Dynamic body mass index trajectories are associated with survival outcomes in advanced Hepatocellular Carcinoma treated with immunotherapy.

BACKGROUND AND OBJECTIVE: Obesity paradoxically increases both cancer risk and anti-tumor efficacy. However, the prognostic association of dynamic body mass index (BMI) trajectories in advanced hepatocellular carcinoma (HCC) patients receiving immunotherapy remains underexplored. MATERIALS AND METHODS: This retrospective cohort study included 563 advanced HCC patients treated with anti-PD-1/PD-L1 immunotherapy at the First Affiliated Hospital of Sun Yat-sen University from January 2019 to May 2023. Longitudinal BMI trajectories were modeled using latent class mixed model (LCMM). Survival outcomes were analyzed using Kaplan-Meier, Cox regression, and landmark analyses. RESULTS: A four-class BMI trajectory model was identified: high-stable (10.3%, median baseline BMI [bBMI]: 27.8 kg/m²), moderate-stable (53.1%, 23.0 kg/m²), moderate-rapid-decline (10.7%, 23.0 kg/m²), and low-slow-decline (25.9%, 20.0 kg/m²). The high-stable group showed superior survival outcomes (median progression-free survival [PFS]: 13.17 months; overall survival [OS]: 32.36 months), while the moderate-rapid-decline group had the worst prognosis (PFS: 4.40 months; OS: 10.87 months). After adjusting for lymph node metastasis, Child-Pugh classification, bBMI, combination therapy, and alpha-fetoprotein (AFP), the hazard ratios (HRs) for PFS in high-stable, moderate-stable, and low-slow-decline groups compared to moderate-rapid-decline were 0.37 (P < 0.001), 0.60 (P = 0.003), and 0.75 (P = 0.151), respectively; for OS, the HRs were 0.25 (P < 0.001), 0.44 (P < 0.001), and 0.73 (P = 0.210). BMI trajectory contributed substantially to survival models. CONCLUSION: Dynamic BMI trajectories were associated with survival outcomes in advanced HCC patients receiving immunotherapy, with high-stable patterns showing favorable outcomes. Continuous BMI monitoring may provide a practical prognostic indicator, but prospective validation is required before clinical intervention thresholds can be recommended.