| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of Cardiac Surgery, University of Maryland Medical Center, Baltimore, Maryland
b North Shore Medical Center, Salem, Massachusetts
Accepted for publication February 3, 2009.
* Address correspondence to Dr Gammie, Division of Cardiac Surgery, University of Maryland Medical Center, N4W94, 22 S Greene St, Baltimore, MD 21201 (Email: jgammie{at}smail.umaryland.edu).
Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5–8, 2008
| Dr Gammie discloses a financial relationship with ATS, Inc.
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Methods: Between July 2002 and November 2005, 119 consecutive patients underwent surgical AF correction with the CryoMaze procedure. Mitral valve disease was the primary indication for operation in 66%. AF was continuous in 65%. Rhythm assessment was with 2-week continuous electrocardiographic (ECG) monitoring in 75% of patients and by noncontinuous ECG in the remainder. Median follow-up was 3.2 years and was 98% complete.
Results: There was one hospital (0.8%) death. Survival at 3 years was 84%. One perioperative stroke resolved completely. No late strokes occurred. In 4 of 119 patients (4 (3.4%), pacemakers were inserted during the index hospitalization. Median length of stay was 7 days. Overall freedom from AF more than 3 years after operation was 60%. Among patients with preoperative intermittent AF, 85% (28 of 33) were in normal sinus rhythm, and 47% (27 of 58) with continuous AF were in normal sinus rhythm (p < 0.001).
Conclusions: CryoMaze AF correction is safe and is associated with a very low risk of stroke. Rates of normal sinus rhythm at more than 3 years postoperatively were high for patients with intermittent AF and acceptable for those with continuous AF. This experience supports wider application of the CryoMaze to all patients with AF who need cardiac operations.
Surgical atrial fibrillation (AF) correction was first performed clinically in 1987 by Dr James Cox [1]. The cut-and-sew Maze procedure consists of a series of predetermined incisions and focal cryolesions in the left and right atria [2], and is thought to cure AF in 85% to 95% of patients [3–5]. The Maze procedure addresses symptoms by establishing sinus rhythm, improves cardiac performance by restoring atrioventricular synchrony, and reduces the risk of thromboembolism [6, 7]. Technical complexity has limited adoption of the cut-and-sew Maze procedure. A variety of devices that use several different energy sources have been introduced into clinical practice that permit the surgeon to rapidly perform AF correction with few suture lines.
Early follow-up of the Maze procedure performed with argon-powered cryoenergy (the CryoMaze procedure) suggests safety and efficacy for the treatment of AF [8]. Recently published guidelines have defined uniform reporting guidelines and have embraced the concept of continuous electrocardiographic (ECG) monitoring as the appropriate follow-up technique for surgical or catheter-based AF correction procedures. This study reports late clinical outcomes of an initial experience with the CryoMaze procedure.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
All patients underwent preoperative echocardiography to establish left atrial size, left ventricular ejection performance, and the presence and extent of heart valve dysfunction.
AF was classified as continuous if it was present all of the time and intermittent if it was not present all of the time, based on the classification system described by Cox [9].
Operative Technique
The CryoMaze procedure consisted of a lesion set identical to the classic Cox-Maze III operation [2], with two exceptions: the atrial septal lesion was omitted, and the internal right atrial lesions were variably performed. The CryoMaze procedure was always performed before valve operations. An argon-powered cryoprobe (SurgiFrost 10, ATS, Inc, Minneapolis, MN) was used to create all lesions. The exterior right atrial lesions (lateral aspect of the superior vena cava (SVC) to the inferior vena cava (IVC) and a perpendicular lesion from first lesion to right atrioventricular (AV) junction) (Fig 1), were performed on the beating, nonworking heart immediately after the initiation of cardiopulmonary bypass. The duration of cryothermal application for these lesions was 2 minutes. If a full right-sided Maze was performed, the right atrium was entered with a standard Cox incision from the IVC-SVC line obliquely up to the right AV groove. The remainder of the internal right lesions were created with a cryoprobe, duplicating the standard lines of the Cox III procedure.
|
|
Follow-Up
After hospital discharge, patients were managed by their referring cardiologists. Follow-up was by phone interview, physician examination, or review of outside medical records. We used a cross-sectional (common closing date) method of follow-up, and stringent efforts were made to contact all surviving patients [10].
All patients were asked to participate in continuous ECG monitoring for 14 days. Two-lead ECG strips were continuously acquired by a mobile outpatient telemetry system (Cardionet Inc, Conshohocken, PA) and transmitted wirelessly to a central monitoring station [11]. Reports were independently reviewed by an electrophysiologist (S.S.) who was blinded to the patient's treatment status. AF or flutter was diagnosed if present for more than 30 seconds [12]. AF burden was defined as the fraction of monitored time that the patient was in AF. Patients with permanent pacemakers underwent device interrogation, and for all other patients, electrocardiograms (ECGs) were available for analysis. Results are reported using recommendations for data analysis and reporting after AF operations [12, 13].
Statistical Analyses
Data were entered into a database and analyzed with JMP 6.0 software (SAS, Cary, NC). Descriptive data included means, standard deviations, and ranges for continuous variables, and proportions for categoric variables. Means of continuous variables were compared using two-sample t test or Wilcoxon rank sum test, depending on the distribution. The 
2 test or Fisher exact test was used to compare proportions among different groups. Kaplan-Meier estimates of survivor functions were obtained for survival data. Predictors of sinus rhythm were identified with univariate and multiple logistic regression analysis.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Completeness of Follow-Up
Two patients (1 homeless, 1 who left the country) were lost to follow-up. Among 96 patients alive and available for cardiac rhythm assessment, 72 (75%) agreed to undergo continuous ECG monitoring. The remaining 25% refused but all had ECGs available for analysis. Among patients assessed with continuous ECG monitoring, the mean duration of monitoring was 11 days (range, 1 to 14; median, 14 days). The median duration of heart rhythm follow-up was 3.2 years (range, 1.2 to 5.2 years; Fig 3).
|
|
|
The median postoperative length of stay was 7 days. During the index hospitalization, 4 of the 119 patients (3.4%) underwent pacemaker insertion. Three underwent pacemaker implantation for symptomatic bradycardia on postoperative days 3, 7, and 19, and another with poor left ventricular function underwent placement of epicardial biventricular leads during the operation and continues to have a paced rhythm at late follow-up. An additional 7 patients have required postoperative pacemaker implantation between 1.5 and 36 months after operation. Freedom from permanent pacemaker implantation is demonstrated in Figure 5. Overall freedom from new pacemaker insertion was 88% at 3 years after operation.
|
|
|
|
Key Findings
Key findings of this study include a low perioperative risk of CryoMaze AF correction, a total absence of permanent perioperative and late neurologic events, and late (> 3 years) freedom from AF in 85% of patients with intermittent AF and just under 50% in patients with continuous AF.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Cryothermy was the first energy source used to directly modify the heart's conduction system both experimentally and clinically, and is an integral part of the classic cut-and-sew Maze procedure [6, 15, 16]. Potential advantages of cryoenergy compared with other energy sources include:
The observed low operative mortality and absence of collateral injury to the esophagus or phrenic nerves strongly supports the safety of argon-based cryoenergy for surgical AF correction. Although the rate of normal sinus rhythm (85 %) for patients with preoperative intermittent AF was gratifying, the rate for patients with continuous AF (47%) was less so. Possible explanations for this finding include (1) cryothermy is less effective than other energy sources, (2) the duration and intensity of follow-up in this study have exposed less optimal results (the prevalence of AF is related to intensity of looking for it) and/or (3) the lesion set performed with cryoenergy was not optimal.
Moten and colleagues [21] described 41 patients who underwent a stand-alone CryoMaze intervention. They reported no operative deaths and a low stroke rate. With a follow-up of less than 1 year and rhythm assessment by spot ECG, 87% of patients were in normal sinus rhythm. Rahmanian and colleagues [22] used cryoenergy to perform left-sided lesions only in 141 patients. Hospital mortality was 2.8%, the stroke rate was 4.3%, and 12% of patients required pacemaker insertion. Mean follow-up was less than 1 year, rhythm assessment was done with spot ECGs, and freedom from AF was 79% at 12 months and 71% at 24 months. Comparison of rates of normal sinus rhythm in the present study and other published reports of argon-based cryoablative surgical AF correction is hampered by a lack of long-term follow-up and infrequent use of continuous ECG monitoring.
Radiofrequency is the most commonly used energy source for nonincisional AF correction [14]. A recent report of 137 patients undergoing bipolar radiofrequency surgical AF correction used spot ECG rhythm assessment at a median follow-up of 9.5 months and found a freedom from AF recurrence without antiarrhythmic drugs of 60% to 83%, depending on the lesion set used [23]. Gillinov and colleagues [24] described results of AF correction using radiofrequency ablation in 333 patients with permanent AF. Rhythm assessment was also by spot ECGs, and fewer than 20% of patients had ECG assessment 3 or more years postoperatively. Results were reported up to a maximum of 18 months postoperatively and demonstrated a prevalence of AF of just over 20%. Doukas and colleagues [25] reported a randomized trial of left atrial radiofrequency AF correction among patients with continuous AF having concomitant mitral valve procedures. Rhythm assessment was primarily by spot ECGs, with 24-hour Holter monitoring reserved for patients with symptoms related to AF. Maximum follow-up was 12 months, and the rate of normal sinus rhythm was only 44%. Groh and colleagues [26] reported results for 129 patients undergoing surgical AF correction with high-intensity focused ultrasound. Routine Holter monitoring was used. With a median follow-up of 1 year, the rate of freedom from AF was 84% [26].
Khargi and colleagues [27] performed a systematic review of AF surgical correction and used multivariable regression analysis to conclude that there was no difference in rates of sinus rhythm between patients undergoing cut-and-sew Maze procedures and those treated with a variety of alternative energy sources. A key limitation of their review—and of most published outcome data of AF surgical correction—was that rhythm was compared at a limited interval of 6 months from the procedure.
Long-term follow-up after the cut-and-sew Maze procedure is available. Prasad and colleagues [28] reported long-term results from the original Washington University series and found that a remarkable 96% of patients were free of AF at a mean follow-up of 5.4 years. Rhythm assessment included mail and phone questionnaires and spot ECGs. Stulak and colleagues [29] reported long-term results from the Mayo Clinic. In this series, with approximately half of patients presenting with intermittent AF, the rate of normal sinus rhythm at a median follow-up of 3 years was 76%. The rate of normal sinus rhythm declined over time and was estimated to be only 50% at 10 years postoperatively. Rhythm assessment techniques were similar to those of Prasad and colleagues. When more rigorous methods of rhythm assessment are applied, sinus rhythm rates after the cut-and-sew maze procedure are lower. Ballaux and colleagues [30] used routine Holter monitoring to assess late results after the Cox-Maze III procedure. In a series of 203 patients, normal sinus rhythm was present in 57% at a mean follow-up of 4 years.
Given the general lack of long-term follow-up and variable rhythm monitoring techniques among currently published series of surgical AF correction, it is not possible to draw definitive conclusions about the relative efficacy of CryoMaze AF correction compared with either cut-and-sew or other energy source-based procedures.
Although the present series does not include a control group having heart operations without concomitant AF correction, several small prospective randomized trials of AF correction surgery define AF prevalence among control patients [31]. In these trials, sinus rhythm rates for control patients with preoperative AF not having concomitant AF correction, assessed at relatively short follow-up of 12 to 18 months, was 4.5% to as high as 40% [25, 31]. The rate of normal sinus rhythm in the present study is higher, suggesting that CryoMaze intervention is associated with higher rates of normal sinus rhythm than untreated controls, even among patients with continuous AF.
There is growing evidence that the method of rhythm assessment strongly influences reported success rates of AF correction procedures. Edgerton and colleagues [32] have shown that spot ECG follow-up alone, compared with continuous ECG monitoring, underestimates the prevalence of AF by 16% to 25%. Another study that compared AF detection rates by ECGs with implantable AF detection devices (in patients with AF treated with antiarrhythmic drugs) found that the prevalence nearly doubled, from 46% to 88% [33]. In another study, an implanted pacemaker capable of continuous rhythm assessment was used to examine heart rhythm after catheter-based pulmonary vein isolation [34]. Routine ECG monitoring resulted in a 71% freedom from AF. This dropped to 43% with the implanted device, and only 21% were considered completely free of AF.
An underappreciated benefit of the Maze procedure is stroke prevention. Cox [4] documented a very low risk of 0.1% per year of late stroke after the traditional Cox-Maze III operation. The current study supports the conclusion that stroke is very uncommon after AF surgical correction. Although the patients who underwent operations had a high-risk profile, with 14 of 119 patients (12%) having a history of a neurologic event preoperatively, the perioperative and late events were strikingly low, comprising one neurologic event that was completely reversible and no strokes. We may never know the relative contributions to stroke prevention of left atrial appendage exclusion and restoration of left atrial function [35], but the consistent finding in this and other observational studies of very low stroke rates after surgical AF correction is a powerful argument for continued performance of CryoMaze AF correction in all patients with a history of AF.
Limitations
This series was performed in a retrospective fashion using cross-sectional follow-up rather than in a prospective fashion using anniversary methods. Our observation that the rate of sinus rhythm increased with longer follow-up may be biased by increased mortality among patients with AF compared with those in normal sinus rhythm and is at odds with other clinical experiences that describe a decline in rates of sinus rhythm with time. We are currently enrolling patients in a prospective trial and performing preoperative and postoperative continuous ECG monitoring. This approach will allow firm conclusions about late recurrence of AF, and increase understanding of the efficacy of the CryoMaze procedure as a function of preoperative AF burden.
Not all patients were agreeable to undergoing continuous ECG monitoring when contacted late after CryoMaze AF correction. It is possible that sampling error among the 25% of patients in this study who had only spot ECG monitoring caused an underestimate of AF prevalence. However, only 4 of 72 patients (5.5%) in this study assessed with continuous ECG monitoring had intermittent AF (ie, most patients were in AF all of the time or none of the time). Because the AF burden was 0.3 for these patients, spot ECG monitoring would miss two-thirds of them, so less than 4% of patients examined with spot ECGs might have been misclassified as not having AF when it actually was present.
The variability in performance and pattern of the right-sided lesions may have influenced success, although the multivariate analysis did not identify this as a significant predictor of sinus rhythm.
Conclusions
This series demonstrates that CryoMaze AF surgical correction is safe when performed alone or combined with other cardiac operations, that a substantial number of patients enjoy long-term freedom from AF, and that the rate of perioperative and long-term stroke is negligible. There is an opportunity to increase the efficacy of the CryoMaze procedure among patients with continuous AF, and efforts to modify the procedure and understand the basic science and pathophysiology of AF are necessary. There is a pressing need for prospective randomized trials of surgical AF correction to assess the effect of this intervention on stroke rates, quality of life, heart rhythm at late follow-up, and survival. Given its safety and efficacy, as well as the provocatively low stroke rate in long-term follow-up, we recommend performance of the CryoMaze procedure for all patients with AF in whom a cardiac operation is planned.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
When you look at this study group compared to the group that was presented at the Fifty-first Meeting, you see that there are similar demographics related to the patient's age, the continuous vs the intermittent rate of atrial fibrillation, the LA [left atrial] size, and the procedure was essentially the same. There are some significant differences between the two groups, mainly that the current group, the duration of atrial fibrillation that predated the procedure was shorter, at 18 months, vs the prior group that was described at 48 months. Despite that, we see a significant falloff from the 1-year data to the 3-year data of overall freedom from atrial fibrillation of 95% from your first presentation to 60% in this presentation.
The possible mechanisms for this I think are elucidated somewhat in your manuscript and would include increased surveillance, and I think that maybe you could elucidate that a little bit further. Were the patients who were found to have AF more frequently the patients who had continuous monitoring, or was it in the sporadic ECG [electrocardiographic] follow-ups?
Secondly, the resurrection of structural heart disease, either moderate MR [mitral regurgitation] that may indeed lead to atrial fibrillation later on or tricuspid regurgitation that develops in patients who have a previously repaired mitral valve, may indeed impact as atrial stretches the incidence of A-fib in late follow-up. Furthermore, the presence of hypertension or new ischemic heart disease, are any of these factors that may be culprits for a higher risk of A-fib later on?
And then finally, the other possible mechanism is that either the lesion set or the ablative energy technology over long-term follow-up is not as efficient. I would be interested in your comments regarding these three potential mechanisms.
Secondly, there has been a fair amount of interest in the last year or two about postablation testing for conduction block. Did you perform postablation testing, and if so, what were the results with that? And if you did not, why did you not perform this? Thank you.
DR GAMMIE: Thank you very much, Dr Groh. I appreciate your thoughtful comments. We have not routinely checked for exit conduction block after completion of the operation. We have in a handful of patients, and we have always found that there has been conduction block, and I think that the main reason we do not is that we have felt that it would not change what we do. These are endocardial ablations and we would not go back on bypass to ablate further.
I appreciate your comparison with our earlier experience, and, quite frankly the quality of our assessment has improved substantially from that time. Back then we were using spot ECGs and we really did not have long-term follow-up on a substantial number of patients. I would not necessarily conclude that the effectiveness of this operation falls off over time.
This report is a cross-sectional view. So we have taken that "freshman class," if you will, and looked at them at one point in time. What we are currently doing is enrolling patients in a prospective fashion. We are assessing their A-fib burden for a week before surgery so that we know ahead of time what that is, and then we are systematically following them out to 6, 12, and 24 months, and we have started to adapt our lesion set a bit. We will see what that shows. I think our earlier report reflected less complete follow-up, and I don't know that we can draw any firm conclusions about whether the rate of AF increases over time, stays the same, or decreases.
DR W. RANDOLPH CHITWOOD, JR (Greenville, NC): I rise to congratulate Dr Gammie on his early work, and he is continuing this important study of ways to surgically ablate atrial fibrillation. I heard his first presentation in Cancun several years regarding this important work. We embraced this technology several years ago as well. I have two questions and a moderately long comment.
We have to realize is that with atrial fibrillation, we are dealing with multiple morphologies: there are patients with large atria, ones with small atria, some with little mitral regurgitation, and some with significant mitral regurgitation. Onlaid with these morphologies are other factors including age, ischemia, and the presence of pulmonary hypertension. Thus, really there is a protean mismatch between different patients.
In our 250 patients that we have done minimally invasive ablative operations using endoscopic and robotic methods, we believe that we have gotten really good results. We have had about in the same type patient mix that you presented with 85% free of atrial fibrillation with about a 5% pacemaker rate. In patients with lone atrial fibrillation, of which we have 90 patients with continuous monitoring after a year of follow-up, for this group, 95% of patients are out of atrial fibrillation even though 5% of those patients are still on drugs. In other words, 90% of patients are off of warfarin [Bristol-Myers Squibb, Princeton, NJ] and drugs and are atrial fibrillation-free. So there must be a difference in the way we are doing the operation.
I applaud you for doing it on pump and the right way to recapitulate the classic Cox- Maze III operation. I question why all of us thought that trying to encircle and isolate the pulmonary veins a few years ago would work and say that was going to be it? We should have known better.
The questions I have for you are: Why didn't you freeze the coronary sinus? Broca suggested a more complete operation in the mid-1800s. He showed liberal muscle fibers in the coronary sinus, especially near the orifice. What did you do with the left atrial appendage? What about the length of time of freezing? I know your answer about exit block, because these lesions often mature later. The time of freezing does mean something, as we found that with a minute failed and two minutes was effective. Thank you for the opportunity to discuss this paper. I congratulate you on this on going study and encourage you to press on for the answers.
DR GAMMIE: Dr Chitwood, I appreciate your questions and I know that our group and your group do take a fairly similar approach. We do freeze the coronary sinus. One would ask the question: if you are seeing a 90% success rate and we are seeing less than that, what is the difference in terms of what we are doing? There are several potential explanations. One of them may be the duration of follow-up, one may be the intensity of follow-up. And I think a good example of that is that we think of the cut-and-sew as a traditional gold standard, and you can look at some good institutions and see some different outcomes. For example, the original Washington University series with a 5-year follow-up has a 96% freedom from AF, yet, many of those follow-ups are telephone questionnaires and spot ECGs. The Mayo Clinic reports a 50% AF-free rate 10 years after the cut-and-sew Maze procedure. And then you look at the cut-and-sew procedure from Dr Ballaux and colleagues in Europe. They have more than 200 patients followed up with Holter monitors at 4 years, and they have a normal sinus rhythm rate of 57%. So I think it depends how you look for it and how late after surgery you are looking for it.
We fully believe in freezing the coronary sinus. Our freeze is endocardial and we freeze for two minutes, and we believe that that is full thickness. One difference between our group and your group is that you are freezing from both the outside and the inside and sandwiching the coronary sinus, and that may explain the discrepant results.
DR CHITWOOD: I think it is important to freeze the coronary sinus. Now, some people are putting it in the coronary sinus. Freezing it endocardially I am not sure is doing it. Epicardially you freeze the hell out of that thing; it is frozen solid dead.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. F. Camm, M. Nagendran, P. Y. Xiu, and M. Maruthappu How effective is cryoablation for atrial fibrillation during concomitant cardiac surgery? Interact CardioVasc Thorac Surg, October 1, 2011; 13(4): 410 - 414. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Nakamura, H. Izutani, and Y. Sawa Mid-term outcomes of the modified Cox-maze procedure for elderly patients: a risk analysis for failure Interact CardioVasc Thorac Surg, June 1, 2011; 12(6): 924 - 928. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Cox The longstanding, persistent confusion surrounding surgery for atrial fibrillation J. Thorac. Cardiovasc. Surg., June 1, 2010; 139(6): 1374 - 1386. [Full Text] [PDF]
|
|
|
|
 |
|
 L. Sternik, D. Luria, M. Glikson, A. Malachy, M. First, and E. Raanani Efficacy of Surgical Ablation of Atrial Fibrillation in Patients With Rheumatic Heart Disease Ann. Thorac. Surg., May 1, 2010; 89(5): 1437 - 1442. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hanke, E. I. Charitos, U. Stierle, H.-H. Sievers, and M. Misfeld Intermediate Outcome of Surgical Atrial Fibrillation Correction With the CryoMaze Procedure Ann. Thorac. Surg., January 1, 2010; 89(1): 340 - 341. [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 |
