Abstract P2-18-15: Primary site surgery for metastatic breast cancer: Patterns of care and cause-specific survival

Background : The impact of surgical interventions for women with a primary diagnosis of metastatic breast cancer has been debated throughout the literature. Much of the impact on survival has been attributed to a selection bias of healthier and younger individuals. Utilizing the SEER registry, we evaluate the patterns of surgical care and survival of patients with de novo Stage IV breast carcinoma with consideration of immortal time bias and potential selection bias for surgical intervention. Methods : Utilizing the SEER database we completed a retrospective, population-based cohort study of all women with a primary diagnosis of Stage IV, pathologically confirmed invasive breast carcinoma diagnosed in 16 SEER registries between 2004 and June 2009. Multivariate analysis of patterns of surgical care was performed using logistic regression, and of cause specific survival (CSS), using Cox Proportional Hazards model. Results : 13,410 patients selected from the SEER database fit the inclusion criteria for this study, of whom 40% had surgery to the primary site, 59% did not, and 0.8% had missing surgical data. On multivariate analysis, receipt of surgery was less likely with increasing age (OR = 0.99, p 5cm (OR = 0.91, p = 0.042), clinical stage T4 (OR = 0.73, p Conclusions : Surgery to the primary site is associated with increased CSS. Surgery was used less commonly in the latter years and in older patients, those with higher clinical stage, and in white collar and large metropolitan areas. Immortal time bias and inability to account for relevant prognostic factors in the metastatic setting preclude any conclusion regarding a causative role of surgery on survival. Randomized trials are necessary, however the poor early survivorship of patients with stage IV who might be destined to benefit from surgery makes accrual to such trials difficult. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-18-15.