BACKGROUND AND OBJECTIVE: BI-RADS Category classification is the most powerful predictor of breast cancer (BC). However, BI-RADS Category 4 lesions are associated with a highly variable rate of BC. The purpose of this study was to develop and validate a nomogram for the prediction of individual probability of BC in patients with BI-RADS Category 4 lesions. METHODS: The study included all patients with BI-RADS Category 4 lesions at screening mammogram, who underwent diagnostic cytology or biopsy and, as needed, surgery or follow-up. Univariate and multivariate logistic regression analyses were used to develop the model and build the nomogram. This nomogram was evaluated on a training set of 170 patients treated at IGR Cancer Center, Paris, France. Nomogram performance was evaluated on an external independent dataset of 188 patients from MDA Cancer Center, Houston, Texas. RESULTS: A total of 51 (28.5%) patients in the training set and 73 (42.4%) patients in the validation set were diagnosed with BC. The final, most informative, nomogram included information on patient age (P = 0.04), palpable tumor (P = 0.002), menopausal status (P = 0.32), lesion size (P = 0.81), HRT (P = 0.09), and Gail risk (P = 0.58). The predictive accuracy of the nomogram was 0.716, respectively. The concordance index of the model was 0.66 in the validation set. CONCLUSION: The nomogram based on clinical and radiological findings may help inform the patients before surgical explorations, to decrease the number of missed cancer cases but currently cannot replace FNA or biopsy. 2010 Wiley-Liss, Inc.
BACKGROUND AND OBJECTIVE: BI-RADS Category classification is the most powerful predictor of breast cancer (BC). However, BI-RADS Category 4 lesions are associated with a highly variable rate of BC. The purpose of this study was to develop and validate a nomogram for the prediction of individual probability of BC in patients with BI-RADS Category 4 lesions. METHODS: The study included all patients with BI-RADS Category 4 lesions at screening mammogram, who underwent diagnostic cytology or biopsy and, as needed, surgery or follow-up. Univariate and multivariate logistic regression analyses were used to develop the model and build the nomogram. This nomogram was evaluated on a training set of 170 patients treated at IGR Cancer Center, Paris, France. Nomogram performance was evaluated on an external independent dataset of 188 patients from MDA Cancer Center, Houston, Texas. RESULTS: A total of 51 (28.5%) patients in the training set and 73 (42.4%) patients in the validation set were diagnosed with BC. The final, most informative, nomogram included information on patient age (P = 0.04), palpable tumor (P = 0.002), menopausal status (P = 0.32), lesion size (P = 0.81), HRT (P = 0.09), and Gail risk (P = 0.58). The predictive accuracy of the nomogram was 0.716, respectively. The concordance index of the model was 0.66 in the validation set. CONCLUSION: The nomogram based on clinical and radiological findings may help inform the patients before surgical explorations, to decrease the number of missed cancer cases but currently cannot replace FNA or biopsy. 2010 Wiley-Liss, Inc.
Authors: A-S Hamy; S Giacchetti; M Albiter; C de Bazelaire; C Cuvier; F Perret; S Bonfils; P Charvériat; H Hocini; A de Roquancourt; M Espie Journal: Eur Radiol Date: 2011-07-16 Impact factor: 5.315
Authors: J M H Timmers; A L M Verbeek; J IntHout; R M Pijnappel; M J M Broeders; G J den Heeten Journal: Eur Radiol Date: 2013-04-18 Impact factor: 5.315
Authors: Anne Marie McCarthy; Brad Keller; Despina Kontos; Leigh Boghossian; Erin McGuire; Mirar Bristol; Jinbo Chen; Susan Domchek; Katrina Armstrong Journal: Breast Cancer Res Date: 2015-01-08 Impact factor: 6.466