Age is the most important demographic risk factor for most of our life-threatening human malignancies (1), including breast cancer (2). Conventional wisdom among physicians has it that “old people get old tumors.” Among scientists studying the biology of aging, current explanations given for the overall increase in cancer incidence associated with aging include (3): i) longer exposure interval to carcinogens or tumor promoters, ii) increased susceptibility to malignancy due to impaired macromolecular repair and/or xenobiotic detoxification, and iii) diminished immune surveillance and stromal tissue response to malignant transformation. These explanations have led to the widely accepted paradigm that cancers associated with aging possess a “mutator” phenotype predisposing to increased genetic instability and mutational load, accelerated tumor cell proliferation, and a generally more invasive and metastatic behavior (3,4). However, clinical studies and direct evidence supporting this paradigm are lacking for specific human cancers, particularly when patient age at diagnosis is correlated with established prognostic or predictive tumor biomarkers. In fact, studies of our “oldest old,” including centenarians, indicate that cancer risk actually declines with advanced age (5,6); and this apparent cancer-aging paradox has also been observed in other mammals (7). To determine how aging affects breast cancer biology, we conducted a retrospective analysis of nearly 4,000 primary breast cancers characterized by multiple biomarkers representing surrogate measures of genetic instability, tumor growth, invasiveness and metastatic potential. While our results support the premise that breast cancer clinical behavior and biology are significantly affected by age at diagnosis, they also call into question key aspects of the current cancer-aging paradigm (8). Numerous studies indicate that breast cancers arising before age 40 are clinically more aggressive and likelier to metastasize and reduce patient survival than those arising in older patients. After age 40, there is no consistent age relationship for tumor stage (TNM staging system), nodal involvement or risk for distant metastasis. Validated prognostic and predictive tumor markers associated with risk for local, regional and systemic dissemination include the angiogenic growth factor VEGF, and secreted proteases such as urokinase-like plasminogen activator (uPA) and cathepsin-D; these show no consistent association with aging. Other biomarkers reflecting genetic instability (e.g. high nuclear grade, aneuploidy, p53-positivity) or tumor growth rate and cell turnover (e.g. mitotic or Ki-67/MIB-1 index, apoptotic index) show strong inverse correlations with patient age at diagnosis. Consistent with age-related declines in breast tumor growth rates, the proportion of newly diagnosed tumors overexpressing growth factor receptors like ErbB2/HER2 or ErbB1/EGFR also declines significantly after age 40. Since these same cell surface growth factor receptors are predictors of clinical responsiveness to receptor-targeted therapeutics like trastuzumab and some small molecule receptor tyrosine kinase inhibitors, it is now apparent that age-associated differences in breast cancer biology will determine age-specific trends in breast cancer therapy and even chemoprevention strategies– a pattern that is already emerging for the treatment of breast cancers overexpressing estrogen receptor (ER). ER-positive breast cancers are effectively managed and even prevented by endocrine treatment modalities that include estrogen deprivation (ovarian ablation, aromatase inhibitors) and the administration of antiestrogens (e.g. tamoxifen) (2). We compared breast cancers with adjacent non-malignant breast tissue and found that the level of ER overexpression in tumors from patients aged <40 to >80 years mirrors the 10-fold lower age-associated increase in ER content that occurs in non-malignant breast tissue up to age 60; thereafter, tumor ER content rises faster and reaches a near 25-fold differential over that expressed in non-malignant breast tissue by the age of 80 (9). Using the NCI/SEER database of >80,000 breast cancers diagnosed across the USA from 1992-1997 with known ER and PR status, we confirmed this strong age-associated increase in ER-positive breast cancer incidence (9). Geographic studies also indicate that while highly dependent on such population demographics as race and socioeconomic status (10), the development of ER-positive breast cancer accounts for most of the increase in population breast cancer incidence rates after age 50, suggesting that endocrine chemoprevention strategies beginning after age 40 could reduce up to 75% of all breast cancers in higher incident geographic populations.
[Proc Amer Assoc Cancer Res, Volume 46, 2005]