Temporal trends in population-level cure of cancer: the Australian context.

BACKGROUND: With the improvements in cancer diagnosis and treatment, more cancer patients are surviving for longer periods than before. This study aims to quantify the proportion cured and median survival time for those who are not cured for major cancers in Australia.

METHODS: Australian population-based cohort of 2,164,172 cases aged 15-89 years, whose first cancer diagnosis between 1982 and 2014 was one of 22 leading cancers, were followed up to December 2014. Flexible parametric cure models were used to estimate the proportion cured and median survival time for those uncured by age, sex and spread of disease, and temporal trends in these measures.

RESULTS: Cure estimates could be generated for 19 of the 22 cancer types. The unadjusted proportion cured ranged from 5.0% for pancreatic cancer to 90.0% for melanoma. Median survival time for those uncured ranged from 0.35 years for pancreatic cancer to 6.05 years for prostate cancer. Cancers were divided into four groups according to their proportion cured in 1980s and the degree of improvement over 28 years. Oesophageal, stomach, pancreatic, liver, gallbladder, lung, and brain cancer had lower proportion cured and smaller improvements over time.

CONCLUSIONS: For cancers with poor survival where little has changed over time either in prolonging life or achieving statistical cure, efforts should be focused on reducing the prevalence of known risk factors and earlier detection, thereby enabling more effective treatment.

IMPACT: Cure models provide unique insights into whether survival improvements are due to prolonging life, or through curing the disease.