Dynamic Profiling of Angiogenic, Metabolic, and Immune Checkpoint Biomarkers Predicts Chronic Critical Illness and...

Many patients who survive abdominal sepsis develop Chronic Critical Illness (CCI), characterized by prolonged intensive care unit (ICU) stay, ongoing organ dysfunction, and poor one-year outcomes. Current severity scoring systems, such as SOFA and APACHE II, are not effective in identifying patients at high risk for this trajectory, limiting the opportunities for early intervention.

We conducted a prospective longitudinal study of adults diagnosed with abdominal sepsis at a tertiary acute care center. The serum levels of 13 biomarkers were measured on days 1, 4, 7, and 14 after diagnosis. Patients were grouped into two categories: Rapid Recovery (RAP) and CCI. Predictive models were developed by combining clinical data based on the Predisposition, Insult, Response, and Organ dysfunction (PIRO) framework with biomarker trends to assess the evolving host response.

Patients with CCI showed poorer clinical outcomes than those with RAP. Among the 252 patients included in the analysis, dynamic profiling revealed significant differences between the RAP and CCI groups. Patients who developed CCI exhibited persistently elevated levels of immunosuppressive biomarkers (sPD-L1, Arg-1, and TGF-β) and inflammatory markers (IP-10), along with dysregulation of the angiogenic and metabolic pathways. Multivariate models incorporating PIRO variables showed strong performance in predicting CCI by day 4 (AUC = 0.899) and poor one-year outcomes (Zubrod 4/5) by day 7 (AUC = 0.851). The addition of key biomarkers significantly enhanced the predictive accuracy for both CCI (AUC = 0.926) and functional outcomes (AUC = 0.910).

Our findings suggest that Persistent Inflammation, Immunosuppression, and Catabolism Syndrome (PICS), a maladaptive pathophysiological state, plays a central role in the development of CCI following abdominal sepsis. The combination of serial biomarker data and clinical variables enables earlier and more accurate risk stratification. This approach may support timely and targeted interventions to improve patient survival and long-term recovery.