Duration of left ventricular assist device support: Effects on abnormal calcium cycling and functional recovery in the failing human heart

Abstract

Intracellular Ca(2+) handling is abnormal in human heart failure. Studies have demonstrated that left ventricular assist device (LVAD) support reverses phenotypic alterations, suggesting that, in select patients, LVAD support may be a bridge to recovery. Few studies have related support duration to phenotypic recovery. We hypothesized that reversal of impaired sarcoendoplasmic reticulum (SR) Ca(2+) cycling following LVAD implantation is duration-dependent. We used post-rest potentiation to assess SR function, and Western blot analysis to measure Ca(2+)-cycling proteins. Left ventricular tissue from 10 non-failing hearts, 8 failing hearts and 10 LVAD-supported hearts was analyzed. Support ranged from 7 to 334 days. The median duration, 115 days, divided patients into short- and long-term support groups. Post-rest potentiation (PRP) response recovered after short-term LVAD support to a level (116.8 +/- 12.1%; n = 5) close to non-failing (123.4 +/- 12.0%; n = 10) hearts, but recovery after long-term support (23.5 +/- 7.0%; n = 5) remained closer to that of failing hearts (13.5 +/- 5.6%). We found a similar pattern of normalization for SR Ca(2+)-ATPase protein and the phospholamban/SR Ca(2+)-ATPase ratio (non-failing: 0.66 +/- 0.11; failing: 1.21 +/- 0.23; short-duration LVAD: 0.68 +/- 0.14; long-duration LVAD: 1.67 +/- 0.30; correlation p < 0.001; r = 0.93). The ratio also tended to correlate with the PRP response after unloading (p = 0.05; r = -0.65). SR Ca(2+) handling improved during early LVAD support, but long-term support was associated with abnormal Ca(2+) cycling. These findings cast doubt on strategies designed to wean patients after complete unloading with an LVAD.