| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1994;108:1049-1055
© 1994 Mosby, Inc.
SURGERY FOR ACQUIRED HEART DISEASE |
Osaka, Japan
Supported in part by the Research Grant for Cardiovascular Diseases (6C-4) from the Ministry of Health and Welfare and by a grant from the Japan Cardiovascular Research Foundation (1992).
Address for reprints: Yoshio Kosakai, MD, National Cardiovascular Center, Fujishirodai 5-7-1, Suita 565, Osaka, Japan.
Abstract
Between April 1992 and October 1993, we combined a modified maze procedure with mitral valve repair (n = 26) or replacement (n = 36) in 62 patients with atrial fibrillation, including 16 patients undergoing reoperation. Associated procedures included aortic valve operation (n = 22), tricuspid annuloplasty (n = 28), atrial plication (n = 10), and others (n = 3). Duration of atrial fibrillation varied from 0.1 to 23 years (average 8.3 ± 6.4 years), the f-wave voltage ranged from 0 to 0.45 mV (0.16 ± 0.09 mV), and cardiothoracic ratio varied from 46% to 85% (64% ± 9%). We modified the maze atriotomies to preserve the sinus node artery and used cryoablation to simplify procedures. Aortic crossclamp time was 142 ± 25 minutes and cardiopulmonary bypass time 226 ± 34 minutes. No early or late deaths occurred in a total of 783 patient-months of follow-up. In 52 patients (84%) who regained atrial rhythm, an atrial A-wave was detected in 84% for transtricuspid flow and in 71% for transmitral flow. One patient with sinus rhythm had an episode of transient neurologic ischemia 4 months after mechanical valve implantation. The 10 patients who remained in atrial fibrillation had preoperative fibrillation for a significantly longer time than the rest of the patients with atrial rhythm (14.8 versus 7.2 years, p < 0.001) and a larger left atrial dimension (70 versus 58 mm, p < 0.01). Nonetheless, no variable alone could have predicted postoperative rhythm for individual patients. The results suggest that the maze procedure is safe and effective and therefore should be considered for patients with chronic atrial fibrillation undergoing mitral valve operations. (J THORACCARDIOVASCSURG1994;108:1049-55)
Atrial fibrillation (AF) not only increases the risk of thromboembolism but also undermines hemodynamics, especially in patients with mitral valve disease. With persistent AF, patients often remain symptomatic and require permanent anticoagulation even after an otherwise successful mitral valve operation. Cox and colleagues 
1,2 developed the maze procedure and proved its safety and efficacy in patients mainly with isolated AF. 3 However, in patients undergoing concomitant mitral valve operation, this procedure may not necessarily be as safe because of its complexity. 1-3 Also it may not be as effective because of various factors 4-9 inherent in mitral valve diseases that predispose the patients to persistent AF. Because of the large number of patients who have this vicious combination, we started combining the maze procedure to control secondary AF with a mitral valve operation. This study describes our experience with a modified maze procedure in the first 62 patients with chronic AF undergoing mitral valve surgery.
PATIENTS AND METHODS
From April 1992 to October 1993, 62 patients with AF caused by mitral valve disorders underwent a modified maze procedure combined with a mitral valve operation, replacement in 36 and repair in 26. The group comprised 30 men and 32 women, their ages ranging from 32 to 75 years with an average of 56.3 years. The causes of the mitral valve disorder were rheumatic (n = 48), degenerative (n = 12), and congenital (n = 2) disease. Thirteen patients had a giant left atrium. 10 Duration of documented AF ranged from 0.1 to 23.0 years with an average of 8.3 ± 6.4 years (mean ± standard deviation); one (1.6%) less than 3 months and six (9.7%) less than 1 year with 45 (72.6%) having had AF for longer than 3 years. Maximum f-wave voltage in the V1 lead varied from undetectable, or 0, to 0.45 mV with an average of 0.16 ± 0.09 mV. Concomitant procedures included aortic valve procedures (n = 22), tricuspid annuloplasty (n = 28), left atrial plication (n = 10), atrial septal defect closure (n = 1), endocardial cushion defect repair (n = 1), and coronary artery bypass grafting (n = 1). Sixteen patients underwent reoperation (25.8%), 10 because of prosthetic valve failure and six after mitral valve repair.
We modified the Cox maze 2 procedure (first modification 11)and applied it in the first 11 patients 12 selected on the basis of high f-wave voltage. We further modified the atriotomies and used cryoablation to preserve the sinus node artery, 13 asin the current procedure 12 (Kosakai's modification), for the next 41 patients. Ten additional patients underwent the Cox maze 3 procedure (second modification 11) combined with mitral valve operations.
Our current modification is illustrated in Fig. 1. Cardiopulmonary bypass is instituted with bicaval venous drainage, direct cannulation of the superior vena cava, and cannulation of the inferior vena cava via the lower right atrium (Fig. 1, A). With the patient supported by total bypass, direct-current cardioversion is used to examine sinus node function for patients without left atrial thrombi. The right atrial appendage is amputated. From the incision the right atriotomy is extended in a curvilinear fashion to the junction with the inferior vena cava, which is cryoablated after institution of cardiac arrest (Fig. 1, A). From the midpoint of this atriotomy, an additional incision is started toward the tricuspid anulus, which is cryoablated later. The superior vena cava is transected 14 distal to the junction with the right atrium. After institution of cardiac arrest, the left atrium is entered in front of the right pulmonary veins as in the regular mitral valve operation. The left atriotomy is then extended to encircle the orifices of the pulmonary veins. At the circumferential atriotomy, the left ventricle is disconnected from the pulmonary veins and suspended with a left atrial cuff to facilitate atriotomy and cryoablation toward the mitral anulus instead of dissection and reanastomosis (Fig. 1, B).
|
Cardiac rhythm was continuously monitored after the operation until stable rhythm returned. Temporary wires were used to pace the heart, to monitor the rhythm, or to overdrive the atrium in case of atrial tachyarrhythmia. External direct-current cardioversion was carried out to terminate hemodynamically deleterious atrial tachyarrhythmia only after overdrive pacing and pharmacologic suppression had failed. Patients with persistent AF were first treated with antiarrhythmic agents, usually class Ia. If they did not respond, electrical cardioversion was done before discharge and medications were continued.
After discharge, patients were followed up monthly so that cardiac medication, rhythm, signs of myocardial ischemia, and anticoagulation could be monitored. Doppler echocardiography and chest roentgenograms were repeated 1, 3, 6, and 12 months after operation. Anticoagulation was discontinued only in patients with atrial rhythm and contraction 3 to 6 months after the reparative operation. Antiarrhythmic agents were tapered gradually after anticoagulation was discontinued or after atrial rhythm was considered stable.
Continuous variables were compared by the t test and discrete variables were analyzed with the 
2 test. Differenceswere considered statistically significant when the p value was less than 0.05.
RESULTS
No operative or hospital deaths occurred. Intraoperative direct cardioversion with the patient supported by bypass failed in 27 patients and resulted in defibrillation in 23 patients. Three of the 23 (4.8%) had sinus node dysfunction but eventually became symptom-free. Cardiac arrest time ranged from 92 to 212 minutes with an average of 142 ± 25 minutes, necessitating cardiopulmonary bypass runs varying from 148 to 295 minutes with an average of 226 ± 34 minutes. Immediately after the operation, AF disappeared in 60 patients (96.7%): 36 had sinus rhythm and 24 had junctional rhythm, with two patients never having had defibrillation. Most patients required right atrial pacing to be weaned from bypass and for stabilization of hemodynamics. Three patients (4.8%) required intraaortic balloon pumping, and five (8.1%) underwent reexploration either for hemostasis (n = 4) or for relief of late cardiac tamponade (n = 1). No patients had postoperative conduction problems, but two (3.2%) with sinus node dysfunction after the modified Cox maze 2 procedure required atrial pacemaker implantation; their preoperative function was unknown because preoperative cardioversion was unsuccessful.
No late deaths have occurred in follow-up times ranging from 6 to 25 months, a total of 783 patient-months. One patient in sinus rhythm had a transient neurologic ischemic attack 4 months after mitral valve replacement with a mechanical valve. Another patient had cerebral bleeding 1 month after mitral valve replacement. Restoration of atrial rhythm or freedom from AF is illustrated for every consecutive 3-month period (Fig. 2). Twenty-seven patients (43.5%) had temporary or persistent AF or atrial flutter after the operation, mainly for the first 2 weeks after the operation. Seventeen of them eventually regained atrial rhythm after pharmacologic or electric cardioversion. Ten patients (16%) had persistent AF, seven after the current modification and three after the Cox maze 3 procedure. These 10 patients were compared with the 52 patients (84%) with atrial rhythm (Table I). Significant differences were noted in history of AF, echocardiographic left atrial dimension, and association of giant left atrium. However, none of the preoperative variables alone could have predicted restoration of atrial rhythm for individual patients.
|
|
Nineteen patients underwent selective coronary angiography 1 month after the operation; none of them had abnormalities around areas 2 and 13, close to where cryoablation was performed. The sinus node artery variation was the right in 25, the left in 12, the posterior in 9, and unidentified in 16. Postoperatively, arterial supply changed to the right sinus node artery in 3 patients, 2 from the posterior and 1 from the left sinus node artery. In later follow-up, no patients had signs of myocardial ischemia on the electrocardiogram, blood chemistry results, or symptoms indicating coronary angiographic examination.
DISCUSSION
Patients with a mitral valve disorder often have factors predisposing to AF, such as advanced age, 8 cardiomegaly, 8 increased atrial size, 9 and long history of AF. 4-8 These factors tend to result in persistent AF even after an otherwise successful mitral valve operation 4-7 and electric cardioversion. 6 Except for patients with recentonset of AF, 4-8 therefore, specific surgical intervention is mandatory to reverse AF resulting from mitral valve disease. Because restoration of atrial rhythm has long been desired 4-7 and is expected to improve long-term prognosis, we started including the maze procedure for patients undergoing mitral valve operations.
To lessen the risk by reducing operating time, we used cryoablation to replace atriotomy, reanastomosis, and dissection at the valvular anuli. 1-3 Despite concerns about cryothermia close to anonperfused coronary artery, 15,16 no early coronary lesions were detected in the 19 patients examined. Although no cases of myocardial ischemia suggestive of delayed lesions 16 have been identified in later follow-up, it may be prudent to rule out such lesions by coronary angiography, because the heart has been found to be denervated in such patients. 17 Transection of the superiorvena cava 14 and detachment of the left ventricle improved exposure and facilitated manipulation of the mitral valve. Moreover, tricuspid valve lesions are often associated with mitral disorders, necessitating a right atriotomy for surgical intervention. Thus combining the maze procedure with a mitral valve operation does not necessarily require as much added time and complexity as would performing each operation separately. Nonetheless, the combined maze procedure necessitated an additional 40 minutes for cardiac arrest and an extra 70 minutes for cardiopulmonary bypass as compared with the valvular procedures alone. 18 Although mortality has fortunately been avoided in the current series, the added complexity and length of operation naturally increase the risk, which has to be weighed against the benefits.
The original maze procedure was developed in the dog, 1,2 in which the sinus node artery usually arises from the right coronary artery coursing along the right atrial free wall to the sinus node. 13 Because atriotomies in theoriginal maze procedure 1,2 transect all major variations of the human sinus node artery, 13 we modified atriotomies and used cryoablation to avoid disrupting the arterial blood supply to the sinus node. In 16 of 19 patients with paired coronary angiograms the preoperative variations were preserved. In three patients the sinus node artery might have been transected because the posterior and left variations are close to the left atriotomies. With the current modification, recovery of atrial electric activity and contraction occurred in the operating room, earlier than in our experience with the modified Cox maze 2 procedure for the initial 11 cases. The risk of sinus node dysfunction necessitating pacemaker implantation was significantly less with the current modification (0/41) than with our experience with the Cox maze 2 procedure (2/11) or with that of Cox's group. 3 Although preoperative sinus node dysfunction was less prevalent in our patients, preservation of arterial supply may account at least in part for the reduced incidence of postoperative sinus node dysfunction.
The current series consisted mainly of patients with rheumatic AF (77%), which may account for the lower defibrillation rate and less frequent detectable atrial contraction than reported by Cox and colleagues. 3 Three of 10 patients undergoing the modified Cox maze 3 procedure 11 in the present series had persistent AF, suggesting that the differences mainly derive from underlying disease rather than modification in atriotomies or use of cryoablation. Doppler echocardiography failed to detect atrial contraction in about one fourth of our patients who regained atrial rhythm. We assume that fibrotic and calcific degeneration of the atrial myocardium as a result of severe dilatation or rheumatic myocarditis is responsible for the lack of forceful contraction detectable by transthoracic Doppler technique, despite recovery of electric atrioventricular synchrony. Even for these patients, we consider it worthwhile to have performed the maze procedure, because the atrium functions more as a conduit after restoration of atrial rhythm. An 84% defibrillation rate is highly significant when compared with the defibrillation rate of 15% among 176 patients undergoing mitral valve operations alone in our institution from 1988 through 1989. The combined approach also appeared worthwhile as compared with a case-control group of patients undergoing valvular operation alone without the maze procedure 18 and other studies reportingresults from surgical 4-7 and electric cardioversion 8,9 alone.
More than half of the patients had continued to need antiarrhythmic drugs either to defibrillate or to stabilize atrial rhythm before anticoagulation could be discontinued. Then we consider tapering the antiarrythmic agents, which are associated with a lower risk and are better tolerated. As a result, 14 patients with atrial rhythm and contraction, representing 67% of 21 patients who had had reparative surgery, stopped receiving anticoagulation, including two who received no medication.
In the beginning we selected patients for the maze procedure only on the basis of efficacy and safety. Increasing experience, however, encouraged us to expand the indication to include reoperation, emergency cases, and patients who were judged able to tolerate extended cardiac arrest and bypass periods, although adhesions and calcified atrial walls continue to pose technical difficulties. Univariate analysis on preoperative variables identified longer history and larger atrial dimension as predisposing factors. Although exclusion of patients with a high probability of persistent AF is reasonable to reduce unsuccessful attempts and associated risks, concomitant reduction in atrial size or further modification may prove effective. In longer follow-up, restoration of atrial rhythm and contraction may abolish stagnation and necessity for systemic anticoagulation, resulting in a reduced risk of thromboembolism and bleeding complications. When these results and perspectives are weighed against surgical risks and technical difficulties, we believe that every patient with AF undergoing a mitral valve operation can be considered for the combined maze procedure.
Appendix: DISCUSSION
Dr. Randas J. V. Batista (Curitiba, Brazil)
Dr. Kosaki, our experience has been somewhat different from yours. In the past 4 years, our patients with mitral valve disease and AF have been treated differently.
We have done "cardiac autotransplantation" with reduction of the left atrium in 45 patients. All patients had enlarged or giant left atrium. To reduce the size of the left atrium and repair the mitral valve with the heart in a basin is simple [video]. Most of the 45 patients treated were having reoperations. We used continuous normothermic retrograde blood cardioplegia in all of them. All of the patients left the operating room in sinus rhythm and none required antiarrhythmic or inotropic drugs.
In your experience only 40% of your patients are in sinus rhythm and free of antiarrhythmic drugs. Perhaps you can explain to me the reason for this difference.
Dr. Kosakai.
I thank Dr. Batista for showing us the excellent results of his autotransplantation procedure.
Because we realize that residual AF relates to the postoperative size of the left atrium, we resect as much of the left atrial wall as possible. However, even with this procedure, 44% of our patients had various kinds of atrial arrhythmias during the first postoperative month. Therefore, we continue using class Ia antiarrhythmics until discontinuation of anticoagulant. This is the reason we are treating a large number of our patients with antiarrhythmic drugs after the operation, although the dosages are tapered thereafter.
One of the reasons for your excellent success in eliminating AF may be the reduction of the atrial size. However, if you are treating the same kind of patients that we are treating, I cannot explain the difference in the results.
João Q Melo (Lisbon, Portugal).
We have been performing the maze procedure together with mitral operations for the past 18 months. We have operated on eight selected patients, meaning that those patients had chronic AF for more than 1 year and had previous multiple thromboembolic episodes.
On these patients we did four mitral repairs and four mitral replacements. After a mean follow-up of 7 months, two patients have sinus rhythm, five have atrial rhythm, and one patient, who had a giant left atrium, has AF. One patient had a pacemaker implant, and only the patient with AF is receiving antiarrhythmic drugs.
By Doppler echocardiography we have seen biphasic flow across the tricuspid valve in diastole. We have also seen biphasic flow through the mitral valve, but across the mitral valve the a wave is very small and very delayed. It is really a small protosystolic left atrial contraction.
I have one question about anticoagulation. Do you stop it in every patient or do you document the left atrial contraction before stopping it? If you see a contraction such as I described, do you stop anticoagulants?
Dr. Kosakai.
We usually discontinue anticoagulation 3 to 6 months after repair in patients with stable atrial rhythm and demonstrable contraction. As you mentioned, the left atrial contraction is less frequent than the right in rheumatic patients. For patients with a tiny or undetectable left atrial contraction, I try to continue anticoagulation.
Dr. James L. Cox (St. Louis, Mo.).
I want to congratulate Dr. Kosakai on what I think is an excellent series of patients. His results are a bit different from those in our own series, and I suspect that the reason is his extensive use of cryosurgery in the area of the coronary sinus.
Dr. Kosakai, it is my understanding from your associates, Dr. Isobe and Dr. Kawashima, that you actually use cryosurgery rather than a longitudinal incision over the coronary sinus posteriorly. That would seem to be a good idea from the standpoint of cutting down the amount of time required for the operation. However, I think that is probably the reason that postoperative atrial flutter is more prevalent in your patients. We have had a few recurrences of atrial flutter, and it is usually because of a return of conduction along the coronary sinus.
With regard to the overall results, it is my understanding that none of your patients operated on more than 10 months ago has AF. Is that correct?
I would like to make two other comments. Dr. Batista has described an autotransplantation operation that is quite an accomplishment but a bit too daring for me. I suspect that the operation is working for two reasons: One is that he is excising a significant portion of the atrial muscle, and the atria fibrillate because of the availability of a critical mass or area; the second is that he is totally denervating the heart. Neither one of those two factors alone will stop the atrium from fibrillating, but the two together apparently do in Dr. Batista's experience.
Finally, Dr. Melo mentioned a late a wave on the left side. We found the same thing with the original maze procedure, and that was one of the main reasons we modified it. We have seen it in only two patients of 60 since modifying the maze procedure.
Dr. Kosakai.
Dr. Cox, I must apologize for having presented a somewhat misleading figure. We have followed 11 patients for more than 18 months. The initial series of patients were selected rather strictly on the basis of the F-wave voltage. All those patients remained in sinus rhythm 18 to 24 months after the operation. Encouraged by these excellent results in the initial series, we have expanded our indications for this operation to those with low F-wave voltage, and we assume this is the reason for our rather high percentage of patients who are remaining in AF, although all of them are in much better condition because of concomitant mitral valve surgery.
Before closing, I would like to express our appreciation to Dr. Cox for his help in starting our program with this procedure at the National Cardiovascular Center in Osaka.
Acknowledgments
We appreciate the editorial help of Dr. Leonard M. Linde, Professor of Pediatrics (Cardiology), University of Southern California, Los Angeles, California.
Footnotes
Read at the Seventy-fourth Annual Meeting of The American Association for Thoracic Surgery, New York, N.Y., April 24-27, 1994. 
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
). Freedom from atrial fibrillation illustrated in each 3-month period after operation. Figures depict the number of patients free from atrial fibrillation (numerator) and the number of patients at risk in each period (denominator). The initial 11 patients were all in atrial rhythm, probably because they were selected for the procedure.