Prediction of left atrial fibrosis in persistent atrial fibrillation using non-invasive methods for planning an ablation strategy

Abstract Background With regard to the recurrence rate after ablation, estimation of the extent of left atrial (LA) fibrosis is essential for planning the best treatment of persistent atrial fibrillation (AF). Currently, the only invasive measure available is electroanatomical mapping. The MRI examination is not yet sufficiently groundbreaking. Non-invasive tools provide tendencies for estimating fibrosis, but there is no summarizing score system so far. Purpose The aim of the study was to combine non-invasive tools to provide a better assessment of LA fibrosis. Methods Patients with persistent AF were examined using contrast-enhanced atrial-CT, TTE including LA strain and comprehensive P-wave analysis in sinus rhythm. Prior to ablation, electroanatomical mapping was performed in sinus rhythm if possible. Results Of the included patients, 19 of 90 (21%) patients had LA fibrosis defined as >10% low voltage area (LVA). Seven of these patients received substrate modification in addition to PVI. These patients were older (66 vs 69 years) and woman are more likely to have LVA (20% vs 47%). Neither pre-existing conditions nor LA characteristics from CT and TTE showed a correlation with the presence of LVA. Of the established P-wave parameters, only the P-wave voltage in lead I showed differences in the groups (0.05 mV vs 0.03 mV). It was striking that there were many patients with LVA<10% who nevertheless received substrate modification. These observations can be explained by the isolated LVA in the anterior and roof areas. After a more detailed analysis, 34 (37%) of the patients in the total cohort had LVA anterior and roof of >10% and 17 of these patients received substrate modification. In comparison, these patients are older (65 vs 69 years), female, have a larger atrium (LAVI 77 vs 86 ml/m²) and a worse LA strain (20 vs 14%). They are more likely to have COPD, renal insufficiency and at least moderate valve disease. Also in this group, only the voltage of the P-wave in lead I is associated with the extent of LVA anterior and roof. Interestinglythese patients have less tachycardia- induced myopathy compared to the global LA fibrosis group. Discussion In the subgroup of patients with persistent atrial fibrillation, characteristics from CT, TTE and ECG do not appear to be suitable for predicting global LA fibrosis. However, the LVA in the anterior and roof regions are particularly relevant with regard to necessary substrate modification. In this subgroup we see trends in some non-invasive parameters.