Trends in survival of stage IV breast cancer (BC) among Caucasian and African American (AA) BC patients (pts)

Abstract

1000 Background: The aim of this review was to look at the trends in survival over time in patients(pts) with de novo stage IV BC and to identify factors that impact survival. Methods: We searched the SEER registry to identify pts with de novo stage IV breast cancer diagnosed between 1988 and 2003. Pts who were male, did not have microscopic confirmation of disease, had more than one primary, and who had a prior malignancy were excluded. Pts were divided into three groups based on their year of diagnosis separated into 5-year intervals: 1988 - 1993, 1994 - 1998, and 1999 -2003. Overall survival (OS) and BC specific survival (BCS) were calculated from the date of diagnosis to the date of death, last follow-up, or 12/31/03. Survival outcomes were estimated by the Kaplan-Meier method, and Cox models were fit to determine the characteristics that were independently associated with survival. Follow-up time was censored at 3 years for Cox analyses. Results: 15,438 pts were identified. 3796 pts were diagnosed from 1988–1993, 3954 from 1994–1998, and 7688 from 1999–2003. Median age was 62 years. Median follow-up was 16 months (mos) (range 0–191), 18 mos (range 0–199), and 11 mos (range 0–59) in periods 1988 - 1993, 1994 - 1998, and 1999 - 2003 respectively. Median OS was 18 mos overall and 16, 18, and 20 mos respectively for each time period. Median BCS was 23 mos overall and 20, 21, and 25 mos for each time period. In the multivariable model, more distant year of diagnosis, Grade 3 disease, higher number of positive LN, increasing age, being unmarried, ER- disease, PR- disease, and no surgery were all independently associated with worse BCS. The interaction term between ER and year of diagnosis was of borderline significance, indicating that over time, pts with ER+ disease had decreasing risk of death compared to patients with ER- disease (HR 0.98, 95% CI 0.96–1.00). An interaction term between race and year of diagnosis was significant, such that with each increasing year of diagnosis AA pts had increasing risk of death compared to whites (HR 1.04, 95% CI 1.01–1.07). Conclusions: The survival of de novo stage IV BC pts has modestly improved over time. Of concern, survival disparities between AA and white patients increased rather than diminished over time. No significant financial relationships to disclose.