Dynamic changes in non-invasive fibrosis scores as early predictors of deterioration in chronic hepatitis B

Non-invasive scores such as APRI, FIB-4, GPR and ALBI are commonly used to evaluate liver fibrosis and functional reserve in patients with chronic hepatitis B (CHB). However, interpretation has predominantly focused on single-time measurements, and the prognostic value of short-term changes (delta scores) remains uncertain. This study evaluated whether one-year changes in these scores are associated with laboratory-based composite deterioration during routine follow-up of CHB patients.

This retrospective cohort included adult CHB patients with two laboratory assessments approximately 12 months apart. APRI, FIB-4, GPR and ALBI were calculated at baseline (T₀) and follow-up (T₁); delta values were defined as ΔScore = T₁ - T₀. Laboratory-based composite deterioration was defined as meeting ≥1 criterion based on albumin, total bilirubin, INR, or platelet count. Logistic regression models, adjusted for age, gender, HBV-DNA level, HBeAg status, and antiviral therapy assessed associations between delta scores and deterioration. ROC analysis evaluated discriminative performance.

A total of 297 patients were included (56.9% male; median age 45 years). Laboratory-based composite deterioration occurred in 23.6%. Baseline non-invasive scores were not associated with deterioration, while ΔAPRI, ΔFIB-4 and ΔGPR were significant predictors. In multivariable analyses, each 0.1-unit increase corresponded to higher risk: ΔAPRI (aOR 1.55; p = 0.004), ΔFIB-4 (aOR 1.29; p < 0.001) and ΔGPR (aOR 1.86; p = 0.001). When standardized, ΔFIB-4 demonstrated the strongest association (aOR 7.32 per 1-SD increase). ΔFIB-4 also showed the best discrimination (AUC 0.730).

One-year dynamic changes in non-invasive scores, particularly ΔFIB-4, provide additional and clinically relevant prognostic information for identifying early deterioration during routine follow-up. Incorporating delta score monitoring into clinical practice may help to detect laboratory-based deterioration at an earlier stage.