# Breast cancer risk prediction with a modified BOADICEA model in Danish women

## TL;DR

A modified BOADICEA model was used to predict breast cancer risk in Danish women, showing good accuracy and potential for clinical use.

## Contribution

The model integrates polygenic risk scores, lifestyle factors, and mammographic density to improve breast cancer risk prediction.

## Key findings

- The 5-year model achieved an AUC of 0.80 and correctly identified 94.8% of breast cancer cases in the highest-risk group.
- Sensitivity was 0.46 in the 10-year model for women aged 50-69.
- The model demonstrated good calibration and specificity in predicting breast cancer risks.

## Abstract

Breast cancer risk prediction approaches clinical practice. The BOADICEA risk model has been updated to consider common breast cancer risk variants, lifestyle/hormonal risk factors and mammographic density (MD).

49,494 women from the Danish Blood Donor Study were followed for up to 10 years. Modified BOADICEA risks within 5 and 10 years were calculated based on a polygenic breast cancer risk score combined with lifestyle/hormonal risk factors. MD was only known for 4608 women. Calibration was assessed by comparing observed and predicted risks. AUC and Harrell’s concordance index (C-index) were used to assess discriminative ability and sensitivities and specificities were obtained for high and low-risk groups.

Within 5 and 10 years, 367 and 617 women had breast cancer. The 5-year model achieved an AUC of 0.80 (95% CI:0.78–0.81), sensitivity of 0.34 and specificity of 0.92 for all and an AUC of 0.61 (95% CI:0.58–0.65) for the 50-69-year-aged. For this age-group, the sensitivity was 0.46 in the 10-year model. 50% of women with the highest 5-year risk predictions, identified 94.8% of those with incident breast cancers.

The modified BOADICEA risk model provided valid risks among a large retrospective cohort of Danish women.