Left Ventricular Function in Patients with Centrifugal Left Ventricular Assist Device

We used blood pool radionuclide angiography to study the left ventricular (LV) ejection fraction (EF) in 17 patients with a centrifugal assist device (AD) placed because of severe postoperative LV dysfunction. During maximal LVAD flow, the 12 patients who could be weaned had a higher LV-EF than the 5 who could not be weaned (18 ± 12% vs. 9 ± 2%, p = 0.04). Sequential studies during variable AD flows in 12 patients revealed an increase in LV-EF from 15 ± 7% at maximal flows to 33 ± 8% during minimal flows (p < 0.005) in the 10 patients who could be weaned and no change in LV-EF in 2 patients who could not be weaned. The LV-EF during maximal LVAD flow rates was similar in the 10 patients with long-term survival (192 ± 129 days) and in the 7 patients with only short-term (9 ± 6 days) survival (LV-EF 17 ± 12% vs. 12 ± 6%, p = ns). The long-term survivors, however, had a substantial increase in LV-EF from 20 ± 13% to 34 ± 9% (p < 0.01), as the LVAD flow was decreased from maximal to minimal, whereas the short-term survivors had an insignificant increase in LV-EF from 12 ± 7% to 21 ± 12% (p = ns). The long-term survivors increased the LV-EF from maximal to minimal LVAD flows by 182%, in contrast with the short-term survivors, who increased the LV-EF by only 44%. Furthermore, during this intervention, seven of eight long-term survivors were able to increase the LV-EF by greater than 100%, in contrast to none of the short-term survivors. Thus, the LV-EF response to manipulations in the AD flow is a major element determining the feasibility of weaning a patient from the LVAD and of long-term survival.