Advantages, disadvantages and rationale for ECMO

 ECMO is half of "Mechanical supports: ECMO/IABP", an L1 topic of Section 2.1.4 in the second edition of the CICM Syllabus for the Second Part Examination. Considerting that each of those technologies is a massive topic all to itself, they each occupy a large chunk of the Cardiovascular Intensive Care section.  The modest goal of this chapter is to focus on briefly answering the question, "what is ECMO actually for?" 

Where ECMO has made appearances in the CICM past papers, this has never been explicitly asked, except in Question 17 from the second paper of 2024 where advantages and disadvantages of VV ECMO were questioned. Question 23 from the first paper of 2014 asked something about the "relative merits" of VV vs VA ECMO (in Legionella pneumonia), which sounds a bit like advantages and disadvantages, and Question 11 from the second paper of 2010 asked the candidates to "outline the role" of ECMO, which resembles rationale. 

In short:

Interestingly, there is basically no literature to recommend here; every article is either a review which briefly skims the technological aspects before diving directly into patient selection, or a clinical study describing the advantages and disadvantages in terms of specific outcomes or complications. In terms of references, the author's modest attention span was best suited to appreciate the details in the VV paper by Akoumianaki et al (2021), but the VA material was distributed so diffusely across so many papers that it became impossible to cite properly.

Rationale for the use of VV ECMO

One may be tempted to quip that the rationale of a parachute or crash helmet is self-evident on first inspection, and so "prevent death and organ failure" might seem like a reasonable thing to list as the "rationale" for the use of ECMO, but in fact this technology has some very specific applications. it would important to list these for the exams, even though they could all be summarised as "do gas exchange without flogging the lungs". 

• Improve oxygenation without using oxygen concentrations that could add to pulmonary toxicity or using positive pressure that could cause barotrauma
• Improve CO₂ removal without resorting to ventilation with injurious tidal volumes or rapid cyclical pressure changes that could exacerbate lung injury.
• Permit the airway structures (eg. tracheal, bronchial, etc) to remain free from airway devices, to allow access for surgery or healing following surgery
• Permit minimal ventilation or persistent apnoea to allow the healing of otherwise inaccessible defects (eg. in bronchopleural fistula)
• Permit procedures that would otherwise be impossible because of poor tolerance of general anaesthesia, one-lung ventilation, or impossible intubation (eg. severe subglottic stenosis).

It may at this stage help to digress somewhat on the topic of that common CICM exam trope, the highly ritualised "critically evaluate" question, where people seem to forever misunderstand what is meant by "rationale" and what belongs in "advantages".  Rationale is also not the same as "indications", though indications are really examples of the rationale applied to specific circumstances. The rationale for the use of VV ECMO is the reasoning behind the technology, the logic that explains the choice of scenarios where it is implemented; whereas the advantages of VV ECMO are the beneficial effects of ultra-protective ventilation and the flow-on effects of not using high positive pressure.

Advantages of VV ECMO

Where by "advantages", we mean "as compared to persevering with conventional ventilation, where the respiratory failure is severe enough to merit ECMO".

So:

• ECMO allows ventilation with protective volumes and pressures
• This prevents further lung injury and allows healing
• The modest pressures permitted by ECMO allow for adequate venous return and better haemodynamics, which in turn improves organ function 
• The ability to remove CO₂ removes the need for "permissive hypercapnia" which has adverse neurological and haemodynamic effects
• The patient can be weaned from sedation and paralysis earlier, which decreases the risk of ICU-acquired weakness and delirium (plus the transition to spontaneous breathing is also faster)
• The lack of dependence on high pressures and heavy-handed sedation makes cough safer and easier, which reduces the risk of pneumothorax and bronchopleural fistula (plus improves sputum clearance)
• Awake patients supported on ECMO can participate in chest physiotherapy without breathlessness-related exercise intolerance

Disadvantages of VV ECMO

It is unclear whether a question on the advantages or disadvantages of VV ECMO would accept generic answers on the risks of all ECMO, or whether only specifically the disadvantages of venovenous circuit configurations would be expected. Just in case, it's be worth including twenty words on it, like: 

• ECMO in general is resource-expensive and carries a risk of vascular injuries, DVT, PE, pressure injuries, and anticoagulation-related bleeding risks. 

Beyond this boilerplate, the disadvantages of VV ECMO would have to be focused on the specific problems that are faced with this modality which are not seen with traditional ventilation, such as:

• Unsuitable for patients with very high cardiac output (eg. young patients with respiratory sepsis and hyperdynamic circulation)
• Unsuitable for patients with a very low cardiac output (eg. where the respiratory failure is complicated by cardiogenic shock)
• The use of ultralow tidal volumes and modest pressures can lead to the development of atelectasis and pneumonia in previously viable lung
• The risk of post-decannulation VTE from DVTs that formed in partially obstructed femoral veins can undo a lot of the positive gas exchange benefits of protective ventilation
• Reduced mobility (or, increased risk of cannula dislodgement with mobility)
• Recirculation can limit the efficiency of the circuit
• Repositioning the patient becomes more dangerous, which leads to a reluctance to prone or regularly rotate the patient's position
• Increased complexity of care leads to the increased propensity towards error, whereas conventional ventilation is simpler and less error-prone

And, from a more organisational/logistic perspective,

• The troubleshooting and maintenance of the circuit requires more nursing staff training
• Often two nurses need to be rostered to care for one ECMO patient, which limits the total pool of staff
• ECMO-specific training may be necessary for staff who do not have routine exposure
• For senior medical staff, low rates of ECMO exposure can make this unfamiliar technique more risky than the relatively better-practiced ventilator techniques

Rationale for VA ECMO

Having largely resisted the urge to invent a statement explaining the rationale for having a functioning circulatory system, one is still tempted to point out that VA ECMO is, at a fundamental level, an extraordinary step taken to delay circulatory death. But let's say one is posed with several choices to achieve this, where VA ECMO is a reasonable option. This makes it easier to think of the rationale for choosing ECMO by forcing one to consider the unpleasant consequences of the alternatives. The rationale for choosing it would be something like:

• Prevent counterproductive systemic inflammatory changes and organ system dysfunction associated with a persistently low cardiac output
• Prevent the translocation of gut bacteria and other complications associated with high dose vasopressors
• Avoid cardiac inotrope (eg. catecholamine) toxicity by reducing the requirements for inotropes 
• Improve the perfusion of coronary arteries by retrograde aortic flow
• Improve right and left ventricular function by reducing preload
• Permit high risk rescue procedures (eg. angiography) that would otherwise be awkward or impossible; and/or delay these until the patient's condition is more favourable
• Avoid trauma from CPR or sternotomy (for peripherally configured VA ECMO circuits)

Next, the advantages and disadvantages mainly refer to peripherally cannulated and conventionally configured VA ECMO, rather than hybrid techniques or percutaneously inserted VADs like Impella, as the whole point of those is often to overcome some of these disadvantages. 

Advantages of VA ECMO

The reader will note that this list was assembled very carefully stepping around the absolutely natural urge to use "not being dead" as an advantage:

• As a bridge to transplant for eligible patients, VA ECMO permits patient recovery from causes of cardiac arrest which would otherwise be irreversible
• Offers an option for inotrope-unresponsive reversible causes of cardiac failure (eg. Takotsubo, acute valve failure, cardiodepressnt overdose, etc) 
• Introduces the option of stabilising the patient and reversing various modifiable risks prior to some definitive procedure
• Increases the patient's access to rescue procedures by reducing the risk of multiorgan system failure that can accompany high dose vasopressors and inotropes
• Decreases the bias against further active management which may develop where a patient has unfavourable-looking trends in inotrope/vasopressor support requirements
• For patients who are neurological nonsurvivors (specifically for the context of ECPR), VA ECMO preserves organ function enough for the option of organ donation to be discussed (though one would need to point out the morally grey nature of any management that is intended for something other than the benefit of the patient, i.e. ethically one should never put patients on VA ECMO with this specific goal in mind)

Disadvantages of VA ECMO

Again, rather than repeating an entire span of text full of handwringing about cost, anticoagulation, stroke, vascular injury, pressure areas etc, it is enough to remind the CICM second part exam candidate that they are at the end of their training and should be expected to recognise that these are common ECMO problems. Yes, to list them all would be tedious, and may not score many marks, as it does not require the spark of genuis to identify the potential for bleeding when one places a 19Fr arterial catheter and then anticoagulates the patient. If the question is asking something more specific, one may even be able to omit these. The non-boring VA ECMO disadvantages are as follows:

• Dangerous for patients with severe aortic atheromatous disease
• Unsuitable in patients with severe aortic regurgitation
• Potentially very inefficient in patients with large AV shunts
• Increases LV afterload, which is counterproductive to the process of LV recovery following injury
• Can create LV stasis leading to thrombosis
• Bypassing the pulmonary circulation may reduce pulmonary blood flow and promote intrapulmonary thrombosis
• Can produce differential oxygenation with recovering LV function, which could have neurological sequelae
• Competes for space in the aorta with other rescue devices that could reduce afterload (eg. IABP)
• By introducing what seems like a favourable delay for stabilisation of the patient, VA ECMO may allow reversible pathology to consolidate, eg. a viable myocardium which could have been rescued will fully infarct while waiting.
• Weaning from VA ECMO is much more complex than weaning from vasopressor/inotrope/IABP support

And, even though we agreed not to mention these,

• Embolic phenomena are common to all ECMO scenarios but in VA ECMO these are particularly devastating
• Anticoagulation for VA ECMO is much more important than for VV because of this and other aspects, which increases the risk of haemorrhagic complications

Lastly, peripheral concerns at an organisational/societal level include:

• The level of expertise required to safely coordinate the care for such patients often limits the access to such services to major metropolitan centres, which raises questions about the ethics of such resource allocation in terms of distributive justice
• The expectation of the availability of VA ECMO could lead to the establishment of unrealistic care standards
• Again, same as for VV ECMO, senior staff need to be trained, multiple nurses need to be available, mobility is restricted except for where the entire department is