Pulmonary vein isolation (PVI) is the mainstay of atrial fibrillation (AF) ablation. At present, it is unclear whether or not the ablation of additional substrates targeting electrical isolation in other atrial areas, such as the left atrium (LA) or left atrial posterior wall (LAPW), provides the same benefit, as no randomized clinical trials have shown improved benefits (1, 2). Macroreentrant atrial arrhythmias can occur late after mitral valve replacement due to the scarring associated with atriotomy. Extensive scarring may also occur in remote areas of the LA, particularly in patients with a history of rheumatic heart disease (3). The case report presented herein describes the case of a patient with a long history of AF, who underwent mitral and aortic valve replacement in conjunction with tricuspid ring annuloplasty for the treatment of rheumatic valvular heart disease. The patient underwent a successful isolation of a large low-voltage area of the LAPW through PVI and the creation of a linear lesion on the roof of the LA, which resulted in the restoration of sinus rhythm (SR) to the rest of the atria, although fibrillatory activity continued in the isolated area.

A 76-year-old male was referred to a tertiary arrhythmia center with palpitations and shortness of breath. The patient had a childhood history of numerous bouts of rheumatic fever, which resulted in severe mitral regurgitation and moderate stenosis, along with moderate aortic and severe tricuspid regurgitation. Subsequently, the patient underwent aortic and mitral valve replacement with tissue valves, as well as tricuspid ring annuloplasty, at 71 years of age. Over time, the patient developed AF, which had already been staged as permanent at the time he had the aforementioned surgery (Supplementary Table).

Five years after his valve replacement surgery, the patient presented to his general cardiologist with symptomatic AF (European Heart Rhythm Association class III), which required rate control optimization. Echocardiography demonstrated a severely dilated LA (59 mm diameter, LA volume index calculated at 52 ml/m²); although the left ventricular size and ejection fraction were normal, as were the tissue valves and the pulmonary artery systolic pressure, which was estimated to be 43 mmHg. Due to the difficulty in achieving rate control, the patient was fitted with an implanted single-chamber pacemaker for physiological pacing of the left bundle branch area in preparation for an atrioventricular node ablation, which was scheduled for 6 weeks later. Surprisingly, upon admission for the ablation, the patient presented with SR); however, despite the initiation of amiodarone therapy, the patient experienced a recurrence of atrial fibrillation shortly thereafter. A decision involving the patient and heart team was made to pursue rhythm control, and after providing written informed consent, the patient was scheduled for a complex catheter ablation.

After securing double transseptal access, a high-density electroanatomical map of the LA was created using a commercially available mapping system (CARTO3; Biosense Webster, Irvine, CA, USA) and a multipolar mapping catheter (PentaRay; Biosense Webster, Irvine, CA, USA). Created while the patient was in AF, the voltage map demonstrated large areas of low voltage, primarily along the posterior wall of the LA, sparing the vestibular portion and the left atrial appendage (Figure 1). Voltage cut-off was 0.24 mV, as suggested by Rodríguez-Mañero et al. (4). PVI was performed with an irrigated-tip catheter (Thermocool Smarttouch SF; Biosense Webster, Irvine, CA, USA) following the CLOSE protocol (5).

During delivery of energy to the LA ridge, the AF was converted into atrial flutter with a cycle length of 240 ms, demonstrating proximal-to-distal coronary sinus activation. Biatrial mapping revealed passive left atrial activation and clockwise peritricuspid atrial flutter circuit (Figure 2A). During ablation at the cavotricuspid isthmus, there was a sudden change in the coronary sinus activation pattern, which reversed to a distal-to-proximal direction, suggesting a left atrial origin (Figure 2B). At the same time, the LAPW was activated with a shorter cycle length, which varied between 200 and 230 ms. Posterior wall isolation was then performed, starting with the roof line. A few lesions on the roof of the LA in the zone of fragmented signals resulted in a progressive delay of conduction from the LAPW to the remaining atria, while SR was eventually restored and the atrial arrhythmia was fully contained in the LAPW (Figure 3). Isolation of the LAPW with persistent atrial arrhythmia and an exit block to the rest of the LA was maintained through the end of the procedure. No attempt at restoration of normal rhythm in the isolated area was made. The isolated area was measured offline, and found to be 72.1 cm², accounting for 28.2% of the surface area of the LA.

The day after the procedure, when he was discharged, the patient was confirmed to be in SR, and no periprocedural complications were observed. Over the course of a 12-month follow-up, the patient remained in SR, and had a significantly improved quality of life. Life-long oral anticoagulation therapy was recommended.

The LAPW stems from the same embryological origin as the pulmonary veins, and the heterogeneous arrangement of its myocardial fibers may result in non-uniform anisotropy with delayed conduction, a unidirectional block, and localized reentry. Compared to other areas of the LA, the cells in the LAPW have larger late sodium currents, more calcium content, and smaller potassium currents, resulting in a low resting membrane potential and short action potential duration (6, 7). In patients who experience rheumatic heart disease following mitral valve surgery, significant scarring at the LAPW, independent from the atriotomy lines, has been described as a marker of atrial cardiomyopathy (8).

In the case presented herein, LAPW isolation was achieved via PVI and a roof line only, without creating an inferior line, a finding which can be explained by several mechanisms. Using a right lateral approach to the left atrium, atriotomy for mitral valve replacement can be extended to run parallel to the mitral annulus. This, in turn, may explain how point ablation could lead to a delay in conduction from the LAPW across the roof, and why LAPW isolation could be accomplished via a roof line only (Figure 4). An alternative explanation for this would be the presence of a pre-existing functional block at the inferior periphery of the scar, which is primarily due to abrupt changes in myocardial strand orientation (9), scarring by the rheumatic process itself (3), or both. The large surface area of the box created during the ablation, which extended to the anterior wall and atrial septum, could have also contributed to the success of the procedure, due to LA compartmentalization (10).

Similar case reports are scarce, as we found only one relevant case report, describing the final isolation of the posterior LA and restoration of SR with single-point ablation to the roofline gap after an unsuccessful thoracoscopic ablation in a patient with longstanding persistent AF (11).

In conclusion, the case presented herein supports the concept that a large and completely isolated box lesion is the key to a successful ablation in patients with advanced atrial cardiomyopathy due to a history of rheumatic disease. Additionally, the present case report indicated that scarring due to a previous atriotomy affects the ablation strategy.