Association of Clinicopathologic Characteristics with IHC-Based Breast Cancer Subtypes.

Background Invasive breast cancers have been classified by gene expression into the following major subclasses – Luminal A, Luminal B, Triple Negative and Her2+. Recently, there have been reports to characterize these subclasses of cancer based on immunohistochemistry (IHC) assays of ER, PR and HER2. In the Clinical Breast Care Project (CBCP), we have enrolled >4000 subjects with comprehensive clinicopathologic data collected using a core questionnaire and a pathology checklist. The latter contains 131 pathologic conditions, including 66 Benign Breast Diseases (BBDs). IHC assay results for ER, PR, HER2, Ki67 and p53 were available on most biopsy-confirmed invasive breast cancer tissues. This complete data set was used in this study to understand the association between subclasses and clinicopathological characteristics. Methods The study consists of a total of 419 invasive breast cancer patients classified based on IHC results for ER, PR and HER2, with a total of 240 Luminal A (ER and/or PR+, HER2-), 89 Luminal B (ER and/or PR+, HER2+), 40 Her2+ (ER-, PR-, HER2+) and 50 Triple Negative (ER-, PR-, HER2-). Clinical characteristics of age, ethnicity, menopausal status, tumor size, lymph node status, metastatic status, tumor stage, age of menarche, birth control pill usage, and history of hormone replacement therapy. Of the 66 BBDs, 21 were studied and 45 were removed due to null values. All the data were analyzed using chi-square test. Results Majority of the cases were Luminal A subtype (57.3%), followed by Luminal B (21.2%), Triple Negative (12%) and Her2+ (9.5%). Luminal A had 54% patients older than 50 yrs of age in comparison to 42% in the non-Luminal A groups (p=0.01). Tumor size distributions are also different among the subclasses (p=0.002); 77% of all Luminal A tumors were in T1, Luminal B had 59% tumors in T1, and Her2+ had higher number of tumors in T4 in comparison to T3. Luminal B subtype had more post menopausal women when compared to the rest (p=0.05). The protein expression of p53 showed a significant difference in the subclasses (p=0.03), but no significance was obtained for Ki67. A trend was detected that there was a lower lymph node positive rate in Luminal A (28%, n= 224) when compared to the rest (36%, n=167; p=0.10). These cancer subclasses had different co-occurrences with Sclerosing Papilloma (p=0.009), Juvenile Fibroadenoma (p=0.025) and Intraductal Hyperplasia (mild) (p=0.016). Triple Negative subtype also co-occurred with Apocrine Adenosis (p=0.021) and Columnar Cell hyperplasia (p=0.018), whereas Luminal B subtype co-occurred with Multiple (peripheral) Papillomas (p=0.014) and Phyllodes Tumor (p=0.004), and Her2+ co-occurred with Lactational Metaplasia (p=0.048). Discussion This study confirms reported differential association of subclasses with T-stage, menopausal status and age. We further found that in this patient population, there is a trend that Luminal A patients are less likely to be lymph node positive. Positive p53 expression is significantly associated with Luminal B, Her2+ and Triple Negative but not Luminal A subtypes. In addition, we found that several BBDs differentially associate with subtypes. This preliminary study points to new directions for further understanding of breast cancer subtypes. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 2133.