Factors predicting long-term survival of hepatocellular carcinoma in developing country.

231 Background: Hepatocellular carcinoma (HCC) is a common cancer globally, especially in Asian Pacific countries with survival time limit. Previous data on long-term survival (LS) in HCC patients has been limited. This study aims identifying factors that predict LS for HCC patients treated in hospitals with restricted access to advanced therapies. Methods: This study utilized a retrospective cohort design, analyzing electronic medical records of patients diagnosed with HCC at Sawanpracharak Hospital from October 2018 to September 2022. LS for HCC was defined as survival for two years or more. Univariable and multivariable with stepwise backward regression analyses were employed to identify independent prognostic factors. Results: The analysis included 398 HCC patients, with 75 (18.84%) classified as long-term survivors (LS HCC). Stepwise backward regression identified that ALP levels <172 U/L were associated with a significantly increased likelihood of LS (OR = 3.667, 95% CI 1.443-9.317, p=0.006). Additionally, curative surgery/RFA treatment (OR = 1.941, 95% CI 1.128-3.342, p=0.017) and Albumin >3.5 g/dL (OR = 9.898, 95% CI 1.160- 84.438, p=0.036) were significant positive predictors. ECOG performance status (3-4 vs 2 vs 0-1) and AJCC stage (early vs locally vs advance) also emerged as significant factors, with better performance status (OR = 8.723, 95% CI 1.046-72.741, p=0.045) and AJCC stage 3 and 4 (OR 0.200 95% CI= 0.075-0.532, P= 0.001 and OR = 0.204, 95% CI 0.0547-0.760, p=0.018 respectively) associated with LS. Notably, alcohol consumption was associated with a decreased likelihood of LS (OR = 0.399, 95% CI 0.171-0.929, p=0.033). The model demonstrated good discriminative ability with an AUC of 0.912. Conclusions: Lower ALP levels, curative treatment, higher albumin levels, better ECOG performance status, earlier AJCC stage, and abstinence from alcohol consumption were all associated with a significantly increased likelihood of LS HCC. The high AUC of the model suggests its potential for predicting long-term survival in this patient population. [Table: see text]