#6391 MARKERS OF SUBCLINICAL LEFT VENTRICULAR DYSFUNCTION ARE ASSOCIATED WITH MORTALITY IN CHRONIC KIDNEY DISEASE PATIENTS WITH PRESERVED EJECTION FRACTION

Abstract Background and Aims Cardiac dysfunction is frequent in chronic kidney disease (CKD) patients and contributes to high morbidity and mortality in this category. The aim of our study is to evaluate the association between markers of subclinical left ventricular (LV) dysfunction and mortality in CKD patients with preserved left ventricular ejection fraction. Method We prospectively enrolled CKD patients (pts) in pre-dialysis. Parameters of cardiac structure and function were evaluated by echocardiography: systolic and diastolic left ventricular (LV) volumes, LV mass index (LVMI), left atrial volume index (LAVI), LV ejection fraction (LVEF), global longitudinal strain (GLS), ratio of mitral velocity to early diastolic velocity of the mitral annulus (E/E`) and ratio between the early maximal ventricular filling velocity and the late filling velocity (E/A). Cardiac functional status was stratified according to NYHA criteria. Al patients were followed until death or loss from records. Primary endpoint was all-cause mortality. Secondary endpoints were cardiovascular mortality and a composite cardiac endpoint comprising coronary heart disease, heart failure hospitalization and new-onset arrhythmia. Results We prospectively enrolled 39 CKD patients (5 pts CKD stage 2, 27 pts CKD stage 3, 7 pts CKD stage 4) mean age 64.3 ± 10.2 years, men = 28 (71.8%). At baseline all pts had LVEF ≥ 50%; 11 pts (28.2%) were classified as NYHA 1, 18 pts (46.2%) were NYHA 2 and 10 pts (25.6%) were NYHA 3. Mean follow up period was 98.1 ± 43.2 months (range 6 - 144 months). GLS, as a marker of subclinical LV systolic dysfunction, was significantly less negative in males (-16.6 ± 2.5 vs −18.4 ± 3.2, p< 0.05), in pts with symptomatic heart failure (−16.5 ± 2.6 vs -18.9 ± 2.8, p< 0.05), and in pts with diabetes mellitus (−15.8 ± 1.9) and hypertensive nephropathy (−16.8 ± 2.8) compared with patients with tubulointerstitial diseases (−18.1 ± 2.5) and chronic glomerulonephritis (vs −21.1 ± 3.6) (p< 0.05). E/E`, as a marker of subclinical LV diastolic dysfunction, was significantly higher in pts with symptomatic heart failure (12.2± 3.4 vs 9.5 ± 1.9, p< 0.05), in pts with diabetes mellitus (13.3± 3.5 vs 10.2 ± 2.5, p< 0.05), All-cause mortality was 17.9% (7 events), while cardiovascular mortality was 10.3% (4 events). Composite cardiac endpoint was recorded in 13 pts (33.3%). Baseline profile in patients who died during follow up period included significantly lower eGFR (30.1 ± 10.9 vs 46.1 ± 16.9 ml/min/1.73 m2, p<0.05), higher LAVI (55.5 ± 11.1 vs 37.4 ± 11.9 m/m2, p<0.05), higher LVMI (149.8±36.1 vs 115±28.5 g/m2, p < 0.05), and higher E/E` (15.2±3.3 vs 10.7 ± 2.8, p < 0.05). Death was significantly correlated with GLS (p = 0.042), E/E` (p = 0.002), LVMI (p = 0.008) and LAVI (p = 0.007), while CV mortality was correlated with E/E` (p = 0.019) and LVMI (p = 0.041). Composite cardiac endpoint was significantly correlated with LVMI (p = 0.01) and LAVI (p = 0.006). Cox regression analysis identified only E/E` as a predictor for all-cause mortality (p = 0.002) and also for cardiovascular mortality (p = 0.015). Conclusion In our study population of CKD patients, cardiac imaging parameters of subclinical LV dysfunction (GLS and E/E`) were associated with all-cause mortality. Marker of LV diastolic dysfunction E/E` proved to be the only predictor of all-cause and cardiovascular mortality. Thus, in CKD patients, evaluation of the cardiac function should focus not only on ejection fraction measurement, but also on the parameters of subclinical LV dysfunction.