Abstract / Summary
Objective Sepsis remains associated with substantial short-term mortality, and conventional severity indicators may not fully capture biological heterogeneity related to tissue injury, inflammation, and immune suppression. This study evaluated whether early plasma cell-free DNA and monocyte human leukocyte antigen-DR (mHLA-DR) expression could improve the prediction of 28-day mortality in adult intensive care unit patients with sepsis. Methods This single-center prospective observational cohort study enrolled consecutive adult intensive care unit patients meeting Sepsis-3 criteria. Plasma cell-free DNA and mHLA-DR were measured within 24 h of intensive care unit admission. The primary outcome was 28-day all-cause mortality. Multivariable logistic regression was used to assess associations adjusted for age, Sequential Organ Failure Assessment score, blood lactate levels, and septic shock. Clinical baseline, single biomarker, and combined biomarker models were compared using the area under the receiver operating characteristic curve, DeLong test, net reclassification improvement, integrated discrimination improvement, Brier score, calibration, bootstrap internal validation, and decision curve analysis. Results The final combined analysis set included 307 patients, of whom 89 died within 28 days (28.99%). Nonsurvivors had higher cell-free DNA and lower mHLA-DR than survivors. After adjustment, each 1-SD increase in ln( cell-free DNA) was associated with higher mortality risk, whereas each 1-SD increase in ln(mHLA-DR) was associated with lower mortality risk. Adding both biomarkers to the clinical baseline model increased the area under the receiver operating characteristic curve from 0.79 to 0.88, improved reclassification, reduced prediction error, and provided higher estimated net benefit across part of the evaluated threshold-probability range. Conclusions Early cell-free DNA and mHLA-DR provided complementary prognostic information for 28-day mortality in patients with sepsis. The combined model demonstrated improved internally validated prediction performance. However, external validation is required before clinical implementation.