Assessment of left ventricular diastolic function using color kinesis: differentiation between normal and pseudonormalized patterns

Doppler examination of transmitral flow has been widely used to noninvasively assess left ventricular (LV) diastolic function. However, it has been demonstrated that transmitral flow velocity is dependent on LV relaxation and left atrial pressure. Increases in left atrial pressure compensate for the effects of impaired LV relaxation, frequently resulting in a "pseudonormalization" of the transmitral flow pattern. The purpose of this study was to assess whether analysis of diastolic color kinesis (CK) can be applied to differentiation between normal and pseudonormalized (PN) patterns of LV inflow. We studied 60 subjects with a ratio of early to late transmitral peak velocities (E/A) greater than 1.0 according to conventional Doppler echocardiography. All subjects simultaneously underwent measurement of the early diastolic mitral annular velocity (e'), which was measured by tissue Doppler imaging, and LV ejection fraction (EF), which was calculated by the modified Simpson method. Study subjects were classified into the following three groups according to the value of e' and EF: (1) the normal group (e' > 10 cm/s, EF > 60%), including 20 subjects (mean age 35 +/- 10 years); (2) the PN1 group (e' < 7 cm/s, EF > 50%), consisting of 20 patients [mean age 63 +/- 11 years, 15 patients with hypertensive heart disease (HHD), 5 patients with aortic valve stenosis]; and (3) the PN2 group (e' < 7 cm/s, EF < 50%), consisting of 20 patients (mean age 61 +/- 17 years, 18 patients with dilated cardiomyopathy, 2 patients with HHD). Diastolic CK images were obtained for each subject from the LV midpapillary short-axis view. Analysis of CK diastolic images was performed using ICK software. The CK-diastolic index (CK-DI) was defined as the calculated LV segmental filling fraction during the first 30% of diastole, expressed as a percentage. The mean CK-DI was determined from the average CK-DI of six LV segments. The mean CK-DI was 70.9% +/- 6.5% in the normal group, 46.3% +/- 10.4% in the PN1 group, and 36.3% +/- 5.1% in the PN2 group. The mean CK-DI was significantly reduced in the PN1 and PN2 groups compared with the normal group (P < 0.0001). Although there was no difference in e' (PN1 group: 4.6 +/- 1.8 cm/s, PN2 group: 4.4 +/- 1.7 cm/s) between the two pseudonormalized patient groups, the mean CK-DI was significantly reduced in the PN2 group compared with the PN1 group (P < 0.005). The reduction in mean CK-DI was seen not only in pseudonormalized patients with LV systolic dysfunction but also in those with preserved LV systolic function. The analysis of diastolic CK with ICK software is a useful method for detecting delayed early diastolic relaxation. We concluded that diastolic CK images may be applied to differentiating between normal and pseudonormalized patterns of LV inflow.