording the surface of the right atrium.
Cardiopulmonary bypass
Each dog was cannulated in the right femoral artery and the superior and inferior vena cava. Heparin chloride (3 mg/kg) was given before cannulation. Cardiopulmonary bypass was initiated with an activated clotting time (ACT) of 480 seconds and the ACT was maintained at more than 400 seconds.
All dogs were placed on an identical cardiopulmonary bypass circuit, which included a hollow fiber membrane oxygenator with an integrated heat exchanger and cardiotomy reservoir, a 3/8-inch silicone elastomer pump head, and a 3/8-inch polyvinyl chloride tubing system (ie, an arterial filter, prebypass filter, and double venous lines). The oxygenator was primed with plasmalyte (1500 mL), mannitol (25 g), and sodium bicarbonate (50 mEq). The blood flow of each animal was maintained more than 40 mL · kg-1 · min-1. Initial cardiac arrest was achieved with 500 mL of a high-potassium crystalloid cardioplegia solution at 4°C injected into the aortic root. Myocardial preservation was maintained during aortic occlusion by administration of a low potassium cardioplegia solution at 4°C every 20 minutes or when activity occurred.
All dogs were maintained at normothermia for the entire procedure. Arterial blood gas samples were obtained and treated when appropriate, as follows: (1) before anesthesia; (2) after anesthesia and heparin administration, before the start of extracorporeal circulation; (3) after 10 minutes of extracorporeal circulation; and (4) at 30-minute intervals during extracorporeal circulation.
Noncontact mapping of atrial activation
Noncontact mapping of the left and right atria during chronic atrial fibrillation was performed in 8 dogs with AF arising from the above model before administration of any lesions. The use of the EnSite system (Endocardial Solutions, Inc., St. Paul, MN) has been described previously [15–18]. Briefly, the noncontact mapping system was used in a two-step process.
In step 1, the 64-channel multielectrode array catheter was positioned in the chamber of interest. A second, conventional electrophysiology catheter was placed in the same chamber. The mapping system is able to track and record the position of the conventional catheter with respect to the multielectrode array catheter. The conventional catheter was systematically moved throughout the chamber to build a three-dimensional model (geometry).
During step 2, the cardiac chamber electrical activation data were acquired by the 64 channel multielectrode array catheter during atrial fibrillation. The data were processed by the system using an inverse transform of the Laplace equation, allowing simultaneous computation of more than of 3000 "virtual" electrograms of the endocardial surface. The mapping system applied the virtual electrogram data from the more than 3,000 points to their corresponding locations on the three-dimensional geometry model, with colors proportional to the voltage on the endocardium. The result was extremely high-resolution activation maps of the right and left atria.
Multielectrode array catheters were positioned in both the right and left atria and remained in position without moving until mapping was completed. Data segments were acquired during atrial fibrillation from the right and left atria before administration of any lesions. Using 2-Hz low-pass filter settings, and the systems auto-focus feature at the nominal setting, 10 maps of atrial fibrillation at randomly selected intervals were analyzed from both the right and left atrium of each dog, avoiding parts of the cardiac cycle with far field depolarization or repolarization interference. The number of independent wavelets on each map was recorded. Each data segment was also analyzed to determine whether rapid, repeated focal activation was present.
Effect of cardioplegia
In 3 animals, we tested whether the left atriotomy and cardioplegia themselves had an effect on the ability to induce sustained AF. In these animals, once on cardiopulmonary bypass, the aorta was cross-clamped, cardioplegia given, and the left atrium opened. After 10 minutes of cardioplegic arrest, the heart was reperfused. Induction of AF was then attempted according to the protocol in Table 2.
The remainder of the experiment was then conducted identically to that in the other animals, starting with the first lesion in the randomization sequence.
|
Lesions were created to isolate the pulmonary veins and the left atrial appendage as well as connecting lesions between these sites (Fig 1). The left atrial appendage was everted to aid in ablation around its base. Various combinations of contiguous electrodes were used, usually one or two at a time. Particular attention was given to maintain good electrode-tissue contact and to avoid leaving gaps or skipping areas. Lesions were repeated if gaps or nontransmural lesions were suspected.
|
|
Determination of lesion depth
After completion of the lesion randomization protocol, the RV free wall was excised and lesions were created in the intraventricular septum of 6 dogs to determine the gross lesion size during normothermic bypass in cardiac muscle. Four different lesions were created in each dog by delivering RF energy for 30 seconds and 60 seconds at temperature set points of 70°C and 80°C.
The dogs were then euthanized, the hearts was removed, and all lesions were examined visually and photographed. Lesion depths were measured to assess lesion homogeneity and transmurality. Samples of the lesions were removed, fixed in 10% formalin, sectioned, and stained using hemotoxylin and eosin. A pathologist examined these slides to correlate the grossly observed measurement of tissue damage.
Statistical methods
Results are reported as mean ± standard deviation. Preoperative and operative variables were compared between groups using analysis of variance with Tukey’s Studentized Range Test. A p value of less than 0.05 was considered significant.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Procedure time
Total RF time ranged from 11 to 82 minutes with a mean of 29.5 ± 23 minutes, mean total time on bypass was 142 ± 43 minutes, and mean cross-clamp time was 74.3 ± 29 minutes. Total RF time was not significantly longer in any of the four groups.
Results of noncontact mapping
A mean of 1.4 simultaneous wavelets were observed in the right atrium and 2.6 wavelets were observed in the left atrium. Representative maps are shown in Figure 2.
No sites of repetitive focal activation were observed in any of the evaluated segments.
|
Results of lesion sequence
All 12 dogs in the randomized lesion sequence were successfully rendered noninducible for sustained atrial fibrillation (Fig 3). Three dogs (group 4) had all three lesion patterns (pulmonary vein, connecting lesions, and left atrial appendage) at the outset, after which AF was noninducible. Each of the other three groups had animals that remained inducible after the initial or second component lesion (Fig 4).
Once the entire lesion pattern was applied by adding the other components, all animals in these three groups were rendered noninducible.
|
|
|
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
In patients with AF associated with mitral regurgitation, Harada and colleagues [20] noted sites of repetitive atrial foci in the majority of patients, presumably serving the role of initiator and LA dilation with ischemia playing the role of substrate. It is reasonable to hypothesize that application of our lesion pattern would successfully ablate AF in those patients in whom such preexisting automatic triggers are eliminated or electrically isolated by the lesion pattern. Although a focal source may often be the trigger or initiating event for AF and may even play a role in creating the substrate for sustaining AF, a focal driver is not necessarily required for the maintenance of AF once it is initiated. It appears that the lesion pattern described in this study also disrupts the substrate necessary for the maintenance of AF.
Noncontact mapping also revealed that, not surprisingly, left atrial activation patterns during atrial fibrillation are more complex and involve a larger number of wavelets than the right atrium. The number of simultaneous wavelets in each chamber was lower than observed by Jalife and colleagues [21] when performing high-density optical mapping of the atria during atrial fibrillation. This difference may be due to the workings of the auto-focus feature of the EnSite system. This feature uses algorithms that determine the peak negative voltage on the entire map, and then automatically assign the map color scale based on this. Wavelets whose peak amplitudes were less than 80% of peak when using the nominal setting would therefore not show on the color map.
Choice of lesion pattern
This study validates a lesion pattern consisting only of left atrial lesions to restore sinus rhythm and to render AF noninducible using a randomized sequence of "building block" lesions and a rigorous rapid atrial pacing protocol to test their efficacy. There is now data to support the concept that atrial fibrillation begins in the left atrium in the great majority of patients, in the sense that both the triggers and the substrate for the perpetuation of atrial fibrillation are concentrated in the left atrium.
Not all possible components of the Maze 3 lesion pattern were tested in this study. A randomization sequence including RA, septal, coronary sinus (CS) and MV annulus connecting lesions would have necessitated an unwieldy number of animals. Therefore, lesions of the Maze 3 that either could not be supported from other independent investigators or that had relative contraindications due to either previous case reports of adverse events or the potential for adverse events based on anatomical grounds were excluded.
Specifically, RA lesions were excluded because of data from Haissaguerre and associates [22] and Harada and colleagues [20], showing AF triggers to be concentrated in the pulmonary veins in paroxysmal AF and in the LA appendage in mitral valve disease associated AF and to be relatively rare in the right atrium.
The mitral valve connecting lesion was excluded because of reports of coronary artery damage during the Maze 3 [23, 24] and the proximity of the circumflex artery to the MV annulus, particularly in patients with left dominant and codominant coronary circulations. We avoided CS lesions because of the potential for thermal injury-induced stricture, as well as the lack of proof that focal circumferential CS lesions delivered distal to the actual CS ostium effectively block RA-LA conduction. Septal lesions were left out because they are considered a nonessential part of the Maze 3 lesion pattern [25] and do not isolate known trigger sites.
The success of the lesion pattern shown in Figure 1 is likely due to an interruption of the left atrial macroreentrant circuits found in the electrophysiological mapping of these animals. The pulmonary veins and LA appendage are common sources of clinically relevant triggers in human patients, but in this model no triggers were seen during mapping. The pulmonary vein and LA appendage isolating lesions therefore probably serve simply to allow the gapless blockade of reentrant circuits by the connecting lesions. In patients, these lesions (Fig 7) would also be expected to remove potential sources of triggers.
|
The lesion set was successful despite absence of a specific connecting lesion to block reentry around the mitral valve annulus, or ablation of the coronary sinus. This suggests that in this model of AF, reentry circuits involving these structures are not particularly common and are not likely to be a byproduct of the lesion pattern.
Clinical use of lesion pattern
A modification of this lesion pattern in patients would take advantage of the fewer pulmonary veins (four compared to six in dogs) and the ability to isolate the left and right pulmonary veins as pairs with a single connecting lesion (Fig 7). This also avoids the complete isolation of the posterior atrium caused by the Maze 3, which accounts for 30% of the total atrial mass [4]. Obliteration of the LA appendage in patients either by LA appendectomy or oversewing the base after ablation will be mandatory to reduce the risk of thromboembolic complications.
A strategy that limits the initial operative lesion pattern to left atrial lesions only with noncontact mapping-guided catheter ablation for postoperative typical right atrial flutter has the advantage of avoiding a lesion that will be necessary in fewer than 10% of patients not presenting with a coexisting history of atrial flutter [26]. On the other hand, patients that are known or suspected to have a typical right atrial flutter circuit preoperatively can undergo right atriotomy and RF ablation immediately after cross clamp removal with the benefit of visible anatomic landmarks that allow the isthmus lesion to be created in less than 5 minutes.
Study limitations
No model of chronic atrial fibrillation can be generalized to all patients, particularly those with very large left atria or atrial fibrosis. The favorable acute results achieved in this study are encouraging but will require confirmation in clinical trials with long term follow-up. Patients with coexisting right atrial flutter circuits may be expected to require an additional intraoperative or subsequent catheter based isthmus ablation lesion.
The proposed clinical lesion pattern may be less effective in eliminating macroreentrant left atrial circuits, and therefore have a higher failure rate, in patients as the size of the left atrium and its surface area increases. Atrial thickness in the clinical setting may be substantially greater than that seen in these animals and require radiofrequency energy delivery at a higher set temperature or greater duration than that used in this study. The septal lesion dosimetry data presented here may be useful in determining clinical settings to deliver transmural lesions.
In conclusion, we have characterized a reliable canine model of chronic atrial fibrillation using bioptome created mitral regurgitation combined with asynchronous rapid atrial pacing. A lesion pattern comprised of electrical isolation of the pulmonary veins, left atrial appendage, and left atrial connecting lesions was applied using a surgical RF ablation probe. This lesion pattern was broken down into "building block" lesions that were randomly applied in sequential order to determine a necessary and sufficient subset of component lesions needed to terminate chronic atrial fibrillation.
Chronic atrial fibrillation was terminated and rendered noninducible in all 12 animals given the entire lesion pattern, with treatment failures if component lesions were omitted. This lesion pattern is clinically relevant to patients with paroxysmal or chronic atrial fibrillation undergoing mitral valve surgery.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. H. Seo, Y. Shi, S.-W. Huang, K. Kim, and M. O'Donnell Thermal strain imaging: a review Interface Focus, August 6, 2011; 1(4): 649 - 664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. K. Voeller, R. B. Schuessler, and R. J. Damiano Jr. Surgical Treatment of Atrial Fibrillation , January 1, 2008; 3(2008): 1375 - 1394. [Full Text]
|
|
|
|
 |
|
 T. Ota, D. Schwartzman, D. Francischelli, D. A. Hettrick, and M. A. Zenati Impact of beating heart left atrial ablation on left-sided heart mechanics. J. Thorac. Cardiovasc. Surg., October 1, 2007; 134(4): 982 - 988. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Guiraudon, D. L. Jones, A. C. Skanes, D. Bainbridge, C. M. Guiraudon, S. M. Jensen, X. Yuan, M. Drangova, and T. M. Peters En Bloc Exclusion of the Pulmonary Vein Region in the Pig Using Off Pump, Beating, Intra-Cardiac Surgery: A Pilot Study of Minimally Invasive Surgery for Atrial Fibrillation Ann. Thorac. Surg., October 1, 2005; 80(4): 1417 - 1423. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Halkos, J. M. Craver, V. H. Thourani, F. Kerendi, J. D. Puskas, W. A. Cooper, and R. A. Guyton Intraoperative Radiofrequency Ablation for the Treatment of Atrial Fibrillation During Concomitant Cardiac Surgery Ann. Thorac. Surg., July 1, 2005; 80(1): 210 - 216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B.-K. Lam, M. Boodhwani, J. P. Veinot, P. J. Hendry, and T. G. Mesana Surgical treatment of atrial fibrillation with diathermy: an in vitro study Eur J Cardiothorac Surg, March 1, 2005; 27(3): 456 - 461. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Park, C.-C. Chou, P. C. Drury, Y. Okuyama, A. Peter, A. Hamabe, Y. Miyauchi, R. M. Kass, H. S. Karagueuzian, M. C. Fishbein, et al. Thoracic vein ablation terminates chronic atrial fibrillation in dogs Am J Physiol Heart Circ Physiol, June 1, 2004; 286(6): H2072 - H2077. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Bonanomi, D. Schwartzman, D. Francischelli, K. Hebsgaard, and M. A. Zenati A new device for beating heart bipolar radiofrequency atrial ablation J. Thorac. Cardiovasc. Surg., December 1, 2003; 126(6): 1859 - 1866. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Raman, S. Ishikawa, M. M. Storer, and J. M. Power Surgical radiofrequency ablation of both atria for atrial fibrillation: results of a multicenter trial J. Thorac. Cardiovasc. Surg., November 1, 2003; 126(5): 1357 - 1365. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Akpinar, M. Guden, E. Sagbas, I. Sanisoglu, U. Ozbek, B. Caynak, and O. Bayindir Combined radiofrequency modified maze and mitral valve procedure through a port access approach: early and mid-term results Eur J Cardiothorac Surg, August 1, 2003; 24(2): 223 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.M. Gillinov and P. M. McCarthy Atricure bipolar radiofrequency clamp for intraoperative ablation of atrial fibrillation Ann. Thorac. Surg., December 1, 2002; 74(6): 2165 - 2168. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Gillinov, E. H. Blackstone, and P. M. McCarthy Atrial fibrillation: current surgical options and their assessment Ann. Thorac. Surg., December 1, 2002; 74(6): 2210 - 2217. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
