Breast cancer risk estimates using the Gail Model in women from an annual screening program in Southern Brazil

Abstract

20045 Background: The Gail model is widely used to estimate breast cancer (BC) risk. It has been validated as a reliable risk predictor in North America, but very few studies have been done in other countries. This study intends to examine the estimated BC risk using the Gail Model in a sample of women from Southern Brazil a region with the highest BC incidence and mortality rates of the country. Methods: Lifetime and 5-year BC risk estimates were obtained for the first 1002 asymptomatic women (ages 40–69 years) enrolled in an annual BC screening program. The frequency of each of the model’s variables was recorded and compared to other studies. Information about family history (FH) included: presence of bilateral BC, male BC, first degree relative with BC and/or ovarian cancer (OC), relative with BC under age 50 and ≥ 2 relatives with either BC, OC or colorectal cancer (CRC). Other potential risk factors for BC such as body mass index and smoking were recorded. Results: Mean (± SD) values for age, age at menarche and age at birth of the first live child were 50.4 (± 7.75), 13.0 (± 1.80), and 21.6 (± 5.00) years, respectively. Only 50 (5.0%) women were nulliparous and 62 (6.2%) reported their first live birth after age 30. History of at least one first-degree relative (FDR) affected with BC was reported by 52 (5.2%), and 31 (3.1%) had a previous breast biopsy. The mean estimated BC risk in 5 years was 0.92% (± 0.49); for those under age 60, 24 (2.8%) had an estimated 5-yr risk over 1.66%. The mean estimated lifetime BC risk was 7.80 (± 3.2). Interestingly, a history of cancer in a FDR was reported by 32.6% of the women, and evidence of familial BC was observed in 20.4%. The estimated BC risk using the Gail model was significantly higher in women with a family history of BC < 50ys and with ≥ 2 relatives with either BC, OC or CRC. Smoking and overweight were reported by 28% and 65% of the women, respectively. Conclusions: BC risk estimates obtained with the Gail Model did not differ significantly from those described in other populations. However, specific findings in cancer FH were associated with higher risk estimates. The importance of FH and overweight will be further explored in a larger sample. Population-specific risk factors for BC should be sought in different communities to ensure proper risk estimates. No significant financial relationships to disclose.