1.1

Preoperative ECMO bridging to lung transplantation

ECMO as a bridge-to-transplantation (BTT) is a life-saving option for patients suffering from an etiology of end-stage lung disease (ESLD) presenting with medically refractory hypercapnia, hypoxemia, or hemodynamic instability. The type of ECMO most commonly used is veno-venous (VV) but may also be veno-arterial (VA). The key difference between these two types of ECMO, in addition to cannulation configuration [ ], is that VV ECMO provides primarily respiratory support while VA ECMO provides cardiac and respiratory support. Prior to cannulation for BTT, cardiothoracic transplant anesthesiologists should be trained to participate in the core multidisciplinary team (MDT) responsible for determining the institutional criteria, discussing ethical considerations for placing recipients on ECMO [ ], optimal timing for the individualized recipient to be BTT, and cannulation strategy [ ]. Once the decision to BTT has been made, anesthesiologists are directly involved in the cannulation of these patients, which may include echocardiographic guidance [ ] or general anesthesia for placement. Ideally, an “awake ECMO” strategy should be employed, which refers to placement of ECMO support in a spontaneously breathing patient with little to no sedative medications [ ]. This approach promotes an intensive rehabilitation program as soon as possible, with benefits of this strategy including early ambulation, reduction in ventilator-associated pneumonia, and prevention of skeletal muscle deconditioning [ ].

A recent study from the United Network for Organ Sharing (UNOS) database examined outcomes of a cohort of 1387 patients who were BTT with ECMO, with stratification of survival eras into four 5-year categories from 2000 to 2019 [ ]. Survival improvement was noted as time progressed, despite increased recipient morbidity and age in more recent eras in both the ECMO (aHR 0.59, 95% CI 0.37–0.96) and non-ECMO (aHR 0.74, 85% CI 0.70–0.79) groups as compared to the initial eras. However, the multivariable Cox regression model found that ECMO was associated with a 52% higher hazard of death (aHR 1.5, p < 0.001) [ ]. Progressive research has shown divergence in outcomes of patients with ECMO BTT based on their specific etiology of ESLD, as recipients with primary pulmonary hypertension (PPH) had a lower transplantation rate than those with obstructive lung disease, cystic fibrosis (CF), and interstitial lung disease [ ].

Similarly, anesthetic management of patients presenting for BTT with ECMO should be tailored to recipient ESLD. Due to lack of specific studies on optimal anesthetic management for BTT yet keeping in mind the goal of a cardiopulmonary stable placement with minimal sedation, principles of management are primarily extrapolated from expert review [ ] and consensus statements driven by expert opinion [ , ]. The highest risk of a challenging sedation is seen in recipient ESLD that is either primary pulmonary hypertension [ ] or high-risk of secondary pulmonary hypertension [ ]. The most important predictor for tailoring anesthetic management of BTT is the ability of the patient to lay flat, as this subjective preoperative exam finding informs the hemodynamic status. Other key predictors for tailoring anesthetic management of BTT includes oxygenation requirements, right ventricular function, right ventricular size, and the presence or absence of esophageal dysmotility [ ]. Maintenance of adequate minute ventilation, adequate perfusion pressure of the right ventricle, and avoidance of positive pressure ventilation or sudden increases in right ventricular afterload are paramount [ ]. ECMO cannulation approach, which can vary significantly between VV ECMO and VA ECMO, may impact anesthetic approach if cannulation vessels require surgical cutdown, high neck access, or require invasive echocardiographic imaging modalities for placement [ ]. Detailed plans should include not only the primary approach for awake ECMO, but also have redundancy for an unplanned induction of general anesthesia.

1.2

Intraoperative management of ECMO in lung transplantation

Whether a patient is BTT or planned for intraoperative ECMO only, there are many anesthetic considerations for the intraoperative management of ECMO during lung transplantation including hemodynamic monitoring, anesthetic, respiratory, and hematological. The coordination of these approaches is similar for patients who are placed on VV or VA ECMO, although VA ECMO is trending as the preferred modality of intraoperative support during lung transplantation.

1.3

Intraoperative management of hemodynamic monitoring

Intraoperative hemodynamic monitoring considerations for intraoperative ECMO should include both macrocirculatory and microcirculatory monitoring. For macrocirculatory monitoring, the use of standard American Society of Anesthesiologists (ASA) monitoring is mandatory, and the placement of a right arm invasive arterial line, central venous catheters, pulmonary venous catheters, and transesophageal echocardiography (TEE) is nearly always mandatory as well. Preoperative placement of pulse oximetry should preferentially be placed on the right hand, as it assesses not only the adequacy of systemic oxygenation but also allows for monitoring of North-South syndrome in patients on peripheral VA ECMO, which has an estimated frequency of 10% [ ]. North-South syndrome, which is an inadequacy of oxygen in the upper part of the body due to competition of the relatively deoxygenated trans -pulmonary flow as compared to the retrograde femoral arterial ECMO flow [ ], should also be monitored with the placement of the arterial line in the right radial artery. Even if central VA ECMO is planned [ ], the utilization of an awake placement of a right radial arterial line is recommended prior to induction of anesthesia to ensure both beat to beat monitoring and monitoring for North-South syndrome in case of switch to peripheral VA ECMO for postoperative prolongation.

The Vienna group suggests maintaining an VA ECMO flow at half of the calculated cardiac output [ ], while also suggesting that pulsatility within the systemic arterial and pulmonary arterial waveforms be utilized as further qualitative measures of macrocirculatory transpulmonary flow during intraoperative VA ECMO [ ]. In the absence of TEE, the use of end-tidal carbon dioxide (ETCO2) monitoring is useful for monitoring the macrocirculation of transpulmonary flow during VA ECMO [ ]. Due to the diversion of cardiac output by VA ECMO support, the qualitative presence of ETCO2 confirms transpulmonary flow, while the quantitative reading allows for a relative trend measuring increases or decreases in transpulmonary flow. If a TEE is available, a structured intraoperative measurement of baseline cardiac output with the measurement of the aortic time velocity integral (VTI) or the VTI of flow across the pulmonary vein ratioed to VA ECMO flow may theoretically provide a superior method of setting VA ECMO flow [ ]. Given the lack of cardiac output diversion in VV ECMO, the same considerations for measuring transpulmonary flow are not indicated, however, these monitors should continue to be used to ensure warm ischemia time is avoided.

Beyond quantitative measurement of transpulmonary flow, echocardiography plays an important role in evaluating the overall macrocirculation during lung transplantation. A complete exam should always be performed before anesthetic induction (transthoracic) and throughout the surgery (TEE), with particular focus on minimizing technical complications [ ] related to ECMO wire/cannulae placement, baseline right ventricular anatomy, right ventricular function, adequacy of left ventricular function, pulmonary vasculature anastomoses [ ], presence of a patent foramen ovale and aorta structure. Modalities to be utilized should include 2D, pulse wave Doppler, continuous wave Doppler, color flow Doppler, and 3D echocardiography [ ], as a combination of both quantitative and qualitative echocardiographic measurements are required to ensure appropriate management of intraoperative ECMO in lung transplantation.

Hemodynamic coherence is the connection of the adequacy of microcirculation and the adequacy of macrocirculation [ ]. Monitoring the microcirculation can be achieved during the intraoperative use of ECMO through the use of sublingual microcirculation [ ], peripheral near-infrared spectroscopy (NIRS) [ ], cerebral oximetry [ ], central venous-to-arterial carbon dioxide difference [ ] or laboratory markers such as lactate [ ]. Given the impracticality of sublingual microcirculatory monitoring in a patient with a double-lumen endotracheal tube and TEE probe, it is not recommended for intraoperative use. However, the use of peripheral NIRS is recommended for the detection of ischemia in peripheral VA ECMO, as it has been shown to promote earlier detection of ischemia and avoidance of surgical morbidity in this population [ ]. Cerebral oximetry provides monitoring for both North-South Syndrome as well as cerebral microcirculation [ ], and should be considered a mandatory anesthetic monitor when utilizing VA ECMO intraoperatively. Intraoperative levels of lactate, which has been shown to correlate with outcomes in lung transplantation, should be drawn on either an event [ ] or time structured [ ] schedule to provide ongoing assessment of microcirculation while on VA ECMO support.

1.4

Intraoperative management of anesthetic medications

Studies examining optimal anesthetic choice during lung transplantation are very rare, with the majority focusing on intravenous versus inhaled anesthetics in animal studies [ ]. Cardiac surgical models have shown that propofol may attenuate the release of inflammatory mediator IL-8 [ ], while studies in thoracic surgical models have studies which show suppression of alveolar inflammatory response with inhaled anesthetics [ , ] and suppression of IL-6 and IL-10 with a propofol anesthetic approach [ ]. A randomized controlled trial examined the potential myocardial effect of propofol in a normothermic cardiac pulmonary bypass (CPB) model, finding that its preemptive use may be protective in patients with type 2 diabetes [ ]. The significant amount of confounders combined with the pharmacological impact of mechanical circulatory support and ventilation-perfusion inadequacy has resulted in lack of actionable anesthetic recommendation data to date, however, a strong recommendation is the utilization of a processed electroencephalogram (EEG) monitor for the depth of anesthesia during the use of intraoperative ECMO in lung transplantation [ ].

1.5

Intraoperative management of protective ventilation

Most of the recommendations for lung transplantation ventilation strategy are derived from studies of acute respiratory distress syndrome and elective thoracic surgery requiring one-lung ventilation [ ]. There are limited observational studies examining optimal perioperative ventilatory strategies in lung transplantation, and the influence of different strategies on clinical outcomes remains largely unknown [ ]. The key respiratory considerations during the intraoperative phase of lung transplantation includes the use of a fractional inspired oxygen (FiO2) level <40% [ ] and a lung protective strategy, both of which can be easier achieved when the patient is on VV or VA ECMO support. The use of an FiO2 at reperfusion of >40% is associated with the development of primary graft dysfunction (PGD) [ ]. The only randomized controlled trial regarding optimal intraoperative protective lung ventilation strategies was performed in an analysis of 30 patients post-reperfusion of the first implanted lung [ ]. One arm was composed of patients with a stepwise recruitment, positive end-expiratory pressure (PEEP), and pressure-controlled ventilation with tidal volume based on ideal body weight. This cohort was compared to a volume-controlled ventilation strategy with a PEEP of 5 cmH2O, and combination of 6 ml/kg tidal volume on two lungs and 4 ml/kg tidal volume on single lung ventilation [ ]. The results showed no significant difference between the two strategies as measured in their primary outcome of PaO2/FiO2 ratio at 24 h (p = 0.26) [ ]. Despite lack of data supporting an ideal practical strategy, general consensus is that a lung protective strategy that avoids volutrauma, barotrauma, and atelectrauma is recommended. Volutrauma is best avoided by utilizing a 4–6 ml/kg delivered volume limit based on donor ideal body weight [ ], with a maximal plateau pressure of <25 cmH2O set as the goal to avoid barotrauma as the chest wall is open [ ]. Similar to tailoring hemodynamic management to the underlying ESLD, the heterogenous nature of lung tissue depending on the recipient ESLD demands a tailored respiratory strategy during the explanation phase. Therefore, a lung protective strategy should be designed for both the explantation and post-implantation stages, with the anesthesiologist utilizing the oxygenation/decarboxylation benefit of VV or VA ECMO to optimize this approach.

A final respiratory consideration for patients on intraoperative ECMO is the decision of whether to fast-track extubate the patient in the operating room. Extubation at the end of a lung transplantation in the operating room has been associated with a good prognosis [ ], and studies examining predictor of intraoperative extubation report that the presence of intraoperative ECMO is not a contraindication to intraoperative extubation [ ]. Early extubation, similar to the preoperative phase, allows for improved patient mobility and rehabilitation. This early extubation strategy may also apply to patients who are on VA ECMO for hemodynamic as opposed to pulmonary reasons, such as those with primary pulmonary hypertension [ ].

1.6

Intraoperative blood management

Hematological considerations for intraoperative ECMO management include transfusion of blood products and maintenance of appropriate anticoagulation. The management of perioperative hemostasis remains an important issue in lung transplantation since ECMO itself may induce impairment of primary hemostasis [ ]. While blood product transfusion has been shown to be associated with poor outcomes, targeted transfusion tailored to the patient’s needs and guided by point-of-care coagulation tests may be of interest. Perioperative transfusion of allogeneic blood products in both donor [ ] and recipient has been widely associated with an increased risk of PGD and 1-year mortality [ ]. Distinguishing between PGD and transfusion-related lung injury is difficult [ ], but transfusion should be kept at a minimum if possible. Antifibrinolysis and coagulation factors are commonly used during lung transplantation and are not contraindicated for use with intraoperative ECMO [ , ]. Therefore, a goal-directed transfusion strategy based on point-of-care coagulation testing (POCCT) is recommended. A POCCT-based strategy has shown to have benefit in both cardiac surgery [ ] and liver transplantation [ ], but limited studies are available for this strategy in lung transplantation.

A retrospective study examining the use of a POCCT-based strategy in a CPB model of lung transplantation showed a significant decrease in blood transfusion with minimal changes in outcomes [ ]. Similarly, a randomized controlled trial within an ECMO model of lung transplantation showed significantly favorable results in a POCCT-based strategy with resultant decrease in perioperative blood loss, transfusion of red blood cells, and transfusion of fresh frozen plasma [ ]. A bias within this study was a difference between the two arms in terms of intraoperative resuscitative fluids, with one receiving 5% albumin and another group receiving a mix of crystalloids and colloids [ ].

Anticoagulation considerations for intraoperative ECMO vary between centers. The primary method of anticoagulation intraoperatively is the administration of heparin, although perioperative anticoagulants that may be administered include bivalirudin or argatroban [ ]. Intraoperative VV ECMO may be performed with no anticoagulation at all, and literature for intraoperative VA ECMO is controversial [ ]. It may include the monitoring of activated clotting time (ACT) with a range of 180–220 s [ , ] or the use of alternative tests including anti -Xa assay [ ], activated partial thromboplastin time (APTT) [ ] or viscoelastic methods such as rotational thromboelastometry (ROTEM) [ ]. The principle of balance between prevention of thrombosis and risk of increased bleeding should guide the intraoperative hematological care of ECMO patients, and minimizing the dose of anticoagulation in a situation of significant bleeding is recommended [ , , ].

1.7

Postoperative utilization of ECMO

ECMO may be used in the postoperative phase of care either as a continuation of the intraoperative phase for immediate graft failure [ ], baseline recipient ESLD [ ], or deployed in the intensive care unit (ICU) as a rescue device. There are two approaches regarding the maintenance of the ECMO into the postoperative phase: routine maintenance or withdrawal in case of a favorable weaning trial. The approach of routine maintenance for all patients is falling out of favor, and the Vienna Lung Transplant Group has change its practice over the past few years from advocating routine postoperative ECMO maintenance to ECMO continuation in selected patients at the end of surgery (8% of all cases) [ , ]. Groups have advocated in the literature for prolongation of all PPH patients [ ], but this too has changed in favor of a more nuanced approach as evidenced by discussion from the most recent surgical consensus statement on postoperative ECMO [ ]. A formal weaning trial from ECMO is recommended, with a retrospective analysis of 74 patients showing positive predictive factors include younger donor age, high donor PaO2, and shorter operating time [ ]. Anesthetic considerations for weaning from ECMO postoperatively, whether VV or VA ECMO, include the balance of a recipient’s graft function and ventilation requirements. However, VA ECMO weaning considerations also include the optimization of cardiac function, as the removal of VA ECMO flow results in a greater preload return to the ventricles. Therefore, the continuation of a lung protective ventilation strategy combined with echocardiographic guidance of a structured weaning exam is necessary for the anesthesiologist to guide postoperative VA ECMO withdrawal.

2.1

Practice points

• •
  

  Anesthesiologists should engage as members of the multidisciplinary selection committee to aid in comprehensive perioperative planning for ECMO management

  
  
• •
  

  Preoperative bridging of ECMO should preferably be achieved in the awake state, with minimal to no sedation, to avoid consequences of cardiopulmonary instability observed with positive pressure ventilation and general anesthesia

  
  
• •
  

  Intraoperative management of ECMO requires a diverse set of hemodynamic monitors to achieve appropriate macrocirculatory and microcirculatory support

  
  
• •
  

  Goals for intraoperative management of ECMO include avoidance of primary graft dysfunction and technical complications associated with ECMO use

  
  
• •
  

  Postoperative anesthetic considerations include optimization of cardiopulmonary function and structured echocardiography to guide weaning

2.2

Research agenda

• •
  

  Further research into etiology of end stage lung disease specific approaches for perioperative ECMO are needed

  
  
• •
  

  Echocardiographic indices to assess optimal ECMO management during intraoperative and postoperative phases of care should be investigated

  
  
• •
  

  Optimal hematological monitoring and transfusion thresholds during intraoperative ECMO for lung transplantation should be researched

  
  
• •
  

  Research examining the pros/cons of a total intravenous anesthetic versus inhalational technique on outcomes in ECMO for lung transplantation is needed