Pulmonary Flow as an Improved Method for Determining Cardiac Output in Mice after Myocardial Infarction

Abstract

Echocardiography is a valuable noninvasive technique to estimate cardiac output (CO) from the left ventricle (LV) not only in clinical practice but also in small-animal experiments. CO is used to grade cardiac function and is especially important when investigating cardiac injury (e.g., myocardial infarction [MI]). Critically, MI deforms the LV, invalidating the assumptions fundamental to calculating of cardiac volumes directly from the LV. Thus, the purpose of this study was to determine if Doppler-derived blood flow through the pulmonary trunk (pulmonary flow [PF]) was an improved method over conventional LV-dependent echocardiography to accurately determine CO after MI. Variations in CO were induced either by transverse aortic constriction or MI. Echocardiography was performed in healthy (n=27), transverse aortic constriction (n=25), and MI (n=41) mice. CO calculated from PF (pulsed-wave Doppler) was internally compared with CO calculated from left ventricular images using M-mode (Teichholz formula) and the single-plane ellipsoid two-dimensional (2D) formula and externally compared with the gold standard, flow probe CO. In healthy mice, all three echocardiographic methods (M-mode, 2D, and PF) correlated well with flow probe-derived CO. In MI mice, only PF CO values correlated well with flow probe values. Bland-Altman analysis confirmed that PF was improved over M-mode and 2D echocardiography. Inter- and intrauser variability of PF CO was reduced, and both inter- and intraclass correlation coefficients were improved compared with either M-mode or 2D CO calculations. PF CO calculated from pulsed-wave Doppler through the pulmonary trunk was an improved methodof estimating CO over LV-dependent formulas after MI.