Utility of Functional and Volumetric Left Atrial Parameters Derived From Preprocedural Cardiac CTA in Predicting Mortality After Transcatheter Aortic Valve Replacement.

Abstract

<b>BACKGROUND.</b> Cardiac CTA is required for preprocedural workup before transcatheter aortic valve replacement (TAVR) and can be used to assess functional parameters of the left atrium (LA). <b>OBJECTIVE.</b> We aimed to evaluate the utility of functional and volumetric LA parameters derived from cardiac CTA to predict mortality in patients with severe aortic stenosis (AS) undergoing TAVR. <b>METHODS.</b> This retrospective study included 175 patients with severe AS (92 men, 83 women; median age, 79.0 years) who underwent cardiac CTA for clinical pre-TAVR assessment. A postdoctoral research fellow calculated maximum and minimum LA volumes using biplane area-length measurements; these values were indexed to body surface area, and maximum and minimum LA volume index (LAVI_max and LAVI_min, respectively) values were calculated. The LA emptying fraction (LAEF) was automatically calculated. All-cause mortality within a 24-month follow-up period after TAVR was recorded. To identify parameters predictive of mortality, Cox regression analysis was performed, and results were summarized by hazard ratio (HR) and 95% CI. The Harrell C-index was used to assess model performance. A radiology resident repeated the measurements in a random sample of 20% (<i>n</i> = 35) of the cases, and interobserver agreement was computed using the intraclass correlation coefficient (ICC). <b>RESULTS.</b> Thirty-eight deaths (21.7%) were recorded within a median follow-up of 21 months. LAVI_max (HR, 1.02 [95% CI, 1.01-1.04]; <i>p</i> = .01), LAVI_min (HR, 1.02 [95% CI, 1.01-1.04]; <i>p</i> < .001), and LAEF (HR, 0.97 [95% CI, 0.95-0.99]; <i>p</i> = .002) were predictive of mortality in univariable analysis. After adjusting for clinical parameters, only LAEF (HR, 0.97 [95% CI, 0.94-0.99]; <i>p</i> = .02) independently predicted mortality. The C-index of the Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) significantly increased from 0.636 to 0.683, 0.694, and 0.700 when incorporating into the model LAVI_max, LAVI_min, and LAEF, respectively. The ICC for maximum and minimum LA volumes and LAEF ranged from 0.94 to 0.99. <b>CONCLUSION.</b> LAEF derived from preprocedural cardiac CTA independently predicts mortality in patients with severe AS undergoing TAVR. <b>CLINICAL IMPACT.</b> Cardiac CTA-derived LA function, evaluated during pre-TAVR workup, can be used to assess preprocedural risk and may improve risk stratification in post-TAVR surveillance.