Incidence trends and age-period cohort effect on ovarian cancer in high-income countries

Abstract

Background Ovarian cancer is the seventh most common cancer worldwide. Historically, the incidence of ovarian cancer in Europe and North America had been higher compared to other regions of the world. Among European countries, the incidence of ovarian cancer has steadily declined over the years, nevertheless, countries in Europe continues to have the highest incidence of ovarian cancer. The aim of the study is to provide an overview of ovarian cancer incidence trends for high-income countries that are currently part of the Cancer Survival in High-Income Countries (SURVMARK-2) project. These countries include Norway, Denmark, Ireland, the United Kingdom, Canada, Australia and New Zealand. In addition, an age-period-cohort (APC) analysis was also conducted utilizing data from population-based cancer registries from these seven countries. Methods Ovarian cancer incidence rates were calculated using data from three sources, namely, SURVMARK-2, the Cancer Incidence in Five Continents database (CI5plus), and the European cancer registries (EUREG) database. The study contains all available years included in the databases until 2014, with Norway and Denmark having the longest study periods (1953–2014). The overall age-standardized incidence rates for ovarian cancer in women aged 20 and above were computed per year. SURVMARK-2 was utilized to calculate the incidence rates from 1995 to 2014. Incidence rates prior to 1995 were then derived from the CI5plus and the EUREG databases. Additionally, the estimated annual percent change (EAPC) between 1999 and 2013 was calculated for all women age 20 and above, women age 25–49 years and 50–74 years. Lastly, the data was grouped into 5-year age groups starting with 20–24 through 70–74 years, and APC analysis was performed to examine the effects of birth cohort and period. In addition, the goodness-of-fit of the models were assessed and the likelihood ratio test was used for APC model comparison. Results In general, Norway, Denmark, Ireland and the United Kingdom, consistently had higher incidence of ovarian cancer compared to non-European countries. Nevertheless, the incidence trends revealed decreasing incidence trends of ovarian cancer in all countries evaluated. Additionally, in the last 15-year period of the study, the overall incidence rate of ovarian cancer has generally been stable. Between 1999 and 2013 the highest decline of ovarian incidence rate was observed in Norway (EAPC: −1.9, 95% CI: −4.0, 0.4). Incidences of ovarian cancer in 25–49 years and 50–74 years age groups have also declined for most countries. Notably, among the age group 25–49 years, a statistically significant decline of ovarian cancer incidence rate was observed in Norway (EAPC: −4.2, 95% CI: −7.7, −0.6). In contrast, a modest increase of incidence rate in Canada (EAPC: 1.8, 95% CI: −2.2, 5.9) was observed in this age group. Moreover, the United Kingdom (EAPC: −1.8, 95% CI: −3.3, −0.3) had statistically significant decline of incidence rates among women 50–74 years. In addition, the APC analysis yielded the full APC model as the best fitting model for all countries. After adjusting for the period effect, a statistically significant cohort effect (P Conclusion In summary, a gradual decline of ovarian cancer was observed for all countries in the study. The birth cohort effect observed in the study may be linked to the changes in the prevalence of ovarian cancer risk factors, such as the use of oral contraceptive pill. In the other hand, the period effect observed may be explained partly by changes in disease classifications and cancer registry practices.