Improved use of clinical variables for prognosis of distant recurrence in patients with ER+ breast cancer treated with 5 years endocrine therapy.

566Background: We have previously reported the development and prognostic performance of the Clinical Treatment Score (CTS) within the transATAC cohort of 1125 patients. In the old model, tumour size was categorised into three and nodal status into two categories, and it became clear that prognostic information was lost. The aim of this analysis was to re-model the CTS by allowing for finer gradation of tumour size and number of positive nodes in postmenopausal women with ER+, HER2-negative disease treated with tamoxifen or anastrozole. Methods: 1009 postmenopausal women with ER+ and HER2-negative breast cancer were included. Women with HER2-positive disease were excluded due to non-applicable prediction value for this subgroup. Distant recurrence was the primary endoint. Proportional hazard models were used to re-estimate the new CTS and Akaike Information Criterion (AIC) was used to compare models. Results: The new CTS model included five categories for nodal status, a continuous term for tumour size, a...