**4.1 Ablation index-guided PVI with HPSD applications**

According to previous studies, ablation index (AI) is a valuable marker of lesion creation during PVI procedure, minimizing the risk of AF recurrence after ablation [14, 19]. AI is calculated by a weighted formula including CF, RF time, and power so that a higher-power application can reach target AI with a shorter duration and even with high energy, RF applications' lesion creation can be monitored in real time properly. Of note, the AI is validated up to 45 W; thus, lesion creation can be reliably monitored with AI at a maximal power of 45 W. The decrease of RF time per application theoretically makes the maintenance of a stable catheter position at the given site easier, resulting in a better lesion quality. It is well known that HPSD ablation results in a different lesion geometry (e.g., larger diameters but smaller depth) compared to conventional lower-power ablation. Still, the depth of the HP applications is sufficient to reach transmural lesions in the atria. The power used for ablation can be varied based on the operating physician's decision. As we will show in this subchapter, 45–50 W was used in the majority of trials.

The use of high-power (HP) RF applications can reduce procedural time and seemed to be non-inferior to low-power (LP) ablation in a multicenter study [20]. Vassallo et al. investigated patients who underwent AF ablation with HP (50 W on the anterior wall and 45 W elsewhere in the left atrium) or LP (30 W) RF ablation power settings. HPSD was safe and efficient compared with LP ablation and was associated with a reduced procedural time and total RF time. They also concluded that HPSD might reduce the chance of esophageal injury and it may also reduce the recurrence of atrial tachyarrythmias [21]. The PVI procedure time was also decreased significantly with HP (50 W) ablation compared to conventional LP (30 W) ablation settings in a study published by Bunch et al. [22].

We would like to highlight a prospective randomized trial conducted by Wielandts et al. [23]. They randomized 96 AF patients to HPSD (45 W) or LPLD (35 W), CLOSE protocol-guided PVI and found that fluoroscopy dose and RF time are lower in case of HPSD ablation. There was no difference in terms of six-month AF-free survival between the two groups. On the other hand, postprocedural endoscopic evaluation of esophageal lesions drew attention to a narrower safety margin at the posterior wall using high power, especially when applying higher CF and reaching higher AI values [23].

Finally, we would like to mention a recent meta-analysis of 15 studies evaluating PVI with HPSD *versus* LP ablation technique. Overall, data of 3718 patients were included in the analysis. The main result is that freedom from atrial arrhythmias was higher in case of HPSD RF ablation when compared with conventional LP RF ablation. Acute PV reconnection was lower and first-pass isolation was higher with HPSD. There was no statistically significant difference in total complications between the two groups. Total procedure duration, fluoroscopy duration, and RF ablation time were all significantly lower in HPSD ablation [24].

#### **4.2 HPSD safety endpoints: esophageal lesions**

Esophageal lesions are not rare after RF point-by-point PVI, even with the use of CF sensing catheters [25, 26]. A rare but potentially lethal complication of pulmonary vein isolation is atrio-esophageal fistula [6]. Thus, besides improving efficacy, reducing the possibility of causing esophageal lesions should be the main goal of technological developments. High-power RF applications have a larger resistive heating zone, but conductive heating does not significantly affect the lesion

creation if the application is kept short. On the other hand, one has to keep in mind that applying high power for a longer duration can cause extensive tissue injury damaging peri-cardiac structures such as the lung or the esophagus.

A study involving 85 patients who underwent PVI with 35-W power using CF-sensing catheters and AI guidance showed that the occurrence of esophageal injury after PVI is markedly low (1.2%), even in those cases where an intraesophageal temperature rise was detected during the procedure. They concluded that their strategy of delivering contiguous, relatively high-power, and short-duration radiofrequency applications is safe even at the posterior wall [27]. Two other larger trials involving 355 and 271 patients who underwent AI-guided PVI with 45–50-W power also concluded that HPSD technique is relatively safe. Esophageal lesions had a similar incidence in HPSD group as in the low-power group, and there was a low incidence of esophageal temperature elevation in the HPSD group [28]. They concluded that HPSD might even have a protective effect avoiding incidental esophageal injury due to the smaller lesion depth [29]. Of note, this is only true if high-power applications are kept very short on the posterior wall, but longerduration ablations might lead to severe complications.
