Melby, Spencer J.; Zierer, Andreas; Voeller, Rochus K.; Lall, Shelly C.; Bailey, Marci S.; Moon, Marc R.; Schuessler, Richard B.; Damiano, Ralph J. Jr
Atrial fibrillation (AF), the most common arrhythmia, affects more than 2 million people in the United States and has many detrimental sequelae.1,2 A number of interventions have been proposed for the treatment of AF, including antiarrhythmic medicines, catheter ablation, and surgery. The surgical treatment of AF has been most successful with the Cox-Maze III procedure, which has long-term cure rates greater than 90%.3,4 Because of the technical difficulty in completing a “cut-and-sew” Cox-Maze procedure, many surgeons have begun using energy sources to create linear lines of ablation on the atria to replace the incisions. These energy sources include cryoablation, unipolar and bipolar radiofrequency energy, high-intensity focused ultrasound, laser, and microwave energy.5,6 Because of its ability to create rapid and reliable transmural lesions,7,8 bipolar radiofrequency (BPRF) ablation has been one of the primary energy sources used at our institution. After extensive laboratory investigation, our group has developed a modification of the original cut-and-sew procedure, using BPRF, which has been termed the Cox-Maze IV.7–10 In this procedure, many of the lesions of the conventional operation are replaced with linear lines of ablation. Early experience with the Cox-Maze IV procedure has shown success rates comparable to the Cox-Maze III procedure, with freedom from AF at last follow-up of greater than 90%.9
The BPRF device was unique in that it used an algorithm based on tissue impedance to determine lesion transmurality. Many of the other available technologies rely on predetermined time and/or temperature criteria to determine ablation duration.11–14 However, the BPRF sensing device tailored energy delivery by determining on-line measurements of tissue conductance every 20 milliseconds during ablation. The power delivered was changed, depending on the measured conductance at each time point. By keeping either the voltage or the current constant and measuring the returned current or the voltage across the electrodes, respectively, the tissue conductance (1/impedance) could be calculated precisely. The appropriate power to the tissue was then delivered on the basis of the calculated conductance value for the next 20 milliseconds, and the cycle continued. The ablation sensing unit (ASU) ablation algorithm had predetermined conductance value ranges that dictated whether the voltage or the current should be held constant and at what magnitude as it calculated the conductance. In laboratory studies, this algorithm was shown to predict transmural lesions 100% of the time in more than 300 tissue sections of ablation lesions, as confirmed by chronic histologic examination.7,8,15
It was hypothesized that the different parameters of ablation would be significantly different for each region of the atria due to the inhomogeneities inherent in pathologic human atrial tissue. The aim of this study was to evaluate the energy and time of ablation required in different atrial structures to create lesions and to evaluate the ability of a novel impedance algorithm to tailor energy delivery to the particular tissue geometry.
Patient Accrual and Statistics
Thirty-eight consecutive patients were enrolled after informed consent was obtained for their procedures, in accordance with the ethical standards of the Human Studies Committee at our institution. Members of the Division of Cardiothoracic Surgery at a single institution carried out the surgical procedures from December of 2003 through March of 2005. Patient demographics and surgical variables, including initial impedance, total energy, temperature, and ablation time, were collected prospectively. Concomitant surgeries were performed for recognized indications. Lesions were categorized into the following groups: right atrial free wall (which included the caval lesions), left atrial free wall, atrium up to mitral valve annulus, atrium up to tricuspid valve annulus, right pulmonary veins, and left pulmonary veins. For the purpose of data analysis, lesions were further grouped as right atrial lesions (which included right atrial free wall and atrium up to tricuspid valve annulus) and left atrial lesions (left atrial free wall, atrium up to mitral valve annulus, and right and left pulmonary veins).
All data are expressed as mean ± SD. Statistical analysis was carried out by using an unpaired t test to evaluate differences between two groups. Analysis of variance was used to analyze groups of three or more, with differences between groups evaluated with the use of a Fisher’s least significance difference test. A Pearson correlation coefficient was calculated to evaluate the relation between initial impedance measurement and time of ablation. Data were analyzed with the use of Systat for Windows version 11.00.1 (Systat Software Inc, 2004, Point Richmond, CA).
Bipolar Radiofrequency Ablation
The BPRF system consisted of the ablation sensing unit/generator and the AtriCure Isolator (AtriCure, Inc, Cincinnati, OH). The AtriCure device consisted of two 0.3-mm-wide electrodes embedded within the jaws of a clamp (Fig. 1). The electrodes in the clamp were either 5.4 cm long (Isolator) or 6.4 cm long (Isolator Long). The AtriCure system operated at power level typically between 15 and 20 W, with a maximum of less than 30 W and maximum voltage of less than 60 V. The generator continuously monitored voltage, current, temperature, time, and conductance. Tissue temperature was measured 1.3 mm from the electrode edge. Total ablation time and maximum tissue temperature were recorded for every lesion.
The power delivered during ablation was based on the impedance of the tissue and did not exceed 28.5 W. The ablation and sensing unit used an algorithm that changed the amount of power applied to the tissue during ablation, based on the changing tissue impedance. The algorithm cycled every 20 milliseconds, and the current or voltage was adjusted at each time interval, based on the measured impedance of the tissue. As the tissue was desiccated during ablation, the conductance fell as the tissue properties changed, and the power applied was decreased until conductance remained stable for 3 seconds below 0.0025 Siemens. This was used as an indicator of complete transmural ablation, and energy application was terminated (signaled by a change in the audible tone from the generator).
After either median sternotomy or right thoracotomy, patients underwent a pericardotomy and were placed on cardiopulmonary bypass. If patients were not in normal sinus rhythm, intraoperative direct-current cardioversion was performed. The right and left pulmonary veins were carefully dissected and isolated with the use of a blunt technique. Pacing thresholds from the right and left pulmonary veins were obtained with bipolar epicardial pacing. The bipolar radiofrequency clamp was then placed such that a rim of atrial tissue surrounding the pulmonary veins was ablated. Two ablations were routinely performed on each of the right and left sides to ensure isolation. After ablation, electrical isolation was confirmed by bipolar pacing at 20 mA from both the superior and inferior pulmonary veins. If atrial capture was present, the ablation was repeated until electrical isolation was achieved. Two of the 38 patients underwent pulmonary vein isolation alone, with no further ablations. In the remainder of patients, the right-sided9 lesions were created by making a simple atriotomy, which extended from the intra-atrial septum, up to near the atrioventricular groove at the acute margin of the heart. All the other incisions of the traditional cut-and-sew method were replaced with BPRF ablation lines (Fig. 2). In some patients, two cryolesions were placed at the tricuspid annulus, using a linear cryoprobe cooled to –60°C for 2 minutes. Four of 38 patients underwent right-sided lesions only because of isolated right heart pathology. The remainder (32/38) of patients underwent the full Cox-Maze procedure.
A single left-sided incision was extended onto the dome of the left atrium and inferiorly around the orifice of the right inferior pulmonary vein. It intersected the encircling right pulmonary vein ablation. A connecting ablation lesion was performed from the inferior aspect of the left atrium into the left inferior pulmonary vein. In atria larger than 5 cm in diameter, a second connecting ablation was placed from the superior aspect of the incision into the left superior pulmonary vein. Finally, a BPRF ablation line was performed from the inferior end of the incision down to the mitral annulus at a point in between the circumflex and right coronary circulations. A cryolesion was placed at the mitral annulus with a 15-mm bell probe (Frigitronics, CCS200, Trumbull, CT) cooled to –60°C for 3 minutes. The left atrial appendage was amputated, and a bipolar radiofrequency ablation was performed between the left atrial appendage and the left superior pulmonary vein. The left atrial appendage was oversewn (Fig. 2).
A total of 656 ablations were made on 38 patients. The Isolator was used in 13 patients and the Isolator Long was used in 25. Concomitant procedures performed in this study included mitral valve repair/replacement (n = 12), mitral valve and tricuspid valve replacement or aortic valve replacement (n = 7), coronary artery bypass grafting (n = 4), mitral valve replacement and coronary bypass grafting (n = 1), aortic valve replacement (n = 2), aortic valve replacement and coronary bypass grafting (n = 1), and septal myectomy (n = 1).
The patient demographics are shown in Table 1. The mean age of this population was 61 ± 17 years. Paroxysmal AF comprised 53% (20/38) of patients and permanent AF consisted of 37% (14/38). Mean left atrial diameter was 6.0 ± 1.3 cm.
By pacing all four pulmonary veins, electrical isolation was confirmed in each of the 38 patients. A total of 656 ablations were done in 38 patients. The time of ablation varied from 2.0 seconds to 29.9 seconds (mean, 11.1 ± 4.3 seconds). Mean maximum tissue temperature 1 mm from the electrode for all ablations was 45.7 ± 7.8°C (range, 23.7 to 69.3). Total energy delivered was 101.8 ± 54.6 Joules, but the range was very wide (15.4 to 457.3 Joules). The mean total energy was widely variable between groups (Table 2). An average of 2.8 ± 1.3 ablations were done on the left pulmonary veins (range, 2 to 7) and 2.8 ± 1.5 ablations on the right pulmonary veins (range, 1 to 9). A summary of comparison of total energy delivery, time of ablation, maximum temperature, and initial impedance is given in Tables 2, 3, 4, and 5, respectively.
The mean ablation duration for all lesions on the right atria was 10.5 ± 4.8 seconds (range, 2.0 to 29.9 seconds), which was significantly shorter than the left atrial ablations (11.8 ± 3.6 seconds; range, 5.6 to 25.8 seconds, P < 0.001). Mean total energy delivered was also significantly less on the right side (92.7 ± 59.6 Joules; range, 15.4 to 457.3 Joules) versus the left (110.2 ± 43.1 Joules; range, 26.3 to 340.0 Joules, P < 0.001). The lower mean total energy correlated with a higher mean initial impedance (127.0 ± 83.7 Ohms on the left versus 96.2 ± 46.6 Ohms on the right, P < 0.001). Mean maximum temperature, however, was not different between the right (45.8 ± 7.6°C; range, 25.1 to 69.3°C) and the left (45.5 ± 7.9°C; range, 23.7 to 63.3°C, P = 0.62).
The lesions involving the right atrium to the tricuspid valve annulus had the shortest duration of all groups (7.8 ± 4.4 seconds; range, 5.4 to 29.9 seconds, Table 3) and the lowest mean maximum temperature (42.6 ± 6.3°C; range, 26.6 to 57.9°C, Table 4). The total energy delivered was also the lowest among groups (56.4 ± 65.9 Joules; range, 19.6 to 457.3 Joules, Table 2), and the corresponding initial impedance was the highest (162.3 ± 86.5 Ohms; range, 41.6 to 450.5 Ohms, Table 5).
The left atrial free wall ablations had a mean duration of 12.2 ± 3.8 seconds (range, 5.6 to 25.8 seconds). The mean total energy delivered was 114.7 ± 46.2 Joules (range, 26.3 to 340.4 Joules), and the mean initial impedance was 85.1 ± 48.5 Ohms (range, 36.5 to 333.8 Ohms). Mean maximum temperature was 43.7 ± 8.3°C (range, 23.7 to 62.0°C). The mean right atrial free wall duration was shorter (10.9 ± 4.8 seconds; range, 2.0 to 29.9 seconds, P = 0.004), and the mean energy delivery was less (93.2 ± 52.8 Joules; range, 15.4 to 398.1 Joules, P < 0.001) compared with left atrial free wall ablations. The mean maximum temperature was higher on the right atrial free wall ablations than the left atrial free wall ablations (46.1 ± 8.0°C; range, 25.1 to 69.1°C, P = 0.003).
There was a wide range of initial impedance of all tissues (32.3 to 760.7 Ohms). When a Pearson correlation coefficient was calculated, the association of initial measured impedance with total energy delivered was statistically significant (r =–0.31, P = 0.002, Fig. 3).
The size of a lesion created by radiofrequency ablation is dependent on many factors, including temperature, time of ablation, and tissue properties of conductivity. The tissue effects of radiofrequency ablation can be accurately based on study of the specific absorption rate of tissue and the temperature profiles through the use of an analytical model Femlab (Comsol, Burlington MA) to calculate temperature and tissue damage.16 The AtriCure ablation technology has been modeled and optimized by using the same Femlab physics–based finite element tool and making adjustments for the bipolar nature of the AtriCure technology. In silico analysis has been correlated to ex vivo tissue studies as well as in vivo preclinical and clinical performances. Based on these tests, the algorithm used by the AtriCure system maintained balance of energy delivery with the impedance of the tissue. As the tissue heated up, it became more conductive; this signaled the system to decrease power to avoid overheating the tissue at the surface; then, as the cells began to break down and the impedance increased, the system increased power delivery to maintain the rate of work. The system monitored the tissue 50 times per second, recognizing impedance, adjusting power, and assuring proper tissue temperature each time. Because the system had a fixed electrode and always maintained either fixed current or voltage, it was able to accurately measure the impedance and determine if a stable, high impedance (low conductance) level had been attained. The AtriCure analytical models benefit from the fact that mechanical forces delivered by the clamp are generally uniform along the jaw length and that pressure is proportional to the volume compressed; both of these factors help improve the consistency of the tissue-specific absorption rate. Based on the models tested in both Femlab as well as ex vivo and in vivo tissue, the algorithm proved to be 100% effective at creating transmural ablations in dogs and pigs with relatively short ablation times.8,15 Because each area of the atrium has different properties such as thickness, amount of muscular trabeculations, and varying amount of fibrosis, an algorithm that monitors the delivery of energy and tailors it to the conductance of the tissue will deliver different amounts of energy to each area yet maintain the ability to create a transmural lesion.
This study evaluated the effect that different areas of the atria had on total energy delivered, time of ablation, and temperature, using real-time conductance measurement. Total energy of ablation was controlled by the sensing unit that cycled through an algorithm, which allowed for accurate and controlled current density; this in turn improved the efficiency of the ablation and allowed for accurate measurement of impedance. The algorithm terminated the ablation after a drop in conductance to a steady level below 0.0025 Siemens. Ablation times were significantly longer on the left atria versus the right, and the total energy delivered was likewise increased when ablating the left side of the atria. This is what would be expected, as the left atria are generally thicker than the right atria, particularly in patients with valvular heart disease and systemic hypertension. Thus, it would be anticipated that more energy would be needed to complete these lesions. However, the mean maximum temperature, regardless of time of ablation or total energy delivered, was below 50°C for all lesion sets. This was measured only 1.3 mm from the electrode, which was still within the jaws of the clamp. This is important, since irreversible tissue injury occurs only at temperatures greater than 50°C.17,18 The bipolar clamp is able to achieve temperatures great enough to create a narrow transmural lesion, yet BPRF minimized thermal spread, thus almost eliminating the chance for collateral damage to other cardiac or noncardiac structures.
There was a correlation between the total energy delivered and the initial measured impedance. The sensing unit determined the impedance by maintaining either the voltage or current fixed and measuring the resulting conductance. Depending on whether impedance increased or decreased, the voltage or current would be changed according to the algorithm. The correlation was not strong between total energy and initial impedance, probably because of the varying types of tissue (some tissue had an initial rise in impedance after ablation began, whereas other tissue had a decrease), yet there was a statistically significant association that demonstrated that the majority of tissue reacted in a similar manner to ablation.
Both the time of ablation and total energy delivered were minimized when less tissue was ablated. The right atrium–to–tricuspid valve annulus lesion, which is done on a small amount of tissue (usually only 1 to 2 cm of tissue in the clamp), had the least amount of total energy delivered, and the time of ablation was the shortest of all lesion groups. Most of the other ablations used the entire length of the clamp.
Many other ablation devices have relied on fixed time and/or temperature ablation for lesion creation.11–14 Because the atria are widely variable, with nonuniform thickness and geometry, this method of “one setting fits all” has potential downfalls, including the possibility of ablating for too long and creating damage to adjacent structures (such as the esophagus or coronary arteries).19 Likewise, in thicker atrial tissue, the ablation time might not be sufficient, and viable atrial myocardium could remain with its potential increased risk for treatment failure. An algorithm that senses the conductance during real-time ablation allows for energy delivery to be tailored to the specific tissue thickness and geometry, which helps ensure complete transmural lesions without delivering excess energy.
Our data have demonstrated the wide variability of energy delivery (a 50-fold variation) required to ablate human atrial tissue. In devices that rely on fixed-time ablation to ensure complete lesions are being formed, it would be essential that a dose-response curve be developed under conditions equivalent to the surgical setting and that these parameters be followed closely in the clinical setting. Having a device that uses an algorithm to tailor energy delivery to the specific tissue characteristics eliminates the necessity to follow such guidelines and gives the surgeon confidence that regardless of tissue nonuniformity, the lesion created would be adequate.
The principal limitation of this study was that it was impossible to collect histologic samples of ablation lines to confirm transmurality. In fact, these lesions occasionally did not create conduction block around the pulmonary veins, and multiple ablations were sometimes required. In the majority of patients, however, two ablations, which were done routinely in every patient per our clinical protocol, were sufficient to create conduction block. Even if some of our lesions were not transmural, our data still testify to the wide variety of energy needed to ablate pathologic human atria. Our 1-year freedom from atrial fibrillation of greater than 90% in these patients9 suggests that most of the lesions created conduction block, or one would have expected more late recurrences.
In summary, bipolar ablation of different atrial structures required widely different amounts of energy and ablation times. This variation probably was due to the inhomogeneity of atrial geometry and tissue impedance. Impedance-controlled ablation optimized the delivery of BPRF energy, minimizing energy requirements, ablation time, and thermal spread.
1. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983–988.
2. Furberg CD, Psaty BM, Manolio TA, et al. Prevalence of atrial fibrillation in elderly subjects (the Cardiovascular Health Study). Am J Cardiol. 1994;74:236–241.
3. Prasad SM, Maniar HS, Camillo CJ, et al. The Cox maze III procedure for atrial fibrillation: long-term efficacy in patients undergoing lone versus concomitant procedures. J Thorac Cardiovasc Surg. 2003;126:1822–1828.
4. Cox JL, Schuessler RB, Boineau JP. The development of the Maze procedure for the treatment of atrial fibrillation. Semin Thorac Cardiovasc Surg. 2000;12:2–14.
5. Melby SJ, Lee AM, Damiano RJ. Advances in surgical ablation devices for atrial fibrillation. In Wang PJ, ed. New Arrhythmia Technologies: Blackwell Futura. Malden, MA: Blackwell Publishing Inc., 2005; 233–241.
6. Viola N, Williams MR, Oz MC, Ad N. The technology in use for the surgical ablation of atrial fibrillation. Semin Thorac Cardiovasc Surg. 2002;14:198–205.
7. Prasad SM, Maniar HS, Schuessler RB, Damiano RJ Jr. Chronic transmural atrial ablation by using bipolar radiofrequency energy on the beating heart. J Thorac Cardiovasc Surg. 2002;124:708–713.
8. Prasad SM, Maniar HS, Diodato MD, et al. Physiological consequences of bipolar radiofrequency energy on the atria and pulmonary veins: a chronic animal study. Ann Thorac Surg. 2003;76:836–841.
9. Gaynor SL, Diodato MD, Prasad SM, et al. A prospective, single-center clinical trial of a modified Cox maze procedure with bipolar radiofrequency ablation. J Thorac Cardiovasc Surg. 2004;128:535–542.
10. Mokadam NA, McCarthy PM, Gillinov AM, et al. A prospective multicenter trial of bipolar radiofrequency ablation for atrial fibrillation: early results. Ann Thorac Surg. 2004;78:1665–1670.
11. Gammie JS, Laschinger JC, Brown JM, et al. A multi-institutional experience with the CryoMaze procedure. Ann Thorac Surg. 2005;80:876–880.
12. Gaynor SL, Byrd GD, Diodato MD, et al. Microwave ablation for atrial fibrillation: dose-response curves in the cardioplegia-arrested and beating heart. Ann Thorac Surg. 2006;81:72–76.
13. Molloy TA. Midterm Clinical Experience With Microwave Surgical Ablation of Atrial Fibrillation. Ann Thorac Surg. 2005;79:2115–2118.
14. Ninet J, Roques X, Seitelberger R, et al. Surgical ablation of atrial fibrillation with off-pump, epicardial, high-intensity focused ultrasound: results of a multicenter trial. J Thorac Cardiovasc Surg. 2005;130:803–809.
15. Gaynor SL, Ishii Y, Diodato MD, et al. Successful performance of Cox-Maze procedure on beating heart using bipolar radiofrequency ablation: a feasibility study in animals. Ann Thorac Surg. 2004;78:1671–1677.
16. Chang IA, Nguyen UD. Thermal modeling of lesion growth with radiofrequency ablation devices. Biomed Eng Online. 2004;3:27.
17. Haines DE, Watson DD. Tissue heating during radiofrequency catheter ablation: a thermodynamic model and observations in isolated perfused and superfused canine right ventricular free wall. Pacing Clin Electrophysiol. 1989;12:962–976.
18. Nath S, DiMarco JP, Haines DE. Basic aspects of radiofrequency catheter ablation. J Cardiovasc Electrophysiol. 1994;5:863–876.
19. Schwartzman D, Michele JJ, Trankiem CT, Ren J-F. Electrogram-guided radiofrequency catheter ablation of atrial tissue comparison with thermometry-guide ablation. J Interv Card Electrophysiol 2001;5:253–256.
© 2007 Lippincott Williams & Wilkins, Inc.