P5-14-16: Molecular Classification in Primary Breast Cancer and Corresponding Lymph Node Metastasis Show Impaired Prognostic Profile in the Metastatic Node.

Abstract Introduction The heterogeneity of breast cancer disease has emphasized the need to optimize prediction of outcome by molecular classification of the primary tumor. Clinical management of primary breast cancer is based on biomarkers measured in the primary tumor, whereas biomarker analysis in metastatic lymph nodes is not used for clinical decision making. Recently, microarray-based expression profiling and subsequent immunohistochemical studies have identified biologically distinct subtypes using a combination of biomarkers. The present study aimed to classify primary breast cancer tumors and corresponding lymph node metastases by multiple biomarkers into Luminal A, Luminal B, HER2 type and Triple negative phenotype and compare outcome. Material and method The study is based on a cohort of patients where biomarker expression (hormone receptor status (ER, PR), human epidermal growth factor receptor 2 (HER2) and Ki67) in primary breast cancer and ipsilateral lymph node metastasis recently were reexamined and individually related to prognosis. In the present study, 85 patients of the original cohort were possible to classify into the four subtypes: Luminal A (ER+, PR+/− and Ki67 ≤ 20%), Luminal B (ER+, PR+/−, HER2+/− and Ki67 > 20%), HER2− type (ER-, HER2+) and Triple negative (ER-, PR-, HER2−). The classifications of primary tumors and corresponding lymph node metastases were compared using the McNemar-Bowker test of symmetry and the molecular subgroups were related to clinical outcome in terms of 5-years distant disease-free survival (DDFS) by Cox analysis with luminal A as the reference group. In 7 cases (16%) the biomarker classification was discordant between the primary tumor and the lymph node metastasis. All of the discordant cases shifted to a subtype with a worse prognosis according to the lymph node metastasis (p=0.06, Mc Nemar-Bowker test of symmetry). When comparing Luminal A to non-Luminal A in the primary tumor and the lymph node, the deviation from symmetry was significant, p=0.02. DDFS was worse in all subtypes compared to the prognosis for Luminal A using Cox analysis in both primary tumors and lymph node metastasis. Conclusion Our data shows discordance in classification according to a multiple molecular phenotype between primary tumors and lymph node metastasis in breast cancer patients. The discordant cases were shifted from Luminal A to a subtype where survival analysis showed an impaired prognosis compared to this subgroup. The finding that the molecular phenotype is shifted from a “good” prognostic signature in the primary tumor to a “bad” prognostic signature in the metastatic node can be of clinical importance for the choice of adjuvant systemic therapy in primary breast cancer with lymph node positive disease. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-14-16.