Abstract
Objective: To establish and explore a novel model and its clinical application value based on abnormal des-gamma-carboxy prothrombin (DCP) for the early-stage diagnosis of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). Methods: A total of 420 cases with chronic HBV infection with nodular liver lesions examined by imaging at the Third Hospital of Hebei Medical University from January 2021 to June 2024 were retrospectively selected. They were divided into the HBV-HCC group (182 cases) and the control group (238 cases) according to the current HCC diagnostic criteria. The basic information of patients, liver-related biochemical indicators, serum DCP, alpha-fetoprotein (AFP) levels, and the efficacy of combined detection in diagnosing early-stage HCC were collected and analyzed. A DSGAA model based on DCP (D) combined with gender (S), γ-glutamyl transferase (GGT, G), AFP (A) and age (A) as independent variables was constructed. The diagnostic performance of the novel model was compared with that of the traditional model through nomogram visualization output and calibration curve. Results: The age, sex, hemoglobin, albumin, alanine aminotransferase, alkaline phosphatase, and GGT levels were significantly higher in patients with HCC than those of the control group (P<0.05). The positivity detection rate in patients with HBV-HCC was significantly higher in DCP than that of AFP (85.71% vs. 59.89%,P<0.05). The abnormal detection rate of DCP in patients with AFP-negative was 76.7%. The sensitivity for diagnosing HCC was significantly higher in DCP than AFP (73.63% vs. 64.29%,P<0.05), with specificity of 83.6% in all. The specificity for diagnosing early-stage HCC was 89.09%, surpassing that of AFP at 68.06% (P<0.05). The area under the receiver operating characteristic curve (AUC) for the constructed DSGAA diagnostic model was 0.8841, with an optimal cutoff value of 0.377, a sensitivity of 80.22%, and a specificity of 86.13%. The AUC for diagnosing early-stage HCC was 0.8122, with a sensitivity of 66.18%, and a specificity of 86.13%, and the diagnostic efficacy was higher than other models (P<0.05). Conclusion: DCP has superior diagnostic efficacy for HBV-related HCC, and the DSGAA model is expected to be used as a new method for screening and diagnosing early-stage HBV-related HCC.