Additional value of lateral tissue Doppler imaging in the assessment of diastolic dysfunction among subjects with pseudonormal pattern of mitral inflow

This is a single center study performed at Seoul National University Boramae Medical Center (Seoul, Korea). Between March and October 2012, a total of 661 consecutive subjects in sinus rhythm without significant structural abnormalities in transthoracic echocardiographic (TTE) examination were prospectively recruited. Subjects with LV ejection fraction < 50%, valvular stenosis or regurgitation of more than mild degree, regional wall motion abnormalities, LV hypertrophy, mitral annular calcification, prosthetic heart valve, and intra-cardiac shunt or pericardial effusion were excluded from the study. Subjects with poor sonic window or poor cooperation during the exam were also excluded. These 661 subjects were grouped according to mitral inflow patterns: a group with E/A ≥ 1 (n = 385) and the other group with E/A < 1 (n = 276). Among 385 subjects with E/A ≥ 1, the present study focused on 122 subjects with septal e’/a’ < 1, who were frequently considered to have a pseudonormalization suggesting advanced stage of diastolic dysfunction [5, 6]. From clinical experience, we have noticed that these subjects can be stratified into 2 groups with different diastolic dysfunction severities by the lateral TDI pattern. In order to investigate the usefulness of lateral TDI in the assessment of diastolic function, these 122 subjects were then classified according to the pattern of lateral TDI. One group of subjects had lateral e’/a’ < 1, and designated as “concordance (to medial TDI pattern) group” (n = 50). The other group had lateral e’/a’ ≥ 1, and this group was designated as “discordance (to medial TDI pattern) group” (n = 72). Study flow for enrollment is shown in Figure 2. Approval of the study protocol was obtained from the Institutional Review Board of Seoul National University Boramae Medical Center (Seoul, Korea). Informed consent was waived due to the routine nature of the information collected.

TTE was performed according to the recommendations of current guidelines [9‐11] using commercially available equipments (Sequoia, Siemens Medical Solutions and Vivid 7, GE Medical Systems) in the left lateral decubitus position. M-mode echocardiographic features were used to measure LV and LA dimensions, with LV ejection fraction from the parasternal short axis view at the mitral chordae level. LA volume was calculated using the prolate ellipsoid method [9]. Peak early transmitral filling velocities during early diastole (E), late diastole (A) and deceleration time (DT) were imaged in the apical four chamber view at the tip of mitral leaflets. Color-coded TDI was applied to the apical four chamber view to determine the mean early (e’) and late (a’) velocities at both septal and lateral mitral annuli. Two experienced cardiosonographers performed the echocardiography. Interobserver agreement for lateral E/e’ was evaluated by Spearman’s correlation among 50 subjects. The correlation coefficient was 0.905 in our laboratory.

LV diastolic function was graded according to the current guidelines [10]. Normal LV diastolic function was defined as septal e’ ≥ 8 cm/s and LA volume index (LAVI) < 34 ml/m². Grade II diastolic dysfunction was defined as septal e’ < 8 cm/s, LAVI ≥ 34 ml/m², DT 160–200 ms and septal E/e’ ≥ 9. Because we had only enrolled subjects with E/A ≥ 1, there were no subjects with grade I diastolic dysfunction, which is defined as E/A < 0.8. Grade III diastolic dysfunction was defined as septal e’ < 8 cm/s, LAVI ≥ 34 ml/m², E/A ≥ 2, DT < 160 ms and septal E/e’ ≥ 13. There were no subjects with grade III diastolic dysfunction in the study.

Continuous variables were presented as mean ± standard deviation (SD), and categorical variables were expressed as percentages. Continuous variables were compared using the Student t test, and categorical variables were compared using the Chi-square test. Pearson’s correlation method was used to assess the relationship between two continuous variables. Multiple linear regression analysis was performed to assess independent association between LAVI and lateral e’/a’ by adjusting for age, sex, body mass index and history of hypertension and diabetes. To identify the best cut-off value of lateral e’/a’ as a predictor of LAVI ≥ 34 ml/m²[10], receiver operating characteristic (ROC) curve analysis was used. A p- value < 0.05 was considered statistically significant. All statistical analyses were conducted using SPSS 18.0 (Chicago, IL, USA).

The results of comparative analyses between the concordance and discordance groups are also shown in Table 1. Subjects in the concordance group were older (59.6 ± 14.1 versus 53.5 ± 12.0 years, p = 0.012). Otherwise, there were no significance differences in terms of cardiovascular risk factors including sex, body mass index, hypertension, diabetes and dyslipidemia between the 2 groups (p > 0.05 for each). LV end-diastolic (LVEDD) and end-systolic dimensions (LVESD) were greater for the concordance group than those of discordance group (p < 0.05 for each). LAVI was significantly higher for the concordance group than that for the discordance group (26.4 ± 8.8 versus 17.9 ± 6.6 ml/m², p < 0.001). Septal e’ wave velocity was significantly lower (6.4 ± 1.7 versus 7.5 ± 1.6 cm/s, p < 0.001) and E/e’ ratios were significantly higher (14.5 ± 6.4 versus 10.7 ± 3.5, p < 0.001) in the concordance groups than those in the discordance group.

According to current guidelines [10], study subjects were grouped into normal and grade II diastolic dysfunction. If big discrepancies existed between parameters for diastolic dysfunction grading, this group of subjects was regarded as undetermined. The number of subjects in the normal, grade II and undetermined groups was 53 (43%), 29 (24%) and 40 (33%), respectively. Among subjects with lateral e’/a’ ≥ 1, only 4 subjects (9.3%) had grade II diastolic dysfunction, whereas among subjects with lateral e’/a’ < 1, the majority of subjects (n = 25, 64.1%) had grade II diastolic dysfunction (p < 0.001) (Figure 3).

LAVI is a useful indicator of LV diastolic function [12]. We tested whether lateral e’/a’ had an association with LAVI. In a simple correlation analysis, lateral e’/a’ was significantly associated with LAVI: increases in lateral e’/a’ was associated with decreases in LAVI (r = −0.469, p < 0.001) (Figure 4A). It is well known that Doppler tissue e’ velocity decreases with age [13]. Therefore, age was a powerful confounder, and this was addressed in this study. As expected, lateral e’/a’ had a significant correlation with age, and decreased with advancing age (r = −0.429, p < 0.001) (Figure 4B).

Subsequently, potential confounders, including age, were addressed using a multivariable model. A multiple linear regression analysis showed that lateral e’/a’ was independently associated with LAVI (β = −0.484, p < 0.001), even after adjusting for potential confounders including age, sex, body mass index, hypertension and diabetes (Table 2). In the ROC curve analysis, the sensitivity and specificity for detection of LAVI ≥ 34 ml/m² were 88.9% and 67.0%, respectively, with lateral e’/a’ of 0.89 as the best cut-off value (Figure 5).

Lateral e’ has been known to be a maker of diastolic dysfunction [10]. When the data was analyzed using lateral e’ velocity instead of lateral e’/a’, the lateral e’ velocity was significantly lower in subjects with grade II diastolic dysfunction than that in those with normal diastolic function (7.4 ± 1.9 versus 11.6 ± 2.0 cm/s, p < 0.001). Lateral e’ velocity was also negatively associated with LAVI even after adjusting for potential confounders (β = −0.360, p = 0.011).

The data was analyzed for any correlations between lateral e’/a’ and LAVI in subjects with normal diastolic function (n = 263) and grade I diastolic dysfunction (relaxation abnormality, n = 276). In the univariate analysis, there was a significant correlation between lateral e’/a’ and LAVI in subjects with normal diastolic function (r = −0.348, p < 0.001), but this was not an independent association as it was not significant in the subsequent multivariable analysis (β = −0.175, p = 0.086). There were no significant correlations between lateral e’/a’ and LAVI in subjects with grad I diastolic dysfunction in both univariate and multivariable analyses.

The study sample size is small. Enrolled subjects were healthy without cardiac structural abnormalities, and therefore, the results of the study cannot be generalized to other populaitons. The most widely used reference parameter for diastolic function is LV relaxation time; however, LV catheterization was not performed in this study. Other parameters which can be measured during echocardiography, such as mitral inflow pattern with the Valsalva maneuver, pulmonary venous flow pattern and color M-mode propagation velocity, were not evaluated in this study.