ABC | Volume 114, Nº4, April 2020

Original Article Correia et al. CA lesions and AF recurrence - a meta-analysis Arq Bras Cardiol. 2020; 114(4):627-635 In addition to that, animal studies have demonstrated that 80% of AF triggers are located in the posterior wall, including the pulmonary vein (PV) region. 22 A previous meta-analysis has shown that isolation of a part of the posterior LA reduces the recurrence of AF after CA. 23 Therefore, an increase in the extent of LA ablation may promote greater substrate modification, decreasing the amount of viable LA tissue capable of harboring AF by overlapping PV and non-PV triggers with ablation lesions. PV scarring The clinical application of real-time MRI may make it possible to visualize LA scarring during the procedure, making it easier to induce scarring. 24 However, as real-time MRI is still a new and expensive imaging method, alternatives such as driver-guided CA by electroanatomic mapping to visualize LA scars might be an option for optimizing outcomes. A recent meta-analysis by Ramirez et al. 25 reported an association between driver-guided CA for AF and increased freedom from AF, in comparison with conventional strategies. However, this meta-analysis included primarily nonrandomized studies of moderate quality. Future observational studies can help build evidence to prove whether electroanatomic mapping can assist in creating contiguous scar lesions around the PV. Risks of targeting more LA scarring Even though this meta-analysis shows that more extensive ablation reduces the risk of AF recurrence, this strategy is not risk free, given that the procedure may decrease LA compliance, LA volume, and LA systolic function, which may induce the development of the stiff left atrial syndrome (SLAS). 26 SLAS, which was described in 1988 by Pilote et al., 27 is characterized by a decline in LA diastolic function and pulmonary hypertension. 28 Although this may represent a severe consequence of RFCA, in a case series study by Gibson et al., the condition was reported in only 1.4% of patients who underwent RFCA. 28 Furthermore, previous studies found that LA scar volume after CA was associated with depressed LA systolic function. 26,29 Ablation scars in the posterior LA wall, however, had less effect on LA systolic function. 26 Another risk of CA that extensive ablation may increase is the possibility of esophageal injury due to the anatomical relationship between the esophagus and the posterior LA wall. 29 The esophagus is separated from the posterior LA by a thin layer of fat, being prone to injury during AF ablation. 30 Possible esophageal injuries include perforation, atrio-esophageal fistula formation, and peri-esophageal nerve injury. 30 To minimize the potential risks of esophageal Table 1 – Characteristics of the included studies and quality evaluation Study, year Region Type of Study N Paroxys mal, N (%) AF detection method Follow up Quality p thresh old McGann et al., 2008 9 North America Single center, prospective, observational 46 22 (48%) Patient reports, event monitoring, Holter monitoring, and ECG data. 3 months Good 0.05 Peters et al., 2009 10 North America Single center, prospective, observational 35 19 (54%) 7-day event monitor at multiple intervals 6.7 ± 3.6 months Good 0.05 Badger et al., 2010 11 North America Single center, prospective, observational 144 57 (40%) 8-day Holter monitoring and ECG at 3 months, 6 months, and 1 year 10.23 ± 5.14 months (range, 6 to 20 months) Good 0.05 Akoum et al., 2011 12 North America Single center, prospective, observational 120 50 (42%) 12-lead ECG and 8-day Holter monitor at 3 months after ablation and in 3-month intervals thereafter. Additional ECG were obtained when patients reported symptoms. 283 ± 167 days Good 0.05 McGann et al., 2011 13 North America Single center, prospective, observational 37 NR NR 1 year Good 0.05 Hunter et al., 2013 16 Europe Multicenter, prospective, observational 50 50 (100%) 7 days of ambulatory ECG monitoring at 3 and 6 months 6 months Good 0.05 Akoum et al., 2015 14 North America, Europe, and Oceania Multicenter, prospective, observational 177 116 (66%) ECG or ambulatory monitor recordings At least 1 year Good 0.05 Parmar et al., 2015 15 North America Single center, prospective, observational 94 45 (48%) 12-lead ECG and 30-day event monitor at 3 and 6 months and 1 year, and every 6 months thereafter. Patients who experienced symptoms were given additional ECG and Holter monitors Mean follow up of 336 days Good 0.05 AF: atrial fibrillation; ECG: electrocardiogram; LA: left atrium; LGE-MRI: Late gadolinium enhanced magnetic resonance imaging; NR: not reported. 630