Risk Stratification in Breast Cancer Screening

Background Breast cancer represents a significant global health burden, being the most prevalent cancer in women across 157 countries. In 2022, approximately 2.3 million women were diagnosed, leading to approximately 670,000 breast cancer-related deaths, prompting the World Health Organisation to launch the Global Breast Cancer Initiative to reduce cases by 2.5% annually from 2020 to 2040 through enhanced detection and management strategies. In the UK, breast cancer accounts for 30% of all newly diagnosed cancers in women, and estimates indicate an increase in new cases from 56,800 (2017-2019) to 69,900 by 2038-2040. The economic implications are substantial, with costs projected to escalate from £2.6- 2.8 billion in 2024 to £3.6 billion by 2034 without effective interventions. The UK’s National Health Service (NHS) aims to increase early-stage diagnoses by 2028 and prioritises advanced screening methods. The UK NHS Breast Screening Programme (NHSBSP) provides routine mammograms every three years to women aged 50-70 years, with additional enhanced screening for those assessed as high-risk. While screening offers benefits of early detection and reduced mortality, it is also associated with various harms, including discomfort, psychological distress, and false positives. Therefore, it is critical to evaluate potential revisions to the NHSBSP aimed at maximising benefits while minimising risks and ensuring health system sustainability. The current uniform screening strategy does not account for varying levels of breast cancer risk among populations. Increasingly, there is a shift towards personalised risk-based screening models, which categorise women based on their individual risk factors, including demographic, genetic, medical, and lifestyle influences. Various multifactorial risk prediction tools or models have been designed to estimate the likelihood of developing breast cancer, yet challenges arise in their applicability across different demographics. The UK NHS currently endorses several validated risk prediction frameworks, including the IBIS/Tyrer-Cuzick Model and the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm (BODICEA) Risk Model-CanRisk. Despite advancements, concerns about the effectiveness of these models remain, particularly regarding their performance in diverse population settings. A 2022 workshop by the UK National Screening Committee identified critical gaps in the evidence base supporting these models. Implementing a risk-stratified screening approach could optimise healthcare resources, enhance efficiency, and establish a cost-effective framework. A systematic analysis of existing breast cancer risk prediction models is imperative to evaluate their methodological quality and predictive performance, providing essential insights for UK policy development. Moreover, a review of economic evaluations using decision-analytic models for breast cancer screening risk stratification will allow policymakers to determine the cost-effectiveness, consequences and resource implications. Assessing the feasibility and implementation of risk-stratified breast cancer screening is beyond the scope of this work. Objectives We plan to conduct two complementary evidence summaries with the following objectives: i. To summarise and evaluate existing breast cancer risk prediction models used to stratify women eligible for breast cancer screening into risk categories (Review 1) ii. To summarise and evaluate existing decision-analytic model-based economic evaluations of risk-stratified breast cancer screening programmes (Review 2).