0157: Echo-based diastolic intraventricular pressure difference as a surrogate of left ventricular untwisting

Abstract

Left ventricular (LV) early diastolic filling is related to an active process of LV relaxation, caused by its rapid untwisting. The untwisting rate (UR) has been reported to be a robust marker of diastolic function but is technically hard to assess in practice. The untwisting phenomenon generates a diastolic intraventricular pressure gradient (IVPGd), which can be assessed by colour Doppler M-Mode. Previous studies showed a close temporal relation between the UR and the IVPGd peaks, but the relation between their magnitudes as well as their load dependency remain unclear. We developed our own Matlab software for IVPGd assessment by colour Doppler M-Mode based on Bernoulli equation. From our research database, we retrieved all echocardiograms where ejection fraction was normal, the colour M-Mode tracing was available and UR was measurable using speckle tracking. Passive leg raising was used in 19 healthy volunteers to study the impact of preload on the IVPGd measurement. We included 104 echocardiograms (83% males, median age 23y [19-69]). The median peak IVPGd was 3.1mmHg [1.6-7.8] and correlated to the peak UR (r=0.57, P<0.0001, figure). There was also a significant correlation between the IVGPd peak and E’ (r=0.62, P<0.0001). For peak IVPGd, correlations between inter- and intra-observer measures were calculated on 10 randomly selected subjects and were 0.83 and 0.99 (P<0.05). Compared to UR and E’, IVPGd was also significantly increased with leg lifting (2.7±0.8mmHg vs 3.3±0.8mmHg, P<0.001). Peak IVPGd based on colour Doppler M mode echocardiography is positively correlated to the LV UR. Despite the observation that IVPGd is not less load-dependent than other parameters, it requires only an echo loop easily obtained during routine clinical studies and is highly reproducible using a semi-automatic software algorithm. This new tool could easily become a new practical echo index for diastolic function assessment. Abstract 0157 - Figure