28-gene classifier model for prediction of local recurrence and distant metastasis risk in HER-2 positive operable breast cancer.

Abstract

e13018 Background: Accurate prediction of postoperative recurrence risk is very important for the formulation of appropriate treatment for breast cancer patients. However, traditional clinicopathological indicators cannot accurately assess the recurrence risk of patients. We aimed to build a novel gene-clinical variable model, which included 28 genetic profiles and clinical variable information, to assess the risk of recurrence in a more individualized and accurate manner. Methods: We retrospectively included a cohort of operable HER-2 positive breast cancer patients treated in Cancer Hospital of the Chinese Academy of Medical Sciences from 2010 to 2016. The genetic and clinical data were used to train and validate the 28-gene classifier model in predicting local recurrence and distant metastasis risk. Results: A total of 260 HER-2 positive breast cancer patients were Included in the study. All patients received standard adjuvant chemotherapy and routine targeted therapy after mastectomy. All patients were followed up for a median of 60.4 months for disease recurrence. Among them, 21 patients had distant metastasis events and 18 patients had local recurrence events. We used the 28 gene-clinical variable model to classify 260 patients into high- and low-risk groups. Kaplan-Meyer survival curves were used to compare the prognostic difference of DFS into high- and low- risk groups. Conclusions: 28 gene-clinical variable model may be a reliable toll to predict local recurrence and distant metastasis risk in HER-2 positive operable breast cancer.