Oct 162018

Jon-Emile S. Kenny MD [@heart_lung]

Once had a love and it was a gas …”



A 56 year old professor returns from a hiking trip in the ‘Four Corners’ area of New Mexico.  She was previously well and decided to rent a secluded desert cabin whilst writing a novel on the ethical obligations we have for one another.  On arrival she swept up dust around the kitchen pantry; once fully-mopped she spent a productive and morally-reaffirming fortnight alone. 

A few days after returning home she is plagued by 72 hours of malaise followed by abrupt fevers, chills and cough.  She finds herself extremely short-of-breath and drives herself to a walk-in clinic where she is noted to have hypoxemia and bilateral crackles on physical examination.  She is transported via ambulance to the closest ED where she rapidly deteriorates and requires intubation; her PaO2:FiO2 ratio is less than 50 despite a PEEP of 10 cmH2O and a FiO2 of 1.0.  The patient is transitioned to APRV by the consulting intensivist.  Initial bloodwork returns significant thrombocytopenia, leukocytosis with 25% ‘immunoblastic lymphocytes’ and polycythemia.  The ED attending is startled by the bloodwork and she immediately calls the intensivist to arrange transfer to a center with ECMO.


In 2016 a young woman lived without lungs at Toronto General Hospital for 6 days; she underwent bilateral pneumectomy because of complications of Cystic Fibrosis.  Then – at the institution of the first successful lung transplantation – she received two new lungs.

Given the above, one may wonder if – whilst on extra-corporeal membrane oxygenation [ECMO] – mechanical ventilation is needed at all.  Without getting into specifics of veno-arterial ECMO, veno-venous ECMO and extra-corporeal carbon dioxide removal [ECCO2R], the idea of using extracorporeal help for lungs experiencing the heartbreak of ARDS is not new.

Indeed, studies in the 1970s evaluated this idea; unfortunately, this was a generation typified not only by hard-hitting music and fashion but also by uncompromising mechanical ventilation.  In fact, this was a time when prophylactic chest tubes were suggested for evolving ARDS because everyone, essentially, developed pneumothoraces on the ventilator.  Soberingly, the mortality rates were 90 – 93% and extracorporeal help did not seem to afford this highly mortal syndrome much benefit.

But in 1986 Gattinoni and colleagues described the use of extra-corporeal carbon dioxide removal which yielded gentler mechanical ventilation in severe ARDS; in this uncontrolled trial, peak airway pressure was limited to ’35-45 cm H2O’ and low respiratory rates [3-5 breaths per minute] were employed.  With this, the mortality rate fell to about 50% from the expected 90%.

Thus, even 30 years ago, the idea of ‘lung rest’ was evolving.  Unfortunately, today the approach to mechanical ventilation during extra-corporeal support remains murky and has been described as a ‘necessary evil.’

With this, a recent report in the Blue Journal sought to relate different ‘doses’ of mechanical ventilation during veno-venous ECMO to markers of lung injury in a porcine model.

What They Did

24 healthy pigs were studied, 6 of which were used as a ‘sham’ group and were ventilated with ‘non-protective’ ventilation [i.e. respiratory rate of 16-18 and tidal volume of 10 mL/kg].  The other 18 had severe lung injury induced by intra-tracheal warm saline lavages to wash out pulmonary surfactant and two hours of injurious mechanical ventilation.  These animals were then randomized to 3 groups: non-protective ventilation as above, conventional protective ventilation [volume control ventilation at 6 mL/kg, PEEP 10 cmH2O and respiratory rate of 20 breaths per minute] and near-apneic ventilation [pressure control ventilation with PEEP of 10 cm H2O, driving pressure of 10 cm H2O – targeting a tidal volume of at most 2 mL/kg – and a respiratory rate of 5 breaths per minute].  Each of the three groups received VV-ECMO.

Following 24 hours of ventilation, the animals were sacrificed and their lungs studied for weight, histologic changes and gene expression associated with fibroproliferative ARDS.

What They Found

While the results were somewhat variable, there was a general trend towards reduced histological damage with decreasing ‘dose’ of mechanical ventilation.  Even conventional protective ventilation induced some degree of histological VILI.  There was poor relationship with wet-to-dry ratio and means of mechanical ventilation; there was linear reduction of both myofibroblast and MMP-2 activities with smaller tidal volumes.  Lastly, when considering each injured animal individually, there was a positive correlation of histological injury, myofibroblast and pro-collagen III scores with both driving pressure and mechanical power applied to the lung.

Of note, mechanical power ranged from 11-13 J/min in the non-protective group, 7-8 J/min in the conventional protective group, and 0.4-0.5 J/min in the near-apneic group.  Driving pressures were 21- 24 cmH2O, 14-15 cmH2O and 9 to 10 cmH2O, respectively.


Overall this interesting physiological study showed a variable but definite trend towards less overall histological lung injury with less mechanical power and driving pressure applied to the lung.  Thus, it may be that there is no ‘safe’ tidal volume or trans-pulmonary pressure that can be applied to the lung in severe ARDS.  This isn’t surprising given the variability of the functional ‘baby lung’ [determinant of lung strain] as well as stress raisers which can locally determine the trans-pulmonary pressure [i.e. the lung stress].

To the credit of the authors, they evaluated the relationship between both the driving pressure and mechanical power applied to the lung and the severity of lung injury.  Importantly, however, the authors did not measure transpulmonary pressure [Ptp] which better approximates actual lung stress [i.e. the pressure across the lung].  Using Ptp, a threshold level of 12 J/min was found to result in ventilator-induced lung injury [VILI] in normal porcine lungs.  Because Ptp was not used in the evaluation by Araos and colleagues, their reported mechanical power are likely overestimates of the actual power applied to the lung skeleton.  In severe ARDS the lung elastance-to-total respiratory system elastance has been reported to be as high as 0.8 [normally, this value is around 0.5].  In other words, 80% of the static plateau pressure is ‘felt’ across the lungs.  As an example, in a 2015 porcine study using a warm saline lavage model of ARDS, an airway driving pressure of 18 cm H2O resulted in a trans-pulmonary driving pressure of only 14 cmH2O.  It’s possible, therefore, that the power reported in the study at hand may be over-estimating the power applied to the lung by ~ 20%.

Further, because the threshold of 12 J/min was reported in normal porcine lungs, it may be that much less mechanical power is tolerated in pre-existing ARDS.  The putative mechanism being that lungs with ARDS have more stress raisers – which may augment local distending pressures by a factor of 4.  Because of this, it would have been interesting to know how prone positioning – which has been shown to diminish stress raisers – interacts with mechanical ventilation whilst on ECMO.

Below is a video describing the basics of stress raisers and prone position - it is the final lecture in a series that I gave last spring in Montreal.

How all of the aforementioned relates to human beings is not entirely clear.  Human lungs have a specific elastance roughly twice that of porcine lungs, thus the power thresholds cannot be wholly compared.  Nevertheless, there may be a direct relationship between driving pressure and/or mechanical power applied to the lungs during severe ARDS with no clear level of ‘safety.’  While still unproven, it may be that near-apnea in severe ARDS offers the lung time to heal – mitigating the effects of VILI and even lung injury associated with spontaneous breathing on ECMO!  Calls to close the lungs have been made and while the Siren Call for extra-corporeal assistance may be technologically-seductive, ICU mariners may be well-served if they first navigate toward the calm waters of the prone position, especially in light of the recent- and controversial - EOLIA.

Return to Case

The patient is urgently moved to a nearby university hospital via air ambulance – epinephrine is initiated to maintain her blood pressure en route; she is met in the cardiothoracic ICU by an ECMO team and is immediately placed veno-arterial ECMO given strong suspicion for Hantavirus Pulmonary Syndrome.  She is also started on ‘lung protective’ ventilation strategies with a tidal volume of 5 mL/kg of ideal body weight.



Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he also the creator and author of a free hemodynamic curriculum at heart-lung.org

Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


Near Apneic Ventilation & Extracorporeal Membrane Oxygenation: close the lungs & keep them closed?