Prognostic value of the left ventricular - left atrial volume ratio assessed using three-dimensional echocardiography with fully automated analytical software.

ABSTRACT

BACKGROUND: we investigated the prognostic value of a new 3D echocardiography (3DE) parameter, the left ventricular-left atrial volume ratio (LVLAVR) in a diverse group of subjects. METHODS: 3DE full-volume datasets were analyzed in 307 patients using fully automated LV and LA quantification software (Dynamic Heart Model, Philips Medical Systems, Andover, MA, USA), which generated LV and LA volume curves using artificial intelligence and 3D speckle tracking technology. We measured LVLAVR at LV end-diastole (edLVLAVR; LV end-diastolic volume / LA minimal volume), LVLAVR at LV end-systole (esLVLAVR LV end-systolic volume / LA maximal volume), and their differences (ΔLVLAVR edLVLAVR - esLVLAVR). No manual editing was performed on data of any patient. The primary endpoint was a major adverse cardiac event (MACE), including cardiac death, heart failure resulting in hospitalization, myocardial infarction, or ventricular tachyarrhythmia. RESULTS: feasibility of LVLAVR measurements was 90%. During a median follow-up of 21 months, 43 patients developed a primary endpoint. Univariate Cox proportional hazard analysis revealed that edLVLAVR [hazard ratio (HR) 0.72, p < 0.01] and ΔLVLAVR (HR 0.62, p < 0.01) were significantly associated with MACE. Median values of both edLVLAVR (4.59) and ΔLVLAVR (2.90) successfully stratified patients into high- and low-risk populations for future MACEs. ΔLVLAVR was still significantly associated with MACEs after adjusting for age, chronic kidney disease (CKD) and LV ejection fraction or after adjusting for age, CKD, and E/ε'. CONCLUSIONS: LVLAVR provided incremental value over traditional LV systolic and diastolic function parameters to predict future adverse outcomes. The analysis was fully automated, thereby eliminating measurement variability.