1203Predictors of treatment and survival in hepatocellular carcinoma patients: A Bayesian parametric survival analysis

Background Culturally and linguistically diverse communities face several challenges to receiving screening and curative treatment of hepatocellular carcinoma (HCC). We reported the predictors of receipt of treatment and survival in a cohort of patients diagnosed with HCC in Southeast Queensland. Methods Data from 1651 HCC patients (147 migrants) followed between January 1, 2007, and December 31, 2016 were analyzed using chi-square statistic to test for the association between socio-economic variables and receipt of treatment. Bayesian Weibull Accelerated Failure Time regression was used to identify predictors of time to death, and we reported time ratios. Results Receipt of surgical resection for HCC was associated with non-metropolitan residence (P=0.02), non-English language (P<0.001), country of origin (P<0.001), and hepatitis B virus etiology (P<0.001). The median survival time after HCC diagnosis was 9.0 months (interquartile range 2.0‒24.0). The strongest predictors of survival were undifferentiated tumor at presentation (time ratio (TR)=0.30, 95% credible interval (CrI) 0.23‒0.39), age ≥70 years (TR = 0.42, 95%CrI 0.34‒0.53), living in a remote area (TR = 0.67, 95%CrI 0.55‒0.80), and presence of ≥ 1 comorbidity (TR = 0.69 95%CrI 0.54‒0.90). Conclusions Improving access to HCC surveillance and uptake of timely curative treatment for persons living in remote areas and having socioeconomic disadvantage may help patients present to clinic with early stage tumor and provide better survival. Key messages Remoteness of residence is associated with late presentation, low rate of receipt of treatment and predicts survival of patients with HCC.