Validity of mouse mammary tumour models for human breast cancer: comparative pathology.

In March 1999, a panel of distinguished pathologists was convened by the U.S. National Institutes of Health Breast Cancer Think Tank to develop a classification of breast lesions based on their examination of 39 models of Genetically Engineered Mice (GEM) associated mouse mammary cancer (Cardiff et al., 2000). The meeting, in Annapolis, Maryland, resulted in a published summary report from the Pathology Panel (Cardiff et al., 2000). The Annapolis consensus report, developed from the Panel's deliberations, pointed out that the mammary lesions of GEM were different from most (spontaneous) mouse mammary tumors and could be divided into three distinct categories: (1) lesions that resemble those found in spontaneous mouse mammary tumorigenesis, (2) lesions that have a unique "signature" tumor phenotype that was specific for the transgene, and (3) lesions that resemble those found in human breast diseases (Cardiff et al., 2000). This review emphasizes the proposed nomenclature and the differences between the models and human breast cancer with the intention of stimulating discussion and the development of new models.