Volume Displacement Pulsatile Veno-Arterial Extracorporeal Membrane Oxygenation: Preliminary Data From In Vitro Tests

Veno-arterial extracorporeal membrane oxygenation (V-A ECMO) is increasingly used to treat refractory cardiogenic shock,1 despite randomized clinical trials reporting no difference in outcomes.2,3 Although traditional ECMO pumps maintain a constant flow by sustaining a consistent impeller rotation speed, innovative designs featuring alternating pump speeds to generate a pulsatile flow pattern have emerged. With several potential benefits reported, such as improved microvascular perfusion and indirect unloading,4 newer pumps aim to improve patient outcomes. Here, we aimed to investigate a novel pressure-sensitive volume displacement pulsatile V-A ECMO device using a validated mock-circulation loop (MCL).5 Results In this experiment, the MCL connected to a V-A ECMO circuit utilizing 19 French arterial and 24 French venous cannulas (Figure 1A) and calibrated with real patient data derived from the IABP SHOCK II trial.6 The volume displacement pulsatile V-A ECMO device (VentriFlo True Pulse Pump; VentriFlo, Inc., Pelham, New Hampshire, USA), with 1:1 counter-pulsation, demonstrated surplus hemodynamic energy (SHE) generation at the post-arterial site across varying pump flow rates: 1 L/min (3,895 ergs/cm3), 1.5 L/min (1,852 ergs/cm3), and 2 L/min (2,161 ergs/cm3) (Table 1). Conversely, no SHE production was observed when utilizing a continuous flow V-A ECMO device at 1, 1.5, and 2 L/min pump flow rates. Table 1. - Surplus Hemodynamic Energy at Different Flow Rates ECMO Flow Rate (L/min) Surplus Hemodynamic Energy (ergs/cm3) at Specific Circuit Sites Pre-Oxygenator Post-Oxygenator Pre-Arterial Post-Arterial 1 169,586 126,168 59,789 3,895 1.5 245,869 182,941 95,700 1,852 2 290,476 216,272 112,865 2,161 ECMO, extracorporeal membrane oxygenation. Figure 1.: Experimental circuit and cardiac stroke volumes during ECMO support. A: The mock-circulation loop, consisting of cylinders that represent four cardiac chambers and the aortic, pulmonary aortic, systemic venous, and pulmonary venous compliance chambers was connected to the V-A ECMO circuit and primed with 40/60 admixture of glycerol-water. The Hilite 7,000 membrane oxygenator (Xenios AG) was used. After simulating a cardiogenic shock condition a baseline condition was captured using ultrasonic flow meters (TS410-10PXL, Transonic Systems, Ithaca, NY) with an accuracy of ±4% and five pressure transducers (PX181B, Omega Engineering, Norwalk, CT) with an accuracy of ±1%, placed at various points along the circuit. All sensors were then connected to a data acquisition system (DS1103, dSPACE GmbH, Paderborn, Germany) where the data for each condition was sampled at 1 kHz and recorded for 20 s after a steady state was achieved. Thereafter continuous flow and volume displacement pulsatile flow pumps were investigated as previously mentioned. B: Left ventricular stroke volume with continuous and pulsatile V-A ECMO devices at different pump flow rates. ECMO, extracorporeal membrane oxygenation; LA, left atrium; LV, left ventricle; P, pressure sensor; RA, right atrium; RV, right ventricle.When comparing continuous to pulsatile flow devices, native (mock) stroke volume increased by 6.3 ml (15%), 2.4 ml (6%), 3.6 ml (9%), and 3.6 (10%) with the volume displacement pulsatile flow V-A ECMO pump at flow rates of 1, 1.5, 2, and 2.2 L/min, respectively (Figure 1B). The relatively smaller effect at 1.5 L/min was due to inefficient counter-pulsation at this flow. Adding to the "native" stroke volumes delivered by the LV, VentriFlo provided augmentative stroke volumes ranging from 12.5 ml at 1 L/min to 27.5 ml at 2.2 L/min, synchronized at the native heart rhythm of 80 beats-per-minute and counter pulsed against the closed aortic valve. The results of this preliminary experiment demonstrate that volume displacement pump design can yield favorable effects such as SHE production and left ventricular unloading, providing combined total flow of up to 5.0 L/min at 80 BPM (LV = 2.8 L/min + VF = 2.2 L/min). The unloading effect of the pump mimics the combination of V-A ECMO and an intra-aortic balloon pump, with the benefit of lower risk of possible complications, as just one extracorporeal device is used to support the patient. The strength of this experiment is the use of real patient data from the largest trial involving patients suffering from CS in our validated MCL. Ex vivo testing with real patient data is a crucial step before animal and human trials to further corroborate the predefined hypothesis and provide guidance when designing in vivo experimentation. Volume displacement pump (VDP) design is particularly important at lower pump flow rates, as there is a concern of thrombotic complications with centrifugal pumps due to smaller size and increased blood-contacting surfaces.7 As such, these devices serve as an alternative in clinical scenarios where lower flow rates may be optimal, such as full support for pediatric and adolescent patients or when performing complex percutaneous coronary interventions. Surplus hemodynamic energy production has been associated with improved capillary recruitment8 and the function of microcirculation is independently associated with patient outcomes in CS.9 An additional benefit of VDPs is their ability to provide a discrete, augmentative stroke volume counter pulsed against the closed aortic valve while the native LV is in diastolic filling. The intermittent extracorporeal arterial delivery (and out-of-phase venous fill) allows for better native LV stroke volumes requiring less myocardial work while providing cyclical arterial pulse pressures needed to enhance recovery of end-organ function. Because these devices do not expose the venous circulation to constant suction, these devices are more fail-safe in comparison with centrifugal pumps, which require increased alertness to detect possible episodes of hypovolemia and resultant collapse of the venous drainage sites, which can lead to inferior end-organ perfusion and hemolysis.10 Although the VDP tested is only able to achieve a maximum flow rate of 2.2 L/min when the "heart failure" LV is producing 2.8 L/min, this should still allow its use in the adult population where higher flow rates are needed to maintain end-organ perfusion because the combined flow was 5.0 L/min. Future studies should explore maximal pump flow capabilities with lower native cardiac outputs within a wide range of systemic vascular resistance profiles and ECMO cannula sizes. Given the fundamental differences in the flow profiles, new and innovative strategies for use, cannulation and timing are likely warranted and should be explored. Whilst our preliminary results are encouraging, further preclinical testing in a validated model of heart failure in vivo would be needed to corroborate these results and assist with clinical translation before research and implementation in humans. Head-to-head comparisons between volume displacement, continuous flow centrifugal, and pulsatile flow centrifugal pumps would be required to delineate pros and cons between these different technologies. The authors also note that pediatric/adult patients often have higher heart rates in CS and thus we recommend future studies to evaluate SHE production and unloading at variable heart rates. Furthermore, future studies should investigate whether the relatively higher SHE yields more blood trauma or not. Conclusion In conclusion, in a preclinical experiment involving a V-A ECMO circuit connected to an MCL, SHE production and left ventricular unloading were demonstrated when using a VDP in comparison with a continuous-flow pump.