Application value of blood-brain barrier and peripheral inflammatory markers in early diagnosis of pancreatic encephalopathy in severe acute pancreatitis.

OBJECTIVE: To investigate the value of blood-brain barrier (BBB) injury markers [S100β protein, matrix metalloproteinase-9 (MMP-9), cerebrospinal fluid/serum albumin ratio (QAlb)] and peripheral inflammatory markers [tumor necrosis factor-α (TNF-α), monocyte chemoattractant protein-1 (MCP-1), interleukin-6 (IL-6)] in the early diagnosis of severe acute pancreatitis (SAP)-related pancreatic encephalopathy (PE). METHODS: A prospective single-center observational cohort study was conducted. A total of 282 SAP patients admitted from March 2024 to February 2026 were consecutively enrolled, with an Acute Physiology and Chronic Health Evaluation II (APACHE II) score ≥ 8, or a CT Severity Index (CTSI) score ≥ 7, or persistent organ failure > 48 h. Patients with pre-existing central nervous system diseases, traumatic brain injury, intracranial infection, alcohol dependence, Wernicke's encephalopathy, severe liver or kidney diseases, or pregnancy and lactation were excluded. Venous blood samples were collected within 24 h after admission (D0) and on day 3 (D3), day 7 (D7), and day 14 (D14) of the disease course. Serum levels of BBB injury markers (S100β and MMP-9) and peripheral inflammatory markers (TNF-α, IL-6, and MCP-1) were measured by enzyme-linked immunosorbent assay (ELISA). Among the 282 enrolled patients, 50 (17.73%) who met clinical indications and provided informed consent underwent lumbar puncture for QAlb measurement. The primary outcome was the occurrence of PE during hospitalization. Differences in each indicator between the PE and non-PE groups at each time point were compared. Generalized estimating equations (GEE) were used to analyze the temporal trajectories of each indicator. Pearson correlation analysis was performed to examine correlations among indicators. The bootstrap method was used to test the mediating effects. Receiver operating characteristic (ROC) curves were constructed to evaluate the diagnostic performance of individual markers and combined models. RESULTS: Among the 282 SAP patients, the incidence of PE was 21.99% (62/282). In the PE group, TNF-α, IL-6, and MCP-1 levels were significantly higher than those in the non-PE group at D0 (P < 0.05), peaked at D3, and remained significantly elevated at D7 and D14 (P < 0.05). In the PE group, MMP-9 and S100β showed no significant differences from the non-PE group at D0 (P > 0.05), but began to increase significantly from D3, remained at high levels at D7, and declined at D14 while still remaining significantly higher than those in the non-PE group (P < 0.01). Among the 50 patients who underwent lumbar puncture, 22 were in the PE group and 28 in the non-PE group; the QAlb level at D3 in the PE group was significantly higher than that in the non-PE group (P < 0.05). GEE analysis of temporal trajectories showed that TNF-α, IL-6, and MCP-1 were significantly elevated at D3 and D7 compared with D0, and significantly decreased at D14 compared with D0. MMP-9 and S100β were significantly elevated at D3, D7, and D14 compared with D0 (P < 0.05). Pearson correlation analysis revealed a strong positive correlation between S100β and MMP-9 (r = 0.742, P < 0.01). Bootstrap mediation analysis confirmed that IL-6 and MCP-1 exerted significant indirect effects on the occurrence of PE through the MMP-9/S100β axis, with mediation effect sizes of 0.336 and 0.312, accounting for 74.50 and 72.90% of the total effects, respectively, while the direct effects were not significant (P > 0.05). ROC analysis showed that the area under the curve (AUC) of the combined model of MMP-9 + S100β + MCP-1 reached 0.901, which was significantly superior to single markers and other combined models. The length of hospital stay and 28-day mortality in the PE group were both higher than those in the non-PE group (P < 0.05). CONCLUSION: The combined model of MMP-9 + S100β + MCP-1 can be used in clinical practice to identify high-risk PE patients at an early stage. Moreover, this study verified the causal chain of inflammation-induced BBB damage leading to brain injury in PE, providing a theoretical basis for MMP-9-targeted prevention strategies for PE.