Authors Hongwu Chen and Bing Yang contributed equally to this work.
This work was supported by the Program of Innovative Research Team in the First Affiliated Hospital of Nanjing Medical University (IRT-004).
Conflict of Interest: None.
Address for reprints: Minglong Chen, M.D., Section of Electrophysiology, Division of Cardiology, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, China. Fax: 0086-25-8371-7168; e-mail: firstname.lastname@example.org
Atrial tachycardia (AT) is a frequent late sequel of surgical valve replacement procedures in patients with rheumatic heart disease (RHD). The aim of this study was to evaluate the acute and long-term outcome of catheter ablation in such patients.
Methods and Results
A total of 21 consecutive RHD patients with AT after a valve replacement were enrolled in this study. The mean interval between the occurrence of symptomatic AT and the surgical intervention was 38.2 ± 48.7 months. The initial procedure was performed 8.4 ± 8.9 months after first onset of AT. During the first procedure, an electroanatomic mapping was completed for 25 ATs, 18 of which were cavotricuspid isthmus-dependent atrial flutter, five that were right atrial free wall AT, and two that were left AT. Acute success was obtained in 95% (20/21) patients. Nine patients with recurrent AT had repeat radiofrequency catheter ablation, and newly developed left AT was identified in five patients after the first right AT ablation. After a mean follow-up of 42.7 ± 17.3 months, only 33% of the patients remained free of ATs, while 14% and 53% of the patients had AT recurrence and the development of atrial fibrillation (AF), respectively.
Right but not left macroreentry is the most common AT postmitral valve replacement in patients with RHD. The incidence of AF is very high after AT ablation in such patients during the long-term follow-up.
The incision on the right atrial free wall (RAFW) is a routine access site to the left atrium for the surgical replacement of the mitral valve in patients with rheumatic heart disease (RHD). Atrial tachycardia (AT) is an important late complication after this procedure[1-4] due to both the surgical substrate and the diseased atria as well. In addition, a heightened inducibility of atrial fibrillation (AF) is also observed in patients with rheumatic mitral stenosis (MS) because of biatrial electrical and structural remodeling. The management of postoperative AT with pharmacological treatment frequently fails to control this tachyarrhythmia and is accompanied by side effects.[4, 6-8] Radiofrequency catheter ablation (RFCA) has evolved as a feasible curative treatment for this AT[4, 9, 10] and is associated with a high-acute success rate.[11, 12] However, to date, there is a paucity of long-term follow-up data in mitral valve replacement patients with RHD after the ablation of AT. Our hypothesis is that the substrate of the RHD atria is complex and nonstatic, and the long-term outcome following the ablation of AT in this population needs to be known.
From March 2006 to September 2011, 60 consecutive RHD patients (37 men; mean age 54.8 ± 11.1 years, range 38–73 years) with supraventricular tachyarrhythmias and a history of a mitral valve replacement were referred to our center, 32 of whom had AF (10 with paroxysmal AF) and were treated for rhythm or rate control, two with paroxysmal supraventricular tachycardia and 26 with AT. There was no significant difference in the left ventricular ejection fraction (LVEF) between the patients with AF and patients with AT (63.3 ± 3.2% vs 63.5 ± 4.8%, P = 0.82). Furthermore, the significant difference in the left atrium diameter (LAD) between these two groups was not observed (46.0 ± 4.2 mm vs 45.9 ± 5.2 mm, P = 0.95). Among all the AT patients, five refused to accept an ablation procedure. In the remaining 21 patients (17 men, mean age 50.8 ± 8.2 years, range 38–65 years), 16 had persistent AT and four had documented AF in addition to AT prior to the procedure. Five of the study population patients had a concomitant aortic valve replacement. The study protocol was approved by the local Institutional Review Board, and all patients provided written informed consent.
All the antiarrhythmic drugs were discontinued for at least five half-lives before the procedure with the exception of amiodarone. Transesophageal echocardiography was performed in all patients to exclude the presence of intracardiac thrombi. An electrophysiological study and catheter ablation was performed under conscious sedation and local anesthesia. A 6F decapolar catheter (St. Jude Medical, St. Paul, MN, USA) was positioned in the coronary sinus via the left subclavian vein, and a 6F quadripolar catheter (St. Jude Medical) was positioned in the right ventricle via a femoral vein. One 8F 65-cm-long sheath (SL1, St. Jude Medical, Inc.) was advanced into the right atrium. If AT was identified from the left atrium by entrainment, then the long sheath was advanced to the left atrium through a standard transseptal puncture. After the transseptal puncture, intravenous heparin was administered to maintain an activated clotting time of 250–300 seconds. A deflectable quadripolar open irrigated catheter (Navi-star, Biosense Webster, Diamond Bar, CA, USA, or Cool Path, IBI, St. Jude Medical, Irvine, CA, USA) was inserted into the right and/or left atrium for mapping and ablation. The intracardiac electrograms were recorded using a digital electrophysiological recording system (Prucka CardioLab, General Electric Health Care System Inc., Milwaukee, WI, USA) and were filtered from 30–300 Hz.
Three-Dimensional (3D) Electroanatomical Mapping and Catheter Ablation
Mapping was performed during sustained tachycardia. Patients in sinus rhythm at baseline underwent programmed stimulation with or without an isoproterenol infusion to induce the tachycardia. Before the ablation procedure, detailed electroanatomic mapping of the right atrium (RA) and/or left atrium (LA) was performed by obtaining contact bipolar electrograms (100–150 points) during the AT guided by 3D electroanatomical mapping using a CARTO system (Biosense Webster Inc.) or EnSite-NavX (St. Jude Medical) system. Voltage and activation mapping were then used to localize the surgical or spontaneous scar, fractionated signals, and double potentials, and to further analyze the direction of the wavefront propagation and the location of the critical isthmus. Radiofrequency energy was delivered to the key isthmus using an irrigated-tip catheter (Navi-star, Biosense Webster Inc., or IBI, St. Jude Medical, Inc.) with a maximal temperature of 45°C at a maximal power of 35 W and infusion rate of 17 mL/min. The ablation endpoint was the termination of the tachycardia during RFCA with noninducibility thereafter. If linear lesions were applied to block the isthmus conduction, then bidirectional conduction block[13-15] of the line needed to be confirmed.
Antiarrhythmic agents were discontinued following a successful procedure and anticoagulation was intensified for 3 days with low molecular weight heparin. Oral anticoagulation with Warfarin was continued for life if a mechanical valve was placed.
A regular outpatient visit was performed in all patients. A surface electrocardiogram (ECG) and 24-hour Holter recording were performed after the procedure at 1 month and then every 6 months thereafter. Furthermore, an ECG was also performed anytime if the patients experienced palpitations. Patients who had a documented recurrence of AT after the procedure underwent a repeat ablation procedure.
The statistical analyses were performed using SPSS software version 15.0 (IBM Corp., Armonk, NY, USA). Continuous variables were expressed as the mean ± standard deviation. A univariate analysis for the predictors of late AT and/or AF was performed using the Cox proportional hazards model. A P < 0.05 was considered statistically significant.
Twenty-one RHD patients with AT after a mitral valve replacement were included in the study, 16 of whom had persistent AT. Four patients had documented AF in addition to AT prior to the ablation procedure. In all patients, a routine surgical access to the left atrium was via an atriotomy incision in the free wall of the right atrium. The valve replacement was performed at a mean of 48.8 ± 10.0 years of age. The patients had an AT occurrence at a mean of 38.2 ± 48.7 months postsurgical intervention. The initial ablation procedure was performed 8.4 ± 8.9 months after the first onset of AT. Among the patients, four (19%) had hypertension and three (14%) had coronary artery disease. Their mean LVEF and LAD were 61.9 ± 8.5% and 46.7 ± 5.4 mm, respectively. The patients could not be controlled by a mean of 1.2 ± 0.8 antiarrhythmic drugs before the ablation procedure, including with the use of verapamil, propofenone, metoprolol, digoxin, and amiodarone.
Outcome of Ablation Procedure
In the entire study population, 26 spontaneous and induced ATs were identified during the first ablation procedure with a mean tachycardia cycle length of 250 ± 37 ms. Complete maps of 25 ATs were obtained by electrophysiologic and electroanatomic mapping. Eighteen patients had cavotricuspid isthmus (CTI)-dependent atrial flutter (AFL), and four of those patients had an induced, coexisting, or automatically transferring AT looping around the RA scar after the CTI ablation. For ablation of a scar-reentrant free wall AT, two patients underwent an ablation line from the inferior atriotomy to the inferior vena cava, and two others from the atriotomy to the tricuspid annulus. Two patients had a LA AT, one of which was going around the mitral annulus (220 ms) with an unsuccessful ablation and the other patient had a focal LA appendage AT (240 ms) with a successful ablation. The remaining one patient only had a RAFW AT going around the scar, which was successfully abolished by placing the lesions in the corridor between the scar and inferior vena cava. An acute success in the initial procedure was obtained in 95% (20/21) of the patients.
Follow-Up Results and the Findings of Redo Procedures
During the follow-up, 15 right AT patients had recurring atrial tachyarrhythmias (six with AF, nine with AT). Eight of them with recurrent AT accepted a repeat ablation procedure and newly developed left AT was identified in five patients after the first right AT ablation, and right AT recurrence was observed in three patients (one with CTI-dependent AFL, and two with scar-reentrant ATs). In the two left AT patients, the one with an unsuccessful ablation degenerated to AF and the other one developed a new left AT which was finally eliminated at the left posterior wall beyond the scar.
Final Follow-Up Results of the Study Cohort
The ablation outcomes and follow-up results are summarized in Figure 1. After a mean follow-up of 42.7 ± 17.3 months since the last procedure, only 33% of the patients (7/21) remained in stable sinus rhythm. However, 67% of the patients (14/21) had recurrent atrial tachyarrhythmias, 14% (3/21) of which were recurrent AT and 53% (11/21) documented AF (Fig. 2). AF was paroxysmal in two patients and permanent in nine. All the recurrence occurred within the first year after the last procedure (Fig. 3).
A univariate analysis of the age, gender, age at the time of the RHD surgery, LVEF, cycle length of the AT, onset time of the AT after the surgery, and time of the first ablation after the AT onset failed to demonstrate any statistically significant predictors of atrial tachyarrhythmias. However, the LA dimension was the predictor of a late atrial tachyarrhythmia occurrence (Table I).
Table I. Clinical Characteristics of Patients with and Without Recurrence
Recurrence (n = 7)
(n = 14)
AT = atrial tachycardia.
52.6 ± 10.4
51.6 ± 7.3
Male, n (%)
Age of the surgical procedure (years)
50.1 ± 11.5
48.2 ± 9.6
Left atrial dimension (mm)
43.8 ± 3.1
48.1 ± 5.8
Left ventricular ejection fraction (%)
64.8 ± 5.1
60.5 ± 9.8
Cycle length of the AT (ms)
250 ± 23.8
254 ± 48.1
Interval between the onset of AT and the surgery (months)
32.8 ± 51.8
41.0 ± 47.5
Time of the first ablation after the onset of AT (months)
9.8 ± 11.2
7.8 ± 8.0
No procedure-related complications were observed in the study population.
This study presents new information on the characteristics of recurrent AT/AF after the ablation of postoperative AT during the long-term follow-up in RHD patients' postmitral valve replacement. The main findings of this study were as follows: (1) In total, most of the postoperative ATs (76%) in such patients were found to be from the right atrium, while only 24% originated from the LA; and (2) the late occurrence of AF was very high (53%) post-AT ablation.
From our study, most of the postoperative ATs (76%) were found to be from the right atrium and 73% of those ATs were CTI-dependent AFL. The mechanisms of these tachycardias might be associated with the incisional scar, surgical suture lines in the right atrium that formed a possible posterior line of blockade, and free wall slow conduction, which created an isthmus-dependent AFL circuit according to our prior and other studies.[12, 16] Such obstacles may prevent a short circuit of the activation wavefront and thereby facilitate the induction and perpetuation of AFL. In this study, we also found that some ATs were located in the LA (24%) in patients with and without a CTI ablation, which may be correlated with both atrial electrical and structural remodeling and fibrosis due to rheumatic inflammation in RHD. To the best of our knowledge, the incidence of left AT in RHD patients with a mitral valve replacement has so far not been reported.
Several studies have demonstrated the occurrence of ultrastructural changes resulting in fibrosis in the atria of patients with RHD.[17, 18] A pathognomonic feature of rheumatic carditis is the Aschoff bodies. Such inflammatory tissue can arise during an acute episode of carditis and remain immunologically active for years. They have been associated with the local release of a variety of cytokines, including many that are implicated in fibrogenesis. Several other immunoactive cytokines have also been observed to be increased in the plasma in patients with RHD. In addition, another study suggested the remodeling of the matrix metalloproteinases as a result of mitral valve disease. Recently, John et al. elegantly described the concept of atrial electrical and structural remodeling in RHD patients using 3D electroanatomic mapping. Their study demonstrated that atrial remodeling in RHD with MS was characterized by LA enlargement, loss of myocardium, and scarring associated with widespread and site-specific conduction abnormalities. These abnormalities were associated with a heightened inducibility of AF. Clinical studies also demonstrated that chronic RHD associated with AF was very common, and almost one-quarter of those individuals had AF, and over 40% of afflicted patients developed AF. There might be reason to consider that the substrate primarily involving the left and right atria could predispose the atria to the occurrence of AF. Our study also confirmed that this population did have a substrate for AF as evidenced by its high occurrence despite successful ablation of right and/or left AT. The other important factor contributing to the high recurrence of atrial tachyarrhythmias was closely related to the progressive electropathological and histological alteration of the atrial tissue in patients with RHD because the rheumatic disease itself was progressive.
In our study, the incidence of atrial tachyarrhythmias after the last ablation procedure during the long-term follow-up was much higher than that of de Groot et al. and Teh et al. This discrepancy indicates that the substrate promoting the atrial tachyarrhythmias in RHD patients is more extensive and progressive than that in congenital heart disease patients in addition to the surgical substrate. Newly developed AT which was not surgical incisional scar-related and a high incidence of postprocedural AF supported this
hypothesis. The other important finding which was unable to be explained was that the incidence of an AF occurrence postmitral valve replacement and AT ablation was much higher than that in the general RHD population.[23, 24] The only explanation is that the patients in our study were highly selective. Patients who developed AT postmitral valve replacement might have had more electrophysiological and histological abnormalities than those without. This speculation is consistent with the univariate analysis results in our study, which revealed that the LA dimension was the single predictor of a late atrial tachyarrhythmia occurrence. The larger the LA dimension, the higher the recurrence rate of atrial tachyarrhythmias.
Recently, clinical study has shown that the circumferential pulmonary vein antrum isolation (PVAI) in combination with substrate modification could achieve 66.7% success rate in patients with persistent AF and RHD, and another study also demonstrates that circumferential pulmonary vein isolation combined with substrate modification 6 months after valvular operation, the cumulative rates of sinus rhythm were 71% after last procedure. The results from these studies have the implications for potential intervention of AF following mitral valve replacement patients with RHD, and the subsequent catheter ablation is safe and effective strategy. Because of the high-recurrence rate of AF after initial AT ablation in our study, it might be necessary to combine with prophylactic PVAI in this population to prevent AF, but this needs further studies in the future.
This study provided information that most of the ATs postmitral valve replacement were surgical incisional scar-related, and were either a scar-mediated AT or scar-promoted AFL. According to our prior study, joining the surgical incision anteriorly to the tricuspid annulus and inferiorly down to the inferior vein cava is necessary to prevent postsurgical ATs during open heart surgery. Because of a higher incidence of AF and left AT postmitral valve replacements, prophylactic maneuvers are strongly recommended in RHD patients with an AF history before surgery. The Maze procedure for persistent AF patients and surgical PVAI might be undertaken when performing a valve replacement in a patient with or without paroxysmal AF.
First, this was a retrospective study and involved only a small number of patients. We need to do a comparative analysis of the difference in the true AF occurrence between RHD patients postmitral valve replacement with AT and those without in a larger population. However, the long-term outcome in this special population might provide useful information to guide clinical practice. Second, the clinical outcome was evaluated by the patients’ symptoms, ECGs at the outpatient clinic, and 24-hour ambulatory monitoring, and therefore, the true atrial tachyarrhythmia recurrence rate post-AT ablation might be underestimated.
Right but not left macroreentry is the most common AT postmitral valve replacement in patients with RHD. The incidence of AF is very high after the AT ablation in such patients during the long-term follow-up.