Stratification of 5-year cancer detection rate in an organized breast screening program based on Gail model risk factors.

Authors

null

Rasika Rajapakshe

British Columbia Cancer Agency, Centre for the Sou

Rasika Rajapakshe , Brent Parker , Cynthia Araujo , Stephanie Ruscheinsky , Steven McAvoy , Tanja Hoegg , Andy Coldman

Organizations

British Columbia Cancer Agency, Centre for the Sou, Population Oncology, BC Cancer Agency

Research Funding

No funding sources reported
Background: The Gail model has been validated in the United States and several European countries, but to our knowledge, it has not been validated in organized breast screening programs in Canada. The Screening Mammography Program of British Columbia (SMPBC) records participant data from a questionnaire based on Gail model parameters (which include family and personal medical history). This study investigates whether the Gail model is a valid tool to predict the breast cancer risk for the population undergoing screening mammography in the province of BC. Methods: Client information of the 223,349 British Columbian women who participated in the year 2000, along with their tumor status from 2000-2004, was extracted from the provincial database. A software program was developed to rapidly calculate the absolute 5-year Gail score from questionnaire data. Participant data was separated into .5% risk intervals and also into quintiles based on increasing Gail scores, and the mean absolute risks were compared to the actual five year rate of cancer as detected by the SMPBC. Results: Overall, goodness of fit between Gail score and SMPBC detection (E/O) across the categories can be rejected (χ2=247.9, df=9, p value < .001). The Gail model significantly underpredicts the cancer detection for risk categories up to 2%, however it provides a sufficient fit for categories 2%-4% as the E/O ratio is not significantly different from 1.0 in these intervals. For the highest risk interval, categorized as greater than 4% risk, the model significantly overpredicts cancer detection. Additionally, when presented in quintiles, the Gail model under-predicts risk in all but the highest quintile (1.77-11.43% risk range). Conclusions: Our results, based on participants of SMPBC, suggest that the Gail model significantly under-predicts cancer detection. Although this model provides a sufficient fit for women with a Gail score between 1.51-4%, it does not predict breast cancer risk accurately for low and high risk women in the Screening Mammography Program of BC.

Disclaimer

This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org

Abstract Details

Meeting

2012 Breast Cancer Symposium

Session Type

Poster Session

Session Title

General Poster Session A

Track

Risk Assessment, Prevention, Detection, and Screening

Sub Track

High Risk

Citation

J Clin Oncol 30, 2012 (suppl 27; abstr 7)

DOI

10.1200/jco.2012.30.27_suppl.7

Abstract #

7

Poster Bd #

B1

Abstract Disclosures

Similar Abstracts

Abstract

2023 ASCO Annual Meeting

Barriers, facilitators, and strategies for improving breast cancer screening in Ukraine.

First Author: Ali Dzhemiliev

Abstract

2023 ASCO Annual Meeting

Beneficial effect of repeated participation in breast cancer screening upon survival.

First Author: Robert A. Smith

Abstract

2023 ASCO Annual Meeting

Mammography in women with sickle cell disease.

First Author: Marwah Wafa Farooqui