Tumour markers are widely used for screening and monitoring cancers. CA 15-3 and CEA are FDA-approved tumour markers for monitoring breast cancer. However, the prognostic efficacy of preoperative elevation of CEA and CA15-3 levels in breast cancer remains controversial.
We retrospectively analyzed the clinicopathological parameters and preoperative serum CEA and CA 15-3 level of 149,238 patients of Korean Breast Cancer Society Registry (KBCSR) Database who underwent surgery between January 2000 and December 2015.
The patients with elevated CA 15-3/CEA level had shown more advanced T stage and N stage, ER negativity, PR negativity and HER2 positivity compared the patients with normal CA15-3/CEA levels. The patients with elevated CA 15-3/CEA level had worse OS than the patients with normal CA 15-3/CEA level. In survival analysis, both elevated group had worst prognosis compared other groups. In subgroup analysis by subtypes, in luminal A, both elevated group had hazard ratio (HR) 2.14 (95% CI 1.01-4.55), only CA 15-3 elevated had HR 2.38 (95% CI 1.58-3.58) and only CEA elevated group had HR 1.79 (95% CI, 1.20-2.68), compared normal group. In luminal B, both elevated group had HR 3.99 (95% CI 2.23-7.16), only CA 15-3 elevated had HR 2.38 (95% CI 1.58-3.58) and only CEA elevated group had HR 1.79 (95% CI, 1.20-2.68), compared normal group. In HER2 subtype group, elevated CEA level was the only independent prognostic factor. However, in TNBC, elevated preoperative CEA and CA 15-3 level were not a significant prognostic factor for OS.
The elevation of preoperative tumor markers was associated with aggressive characteristic and worse overall survival. Preoperative tumor markers predicted the prognosis of the patients and showed differences according to subtypes. In luminal breast cancer, the CA 15-3 elevation has higher hazard ratio than the CEA elevation and the patients with both tumor markers showed the worst prognosis.
Clinical trial identification
Legal entity responsible for the study
Has not received any funding.
All authors have declared no conflicts of interest.