No abstract

No abstract

Background: Breast cancer patients comprise a prognostic heterogeneous group. Stroma-tissue surrounding cancer cells plays an important role in tumor behavior. The prognostic value of the tumor-stroma ratio was evaluated.Material and Methods: A cohort of 574 patients with early breast cancer, primarily treated by surgery between 1985 and 1994 was analyzed. The percentage of intra-tumoral stroma was visually estimated on haematoxylin-eosin-stained histological sections from primary tumor tissue by two observers. Patients with more than 50% intra-tumor stroma were quantified as stroma-rich and patients with less than 50% intra-tumor stroma as stroma-poor.Results: Of all tumors, 68% were classified as stroma-rich and 32% as stroma-poor. Cohen's kappa coefficient revealed an almost perfect agreement in classification (kappa=0.85; 94% concordance in classification) between the two observers. For the total group of patients, stroma-rich tumors had a shorter relapse free period (RFP) (Hazard ratio (HR) 1.62; 95%CI 1.23-2.13; p=0.001) and overall survival (OS) (HR1.29; 95%CI 1.03-1.60; p=0.025) compared to stroma-poor tumors. Tumor-stroma ratio was an independent prognostic parameter for all patients (p<0.001) and in stratified analysis for patients that received systemic treatment. Importantly, within the triple-negative-cancer (TNC) subpopulation, patients with stroma-rich tumors had a 3.19 times higher risk of relapse (p=0.003) compared to patients with stroma-poor tumors, independently of grade and lymph node status. Five year RFP-rates for patients with stroma-rich compared to stroma-poor tumors were 56% and 81% respectively.Discussion: Tumor-stroma ratio was an independent prognostic factor for RFP in breast cancer patients, especially in TNC. The tumor-stroma ratio is easy to apply, of low cost and reproducible. It is a candidate parameter that can easily be implemented in routine daily diagnostics to optimize risk-stratification for breast cancer patients and should therefore be considered to be implemented in standard pathology reports. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 3049.

No abstract

No abstract