Serum Golgi Protein 73 as a Noninvasive Marker of Hepatic Fibrosis in Chronic Hepatitis C Virus Infection
Yun‐Chu Chen, Yu‐Ping Chang, Jenna Bedrava, Laurel Jackson, Ji‐Yuh Lee, Chi‐Yi Chen, Wei‐Yu Kao, Chih‐Lin Lin, Sheng‐Shun Yang, Yu‐Lueng Shih, Cheng‐Yuan Peng, Fu‐Jen Lee, Ming‐Chang Tsai, Shang‐Chin Huang, Tung‐Hung Su, Tai‐Chung Tseng, Chun‐Jen Liu, Pei‐Jer Chen, Jia‐Horng Kao, Claire Chen, Chen‐Hua LiuABSTRACT
The clinical utility of serum Golgi protein 73 (GP73) for evaluating hepatic fibrosis in hepatitis C virus (HCV) infection remains to be determined. We retrospectively enrolled 978 patients with serum GP73, aspartate‐to‐platelet ratio index (APRI), and Fibrosis‐4 (FIB‐4) measurements. Hepatic fibrosis was staged using vibration‐controlled transient elastography (VCTE) with cutoff values of ≤ 7.0, 7.1–9.4, 9.5–12.4, and ≥ 12.5 kPa corresponding to F0–F1, F2, F3, and F4. Distributions of GP73, APRI, and FIB‐4 levels across fibrosis stages were compared using Kruskal–Wallis tests. Areas under the receiver operating characteristic curves (AUROCs) were calculated and compared using DeLong tests, and optimal GP73 cutoff values were determined by maximal Youden indices. Multivariable linear regression was performed for factors associated with GP73 levels. Median GP73, APRI, and FIB‐4 levels increased with advancing fibrosis stage ( p < 0.001). Serum GP73 showed superior performance to APRI in discriminating ≥ F2 (AUROC 0.86 vs. 0.79, p < 0.001), ≥ F3 (0.88 vs. 0.81, p < 0.001), and F4 (0.90 vs. 0.84, p = 0.002). GP73 showed better performance than FIB‐4 for ≥ F2 (0.86 vs. 0.82, p = 0.005) and comparable performance for ≥ F3 (0.88 vs. 0.87, p = 0.63) and F4 (0.90 vs. 0.89, p = 0.72). Optimal GP73 cutoff values were 122, 159, and 174 ng/mL for ≥ F2, ≥ F3, and F4. Regression analysis demonstrated a significant association between fibrosis stage and GP73 levels. In conclusion, serum GP73 is a reliable and accurate noninvasive biomarker for staging hepatic fibrosis for HCV.