Choice of CoreValve prosthesis size in case of disagreement between echocardiography and computed tomography: impact on procedural and 1-year results

The measurement of the aortic annulus is critical for prosthesis selection in transcatheter aortic valve implantation. The choice of the gold standard for noninvasive cardiac imaging between transthoracic echocardiography (TTE), transoesophageal echocardiography (TEE) and multislice computed tomography (MSCT) is still debated. Our aim was to evaluate procedural and device success and 1-year results in patients with discrepancy between TTE (or TEE) and MSCT regarding prosthesis choice. Methods: Aortic annulus diameter was measured with TTE and MSCT, and with TEE in case of discrepancy between TTE and MSCT. When TEE and MSCT were in disagreement, the size of the Medtronic CoreValve was chosen based on TEE measurement. We compared the procedural results, as well as the 1-year clinical and echocardiographic results between patients receiving a small CoreValve based on TEE (in disagreement MSCT) and patients with agreement between TTE/TEE and MSCT on CoreValve size. Results: TTE/TEE and MSCT agreed on prosthesis size choice in 105 out of 131 patients (80.2%) (Group A); in 22 patients (16.8%) a small CoreValve was implanted, although MSCT suggested a large prosthesis (Group B), while a large CoreValve was implanted in 4 patients (6.3%) in spite of MSCT measurements. The aortic annulus was more oval in Group B (large-to-short axis ratio 1.46±0.19 vs 1.40±0.21, p=0.08). Group B patients showed similar procedural success (95.5% vs 95.2%, p=0.99), device success (90.9% vs 89.5%, respectively; p=0.99), and valve-in-valve rate (4.5% vs 5.7%, p=0.99). New left bundle branch block occurred in 27.3% vs 37.1% of Group B vs A, respectively (p=0.38), while new pacemaker implantation was required in 18.2% vs 23.8% of Group B vs A, respectively (p=0.60). A trend to a higher rate of more-than-mild paravalvular leak was observed in Group B (36.4% vs 21.9%, p=0.16). At 1-year follow-up, more-than-mild paravalvular leak was observed in 4 out of 12 patients (33.3%) of Group B vs 6 out of 51 patients (11.8%) in Group A (p=0.05). No significant difference in terms of NYHA functional class and of left ventricular mass and volumes was observed between groups. Conclusions: Disagreement between TTE/TEE and MSCT on CoreValve size choice occurred frequently (19.8%). Those patients who, according to TEE, received a smaller CoreValve size than indicated by MSCT had similar procedural success than patients for whom TTE/TEE and MSCT were in agreement. However, a trend to a higher rate of postprocedural paravalvular leak was observed, reaching statistical significance at 1 year.