Clinical relevance of conditional survival of cancer patients in europe: age-specific analyses of 13 cancers.

PURPOSE When cancer survivors wish to receive accurate information on their current prognosis during follow-up, conditional 5-year relative survival may be most suitable. We have estimated conditional 5-year relative survival for 13 cancers using a large European database-European Network for Indicators on Cancer (EUNICE)-of 10 dedicated long-standing cancer registries across Europe. PATIENTS AND METHODS Patients age 15 years and older diagnosed between 1985 and 2004 were included. Conditional 5-year relative survival for each age group was computed for every additional year survived up to 10 years. Period analysis with follow-up period 2000 to 2004 was used. RESULTS All patients with cutaneous melanoma or colorectal, endometrial, or testis cancer and younger patients with stomach, glottis, cervix, ovary, or thyroid cancer or non-Hodgkin's lymphoma exhibited hardly any excess mortality (conditional 5-year relative survival > 95%) given that they were alive at a defined time point within 10 years of initial diagnosis. However, patients with supraglottis, lung, breast, and kidney cancer, as well as older patients with most cancers exhibited substantial excess mortality (conditional 5-year relative survival < 90%). Initial differences in relative survival at diagnosis between age groups largely disappeared with time since initial diagnosis for melanoma, or stomach, colorectal, corpus uteri, or testicular cancer but persisted for patients diagnosed with other tumors. Differences between stage groups became smaller over time or disappeared. CONCLUSION Conditional relative survival shows clinically relevant variations according to time since diagnosis, type of cancer, and age, and can help serve as a guide for cancer survivors in planning for their future and for doctors in planning schedules for surveillance.