Benefits and risks of breast cancer screening among women with a familial or genetic predisposition: validation of a simulation model using published data
G.H. De Bock1, M.C. Jansen-van der Weide, M.J.W. Greuter, M. Oudkerk -Department of Epidemiology, University Medical Center Groningen1
Background: Women at increased risk of breast cancer are often screened with mammography at relatively young ages when the sensibility for x-rays is high and the risk of radiation-induced tumours is increased. It is important to know, whether screening with mammography at young ages provides a balanced benefit versus risk ratio. For that, we constructed a simulation model to provide accurate benefits and risks of screening scenarios. Radiation risk was calculated using data from BEIR VII phase 2 report. The purpose of this study is to validate the outcomes of our model using published data on women at increased risk of breast cancer.
Materials and Methods: A study simulating screening was conducted among a simulated population of women with breast cancer family histories. The model parameters were derived from published estimates of population incidence and relative risks. The outcomes of our model were validated by comparing these to data reported in a study on breast cancer screening among high-risk women in Italy (Cortesi, BMC Cancer, 2006).
Preliminary results: During the study period of Cortesi 5 tumours developed (95%CI: 0.85-9.15) among mutation carriers (n=48), of which 2 were found through the screening program. Among women with a family history of breast cancer (n=931) 33 tumours (95%CI: 21.9-44.1) were found, of which 23 were detected with screening. When using our model, the same screening scenario revealed 2.4 tumours among mutation carriers (SE: ±0.02) of which 1.0 was found through the screening program. Among women with a family history of breast cancer, 33.4 tumours (SE: ±0.3) developed, of which 20.0 were detected through screening.
Conclusions: Our model outcomes are comparable with the results of data published by Cortesi. Therefore, our model seems to be suitable for the provision of accurate benefit-risk ratios, useful for the refinement of the screening guidelines, concerning women at increased risk of breast cancer.