Abstract P5-08-02: Improving the AJCC breast cancer staging system by incorporating tumor biomarkers

Abstract

Abstract BACKGROUND: The current American Joint Committee on Cancer (AJCC) breast cancer staging system provides important prognostic information, however its use is limited by the lack of data incorporating prognostic and predictive biological markers. In this study we sought to determine the relationship between stage, breast cancer subtype, grade and outcome in a large population-based cohort, and to develop a risk score point-based system incorporating biological factors to the current AJCC staging system. METHODS: Patients diagnosed with primary breast cancer stage I-IV, between 2005-2008 were identified in the California Cancer Registry. For patients with stage I-III disease, pathological stage was recorded. For those with stage IV, clinical stage was used. 5 year-breast cancer specific survival (BCSS) and overall survival (OS) rates were determined for each potential TNM combination according to breast cancer subtype. Cox proportional hazard models were used to identify independent predictors of outcome. A risk score point-based system (range 0-3 points) was created to complement the current anatomic AJCC staging system. One point was assigned for each one of the following tumor characteristics: hormone receptor (HR)-negative status, HER2-negative status and grade 3. Survival probabilities between groups were compared using log-rank test. Multivariable analysis models according to stage and risk score were performed for BCSS and OS. RESULTS: A total of 43,938 patients were included. The 5-year BCSS and OS for each TNM combination differed according to breast cancer subtype. The best outcomes were seen among HR-positive patients followed closely by those with HER2-positive and HR-positive tumors with the worst outcomes observed among patients with triple negative tumors. In a multivariable model, after adjusting for stage, treatment variables and other important confounders, ER negative status (OR 2.14; 95%CI 1.98-2.30), HER2-negative status (OR= 1.24; 95%CI1.14-1.34) and grade 3 (OR=2.03; 95% CI 1.88-2.20) were independent predictors of BCSS. Our risk score system separated patients into 4 risk groups within each stage category (all P<0.05). Similar results were seen for OS. The results in the table show that combining stage and risk score provides improved prognostic information. Hazard ratios for BCSS according to stage and risk score.Stage/Risk ScoreHazard Ratio95%CII-0Reference I-10.630.42-0.97I-22.811.87-4.23I-34.902.79-8.61IIA-03.652.26-5.91IIA-12.241.50-3.33IIA-25.873.94-8.73IIA-39.356.24-13.99IIB-04.763.13-7.24IIB-15.043.37-7.53IIB-29.465.63-15.92IIB-314.599.78-21.79IIIA-09.136.12-13.62IIIA-17.384.92-11.06IIIA-212.017.14-20.19IIIA-332.5421.59-49.04IIIB-018.9712.62-28.52IIIB-119.0512.68-28.61IIIB-228.3418.91-42.47IIIB-335.6121.33-59.43IIIC-011.525.84-22.74IIIC-115.766.07-40.94IIIC-226.7716.74-42.83IIIC-356.6336.60-87.60IV-049.9330.64-81.36IV-167.4545.36-100.23IV-2107.6272.36-60.07IV-3164.04107.18-251.04 CONCLUSIONS: Incorporating biological factors to the current AJCC staging system provides more accurate prognostic information and reflects current treatment practice. The proposed risk score improves the AJCC breast cancer staging system and should be incorporated in the upcoming revision. Citation Format: Chavez-MacGregor M, Mittendorf EA, Clarke CA, Lichensztajn DY, Hunt KK, Giordano SH. Improving the AJCC breast cancer staging system by incorporating tumor biomarkers [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P5-08-02.