Author + information
- Received June 4, 2009
- Revision received August 19, 2009
- Accepted August 24, 2009
- Published online November 24, 2009.
- José Dizon, MD⁎,⁎ (, )
- Kimberly Chen⁎,
- Matthew Bacchetta, MD†,
- Michael Argenziano, MD†,
- Donna Mancini, MD⁎,
- Angelo Biviano, MD⁎,
- Joshua Sonett, MD† and
- Hasan Garan, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. José M. Dizon, Division of Cardiology, Department of Medicine, Columbia University, 222 Westchester Avenue, White Plains, New York 10604
Objectives We compared the incidence of atrial arrhythmias in double-lung transplant patients versus heart transplant patients to gain insight into factors that contribute to post-operative atrial fibrillation (AF).
Background Atrial fibrillation is a common complication after thoracic surgery. Pulmonary vein isolation is an effective treatment for AF. Heart or double-lung transplantation surgery both involve pulmonary vein isolation because of suture lines.
Methods We reviewed the records of 174 consecutive heart transplant patients and 122 double-lung transplant patients at the Columbia Presbyterian Medical Center between January 2005 and June 2008. Electrocardiograms during atrial arrhythmia episodes were reviewed by an electrophysiologist. Clinical variables, biopsy results, immunosuppressive regimens, and echocardiographic measurements were collected from the perioperative time period and at the time of arrhythmia occurrence.
Results In the heart transplant group, 8 (4.6%) patients had AF (group A). In the lung transplant group, 23 (18.9%) patients had AF (group B; p < 0.001). The incidence of AF in a comparison group of 131 patients with normal left ventricular function who underwent coronary artery bypass graft surgery was 19.8%. Immunosuppressive regimens and clinical variables were similar for both groups. Echocardiographic data revealed no significant cardiac abnormalities in 74% of group B compared with 25% of group A (p < 0.05), and 78% of biopsy results in group B were normal, whereas only 25% of group A results were normal (p < 0.05).
Conclusions In heart transplant recipients, AF is uncommon and occurs in the setting of myocardial dysfunction and graft rejection. In contrast, AF is more common after lung transplantation despite the absence of graft rejection and cardiac dysfunction. Pulmonary vein isolation alone cannot explain the discrepancy in AF incidence between heart transplant recipients and double-lung transplant recipients. Cardiac autonomic denervation may have a protective effect for heart transplant patients in the post-operative setting.
- post-operative atrial fibrillation
- pulmonary vein isolation
- cardiac transplantation
- pulmonary transplantation
Atrial fibrillation (AF) is a common complication after thoracic surgery, with an incidence approaching 50% in some series (1,2). Multiple mechanisms may underlie post-operative AF, and various studies have aimed at predicting its occurrence (3–5). Recently, pulmonary vein isolation (PVI) using catheter and surgical approaches has been shown to be effective in treating paroxysmal or persistent AF (6–8). The efficacy of these techniques approaches 90% in some series (9–12).
The most common techniques in heart or double-lung transplantation effectively cause PVI due to suture lines. The incidence of post-operative AF in heart transplant series is known to be low (13–16), whereas the same incidence in post–lung transplant patients appears to be higher (17,18). It is not known whether differences in pre- or post-transplant clinical variables, cardiac function, surgical technique, immunosuppressive regimens, or allograft rejection explain the difference in AF rates after heart transplantation compared to lung transplantation.
The purpose of this study is to analyze the incidence of AF after heart or double-lung transplantation in a single institution with a large volume of both procedures. Clinical variables, echocardiographic measurements, immunosuppressive regimens, and biopsy results of patients with AF were compared to gain insight into the mechanisms of post-operative AF. Because atrial flutter (AFL) is often associated with AF and also occurs commonly after thoracic surgery, we recorded episodes of AFL as well.
The records of 174 consecutive adult heart transplant and 122 double-lung transplant patients at the Columbia Presbyterian Medical Center from January 2005 to June 2008 were reviewed. The review included inpatient and outpatient charts, discharge summaries, operative reports, and all other available records in paper or electronic form. Any notation of an atrial arrhythmia such as AF or AFL was investigated, and the corresponding electrocardiogram (ECG) was located and copied. If the ECG corresponding to a noted arrhythmia episode could not be located, that episode was not included in the analysis. An arrhythmia episode is defined as the occurrence of AF or AFL lasting >30 min. If episodes were separated by >24 h, they were counted as individual episodes.
The ECGs were reviewed in a blinded fashion by a board-certified electrophysiologist and classified as AFL, AF, or neither. When the ECG reading was uncertain, a second electrophysiologist was asked to read the ECG and a consensus reading was determined. Perioperative and post-operative data from the group of transplant patients with confirmed AF or AFL were then collected. These included clinical variables, echocardiographic data, immunosuppressive regimens, and biopsy results. Clinical data at the time of arrhythmia occurrence were particularly scrutinized.
As a control, data from a contemporary group of 131 patients with preserved left ventricular function who had undergone coronary artery bypass graft surgery (CABG), and who also did not have a pre-operative history of AF or AFL, were analyzed. The number of post-operative atrial arrhythmias in this group would represent the approximate incidence expected to occur in a group of patients with preserved cardiac function undergoing thoracotomy and pericardiectomy.
The double-lung transplant technique for the patient population was a sequential single-lung implant procedure using a bilateral anterolateral fourth interspace thoracotomy, or a clamshell incision in which the sternum was divided transversely. After the bronchial anastomoses, each donor lung's pulmonary veins with atrial cuff were sutured to an atriotomy in the recipient left atrium. Therefore, 2 separate atriotomies were created (Fig. 1).The pulmonary artery anastomoses were completed after the bronchial and pulmonary vein anastomoses. The orthotopic heart transplant technique for this population consistently involved a bicaval right atrial anastomoses. The recipient pulmonary veins with atrial cuff were sutured to a single atriotomy in the donor left atrium (Fig. 2).
Data are presented as absolute values, percentages, or mean ± SD. Continuous variables were compared using a 2-tailed ttest. Nominal variables were compared using the Fisher exact test (2-tailed). The Bonferroni correction was applied when comparing multiple means. A value of p < 0.05 was considered significant.
The demographics of the groups, along with pre-operative diagnoses, are listed in Tables 1 and 2.⇓⇓Our major findings are displayed in Table 3,which lists the incidence and timing of AF for the transplant and CABG groups. There was a more than 4-fold higher incidence of post-operative AF in the lung transplant group and the control group of CABG patients compared with the heart transplant group (p < 0.001). The incidence of post-operative AF in the CABG group was similar to that of the lung transplant group. Of the lung transplant patients with AF, only 2 were known to have AF before transplant, and 78% had AF within 1 week after transplant. In contrast, 50% of heart transplant patients with AF had AF in the first week after transplant. Three arrhythmia episodes occurred later than 1 month post-operatively in the heart transplant group, the latest occurring 9 months after transplant. Nine atrial arrhythmia episodes occurred after 1 month in the lung transplant group; the latest was 10 months after transplant. The average duration of follow-up was 490 ± 316 days (range 9 to 1,170 days) for the lung transplant group and 464 ± 240 days (range 3 to 896 days) for the heart transplant group. Five patients in the heart transplant group (2.9%) had AFL, as compared with 9 (7.4%) patients in the lung transplant group and 5 (3.8%) patients in the CABG group. Nine episodes of an atrial arrhythmia could not be corroborated with an ECG in the lung transplant group, as compared with 2 in the heart transplant group.
Table 4lists selected clinical variables for the atrial arrhythmia patients, chosen for potential impact on the incidence of post-operative AF and as general measures of severity of clinical course. There was a 4-fold higher rate of the use of pre-operative beta-blockers in the heart transplant patients, but not the use of post-operative beta-blockers before the onset of an atrial arrhythmia. Three patients in the lung transplant atrial arrhythmia group were taking antiarrhythmic agents before transplant (2 amiodarone, 1 flecainide), and all 3 continued these medications after transplant. Five heart transplant atrial arrhythmia patients were taking amiodarone before transplant, but none continued antiarrhythmic agents after transplant. Both groups had similar lengths of admission, prolonged use of inotropic/pressor agents, complication rates, comorbidities, and deaths.
Table 5lists the immunosuppressive regimens for the transplant patients, which were similar in terms of dose and type of regimen. Table 6lists the echocardiographic data for the transplant patients with AF. These data were from post-operative echocardiograms selected within days of the occurrence of the atrial arrhythmias. Where quantitative data were not available (for right ventricle function and right atrium size), the percent of studies listed as “normal” is denoted. The lung transplant arrhythmia patients, despite their higher incidence of AF, had a more than 3-fold higher percentage of normal echocardiographic indexes.
Finally, Table 7lists the biopsy results of the patients with atrial arrhythmias, stratified by International Society for Heart and Lung Transplantation grade. Overall, only 18% of the lung transplant patients who developed an atrial arrhythmia had rejection as defined by International Society for Heart and Lung Transplantation grade >0 or A0, compared with 63% of the heart transplant arrhythmia patients.
The incidence of AF is high after thoracic surgery, reaching >50% for procedures involving valve replacement (1,2). However, the incidence of AF is well known to be low after cardiac transplantation (13–16). Given the recent emphasis on PVI as therapy for AF, one may postulate PVI as being protective against AF in the heart transplant recipient. Because double-lung transplant procedures also involve effective isolation of the pulmonary veins, we sought to contrast the incidence of post-operative AF in this population against the incidence after heart transplant. The differences in the incidence of AF between these transplant groups might point toward alternative explanations for post-operative AF other than pulmonary vein sources, since PVI was a common component of both procedures. The remaining differences between the groups, such as autonomic denervation in the transplanted heart, could highlight other potential mechanisms for post-operative AF such as autonomic modulation.
Our results, which are similar to those reported in previous lung and heart transplant series (13,14,16–18), reveal a much higher incidence of AF in the post–lung transplant group, similar in fact to the incidence in our post-CABG patients with preserved left ventricular ejection fraction. There was a preponderance of males in the heart transplant group, reflecting the general higher incidence of heart disease among males, but there is no obvious reason why this could explain the difference in AF incidence. Neither does the higher incidence in the lung transplant group appear to be explained by a higher incidence of pre-operative AF, which was very low.
We also recorded the incidence of AFL, given its association with AF and known occurrence after thoracic surgery. We could not confirm with certainty that the AFL episodes were not left atrial in origin, and hence potentially impacted by PVI. Nevertheless, including the AFL episodes in our analysis does not alter our main results.
Both groups had similar average post-transplant lengths of stay, complication rates, prolonged use of inotropic or pressor agents, and comorbidities. The heart transplant group had a higher use of pre-operative beta-blockers and amiodarone, due probably to the frequent use of these agents for end-stage heart failure, and the reluctance to use beta-blockers for advanced pulmonary disease. However, the use of pre-operative beta-blockers and antiarrhythmic agents is unlikely to impact arrhythmias in a newly transplanted heart not previously exposed to such agents. The use of these agents in the post-operative period before the onset of any atrial arrhythmias was not significantly different between the 2 transplant groups.
The higher incidence of AF in the lung transplant group occurred despite our observation that the lung transplant arrhythmia patients as a whole did not have larger post-operative right atrium size. The normalized echocardiographic indexes after lung transplant may reflect the selection of patients without cor pulmonale and hence reversible cardiac function and chamber enlargement. However, even with normalization of right atrial size, subclinical fibrosis and hence altered electrophysiologic properties may persist in the atria of the lung transplant patients. The lung transplant group with atrial arrhythmias also did not tend to have pulmonary graft rejection.
Heart transplant patients with atrial arrhythmias in our series more commonly had evidence of atrial abnormalities, right ventricle dysfunction, or abnormal myocardial biopsy results. These results are consistent with prior studies reporting atrial tachyarrhythmias in heart transplant recipients (13–15,19). Graft rejection resulting in fibrosis or edema may set up heterogeneities in conduction or repolarization within the atria, leading to AF or AFL. The cumulative data suggest that atrial arrhythmias in post–heart transplant patients are more often a sequelae of underlying myocardial dysfunction, whereas atrial arrhythmias after lung transplant are less related to intrinsic cardiac causes, similar to patients after CABG with normal left ventricle function. Since, in our study, PVI occurred in both transplant groups, pulmonary vein sources for AF do not appear to be the primary cause for post-operative AF, nor does PVI appear to be the primary protective mechanism against AF for the heart transplant recipients.
Mechanisms proposed for post-operative AF include inflammation, and various studies have pointed to a beneficial effect of anti-inflammatory agents such as steroids (20–22). It is possible that the low incidence of AF after heart transplant may be related to the anti-inflammatory effect of immunosuppressive agents. However, as noted in Table 5, the immunosuppressive regimen for lung transplant recipients is nearly identical to that for heart transplant recipients, arguing against an anti-inflammatory explanation for the difference in AF incidence.
Two possible explanations may be proposed for the difference in AF incidence between the 2 groups. One explanation is the slightly different surgical technique used between the 2 groups. In both the double-lung and heart transplant cases, the right atrium is left intact because the bicaval heart transplant technique does not involve a right atriotomy (Fig. 2). However, because the pulmonary veins are often attached to the extreme lateral ends of the posterior left atrium, the mass of left atrium excluded by the heart transplant procedure is larger and includes a larger section of the posterior wall than that included during the lung transplant procedure (Figs. 1 and 2). That might result in more potentially arrhythmogenic tissue being excluded by the heart transplant procedure, especially in view of the recent focus on the posterior wall of the left atrium as a site for initiators and drivers of AF (23). Post-lung transplant patients may also have sources of AF that are distant from the pulmonary veins, and hence are unaffected by PVI.
Another explanation is cardiac autonomic denervation in the heart transplant patients. Autonomic factors are known to be important in arrhythmogenesis, particularly the role of ganglionic plexi in the pathogenesis of AF (24,25). The orthotopic heart transplant procedure results in complete denervation of the implanted heart, which has been well documented (14,26–28). Late cardiac autonomic reinervation may occur, but is likely a slow, incomplete process (28,29). In contrast, previous studies using exercise testing or measures of autonomic nervous system activity support the preservation of autonomic cardiac innervation in single- and double-lung transplant recipients (30–34). It is possible that cardiac autonomic denervation may be another important mechanism that accounts for the difference in AF incidence between heart and lung transplant recipients in the post-operative setting.
This study is a retrospective review of clinical and laboratory data, including patients with complex and prolonged hospital courses. It is therefore possible that the observed differences between the transplant groups are a chance occurrence. Another limitation is the possible omission of relevant clinical data. However, the electronic medical records in our hospital system make relatively complete access and review of records feasible. Although we were unable to locate some ECGs to confirm AF or AFL in a small group of patients, there were more nonlocalized ECGs in the lung transplant group, which if confirmed, would sway the results even more toward a higher incidence of atrial arrhythmias in the lung transplant patients. Another limitation is our assumption that the pulmonary veins are electrically isolated in the transplant patients, which was not confirmed with electrode catheters.
In heart transplant recipients, AF is uncommon and occurs in the setting of myocardial dysfunction and graft rejection. In contrast, the incidence of AF in double-lung transplant recipients is similar to that in post-CABG patients, despite the absence of graft rejection and cardiac dysfunction. Moreover, PVI alone cannot explain the differences in AF incidence between these groups, although the heart transplant procedure may be more antiarrhythmic because of the larger mass of atria isolated. Cardiac autonomic denervation may have a protective effect for heart transplant patients in the post-operative setting.
Dr. Argenziano serves as a consultant for ESTECH, Inc.
- Abbreviations and Acronyms
- atrial fibrillation
- atrial flutter
- coronary artery bypass graft
- pulmonary vein isolation
- Received June 4, 2009.
- Revision received August 19, 2009.
- Accepted August 24, 2009.
- American College of Cardiology Foundation
- Kaireviciute D.,
- Aidietis A.,
- Lip G.Y.H.
- Echahidi N.,
- Pibarot P.,
- O'Hara G.,
- Mathiue P.
- Haissaguerre M.,
- Jais P.,
- Shah D.C.,
- et al.
- Pappone C.,
- Rosanio S.,
- Oreto G.,
- et al.
- Pappone C.,
- Oreto G.,
- Rosanio S.,
- et al.
- O'Neill M.D.,
- Jais P.,
- Hocini M.,
- et al.
- Vaseghi M.,
- Boyle N.G.,
- Kedia R.,
- et al.
- Ruffin R.T.,
- Kluger J.,
- Baker W.L.,
- Willis S.M.,
- White C.M.,
- Coleman C.I.
- Baker W.L.,
- White C.M.
- Sanders P.,
- Hocini M.,
- Jais P.,
- et al.
- Morgan-Hughes N.J.,
- Corris P.A.,
- Healey M.D.,
- Dark J.H.,
- McComb J.M.
- Hasan A.,
- Corris P.A.,
- Healy M.,
- et al.