Cancer survival in sub-Saharan Africa (SURVCAN-3): a population-based study

Abstract

Background: The Cancer Survival in Africa, Asia, and South America project (SURVCAN-3) of the International Agency for Research on Cancer aims to fill gaps in the availability of population-level cancer survival estimates from countries in these regions. Here, we analysed survival for 18 cancers using data from member registries of the African Cancer Registry Network across 11 countries in sub-Saharan Africa.

Methods: We included data on patients diagnosed with 18 cancer types between Jan 1, 2005, and Dec 31, 2014, from 13 population-based cancer registries in Cotonou (Benin), Abidjan (CÔte d'Ivoire), Addis Ababa (Ethiopia), Eldoret and Nairobi (Kenya), Bamako (Mali), Mauritius, Namibia, Seychelles, Eastern Cape (South Africa), Kampala (Uganda), and Bulawayo and Harare (Zimbabwe). Patients were followed up until Dec 31, 2018. Patient-level data including cancer topography and morphology, age and date at diagnosis, vital status, and date of death (if applicable) were collected. The follow-up (survival) time was measured from the date of incidence until the date of last contact, the date of death, or until the end of the study, whichever occurred first. We estimated the 1-year, 3-year, and 5-year survival (observed, net, and age-standardised net survival) by sex, cancer type, registry, country, and human development index (HDI). 1-year and 3-year survival data were available for all registries and all cancer sites, whereas availability of 5-year survival data was slightly more variable; thus to provide medium-term survival prospects, we have focused on 3-year survival in the Results section.

Findings: 10 500 individuals from 13 population-based cancer registries in 11 countries were included in the survival analyses. 9177 (87·4%) of 10 500 cases were morphologically verified. Survival from cancers with a high burden and amenable to prevention was poor: the 3-year age-standardised net survival was 52·3% (95% CI 49·4-55·0) for cervical cancer, 18·1% (11·5-25·9) for liver cancer, and 32·4% (27·5-37·3) for lung cancer. Less than half of the included patients were alive 3 years after a cancer diagnosis for eight cancer types (oral cavity, oesophagus, stomach, larynx, lung, liver, non-Hodgkin lymphoma, and leukaemia). There were differences in survival for some cancers by sex: survival was longer for females with stomach or lung cancer than males with stomach or lung cancer, and longer for males with non-Hodgkin lymphomas than females with non-Hodgkin lymphomas. Survival did not differ by country-level HDI for cancers of the oral cavity, oesophagus, liver, thyroid, and for Hodgkin lymphoma.

Interpretation: For cancers for which population-level prevention strategies exist, and with relatively poor prognosis, these estimates highlight the urgent need to upscale population-level prevention activities in sub-Saharan Africa. These data are vital for providing the knowledge base for advocacy to improve access to prevention, diagnosis, and care for patients with cancers in sub-Saharan Africa.

Funding: Vital Strategies, the Martin-Luther-University Halle-Wittenberg, and the International Agency for Research on Cancer.

Translations: For the French and Portuguese translations of the abstract see Supplementary Materials section.

Figure 1

Observed all-cause cancer survival by sex for all registries combined The log-rank tests the difference in survival curve by sex based on a pre-determined significance level of 5%.x

Figure 2

Observed all-cause cancer survival by country-level HDI for all included registries At the country level, Seychelles and Mauritius were categorised as high HDI, Eldoret (Kenya), Nairobi (Kenya), Namibia, and the Eastern Cape Cancer Registry (South Africa) as medium HDI, and Cotonou (Benin), Abidjan (Côte d’Ivoire), Addis Ababa (Ethiopia), Bamako (Mali), Kampala (Uganda), Bulawayo (Zimbabwe), and Harare (Zimbabwe) as low HDI. The log-rank tests the difference in survival curve by country-level HDI based on a pre-determined significance level of 5%. All cases of cancers of the larynx, Kaposi sarcoma, and corpus uteri were from the Eastern Cape Cancer Registry (South Africa). HDI=Human Development Index. NA=not applicable.

Figure 3

3-year age-standardised net survival by cancer type, Human Development Index, and registry Error bars show 95% CIs. ECCR=Eastern Cape Cancer Registry.