Until recently, the interplay between hereditary and environmental factors in the causation of colorectal cancer in sub-Saharan Africa has been poorly understood. We have carried out a series of collaborative studies between the Department of Medicine, University of Zimbabwe, and the MRC/UCT Research Unit for Genomic and Precision Medicine, Division of Human Genetics, University of Cape Town, whose findings have a major impact on screening strategies for colorectal cancer among Africans. In brief, we found that the incidence of colorectal cancer among Africans is rising, and young individuals are over-represented. The major risk factors are familial, diabetes mellitus, prior schistosomiasis, and a shift away from traditional dietary patterns. The prevalence of Lynch syndrome among Africans is approximately 3.3%, and identification of the pathogenic mutations is feasible with collaborative approaches. (We identified two; MLH1 c.(1896+1_1897-1)_(*193_?)del and MSH2 c.2634G>A). We councluded that the primary prevention of colorectal cancer should focus on maintenance of traditional dietary practices. Screening should be considered in individuals with diabetes mellitus above the age of 50 years, and in those with colorectal cancer in first degree relatives, possibly using faecal immunochemical tests. Identifying cases of Lynch syndrome is feasible and can guide screening among families. In particular, resources for immunohistochemistry for mismatch repair proteins should be availed, and from our findings, this can be targeted at all individuals with colorectal cancer under the age of 50 years. Finally, BRAF V600E testing has a limited role in clinical practice in Africa.

BACKGROUND The Eldoret Cancer Registry (ECR) is both a Population–based and Hospital-based Cancer Registry located in Uasin Gishu County in the Western region of Kenya. ECR is a member of the African Cancer Registries Network (AFCRN). As a hospital-based cancer registry (HBCR), ECR collects and processes all cancer cases that are diagnosed, referred to, or managed at the Moi Teaching and Referral Hospital (M.T.R.H.). MTRH is the second largest public referral hospital in Kenya, and has robust cancer treatment and diagnostic facilities as well as a wide range of medical specialists. MTRH is an accredited National Health Insurance Fund (NHIF) hospital, therefore, patients can access cancer management through this affordable insurance cover. Over the years, through donor funding at the AMPATH Oncology and Haematology, needy cancer patients have been supported on NHIF enrolment, cancer treatment and management. The catchment population for MTRH is estimated at 20 million.
The aim of this study is to design evidence-based methods for determining the establishment of new population-based cancer registries in Western region of Kenya, and Africa in general, by using hospital-based cancer registry data, from national cancer treatment facilities as a critical indicator of priority by cancer control programs.
METHODS Case finding by ECR is an active process. It is the systematic process by which cancer cases eligible for registration or inclusion in the database are identified. The E.C.R seeks clearance to collect data from the various laboratories, screening programs, cancer center, medical records areas, inpatient and outpatient facilities in MTRH. Cancer registrars visit these units on a regular basis to abstract and update cancer data into CanReg5 database. Analysis of ECR data is mainly done using CanReg5 data management software.
RESULTS 20 years of hospital-based data was analyzed. Out of a total of 23,829 cases, Uasin Gishu County registered 41% of all cancers. This is where there is a structured population-based cancer registry, ECR. Of the 23,829 cases, 14, 101 cases were from other counties in western region of Kenya, as follows: Nandi (13%), Bungoma (11%), Kakamega (11%), Trans Nzoia (9%) and Elgeiyo Marakwet (7%). Busia, Baringo, Kericho, Kisii, Kisumu and West Pokot Counties all had between 6% - 4% of cases, while other counties registered less than 3% of all cancer cases registered for the period of analysis.
CONCLUSION Notably, there was a high number of cancer patients registered from counties bordering Uasin Gishu County. The assumption is that, due to Socio-economic factors, Cultural factors, Preferences, Health status, Distance, among other factors, not all cancer patients would be able to access the MTRH. It is highly likely that these counties have more cancer incidence, prevalence and mortality. Hospital-based Cancer Registration data from difference large cancer facilities can therefore be analyzed to inform prioritizing needs in establishment of new PBCR and encourage targeted concerted efforts by Cancer Control Programs.
Despite this evidence, counties and countries through Ministries of Health, must be ready and willing to allocate resources to supporting the establishment of PBCRs in their population. Without systematic cancer data collection, processing and analysis, accurate interventions and planning for screening programs, staffing, diagnostics, treatment and management would be virtually impossible.