634: Biomarker-Concordant Steroid Use in Hospitalized Patients With Community-Acquired Pneumonia

Introduction: The Neutrophil-Lymphocyte ratio (NLR) is an emerging inflammatory biomarker in community-acquired pneumonia (CAP) with potential prognostic implications1,2. Biomarker-concordant corticosteroid dosing strategies have been described with other inflammatory markers, including C-reactive protein (CRP), and found to be associated with improved clinical outcomes3,4. The NLR is more readily available and affordable compared to other inflammatory biomarkers; therefore, it could be easily scalable as a predictive enrichment tool for corticosteroid use in CAP. Our goal is to evaluate the frequency and consequences of prescribing corticosteroids for pneumonia in a biomarker-concordant manner utilizing the NLR. Methods: Single-center prospective cohort study of adults with CAP admitted to the Mayo Clinic in Rochester, Minnesota, between December 10, 2023, and March 11, 2025. Steroid use was ‘biomarker-concordant’ if given when NLR > 12 or withheld when NLR≤12, and vice versa for ‘biomarker-discordant’. Results: Of 545 admissions with CAP, 465 (85 %) had an NLR measured within 24 hours of admission. Steroid use was biomarker-discordant in 222 (47.7%) patients and biomarker-concordant in 243 (52.3%) patients. Systemic corticosteroid use within 24 hours was more common in the discordant group (127 (57.2%) versus 77 (31.7%). After adjusting for the pneumonia severity index (PSI), there was no significant difference in clinical outcomes including in-hospital mortality (odds ratio [95% C.I.] = 1.120 [0.463, 2.706]), 30 – day mortality (odds ratio [95% C.I.] = 1.423 [0.747, 2.710]), oxygen-free days (0.612, IQR –0.655 - 1.879, p=0.343), hospital-free days (0.222, IQR –0.943 - 1.386, p=0.709) between biomarker-discordant versus concordant corticosteroid use. Conclusions: In this observational cohort of hospitalized CAP patients, a biomarker-concordant corticosteroid treatment strategy based on the neutrophil-lymphocyte ratio was not associated with improved clinical outcomes.