Mechanical assist in cardiac arrest: Optimising circulatory support. Experimental studies.

Abstract

Introduction: Mechanical circulatory support (MCS) may be useful in cardiac arrest (CA), both in- and out- of hospital. However, efficacy and survival benefit has been difficult to evaluate compared to standard cardiopulmonary resuscitation. In three experimental studies we aimed to assess different modes of MCS during CA in providing adequate organ perfusion and systemic circulation and identify predictors of sustainable post-CA heart function. Different theoretical assumptions were the background for analysis in the three study protocols performed as acute experiments in anaesthetized pigs: Paper I: A major limitation to the effectiveness of a LVAD alone during CA is the lack of left ventricular (LV) filling due to minimal pulmonary circulation. We therefore wanted to assess if the combination of a left- and right ventricular assist device (BIVAD/BiPella) was beneficial as circulatory support versus a LVAD alone. Paper II: ECMO has the potential to replace systemic circulation during CA. However, concerns have been voiced regarding retrograde flow-delivery and effect on the myocardium during circulatory collapse. Based on results from Paper I we optimized BiPella support aiming to improve and maintain acceptable coronary perfusion pressure, believing this could potentially rectify the poor outcome of BIVAD/BiPella in Paper I if successful. Thus, in Paper II we compared the efficacy of balanced biventricular circulatory assist with extracorporeal membrane oxygenation (ECMO). Paper III: Pressure build-up in the left ventricle during cardiac arrest may be detrimental during extracorporeal cardiopulmonary resuscitation (ECPR) as indicated in Paper II. Therefore, we wished to investigate if unloading (venting) the left ventricle using add-on LVAD could be of benefit. However, the ideal flow-contributions of each assist device when combining LVAD and ECMO during ECPR in is not known. We therefore wanted to compare ECMO with standard or reduced flow and add-on LVAD versus ECMO alone. Finally, we wished to assess the contribution of add-on LVAD regarding pulmonary flow.

Materials and methods: The animal experiments were performed at the Vivarium, University of Bergen, and protocols were approved by the Norwegian Animal Research Authority or by the Norwegian Food Safety Authority. Paper I and II were performed with percutaneous techniques. The final experiment was an open chest model. All protocols followed a similar timeline: 1. Anaesthesia and instrumentation of the pig. 2. Baseline evaluation. 3. Induction of CA by application of a 9V DC battery to the myocardium. 4. Immediate initiation of mechanical circulatory support (MCS). 5. Three attempts of cardioversion at the end of the CA period. 6. If successful return of spontaneous circulation (ROSC) was achieved, unsupported observation (Paper II and Paper III). Comparisons between intervention groups: 1. Haemodynamics (during and after CA). 2. Organ tissue blood flow rate (organ perfusion) and device output as calculated from fluorescent microspheres. 3. Arterial blood gases and biomarkers. 4. ROSC. 5. Sustained cardiac function post-ROSC (Paper II and Paper III). In Paper I, twenty animals were randomized in two groups receiving circulatory support either by the Impella CP alone (LVAD) or in combination with the Impella RP (BIVAD/BiPella) during 30 minutes of CA. In Paper II, twenty pigs were randomized to receive MCS either by BiPella or by extracorporeal membrane oxygenation (ECMO) during 40 minutes of CA. If ROSC was successful, animals were observed for 60 minutes unsupported. In Paper III, twenty-four animals were randomized in three groups. Extracorporeal cardiopulmonary resuscitation (ECPR) in Group 1 was provided by ECMO with standard-flow and add-on Impella CP. In Group 2: ECMO with reduced flow combined with Impella CP. In Group 3, animals were supported by standard-flow ECMO alone. ECPR lasted for 60 minutes. If ROSC was successful, 180 minutes unsupported observation followed.

Results: Paper I demonstrated that BIVAD/BiPella provides superior circulatory support and perfusion for peripheral organs (including the brain) related to higher LVAD output and increased central aortic pressure compared to LVAD alone. However, myocardial perfusion was related to the pressure difference between mean aortic pressure and mean left ventricular pressure during cardiac arrest. Myocardial perfusion was inferior with BiPella resulting in significantly fewer ROSC (5/10 vs 10/10, p = 0.033) despite significantly higher etCO2 (p = 0.029). Paper II showed that balancing RVAD and LVAD to ensure acceptable coronary perfusion pressure and concomitant LVAD output was feasible, also sustaining vital organ perfusion. However, ECMO provided a more optimal systemic circulatory support. Device output and mean aortic pressure were increased with subsequent improved peripheral tissue perfusion reflected by reduction of s-lactate. In animals where sufficient myocardial perfusion pressure (mean aortic pressure – mean LV pressure > 10-15 mmHg) could not be achieved, perfusion (ml/min/g) was reduced in the subendo- and midmyocardium, averaging 0.59 ± 0.05 vs. 0.31 ± 0.07, (p = 0.005) and 0.91 ± 0.06 vs 0.65 ± 0.15 (p = 0.085), but not in the subepicardium (1.02 ± 0.07 vs 0.86 ± 0.17, p = 0.30) irrespective of group. These subjects also had inferior post-ROSC cardiac function. Paper III showed that add-on LVAD improved haemodynamics compared with ECMO alone during refractory CA. Add-on LVAD could not substitute a reduced ECMO-flow. Three animals with reduced ECMO flow and adjunctive Impella support did not achieve ROSC. With ECMO alone, ROSC was obtained in all animals. However, 4/8 died post-ROSC due to development of cardiogenic shock. In the remaining 21 animals, 17 animals had sustained cardiac function at study termination 3 h after ROSC. Animals without sustained cardiac function (7/24) had reduced mAP (p < 0.001), CPP (p = 0.002) and mPAf (p = 0.004) during CA and ECPR.

Conclusions: Paper I: Biventricular support during cardiac arrest was associated with high intraventricular pressure in the left ventricle resulting in decreased myocardial perfusion pressure, reduced myocardial tissue blood flow rate and subsequent reduction in ROSC. Paper II: Myocardial perfusion and sustained cardiac function were related to myocardial perfusion pressure during VF irrespective of MCS (ECMO and balanced biventricular support). Balanced biventricular support maintained lower intraventricular pressure compared to ECMO. Paper III: Add-on LVAD improved haemodynamics compared to ECMO alone. An add-on Impella could not substitute a reduction in ECMO flow. Increased mean aortic pressure, myocardial perfusion pressure and mean pulmonary artery flow were related to sustained cardiac function and ROSC.