Five‐year conditional survival for patients with hepatocellular carcinoma in Queensland, Australia

Abstract

Summary Background Conditional survival provides a dynamic assessment of a survivor's ongoing survival probability over‐time post‐diagnosis. Aim To describe 5‐year survival of patients with hepatocellular carcinoma (HCC) conditional on having survived 1, 3 or 5 years. Methods Using a population‐based‐linkage cohort of HCC patients from Queensland, Australia we reported cumulative overall survival by treatment at diagnosis (Kaplan‐Meier method) and conditional 5‐year overall survival after diagnosis (flexible parametric modelling). We also examined disparities in receipt of treatment (Poisson regression). Results 1393 cases were diagnosed during 1996‐2010: 28% received treatment (resection, RFA and/or TACE), and 72% did not receive any of these treatments. Median survival over the 15‐year period was 184 days (IQR 61‐669 days), and improved over time. The overall probability of 1, 3 and 5‐year survival were 38.5% (95% CI 35.9‐41.0), 20.9% (18.8‐23.2) and 14.8% (12.8‐16.9). The probability of surviving 5‐years improved with each additional year that they survived. Patients who survived 1 year had twice the 5‐year survival of those at baseline (33.4%, 95% CI 28.6‐38.2), if they survived 3 years, that approximately tripled (47.9%, 41.5‐54.3), and reached 55.1 (48.5‐61.6) if they survived 5 years after diagnosis. Patients diagnosed during 2006‐2010 were more likely to receive treatment within 12 months (RR = 3.05, 95% CI 2.28‐4.08) compared to 1996‐2000. Higher hepatitis B prevalence in country of birth was associated with receipt of treatment within 12 months after diagnosis (RR = 1.28, 1.03‐1.60). Conclusions Despite improvements in treatment, and subsequent improvement in survival, even for a patient surviving the initial HCC diagnosis, the prognosis remains poor.

Depressing data and grounds for optimism

“…Valery et al presented 5‐year conditional survival for patients with hepatocellular carcinoma in Queensland, Australia . Epidemiological data of this nature are helpful to a variety of practicing clinicians as well as public health decision makers.…”