Early Diagnosis of Breast Cancer by Identifying Malignant Cells Within Neoplasias Histologically Classified as Benign
Abstract
Current diagnostic tools permit the classification of breast neoplasias into categories that represent different relative risks of developing cancer, but they do not indicate which particular lesion of an individual will lead to cancer. The lack of precise classification of breast lesions results in patients being under or overtreated. In three-dimensional (3D) cell cultures that mimic different breast phenotypes, we can correlate specific distributions of certain nuclear proteins to a particular cell behavior and/or degree of malignancy using an automated image analysis, referred to as local bright feature (LBF) analysis. We have proposed the novel concept that in true precancerous diseases, a small fraction of malignant cells might be detectable admixed within a cell population globally classified as benign. Using 3D cell culture, we show that the LBF analysis of proteins NuMA and H4K20m distinguishes premalignant from normal and malignant cells and that subpopulations of cells with different behaviors can be identified within a premalignant cell population. We will now use the LBF method to identify cell subpopulations within premalignant and preinvasive lesions on archival biopsy sections. The ability to recognize malignant cells in these lesions should bring critical improvement for the identification of lesions of adverse and good prognoses.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 2005
- Accession Number
- ADA443178
Entities
People
- Sophie A Lelièvre
Organizations
- Purdue University