A33-19 Characterizing Peripheral Inflammatory Patterns After Inhalation Injury and Their Relationship to Pulmonary Infection

Abstract Rationale Burn-related inhalation injury increases the risk of pulmonary infection via epithelial damage, causing bacterial translocation as well as mucociliary clearance disruption and bronchial plugging. Prior studies demonstrated increased alveolar leukocyte infiltration, but systemic inflammatory patterns have been less extensively explored and largely focus on neutrophilic involvement. Recent research, however, has revealed eosinophilic involvement in burn-related inflammation. We aimed to determine whether admission eosinophil and neutrophil indices were associated with inhalation injury and to assess their relationships with subsequent pulmonary infections. Methods Retrospective analysis was conducted on a cohort of 960 burn-injured patients admitted to a burn center. Associations with pulmonary infection, defined by positive respiratory culture, were modeled using logistic regression and prediction model discrimination was completed with AUC analyses (DeLong test). Multivariable logistic regression was used to model the associations between inhalation injury with eosinophil and neutrophil indices, adjusting for patient variables such as total body surface area burned (TBSA) and sepsis, defined by a documented diagnosis or blood or respiratory culture. Results Inhalation injury was positively associated with the development of a pulmonary infection (aOR 3.5, p 0.001), independent of TBSA. Admission neutrophil-to-lymphocyte ratio (NLR) was modestly associated with pulmonary infection (OR 1.02 per unit, 95% CI 1.005-1.040, p = 0.011) and TBSA (OR 1.1, 95% CI 1.04-1.16, p 0.001) but demonstrated no relationship with inhalation injury (OR 0.97, 95% CI 0.94-10.1, p = 0.249). In contrast, higher eosinophil counts were associated with the presence of inhalation injury: absolute eosinophil count (AEC, OR 1.3 per 0.1x109/L, 95% CI 1.04-1.52, p = 0.02) and eosinophil-to-lymphocyte ratio (ELR, OR 2.7 per 0.1, 95% CI 1.62-4.82, p 0.001). These relationships persisted after adjusting for TBSA and sepsis. Notably, elevated admission eosinophil counts were not predictive of pulmonary infection (AUC unchanged; DeLong p = 0.61). Conclusion Both the presence of inhalation injury and elevated admission NLR increased the odds of pulmonary infection. NLR was also higher in those with greater burn TBSA but had no relationship with inhalation injury. We identified a novel association-higher peripheral eosinophil counts conferred greater odds of inhalation injury. In contrast, eosinophil counts were not associated with pulmonary infection. These findings suggest that while inhalation injury substantially increases infection risk, it does so via mechanisms not fully captured by peripheral inflammatory ratios. While eosinophil indices did not predict infection risk, they may reflect a hypersensitivity-type immune response and could provide guidance on the presence of inhalation injury and the need for proactive therapies minimizing its consequences. This abstract is funded by: None