In vitro characterization of aortic retrograde and antegrade flow from pulsatile and non-pulsatile ventricular assist devices

Background: Many advances have been made in left ventricular assist device (LVAD) development including the introduction of smaller, non-pulsatile pumps. However, controversy exists over the potential significance of non-pulsatile blood flow. In addition, some newer LVADs incorporate descending aortic anastomosis (and therefore retrograde ascending aortic flow) for outflow rather than the traditional ascending aortic anastomosis. This, combined with non-pulsatile flow, may significantly increase the risks of ascending aortic thrombus formation, especially if native cardiac function is negligible and the aortic valve remains closed. The purpose of this study was to compare pulsatile and non-pulsatile flow generated by LVADs with outflow to the ascending aorta and descending aorta. Methods: An in vitro mock circulatory loop, driven by either a pulsatile or a non-pulsatile LVAD, was anastomosed to transparent aortic models at either the ascending or descending aortic position. The aortic valve was kept closed, modeling no native cardiac output. Normal saline was used as a blood analog. Methylene blue dye was injected into the ascending aorta and aortic arch to demonstrate flow patterns. Dye washout time (in seconds) was used as a marker of flow stagnation and potential thrombogenicity. LVAD flow, rate, after-load and coronary flow were measured. Results: Dye washout times at a flow rate of 5 liters/min were 1.7 +/- 0.75, 2.1 +/- 0.71, 4.7 +/- 0.82 and 9.9 +/- 4.4 seconds for pulsatile ascending (PA), non-pulsatile ascending (NPA), pulsatile descending (PD) and non-pulsatile descending flow (NPD), respectively. Coronary flow averaged 294 ml/min over all set-ups. Dye washout times at a flow rate of 4-liters/min were 3.0 +/- 1.0, 3.0 +/- 0.8, 14.0 +/- 3.8 and 25.0 +/- 9.1 seconds for PA, NPA, PD and NPD, respectively. Coronary flow averaged 227 ml/min over all set-ups. Ascending aortic anastomoses were associated with shorter dye washout times compared with descending aortic anastomoses, regardless of flow type (P < 0.001). There was no difference in washout time between pulsatile and non-pulsatile flow in the ascending aortic position (p = 0.23 and 0.12 for 5 and 4 liters/min, respectively). Pulsatile flow, in the descending aorta had shorter washout times than non-pulsatile flow in the descending aorta (p < 0.001 and p = 0.004 for 5 and 4 liters/min, respectively). Conclusions: LVAD descending aortic anastomosis and retrograde aortic flow is associated with increased flow stagnation in the ascending aorta. This may increase the risk for thrombus formation in patients relying solely on retrograde aortic flow, especially if cardiac function and antegrade blood flow returns.