RESUMO
Background: Acute Respiratory Distress Syndrome (ARDS) is a devastating pulmonary inflammatory disorder, commonly precipitated by sepsis. Glucocorticoids are immunomodulatory steroids that can suppress inflammation. Their anti-inflammatory properties within tissues are influenced by their pre-receptor metabolism and amplification from inactive precursors by 11ß-hydroxysteroid dehydrogenase type-1 (HSD-1). We hypothesised that in sepsis-related ARDS, alveolar macrophage (AM) HSD-1 activity and glucocorticoid activation are impaired, and associated with greater inflammatory injury and worse outcomes. Methods: We analysed broncho-alveolar lavage (BAL) and circulating glucocorticoid levels, AM HSD-1 reductase activity and Receptor for Advanced Glycation End-products (RAGE) levels in two cohorts of critically ill sepsis patients, with and without ARDS. AM HSD-1 reductase activity was also measured in lobectomy patients. We assessed inflammatory injury parameters in models of lung injury and sepsis in HSD-1 knockout (KO) and wild type (WT) mice. Results: No difference in serum and BAL cortisol: cortisone ratios are shown between sepsis patients with and without ARDS. Across all sepsis patients, there is no association between BAL cortisol: cortisone ratio and 30-day mortality. However, AM HSD-1 reductase activity is impaired in patients with sepsis-related ARDS, compared to sepsis patients without ARDS and lobectomy patients (0.075 v 0.882 v 0.967 pM/hr/106 AMs, p=0.004). Across all sepsis patients (with and without ARDS), impaired AM HSD-1 reductase activity is associated with defective efferocytosis (r=0.804, p=0.008) and increased 30-day mortality. AM HSD-1 reductase activity negatively correlates with BAL RAGE in sepsis patients with ARDS (r=-0.427, p=0.017). Following intra-tracheal lipopolysaccharide (IT-LPS) injury, HSD-1 KO mice demonstrate increased alveolar neutrophil infiltration, apoptotic neutrophil accumulation, alveolar protein permeability and BAL RAGE concentrations compared to WT mice. Caecal Ligation and Puncture (CLP) injury in HSD-1 KO mice results in greater peritoneal apoptotic neutrophil accumulation compared to WT mice. Conclusions: AM HSD-1 reductase activity does not shape total BAL and serum cortisol: cortisone ratios, however impaired HSD-1 autocrine signalling renders AMs insensitive to the anti-inflammatory effects of local glucocorticoids. This contributes to the decreased efferocytosis, increased BAL RAGE concentrations and mortality seen in sepsis-related ARDS. Upregulation of alveolar HSD-1 activity could restore AM function and improve clinical outcomes in these patients.