Current overview of breast cancer risk assessment models - practical application in outpatient gynecological practice.

Abstract

The rising incidence of breast cancer represents a serious medical problem. Women diagnosed with breast cancer before the age of 45 - that is, those who were not included in mammography screening programs - constitute a specific patient group. Identification of risk factors for the pathogenesis of breast cancer has led to the development of triage models based on both medical history data and genetic testing, as well as imaging methods; their updated versions are based on a combination of these modalities. The oldest empirical Gail model relies solely on basic medical history. The IBIS model, on the other hand, incorporates an expanded family history, as well as mammographic breast tissue density. The BOADICEA model enables the calculation of lifetime risk and risk of being a carrier of BRCA1/2 and other major mutations (BRCA1/2, TP53, PTEN, CHEK2, ATM). The Pecný model represents a Czech contribution. With new possibilities in genetic testing, the latest versions of these models now include the polygenic risk score (PRS), which increases their predictive value. With the growing integration of artificial intelligence (AI) and deep learning into clinical medicine, we can expect the emergence of new, AI-dependent triage models. A major limitation of the aforementioned models is their restriction to the Caucasian population. Calibration based on statistical data for other populations often does not work. This article will present current statistical data for the GAIL model, IBIS and its variants, and the BOADICEA model. The aim of this article is to present predictive models to the professional community and to highlight their potential applications in routine gynecological outpatient practice.