Quantification of mitral regurgitation by the automated cardiac output method: an in vitro and in vivo study

Recently, the automated cardiac output method (ACM) was introduced for the calculation of blood flow at the left ventricular outflow tract (LVOT). This study was performed to examine the possibility of using ACM for flow calculation at the level of the mitral valve and for the quantification of mitral regurgitation (MR) in vitro and in vivo.

In a computer-controlled in vitro model of the human heart, aortic and mitral normal bioprosthetic valves were inserted. ACM and electromagnetic probe flow measurements correlated well at the LVOT and at the mitral level (r2 = 0.79 and 0.77, respectively). For stroke volumes ranging from 30 to 100 ml/beat, there was no statistically significant bias between ACM and electromagnetic flow probe (-1.5 and 1.3 ml for LVOT and mitral level, respectively). Limits of agreement were [-14; +11] ml and [-18; +16] ml, respectively. We evaluated 68 patients in our in vivo study. They were divided into three groups according to the results of "standard" echocardiographic Doppler methods for the semiquantification of MR: echocardiographic color Doppler cartography, intensity of the continuous wave Doppler spectra, and in some patients, pulmonary venous flow, conventional Doppler, and proximal isovelocity surface area quantitative data. Group 1 consisted of 35 patients without MR or a physiologic one; the 17 patients in group 2 had a mild MR (1-2/4) and in group 3, 16 patients with MR 3-4/4 were included. Regurgitant volume (RV) was calculated as the difference between ACM mitral flow and ACM aortic flow, and regurgitant fraction (RF) was defined as the ratio between RV and ACM mitral flow. When mitral flow was measured only from the four-chamber view, we found in group 1, RV = -0.57 (0.67) L/min and RF = -16% (19%); in group 2, RV = -0.31 (1.06) L/min and RF = -8% (19%); and in group 3, RV = 1.53 (0.94) L/min and RF = 23% (13%). RV and RF were statistically higher in group 3 compared with group 2 or group 1 (p < 0.0005), but no significant difference was found between groups 1 and 2. When mitral flow was measured by the mean value of ACM four-chamber and two-chamber views, this resulted in group 1, RV = -0.26 (0.63) L/min and RF = -8% (15%); in group 2, RV = 0.01 (1.04) L/min and RF = -2% (18%); and in group 3, RV = 2.07 (1.21) L/min and RF = 34% (19%). RV and RF were again significantly higher in group 3 (p < 0.0001). There was no significant difference between group 1 and group 2, but in group 1 RF was no longer statistically different from 0%.

(1) In our in vitro setting, ACM is reliable both at the LVOT and at the mitral valve. (2) In the in vivo situation, some overlapping does exist between the three groups of MR. However, ACM is a very easy, rapid, and objective method to differentiate hemodynamic nonsignificant (<3/4) from significant (> or =3/4) MR. Together with other well-known methods for the quantification of MR, it should facilitate the gradation of MR in the clinical setting. The absence of significant differences between group 1 and group 2 proves that the accuracy of ACM measurements at the mitral valve needs to be ameliorated before ACM can be used as a gold standard for the noninvasive measurement of RV and RF.