Aug 302017

Jon-Emile S. Kenny MD [@heart_lung] with illustrations by Carla M. Canepa MD

“Our life consists partly in madness, partly in wisdom: whoever writes about it merely respectfully and by rule leaves more than half of it behind.”


Marking the 50 year anniversary of the first description of the adult respiratory distress syndrome – later renamed the acute respiratory distress syndrome [ARDS] – both the New England Journal of Medicine and Annals of Translational Medicine have contributed terrific reviews on this prevalent and deadly condition.  For additional reading, please refer to previous posts on the physiology of mechanical power, stress raisers and the pulmonary vasculature within the context of ARDS here at pulmccm.

The molecular underpinnings of ARDS are continuously clarified and may be appreciated when considering the different phases of ARDS: exudative, proliferative and – sometimes – fibrotic.  Yet, for the busy clinician, the value of knowing the spilling of cells and cytokines may feel far-flung; the ‘wisdom’ of the molecular mechanisms being divorced from the ‘madness’ of clinical practice.  May there be matrimony between the respectful rules of biology and the rumbling realities at the bedside?

Recently, re-analysis of the ALVEOLI, ARMA and FACTT trials have uncovered two sub-phenotypes of ARDS which carry management implications.  These findings are especially important for both the application of PEEP and the provision of intravenous fluids and will be touched upon below.

The Phases

The initial, or exudative phase, illustrated below in figure 1, begins with an insult which may arise either directly to the alveolar epithelium [e.g. pneumonia or gastric acid] or indirectly to the pulmonary capillary endothelium [e.g. pancreatitis, extra-pulmonary sepsis].  In either case, alveolar macrophages sound the inflammatory alarm with recruitment of additional monocytes/macrophages and polymorphs.  Destruction of both epithelial and endothelial surfaces progress and there is alveolar flooding with protein-rich fluid and inflammatory cells.  Importantly, markers of alveolar epithelial injury [e.g. RAGE, SP-D] and capillary endothelial injury [e.g. VEGF, von Willebrand factor and, Ang-II] are elaborated and depicted in the cartoon.

Figure 1: Exudative phase: M1 alveolar macrophages are enraged by direct and/or indirect insults, inflammatory cascades are triggered. PMNs [polymorphs] arrive within the alveolus and there is breakdown of the epithelial and endothelial barriers with interstitial and alveolar edema. Markers of the endothelium are indicated as: VEGF is vascular endothelial growth factor, Ang-2 is angiopoietin 2; vWF is von Willebrand Factor. Markers of the epithelium are indicated as: RAGE is receptor of advanced glycation end-product; SP-D is surfactant protein D

As the insult abates, reparative processes are required for survival.  As seen in figure 2, inflammatory M1 macrophages morph into M2 macrophages which help along alveolar clean-up.  These M2 cells assist in the destruction of alveolar polymorphs; as well, the M2 macrophages send signals to alveolar epithelial type II cells [AEC] which stimulate fibrin matrix scaffolds, regenerate alveolar type I cells and synthesize surfactant.  As AEC type I cells reconstitute the alveolar epithelium, tight junctions between them are re-established and protein channels [e.g. CFTR, Na+/K+ ATPase] help remove alveolar fluid.

Figure 2: Proliferative phase: M2 alveolar macrophages aide in clean-up. They endocytose polymorphs and trigger AEC cells to regenerate the alveolar epithelium. Tight-junctions form and alveolar edema is drained via the lymphatics. AEC cells begin to elaborate surfactant; as well, they stimulate an interstitial scaffold upon which endothelial and epithelial repair occur; markers of endothelial damage return home; AEC is alveolar epithelial cell. VEGF is vascular endothelial growth factor, Ang-2 is angiopoietin 2; vWF is von Willebrand Factor.

The final, fibrotic phase, is not a universal outcome in ARDS, but is linked with increased mortality and prolonged mechanical ventilation.  Whether or not the fibrotic stage is reached may be related to the degree of damage of the basement membrane and an overzealous assembly of pro-fibrotic mediators.

Pulmonary versus Extra-pulmonary ARDS

Wherein lies the utility of knowing the aforementioned biochemical orchestrations?  The pathways above may provide a rich scaffold upon which the intensivist can construct an understanding for different ARDS phenotypes.  One clinical approach has been to parse the inciting events of ARDS.  For example, direct, airway-centric insults such as pneumonia and gastric acid pneumonitis may have slightly different management implications as compared to indirect, extra-pulmonary insults such as systemic sepsis, pancreatitis, etc.

Indeed, the pathology, radiology and mechanics of direct versus indirect ARDS have been shown to be relatively distinct.  For example, lung biopsies from patients with direct causes of ARDS tend to reveal more alveolar collapse with thick, discontinuous hyaline membranes.  By contrast, indirect insults disclose more interstitial edema and vascular congestion.  Similarly, patients deemed to have direct insults causing ARDS were noted to have asymmetrical consolidation with similar fractions of ground glass to consolidation pattern on chest CT; by contrast, indirect ARDS was typified by symmetrical, bilateral and centrally-located ground glass.

The mechanics of pulmonary [or direct] versus extrapulmonary [or indirect] ARDS were first described in a classic investigation by Gattinoni and colleagues.  In this study, pulmonary ARDS was exemplified by poor recruitment in response to PEEP and increased lung elastance [i.e. decreased compliance].  By contrast, extra-pulmonary ARDS patients were noted have an increased chest wall elastance [i.e. decreased chest wall compliance] which correlated with intra-abdominal pressure.  Extra-pulmonary ARDS patients were responsive to recruitment by PEEP.

Implications for Management – Fluids & PEEP

Notably, re-analysis of the original ARMA trial detected two sub-phenotypes of ARDS based on serum biomarkers.  In this analysis, ‘sub-phenotype 2’ displayed biochemical evidence of more intense systemic inflammation including very high levels of IL-6; these patients were more likely to require vasopressors, had higher mortality and were more likely to have ‘sepsis-associated’ ARDS.  While it was not clear whether the origin of sepsis in sub-phenotype 2 was pulmonary or extra-pulmonary, previous analyses have shown sepsis-associated ARDS to be roughly 50% of pulmonary versus extra-pulmonary in origin.

Subsequently, when the above sub-phenotypes were used to – retrospectively – predict response to PEEP in the ALVEOLI trial, those with the hyper-inflammatory sub-phenotype 2 randomized to high PEEP had a 9% absolute risk reduction in mortality.  Additionally, an ensuing investigation – where patients in the FACTT trial were retrospectively parsed into sub-phenotype 1 & 2 based on serum biomarkers – showed that those with the hyper-inflammatory sub-phenotype 2 had reduced mortality when randomized to a conservative fluid strategy.

Given the above, it is tempting to speculate as to whether the hyper-inflammatory, sub-phenotype 2 is akin to indirect, or extra-pulmonary ARDS.  Why?  As described, indirect ARDS – like sub-phenotype 2 –  is typified by favourable mechanical response to PEEP.  Further, because the histology of indirect or systemic ARDS is primarily that of interstitial edema, it is plausible that a fluid conservative approach to resuscitation in such patients would be advantageous – as observed in the sub-phenotype 2 population.  Lastly, that indirect ARDS is, by definition, a systemic insult, cytokine expression from a distant source may be expected to incite a sweeping, sub-phenotype 2 physiology.

Consider that IL-6 was much higher in the hyper-inflammatory, sub-phenotype 2 ARDS cohort.  How does this relate to direct versus indirect pulmonary insults?  In a murine model of ARDS, direct airway injury resulted in serum IL-6 levels of, on average, just over 1000 pg/mL whereas in indirect ARDS [i.e. from intra-peritoneal endotoxin], IL-6 levels were well-above 5000 pg/mL.   Further, in a separate analysis of both the VALID and FACTT trials, indirect ARDS had a significantly higher serum IL-6 than direct ARDS; numerically, however, the difference was not as pronounced as the murine model.

Given the cartoons above, one would also expect RAGE and SP-D levels to be higher in direct ARDS because they are markers of pulmonary epithelium; by extension, these markers might be greater in the less-inflammatory sub-phenotype 1.  Conversely, it is expected that Ang-II and vWF would be higher in indirect ARDS, and therefore greater in the hyper-inflammatory, fluid-sensitive, PEEP-responsive sub-phenotype 2.  Unexpectedly, serum RAGE was higher in sub-phenotype 2, and could not discriminate between direct and indirect ARDS which may be due to the role of RAGE in the innate immune system.  As expected, however, SP-D was higher in both less-inflammatory sub-phenotype 1 and in direct ARDS.  Further, markers of endothelial injury, Ang-II and vWF were higher in both indirect ARDS and in the hyper-inflammatory sub-phenotype 2 population.

Despite a seeming connection between the direct versus indirect and sub-phenotype 1 versus sub-phenotype 2 dichotomies, it is not perfect.  For example, if indirect ARDS was completely akin to sub-phenotype 2, one would expect traumatic causes of ARDS to over-represent the hyper-inflammatory sub-group – that was not observed.  Obviously, clinical medicine is messy and at least 1 in 5 ARDS patients have ‘mixed’ etiologies.  For example, direct injury to one lung from a bad pneumonia may have elements of indirect injury in the contra-lateral lung.  Genetic diathesis almost certainly plays some role in the complex evolution of ARDS.

Nevertheless, there may some linkage between the cytokines and biomarkers elaborated in ARDS and its underlying, clinical etiology; further investigation can better clarify this connection.  Until then, it may be prudent to rethink an extra fluid challenge in the inflamed, extra-pulmonary ARDS patient on vasopressors.  It is also expected that such a patient may better respond to increased PEEP and, potentially, pronation.  Use of the stress-index and driving pressure may help with said titration.

Happy Anniversary Mum & Dad,


Get our weekly email update, and explore our library of practice updatesreview articles. and board review questions.

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

Authors: contribute your work in a guest post.


An Illustrated Guide to the Phases of ARDS: Implications for management