BRIDGES has completed the analysis of multigene panel sequencing of a very large case-control cohort with >110,000 individuals, the largest case-control study to date. The results identified nine genes as clinically most useful for inclusion on panels for breast cancer risk prediction. The study-size allowed us to provide the narrowest confidence intervals on risk estimates so far, further reducing uncertainty during risk counselling in the clinic. In addition, we were able to improve insights into breast tumour subtype-specificity of the risks conferred by some genes. Since it is known that some subtypes have a poor prognosis (such as oestrogen receptor-negative breast cancer) and require other treatments, this information could become relevant when considering preventive options. Finally, we were able to demonstrate that (subsets of) missense variants are associated with risk for some genes.
The BOADICEA breast cancer prediction model, realised by the joint efforts of BRIDGES and B-CAST EU-programs, was externally validated and prospectively calibrated. The online web-interface for BOADICEA was developed into a user-friendly version called CanRisk and successfully launched (
https://canrisk.org/(si apre in una nuova finestra)). The tool received CE marking from the EU in December 2019.
We have surveyed the use of the CanRisk webtool by healthcare providers, mostly clinical geneticists, early and late in the BRIDGES project. The results show that uptake has increased, not only in number of users, but also in the number of countries where MDs are using it.
Laboratory assays to support the pathogenic classification of variants of uncertain clinical significance (VUS) have been developed for all genes with good evidence for being associated with breast cancer. These assays have examined rare missense variants and potential splice-variants revealed by BRIDGES. An encyclopaedia of ex vivo validated alternative mRNA-splice products has been generated for all DNA repair genes in the BRIDGES gene-set. In addition, a novel in silico tool for variant classification was developed with improved accuracy relative to other publicly available tools.
We prepared validated questionnaire instruments for the identification of patients with psychological distress after genetic counselling for familial breast cancer, as well as a template for psychological intervention. These were used in a prospective cohort of 450 counselees from France, Germany, Netherlands. The results revealed the psychosocial impact on women from breast cancer families upon being given a personalised risk estimate based on CanRisk, as well as how their medical counsellors perceive this.
An evidence-based patient decision aid for the uptake of preventive measures was designed for BRCA1 and BRCA2 carriers in the German language. For women at moderate risk of breast cancer, information leaflets were conceived, as the evidence-base for these genes was still too weak to define optimal clinical management decisions.
Finally, an e-learning tool for health care providers, based on the concept of shared decision-making, was designed to train MDs how to use CanRisk (in German language).
