Dr. McAuliffe uses our data to invoke the Emax method for calculating myocardial contractility. Emax, a load-independent measure of left ventricular contractility, is defined as Pes/(Ves − Vd), where Vd is the volume-axis intercept of a linear regression line of multiple end-systolic ventricular pressure–volume data in a given contractile state, Pes is left ventricular (LV) pressure at end systole, and Ves is LV volume at the same time. 1Emax is determined invasively by measuring Pes with a catheter in the left ventricle. Noninvasive estimates of Emax have been described, 2,3but to our knowledge, none have been validated by comparison with standard, accepted pediatric echocardiographic methods to quantitate myocardial contractility, such as the Simpson biplane method for ejection fraction (EF), 4load-independent methods to assess contractility (stress-velocity and stress-shortening indices), 5or angiographic methods in pediatric patients with normal hearts 4or with congenital heart disease. In fact, the reliability of Vd and thus Emax as indexes of load independent contractility has been questioned. 6
The Simpson biplane method has been shown to correlate well with angiographic EF calculations in pediatric patients with congenital heart disease. 4Using this methodology, we found that sevoflurane changed contractility only at 1.5 minimum alveolar concentration (MAC), resulting in a 15% decrease in EF from baseline. EF at 1 MAC sevoflurane and 1 and 1.5 MAC isoflurane did not change from baseline, nor did the shortening fraction (an M-mode measure of contractility) at any concentration of either agent. Our conclusion that isoflurane and sevoflurane maintain systemic cardiac index (CI) with little change in contractility is based on this data (table 4). Dr. McAuliffe offers neither statistical analysis nor similar calculations for halothane.
Although Dr. McAuliffe's calculations using Emax show that LV ejection into the aorta may decrease, his analysis does not provide consideration of the effect of intracardiac shunting. Twenty-one of the patients in our study had left-to-right intracardiac shunting at the ventricular level, with some of the LV stroke volume ejected into the right ventricle, not the aorta. Ten patients had right-to-left intracardiac shunting at the ventricular level, resulting in a net reduction in the total amount of blood ejected from the left ventricle during systole. This presumably is the reason that left ventricular end-diastolic volume (LVEDVI) × heart rate (HR) × EF from our data equals only 70% of the calculated CI in our study. In our group of patients as a whole, a significant portion of the LV stroke volume is not ejected into the aorta. This is why we chose to calculate CI by measuring stroke volume (SV) into the proximal aorta by the pulse-wave, Doppler-derived, velocity–time integral (VTI) method: CI = SV × HR, where SV = VTI × cross-sectional area of aorta (see Web Enhancement, www.anesthesiology.org, February 2001, for all formulae for hemodynamic calculations). This method measures LV ejection into the aorta only and has been shown to have a strong correlation (r = 0.98) to Fick CI values during cardiac catheterization 7in children.
Although Dr. McAuliffe is correct in contending that contractility in normal hearts must decrease if SVRI decreases and CI and preload (LVEDVI) are unchanged, this again may not be the case for patients with intracardiac shunting, in whom under different conditions varying proportions of the LV stroke volume may be ejected into the aorta, allowing contractility to be preserved with lower SVRI.
We appreciate Dr. McAuliffe's perspective but maintain that our methods of calculating contractility and CI with well-validated echocardiographic methods give an accurate idea of the effects of anesthetics on myocardial contractility and hemodynamics in patients with congenital heart disease.