Trends in national and ethnic burden of ovarian cancer mortality in South Africa (1999-2018): a population based, age-period-cohort and join point regression analyses.

Abstract

Ovarian cancer is the most lethal and third leading cause of gynaecological cancers globally and in South Africa (SA). However, its current mortality trends have not been evaluated in most sub-Saharan African Countries including South Africa that is currently undergoing epidemiological and health transitions. We evaluate the trends in the ovarian cancer mortality rates in SA over 20 years (1999-2018).

Methods: Crude (CMR) and age standardised mortality rates (ASMR) of ovarian cancer was calculated based on national mortality data of South Africa. The overall and ethnic trends of ovarian cancer mortality among women aged 15 years and older from 1999 to 2018 was assessed using the Join point regression model, while Age-period-cohort regression analysis was conducted to evaluate the underlying impact of age, period and cohort on ovarian cancer mortality.

Results: In all, 12,721 ovarian cancer deaths were reported in South Africa from 1999 to 2018 and the mortality rates increased from 2.34 to 3.21 per 100,00 women at 1.8% per annum. In 2018, the overall mean age at ovarian cancer death in South Africa was 62.30 ± 14.96 years while the mean age at death among Black women (58.07 ± 15.56 years), was about 11 years earlier than among White women (69.48 ± 11.71 years). In 2018, the White ethnic group (4.93 deaths per 100,000 women) had about doubled the ovarian cancer ASMR for the non-Whites (Indian/Asians, 2.92/100,000 women, mixed race, 2.49/100,000 women and Black women (2.36/ 100,000 women). All the ethnic groups had increased ASMR with Black women (Average annual percent change, [AAPC]: 4.7%, P-value < 0.001) and Indian/Asian women (AAPC: 2.5%, P-value < 0.001) having the highest rise. Cohort mortality risk ratio of ovarian cancer increased with successive birth cohort from 0.35 among 1924-1928 birth cohorts to 3.04 among 1999-2003 cohort and the period mortality risk increased by about 13% and 7.5% from 1999 to 2003 to 2004-2008 (RR: 0.87, 95% CI: 0.80-0.94), and from 2004 to 2008 to 2009-2013 (RR: 1.075, 95% CI:1.004-1.152) respectively. The longitudinal age analysis revealed that ovarian cancer increased with age, but there was an exponential increase from 55 years.

Conclusions: Our study showed that there was increasing trends in ovarian cancer mortality among all the South African ethnic groups, driven partly by increasing cohort and period mortality risks. We therefore highlight the huge burden of ovarian cancer in SA and the need for targeted intervention. Public health interventions geared towards reducing ovarian cancer mortality should be instituted and ethnic disparity should be incorporated in the cancer control policy.