Comparison of biplane and single plane left ventricular volumes in atrial septal defect

Abstract

The close agreement between biplane (BP) and single-plane (SP) angiographic estimates of left ventricular (LV) volumes results from the slmilarity of the minor axes measured in the right anterior oblique (RAO) and left anterior oblique (LAO) views. Disease states that alter LV geometry may change the length of one minor axis more than the other, producing a discrepancy between BP and SP volumes. To examine this hyposthesls, angiographically derlved volumes In 21 patients with atrial septal defects (ASD) in which the LV appears to be compressed and flattened by an enlarged right ventricle, were compared to 100 normal control patients. In the control patients, the median SPestlmate of end-diastolic volume(EDV) was7.6% larger than the BP determination, whereas in patients with ASD, the median SP EDV estimate was 16.7% larger than the BP EDV (P<O.OOOl). The SP end-systolic volume (ESV) underestimated the BP value by 3.4% in controls but overestimated the EP ESV in patients with ASD by 4.3% (Pc0.02). The overestimate of the SP EDV and SP ESV when compared to the BP volumes may be due to changes in either the minor axes or the appearance of the longest major axis in the LAOview. The longest major axis was found in the RAOview in 99% (99/100) of normals and 95% (20/21) of ASD patients (P=NS). The median ratio of RAO to LAO end-diastolic minor axes, however, was 1.07 in the normals and 1.17 for ASD patients. The median ratio of end-systolic minor axes was 0.97 for controls and 1.04 for ASD patients.Compression of the LV in patients with ASD shortens the LAO minor axis, resulting in a significantly greater SP overestimation of LV volume than occurs In normals. The degree of SP volume overestimate was not predicted by the magnitude of the left-tce right shunt or pulmonary pressure. This source of error affects all SP methods for determining left ventricular volume, including radionuclide techniques using static images.