RESUMO
Background: Veno-arterial extracorporeal membrane oxygenation (V-A ECMO) is one of the most frequently used mechanical circulatory support devices. Distribution of extracorporeal membrane oxygenation flow depends (similarly as the cardiac output distribution) on regional vascular resistance. Arteriovenous fistulas (AVFs), used frequently as hemodialysis access, represent a low-resistant circuit which steals part of the systemic perfusion. We tested the hypothesis that the presence of a large Arteriovenous fistulas significantly changes organ perfusion during a partial and a full Veno-arterial extracorporeal membrane oxygenation support. Methods: The protocol was performed on domestic female pigs held under general anesthesia. Cannulas for Veno-arterial extracorporeal membrane oxygenation were inserted into femoral artery and vein. The Arteriovenous fistulas was created using another two high-diameter extracorporeal membrane oxygenation cannulas inserted in the contralateral femoral artery and vein. Catheters, flow probes, flow wires and other sensors were placed for continuous monitoring of haemodynamics and organ perfusion. A stepwise increase in extracorporeal membrane oxygenation flow was considered under beating heart and ventricular fibrillation (VF) with closed and opened Arteriovenous fistulas. Results: Opening of a large Arteriovenous fistulas (blood flow ranging from 1.1 to 2.2 L/min) resulted in decrease of effective systemic blood flow by 17%-30% (p < 0.01 for all steps). This led to a significant decrease of carotid artery flow (ranging from 13% to 25% after Arteriovenous fistulas opening) following VF and under partial extracorporeal membrane oxygenation support. Cerebral tissue oxygenation measured by near infrared spectroscopy also decreased significantly in all steps. These changes occurred even with maintained perfusion pressure. Changes in coronary artery flow were driven by changes in the native cardiac output. Conclusion: A large arteriovenous fistula can completely counteract Veno-arterial extracorporeal membrane oxygenation support unless maximal extracorporeal membrane oxygenation flow is applied. Cerebral blood flow and oxygenation are mainly compromised by the effect of the Arteriovenous fistulas. These effects could influence brain function in patients with Arteriovenous fistulas on Veno-arterial extracorporeal membrane oxygenation.
RESUMO
Background: A large arteriovenous fistula (AVF) is a low-resistant circuit that affects organ perfusion and systemic hemodynamics even in standard conditions. The extent of its' effect in critical states has not been elucidated yet. We used norepinephrine to create systemic vasoconstriction, dobutamine to create high cardiac output, and rapid right ventricle pacing as a model of acute heart failure in a porcine model of high-flow AVF circulation. Methods: The protocol was performed on nine domestic female pigs under general anesthesia. AVF was created by connecting two high-diameter ECMO cannulas inserted in the femoral artery and vein. Continuous hemodynamic monitoring was performed throughout the protocol. Three interventions were performed-moderate dose of norepinephrine (0.25 ug/kg/min), moderate dose of dobutamine (10 ug/kg/min) and rapid right ventricle pacing to simulate low cardiac output state with mean arterial pressure under 60 mmHg. Measurements were taken with opened and closed arteriovenous fistula. Results: Continuous infusion of norepinephrine with opened AVF significantly increased mean arterial pressure (+20%) and total cardiac output (CO) (+36%), but vascular resistance remained virtually unchanged. AVF flow (Qa) rise correlated with mean arterial pressure increase (+20%; R = 0.97, p = 0.0001). Effective cardiac output increased, leading to insignificant improvement in organ perfusion. Dobutamine substantially increased cardiac output with insignificant effect on AVF flow and mean arterial pressure. Carotid artery blood flow increased significantly after dobutamine infusion by approximately 30%, coronary flow velocity increased significantly only in closed AVF state. The effective cardiac output using the heart failure model leading to decrease of carotid artery flow and worsening of brain and peripheral tissue oximetry. AVF blood flow also dropped significantly and proportionally to pressure, but Qa/CO ratio did not change. Therefore, the effective cardiac output decreased. Conclusion: In abovementioned extreme hemodynamic conditions the AVF flow was always directly proportional to systemic perfusion pressure. The ratio of shunt flow to cardiac output depended on systemic vascular resistance. These experiments highlight the detrimental role of a large AVF in these critical conditions' models.
