Prognostic Factors in Triple Negative Breast Carcinomas: Is Sub-Classification Using Basal Markers Warranted?
MJ Nystrom, ES Hwang, A Griffin, W Devine, A Lal, I Mehdi, S DeVries, F Waldman, JT Rabban, YY Chen. University of California, San Francisco, SF, CA
Background: While most triple negative (ER-/PR-/HER2-, TN) breast carcinomas demonstrate aggressive clinical behavior and high mortality, a subset do not. Although heterogeneity of morphology and biomarkers exist within the TN family of breast cancer, it is not well established whether any such features stratify risk. The subset of basal phenotype TN cancers (defined as CK5/6, CK14 or EGFR+ TN cancers) has received particular attention. In this study we examined whether morphology and biomarkers can distinguish high-risk and low-risk patients in TN breast cancer.
Design: The study population consists of 104 primary TN tumors treated and followed at our institution between 1998-2009 with slides and blocks available from which a tissue microarray (TMA) was constructed. Pathologic features were reviewed. Biomarker expression was evaluated using immunostaining for androgen receptor, mammaglobin, gross cystic disease fluid protein, CK5/6, CK14, EGFR, Ki67, p53, retinoblastoma protein (Rb) and p16. Statistical analysis was performed using STATA statistical software.
Results: All patients were female with median age at diagnosis of 57 years (range 25-92 years). The group was comprised of 92 ductal cancers, 4 each of metaplastic and adenoid cystic-like cancers, and 2 each of lobular and small cell carcinomas. 41% of cases were node-positive. 79/104 cases (76%) were basal-type based on the above criteria; 16/104 (15%) were non-basal and 9/104 (9%) were uncategorized. Median follow up time for the cohort was 38 months (95% CI 29-47 months). Univariate analysis showed that p16 conferred an improved disease-free survival (DFS), whereas LVI and number of positive lymph nodes were associated with lower likelihood of DFS (OR 4.2, p=0.01 and 1.3, p=0.007 respectively). Basal subtype, age, grade, Ki67 labeling index, and Rb were not predictive of DFS. On multivariate analysis including basal subtype, p16, LVI, number of positive nodes, age, grade, and tumor size, only tumor size and number of positive lymph nodes remained independent predictors of breast cancer-specific survival.
Conclusions: Independent prognostic markers of TN breast cancer are tumor size and nodal status. There is no prognostic value in subclassifying TN cancers using basal markers, Ki67, p53, RB or p16. While the basal phenotype may be of interest from pathogenetic perspective, it does not appear to carry clinical significance.
Tuesday, March 23, 2010 9:30 AM
Poster Session III # 54, Tuesday Morning