Sustained cardiac diastolic changes elicited by ultrafiltration in patients with moderate congestive heart failure: pathophysiological correlates.

OBJECTIVE To investigate the pathophysiological (cardiac function and physical performance) significance of clinically silent interstitial lung water accumulation in patients with moderate heart failure; to use isolated ultrafiltration as a means of extravascular fluid reabsorption. DESIGN Echocardiographic, Doppler, chest x-ray evaluations, and cardiopulmonary tests at baseline, soon after ultrafiltration (veno venous extracorporeal circuit), and four days, one month, and three months later. SETTING University institute of cardiology. SUBJECTS 24 patients with heart failure due to idiopathic dilated cardiomyopathy or ischaemic myocardial disease with sinus rhythm and ejection fraction less than 35%. Twelve were randomised to ultrafiltration and 12 were taken as controls. MAIN OUTCOME MEASURES Left ventricular systolic function (from ultrasonography); Doppler evaluation of mitral, tricuspid, and aortic flow and echo-Doppler determination of cardiac output; radiological score of extravascular lung water; right and left ventricular filling pressures; oxygen consumption at peak exercise and exercise tolerance time in cardiopulmonary tests. RESULTS Soon after ultrafiltration (1976 (760) ml of fluid removed) the following was observed: a reduction in radiological score of extravascular lung water (from 15(1) to 9(1)) and of right (from 7.1 (2.3) to 2.3 (1.7) mm Hg) and left (from 17.6 (8.8) to 9.5 (6.4) mm Hg) ventricular filling pressures; an increase in oxygen consumption at peak exercise (from 15.8 (3.3) to 17.6 (2) ml/min/kg) and of tolerance time (from 444 (138) to 508 (134) s); a slight decrease in atrial and ventricular dimensions; no changes in the systolic function of the left ventricle; a reduction of the early to late filling ratio in both ventricles (mitral valve from 2 (2) to 1.1 (1.1)); (tricuspid valve from 1.3 (1.3) to 0.69 (0.18)) and an increase in the deceleration time of mitral and tricuspid flow, reflecting a redistribution of filling to late diastole. Variations in the ventricular filling pattern, lung water content, and functional performance persisted for three months in all cases. None of these changes was detected in the control group. CONCLUSIONS Reduction of interstitial lung water was probably the mechanism whereby ultrafiltration modified the pattern of filling of the two ventricles and improved functional performance.