Transcatheter tricuspid valve replacement: Determining the easiest venous approach and optimal prosthesis sizing with CT-scan

Abstract

Because of its high mortality, isolated tricuspid regurgitation (TR) surgery is often not performed. In this setting, transcatheter tricuspid valve replacement is currently undergoing development. However, optimal venous approach and prosthesis size have yet to be optimized. We sought to assess the determinants of venous route and pitfalls of tricuspid prosthesis implantation using CT-scan analysis. We analyzed at end-diastole 89 cardiac CT-scans (performed prior to transcatheter aortic valve replacement) of patients with ( n = 19) and without ( n = 70) moderate to severe TR. Tricuspid annulus (TA) area, anteroposterior (AP) and septolateral (SL) diameters were measured. Then, minimal lengths between the TA and the right-ventricle outflow tract (RVOT) and coronary sinus were measured to assess the risk of obstruction. Last, angles of the TA center to the superior and inferior vena cava were measured to determine the easiest veinous route. Patients with TR had larger TA area ( P = 0.02), SL ( P = 0.02) and AP ( P = 0.04) diameters than patients without TR. In patients with TR, the eccentricity index averaged 0.88 ± 0.08 with the SL diameter greater than the AP diameter [44.1 mm (95% CI, 29.7–58.5) vs. 40.3 mm (95% CI, 24.1–56.5), P < 0.01]. In the whole population, the minimal length from the TA to the coronary sinus was 11 mm (95% CI, 3.6–18.0) and was 17 mm (95% CI, 10.1–24.5) to the RVOT. The angle from superior vena cava to TA center averaged 143° (95% CI, 123–163), while the angle from inferior vena cava was not only tighter (mean = 94°) but with a broader range of value (95% CI, 52° to 136°). Coronary sinus and RVOT positions relative to the TA and vena cava angulations were not modified by the presence of TR. Development of percutaneous valve in tricuspid position may be challenged by the close position of the coronary sinus. The transjugular approach appears straighter with less individual variability than the transfemoral approach.