Time trends and short term projections of cancer prevalence in France.

BACKGROUND: This study analyzes time trends in cancer prevalence in France and provides short-term projections up to the year 2017. The 15-year prevalence for 24 cancers was estimated from the French cancer registries network (FRANCIM) incidence and survival data.

METHOD: We estimated prevalence using the P = I × S relationship, with flexible modeling of incidence and survival. Based on observations of the incidence and survival up to 2010, different scenarios for evolution up to 2017 were studied, combining stable and dynamic incidence and survival. The determinants of variations in prevalence (incidence, survival and demography) were quantified.

RESULTS: At the end of 2017, an estimated 1,396,000 men and 1,359,000 women having had cancer in the previous 15 years were alive, respectively 5.4% and 4.8% of the population. Twelve percent had been diagnosed in the preceding year and 23% between 10 and 15 years. Between 2010 and 2017, changes in incidence and survival depended on the cancer site. The effect of the demographic change was null for those under age 65, whereas above age 65, the contribution of this factor was 20% in men and 17% in women at 15 years. The different projection scenarios led to very different estimates for some cancers for which incidence strongly varied in the last decades.

CONCLUSION: Prevalent cases are numerous in a country such as France, where incidence and survival are high. Due to the sensitivity of prevalence to changes in incidence and survival, we recommend that the results of projections are presented under different scenarios. We propose a robust and flexible prevalence estimate.