Article Text
Abstract
Background Patients with atrial fibrillation (AF) frequently have atrial scarring characterised by discrete regions of low voltage. Pre-existing left atrial scarring is an independent predictor of pulmonary vein isolation (PVI) failure. Novel mapping algorithms have also been developed to assess the degree of atrial fibrosis. This may be expressed as a percentage of the total left atrial (LA) volume mapped. In addition to the electrical remodelling seen, structural remodelling occurs, with dilatation and reduced function. The most accurate determinant of LA function is debated, but the most frequently used method is transmitral A Wave velocity on pulsed-wave Doppler. The relationship between LA function and electrical changes seen in AF has not been defined.
Methods This was a single centre observational study. Left atrial voltage maps were created in patients undergoing PVI for the first time in the Mater Private Hospital between August 2016 and April 2017. LA voltage maps were initially created with a Lasso catheter with some further points taken with a Smarttouch ablation catheter (both Biosense Webster, Diamond Bar, California). Voltage greater than 0.5 mV was accepted as normal tissue and voltages < 0.2 mV scar. After creation of the voltage maps, the percentage scar was assessed using a novel computer algorithm (Biosense Webster). Pre-ablation echocardiograms were studied and the LA function was assessed by measuring the trans-mitral pulse wave Doppler. Assessments were only made in sinus rhythm.
Results Out of 96 patients who had undergone PVI, only 24 were found to have had sinus rhythm on pre-procedural echo. The mean age was 63.5 (standard deviation or SD 10.6) years. 66% of the group were men. 58% had paroxysmal AF. The mean amplitude of the A-wave in the study was 0.61 (SD 0.16) ms-1. An average of 1269.7 (SD 857.0) mapping points were taken. The mean LA percentage scar was 25.1 (SD 20.3) %. Using linear regression, adjusted for age at time of procedure, there was a significant negative association between a wave (in ms-1) and % LA scar (Beta coefficient -72.44, 95% CI -122.99 to -21.88, p=0.007). Using pairwise correlation, the correlation coefficient between LA scar and A wave was -0.38, p=0.06.
Conclusion Our study found an inverse correlation between transmitral A wave and degree of left atrial scarring when LA function was adjusted for patient age, indicating that LA electrical remodelling as measure by percentage scar is associated with decreased LA function in patients with AF. However, there were only 24 patients in this study and ongoing research with more patients is warranted to further substantiate this.