Optimal Ablation Settings Predicting Durable Scar Detected Using LGE-MRI after Modified Left Atrial Anterior Line Ablation

Abstract

(1) Background: The modified anterior line (MAL) has been described as an alternative to the mitral isthmus line. Despite better ablation results, achieving a bidirectional line block can be challenging. We aimed to investigate the ablation parameters that determine a persistent scar on late-gadolinium enhancement magnet resonance imaging (LGE-MRI) as a surrogate parameter for successful ablation 3 months after MAL ablation. (2) Methods: Twenty-four consecutive patients who underwent a MAL ablation have been included. The indication for MAL was perimitral flutter (n = 5) or substrate ablation in the diffuse anterior left atrial (LA) low-voltage area in persistent atrial fibrillation (AF) (n = 19). The MAL was divided into three segments: segment 1 (S1) from mitral annulus to height of lower region of left atrial appendage (LAA) antrum; segment 2 (S2) height of lower region of LAA antrum to end of upper LAA antrum; segment 3 (S3) from end of upper LAA antrum to left superior pulmonary vein. Ablation was performed using a contact force irrigated catheter with a power of 40 Watt and guided by automated lesion tagging and the Ablation Index (AI). The AI target was left to the operator's choice. An inter-lesion distance of ≤6 mm was recommended. The bidirectional block was systematically evaluated using stimulation maneuvers at the end of procedure. All patients underwent LGE-MRI imaging at 3 months, regardless of symptoms, to identify myocardial lesions (scars). (3) Results: Bidirectional MAL block was achieved in all patients. LGE-MRI imaging revealed scarring in 45 of 72 (63%) segments. In all three segments of MAL, ablation time and AI were significantly higher in scarred areas compared with non-scar areas. The mean AI value to detect a durable scar was 514.2 in S1, 486.7 in S2 and 485.9 in S3. The mean ablation time to detect a scar was 20.4 s in S1, 22.1 s in S2 and 20.2 s in S3. Mean contact force and impedance drop were not significantly different between scar and non-scar areas. (4) Conclusions: Targeting optimal AI values is crucial to determine persistent left atrial scars on an LGE-MRI scan 3 months after ablation. AI guided linear left atrial ablation seems to be effective in producing durable lesions.

Keywords: atrial fibrillation; cardiac arrhythmias; catheter ablation.

Figure 1

Three segments of a MAL. Voltage map of the left atrium. Violet areas reflect normal voltage (>0.5 mV), red areas a voltage <0.1 mV. VisiTags colour-coded by AI value. S1, segment 1; S2, segment 2; S3 segment 3.

Figure 2

Three-dimensional model of the left atrium. Blue reflects healthy LA tissue, whereas scar areas are shown in red (Merisight TM).

Figure 3

Median AI values for each of the 3 MAL segments for detecting a scar.

Figure 4

Comparison of AI values. In all segments of the MAL AI values were significantly higher when a scar was identified. (A) Total modified anterior line. (B) Segment 1 of modified anterior line. (C) Segment 2 of modified anterior line. (D) Segment 3 of modified anterior line.

Figure 5

Receiver-operating curves (ROC) for ablation index, ablation time and contact force. AI had the highest accuracy to predict scar (AUC 0.816; p < 0.001) with a cut-off value of 421 (sensitivity 0.87; specificity 0.63). Additionally, ablation time showed a high accuracy for predicting scar (AUC 0.784; p < 0.001). The point on the ROC curve associated with the greatest discriminatory potential was 14.7 s (sensitivity 0.867; specificity 0.44). CF did not predict scarring, with an AUC of 0.501.