Abstract 3717: Evaluation and comparison of breast cancer risk prediction models for Hispanic and non-Hispanic white women in the Women's Health Initiative

Abstract

Abstract Introduction: The National Cancer Institute (NCI) Breast Cancer Risk Assessment Tool (BCRAT) has not been previously validated for use in Hispanic women. We evaluated and compared BCRAT risk projections for Hispanic and non-Hispanic White (NHW) participants in the Women's Health Initiative (WHI). Methods: We used data on 6,248 Hispanic and 127,426 NHW post-menopausal WHI participants aged 50 years and older. Absolute risks of developing invasive breast cancer over the WHI main and extension studies were calculated using the BCRAT and an updated BCRAT using 1993-2007 SEER breast cancer incidence rates. We assessed model calibration by the ratio of observed to expected number of breast cancers (O/E ratio) and discriminatory accuracy with the concordance statistic (c-statistic). Results: In the WHI main study, the BCRAT underestimated the number of breast cancers in Hispanics (O/E ratio=1.18, 95% CI = 0.99, 1.40) and NHWs (O/E ratio=1.18, 95% CI = 1.14, 1.21). The updated BCRAT was well calibrated for Hispanics (O/E ratio=1.08, 95% CI = 0.91, 1.28) and NHWs (O/E ratio=0.98, 95% CI = 0.96, 1.01). The c-statistic was 0.629 (95% CI = 0.582, 0.676) for Hispanics and 0.575 (95%CI = 0.566, 0.583] for NHWs. In the WHI extension study, where women reported being screened less frequently, BCRAT overestimated risk in both Hispanics and NHWs. Conclusion: Underestimation of breast cancer risk by the BCRAT in Hispanic and NHW women was corrected by using updated SEER breast cancer incidence rates. Recalibrating the BCRAT with current SEER incidence rates and racial/ethnic specific relative risk estimates may improve the predictive accuracy for Hispanic women. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3717. doi:10.1158/1538-7445.AM2011-3717