Venoarterial Extracorporeal Membrane Oxygenation in Cardiogenic Shock

Evolution in ECMO Technology

The death of a patient from a massive pulmonary embolism at Massachusetts General Hospital in February 1931 inspired the initial use of extracorporeal circulation (4). Advances in pump and oxygenator technology, percutaneous cannulation techniques, critical care management, and durable MCS options synergized to foster the maturation of ECMO as a viable lifesaving modality. Currently over 87,000 patients have been enrolled in the Extracorporeal Life Support Organization registry, including 12,566 adults with VA-ECMO, with the number of VA-ECMO centers increasing markedly in the last decade.

In a VA-ECMO circuit, deoxygenated blood is pulled from the venous circulation by a pump via a large-bore cannula. Patients may be cannulated centrally (Central Illustration, A) or peripherally (Central Illustration, B). Blood passes through the pump into an oxygenator where gas exchange occurs. Oxygenated blood returns via another large-bore cannula to the arterial circulation.

• Download figure
• Open in new tab
• Download powerpoint

Central Illustration

Basic Cannulation for VA-ECMO Support

(A) Central cannulation. (B) Peripheral cannulation. (C) Peripheral cannulation with distal perfusion catheter. (D1) Upper extremity cannulation with internal jugular venous cannula and axillary artery arterial cannula. (D2) Patients with this configuration may be able to ambulate if clinically appropriate. ECMO = extracorporeal membrane oxygenation; VA = venoarterial.

Although an exhaustive summary of circuit technology is beyond the scope of this review, it is important to note 2 critical advances. First, the development of hollow tube fiber membranes in the oxygenator allowed low resistance and improved blood compatibility characteristics (5). Second, redesigned centrifugal pumps limited heat generation and thrombogenicity, making extended duration of support feasible (6). These advances have recently been coupled to miniaturized circuits, facilitating transport of patients on ECMO.

Although central cannulation remains the primary approach in post-cardiotomy patients, novel percutaneous approaches have resulted in wider utilization of ECMO, including in-hospital based programs that place patients in cardiac arrest on ECMO support (extracorporeal cardiopulmonary resuscitation), service delivery programs with “in the field” ECMO cannulation, and periprocedural ECMO in cardiac catheterization laboratories. Distal perfusion catheters that direct a proportion of the returned oxygenated blood flow from the ECMO circuit to the distal limb of the cannulated leg significantly limit risks for critical limb ischemia in femoral arterial cannulation (Central Illustration, C) (7). More recently, upper extremity peripheral cannulation approaches allow increased mobility for some patients (Central Illustration, D1 and D2).

Hybrid ECMO configurations are increasingly used in patients with severe lung injury or in those inadequately supported with VA- or VV-ECMO. The veno-arteriovenous (V-AV) configuration is one of the more commonly used approaches. Venous blood returns to the oxygenator in the usual fashion and is reinfused via an arterial cannula to the femoral artery and a second venous cannula to the right heart at the level of the tricuspid valve, providing supra-oxygenated pulmonary blood flow. This configuration avoids harlequin (north/south) syndrome, in which deoxygenated cerebral blood flow occurs during retrograde perfusion with peripheral cannulation, discussed in the section Prevent Upper Body Hypoxia [Harlequin (North/South)] Syndrome. VV-ECMO can also be converted to V-AV ECMO when cardiac function deteriorates in a patient initially presenting with isolated pulmonary failure (8).

Advances in circuit technology required parallel advances in bedside management of ECMO patients, including development of multidisciplinary ECMO teams including cardiac surgeons, cardiologists, intensivists, ECMO specialist nurses, perfusionists, and pharmacists.

Device Selection

Percutaneous devices used in CS are compared in Online Table 1. ECMO is the only form of support useful in cases of hypoxemia due to pulmonary failure and the only device that simultaneously supports the right ventricle (RV).

Patient Selection and Clinical Outcomes

Despite advances in technology, survival among patients on VA-ECMO support remains modest, with in-hospital mortality of 50% to 60% and 6-month survival as low as 30% (1,9,10). Just as ECMO technology has evolved, so too has our understanding of the importance of appropriate patient selection to optimize outcomes, efficiently allocate resources, and avoid medical futility. Indications and contraindications for ECMO application are outlined in Table 1.

Several studies have highlight the importance of the underlying diagnosis in determining survival, as summarized in Table 2. Patients with potentially reversible causes of myocardial injury, such as fulminant myocarditis or primary graft failure, have better survival than patients with CS after surgery or acute myocardial infarction (11–14). Patients for whom ECMO is deployed during or immediately after cardiac arrest have an especially poor prognosis (1,15–17). Whether extracorporeal cardiopulmonary resuscitation is superior to conventional cardiopulmonary resuscitation for out-of-hospital arrest remains unclear (16,17).

In addition to underlying diagnoses, pre-ECMO risk factors independently associated with poor outcomes include older age, female sex, and higher body mass index, as well as markers of illness severity including renal, hepatic, or central nervous system dysfunction, longer duration of mechanical ventilation before deployment, elevated serum lactate levels, and reduced prothrombin activity (1,12,14,18). Risk scores have been developed that incorporate these variables to aid in decision-making regarding utility (vs. futility) of ECMO deployment (14,18,19). Most recently, the survival after veno-arterial-ECMO (SAVE) score was developed using data from 3846 adult patients enrolled in the international Extracorporeal Life Support Organization registry to stratify patients into 5 risk categories that correlated with post-ECMO survival (Table 3) (14).

Given the importance of end-organ function in determining outcomes, timing of ECMO initiation is key. Just as premature utilization may expose a patient to undue risks and complications, delayed initiation may be medically futile. The ideal window for deployment is after other, less invasive treatments have been considered or exhausted but before the onset of significant end-organ dysfunction. Recognizing this, some centers developed mobile ECMO services, in which ECMO teams travel to initiate ECMO remotely and return to the center for ongoing management (20).

Ultimately, patient selection for ECMO utilization must take into consideration the underlying diagnosis, patient-specific risk factors, anticipated duration of support, and, perhaps most importantly, whether a viable exit strategy such as recovery, durable MCS, or HT exists. Given the high mortality and complication rates, early consultation with palliative care specialists should be considered.

Patient Management

Suggested goals of ECMO support are listed in Table 4. The primary goal is restoration of tissue and end-organ perfusion to allow stabilization or recovery of function. Nuances of ECMO management, including left ventricular (LV) venting, focus on achieving myocardial recovery and preventing pulmonary damage. When recovery is unlikely, ECMO provides time to assess neurological function, social barriers, and other disease processes that may prohibit durable LV assist device and/or transplant.

Manage gas exchange

Maintaining appropriate oxygenation is a critical component of ECMO management. Oxygen delivery can be adjusted via the ECMO circuit (Figure 1A) or by mechanical ventilation, using strategies to reduce barotrauma and promote lung rest (21). Although the deleterious effects of prolonged hypoxia are well known, supranormal levels of oxygen (hyperoxia) are uniformly associated with worse outcomes, with 1 multicenter study of adults post-cardiac arrest demonstrating a 24% increase in mortality for every 100 mm Hg increase in Pao₂ (22,23). Hyperoxia on VA-ECMO occurs because of the high efficiency of modern oxygenators and can be avoided by reducing the Fio₂ of gas passing through the oxygenation filter (sweep gas) (Figure 1A) to maintain Pao₂ values between 60 and 100 mm Hg.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

ECMO Circuit Controls and In-Line Continuous Dialysis Configuration

ECMO controls used to adjust flow and Fio₂ of sweep gas (A) and pump speed (B) to achieve appropriate pump flow for end-organ perfusion. Auto mode sets the goal flow, and speed automatically adjusts to maintain the set flow. Speed can also be set manually depending on the patient’s condition. Monitors and alarms are incorporated into the control panel to alert providers of changes in line pressure or air in the circuit. (C) A dialysis filter can be incorporated into the circuit. CRRT = continuous renal replacement therapy; ECMO = extracorporeal membrane oxygenation. (A) Reprinted with permission from Sechrist. (B) Courtesy of LivaNova PLC/Sorin Group. (C) Reprinted with permission from Seczyńska B, Królikowski W, Nowak I, Jankowski M, Szułdrzyński K, Szczeklik W. Continuous renal replacement therapy during extracorporeal membrane oxygenation in patients treated in medical intensive care unit: technical considerations. Ther Apher Dial 2014;18:523–34.

Respiratory acidosis should be avoided. Because of decreased transpulmonary blood flow during VA-ECMO, the ability of the lungs to clear CO₂ is impaired, independent of existing air space disease. CO₂ clearance can be controlled by increasing the sweep gas flow relative to blood flow through the membrane filter to remove excess CO₂ or by decreasing it if alkalosis occurs. In mechanically ventilated patients, ventilator settings may be modified, but high tidal volumes and/or peak and plateau pressures that exceed 25 cm H₂O should be avoided whenever possible to avoid barotrauma (21,24).

Weaning ECMO Support

At the crux of the decision to wean support is the demonstration of adequate myocardial recovery to provide sufficient blood and oxygen delivery to end organs to meet metabolic demands. Therefore, end-organ dysfunction, particularly pulmonary failure, should be either recovered or supported by other means (hemodialysis, mechanical ventilation) before decannulation. Therapeutic bronchoscopy to minimize dead space before weaning may be considered (41). Invasive hemodynamic monitoring and bedside echocardiography are complementary in evaluating hemodynamics and myocardial function as ECMO is being weaned (42). Although the degree of acceptable pharmacological hemodynamic support is debated, data suggest that lower levels of inotropes and vasopressors at the time of weaning are associated with improved outcomes (43), likely reflecting improved intrinsic myocardial function.

Both fast and slow weaning protocols have been described (43,44). Although data supporting a specific strategy are limited, an algorithmic approach is recommended (Figure 2). In patients deemed ready for weaning, a stepwise bedside decrease in ECMO flow increases preload to the heart, allowing the clinician to assess cardiac recovery using hemodynamic and echocardiographic data (42,43). Increasing pulse pressure on the arterial line waveform without concomitant LV or RV distention while ECMO circuit flows gradually decrease indicates improved cardiac contractility (42,45). Typically, a formal bedside wean is performed before the final wean, with pharmacological aid to optimize hemodynamic conditions. If the results are satisfactory, a final wean is scheduled in the operating room, which would allow controlled decannulation or expedited recannulation and reinstitution of support if necessary.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Suggested ECMO Weaning Algorithm

CI = cardiac index; CVP = central venous pressure; ECMO = extracorporeal membrane oxygenation; LV = left ventricle; MAP = mean arterial pressure; MCS = mechanical circulatory support; OR = operating room; PAM = pulmonary arterial mean pressure; RV = right ventricle; VTI = velocity-time integral.

If cardiac recovery is unlikely or cannot be achieved despite medical optimization and recovery of end-organ function, direct HT or durable MCS should be considered (46,47). Direct HT from VA-ECMO support should be approached with caution, however, given poor post-transplant outcomes in this group. Renal insufficiency (estimated glomerular filtration rate <45 ml/min/1.73 m² or on hemodialysis) and mechanical ventilation predict worse prognosis (48). If no viable long-term support options exist, withdrawal of support is inevitable.

Complications With VA-ECMO Support

Use of VA-ECMO is associated with a broad range of complications, some of which significantly impact morbidity and mortality. In addition, patients undergoing ECMO often have pre-existing end-organ damage, making the attribution of adverse events difficult. Given the lack of randomized controlled trials, it is challenging to ascertain the exact prevalence and incidence of ECMO-related complications; however, single-center studies, multicenter registries, and meta-analyses provide valuable insight. Table 5 summarizes some of the common complications encountered during ECMO support.

The studies summarized in Table 5 are heterogeneous. Patient age, ECMO cannulation configuration (VA vs. VV, peripheral vs. central), and ECMO indication varied. Definitions of various complications were not standardized, further limiting the ability to characterize complications in a uniform, comprehensive way. Because most ECMO outcomes data come from small, single-center, observational reports or administrative data, inherent selection bias, regional differences, lack of granularity, and practice variation limit comparisons. Collection of prospective data using standardized protocols is needed.

ECMO Economics

Current trends in resource utilization to develop and maintain high-quality ECMO programs are of significant interest at multiple levels, including local institutions and government and third-party payers. The U.S. Nationwide Inpatient Sample documented a >700% increase in ECMO utilization between 2002 and 2012 (352 to 2715 total ECMO discharges) (2,49), with total national charges for ECMO rising from $109 million to >$700 million over this same general time period (50).

Cost estimates for in-hospital care of ECMO patients vary significantly, and more reliable methods to report costs will be critical as policymakers attempt to maximize value. Although comparison of costs in a private delivery model (such as the United States) and a public system (most systems outside the United States) are challenging, costs for ECMO in the United States generally exceed $100,000 per patient, whereas per-patient costs in one-half of international centers are less (51). Despite cost concerns, data suggest that percutaneous circulatory support utilization, including ECMO, results in decreased mortality and in-hospital costs for patients in CS, possibly due to an avoidance of end-organ dysfunction leading to shorter hospital stays (52).