Tissue microarrays for comparing molecular features with proliferation activity in breast cancer

Tissue microarrays (TMAs) are potentially suited to find associations between molecular features and clinical outcome. Enhanced cell proliferation, as measured by Ki67 immunohistochemistry, is related to poor patient prognosis in many different tumor types. Ki67 expression shows considerable intratumoral heterogeneity. It is unclear if the TMA format is suitable for the analysis of potentially heterogeneous markers because of the small size of TMA spots. We have analyzed a breast cancer TMA containing 2,517 breast tissues, including 2,222 neoplastic and 295 normal or premalignant samples, for Ki67 labeling index (Ki67 LI) and additional markers with a known relationship to Ki67 LI by immunohistochemistry (ER, PR, Bcl-2, Egfr, p16, p53) and Fluorescence in situ hybridization (HER2, MDM2, CCND1, MYC). A high Ki67 LI was linked to tumor phenotype including grade (p < 0.0001), stage (p < 0.0001), nodal stage (p = 0.0018), and patient prognosis (p < 0.0001), elevated protein levels of p53, p16 and Egfr, reduced levels of Bcl2, ER, and PR (p < 0.0001 each), as well as amplifications of HER2, MYC, CCND1 and MDM2 (p < 0.0001 each). In summary, all expected associations between Ki67 and the analyzed molecular markers could be reproduced with high statistical significance using a TMA containing only one tissue sample per tumor, measuring 0.6 mm in diameter. We conclude that associations with cell proliferation can be reliably analyzed in a TMA format.