IntroductionDifferential hypoxia is a pivotal problem in patients with femoral veno-arterial (VA) extracorporeal membrane oxygenation (ECMO) support. Despite recognition of differential hypoxia and attempts to deliver more oxygenated blood to the upper body, the mechanism of differential hypoxia as well as prevention strategies have not been well investigated.MethodsWe used a sheep model of acute respiratory failure that was supported with femoral VA ECMO from the inferior vena cava to the femoral artery (IVC-FA), ECMO from the superior vena cava to the FA (SVC-FA), ECMO from the IVC to the carotid artery (IVC-CA) and ECMO with an additional return cannula to the internal jugular vein based on the femoral VA ECMO (FA-IJV). Angiography and blood gas analyses were performed.ResultsWith IVC-FA, blood oxygen saturation (SO2) of the IVC (83.6 ± 0.8%) was higher than that of the SVC (40.3 ± 1.0%). Oxygen-rich blood was drained back to the ECMO circuit and poorly oxygenated blood in the SVC entered the right atrium (RA). SVC-FA achieved oxygen-rich blood return from the IVC to the RA without shifting the arterial cannulation. Subsequently, SO2 of the SVC and the pulmonary artery increased (70.4 ± 1.0% and 73.4 ± 1.1%, respectively). Compared with IVC-FA, a lesser difference in venous oxygen return and attenuated differential hypoxia were observed with IVC-CA and FA-IJV.ConclusionsDifferential venous oxygen return is a key factor in the etiology of differential hypoxia in VA ECMO. With knowledge of this mechanism, we can apply better cannula configurations in clinical practice.Electronic supplementary materialThe online version of this article (doi:10.1186/s13054-015-0791-2) contains supplementary material, which is available to authorized users.

BackgroundThe addition of an intra-aortic balloon pump (IABP) during peripheral venoarterial extracorporeal membrane oxygenation (VA ECMO) support has been shown to improve coronary bypass graft flows and cardiac function in refractory cardiogenic shock after cardiac surgery. The purpose of this study was to evaluate the impact of additional IABP support on the cerebral blood flow (CBF) in patients with peripheral VA ECMO following cardiac procedures.MethodsTwelve patients (mean age 60.40 ± 9.80 years) received VA ECMO combined with IABP support for postcardiotomy cardiogenic shock after coronary artery bypass grafting. The mean CBF in the bilateral middle cerebral arteries was measured with and without IABP counterpulsation by transcranial Doppler. The patients provided their control values. The mean CBF data were divided into two groups (pulsatile pressure greater than 10 mmHg, P group; pulsatile pressure less than 10 mmHg, N group) based on whether the patients experienced cardiac stun. The mean cerebral blood flow in VA ECMO (IABP turned off) alone and VA ECMO with IABP support were compared using the paired t test.ResultsAll of the patients were successfully weaned from VA ECMO, and eight patients survived to discharge. The addition of IABP to VA ECMO did not change the mean CBF (251.47 ± 79.28 ml/min vs. 251.30 ± 79.47 ml/min, P = 0.96). The mean CBF was higher in VA ECMO alone than in VA ECMO combined with IABP support in the N group (257.68 ± 97.21 ml/min vs. 239.47 ± 95.60, P = 0.00). The addition of IABP to VA ECMO support increased the mean CBF values significantly compared with VA ECMO alone (261.68 ± 82.45 ml/min vs. 244.43 ± 45.85 ml/min, P = 0.00) in the P group.ConclusionThese results demonstrate that an IABP significantly changes the CBF during peripheral VA ECMO, depending on the antegrade blood flow by spontaneous cardiac function. The addition of an IABP to VA ECMO support decreased the CBF during cardiac stun, and it increased CBF without cardiac stun.