Abstract
Screening mammography was assessed in 9 randomized trials initiated between 1963 and 1990, with breast cancer-specific mortality as the primary endpoint. In contrast, breast cancer detection has been the primary endpoint in most screening trials initiated during the past decade. These trials have evaluated digital breast tomosynthesis, magnetic resonance imaging, and ultrasound, and novel screening strategies have been recommended solely on the basis of improvements in breast cancer detection rates. Yet, the assumption that increases in tumor detection produce reductions in cancer mortality has not been validated, and tumor-detection endpoints may exacerbate the problem of overdiagnosis. Indeed, the detection of greater numbers of early stage breast cancers in the absence of a subsequent decline in rates of metastatic cancers and cancer-related mortality is the hallmark of overdiagnosis. There is now evidence to suggest that both ductal carcinoma in situ and invasive cancers are overdiagnosed as a consequence of screening. For each patient who is overdiagnosed with breast cancer, the adverse consequences include unnecessary anxiety, financial hardships, and a small risk of morbidity and mortality from unnecessary treatments. Moreover, the overtreatment of breast cancer, as a consequence of overdiagnosis, is costly and contributes to waste in health-care spending. In this article, we argue that there is a need to establish better endpoints in breast cancer screening trials, including quality of life and composite endpoints. Tumor-detection endpoints should be abandoned, because they may lead to the implementation of screening strategies that increase the risk of overdiagnosis.