[Value of Hepcidin as a diagnostic biomarker of sepsis in critically ill adults].

OBJECTIVE: To investigate the diagnostic value of Hepcidin as a sepsis biomarker in critically ill adults.

METHODS: An observational study was conducted. The patients with suspected or proven infection admitted to intensive care unit (ICU) of Zhoupu Hospital Affiliated to Shanghai University of Medicine and Health Sciences from March 2016 to November 2017 were enrolled. According to the third international consensus definitions for sepsis and septic shock (Sepsis-3), the patients were divided into non-sepsis group and sepsis group, and the septic patients were subdivided into general sepsis subgroup and septic shock subgroup according to the severity of disease. The differences in serum Hepcidin, interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), procalcitonin (PCT), C-reactive protein (CRP), white blood cell (WBC), neutrophil granulocytes (NEUT) and lactic acid (Lac) within 1 hour after ICU admission between non-sepsis and sepsis groups and among the sepsis subgroups were compared. The acute physiology and chronic health evaluation II (APACHE II) within 24 hours after ICU admission and sequential organ failure score (SOFA) were recorded, and the mortality rate was followed up for 28 days. Receiver operation characteristic curve (ROC) was used to evaluate and compare the diagnostic value of Hepcidin and PCT, CRP, WBC for sepsis. Logistic regression model was used to estimate the association between Hepcidin and sepsis. Spearman correlation analysis was used to analyze the correlation between Hepcidin and other parameters of sepsis patients.

RESULTS: A total of 183 patients were enrolled, 93 in the non-sepsis group and 90 in the sepsis group (48 with general sepsis and 42 with septic shock). (1) The levels of Hepcidin, IL-6, TNF-α, PCT, Lac in serum, and APACHE II and SOFA scores in the sepsis group were significantly higher than those in the non-sepsis group. ROC analysis showed that the area under the ROC curve (AUC) of Hepcidin and PCT for sepsis diagnosis were 0.865 [95% confidence interval (95%CI) = 0.807-0.911] and 0.848 (95%CI = 0.788-0.897), respectively, without statistical significance (Z = 0.443, P = 0.657). Furthermore, the AUC of Hepcidin for sepsis diagnosis was significantly higher than that of the conventional biomarkers CRP and WBC [AUC was 0.530 (95%CI = 0.455-0.604) and 0.527 (95%CI = 0.452-0.601), respectively] with statistical significance (both P < 0.01). When Hepcidin > 54.00 μg/L, its sensitivity for sepsis diagnosis was 95.56%, specificity was 66.67%, positive and negative predictive value was 73.51% and 93.94%, respectively. Parallel test was conducted for combination of Hepcidin and PCT, which showed that the AUC was 0.885, and the sensitivity and negative predictive value was significantly improved to 98.96% and 98.36%, respectively. Logistic regression analysis demonstrated that after adjusted for PCT, Hepcidin > 54.00 μg/L was also associated with sepsis independently, with odds ratio (OR) of 1.011 (95%CI = 1.008-1.015, P < 0.001), indicating that Hepcidin and PCT were not completely overlapped in the diagnosis of sepsis. (2) With the increase in infection severity, serum Hepcidin, PCT, IL-6, TNF-α, Lac, APACHE II, SOFA score and 28-day mortality all showed an increasing trend in patients. There was a significantly positive correlation between Hepcidin and IL-6, TNF-α, PCT, APACHE II, and SOFA in the sepsis patients (r value was 0.526, 0.449, 0.591, 0.359, and 0.374, respectively, all P < 0.01), but no correlation was found between Hepcidin and Lac (r = 1.104, P > 0.05).

CONCLUSIONS: Serum Hepcidin is a useful biomarker for the diagnosis of sepsis, and it is correlated to the severity of the sepsis. The combination of Hepcidin and PCT can improve the accuracy of diagnosis of sepsis.

Clinical trial registration: China Clinical Trial Registration Center, ChiCTR-DDD-16008522.