Johannesburg Cancer Study (JCS): contribution to knowledge and opportunities arising from 20 years of data collection in an African setting. (April 2020)

• Record Type: Journal Article
• Authors: Chen, Wenlong Carl; Singh, Elvira; Muchengeti, Mazvita; Bradshaw, Debbie; Mathew, Christopher G.; Babb de Villiers, Chantal; Lewis, Cathryn M.; Waterboer, Tim; Newton, Robert; Sitas, Freddy
• Is Part Of: Cancer epidemiology. Volume 65(2020:Apr.)
• Publication Date: 2020-04
• Subjects: Cancer cohort -- Johannesburg Cancer Study -- JCS -- Cancer Biosample; Cancer -- Epidemiology -- Periodicals; Cancer -- Prevention -- Periodicals; Cancer -- Diagnosis -- Periodicals; Carcinogenesis -- Periodicals; 616.994005
• DOI: 10.1016/j.canep.2020.101701
• Languages: English
• ISSNs: 1877-7821

Abstract

Highlights: Cancer cohort recruitment methodology for low resource environment. 20 years of data to identify risk factors for cancer development in black patients. 24 971 black patients with histologically proven cancer of any type were enrolled. Serum and DNA samples are available for serological and genomic analysis. Viral infections & leading lifestyle risk factors are key in cancer aetiology in SA. Abstract: The Johannesburg Cancer Study (JCS) aims were to examine whether cancer risk factors identified in Western countries applied to black patients in Johannesburg, South Africa and to understand the impact of HIV on cancer risk, with a view to identifying previously unrecognised HIV associated cancers. A total of 24 971 black patients with an incident histologically proven (>95%) cancer of any type were enrolled between 1995-2016. Response rates were >90%. Patients provided informed consent, lifestyle and demographic information using a structured questionnaire; 19 351 provided a serum sample and 18 972 a whole blood sample for genomic analyses. This is currently the largest cancer epidemiological biobank in Africa. JCS uses a cancer case-control method; controls being cancer types unrelated to exposures of interest. Published results show the importance of HIV in several cancers known to be infection associated e.g. Kaposi sarcoma (OR = 1683; CI = 595-5194) in those with high Kaposi-sarcoma-associated-herpesvirus titres; no effect of HIV on lung or liver cancer-in the latter showing a strong association with HBVDNA, sAg and c positivity (OR = 47; CI = 21-104). Comparable data to higher-income country studies include lung cancer ORs in relation to smoking (15+g tobacco/day) (ORMales = 37; CI = 21-67, ORFemales = 18.5; CI = 8-45) and associations between alcohol and oesophageal cancer in smokers (ORM&F = 4.4; CI = 3-6). Relationship between hormonal contraception declined to null 10 or more years after stopping for breast (OR = 1.1; CI = 0.9-1.4) and cervical cancer (OR = 1.0;CI = 0.8-1.2), and protective effects shown, five or more years after stopping for ovarian (OR = 0.6; CI = 0.4-1) and endometrial cancer (OR = 0.4; CI = 0.2-0.9). Preferential access is based on data requests promoting data pooling, equal collaborative opportunities and enhancement of research capacity in South Africa. The JCS is a practical and valid design in otherwise logistically difficult settings.