ABSTRACT
OBJECTIVE: To demonstrate that among patients with acute respiratory distress syndrome (ARDS), the presence of diffuse alveolar damage (DAD) at histological examination, as compared to its absence, defines a specific subphenotype. METHODS: We studied 149 patients who died in our ICU with the clinical diagnosis of ARDS according to the Berlin Definition (BD) and who had autopsy examination. We compared the change over time of different clinical variables in patients with (n = 49) and without (n = 100) DAD. A predictive model for the presence of DAD was developed and validated in an independent cohort of 57 patients with ARDS and postmortem examination (21 of them with DAD). RESULTS: Patients with DAD, as compared to patients without DAD, had a lower PaO2/FiO2 ratio and dynamic respiratory system compliance, and a higher SOFA score and INR, and were more likely to die of hypoxemia and less likely to die of shock. In multivariate analysis, variables associated with DAD [odds ratio, 95% confidence interval (CI)] were PaO2/FiO2 ratio [0.988 (0.981-0.995)], dynamic respiratory system compliance [0.937 (0.892-0.984)] and age [0.972 (0.946-0.999)]. Areas under the ROC curve (95 % CI) for the classification of DAD using the regression model or the BD were, respectively, 0.74 (0.65-0.82) and 0.64 (0.55-0.72) (p = 0.03). In the validation cohort, the areas under the ROC curve for the diagnosis of DAD were 0.73 (0.56-0.90) and 0.67 (0.54-0.81) for the regression model and the BD, respectively. CONCLUSIONS: The presence of DAD appears to define a specific subphenotype in patients with ARDS. Targeting patients with DAD within the population of patients with the clinical diagnosis of ARDS might be appropriate to find effective therapies for this condition.