https://doi.org/10.1007/s00392-025-02625-4
1Klinikum rechts der Isar der Technischen Universität München Klinik und Poliklinik für Innere Medizin I München, Deutschland
Background:
While pulmonary vein isolation (PVI) is the standard ablation treatment for paroxysmal atrial fibrillation (AF), the best approach for persistent AF remains unclear. The TAILORED-AF trial recently showed that combining PVI with ablation targeting AI-identified spatiotemporal dispersion improves outcomes for persistent AF (pers AF). It is uncertain, however, whether this method applies to patients presenting in sinus rhythm after cardioversion, where AF induction would be required before ablation. A critical assumption is that spatiotemporal dispersion extent and distribution are comparable in both spontaneous and induced AF.
Aim:
The aim of our study was to compare the location and distribution of spatiotemporal dispersion areas detected by the Volta AF-Xplorer software in spontaneous AF (spont-AF) and induced AF (ind-AF).
Methods:
Patients with persAF who received a tailored ablation plus PVI were included. In our clinical practice, we first create a left atrial (LA) electroanatomic spont-AF map, followed by electrical cardioversion and creation of a second electroanatomic map during atrial pacing. If pacing induced AF, an ind-AF map was generated. The LA was divided into 16 standard subregions, and the AI-adjudicated dispersion extent and distribution (AI-DED) from both maps were compared visually. Two analyses were performed: (1) a dichotomous comparison for the presence of spatiotemporal dispersion in spont-AF vs. ind-AF, and (2) a semi-quantitative comparison categorizing each subregion’s dispersion extent as 0%, <10%, >10–50%, or >50%.
Results:
Among the 63 patients with persAF who underwent ablation procedures at our institution from April 2024 to November 2024, ten patients (mean age 78±9 years, 50% female, 40% long-standing persAF) received a second electroanatomic map of induced AF. AF termination was achieved in 7 out of 10 patients (70%) during ablation. In those where AF termination was not achieved, the mean AF cycle length measured in the LA appendage increased from 132±12 ms prior to the procedure to 144±13 ms post-procedure (p=0.25). A total of 160 segments were mapped in spont-AF and ind-AF, followed by assessment and analysis of dispersion. Spatiotemporal dispersion was detected in 98 of the 160 analyzed spont-AF subregions and in 97 of the 160 analyzed ind-AF subregions, as identified by the Volta AF-Xplorer software (p=0.909). In only 5 out of 160 cases (3%), the characterization of a subregion during ind-AF differed from its previous characterization in spont-AF. For the semi-quantitative analysis, the subregions showed the following percentages of dispersion: spont-AF: 0% - 62 subregions, <10% - 45, >10 to 50% - 33, >50% - 20; ind-AF: 0% - 63, <10% - 48, >10 to 50% - 28, >50% - 21; p=0.817. In 20 cases (12.5%), there was a one-category discrepancy between the spont-AF and ind-AF maps, while in the remaining 140 cases (87.5%), the classification of the regions remained consistent. There were no cases with a discrepancy of two or more categories.
Conclusion:
We found that artificial intelligence-adjudicated dispersion extent and distribution were highly similar between spontaneous AF and induced AF. Thus, physicians may consider inducing AF prior to ablation and incorporating the ind-AF map into their ablation strategy. Further studies are necessary to assess whether the findings of the TAILORED-AF trial can be extended to persistent AF patients in whom AF is induced prior to an ablation procedure.