| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;82:494-501
© 2006 The Society of Thoracic Surgeons
Division of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minnesota
Accepted for publication March 24, 2006.
* Address correspondence to Dr Schaff, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905 (Email: schaff{at}mayo.edu).
Presented at the Fifty-second Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 10–12, 2005.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
METHODS: Four hundred forty-three patients underwent the Cox-maze procedure from 1993 to 2002. Ninety-nine had atrial flutter or fibrillation without associated valvular or congenital heart disease, and 37 (37%) had decreased left ventricular function (ejection fraction < 0.35 in 11 [severe], ejection fraction 0.36 to 0.45 in 8 [moderate], and ejection fraction 0.46 to 0.55 in 18 [mild]). Ages of these 37 patients (34 male) ranged from 35 to 74 years (median, 55 years).
RESULTS: Atrial flutter or fibrillation was present for 3 months to 19 years (median, 48 months) preoperatively, and 24 patients (65%) exhibited symptoms of heart failure. Preoperative ejection fraction ranged from 0.25 to 0.55 (median, 0.45). At last follow-up (median, 63 months), the Cox-maze procedure eliminated atrial flutter or fibrillation in all but 1 patient, and the greatest improvement was observed in patients with severe preoperative impairment (0.31 to 0.53; p = 0.01, preoperative versus follow-up), and patients with preoperative chronic atrial flutter or fibrillation (0.43 to 0.55; p < 0.05 preoperative versus follow-up). This improvement was observed immediately postoperatively and was sustained at last follow-up. Further, improvement in left ventricular function correlated with enhancement of functional status.
CONCLUSIONS: In some patients, atrial flutter or fibrillation may be the cause rather than the consequence of left ventricular dysfunction. Importantly, systolic function and functional status can be significantly improved with the restoration of sinus rhythm by the Cox-maze procedure.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
In addition, rapid heart rate caused by AF can lead to cardiomyopathy [4–6]. Multiple reports have documented that left ventricular (LV) dysfunction caused by supraventricular tachycardia can be cured or improved by conversion to sinus rhythm [7, 8]. Also, ventricular dysfunction may resolve with control of tachycardia by the ablation of the atrioventricular node and insertion of the pacemaker [9].
The maze procedure developed by Cox and colleagues [10–14] corrects AF in more than 90% of patients, and the operation has been used in patients with symptomatic tachycardia, thromboembolic stroke, and AF associated with valvular and congenital heart disease [15–17]. However, the effect of the Cox-maze procedure on improving cardiac performance has not been well defined. We have encountered several patients with marked reduction in LV ejection fraction (EF) who were referred for operation to correct AF. The purpose of this study was to examine hemodynamic and functional outcome of patients with LV dysfunction undergoing operation for AF.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Preoperative patient characteristics are summarized in Table 1. The 34 men and 3 women had a median age of 55 years (range, 35 to 74 years). Twenty-six (73%) patients were older than 50 years. Preoperative AF was chronic in 23 patients and paroxysmal in 14. Preoperatively, the patients, as a group, complained of a variety of symptoms including fatigue, shortness of breath, palpitations, syncope, transient ischemic attack, stroke, angina, and congestive heart failure. Importantly, 24 patients (65%) had symptoms attributable to heart failure.
|
The patients were stratified according to preoperative EF; 11 patients had severe LV dysfunction (EF < 0.35), 8 had moderate LV dysfunction (EF 0.36 to 0.45), and 18 had mild LV impairment (EF 0.46 to 0.55). In all, preoperative EF ranged from 0.25 to 0.55 (median, 0.45), and the mean was 0.439 ± 0.03.
The major indication for operation was intolerance to arrhythmia, leading to fatigue and palpitations, and this was true for patients with either chronic or paroxysmal AF. Another indication for surgical treatment of AF was prior stroke (2 patients) or transient ischemic attack (1 patient). All patients had antiarrhythmic drug therapy preoperatively (mean of 1.4 ± 0.9 medications per patient); 15 patients had attempted previous electrical cardioversion, and 2 had previous radiofrequency ablation therapy in attempt to control the AF. Thus, patients considered for the Cox-maze operation had failed nonsurgical methods of treating the troublesome AF.
For patients who were in chronic AF, the median preoperative resting heart rate was 88 beats per minute (range, 58 to 144 beats per minute). Thus, many, but not all, had inadequate heart rate control despite close medical supervision and aggressive pharmacologic antiarrhythmic therapy. All patients with chronic AF were in either atrial fibrillation (18 patients) or atrial flutter (5 patients) at the time of surgery.
For patients with paroxysmal AF, the median preoperative resting heart rate was 62 beats per minute (range, 47 to 94 beats per minute), and during paroxysmal episodes of arrhythmia, the median heart rate measured was 84 beats per minute (range, 66 to 163 beats per minute). Four patients of 14 with paroxysmal AF were in either AF (3 patients) or atrial flutter (1 patient) at the time of surgery.
A standard "cut and sew" maze procedure was performed as described by Cox and associates [10–14] using cardiopulmonary bypass with moderate hypothermia (28° to 32°C) in 23 patients and mild hypothermia or normothermia in the remaining 14 patients. The right-sided maze incisions were performed before aortic cross-clamping, and the left atrial incisions were performed after cardiac arrest with cold blood cardioplegia.
We have modified the original Cox-maze procedure in two ways. On the medial aspect of the right atrium, we apply a linear cryolesion from the cut edge of the appendage to the tricuspid valve rather than incising the atrium; this avoids division of the frequently seen branch of the right coronary artery that supplies the sinoatrial node. We have found that the risk of postoperative sinus node dysfunction can be reduced by using a cryolesion instead of an incision in this location. In the left atrium, we prefer to extend the incision that encircles the pulmonary veins to the orifice of the left atrial appendage, and then close the orifice transversely as part of the encircling incision. Alternatively, cryolesions can be used as part of the lesion encircling the pulmonary veins to avoid the conjunction of the encircling suture line and left atrial appendage suture line.
Cardiac rhythm was monitored continuously after operation, and temporary epicardial wires were used for pacing as needed. All patients were dismissed with warfarin for 3 months, after which it was discontinued, except for those with persistent or recurrent atrial arrhythmia.
Demographic and other patient-related data were obtained from Mayo Clinic medical records. Follow-up information was obtained from subsequent clinic visits, written correspondence from local physicians, and mailed questionnaires to patients or families. Data were expressed as mean ± standard error of the mean, and statistical significance was considered at p less than 0.05. Early operative mortality was defined as death occurring within 30 days of operation or at any time during the index hospitalization.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
During the median follow-up period of 48 months (range, 1 to 141 months), 4 patients experienced recurrent atrial arrhythmias; 2 had atrial flutter, 1 had atrial fibrillation, and another had atrial tachycardia. Cardioversion to sinus rhythm in 2 patients led to improvement in their EFs, but EF remained depressed in the patient whose atrial flutter persisted despite electrical cardioversion and attempts at radiofrequency ablation. This was the only patient who had congestive heart failure after operation. This patient subsequently underwent atrioventricular node ablation with placement of a permanent pacemaker. At last follow-up, 28 patients were in sinus rhythm, 4 were in a paced rhythm, 2 patients remained in AF, and 3 were in a junctional rhythm (Table 3). Table 4 demonstrates rhythm at last follow-up for patients separated according to type of preoperative AF. There were 2 late deaths, 1 as a result of lymphoma and 1 as a result of metastatic prostate cancer. Both patients were in sinus rhythm at the time of death.
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Recent studies have addressed the issue of rate versus rhythm control in atrial fibrillation [22–24]. The Atrial Fibrillation Follow-up Investigation of Rhythm Management and Rate Control versus Electrical cardioversion trials demonstrated a trend toward rate control versus rhythm control for the composite end points of death, stroke, and recurrent hospitalization. Results from these trials must be interpreted in the context of the patient population studied. The study population generally comprised patients who were tolerant enough of AF to be randomized. Although there was clear underrepresentation of patients with heart failure, in patients with congestive heart failure in the Atrial Fibrillation Follow-up Investigation of Rhythm Management trial, the two strategies of rate versus rhythm control were similar.
Atrial flutter or fibrillation impairs hemodynamic function by several mechanisms. First, AF results in loss of atrioventricular synchrony and atrial contraction. This may reduce ventricular filling and thereby reduce cardiac output. The consequence of loss of atrial contraction may be especially pronounced in patients with impaired diastolic function as is seen with hypertrophied ventricles, restrictive cardiomyopathy, and mitral valve stenosis. Fluctuation in the R-R interval changes diastolic filling interval, producing a variable stroke volume. In animals, cardiac output is reduced 15% when ventricular rhythm is irregular compared with a regular rhythm at the same heart rate [25].
In addition to these mechanical consequences, AF can lead to tachycardia-induced cardiomyopathy [26–30]. Cardiomyopathy caused by tachycardia is commonly thought to be associated with chronic arrhythmias having rates greater than 120 beats per minute. Our experience demonstrates that ventricular dysfunction may be associated with resting heart rates considerably lower than this, and, further, paroxysmal AF can lead to ventricular dysfunction. Patients in this series with preoperative chronic AF had a median resting heart rate of 88 beats per minute (range, 58 to 144 beats per minute). Further, among patients with preoperative paroxysmal AF, the median preoperative resting heart rate was 62 beats per minute (range, 47 to 94 beats per minute), and during paroxysmal episodes of arrhythmia, the median heart rate measured was 84 beats per minute (range, 66 to 163 beats per minute). It is important to recognize that resting heart rate is not truly reflective of overall heart rate in patients with atrial fibrillation, because the heart rate response to exercise may vary [31]. Other studies have documented the finding that patients with well-controlled resting heart rates may have a rapid ventricular response with minimal activity, and develop tachycardia-induced cardiomyopathy [31].
The effects of tachycardia on ventricular function have largely been derived from studies examining patients with chronic atrial fibrillation and a rapid ventricular response. Adequate rate control varies with age, but ventricular responses between 60 and 80 beats per minute at rest, and between 90 and 115 beats per minute during moderate exercise, are the usual goals. In patients with systolic dysfunction, rates higher than these ranges are considered tachycardic.
The identification of patients with tachycardia-induced cardiomyopathy may be challenging, and careful search for structural heart disease is necessary. The absence of structural anomalies and the documentation of AF before LV dysfunction make the diagnosis of tachycardia-induced cardiomyopathy more likely. Other patients may present with AF and LV impairment, and it is unclear whether the arrhythmia is causative or a consequence of ventricular dysfunction. In such patients medical management is initiated, and cardioversion should be considered. If the EF improves when the rate is more adequately controlled, the diagnosis again is more likely.
If tachycardia-induced cardiomyopathy is indeed identified, the question arises, "how much LV dysfunction is too much?" In a patient with critical impairment of LV function, strict medical rate control with attempted cardioversion is first-line treatment, with catheter-based techniques as necessary to attempt to ablate the arrhythmia. It would be reassuring in this patient situation to once again document improvement during times of successful rate control. If a structural cardiac defect is identified, and improvement in EF can be documented with rate control, surgery should be considered.
It is important to identify patients with tachycardia-induced cardiomyopathy because ventricular dysfunction can be reversed by control of the arrhythmia; this has been documented in patients having cardioversion (medical and electrical) to sinus rhythm and in patients having rate control with ablation of the atrioventricular node coupled with implantation of a transvenous pacemaker. Table 7 summarizes results of changes in EF after conversion of AF or control of heart rate, and it should be noted that patients in this series having the Cox-maze operation had similar degrees of improvement in ventricular function as did patients managed without operation.
|
It is important to use diuretics liberally early after operation. Removal of the atrial appendages during the Cox-maze procedure eliminates an important source of atrial natriuretic peptide, and this, along with elevations of aldosterone and antidiuretic hormone early postoperatively, may predispose the patient to fluid retention.
We have used systemic anticoagulation with warfarin for 3 months postoperatively, but there is no consensus on the need for anticoagulation beyond this interval. Some clinicians prefer to continue warfarin, believing that risk of thromboembolism is not reduced sufficiently to avoid systemic anticoagulation. Others argue that if AF is eliminated and ventricular function is normal, the risk of an intracardiac source of thromboemboli from a postoperative patient without a left atrial appendage is very low, and the additional risk and inconvenience of using warfarin is not justified.
Although a number of patients in our series had a junctional rhythm early postoperatively, we do not routinely use antiarrhythmic medications or stimulants, such as theophylline, as have been suggested. Several weeks or even months may be required for return of a stable sinus mechanism. In such patients, we would defer permanent pacemaker implantation as long as the patient remains asymptomatic, has no prolonged pauses, and has an adequate rate response to exercise. Some patients with atrial fibrillation have underlying sinus node dysfunction, and conversion of the arrhythmia by operation or medical treatment may lead to symptomatic bradycardia requiring permanent pacing.
Generally accepted indications for Cox maze operation include troublesome symptoms as a result of tachycardia, thromboembolic stroke, and intolerance to antiarrhythmic medications. Indeed, it was previously stated that LV dysfunction was the only contraindication to the procedure [12, 32]. Our results demonstrate that even patients with moderate LV impairment benefit in terms of ventricular function immediately after the maze procedure, although this improvement lost statistical significance at late follow-up. This experience suggests that surgical treatment of AF should be considered in select patients with cardiomyopathy, particularly those in whom onset of tachycardia precedes or is known to coincide with development of LV dysfunction. Operation should also be considered in patients with AF and LV dysfunction who are referred for atrioventricular node ablation [33].
This study was retrospective, and patients were, of course, carefully selected. It is unclear how the results can be generalized to larger groups of patients with LV dysfunction and atrial fibrillation. Detailed information on duration and severity of tachycardia was not available in all patients in this study. Nevertheless, our finding that LV dysfunction improves after successful Cox-maze operation should broaden consideration of surgical treatment of AF.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
If I am not mistaken, this is one of the first surgical series to demonstrate that ventricular function actually improves after the maze procedure. Congratulations on another landmark paper that has been presented to the Southern Thoracic Surgical Association.
At Children's Memorial Hospital in Chicago we have some experience with atrial fibrillation in patients with a single ventricle, who tolerate atrial fibrillation poorly. After Fontan revision, the ventricular function in a large number of patients improved after a maze procedure as well.
There is, however, one problem that we and others have had. Namely, do the patients that present with atrial fibrillation have attendant ventricular dysfunction due to atrial fibrillation which is treatable by a maze operation or due to myocardiopathy which might best be treated by cardiac transplantation? We have a very difficult time figuring this out.
Now, I am not so sure that your data will allow you to answer this question because very few of your patients had an ejection fraction less than 35%. That having been said, would you tackle this question and then perhaps give us some idea on how we could actually answer this significant problem going forward, because the maze procedure is going to be applied to more and more patients with decreased ventricular dysfunction. Thank you.
DR STULAK: Thank you, Dr Mavroudis, I appreciate your kind words. To my knowledge, this is the first surgical series documenting improved systolic function after the Cox-maze procedure in patients with tachycardia-induced cardiomyopathy.
Your second comment gets back to the question of, in the setting of congestive heart failure, whether atrial fibrillation is the cause or consequence, and this can be very challenging. A careful search for structural heart disease that may account for the patient's heart failure is the first step. The absence of structural heart disease and the documentation of atrial fibrillation prior to left ventricular dysfunction make the diagnosis of tachycardia-induced cardiomyopathy more likely. Other patients may present with concomitant AF (atrial flutter or fibrillation) and LV (left ventricular) dysfunction, and it is unclear whether the AF is causative or a consequence. In these patients, medical management is initiated, and cardioversion should be considered. If the ejection fraction improves when the rate is adequately controlled, the diagnosis is again more likely.
For example, in many of our patients with paroxysmal AF, when they had that resting heart rate in the 60s, they had a good ejection fraction, but it was during paroxysms of atrial fibrillation that they had decreased LV function. So that again is one way to tell.
DR CLIFFORD VAN METER (New Orleans, LA): Excellent paper, excellent question to look for in the first place.
It stands to reason that the worst ventricles have the most room to improve. So that question sort of answers itself. But moreover, in these two groups, do you have any data on left atrial size, and was reduction in left atrial size part of the operation that you performed, if necessary?
DR STULAK: That is an excellent point. As part of the standard Cox-maze procedure, these patients had their atrial appendages amputated, but no further left atrial reduction plasty was performed. This is definitely something that can be considered in patients with a dilated left atrium, as this attenuated atrial tissue will continue to serve as an arrhythmogenic substrate postoperatively. In terms of data on their left atrial size, a lot of these patients had hypertrophied ventricles, and their atria were not significantly enlarged.
DR VAN METER: As a follow-up to that and whether or not Coumadin (warfarin) is continued or not, a lot of electrophysiologists are looking at whether there is left atrial excursion. Did you examine for that in these groups to see whether that contributed to ventricular function in terms of dynamic filling?
DR STULAK: Unfortunately, this was a retrospective review, and we did not gather information specifically examining atrial function. That is an excellent point.
DR D. GLENN PENNINGTON (Johnson City, TN): I just wanted to know if you looked at mitral valve function and did indeed sinus rhythm improve AV (atrioventricular) regurgitation? To what extent was mitral insufficiency involved in the outcome?
DR STULAK: These patients were highly selected such that patients with congenital and valvular heart disease were not included in this study because this pathology may have contributed to their decreased LV function. We wanted to examine patients who had no other potential cause of heart failure so that tachycardia-induced cardiomyopathy was the only diagnosis responsible. So of the 443 patients who underwent the Cox-maze operation at our institution during the study period, 344 had either valvular or congenital heart disease. We took those out of the equation. There remained 99 patients who had atrial fibrillation without concomitant structural heart disease, and of those, 37 patients had evidence of LV dysfunction. Further, only 7 were found to have coronary artery disease, which was not thought to contribute to decreased LV function. So, these patients were chosen so as to remove all other potential causes for their heart failure. So, to answer your question, these patients didn't have any evidence of mitral valve disease.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Roy, M. Talajic, S. Nattel, D. G. Wyse, P. Dorian, K. L. Lee, M. G. Bourassa, J. M. O. Arnold, A. E. Buxton, A. J. Camm, et al. Rhythm Control versus Rate Control for Atrial Fibrillation and Heart Failure N. Engl. J. Med., June 19, 2008; 358(25): 2667 - 2677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Onorati, A. Esposito, G. Messina, A. di Virgilio, and A. Renzulli Right Isthmus Ablation Reduces Supraventricular Arrhythmias After Surgery for Chronic Atrial Fibrillation Ann. Thorac. Surg., January 1, 2008; 85(1): 39 - 48. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-R. Neuberger, C. Mewis, D. J. van Veldhuisen, U. Schotten, I. C. van Gelder, M. A. Allessie, and M. Bohm Management of atrial fibrillation in patients with heart failure Eur. Heart J., November 1, 2007; 28(21): 2568 - 2577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Onorati, M. Bilotta, F. Borrello, M. Vatrano, A. di Virgilio, M. C. Comi, F. Perticone, and A. Renzulli Successful radiofrequency ablation determines atrio-ventricular remodelling and improves systo-diastolic function at tissue Doppler-imaging Eur. J. Cardiothorac. Surg., March 1, 2007; 31(3): 414 - 422. [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 |
