3.3 Impact of TMVI on LVOT
3D derived LVOT area and dimension were evaluated in 29 patients who underwent TMVI. The LVOT area at the level of the valve stent distal edge was newly created after TMVI. As shown in Figure 4, LVOT dimension from LV long-axis cut plane was the minimum at the valve stent distal edge, became larger at the middle portion, and was the largest at the aortic annulus [valve stent distal edge: 9.2 (5.1-15.3), middle portion: 14.0 (11.9-17.8), aortic annulus: 19.2 (17.6-20.2) mm; P < 0.001], while 3D LVOT area tended to large at the valve stent distal edge, gradually became smaller, and was the minimum at the aortic annulus [valve stent distal edge: 7.0 (4.0-11.6), middle portion: 6.0 (4.2-8.1), aortic annulus: 3.9 (3.3-4.5) cm2; P < 0.001]. Of note, there were 8 patients in whom LVOT area at the level of the aortic annulus was not the minimum; 5 had the minimum area at the valve stent distal edge and the others at the middle portion of the LVOT.
As shown in Figure 5, pre-procedural LVESV, LVEF, and 3D derived AM angle were significantly associated with LVOT area at the valve stent distal edge (LVESV: r = 0.56, P = 0.001; LVEF: r = -0.57, P < 0.001; AM angle: r = 0.68, P < 0.001). In addition, post-procedural LVESV, LVEF, and 3D derived AM angle were also associated with that area (post LVESV: r = 0.49, P = 0.007; post LVEF: r = -0.55, P = 0.002: post AM angle: r = 0.55, P = 0.003). Importantly, LVOT area at the valve stent distal edge was significantly smaller in patients with increase in LVOT gradient compared to those without increase in LVOT gradient (P < 0.001) (Table 4), and the correlation between the LVOT area and the ∆PG was strong (Figure 6). Multivariable regression analysis identified that 3D derived AM angle was independently associated with LVOT area at the valve stent distal edge (standardized regression coefficient of 0.40, P = 0.017) while LVESV and LVEF were not.
3.4 Observer variability and comparison between 3D TEE and cardiac CT
The intraobserver variability and intraclass correlation for the measurements of 3D AM angle and LVOT area were 1.5±1.6% and 8.9±7.6% and 0.96 and 0.98, respectively. The corresponding interobserver variability values were 1.9±1.8% and 9.5±6.9% and 0.94 and 0.96, respectively.
The comparison between 3D TEE and MDCT for the measurement of pre-procedure AM angle was performed. There was a significant although modest correlation between the 2 measurements (r = 0.62, P < 0.001), and Bland-Altman showed the systematic difference was 2.4° with limits of agreement of ±11.7°.