Assessment of diastolic function during exercise echocardiography:: Annulus mitral velocity or transmitral flow pattern?

Objectives: We hypothesize that the change in the left ventricular (LV) diastolic pattern (DP) may be measured with high reproducibility and correlates with exercise echocardiography (EE) better than the ratio of early LV inflow velocity to early diastolic annulus velocity (E/e' index). Background: The E/e' index has been related to LV filling pressures but has not been compared with DP. Methods: We selected 179 consecutive patients who were referred for EE. Early (E) and late (A) LV inflow velocities by conventional pulsed Doppler, and septal annulus e' velocity by pulsed Doppler myocardial imaging were measured at rest (R) and post-exercise (PE). Results: Four LV-DPs were found: abnormal relaxation (AR) at R and PE (E < A) in 110 patients; AR at PE (E > A at R; E < A at PE) in 22 patients; restrictive pattern (RP) at R and PE (E > A) in 18 patients; and RP at PE (E < A at R; E > A at PE) in 29 patients. The more accurate PE cutoff E/e' values to predict abnormal EE, ischemic response, poor functional capacity (<8 Mets in men; <6 Mets in women), and lack of increase in left ventricular ejection fraction (LVEF) were 12, 12, 11, and 11 (areas under the curve were 0.53, 0.53, 0.63, and 0.57, respectively). Corresponding areas under the curve for an RP at R + PE or only at PE were 0.57, 0.55, 0.54, and 0.56 (P = not significant). The sensitivity of an RP at R + PE or only at PE was lower and the specificity was higher than those of the different E/e' cutoff values for predicting abnormal EE, functional capacity, ischemic response, and lack of increase in LVEF. Achieved Mets were lower in patients with an RP at R + PE or only at PE irrespectively of the E/e' values, whereas achieved Mets in patients with AR at R + PE or only at PE were lower if the E/e' was >= 11 (8.2 +/- 2.9 vs. 9.8 +/- 3.1, P =.01). Interobserver and intraobserver concordance were 95% (kappa = 0.86) and 100% (kappa = 1.0) for an RP, and 86% (kappa = 0.73) and 92% (kappa = 0.78) for a PE-E/e' value of >= 11. Conclusions: E/e' does not allow further stratification in patients with exercise RP. We propose both measurement of E/e' and determination of the LV-DP (a quickly assessable variable) for the assessment of diastolic function during EE. However, when an RIP persists or develops with exercise, further assessment may not be more informative.