Comparison of three-dimensional proximal isovelocity surface area to cardiac magnetic resonance imaging for quantifying mitral regurgitation

The aim of our study was to evaluate 3-dimensional (3D) color Doppler proximal isovelocity surface area (PISA) as a tool for quantitative assessment of mitral regurgitation (MR) against in vitro and in vivo reference methods. A customized 3D PISA software was validated in vitro against a flowmeter MR phantom. Sixty consecutive patients, with ≥mild MR of any cause, were recruited and the regurgitant volume (RVol) was measured by 2D PISA, 3D peak PISA, and 3D integrated PISA, using transthoracic (TTE) and transesophageal echocardiography (TEE). Cardiac magnetic resonance imaging (CMR) was used as reference method. Flowmeter RVol was associated with 3D integrated PISA as follows: y = 0.64x + 4.7, r(2) = 0.97, p <0.0001 for TEE and y = 0.88x + 4.07, r(2) = 0.96, p <0.0001 for TTE. The bias and limit of agreement in the Bland-Altman analysis were 6.8 ml [-3.5 to 17.1] for TEE and -0.059 ml [-6.2 to 6.1] for TTE. In vivo, TEE-derived 3D integrated PISA was the most accurate method for MR quantification compared to CMR: r(2) = 0.76, y = 0.95x - 3.95, p <0.0001; 5.1 ml (-14.7 to 26.5). It was superior to TEE 3D peak PISA (r(2) = 0.67, y = 1.00x + 6.20, p <0.0001; -6.3 ml [-33.4 to 21.0]), TEE 2D PISA (r(2) = 0.54, y = 0.76x + 0.18, p <0.0001; 8.4 ml [-20.4 to 37.2]), and TTE-derived measurements. It was also most accurate by receiver operating characteristic analysis (area under the curve 0.99) for the detection of severe MR, RVol cutoff = 48 ml, sensibility 100%, and specificity 96%. RVol and the cutoff to define severe MR were underestimated using the most accurate method. In conclusion, quantitative 3D color Doppler echocardiography of the PISA permits a more accurate MR assessment than conventional techniques and, consequently, should enable an optimized management of patients suffering from MR.