Diagnostic and Prognostic Utilities of Pancreatic Stone Protein in Patients with Suspected Sepsis.

Abstract

Background/Objectives: Pancreatic stone protein (PSP) is an emerging biomarker of sepsis that is secreted from pancreas sensing remote organ damages. We explored the diagnostic and prognostic utilities of PSP in patients with suspected sepsis. Methods: In a total of 285 patients (suspected sepsis, n = 148; sepsis, n = 137), we compared PSP with procalcitonin (PCT) and sequential organ failure assessment (SOFA) score. Sepsis diagnoses were explored using receiver operating characteristic curve analyses with area under the curves (AUCs). Clinical outcomes (in-hospital mortality, 30-day mortality, and kidney replacement therapy [KRT]) were explored using the Kaplan-Meier method and a multivariate analysis with hazard ratio (HR). Results: PCT and PSP were comparable for sepsis diagnosis (AUC = 0.71-0.72, p < 0.001). The sepsis proportion was significantly higher when both biomarkers increased than when either one or both biomarkers did not increase (89.0% vs. 21.3-47.7%, p < 0.001). Each biomarker quartile (Q1-Q4) differed significantly according to their SOFA score (all p < 0.001). Compared with Q1, the Q2-Q4 groups showed worse clinical outcomes (p = 0.002-0.041). Both biomarkers added to the SOFA score showed higher HRs than the SOFA score alone (3.3-9.6 vs. 2.8-4.2, p < 0.001-0.011), with nearly 2.5-fold higher HR (9.6 vs. 4.2) for predicting KRT. Conclusions: Although PCT and PSP did not independently predict clinical outcomes in the multivariate analysis, PSP demonstrated diagnostic and prognostic utilities in patients with suspected sepsis, especially for predicting kidney dysfunction. PSP, alone or in combination with PCT, would be a valuable tool that can be added to clinical assessments.