Comparison of Echocardiographic and Catheter Mean Gradient to Assess Stenosis After Transcatheter Aortic Valve Implantation

Abstract

Discordance exists between Doppler-derived and left heart catheterization (LHC)-derived mean gradient (MG) in transcatheter aortic valve implantation (TAVI). We compared echocardiographic parameters of prosthetic valve stenosis and LHC-derived MG in new TAVIs. In a retrospective, single-center study, intraoperative transesophageal echocardiogram (TEE)-derived MG, LHC-derived MG, and acceleration time (AT) were obtained before and after TAVI in 362 patients. Discharge MG, AT, and Doppler velocity index (DVI) using transthoracic echocardiogram (TTE) were also obtained. MG ≥10 mm Hg was defined as abnormal. During native valve assessment with pre-TAVI TEE and pre-TAVI LHC, Pearson correlation coefficient revealed a nearly perfect linear relation between both methods' MGs (r=0.97, p <0.0001). Intraoperatively, after TAVI, Spearman correlation coefficient revealed a weak-to-moderate relation between post-TAVI TEE and LHC MGs (r=0.33, p <0.0001). Significant differences were observed in categorizations between post-TAVI TEE MG and post-TAVI AT (McNemar test p=0.0003) and between post-TAVI TEE MG and post-TAVI LHC MG (signed-rank test p <0.0001), with TEE MG more likely to misclassify a patient as abnormal. At discharge, 30% of patients had abnormal TTE MG, whereas 0% and 0.8% of patients had abnormal DVI and AT, respectively. Discharge TTE MG was not associated with death or hospitalization for heart failure at a median follow-up of 862days. Post-TAVI Doppler-derived MG by intraoperative TEE was higher than LHC, despite being virtually identical before implantation. At discharge, patients were more likely to be classified as abnormal using MG than DVI and AT. Elevated MG at discharge was not associated with death or hospitalization for heart failure.