Correlation of angiographic and autopsy left ventricular dimensions in children

E A Fisher, F A Eckner, I W DuBrow, A R Hastreiter
1978 Circulation  
To evaluate controlled pressure coronary artery perfusion fixation of hearts as a means of obtaining meaningful cardiac measurements at autopsy, left ventricular (LV) autopsy measurements were correlated with in vivo end-diastolic (D) and end-systolic (S) angiographic (angio) values from biplane cineangiograms in the same patients. Mitral (MV) and aortic valve (AV) circumferences, LV equatorial circumference (Circ), spatial inflow and outflow lengths, wall thickness (W) and LV volume (Vol) were
more » ... measured in 34 children with congenital heart disease. All dimensions showed significant correlation of angiographic and autopsy data from which linear regression equations were derived. The r values obtained,were: MV 0.74; AV 0.85; D Circ 0.83; S Circ 0.82; D inflow 0.92; S inflow 0.90; D and S outflow 0.96; W 0.78; D Vol 0.92; S Vol 0.86. Mean angio and autopsy values were not significantly different for AV, S inflow, and S Vol. Autopsy values were lower than all D angio values and MV (P < 0.001 for all), and higher than angio W and S outflow (P < 0.001 for both). The data show that changes of LV morphology with this fixation method are predictable, making estimation of in vivo values from autopsy measurements possible using the derived linear regression equations. ANGIOGRAPHIC TECHNIQUES for determining in vivo cardiac chamber volumes and dimensions have contributed greatly to the understanding of the pathophysiology of heart disease. Recognizing the limitations and variability of cardiac measurements made at autopsy, Glagov et al.' developed a method of controlled pressure fixation which provided a standardized autopsy preparation. A modification of this method was shown to yield reproducible measurements with little variation in hearts from similar-sized pigs.2 We have compared measurements of the left ventricle from human hearts fixed by this method with in vivo angiographic measurements in the same patients in an attempt to further validate this fixation technique in a clinical setting. Methods Study Group Thirty-four children, primarily with congenital heart disease, who underwent diagnostic cardiac catheterization and cineangiography and subsequently came to autopsy at the University of Illinois Hospital, Chicago, Illinois, comprised the study group. The types of heart defects included are shown in table 1. Age of patients was 900 ± 222 days or 2.5 ± 0.61 years (mean ± SEM), range 04,289 days (11.8 years) at angiography and 960 ± 225 days or 2.6 + 0.62 years at autopsy. The interval between angiography and autopsy ranged from 0-300 days (mean ± SD, 60 ± 87 days). Body surface area (BSA), averaged 0.40 ± 0.04 m2 (mean ± SEM) at angiography and 0.41 ± Circulation 58, No. 4, 1978. 0.04 at autopsy. Table 2 shows the average intervals and changes in body surface areas in various age ranges. Twenty-nine patients underwent cardiac surgical procedures, 16 open and 13 closed, in that interval. Twenty-five died in the immediate postoperative period. Angiographic Methods Biplane antero-posterior (AP) and lateral (L) cineangiograms were filmed using a Philips biplane cineangiography system. Cines were recorded at 64 or 80 frames/sec after high speed injection of 1-2 ml/kg of contrast medium (Conray 400R) into the inferior vena cava or right atrium. In a few patients it was necessary to use selective left ventricular injections in order to visualize ventricular outlines adequately. Cardiac cycles with arrhythmias were excluded. The earliest cycle with sufficient contrast to allow accurate drawing of left ventricular outlines was used. Correction of all measurements for linear x-ray magnification was made utilizing a grid system. Plastic grids etched with 1 x I cm lines filled with lead powder in a glue base were radiographed at midchest level in AP and L planes. The radiographic area of a known number of squares roughly overlying the cardiac image was determined by planimetry and divided by the actual area as determined by the known number of squares to yield a separate magnification factor (MF) for each plane. Projected ventricular areas were corrected by dividing by this factor. Projected linear measurements were divided by the square root of the appropriate MF. Various left ventricular dimensions corresponding to dimensions at autopsy were measured from the drawings. Two angiographic (angio) measurementsend-diastolic (D) and end-systolic (S) -were compared with a single autopsy measurement of each dimension, with the exceptions of valve circumferences and wall thickness, where only D angiographic measurements were made. All 739 by guest on July 22, 2018 Downloaded from
doi:10.1161/01.cir.58.4.739 pmid:688583 fatcat:52z73r3f7fc4lnzwclsmnlv3wq