Abstract A1: Race has no impact on breast cancer outcome: Results of 803 patients treated at a public hospital

Introduction: African-American women (AA) have a lower incidence of breast cancer, yet a higher breast cancer mortality rate than Caucasian women (C). Whether the mortality differences are the result of more aggressive tumors in AA women or the effect of differences in socioeconomic status (SES) is debatable. A clearer understanding of the role of race requires SES to be controlled as a potential confounder, which can be achieved by evaluating outcome within a population with a high proportion of uninsured patients. Our state-run academic medical center serves as a healthcare safety net for the 715,000 uninsured residents of Louisiana. We provide compelling evidence that race is not predictive of outcome for patients with operable breast cancer.

Methods: From our prospective breast cancer database which has been maintained since 1998, we examined the data for all 803 patients with stage 0 to 3 breast cancer. All patients received standard definitive surgical care as well as appropriate adjuvant treatment. Study homogeneity was maintained by standardized treatment, surveillance, and compliance protocols. Primary endpoints were cancer recurrence and death. Statistical analysis performed included Kaplan-Meier survival analysis, log-rank test, Cox proportional hazard model, independent samples t-test, and chi-square test. A p-value ≤ 0.05 was considered statistically significant.

Results: Sixty percent of patients were AA (N= 479 patients) and the mean follow-up time for AA and C patients was 58 months. Almost 70% of patients were classified as either free care or Medicaid. There were no significant differences in tumor size (p=0.83), nodal distribution (p=0.74), stage distribution (p=0.88), or definitive surgery performed (p=0.32) between the races. However, AA tend to be younger (p=0.003) with a higher tumor grade (p<0.0001) than C. The 5-yr overall survival (OS) and disease-free survival (DFS) for the entire cohort was 81% and 68%, respectively. For node-negative disease, the 5-yr OS and DFS was 90% and 78%, respectively (p<0.0001), and for node-positive disease, it was 70% and 54%, respectively (p<0.0001). The 5-yr OS for stages 0, 1, 2, 3 was 100%, 93%, 83%, and 61%, respectively (p<0.0001), and the DFS for stages 0, 1, 2, 3 was 85%, 82%, 71%, and 43%, respectively (p<0.0001). These results were comparable with the National Cancer Database. The impact of race on outcome was as follows: The 5-yr OS for AA and C was 80% and 83% (p=0.21), respectively, and the 5-yr DFS for AA and C was 69% and 65% (p=0.19), respectively. For stage 0, the 5-yr OS was 100% for both AA and C and the 5-yr DFS was 85% for AA and 84% for C (p=0.90). For stage 1, the 5-yr OS was 91% for AA and 93% for C (p=0.41), and the 5-yr DFS was 82% for AA and 80% for C (p=0.32). For stage 2, the 5-yr OS was 81% for AA and 86% for C (p=0.18), and the 5-yr DFS was 72% for AA and 68% for C (p=0.29). For stage 3, the 5-yr OS was 59% for AA and 61% for C (p=0.67), and the 5-yr DFS was 45% for AA and 38% for C (p=0.35). On multivariate analysis, race was not an independent predictor of cancer recurrence (p=0.11) or cancer death (p=0.22).

Conclusion: In a predominantly indigent population, race had no impact on breast cancer outcome. Hence, women who were treated at our academic center with a public hospital can expect to have breast cancer outcome rivaling those reported in the literature. Further study is needed to understand the elements involved in our success.