Development and validation of a GP73-based predictive model for the diagnosis of early-stage hepatocellular carcinoma.

BACKGROUND

Hepatocellular carcinoma (HCC) has become a major global public health concern due to its high incidence, high mortality, and frequent diagnosis at advanced stages. Currently available biomarkers, such as alpha-fetoprotein (AFP), have limited sensitivity and specificity for early-stage HCC, making early screening a significant challenge. Therefore, identifying novel and complementary biomarkers is essential to improve early detection rates and patient prognosis. Golgi protein-73 (GP73) is a secreted glycoprotein that is highly expressed in various malignancies and can be released into the bloodstream. Herein, we evaluated its potential value for early diagnosis of HCC.

METHODS

This retrospective cross-sectional study aimed to evaluate the diagnostic value of serum GP73 for early-stage HCC. A total of 401 patients were included in the study, comprising 63 patients with early-stage HCC and 338 patients with liver cirrhosis. Diagnoses of HCC and cirrhosis were confirmed based on a combination of imaging, pathological findings, and clinical guidelines. Serum levels of GP73 and AFP were measured using enzyme-linked immunosorbent assay (ELISA), and other clinical variables such as age and prothrombin time (PT) were obtained from medical records.

RESULTS

The area under the curve (AUC) for AFP and GP73 alone in diagnosing early-stage HCC was 0.769 [95% confidence interval (CI): 0.700-0.838] and 0.627 (95% CI: 0.539-0.716), respectively. After adjustment in multivariate analysis, GP73 remained an independent diagnostic factor (P<0.05). The developed nomogram achieved corrected C-indexes of 0.812 and 0.918 in the training and validation cohorts, respectively. Among patients infected with hepatitis B virus (HBV), the nomogram demonstrated AUCs of 0.836 in the training cohort and 0.900 in the validation cohort, indicating good discriminative ability. Baseline characteristics were comparable between the groups.

CONCLUSIONS

The nomogram model proposed in this study, which integrates GP73, AFP, age, and PT, may serve as a simple, intuitive, and customizable clinical tool for identifying early-stage HCC among patients with liver cirrhosis. The model exhibits high diagnostic performance, especially in HBV-related populations, and shows promising potential for clinical application.