LAUSR

Critical Care Medicine www.ccmjournal.org 1017 Acute respiratory distress syndrome (ARDS) is the clinical manifestation of inflammatory (i.e., noncardiogenic) pulmonary edema and may result from various insults, either pulmonary or extra pulmonary (1). The diagnosis of ARDS relies on the acute onset (≤ 1 wk) of bilateral chest infiltrates in addition to a variable degree of hypoxemia (Pao 2 /Fio 2 ), the latter being proportional to the severity of the disease (2). Invasive mechanical ventilation is required in approximatively 85% of patients with a median duration of 8 days (interquartile range [IQR], 4–15 d) (3). ARDS patients who do not improve after 5–7 days of mechanical ventilation (i.e., persistence or worsening of hypoxemia, lung compliance, or infiltrates) are categorized as persistent or nonresolving ARDS. These patients, although representing less than 5% of all ARDS, are likely to be diagnosed with atypical infectious disease noncovered by the initial antimicrobial regimen, intense lung fibroproliferation, or other interstitial lung diseases (4). Bronchoalveolar lavage (BAL) fluid analysis constitutes the first step of the diagnosis work-up (5). BAL cytology may be representative of specific lung disease, such as acute eosinophilic pneumonia, or may also indicate the presence of abnormal cells (e.g., malignant) or alveolar hemorrhage. BAL also provides material for extensive microbiological investigation, especially using wide panel of quantitative polymerase chain reaction (qPCR) for the detection of viruses, atypical bacteria, mycobacterium tuberculosis, and fungi. Other “first-line” analyses include serum autoantibodies and various serology or antigenemia. If not yet available, a high-resolution lung CT scan may also provide signs toward usual interstitial pneumonia (4). Once these steps reveal inconclusive, some centers advocate for the use of surgical open lung biopsy (OLB), mainly performed at the bedside (6). Diffuse alveolar damage (DAD) is a histopathologic pattern including capillary congestion, atelectasis, alveolar hemorrhage or edema, and hyaline membrane deposition. Such autopsy findings have been observed in most of ARDS patients at the time of the original description of the disease, linking ARDS to DAD (7). However, more recent literature consistently reported that only half of ARDS patients exhibit DAD either at autopsy or during the course of the disease (using OLB) (8, 9). Hence, DAD is neither predictive nor necessary for the diagnosis of ARDS, and the presence of DAD may also reflect alternative diagnosis such as usual interstitial pneumonia, if the clinical scenario diverges from ARDS (4). Nevertheless, ARDS patients with DAD have a more severe form of the disease, characterized by worse oxygenation, lung compliance, and higher mortality as compared to those without DAD (9). Besides DAD, other pathologic findings include bacterial or viruses pneumonia (frequently involving “herpesviridae”), fibrosis (more frequent after ≥ 7 d), diffuse alveolar hemorrhage, cryptogenic organizing pneumonia, bronchiolitis, malignancies, and drug toxicity (6). Therefore, OLB performed in nonresolving ARDS patients should be seen as a tool to identify a specific lung disease, mimicking ARDS in its initial presentation, but without responding to initial therapy. Among the numerous studies that report on OLB during nonresolving ARDS, most of them highlighted that changes in the current medications occurred in approximatively half of the patients, consisting mainly in initiating corticosteroids (for fibrosis or interstitial disease) or antiviral therapy (for cytomegalovirus pneumonia) or stopping antibiotics (6, 10). As above discussed, the performance of qPCR analysis for pathogens detection in BAL allows clinicians to initiate earlier any appropriate therapy. Thus, most of the “pharmacologic” interventions following OLB are limited to corticosteroids. Unfortunately, not all lung diseases are sensitive to corticosteroids; however, knowing the disease to predict the corticosteroid sensibility makes sense. In this issue of Critical Care Medicine, Gerard et al (11) reported a retrospective case series of 51 ARDS patients for whom an OLB have been performed during the course of their disease. Those patients represent 7% of all ARDS admitted in the ICU among the 10 years period of observation. Indications for OLB were either nonresolving ARDS or suspicion of alternative diagnosis, and the median delay from ARDS onset was 7 days (IQR, 2–13 d). The population has been separated, according to histopathologic findings, into patients with pathologies supposed to be reverse by corticosteroid therapy versus others which are not supposed to be sensitive to steroids. Such separation has been performed retrospectively on the basis of literature findings and pathologists consensus. DAD, considered as a corticosteroid-resistant pattern, was observed in 43% of patients. Overall, a corticosteroid-sensitive pathology was identified in 37% of patients, and all of them received a corticosteroid therapy (prior the OLB for two patients) with a mean duration of 25 days. Conversely, 63% of patients had corticosteroid-resistant pathologies, and 58% of them received corticosteroid therapy for a mean duration of 9 days, awaiting OLB results. The in-hospital and 180-day mortality rates of the *See also p. 907.