Abstract
Background: Mammographic density (MD) is a well-established breast cancer risk factor. Decreasing MD has been demonstrated to be a risk-reduction biomarker in a landmark chemoprevention study in high-risk women. A few small studies suggest that a decreasing MD in women taking adjuvant endocrine therapy (AET) may be associated with a lower risk of recurrence and breast cancer specific death.
Aim: This interim analysis aims to measure inter and intra-reader variability of MD classification within a current study (MEDICI) investigating whether a reduction in MD after one and/or three years in women taking AET is associated with a lower risk of breast cancer recurrence and death compared to no reduction.
Methods: Retrospective cohort study, generated from a subset of participants from the Mammo-50 trial. Eleven readers assessed MD using a 0 to 100% visual analogue scale (VAS), scoring mammograms from time of diagnosis and at one and three years after diagnosis. A subset of four readers re-analysed the mammograms. Inter and intra- reader agreement were estimated using intra-class correlation coefficients (ICC).
Results: Eleven readers classified MD in 50 patients at three timepoints (150 mammograms), using the VAS. Four readers re-interpreted all 150 mammograms. The ICC between all 11 readers (inter-reader agreement) was 0.80 (95% confidence interval (CI): 0.77-0.84). The ICC within readers (intra-reader agreement) was 0.86 (95% CI 0.77-0.91).
Conclusion: Interim analysis demonstrates good inter- and intra-reader agreement for assessment of MD on VAS, suggesting validity of the human MD assessment to be performed within the MEDICI study.
Aim: This interim analysis aims to measure inter and intra-reader variability of MD classification within a current study (MEDICI) investigating whether a reduction in MD after one and/or three years in women taking AET is associated with a lower risk of breast cancer recurrence and death compared to no reduction.
Methods: Retrospective cohort study, generated from a subset of participants from the Mammo-50 trial. Eleven readers assessed MD using a 0 to 100% visual analogue scale (VAS), scoring mammograms from time of diagnosis and at one and three years after diagnosis. A subset of four readers re-analysed the mammograms. Inter and intra- reader agreement were estimated using intra-class correlation coefficients (ICC).
Results: Eleven readers classified MD in 50 patients at three timepoints (150 mammograms), using the VAS. Four readers re-interpreted all 150 mammograms. The ICC between all 11 readers (inter-reader agreement) was 0.80 (95% confidence interval (CI): 0.77-0.84). The ICC within readers (intra-reader agreement) was 0.86 (95% CI 0.77-0.91).
Conclusion: Interim analysis demonstrates good inter- and intra-reader agreement for assessment of MD on VAS, suggesting validity of the human MD assessment to be performed within the MEDICI study.
Original language | English |
---|---|
Publication status | Published - Nov 2024 |
Event | British Society of Breast Radiology Annual Scientific Meeting 2024 - Hilton Hotel, Liverpool, United Kingdom Duration: 3 Nov 2024 → 5 Nov 2024 https://www.delegate-reg.co.uk/bsbr2024/ |
Conference
Conference | British Society of Breast Radiology Annual Scientific Meeting 2024 |
---|---|
Abbreviated title | BSBR 2024 |
Country/Territory | United Kingdom |
City | Liverpool |
Period | 3/11/24 → 5/11/24 |
Internet address |