et al 1 presented an interesting study comparing differences in phenotypes, signaling, and gene expression of preload-and afterload-induced hypertrophy in mice in vivo. Aortocaval fistula in mice (shunt) was performed to generate volume overload with predominantly increased preload, whereas mice with transverse aortic constriction (TAC) were used to generate high afterload. Ventricular wall stress (WS) throughout the cardiac cycle (Laplace law) was similar in both mice strains. Transverse

more »

... constriction mice showed maladaptive fibrotic hypertrophy, more apoptosis, impaired left ventricular (LV) function, and higher mortality compared with shunt mice. Although the study was well performed, we believe that the findings of the article allow alternative interpretations. When analyzing representative pressure volume loops of the different mice groups by approximation in Figure 1 , the shunt model demonstrates ventricular unloading. The effective arterial elastance Ea is a measure of the total mean and pulsatile load the heart is exposed to. 2 Ea can be calculated by the ratio of the end-systolic pressure and the stroke volume 2 ; both values are calculated from the pressure volume loops in Figure 1 . The Ea of the TAC mouse was markedly higher (Eaϭ140 mm Hg/5 Lϭ28 mm Hg/L) than the Ea of the control (Eaϭ90 mm Hg/7.5 Lϭ12 mm Hg/L), whereas the Ea of the shunt mouse was definitely reduced (Eaϭ90 mm Hg/17.5 Lϭ5.1 mm Hg/ L). The marked reduction of Ea in shunt mice is plausible, because the shunt bypassed the high-resistance vessels of the periphery, thereby decreasing mean arterial load of the heart. In addition, the shunt favors an increased venous return to the heart with volume overload of the right heart especially. The heart is embedded in stiff pericardium that is not adapted to the volume overload in the shunt experiments. The stiff pericardium contributes significantly to the LV end-diastolic pressure, 3,4 indicating that it would be more precise to calculate the transmural gradient (LV end-diastolic pressure minus right atrial pressure) to determine unbiased LV preload. Considering these results, 3,4 it becomes obvious that the preload (Ϸ12 mm Hg in Figure I in the online-only Data Supplement) used for WS measurements and, thereby, diastolic WS values of shunt mice, was overestimated. Furthermore, it is not clear why pressure decay of isovolumetric relaxation (time period between T3ϭend-systole and T4ϭmid-diastole in Figure II of the online-only Data Supplement) was not included for calculations of LV diastolic WS. Mean diastolic WS was calculated only by the average value of mid-diastolic and end-diastolic WS (WS diastolic meanϭWS4ϩWS1; online-only Data Supplement). Because of the markedly increased pressures of TAC mice in early diastole, WS during isovolumetric relaxation is probably higher in TAC mice than in shunt mice. The authors investigated an important topic of myocardial disease by comparing the different effect of increased preload or afterload on myocardial function and signaling. Our objections may indicate that WS of TAC mice during the cardiac cycle is higher compared with shunt mice. These arguments are supported by the fact that brain natriuretic peptide expression was elevated only in TAC mice, but not in shunt mice. Taken together, markedly elevated Ea of TAC mice seems to stress the heart more detrimentally than volume overload in shunt mice. Therefore, these results may explain that ventricular performance and survival was significantly improved in shunt mice compared with TAC mice.