METHODS
Study patients. Three different groups of patients who underwent echocardiography at the University Hospital of Ferrara were retrospectively examined and enrolled in the present study.
TEE group . In 53 consecutive patients referred for clinically indicated TEE from January to April 2018, we measured the LVOTD during both TTE and TEE. Both measurements were performed blindly one from the other. TEE indications were: assessment of aortic and or mitral valve disease severity (N=25), exclusion of left atrial appendage thrombi or cardiac masses (N=13), evaluation for patent foramen ovale (N=8), and suspected endocarditis (N=7). There was no exclusion criterion except unmeasurable LVOTD during TTE.
Derivation group . This group was used to derive a regression equation for predicting LVOTD using its independent determinants. It included 340 consecutive patients with heart failure who underwent echocardiography from January to August 2018.
Validation group . This group was used to validate the derived regression equation and compare it with the previously published one based on BSA (14). It included 200 consecutive patients with heart failure who underwent echocardiography from September to December 2018.
For both the derivation and validation groups, exclusion criteria were history of aortic valve or aortic root surgical or transcatheter intervention (N=17, derivation group; N=14, validation group), and poor echogenicity that could affect aortic root, LVOT, LV end-diastolic diameter (LVEDD) and LV volume measurement (N=39, derivation group; N=18 derivation group). Aortic stenosis (aortic valve area ≤1.5 cm2) was not considered an exclusion criterion. Demographic and clinical patients’ characteristics were collected.
TTE examination . A comprehensive two-dimensional echocardiographic, Doppler and color Doppler examination was performed using a GE Vivid 7 or E9 echo scanner (GE Health Care, Milwaukee, US) equipped with a 3.5 MHz transducer. Echocardiographic images were stored in digital format and analyzed using the EchoPAC software v. 201 (GE Health Care, Milwaukee, US). One trained physician did all the echocardiographic measures, according with the American Society of Echocardiography/European Association of Cardiovascular Imaging guidelines (17). LV end-diastolic and end-systolic volumes were calculated from orthogonal apical views using the biplane Simpson method. LV ejection fraction was derived from the standard equation (17). Two-dimensional linear internal measurements of the LV were acquired in the parasternal long-axis view carefully obtained perpendicular to the LV long-axis, and measured at the level of the mitral valve leaflet tips (17). Electronic calipers were positioned at the interface between myocardial wall and cavity and the interface between wall and pericardium (17). The diameter at the level of the sinuses of Valsalva (SVD) was measured at end-diastole, in a strictly perpendicular plane to that of the long-axis of the aorta using the leading-edge-to-leading-edge convention (17). Aortic root dilatation was defined as SVD indexed to BSA (SVDI) \(\geq\)21mm/m2, according with previous adult normal reference studies (17-19). LVOTD was measured using the zoom mode at the insertion of the leaflets (at the annulus level) in midsystole with the inner-edge-to-inner-edge approach from the parasternal long-axis view (7,17). LVOTBSA was calculated with the equation LVOTDBSA=12.1+5.7xBSA (14), with BSA calculated with the Mosteller formula (14,20). For intraobserver variability, measurements were repeated by the same physician one week later; for interobserver variability, measurements were performed by a second physician who used the same criteria and was blind to the results of the first observer (17).
TEE examination . The examination was performed using a GE Vivid E9 echo scanner equipped with a 5 MHz transducer. Echocardiographic images were stored in digital format and analyzed using the EchoPAC software v. 201. The LVOTD was measured using the zoom mode, as described for TTE, on images acquired in the aortic long-axis view at the mid-esophageal position using an image plane with about 120° rotation (21). Measurements were performed by an independent operator who was blind to the measurements performed on the TTE images.
Statistical analysis and study endpoints. Normal distribution was tested with the Kolmogorov-Smirnov test. Continuous variables were expressed as mean and standard deviation or median values with 25th and 75th percentiles if normally or non-normally distributed, respectively. Categorical variables were reported as counts and percentages. For continuous normal and non-normal variables, Student’s t-test and Mann-Whitney U test were respectively used for comparisons between two unpaired groups. Categorical variables were compared by the chi-square test. Pearson correlation was used for normally distributed variables, whereas Spearman correlation was used if at least one variable had non-parametric distribution. Comparison of LVOTD measured during both TTE and TEE was performed using a paired Student’s t test. To assess for error and bias, the Bland-Altman method was used. The correlation between LVOTD, anthropometrics data (patient BSA and height), LVEDD, SVD and age were studied with univariate linear regression analysis in the derivation group. For the primary endpoint of the derivation study, the covariates that were found to be statistically correlated with LVOTD were included in the multivariate linear regression model (Model 1) in order to derive a regression equation for predicting LVOTD (LVOTDRE1). LVEDD and SVD were included as non-indexed values for this analysis to facilitate the utilization of the derived formula. The independent covariates from Model 1 were included as indexed values in a secondary multivariate linear regression analysis with the difference between measured LVOTD (LVOTDM) and LVOTDBSA as endpoint, to corroborate the results of Model 1. Secondary endpoints for the derivation group were: 1) an analogous multivariate linear regression model (Model 2) including LV end-diastolic volume (LVEDV) instead of LVEDD as covariate; 2) a repeated multivariate linear regression analysis with the addition of gender (Model 3, with LVEDD; Model 4, with LVEDV). A regression equation was derived for each secondary multivariate model (LVOTDRE2, LVOTDRE3, LVOTDRE4). For the primary endpoint of the validation study, the Student’s t-test for paired samples was used to compare LVOTDRE1 and LVOTDBSA (14) with LVOTDM in the validation group. As a secondary endpoint, the other regression equations (including LVEDV and/or gender) were also tested. Multicollinearity between the variables in all the multivariate models was assessed by calculation of the variance inflation factor, with a value \(\geq\)5 indicating significant collinearity. The intra- and interobserver variability for SVD, LVEDD, LVEDV, LVOTDM, LVOTDRE1, LVOTDRE2 and their derived LVOT areas were assessed in a random sample of 25 patients. Observer variability was determined as the standard deviation of the mean error and expressed as percentage of the first measure for each variable. Data were analyzed using the IBM SPSS Statistics software, v. 24. Differences were considered statistically significant for P<0.05. The study was approved by the local Ethics Committee.