15.1 VV and VA ECMO Circuit Overview

An extracorporeal membrane oxygenation (ECMO) circuit consists of drainage and return cannulae, a pump and a membrane oxygenator with heat exchanger (Fig. 15.1a) [1, 2]. Venoarterial (VA) ECMO drains deoxygenated blood through a venous cannula to a centrifugal pump, arranged in series with a membrane oxygenator and returns oxygenated blood via an arterial cannula [2, 3]. In contrast, venovenous (VV) ECMO returns oxygenated blood via a second venous cannula [4].

Fig. 15.1
figure 1

Common circuit components and cannulation strategies for VV ECMO. (a) Single site internal jugular cannulation. (b) Dual site femoral-femoral cannulation. (c) Dual site femoral-internal jugular cannulation

Full size image

15.2 VV ECMO

The use of VV ECMO for acute respiratory distress syndrome (ARDS) has expanded dramatically following positive outcomes published in the CESAR trial [5] and the favorable experience during the H1N1 influenza pandemic of 2009–2010 [6,7,8]. More recent evidence clarifying its role in the management of adults with severe ARDS will likely contribute to increased use of VV ECMO in the future [9]. Despite the expanding role of VV ECMO for ARDS in the nontrauma patient population [5,6,7,8], VV ECMO use in the trauma patient population has been somewhat limited due to continued concerns over bleeding complications associated with systemic anticoagulation and the inflammatory response incited by the ECMO circuit [10,11,12], particularly in patients with traumatic brain injury (TBI) [4, 5, 13]. However, recent observational studies have demonstrated promising results for the use of VV ECMO in both the poly-trauma and TBI patient population with very few reported bleeding complications [18,19,20,21,22,23,24,25,26]. Table 15.1 summarizes the evidence for the use of VV ECMO for ARDS specific to the trauma patient population.

Table 15.1 Evidence for use of VV ECMO specific to trauma patients
Full size table

According to the 2017 Extracorporeal Life Support Organization (ELSO) guidelines [14], VV ECMO should be considered when risk of mortality exceeds 50% [14], and indicated when risk of mortality exceeds 80% [14]. 50% mortality in ARDS is associated with: (1) PaO2/FiO2 < 150 on FiO2 > 90% [15]; (2) Murray score of 2–3 (Table 15.2) [15]; (3) Age-adjusted oxygenation index (AOI) >60 [16]; and (4) ARDS prediction score (APPS) ≥5 [17]. 80% mortality in ARDS is associated with: (1) PaO2/FiO2 < 100 on FiO2 > 90% [15]; (2) Murray score of 3–4 (Table 15.2) [15]; (3) AOI >80 [16]; and (4) APSS ≥8 [17]. While ELSO states there are no absolute contraindications to VV ECMO [14], severely injured poly-trauma [10,11,12] or TBI [4, 13] are considered by many to have relative contraindication to the systemic anticoagulation used in VV ECMO. It is worth noting that the CESAR trial [5], the single randomized controlled trial demonstrating a survival benefit for VV ECMO referrals compared to no-ECMO (relative risk (RR) 0.69, [95% confidence interval (CI): 0.05–0.97]; p = 0.03), included trauma patients (6% of the ECMO cohort) but excluded patients with intracranial bleeding or any contraindication (relative or absolute) to systemic heparinization [5].

Table 15.2 The Murray score is obtained by averaging the parameter scores for each of the following four areas [15]
Full size table

15.3 VV ECMO Circuit Management

15.3.1 VV ECMO Cannulation Strategies

The elements of a typical VV ECMO circuit and the three most common cannulation strategies employed in VV ECMO are shown in Fig. 15.1. The cannula orientation should maximize flow and minimize recirculation [14, 27,28,29] and should be placed under fluoroscopic and echocardiographic guidance if at all possible. In all cases, a bolus 5000 units of Heparin should be administered prior to cannulation to minimize the risk of clot formation and possible circuit thrombosis [1, 14].

Single site dual-lumen cannulation (AvalonElite Bi-caval Dual Lumen Catheter; Maquet, Gothenburg, Sweden) is performed with a 27 or 31 French cannula (depending on the patient’s size and cardiac output) typically using the right internal jugular (IJ) vein. The tip of the cannula is positioned in the mid-IVC a few centimeters below the hepatic veins with drainage occurring through side-ports in the SVC and IVC. The return lumen is approximately 10 cm above the distal tip and should be positioned such that the oxygenated return will flow through the tricuspid valve [29]. This cannulation strategy enables early ambulation but can be somewhat difficult to position.

The other cannulation strategies use single lumen catheters. In bilateral femoral cannulation, venous drainage occurs from a cannula introduced into the femoral vein with the tip placed 5–10 cm below the IVC-RA junction within the intra-hepatic vena cava (drainage side-holes positioned above the collapsible intra-abdominal vena cava). Oxygenated return occurs from a cannula introduced into the contralateral femoral vein with the tip in the RA at the level of the tricuspid valve [14, 27]. This strategy is commonly employed in urgent situations where access to the neck is limited and early ambulation is unlikely. This cannula orientation requires a large caliber vena cava to ensure adequate space for two cannulae. The other 2-site strategy is termed “bi-caval cannulation.” In this approach, venous drainage occurs from a cannula introduced into the femoral vein with the tip placed 5–10 cm below the IVC-RA junction, again within the intra-hepatic vena cava. Oxygenated return occurs through a small caliber, short cannula introduced into the right internal jugular (IJ) vein with the tip at the SVC-RA junction [14, 27, 28]. This approach is ideal for controlled cannulation in most trauma patients who will not be candidates for early ambulation.

15.3.2 Monitoring Targets

Following cannulation and heparinization, the VV ECMO circuit should be unclamped and flows gradually increased to the target flow range, typically ≥60% of the calculated cardiac output (CO) (approximately 50–80 mL/kg/min [3.5–5 L/min]) [14]. Inlet saturation (sampled from the drainage cannula immediately prior to the oxygenator) is a surrogate for SvO2 and should be maintained ≥70% [14, 36]. Outlet saturation (sampled from the return cannula immediately after the oxygenator) should be ≥95% with a PaO2 > 300 mmHg [14, 36]. If the outlet saturation is less than 95%, the oxygenator should be investigated for potential clot formation [14, 36]. FiO2 on the VV ECMO circuit should be titrated to achieve a patient-level arterial saturation of ≥88% [14, 36]. Sweep gas flow (oxygen flow through the gas exchange membrane) on the VV ECMO circuit should be titrated to achieve a patient-level PaCO2 between 30 mmHg and 40 mmHg [14, 36]. VV ECMO does not provide hemodynamic support and therefore will not mitigate the need for inotropic and/or vasopressor support. Inotropes are typically titrated to targets such as SvO2 ≥ 65% or cardiac index (CI) ≥2.0 L/min, and vasopressors titrated to a MAP ≥65 mmHg. In many cases, the patient’s hemodynamics will improve with decreased ventilator pressures and increased systemic oxygen levels.

15.4 VV ECMO Patient Management

15.4.1 Anticoagulation Range

In the absence of any contraindications to systemic anticoagulation, a heparin bolus of 5000 units should be administered prior to cannulation to minimize risk of clot formation while the circuit is clamped [1, 14]. A heparin infusion should then be initiated with a goal ACT of at least 160 s, [1, 14] ideally between 180 s and 220 s [14, 22]. Although aPTT may be used, ESLO guidelines do not recommend its use because it is susceptible to derangements in coagulation factor levels and platelet function which commonly occur in VV ECMO patients [14]. If aPTT is used to monitor ECMO anticoagulation, it should be maintained between 40 s and 50 s [14]. In the setting of TBI, heparin-boned circuitry [19, 30,31,32,33] and a period of heparin-free support have led to successful management of VV ECMO for ARDS in several case series [19, 33,34,35].

15.4.2 Ventilator Management

Ventilator FiO2 should be set on “lung rest” settings with an FiO2 ≤ 0.4 [14, 36], a plateau pressure of ≤25 cm H2O [5, 14, 36], and a PEEP between 5 and 10 cm H2O [5, 14, 36]. Although the ELSO guidelines [14, 36] and the CESAR Trial [5] promote pressure control ventilation (PCV) [5, 36], volume controlled ventilation (VCV) is acceptable, as long as tidal volumes are set at 4–6 mL/kg/ideal body weight and plateau pressures are maintained at ≤25 cm H2O [4]. Debate on the safety of allowing the lungs to “white out” by minimizing ventilator support continues. Regardless, PEEP levels should be decreased judiciously to avoid losing recruited alveolar units that may still be contributing to gas exchange.

15.4.3 Sedation Strategies

For the first 24–48 h after VV ECMO initiation, heavy sedation is recommended [14, 36]. After initial stabilization, a tapered sedation plan should be implemented to allow for early and frequent assessment of neurologic status [37]. Pharmacokinetic and pharmacodynamic changes in the critically ill result in significant variability between drug dosing and response [38]. These pharmacologic derangements are further exaggerated in ECMO patients [39]. The ECMO circuit increases the volume of distribution by either hemodilution and/or sequestration of drugs [39, 40], particularly highly lipophilic drugs [39,40,41,42]. Existing data for appropriate anesthesia and analgesia drug choices on ECMO remains somewhat sparse [43]. Initiating a continuous infusion of an opioid (e.g., fentanyl or hydromorphone) and a sedative (e.g., propofol) during VV ECMO is a reasonable first step [43]. Propofol buildup may start to appear as white streaks in the membrane lung after several days, but the impact of this on membrane efficiency is unknown. If hemodynamically stable, daily sedation interruptions are recommended, especially in anticipation of ECMO weaning and ultimately decannulation [44].

15.4.4 Peri-procedural Management

Surgical procedures can be done successfully while on VV ECMO. When possible, the heparin infusion should be discontinued 6 h prior. If urgent or emergent surgery is necessary, fresh frozen plasma (FFP) should be infused prior to and during surgery; however, pharmacologic reversal with protamine is never recommended because of risk of circuit thrombosis [14]. Electrocautery should be used liberally in surgical cases, and even in minor procedures such as chest tube insertion performed on VV ECMO, to minimize bleeding [14]. For patients who require open surgery while on ECMO, we recommended temporary cavitary closure with intermittent washouts until ECMO has been discontinued, as the patient is very likely to bleed significantly into the closed cavity during ECMO support.

15.4.5 Tracheostomy Timing and Technique

According to the 2017 ELSO guidelines [14], both “early” extubation and tracheostomy (i.e., at 3–5 days post-cannulation) are recommended for those on VV ECMO [14]. Candidates for endotracheal extubation (or no endotracheal intubation) [45] while on VV ECMO support are typically pre-operative lung transplantation cases [46,47,48,49,50]. Unlike pre-operative lung transplant patients, severely injured, polytrauma patients with ARDS are more likely to benefit from early tracheostomy airway management. Although early tracheostomy does not necessarily confer a mortality benefit or decreased duration of mechanical ventilation, it can permit decreased sedation and earlier mobilization [51,52,53]. Careful planning and meticulous hemostasis are essential to the success of a tracheostomy in a patient on VV ECMO and the advised technique differs from a standard tracheostomy [14]. A “hybrid” open/percutaneous technique minimizes the risk of bleeding: (1) hold heparin for 6 h, (2) set the ventilator to room air, (3) expose the anterior trachea through a small incision made with an electrocauter, (4) insert the tracheostomy using a percutaneous dilational technique with a Ciaglia Blue Rhino® (Cook Medical, Bloomington, IN) under bronchoscopic guidance, and (5) resume heparin at the previous infusion rate without a bolus once hemostasis is assured.

15.4.6 Early Mobilization and Physical Therapy

The literature for early physical therapy while on VV ECMO is accumulating [54]. Evidence for the efficacy and safety of early mobilization while on VV ECMO is in the pre-operative lung transplantation population [46,47,48,49,50] facilitated largely by using a dual-lumen cannula in the right IJ (AvalonElite Bi-caval Dual Lumen Catheter; Maquet, Gothenburg, Sweden). Recently, the scope of physical therapy during VV ECMO support has expanded and proven to be both efficacious [55] and safe [56].

15.5 VA ECMO

While the evidence for the use of VV ECMO for ARDS in the trauma patient population is accumulating with positive outcomes [18,19,20,21,22,23,24,25,26], the evidence for VA ECMO following cardiothoracic trauma or traumatic cardiac arrest from exsanguination is inadequate. Table 15.3 summarizes two retrospective, observational cohort studies investigating outcomes of a combined VV and VA ECMO cohort [30, 57]. VV ECMO cases in both studies had a survival benefit, but the VA patients in each study were very heterogeneous with respect to their underlying diagnoses [30, 57]. Future randomized controlled trials comparing VA ECMO to the current standard in a select trauma patient population are warranted.

Table 15.3 Evidence for use of VA ECMO specific to trauma patients
Full size table

To address this evidence gap for the utility of VA ECMO following traumatic arrest, Tisherman and colleagues are actively enrolling in a multicenter clinical trial [58]. This trial is an innovative, parallel assignment, interventional clinical trial comparing “usual care” to “emergency preservation and resuscitation (EPR)” in trauma patients who have exsanguinated to the point of cardiac arrest requiring resuscitative thoracotomy [58, 59]. The investigators define usual care as an emergency thoracotomy, open cardiac massage and fluid resuscitation, and EPR as going onto cardiopulmonary bypass (CPB) by central aortic cannulation in the ascending aorta and central venous cannulation in the right atrium for those patients who fail to achieve return of spontaneous circulation after aortic clamping [59]. These investigators plan to enroll 20 trauma patients (10 assigned to each arm) with a primary outcome of survival to hospital discharge without major disability, and secondary outcomes of (1) feasibility, (2) survival, (3) neurologic functional outcome, and (4) multiple organ dysfunction [58]. This trial represents an important first step in understanding how ECMO may be applied to the management of severely injured trauma patients outside of the typical indications of respiratory failure and the surgical management of tracheobronchial injuries.

15. Conclusions

VV ECMO for ARDS is feasible and safe in the trauma patient population and appears to confer a significant mortality benefit based on retrospective data. In the setting severe ARDS refractory to conventional mechanical ventilation, VV ECMO with delayed systemic anticoagulation is acceptable in those with TBI when combined with vigilant monitoring for circuit thrombosis. ECMO alters the pharmacokinetics and pharmacodynamics of lipophilic and protein-bound medications; so sedation strategies often need to be adjusted significantly during ECMO support. Surgical interventions can be performed, but the techniques used require modification to include liberal use of cautery and damage control techniques with open cavitary management. VA ECMO following traumatic arrest is being evaluated in a single pilot study. Taken together, use of both VV and potentially VA ECMO has the potential to substantially improve outcomes in the severely injured.