Multimodality evaluation of aortic valve area for pre-procedural planning of transcatheter aortic valve replacement – a multi-center study

Introduction: Transcatheter aortic valve replacement is an established treatment option for patients with symptomatic severe aortic stenosis (AS), especially at increased risk for surgical complications. In pre-procedural planning, transthoracic echocardiography (TTE) is the recommended modality for aortic stenosis severity grading and subgroup differentiation. In challenging or discordant cases, transesophageal echocardiography (TOE) is often performed for direct aortic valve area (AVA) planimetry in clinical practice. With emerging technical development contrast-enhanced computed tomography angiography (CTA) can non-invasively provide supplementary information, including AVA planimetry (AVA_CTA) and precise annulus plane measurements for optimal valve sizing. However, there is limited knowledge about modality specific thresholds for AS severity grading and the correlation of the AVA_CTA with AVA_TTE or AVA_TOE. In this multi-center study, we aimed to close this gap by directly comparing AVA measurements obtained from CTA and echocardiography. Through its numerous advantages AVA_CTA could improve the diagnostic process, aid in the assessment of challenging or borderline cases of moderate to severe AS and make TOE redundant in many instances.

Methods and Results: A single-center derivation cohort (176 patients; 93 men; mean age 80.0 ± 7.7 years) of patients with moderate to severe AS who received full-cycle CTA, TTE and TOE were included in our retrospective analysis. Planimetry of AVA_CTA was performed by two independent raters with good inter-observer correlation coefficients of 0.80. The mean AVA_CTA was 95.6 ± 23.8 mm² compared to a mean AVA_TOE of 87.6 ± 22.9 cm². The correlation coefficients for AVA_CTA were considered acceptable at a value of 0.73 for TOE and 0.61 for TTE. The best cut-off value for AS severity grading was evaluated by ROC-analysis at an AVA_CTA of 95.7 mm² (area under the curve (AUC) 0.846, sensitivity 71.7%, specificity 89.8%, accuracy 76.8%) and AVA_CTA of 94.7 mm² (AUC 0.774; sensitivity 63.4%; specificity 84.4%, accuracy 68.8%) compared to TOE and TTE classification. Afterwards these cut-off values were applied in a second multi-center included validation cohort (407 patients, 215 men, mean age 80.9 ± 6.7 years) with comparable baseline characteristics but smaller mean AVA_CTA (88.1 ± 21.5 mm²) and mean AVA_TOE (83.3 ± 23.8 cm²). Likewise, correlation coefficients for AVA_CTA were slightly lower. Using the above-mentioned cut-off values resulted in acceptable results for both AVA_CTA of 95.7 mm² compared to TOE (AUC 0.817, sensitivity 78.2%, specificity 72.6%, accuracy 76.6%) and AVA_CTA of 94.7 mm² compared to TTE AS severity classification (AUC 0.698, sensitivity 68.9%, specificity 60.8%, accuracy of 67.3%).

Conclusion: AVA_CTA is a reliable parameter for AS classification, showing the best - albeit only acceptable - correlation with AVA_TOE in both derivation and validation cohort. The threshold for severe aortic stenosis based on AVA_CTA is lower (95.7 mm²) than those for AVA_TTE and AVA_TOE and shows acceptable diagnostic performance in AS severity grading. AVA_CTA can serve as an additional parameter in the multiparametric approach of clinical decision-making, potentially rendering additional TOE planimetry unnecessary for many patients.