Bridging Age Gap

Bridging Age Gap provides prognostic survival probabilities to support primary treatment decisions for older women (aged ≥70 years) with oestrogen receptor-positive early breast cancer by comparing surgery plus adjuvant endocrine therapy versus primary endocrine therapy.

Key Features:

• Target population: Older women aged 70 years or above with oestrogen receptor-positive (hormone receptor-positive) early breast cancer.
• Predicted endpoints: Separate prediction of breast cancer-specific mortality and other-cause mortality.
• Treatment comparison: Direct comparison of survival probabilities for surgery followed by adjuvant endocrine therapy versus primary endocrine therapy.
• Development dataset: Model developed using 23,842 women diagnosed between 2002 and 2010 in the English Northern and Yorkshire and West Midlands regions.
• External validation: Validation performed on 14,526 patients from the Eastern Cancer Registration and Information Centre.
• Calibration: Reported good overall calibration with predicted versus observed mortality differing by less than 1% at 2-year and 5-year intervals.
• Discrimination: Achieved a concordance (c-) index of 0.75 or above across mortality measures.
• Statistical comparison: Reports predicted versus observed event counts and significance testing (example at 5 years: breast cancer-specific mortality predicted 2629 vs observed 2556, P = 0.142; all-cause mortality predicted 6399 vs observed 6320, P = 0.583).
• Subgroup performance: Model performance varies slightly across clinical subgroups.

Scientific Applications:

• Clinical decision support: Inform primary treatment selection between surgery plus adjuvant endocrine therapy and primary endocrine therapy for older women with oestrogen receptor-positive early breast cancer.
• Personalized prognosis: Provide individualized survival probabilities that combine breast cancer-specific and other-cause mortality to guide risk–benefit assessment.
• Oncology and geriatrics research: Evaluate outcomes and treatment effects in older patient cohorts using registry-derived prognostic estimates.

Methodology:

Models were developed using registry data from 23,842 women (2002–2010, Northern and Yorkshire and West Midlands), fitted to predict breast cancer-specific and other-cause mortality and integrated to derive treatment-specific survival probabilities, then externally validated on 14,526 patients from the Eastern Cancer Registration and Information Centre with calibration and discrimination (c-index ≥0.75) assessed and predicted versus observed counts compared with reported P-values.