96 Disease-Free Interval Patients with de novo MBC are used in studies of prognosis, despite the difficulty of extrapolating results from this population to the entire MBC population, because the disease- free interval—the time between the initial diagnosis and the metastatic diagnosis—doesn’t exist in this subgroup and need not be considered. Because the length of time before breast cancer recurs has been confirmed as an independent predictive factor known to impact duration of survival, studies relying on these data can be misleading. Tevaarwerk et al.[110] demonstrated the effect of the disease-free interval in their 2013 analysis of long-term patient outcomes across 11 phase 3 adjuvant chemotherapy trials completed by the Eastern Cooperative Oncology Group over approximately 30 years (1978–2010). In this study of 13,785 breast cancer patients who received adjuvant chemotherapy, 3447 patients (25%) developed distant MBC; the overall median survival after relapse was 20 months. The factor that best predicted duration of survival was disease-free interval, which was 2.44 times higher among patients with relapse 6 or more years after initial diagnosis as compared with those with relapse after 3 or fewer years. By contrast, TN or ER− tumors (vs. ER+ tumors), any involved lymph nodes (vs. none), and black race (vs. other) were much weaker (but statistically significant) predictors of survival. In fact, when this study’s results were stratified to take disease-free interval into account, the increased survival benefit over time all but disappeared—except among ER− MBC patients who had relapse within 5 years after adjuvant treatment. The exception was probably due to the approval of trastuzumab (Herceptin) in 1998. Summary Recent studies on duration of survival of de novo and recurrent MBC generally demonstrate 3 findings: • Over the past few decades, the duration of survival after metastatic diagnosis has increased modestly—by a matter of months, not years. Hospital-based studies generally report a larger survival benefit than population-based studies. • The modest increase in survival has been observed mainly in ER+ and/or HER2+ MBC and is attributable to the wide use of targeted therapies. No survival benefit has been found in TN MBC. • The disparity between survival among black women with MBC and non-Hispanic white women with MBC appears to be increasing. According to SEER data, non-Hispanic white patients with de novo MBC have a survival benefit that is not found in black patients. It is unclear how much of the observed disparity in outcome is related to access to care and related socioeconomic concerns and how much is related to the greater incidence of TN MBC among black women. Modest increase in survival has been observed mainly in ER+ and/ or HER2+ MBC and is attributable to the wide use of targeted therapies. No survival benefit has been found in TN MBC. The disparity between survival among black women with MBC and non-Hispanic white women with MBC appears to be increasing as treatments improve.