[Left ventricular early diastolic filling and atrial contribution assessed by ECG-gated cardiac blood pool scintigraphy].

This study evaluated early diastolic left ventricular (LV) filling and the atrial contribution to ventricular filling in patients (pts) with various heart diseases using ECG-gated cardiac blood pool scintigraphy. Conventional equilibrium list mode ECG-gated cardiac blood pool scintigraphy was performed for 19 normal subjects (N) as controls, 104 pts with old myocardial infarction (OMI), 19 pts with essential hypertension (HT), seven pts with idiopathic hypertrophic subaortic stenosis (IHSS), three pts with non-obstructive hypertrophic cardiomyopathy (HCM), 19 pts with pure mitral stenosis (MS) and one pt with both MS and aortic regurgitation to evaluate early diastolic LV filling. The LV stroke counts corresponding to stroke volume and the early diastolic LV peak filling rate (DdV/dt) were obtained from the LV time-activity curve and its first derivative. Then the DdV/dt was normalized by stroke counts. The DdV/dt was significantly lower in pts with OMI (4.34 +/- 1.02/sec, p less than 0.001), HT (3.93 +/- 0.70/sec, p less than 0.001), IHSS (4.23 +/- 1.59/sec, p less than 0.01) and MS (4.56 +/- 1.05/sec, p less than 0.01) than in N (5.93 +/- 1.26/sec). Then, in OMI, the DdV/dt correlated significantly (r = -0.45, p less than 0.05) with infarct size (% abnormal contracting segment = %ACS) obtained by contrast left ventriculography. Furthermore, in pts with HT, the DdV/dt correlated significantly (r = -0.59, p less than 0.02) with the left ventricular mean wall thickness obtained by M-mode echocardiography. In pts with MS, the DdV/dt also correlated significantly (r = 0.73, p less than 0.001) with the mitral orifice area obtained by two-dimensional echocardiography. However, it has been difficult to assess the atrial contribution to ventricular filling by conventional ECG-gated cardiac blood pool scintigraphy, because the LV time-activity curve in the late diastolic phase was distorted and unreliable, whenever a minimal variation of the R-R interval occurred. Therefore, to produce a more reliable late diastolic LV volume curve, a "two-beat LV volume curve" was constructed using a new method; namely, each cardiac cycle was divided into 20 msec segments in two different ways, i.e., backward and forward of the R wave, and the backward LV volume curve and forward LV volume curve (conventional method) were connected at the R wave. Then, to estimate the atrial contribution, an increment of counts after the beginning of the P wave divided by counts corresponding to the stroke volume (A(P)/SV) was calculated.(ABSTRACT TRUNCATED AT 400 WORDS)