Temporal associations between prognostic indicators and overall survival after breast cancer.
Plichta, JK; Thomas, SM; Fayanju, OM; Rosenberger, LH; Park, TS; Hyslop, T; Greenup, RA; Hwang, E-SS
Published in: Journal of Clinical Oncology
e18144 Background: Breast oncologists have little guidance on predicting long term outcomes after interval survival. We aim to evaluate the association of overall survival (OS) with select factors at multiple time points. Methods: Women ages 18-80 with non-metastatic invasive breast cancer were identified from the NCDB (2004─2014). Using multivariate survival analysis, we estimated the association of OS with tumor and treatment factors for those with follow up surviving at least 2, 5, and 10y post-diagnosis. Results: 685598 women (median age 59) were identified; 573757 were alive with follow up at 2y, 254303 at 5y, and 18640 at 10y. The majority had early stage, hormone receptor (HR) +, invasive ductal carcinoma. 57% underwent lumpectomy; 49% received chemotherapy; 89% of lumpectomies received radiation (RT); and 83% of HR+ tumors received endocrine therapy (ET). Of those alive at 2y, improved OS was associated with ER+, PR+, chemotherapy, RT, and ET. Reduced OS was associated with higher cT/cN stage, grade, and mastectomy. Of those alive at 5y, improved OS was associated with chemotherapy, RT, and ET. Reduced OS was associated with higher cT/cN stage, grade, and mastectomy. Of those alive at 10y, reduced OS was associated with higher cT stage and grade. Conclusions: Tumor and treatment factors are associated with OS, but the association of some factors may change for women surviving at least 5y or 10y. These findings may contribute to more tailored prognostication for patients based on interval survival. [Table: see text]
