Abstract
PURPOSE: With the increasing incidence of acute diquat (DQ) poisoning, this study aims to analyze immune biomarkers in patients and explore their impact on outcomes. METHODS: A retrospective multicenter cohort study included 145 patients with acute DQ poisoning from six hospitals in Zhejiang Province, China. Immune-inflammatory parameters were compared between DQ patients and healthy controls. Patients were then grouped according to their 28-day survival status for prognostic analysis. Clinical, laboratory, and immune parameters were collected. Least absolute shrinkage and selection operator (LASSO) regression, multivariate logistic regression, and receiver operating characteristic analysis were used to identify predictors of mortality. Bioinformatics analysis was performed to explore potential molecular targets and pathways. RESULTS: Of 145 patients, 72 died within 28 days. Non-survivors showed higher procalcitonin (PCT), lactate, and organ injury markers, as well as elevated interleukin-6 and interleukin-10, compared with survivors. Compared with healthy controls, patients with DQ poisoning exhibited increased white blood cells (WBC), neutrophils, monocytes, and cytokines, alongside reduced lymphocytes and T-cell subsets (all P < 0.0001). LASSO and logistic regression identified PCT, WBC, and lactate as independent predictors of mortality, with PCT providing the greatest discriminative value (AUC = 0.88). Bioinformatics analysis further indicated enrichment of immune-related pathways and hub genes associated with immune dysregulation. CONCLUSION: Acute DQ poisoning causes pronounced immune-inflammatory disturbances that are more severe in non-survivors. PCT is the strongest independent predictor of 28-day mortality, while exploratory bioinformatics highlights immune pathways with potential prognostic and therapeutic relevance.