(466) Host Response Subphenotypes Do Not Differ in Response to Trophic Versus Full Enteral Nutrition

Introduction: Hyperinflammatory and hypoinflammatory host response subphenotypes in acute respiratory distress syndrome (ARDS) are associated with differences in prognosis and may differ in response to therapy. The Early Versus Delayed Enteral Feeding to Treat People with Acute Lung Injury (EDEN) trial randomized participants with ARDS to trophic and full dose enteral nutrition arms and found a higher incidence of gastrointestinal (GI) complications with full dose enteral nutrition and no between-group differences in mortality. Heterogeneity of treatment effect (HTE) to enteral nutrition dose strategies by host response subphenotype has not been examined.

Methods: We performed a secondary analysis of 880 EDEN trial participants with available baseline blood biospecimens. We measured angiopoietin-2, procalcitonin, and soluble tumor necrosis factor receptor (sTNFR1) by Luminex assay. We determined subphenotype membership using a previously published 4-variable parsimonious logistic regression classification model (angiopoietin-2, procalcitonin, sTNFR1, and bicarbonate). Differences in baseline characteristics and clinical outcomes (GI complications and 60-day mortality) between subphenotypes were assessed by Mann-Whitney U or chi squared analysis. Interactions between subphenotype and treatment arm for clinical outcomes were assessed by logistic regression. Data are reported as median [interquartile range] or n (%) as appropriate.

Results: 530 participants (60%) were classified as hypoinflammatory and 350 (40%) as hyperinflammatory. Hyperinflammatory participants were older (54 [43, 66] versus 51 [41, 61] years, p=0.003) and had a higher severity of illness (APACHE 104 [86, 122] versus 78 [64, 98], p< 0.001). Hyperinflammatory participants had a higher incidence of GI complications (54% versus 46%, p=0.020) and a higher 60-day mortality (30% versus 18%, p< 0.001). No significant interaction was observed between subphenotype and feeding strategy for GI complications (odds ratio [OR] 0.95, 95% CI [0.55,1.64], p=0.851) or 60-day mortality (OR 0.78 [0.41,1.46], p=0.431).

Conclusions: Host response subphenotypes did not differ in response to trophic versus enteral nutrition. Hyperinflammatory patients had a higher incidence of GI complications irrespective of treatment arm.