Abstract
OBJECTIVE: To evaluate the diagnostic utility of serum cystatin S (CST4) in chemotherapy-treated colorectal cancer (CRC) patients and establish its complementary value to conventional tumor markers. METHODS: This retrospective cohort study analyzed 81 CRC patients receiving chemotherapy and 83 colorectal polyp controls. Serum CST4 levels were quantified by ELISA alongside six conventional tumor markers (CEA, CA125, CA153, CA199, AFP, CA724). Diagnostic performance was assessed through ROC analysis and multivariate logistic regression. Additionally, in vitro experiments with HCT116 CRC cells were conducted to validate the regulatory relationship between CST4 and PDGFRB. RESULTS: CRC patients exhibited significantly elevated CST4 levels compared to polyp controls (median [IQR]: 54.07 [32.18-91.49] vs 37.48 [24.18-49.28] U/mL, P<0.05). CST4 demonstrated superior diagnostic performance with AUC = 0.689 (95%CI:0.607-0.771), outperforming individual conventional markers. Notably, CST4 maintained diagnostic independence across tumor stages (P>0.05) and age groups. A multimodal diagnostic model combining CST4 with CEA, CA724, and CA125 significantly enhanced detection capability (AUC = 0.828, sensitivity 74.1%, specificity 81.9%), representing a 28.4% sensitivity improvement over CST4 alone. In vitro, CST4 knockdown in HCT116 cells led to a 68.3% reduction in PDGFRB expression (P<0.0001), validating a regulatory axis between CST4 and PDGFRB. CONCLUSION: CST4 emerges as a stable post-chemotherapy biomarker that effectively discriminates malignant colorectal lesions. Its integration with conventional markers creates a robust diagnostic algorithm, while functional validation supports a mechanistic role via PDGFRB-mediated pathways. These findings position CST4 as a promising candidate for therapeutic monitoring and recurrence detection in CRC management.