Background: The use of a novel radiofrequency balloon (RFB) for pulmonary vein isolation (PVI) is currently investigated in selected centres. Occurrence of silent cerebral lesions (SCL) detected by magnetic resonance imaging (MRI) after catheter ablation has already been researched for well-established energy sources such as the cryoballoon and irrigated-tip radiofrequency catheters. We investigated the incidence of SCL after RFB ablation. 

Methods: A total of 105 patients were treated with RFB PVI between December 2020 and November 2022. Cerebral MRI was performed in 38 unselected patients after a median of 1 day after the procedure. Neurological symptoms as well as size and location of SCL were assessed. 

The RFB allows for assessment of real time isolation using a spiral catheter which can be inserted through a central lumen of the balloon. During the investigated period the procedure underwent a streamlining process. Initially a 3D high resolution map of the left atrium (LA) was obtained before and after the ablation. To create this post-ablation map, the RFB had to be exchanged for a circumferential mapping catheter. After the first 50 procedures observation of the real time isolation was deemed sufficient for acute ablation success and the post-ablation re-mapping was omitted. This reduced the number of necessary catheter exchanges through the transseptal sheath. 

Results: None of the patients developed neurological symptoms after the procedure. SCLs were detected in 28,9% (11) of patients. Analysis of patient baseline characteristics (gender, age, BMI, comorbidities, left atrial size, left ventricular ejection fraction) as well as procedural data (total procedure length, LA dwell time, ablation time, total number of radiofrequency applications, intraprocedural cardioversion, intraprocedural activated clotting time) showed no significant independent predictor for SCL. 

We compared procedural data and SCL incidence prior to streamlining of the procedure (non-streamlined, NSL) to those after streamlining (SL). Incidence of SCL in NSL was 42,1% (8 of 20) compared to 16,7% (3 of 18) in SL. This difference did not reach statistical significance (p = 0,159). Procedure time (NSL: 69.4 ± 10.8, SL: 51.1 ± 8.6, p = 0.009) and LA dwell time (NSL: 63.8 ± 10.9, SL: 45.8 ± 8.3, p = 0.009) were significantly shorter in the SL group. 

Conclusion: We initially observed a higher-than-expected number of silent cerebral lesions after pulmonary vein isolation with a novel radiofrequency balloon. After streamlining of the procedure and especially reducing the necessary catheter exchanges the incidence of SCL dropped. It is now comparable to data of SCL after PVI with other technologies in our centre. This supports our assumption that the number of left atrial catheter exchanges might be an important factor for the development of SCL. Furthermore, this data emphasize the necessity of caution during the initial learning phase with new ablation devices. Further MRI after the procedure could reveal statistically significant predictors of SCL.