Background: Left ventricular (LV) assist device (LVAD) support has been associated with right ventricular (RV) failure in humans, but the etiology remains unknown. Mechanical LV support apparently does not induce RV pump failure in normal hearts, but controlled studies of LV assistance in hearts with preexistent RV dysfunction have been limited. Therefore, this study was performed to determine if LVAD support impairs RV systolic mechanics during acute RV ischemia.

Methods and results: Five closed-chest, autonomically blocked, sedated dogs were studied after placement of an LVAD (LV-femoral artery bypass), right coronary artery (RCA) occluder, and 27 miniature radiopaque tantalum markers into the LV and RV walls for independent computation of RV and LV volumes. Biplane videofluoroscopic marker images and hemodynamic data were recorded before RCA occlusion with the LVAD off (maximum LV pressure [LVPmax] = 119 +/- 25 mm Hg), after 3 minutes of RCA occlusion with the LVAD off (LVPmax = 84 +/- 18 mm Hg), and then with the LVAD on (LVPmax = 26 +/- 32 mm Hg). Global RV contractility (end-systolic elastance [RV Ees] and preload recruitable stroke work [RV PRSW]), RV power output, and the mechanical (pump) efficiency of converting potential energy to external work (ratio of RV stroke work/total pressure-volume area) were calculated. As expected, with RCA occlusion there were major decreases in RV Ees (from 2.5 +/- 1.2 to 1.4 +/- 0.5 mm Hg/mL, P < .005) and RV PRSW (15 +/- 4 versus 9 +/- 4 mm Hg, P < .001). RV power output (overall pump performance) declined by 39 +/- 20% (P < .025), and mechanical efficiency fell by 38 +/- 13% (P < .001). After initiation of mechanical LV support, however, there was no further impairment of RV contractility or power output (P > .80). Pulmonary artery input impedance (RV afterload) decreased from 848 +/- 628 to 673 +/- 577 dyne.sec-1.cm-5 (P < .01), which led to a 26 +/- 29% improvement in RV pump efficiency (P < .001).

Conclusions: While right coronary artery occlusion significantly reduced RV systolic performance, LVAD support during acute RV ischemia did not further impair RV contractility or power output. Furthermore, since RV afterload fell with LV unloading, the mechanical pump efficiency of the right ventricle actually improved. These observations demonstrate that LVAD support does not directly induce RV failure in canine hearts with acute isolated RV ischemia.