Comparison of Circulatory Unloading Techniques for Venoarterial Extracorporeal Membrane Oxygenation : ASAIO Journal

Secondary Logo

Journal Logo

Adult Circulatory Support

Comparison of Circulatory Unloading Techniques for Venoarterial Extracorporeal Membrane Oxygenation

Stephens, Andrew F.*,†; Wanigasekara, Devindi*,†; Pellegrino, Vincent A.‡,§; Burrell, Aidan J.C.‡,§; Marasco, Silvana F.‡,¶; Kaye, David M.‡,∥; Steinseifer, Ulrich#; Gregory, Shaun D.*,†

Author Information
ASAIO Journal 67(6):p 623-631, June 2021. | DOI: 10.1097/MAT.0000000000001268


Left ventricular (LV) distention and pulmonary congestion are major complications inherent to venoarterial extracorporeal membrane oxygenation (ECMO). This study aimed to quantitatively compare the hemodynamic differences between common circulatory unloading methods for ECMO. Ten circulatory unloading techniques were evaluated on a mock circulatory loop simulating acute LV failure supported by ECMO. Simulated unloading techniques included: surgical and percutaneous pulmonary artery (PA) venting, surgical left atrial venting, surgical and percutaneous LV venting, atrial septal defect, partial support ventricular assist device, intraaortic balloon pump, and temporary VAD with inline oxygenator (tVAD). The most LV unloading occurred with the surgically placed LV vent and tVAD, which reduced LV end-diastolic volume from 295 to 167 ml and 82 ml, respectively. Meanwhile, the PA surgical vent was the most effective at reducing mean PA pressure from 21.0 to 10.6 mm Hg, and the tVAD was most effective at reducing left atrial pressure from 13.3 to 4.4 mm Hg. The major limitation of this study was the use of a mock circulatory loop, which simulated lower left atrial pressure than is typically seen clinically. This study identified clinically significant hemodynamic variability between the different circulatory unloading techniques evaluated. However, the applicability of these techniques will vary with different patient disease etiology. Further studies on ECMO unloading will help to quantify hemodynamic benefits and establish treatment guidelines.

Venoarterial extracorporeal membrane oxygenation (ECMO) is used to treat a wide variety of cardiac illnesses. Inherent to ECMO treatment is an elevation of left ventricular (LV) afterload caused by the blood flow from the ECMO arterial return cannula. In cases of compromised LV function, the heart cannot eject against the additional afterload, leading to LV distention, which, in turn, leads to pulmonary congestion. Pulmonary congestion can lead to irreversible pulmonary failure, while LV distention results in a positive feedback loop, which may cause degradation of LV function and subsequent progression of heart failure.1 Finally, blood stagnation in a nonejecting ventricle can lead to the formation of a ventricular or aortic thrombus.1

In many patients, the first course of unloading is to use partial ECMO flow in combination with inotropes to promote LV ejection.1,2 In cases of severe LV dysfunction, this may prove ineffective, and other mechanical methods of circulatory unloading are used. These methods include transeptal balloon/blade septostomy or sheath;3–5 intraaortic balloon pump;6 partial ventricular assist devices (VADs) such as Impella 2.5 or Impella CP (Abiomed, Danvers, MA);7,8 temporary VADs (tVADs) such as the CentriMag (Abbot Labs., Abbott Park, IL), TandemHeart (CardiacAssist Inc., Pittsburgh, PA), or Impella 5.0 (Abiomed)9,10; and catheter interventions such as LV/left atrial (LA)/pulmonary artery (PA) vents, which connect in parallel to the ECMO drainage circuit.11–14 Each technique for circulatory unloading has shown some clinical evidence for improving patient outcomes3–7,9–13; however, each technique comes with associated risks and may be contraindicated depending on disease etiology.

Previous limited studies comparing ECMO unloading techniques have been conducted in numerical simulations15 and in small numbers of pigs.16 Currently, there are no published studies extensively comparing different circulatory unloading techniques directly in a homogeneous patient cohort. Sophisticated physical cardiovascular simulators can emulate specific patient etiologies and directly compare unloading techniques in a controlled, repeatable manner. Direct, quantitative comparison of unloading techniques and their effects on hemodynamics can be combined with clinical experience, allowing for the establishment of guidelines for determining the best circulatory unloading technique based on patient requirements and the available resources and expertise of the attending institution.

This study aims to directly compare circulatory unloading techniques to quantify their effects on LV loading (LV end-diastolic volume [LVEDV], ventricular stroke work [SW]), pulmonary congestion (LA pressure [LAP], mean PA pressure [MPAP]), and ventricular blood stagnation (cardiac output through the aortic valve [CO]). For comparison, each unloading technique was independently evaluated at different ECMO flow rates in a benchtop cardiovascular system simulating an acute LV heart failure patient, as per acute myocarditis or ischemic cardiogenic shock.

Materials and Methods

Experimental Test Rig

A sophisticated benchtop cardiovascular simulator (mock circulatory loop [MCL]) was used for all experiments. The MCL is described in detail elsewhere.17 Briefly, the MCL is a physical 5-element Windkessel model of the pulmonary and systemic circulations. Four pneumatically driven polyvinylchloride cylinders simulate the left and right ventricles and atria, which are separated by mechanical umbrella valves representing the atrioventricular and semilunar valves. Four larger polyvinylchloride cylinders act as venous and arterial compliance chambers. The ventricles contain a Frank-Starling mechanism, which adjusts ventricular contraction relative to measured end-diastolic volume. Pulmonary and systemic vascular resistances (PVR and SVR) are autoregulated to a user-defined value.

The MCL was set to simulate a severe acute left heart failure condition with a functionally intact right ventricle. Baseline hemodynamics simulating acute left heart failure were mean arterial pressure of 56.0 mm Hg, LAP of 11.1 mm Hg, LVEDV of 222 ml, MPAP of 21.0 mm Hg, CO of 2.1 L/min, and SW of 0.33 W.18 The standard ECMO configuration was venoarterial cannulation (Figure 1), whereby a 25 Fr drainage cannula was attached to the MCL’s right atrium (simulated by a tube with 0.8 cm inner diameter—ID, and 75 cm long) and was connected to the ECMO pump inlet (Bio-Medicus 550 Bio-Console with a Rotaflow adaptor and pump head, Medtronic Inc., Dublin, Ireland) by 200 cm of 3/8th inch (0.85 cm) tubing. Seventy centimeters of 3/8th inch tubing connected the ECMO pump outlet to the oxygenator (Quadrox D, Getinge, Gothenburg, Sweden). Finally, 200 cm of 3/8th inch tubing connected the oxygenator to the 19 Fr return cannula (simulated by a 0.6 cm ID tube, 30 cm long), which was attached to the MCL’s aorta. These values were chosen to represent a typical clinical setup in the authors’ institution.

Figure 1.:
Basic mock circulatory loop supported by VA-ECMO configuration. Illustrative only, not to scale. AOC, PAC, PVC, SVC, aortic and pulmonary arterial, pulmonary and systemic venous compliance chambers; SQ, PQ, EQ, systemic, pulmonary, and ECMO flow meters; Pump, Ox, ECMO pump and oxygenator; LA, LV, RA, RV, left and right atria and ventricles; MV, AV, TV, PV, mitral, aortic, tricuspid, and pulmonary valves; SVR, PVR, systemic and pulmonary vascular resistance pinch valves; ID, inner diameter; VA-ECMO, venoarterial extracorporeal membrane oxygenation.

The working fluid was a glycerol and water solution mixed to a viscosity of 3.6 ± 0.1 cP within the physiologic range of blood viscosity. All data were acquired into a dSPACE 1202 MicroLabBox (dSPACE GmbH, Paderborn, Germany) at a rate of 2 kHz and down-sampled to 200 Hz for postprocessing. All pressures were measured by TruWave disposable pressure sensors (Edwards Lifesciences, Irvine, CA), while all flows were measured by clamp-on ultrasonic sensors (em-tec BioProTT 3/8" × 3/32" for ECMO flow; 3/8" × 1/8" for systemic flow; 1" × 1/8" for pulmonary flow; SonoTT DigiFlow board, em-tec GmbH; Finning; Germany). Left and right ventricular volumes were measured using linear magnetic level sensors (MTL4-650MM, Miran Technology Co., Shenzen, China). Outflow from the ventricle (CO) was calculated as the difference between total systemic flow and ECMO flow or as the difference between total systemic flow and ECMO flow plus VAD flow. SW was calculated as , where MAP is mean arterial pressure, LVEDP is LV end-diastolic pressure, HR is heart rate, and 0.1333 is a constant converting L·mm Hg/min to Pa·m3. For all experiments, the MCL heart rate was maintained at 80 bpm with a systolic duty of 35%. When establishing the heart failure condition, SVR and PVR were initially set to 2000 and 380 Dyne·s·cm−5, respectively, and then reduced to 1200 and 160 Dyne·s·cm−5 following ECMO establishment, representing the clinical use of vasodilators.19

Experimental Protocol

The standard protocol for evaluating each of the circulatory unloading methods was to establish the baseline heart failure condition and then initiate ECMO support. ECMO pump flow was then increased in 1 L/min increments to achieve ECMO flow rates from 1 to 4 L/min, where possible. At each ECMO flow increment, the hemodynamics were recorded following the establishment of the steady-state. Each experiment was conducted only once for each of the unloading techniques due to the highly repeatable nature of the MCL used.17 Due to only one repeat of each data point, statistical significance was forgone, focusing on clinical significance instead.

The experiments compared 10 commonly used types of circulatory unloading techniques:

  1. No unloading
  2. Percutaneous PA vent13
  3. Surgical PA vent20
  4. Surgical LA vent21
  5. Percutaneous LV vent13,22
  6. Surgical LV vent11,12,21
  7. Atrial septal defect (ASD)3–5
  8. Temporary VAD with inline oxygenator (tVAD)10,16
  9. partial assist VAD (VAD)7,8
  10. Intraaortic balloon pump (IABP).6,23

The values for each of the cannulae sizes for the surgical and percutaneous vents were based on the literature, and typical values used within the authors’ institution.20 Details of each circulatory unloading configuration are outlined below, while schematics of the MCL configuration for each technique are supplied in the Supplemental Digital Content, and

The PA cannulae consisted of either 0.5 cm (16 Fr—percutaneous) or 0.8 cm ID (25 Fr—surgical) tubing, 75 cm long, which was then connected via a Y-connector to the ECMO drainage circuit at the interface between the drainage cannula and the 3/8th inch (0.85 cm) tubing leading to the ECMO pump inlet. Similarly, the surgical LA and LV cannulae were simulated using 0.8 cm (25 Fr) ID tubes, which were 75 cm long; these were again connected directly to the drainage circuit using a Y-connector at the drainage cannula interface. A 0.2 cm (6 Fr) pigtail catheter was connected to the LV to represent percutaneous direct LV unloading. The pigtail, which was 110 cm long, was connected to a Luer port on the drainage cannula (at the interface with the 3/8th inch tubing) by a 25 cm × 0.01 cm ID pressure monitoring line. The ASD was simulated by connecting an 80 cm section of ½ inch (1.27 cm) tubing directly between the left and right atria to allow pressure exchange.

The tVAD was established to model a TandemHeart device, a temporary extracorporeal support pump. In clinical practice, the drainage cannula is inserted transeptally into the LA, and the infusion cannula is positioned in the femoral artery. Case studies have shown potential to combine the TandemHeart with an inline oxygenator to create a makeshift ECMO circuit that drains from the LA, theoretically avoiding pulmonary congestion.7,8 In these experiments, the TandemHeart was simulated by switching the standard ECMO circuit to drain from the left atrium using the same drainage and infusion cannulae sizes (25 and 19 Fr, respectively) with the same lengths of tubing connecting the cannula to the pump and oxygenator circuit. In these experiments, the tVAD flow was adjusted between 1 and 4 L/min (maximum achievable flow) in 1 L/min increments.

For simulating partial VAD support (as with Impella 2.5 and Impella CP), a HeartWare HVAD (Medtronic Inc., Dublin, Ireland) was used in a partial support capacity. The VAD cannulation sites were the LV and aorta for the inlet and outlet, respectively. Evaluation of VAD unloading was achieved by setting the MCL to the baseline heart failure condition (as used for all other experiments), turning on the VAD, and adjusting the speed to reach the desired flow rate of 2.5 L/min. Once the VAD flow had been established, ECMO support was then initiated, and the ECMO flow was adjusted, as per the other experiments. The VAD speed was maintained for all ECMO flow rates, despite a distinct drop in VAD outflow due to the increased afterload caused by ECMO support.

The IABP was placed in the MCL’s simulated aorta, after the compliance chamber and proximal to the heart, compared with the ECMO return cannula. The IABP (40 ml displacement) was operated in a 1:1 counter-pulse support mode. Balloon inflation was triggered by the LV pressure waveform, which was consistently pulsatile, unlike aortic pressure, where pulsatility diminished as ECMO support increased. ECMO was then established, and ECMO flow was adjusted as per other experiments.


Hemodynamics were recorded for all ten unloading techniques at ECMO flow rates between 1 and 4 L/min. Unless otherwise specified, all results and discussion hereon compare the specified unloading technique to no unloading technique at an ECMO flow rate of 4 L/min.

The tVAD provided the best reduction in LVEDV from 295 to 82 ml, more than double the LVEDV reduction provided by the surgical LV vent, which reduced LVEDV to 167 ml and was the second-best technique for LVEDV reduction (Figure 2). Meanwhile, the percutaneous LV vent and IABP provided the least ventricular unloading with an LVEDV of 289 and 301 ml, respectively. Interestingly, the concomitant use of IABP with ECMO in the simulated patient resulted in an increased LVEDV from 295 to 301 ml when compared with no unloading and no IABP support, possibly due to an increased afterload caused by IABP treatment.

Figure 2.:
Variations in LVEDV with respect to ECMO flow rates. PA, pulmonary artery; LA, left atrium; LV, left ventricle; ASD, atrial septal defect; VAD, partial support ventricular assist device; IABP, intraaortic balloon pump; tVAD, temporary VAD with inline oxygenator; LVEDV, left ventricular end-diastolic volume; ECMO, extracorporeal membrane oxygenation.

Resulting reductions in SW unsurprisingly mirror the results for reduction in LVEDV with the tVAD and surgical LV vent providing the greatest reductions in ventricular SW from 0.143 W baseline to 0.013 and 0.026 W, respectively. It is interesting to note that although the difference in LVEDV reduction between the tVAD and surgical LV vent was more than double, the differences in SW were relatively minor in terms of cardiac workload. It is also interesting to note that while the reduction in LVEDV from the VAD was only slightly lower than other unloading techniques at low ECMO flow rates (165 ml at 1 L/min), the VAD demonstrated a consistently good reduction in SW (0.065–0.078 W) across the entire ECMO flow range due to the large and immediate reduction in LVEDV provided by the VAD as soon as it was initiated. Meanwhile, the IABP produced an increase in SW compared with no unloading across all ECMO flows (0.156 compared with 0.143 W at 4 L/min), due to the increased afterload provided by the IABP. Additionally, the percutaneous PA and percutaneous LV vents only had a minimal reduction in SW, both resulting in an SW of 0.13 W at 4 L/min ECMO flow rate (Figure 3).

Figure 3.:
Variations in SW with respect to ECMO flow rates. PA, pulmonary artery; LA, left atrium; LV, left ventricle; ASD, atrial septal defect; VAD, partial support ventricular assist device; IABP, intraaortic balloon pump; tVAD, temporary VAD with inline oxygenator; SW, stroke work; ECMO, extracorporeal membrane oxygenation.

The surgical PA vent was the most effective at reducing MPAP, providing a reduction from 15.2 to 10.6 mm Hg, while the tVAD was the second most effective, reducing MPAP to 11.6 mm Hg. Interestingly, the surgical LV vent was among the worst at unloading MPAP (14.6 mm Hg), alongside the VAD (14.8 mm Hg), the percutaneous LV vent (14.9 mm Hg), and the IABP (15.6 mm Hg) (Figure 4). Meanwhile, the tVAD reduced LAP from the baseline 13.3 to 4.4 mm Hg, while the surgical LA vent and ASD were also effective at LAP reduction (8.4 and 8.5 mm Hg, respectively). The least effective methods of reducing LAP were the IABP (13.8 mm Hg), the percutaneous LV vent (12.9 mm Hg), and the VAD (13.8 mm Hg) (Figure 5).

Figure 4.:
Variations in MPAP with respect to ECMO flow rates. PA, pulmonary artery; LA, left atrium; LV, left ventricle; ASD, atrial septal defect; VAD, partial support ventricular assist device; IABP, intraaortic balloon pump; tVAD, temporary VAD with inline oxygenator; MPAP, mean pulmonary artery pressure; ECMO, extracorporeal membrane oxygenation.
Figure 5.:
Variations in LAP with respect to ECMO flow rates. PA, pulmonary artery; LA, left atrium; LV, left ventricle; ASD, atrial septal defect; VAD, partial support ventricular assist device; IABP, intraaortic balloon pump; tVAD, temporary VAD with inline oxygenator; LAP, left atrial pressure; ECMO, extracorporeal membrane oxygenation.

Native cardiac output (CO) reduced linearly for all unloading methods due to the increased ventricular afterload caused by the increasing ECMO outflow. The IABP encouraged the most native cardiac output (CO) with 1.1 L/min at ECMO flow rates of 4 L/min (Figure 6). Conversely, the surgical LV vent, tVAD, and VAD all provided no native CO at 4 L/min ECMO flow rate, with the VAD providing no native CO at ECMO flow rates above 2 L/min.

Figure 6.:
Variations in native CO with respect to ECMO flow rates. PA, pulmonary artery; LA, left atrium; LV, left ventricle; ASD, atrial septal defect; VAD, partial support ventricular assist device; IABP, intraaortic balloon pump; tVAD, temporary VAD with inline oxygenator; CO, cardiac output; ECMO, extracorporeal membrane oxygenation.


This study compared different circulatory unloading techniques for ECMO support in a benchtop cardiovascular simulator. Each circulatory unloading technique was found to have varying levels of effect on different patient hemodynamics. For instance, the surgical LV vent was effective at reducing LVEDV, SW, and LAP, resulting in LV decompression but ineffective at reducing MPAP or encouraging CO through the aortic valve, potentially resulting in pulmonary edema or stagnation of blood within the ventricle. Conversely, the IABP was ineffective at reducing LVEDV, MPAP, LAP, and SW but increased native CO.

From a pure analysis of the hemodynamics evaluated in this study, the tVAD was found to have the best performance, demonstrating the most reduction in LVEDV, LAP, and third most reduction CO. Meanwhile, the surgical LV vent performed the second best with the second-highest reduction in LVEDV and the most considerable reductions in SW and CO. Although both of these unloading techniques performed well hemodynamically, there are many considerations to take into account when choosing a circulatory unloading technique. As a disclaimer, to the author’s knowledge, none of the techniques investigated herein are endorsed by the FDA for concomitant use with ECMO.

Data from studies such as this may be used to identify indications and contraindications of which circulatory unloading technique should be initiated based on patient disease etiology and hemodynamic condition. Although each circulatory unloading technique comes with inherent risk, there is some evidence to support that earlier unloading is an independent factor predicting reduced ECMO treatment times. In a small patient cohort, Kotani et al. described 23 pediatric patients who had LV decompression performed on VA-ECMO via adjustable ASD. They showed that early LV decompression reduced ECMO weaning times in patients with severe LV dysfunction.14 This is corroborated by other studies, which suggest that LV decompression is a factor in improved survival rate, encouragement of myocardial recovery, and decreased ECMO weaning time.18,24,25 Results presented herein demonstrated that some unloading techniques were effective at reducing SW and may, therefore, be useful for enhancing myocardial recovery. Unloading techniques in this study which reduced SW by more than one-third compared with no unloading (0.143 W) at ECMO flow rates of 4 L/min included tVAD (0.0013 W); surgical venting in the LA, PA, and LV (0.091, 0.065, and 0.026 W, respectively); ASD (0.091 W); and VAD (0.091 W).

Meanwhile, a reduction in MPAP and LAP can lead to a reduction in fatal pulmonary edema and right heart failure. In this study, tVAD, percutaneous and surgical PA vent, ASD, and surgical LA vent demonstrated a minor reduction in MPAP, reducing pressure from 15.2 mm Hg to between 10.6 and 13.3 mm Hg. The unloading techniques that notably reduced LAP (from 13.3 mm Hg) were the tVAD (4.4 mm Hg), surgical LA vent (8.5 mm Hg), ASD (8.5 mm Hg), and the surgical PA and LV vents (9.3 and 9.5 mm Hg, respectively). These reductions in pressure are smaller than the ranges presented by Donker et al.,15 who showed reductions in MPAP from 36 to 23 mm Hg (surgical LA vent, surgical LV vent, Impella 2.5), 22 mm Hg (surgical PA vent), and 18 mm Hg (with an ASD of 1.5 cm2) in a numerical simulation study. The lower initial MPAP and LAP in this study are due to the high-volume capacity of the MCL’s left ventricle as further discussed in the limitations of this article.

Surgically placed catheters are among the more commonly used unloading techniques during ECMO support. This study showed that different catheterization sites had different effects on hemodynamics. The surgical LA vent was effective at reducing LAP (13.3–8.5 mm Hg) and LVEDV (295–194 ml), while the surgical PA vent reduced LVEDV (295–210), MPAP (15.2–10.6 mm Hg), and LAP (13.3–8.5 mm Hg); and the surgical LV vent reduced LAP (13.3–9.5 mm Hg) and LVEDV (295–167 mm Hg). The drawback of these unloading techniques is that they require surgical cut-downs. Surgery comes with inherent risks, and such techniques may be contraindicated in patients who are too unstable. Nonsurgical vents can be placed, but these are limited in size, which may affect unloading efficacy.26,27

Diminished unloading efficacy by the smaller diameter percutaneous vents was demonstrated in this study. Comparing the percutaneous LV vent with the surgical LV vent shows a marked difference in LVEDV reduction (289 and 167 ml, respectively), LAP reduction (12.7 and 9.5 mm Hg, respectively), and SW reduction (0.13 and 0.026 W, respectively). Meanwhile, a comparison of the percutaneous PA vent with its surgical counterpart shows differences in LVEDV reduction (244–210 ml, respectively) and SW reduction (0.13–0.065 W, respectively). Interestingly, there were only minimal differences in MPAP and LAP between the percutaneous and surgical PA vents during these experiments (MPAP 12.7 and 10.6 mm Hg and LAP 11.1 and 9.3 mm Hg for percutaneous and surgical, respectively). Non-catheter based percutaneous unloading techniques were also evaluated in this article, each with widely different effects on hemodynamics.

The ASD is a percutaneous unloading technique, which presents as an enticing, high-skill but low-risk solution for circulatory unloading. In this study, the ASD maintained low levels of LVEDV (194 ml), MPAP (13.3 mm Hg), LAP (8.5 mm Hg), and SW (0.091 W), with many of these factors minimally affected by ECMO flow rate. However, the ASD creates a flow shunt between the atria and allows flow to bypass the ventricles, potentially promoting ventricular blood stagnation and stroke.28 In this study, a reduction of blood flow through the ventricles due to the ASD was demonstrated by a somewhat reduced native CO (to a minimum of 0.5 L/min). This reduction in blood flow was clinically significant, but not more so than the surgical LV vent, tVAD, and VAD, which all had a minimum CO of 0 L/min at 4 L/min ECMO flow rate. The use of an ASD with ECMO was supported by Baruteau et al., who in a study of 64 patients saw LA pressures reduce from 24.2 ± 6.9 to 7.8 ± 2.6 mm Hg using intraseptal balloon septostomy with a left to right atrial pressure gradient of 0.09 ± 0.5 mm Hg and no thrombus formation for up to 24 days of ECMO treatment.3

Another percutaneous unloading technique is a partial support VAD like the Impella 2.5. In this study, the VAD demonstrated a reduction in SW from baseline 0.143 W to between 0.065 W at 1 L/min ECMO flow to 0.078 W at 4 L/min ECMO flow. Efficacy of the VAD at unloading LVEDV and MPAP decreased as ECMO flow was increased, due to the increasing pressure head across the VAD cause by the ECMO outflow. The use of an HVAD to mimic an Impella is a limitation of this study due to the differences in pressure-flow characteristics. Regardless, since the FDA approval of the Impella 2.5 for use in cardiogenic shock, several researchers have investigated the use of VADs with ECMO for circulatory unloading. Some of these studies have concluded that VAD concomitant with ECMO is safe and reduces pulmonary edema,7,8,24,29 while other studies have concluded that the use of such devices leads to bleeding and hemolysis, and poses a high risk to patients.30,31 Partial support VADs such as the Impella 2.5 and Impella CP have the distinct benefit of being easy to use, especially in less-experienced ECMO centers.31

In this study, the IABP improved CO slightly compared with no unloading technique (1.1 compared with 1.0 L/min), but was the least effective at reducing LVEDV, MPAP, LAP, and SW; this may have been due to the severely reduced native cardiac function as simulated in the model or otherwise due to the higher afterload created by the IABP in counter-pulsation with the native heartbeat. Efficacy of IABP concomitant with ECMO is highly contentious; in a metaanalysis, Cheng et al.6 found no significant difference in survival with concomitant IABP and ECMO. Alternatively, other studies have found a clinical improvement in survival of ECMO with IABP, particularly in children.22,32,33 Further investigation may be required associating ventricular function with IABP efficacy when combined with venoarterial ECMO.


The main limitation of this study is that it was conducted on a MCL, which is a physical simulation of the heart and circulatory system. Although the MCL has been validated extensively as a hemodynamic model, there are many biologic and anatomical functions, which it does not simulate. For instance, the MCL lacks any mitral or aortic insufficiency, which may be present in some ECMO patients (significant regurgitation is usually a contraindication for ECMO support); this may significantly affect the efficacy of some unloading techniques and will be the subject of future in vitro research. The MCL is also incapable of accurately mimicking ventricular suction, which may be a flow-limiting factor for concomitant ECMO and VAD use. Also, the percutaneous vents in this study were connected to the same cannulation site on the MCL as the surgical vents. In the body, placing a percutaneous cannula will affect the overall flow dynamics within the circulation path, while for venting cannulae or a partial assist VAD placed across valves, some regurgitant flow may also take place, a phenomenon which was not emulated in the MCL.

In this study, an acute systolic heart failure condition was simulated in part due to the MCL’s limited capacity to simulate diastolic heart failure. This is due to the large pneumatically driven ventricle that has a filling capacity of about 480 ml, which is in contrast to the native ventricle, which during diastolic heart failure, will fill to the capacity of the ventricle and then congest causing elevated blood pressures. This limitation of the MCL results in higher LVEDV and lower LAP and MPAP than might be present in a human body with heart failure (13.3 mm Hg LAP here, compared with a typical 16–25 mm Hg in some patients) and to those hemodynamics present in other numerical comparison studies.15,19 These limitations may, in turn, affect the hemodynamics with respect for unloading techniques, particularly those relating to pulmonary edema such as LAP and MPAP.

Additionally, the MCL only simulated a single adult patient with a single etiology combined with only 1 size for the inflow, outflow, and where applicable, venting cannulae. Changing the ratio of sizes between these cannulae will create differences in the resistances and, therefore, efficacy of many of the evaluated unloading techniques at any given flow rate. Further work should be conducted to characterize these relationships, particularly for the catheter-based unloading techniques. Also, according to the Extracorporeal Life Support Organization registry, more than half of ECMO patients are pediatric or neonates whose physiology will vary significantly from the adult population, therefore the results of this experiment may not translate to that population.


This study quantified the effects of different circulatory unloading techniques at different ECMO flow rates and demonstrated that the effect on hemodynamics greatly varied between common circulatory unloading techniques. Despite this, the surgical vents, ASD, and tVAD, all demonstrated adequate circulatory unloading for the measured hemodynamics. A systematic, controlled review of unloading techniques combined with medical experience will help to form clinical guidelines on the best circulatory unloading technique based on each patient’s needs and improve patient outcomes by reducing LV distention, ventricular stagnation, and pulmonary edema.


1. Soleimani B, Pae WE. Management of left ventricular distension during peripheral extracorporeal membrane oxygenation for cardiogenic shock. Perfusion (United Kingdom). 27: 326–331, 2012.
2. Rajagopal K. Left ventricular distension in veno-arterial extracorporeal membrane oxygenation: From mechanics to therapies. ASAIO J. 65: 1–10, 2019.
3. Baruteau AE, Barnetche T, Morin L, et al. Percutaneous balloon atrial septostomy on top of venoarterial extracorporeal membrane oxygenation results in safe and effective left heart decompression. Eur Heart J Acute Cardiovasc Care. 7: 70–79, 2018.
4. Alhussein M, Osten M, Horlick E, et al. Percutaneous left atrial decompression in adults with refractory cardiogenic shock supported with veno-arterial extracorporeal membrane oxygenation. J Card Surg. 32: 396–401, 2017.
5. Seib PM, Faulkner SC, Erickson CC, et al. Blade and balloon atrial septostomy for left heart decompression in patients with severe ventricular dysfunction on extracorporeal membrane oxygenation. Catheter Cardiovasc Interv. 46: 179–186, 1999.
6. Cheng R, Makkar R, Chang D, et al. TCT-70 No survival benefit with use of intra-aortic balloon pump in extracorporeal membrane oxygenation: A pooled experience of 1,517 patients. J Am Coll Cardiol. 64: B21, 2014.
7. Koeckert MS, Jorde UP, Naka Y, Moses JW, Takayama H. Impella LP 2.5 for left ventricular unloading during venoarterial extracorporeal membrane oxygenation support. J Card Surg. 26: 666–668, 2011.
8. Cheng A, Swartz MF, Massey HT. Impella to unload the left ventricle during peripheral extracorporeal membrane oxygenation. ASAIO J. 59: 533–536, 2013.
9. Weymann A, Sabashnikov A, Bowles CT, et al. Central extracorporeal life support with left ventricular decompression for the treatment of refractory cardiogenic shock and lung failure. J Cardiothorac Surg. 9: 60, 2014.
10. Li YW, Rosenblum WD, Gass AL, Weiss MB, Aronow WS. Combination use of a tandem heart with an extracorporeal oxygenator in the treatment of five patients with refractory cardiogenic shock after acute myocardial infarction. Am J Ther. 20: 213–218, 2013.
11. Rao P, Mosier J, Malo J, et al. Peripheral VA-ECMO with direct biventricular decompression for refractory cardiogenic shock. Perfusion (United Kingdom). 33: 493–495, 2018.
12. Fumagalli R, Bombino M, Borelli M, et al. Percutaneous bridge to heart transplantation by venoarterial ECMO and transaortic left ventricular venting. Int J Artif Organs. 27: 410–413, 2004.
13. Avalli L, Maggioni E, Sangalli F, Favini G, Formica F, Fumagalli R. Percutaneous left-heart decompression during extracorporeal membrane oxygenation: An alternative to surgical and transeptal venting in adult patients. ASAIO J. 57: 38–40, 2011.
14. Kotani Y, Chetan D, Rodrigues W, et al. Left atrial decompression during venoarterial extracorporeal membrane oxygenation for left ventricular failure in children: Current strategy and clinical outcomes. Artif Organs. 37: 21–28, 2013.
15. Donker DW, Brodie D, Henriques JPS, Broomé M. Left ventricular unloading during veno-arterial ECMO: A simulation study. ASAIO J. 65: 11–20, 2019.
16. Esposito ML, Shah N, Dow S, et al. Distinct effects of left or right atrial cannulation on left ventricular hemodynamics in a swine model of acute myocardial injury. ASAIO J. 62: 671–676, 2016.
17. Gregory SD, Pauls JP, Wu EL, et al. An advanced mock circulation loop for in vitro cardiovascular device evaluation. Artif Organs. 44: E238–E250, 2020.
18. Truby LK, Takeda K, Mauro C, et al. Incidence and implications of left ventricular distention during venoarterial extracorporeal membrane oxygenation support. ASAIO J. 63: 257–265, 2017.
19. Delle Karth G, Buberl A, Geppert A, et al. Hemodynamic effects of a continuous infusion of levosimendan in critically ill patients with cardiogenic shock requiring catecholamines. Acta Anaesthesiol Scand. 47: 1251–1256, 2003.
20. Rupprecht L, Flörchinger B, Schopka S, et al. Cardiac decompression on extracorporeal life support: A review and discussion of the literature. ASAIO J. 59: 547–553, 2013.
21. Hacking DF, Best D, d’Udekem Y, et al. Elective decompression of the left ventricle in pediatric patients may reduce the duration of venoarterial extracorporeal membrane oxygenation. Artif Organs. 39: 319–326, 2015.
22. Lin YJ, Liu HY, Kuo HC, et al. Left ventricle decompression strategies in pediatric peripheral extracorporeal membrane oxygenation. Acta Cardiol Sin. 35: 335–341, 2019.
23. Sauren LD, Accord RE, Hamzeh K, et al. Combined impella and intra-aortic balloon pump support to improve both ventricular unloading and coronary blood flow for myocardial recovery: An experimental study. Artif Organs. 31: 839–842, 2007.
24. Patel SM, Lipinski J, Al-Kindi SG, et al. Simultaneous venoarterial extracorporeal membrane oxygenation and percutaneous left ventricular decompression therapy with impella is associated with improved outcomes in refractory cardiogenic shock. ASAIO J. 65: 21–28, 2019.
25. Aiyagari RM, Rocchini AP, Remenapp RT, Graziano JN. Decompression of the left atrium during extracorporeal membrane oxygenation using a transseptal cannula incorporated into the circuit. Crit Care Med. 34: 2603–2606, 2006.
26. Hong TH, Byun JH, Yoo BH, Hwang SW, Kim HY, Park JH. Successful left-heart decompression during extracorporeal membrane oxygenation in an adult patient by percutaneous transaortic catheter venting. Korean J Thorac Cardiovasc Surg. 48: 210–213, 2015.
27. Alkhouli M, Narins CR, Lehoux J, Knight PA, Waits B, Ling FS. Percutaneous decompression of the left ventricle in cardiogenic shock patients on venoarterial extracorporeal membrane oxygenation. J Card Surg. 31: 177–182, 2016.
28. Madhavan M, Yao X, Sangaralingham LR, et al. Ischemic stroke or systemic embolism after transseptal ablation of arrhythmias in patients with cardiac implantable electronic devices. J Am Heart Assoc. 5: e003163, 2016.
29. Dixon SR, Henriques JP, Mauri L, et al. A prospective feasibility trial investigating the use of the Impella 2.5 system in patients undergoing high-risk percutaneous coronary intervention (The PROTECT I Trial): Initial U.S. experience. JACC Cardiovasc Interv. 2: 91–96, 2009.
30. Badiye AP, Hernandez GA, Novoa I, Chaparro SV. Incidence of hemolysis in patients with cardiogenic shock treated with impella percutaneous left ventricular assist device. ASAIO J. 62: 11–14, 2016.
31. Lauten A, Engström AE, Jung C, et al. Percutaneous left-ventricular support with the impella-2.5-assist device in acute cardiogenic shock results of the impella-EUROSHOCK-Registry. Circ Heart Fail. 6: 23–30, 2013.
32. Doll N, Kiaii B, Borger M, et al. Five-year results of 219 consecutive patients treated with extracorporeal membrane oxygenation for refractory postoperative cardiogenic shock. Ann Thorac Surg. 77: 1460–1462, 2004.
33. Gass A, Palaniswamy C, Aronow WS, et al. Peripheral venoarterial extracorporeal membrane oxygenation in combination with intra-aortic balloon counterpulsation in Patients with cardiovascular compromise. Cardiology. 129: 137–143, 2014.

ECMO; circulatory unloading; mechanical circulatory support; ventricular decompression; pulmonary hypertension; venting

Supplemental Digital Content

Copyright © ASAIO 2020