Preprocedural computed tomography angiography aortic root geometry and 1-year left ventricular reverse remodeling after transcatheter aortic valve replacement

Figures

• 

  Figure 1. CTA-derived aortic root geometric parameters. (A) Annular eccentricity index (EI) at the virtual basal annular plane: EI = (Dmax−Dmin)/Dmax. EI = 0 denotes a circular annulus; rising values indicate increasing ellipticity. (B) Sinus of Valsalva height: mean perpendicular distance from the annular plane to the sinus tip, averaged across the three sinuses. STJ, sinotubular junction.

• 

  Figure 2. Primary and supportive effect estimates across model specifications. Standardized odds ratios (95% CI) for LVRR are shown in the clinically favorable direction for (A) annular eccentricity index per 1-SD decrease and (B) sinus of Valsalva height per 1-SD increase. The filled diamond is the primary Firth M1 estimate, the open diamond is the conventional maximum-likelihood M1 sensitivity estimate, and the circles are exploratory supportive estimates. The vertical dashed line denotes an odds ratio of 1. Primary Firth M1 estimates were 2.40 (95% CI, 1.52–3.92) for EI and 1.84 (95% CI, 1.23–2.74) for sinus height. The figure is descriptive and should not be interpreted as independent confirmation or proof of robustness. CI, confidence interval; EI, eccentricity index; LVRR, left ventricular reverse remodeling; MLE, maximum-likelihood estimation; PPM, permanent pacemaker; SD, standard deviation.

• 

  Central Illustration. Preprocedural aortic-root geometry and 1-year left ventricular reverse remodeling after transapical TAVR. A lower annular eccentricity index and greater sinus of Valsalva height on preprocedural CTA were associated with higher odds of 1-year LVRR among patients who survived to undergo evaluable 1-year echocardiography. Primary Firth M1 adjusted for baseline LVEF, baseline LVEDVI, age, sex, valve morphology, STS-PROM, and LV mass index. Neutral connectors depict the observed association and do not imply a causal pathway through THV expansion, postprocedural hemodynamics, or sinus flow. EI, eccentricity index; LVEDVI, indexed left ventricular end-diastolic volume; LVEF, left ventricular ejection fraction; LVRR, left ventricular reverse remodeling; STS-PROM, Society of Thoracic Surgeons Predicted Risk of Mortality; TAVR, transcatheter aortic valve replacement; THV, transcatheter heart valve.

• 

  Figure 3. Study flow diagram. Of the 130 patients screened across 11 centers between May 2022 and May 2023, 115 were included in the final analytic cohort and 15 were excluded in mutually exclusive chronological stages: preprocedural or eligibility exclusions (n = 4: non-diagnostic CTA, n = 2; prior cardiac surgery, n = 1; more than mild baseline mitral or aortic regurgitation, n = 1); procedural or early exclusions (n = 8: failure to meet procedural-success criteria, n = 3; death within 30 days, n = 5); exclusions before the 1-year assessment (n = 2: death after 30 days but before the 1-year assessment, n = 1; loss to follow-up, n = 1); and non-diagnostic 1-year echocardiography (n = 1). LVRR was achieved in 78 patients (67.8%); 37 patients (32.2%) did not achieve LVRR. The analytic cohort was conditional on procedural success, survival, and evaluable 1-year echocardiography. CTA, computed tomography angiography; LVRR, left ventricular reverse remodeling.

• 

  Figure 4. Unadjusted continuous associations of CTA geometry with 1-year change in LVEF. Each point represents one patient (blue, LVRR; coral, no LVRR). The solid lines are unadjusted ordinary least-squares fits, and the shaded bands are 95% confidence intervals. (A) Annular eccentricity index vs. change in LVEF. (B) Sinus of Valsalva height vs. change in LVEF. The displayed Pearson r values are exploratory and unadjusted. Adjusted beta coefficients are reported separately in the Results and Table 1 and are not displayed in the figure. CTA, computed tomography angiography; EI, eccentricity index; LVEF, left ventricular ejection fraction; LVRR, left ventricular reverse remodeling.