ICU · Respiratory
Acute respiratory distress syndrome: phenotyping and precision medicine
Also known as ARDS phenotypes · Hyperinflammatory ARDS · Hypoinflammatory ARDS · Hypoimmune ARDS · Subphenotypes of ARDS · Precision medicine in ARDS · Berlin definition ARDS · ARDS biomarkers
ARDS is HETEROGENEOUS — not one disease. TWO reproducible subphenotypes identified by latent class analysis in ARDS Network trials and HARP-2 (Calfee 2014, Famous 2017, Calfee 2018; stability Delucchi 2018): HYPERINFLAMMATORY (~35% in HARP-2): high inflammatory markers (IL-6, IL-8, sTNFr-1, sRAGE, SP-D, angiopoietin-2), low PaO2/FiO2, acidosis, vasopressor-requiring, resembles sepsis, worse prognosis (28-day mortality 39% vs 17% in HARP-2; 90-day mortality 57% vs 33% in LUNG SAFE re-analysis), higher PEEP associated with lower 90-day mortality (Maddali 2022: 54% vs 62%), improved 28-day survival with simvastatin in the HARP-2 re-analysis (Calfee 2018, p=0.008), BUT a conservative fluid strategy INCREASED 90-day mortality in this subphenotype in FACTT (Famous 2017: 50% vs 40% with liberal). HYPOINFLAMMATORY (~65%): lower inflammation, better oxygenation, better prognosis (28-day mortality 17% in HARP-2), conservative fluid strategy lowered 90-day mortality here (18% vs 26%). The Berlin definition (2012) standardised diagnosis (timing within 1 week, bilateral opacities not fully explained by effusion/atelectasis/nodules, non-cardiogenic oedema, severity by PaO2/FiO2: mild 200-300, moderate 100-200, severe <100 with PEEP/CPAP =5) but is purely DESCRIPTIVE — it does NOT capture biological heterogeneity. The LUNG SAFE multinational study (Bellani 2016, 459 ICUs, 29144 patients) showed ARDS is under-recognised (clinical recognition only 51.3% of mild and 78.5% of severe cases) and undertreated (tidal volume 8 mL/kg predicted body weight or less in under two-thirds; prone positioning in 16.3% of severe ARDS). Calfee 2014 Lancet Respir Med applied latent class analysis to ARMA + ALVEOLI (1022 patients) and found 2 subphenotypes with differential PEEP response. Maddali/Pham/Calfee 2022 Lancet Respir Med validated ML-derived classifiers in EARLI and VALID and applied them to LUNG SAFE. This explains why ARDS pharmacological trials (statins, beta-agonists, NO, KGF) showed heterogeneous/neutral results — subphenotype determines treatment response. Precision medicine: stratify patients, tailor treatment.
Practise this topic
On this page
Study tools
Your progress
Saved on this device.
Target exams
Red flags
- Hyperinflammatory ARDS — worse prognosis (28-day mortality 39% vs 17% in HARP-2); higher PEEP associated with lower 90-day mortality (Maddali 2022); simvastatin improved 28-day survival (HARP-2 re-analysis, p=0.008); conservative fluid strategy RAISED 90-day mortality here (Famous 2017: 50% vs 40%)
- Hypoinflammatory ARDS — better prognosis; no mortality benefit from higher PEEP in the LUNG SAFE re-analysis (34% vs 32% 90-day mortality); avoid overdistension
- This explains heterogeneous trial results — subphenotype determines treatment response
- Berlin definition is DESCRIPTIVE only — it does NOT capture biological heterogeneity or predict treatment response
- Subphenotype may SHIFT during illness — reassess if new infection or worsening organ failure develops
- Rising sRAGE and angiopoietin-2 mark alveolar epithelial injury and endothelial dysfunction — hyperinflammatory phenotype
- LUNG SAFE: ARDS is under-recognised (clinically recognised in only 51.3% of mild and 78.5% of severe cases) and lung-protective ventilation is underused (tidal volume 8 mL/kg PBW or less in under two-thirds of patients)
Meet the patient
A 52-year-old man is 36 hours into ICU for pneumococcal pneumonia and septic shock. He is intubated for severe ARDS with PaO2/FiO2 88 on FiO2 0.85 and PEEP 12, on noradrenaline 0.4 mcg/kg/min, lactate 4.2 mmol/L, pH 7.24, bicarbonate 18 mmol/L, platelets 95 (x10^9/L), CRP 290 mg/L. The consultant asks: is this patient the same biological disease as the 28-year-old asthmatic with aspiration pneumonitis two beds down, and does the answer change what we do with the ventilator this morning?[7][2]
The two patients share a Berlin label and an X-ray — but their biology is different, and their response to PEEP, fluids, and a statin trial is different. The bedside question this chapter answers is exactly that.[2][5]
The mantra: Berlin labels severity, biology labels treatment — phenotype first, PEEP second, pharmacology third.[2][1]
Two reproducible phenotypes — the face-off. Same Berlin label, different biology, different PEEP and fluid response. The mnemonic is "sepsis vs pneumonia" — hyperinflammatory looks like systemic inflammation with a lung complication; hypoinflammatory looks like a primary lung injury with quiet biology elsewhere.[2][5]
| Feature | Hyperinflammatory (Type 1) | Hypoinflammatory (Type 2) |
|---|---|---|
| Prevalence | ~35% (HARP-2); 725 of 2813 in LUNG SAFE | ~65% (HARP-2) |
| Driver biology | Sepsis-like, systemic inflammation, endothelial leak | Primary lung injury, less systemic spillover |
| Inflammation panel | HIGH — IL-6, IL-8, sTNFr-1, sRAGE, SP-D, angiopoietin-2 | LOW — near-normal |
| PaO2/FiO2 | Lower (worse oxygenation) | Higher (better oxygenation) |
| Platelets / protein C | Low / low | Normal / normal |
| Bicarbonate | Low (metabolic acidosis) | Normal |
| Vasopressors | Usually required | Often not |
| Organ failure | Multi-organ, sepsis-like | Lung-predominant |
| Mortality | HIGHER — 28-day 39% (HARP-2); 90-day 57% (LUNG SAFE) | LOWER — 28-day 17% (HARP-2); 90-day 33% (LUNG SAFE) |
| Higher PEEP | Lower 90-day mortality (54% vs 62%, Maddali 2022) | No mortality benefit (34% vs 32%) |
| Conservative fluids | HARM — 90-day mortality 50% vs 40% (Famous 2017) | BENEFIT — 90-day mortality 18% vs 26% (Famous 2017) |
| Simvastatin | Higher 28-day survival (p=0.008, HARP-2 re-analysis) | No differential benefit shown |
| Clinical mirror | Septic ARDS, pancreatitis, transfusion, trauma | Pneumonia, aspiration, contusion |
References13ShowHide
- [1]Famous KR, Delucchi KL, Ware LB, et al. Acute Respiratory Distress Syndrome Subphenotypes Respond Differently to Randomized Fluid Management Strategy Am J Respir Crit Care Med, 2017.PMID 27513822
- [2]Calfee CS, Delucchi K, Parsons PE, Thompson BT, Ware LB, Matthay MA. Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from two randomised controlled trials Lancet Respir Med, 2014.PMID 24853585
- [3]Calfee CS, Delucchi KL, Sinha P, et al. Acute respiratory distress syndrome subphenotypes and differential response to simvastatin: secondary analysis of a randomised controlled trial Lancet Respir Med, 2018.PMID 30078618
- [4]McAuley DF, Laffey JG, O'Kane CM, et al. Simvastatin in the acute respiratory distress syndrome N Engl J Med, 2014.PMID 25268516
- [5]Maddali MV, Churpek M, Pham T, Rezoagli E, Zhuo H, Zhao W, He J, Calfee CS, Laffey JG, Bellani G, Laffey JG; LUNG SAFE Investigators. Validation and utility of ARDS subphenotypes identified by machine-learning models using clinical data: an observational, multicohort, retrospective analysis Lancet Respir Med, 2022.PMID 35026177
- [6]Delucchi K, Famous KR, Ware LB, et al. Stability of ARDS subphenotypes over time in two randomised controlled trials Thorax, 2018.PMID 29477989
- [7]Bellani G, Laffey JG, Pham T, et al. Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries JAMA, 2016.PMID 26903337
- [8]Wiedemann HP, Wheeler AP, Bernard GR, et al. Comparison of two fluid-management strategies in acute lung injury N Engl J Med, 2006.PMID 16714767
- [9]Brower RG, Lanken PN, MacIntyre N, et al. Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome N Engl J Med, 2004.PMID 15269312
- [10]Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome N Engl J Med, 2013.PMID 23688302
- [11]Ware LB, Koyama T, Billheimer DD, et al. Prognostic and pathogenetic value of combining clinical and biochemical indices in patients with acute lung injury Chest, 2010.PMID 19858233
- [12]Jabaudon M, Blondonnet R, Pereira B, et al. Plasma sRAGE is independently associated with increased mortality in ARDS: a meta-analysis of individual patient data Intensive Care Med, 2018.PMID 30051136
- [13]Combes A, Hajage D, Capellier G, et al. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome N Engl J Med, 2018.PMID 29791822