Comprehensive Cervical Cancer Prevention in sub-Saharan Africa

Background - Around 266,000 women die of cervical cancer (CC) annually, with more than 85% of these deaths occurring in low- and medium-income countries (LMICs). In response to the growing problem, the World Health Organization (WHO) launched the “90-70-90 targets” for 2030, based on the results of newly acquired scientific evidence that cervical cancer can be eliminated. Primary prevention with HPV vaccination has been shown to reduce the later incidence of cervical cancer through preventing sexual transmission of the HPV virus, but vaccination will still take a long time to achieve a reduction of cancer cases.

Secondary prevention as defined by the WHO, is the early detection and treatment of disease before symptoms are apparent. In the WHO guidelines, three inter-related steps are recommended to effectively identify and prevent cancer, which are to increase the awareness of cancer within the community, to implement effective screening at a reasonable cost, and to provide effective treatment of pre-cancer identified through screening.

The overall objective is to contribute to reducing mortality of CC in line with the WHO Goals. Our specific aims are to finalize and optimize our artificial intelligence (AI) algorithm for use in cervical cancer screening (research task (RT) 1.1); To assess the diagnostic performance of AI for identification precancerous and cancerous lesions and the associated costs (RT1.2), and (iii) to study the acceptability and understanding of a test-triage-treat strategy including the use of artificial intelligence technology (RT1.3).

Method - The proposed research activities will directly link with the existing clinical and associated research platform infrastructure at screening sites, and will address new research questions focusing on the second and third “WHO targets”. The approach will be conducted through the established “CC Screening Unit” at Dschang which recruits 1,500 participants annually, and a new screening center in Bafoussam. We will expand on the findings of the established infrastructure within the network, and finalize and deploy a machine learning approach to identify cervical precancerous lesions (RT1.1 and RT1.2) in clinical practice. Participant and health care worker acceptability of AI technology will also be explored though qualitative methods, with local domain-specific expert knowledge (RT1.3). To ensure sustainability of the overall screening program, we will lead transversal actions in the fields of community engagement (WP2), local capacity building (WP3) and policy engagement (WP4).Expected results and their impact for the field - The research packages build on WHO Global Cancer Initiatives and align with National Cancer Control Plans. The consortium builds upon a longstanding cooperation between Cameroon and Switzerland, and will effectuate high-quality research addressing severe knowledge gaps in cervical cancer prevention, and contribute to generating policy recommendations for effective cervical cancer prevention strategies (WP4). It will also provide academics and young scientists with comprehensive methodological and management competencies enabling them to take on long-term applied research within the framework of national strategies (WP3).