Background: The American Joint Committee on Cancer (AJCC) breast cancer pathological prognostic stage, which incorporates biologic factors, was developed using data from patients undergoing upfront surgery, and its application in patients receiving neoadjuvant chemotherapy (NAC) is unknown. We previously developed the Neo-Bioscore, incorporating clinical and pathological TNM categories with biologic factors, to improve the prognostic stratification of NAC patients.
Objective: This study was undertaken to evaluate the use of available staging models incorporating biologic factors in NAC patients.
Methods: Patients treated with NAC between 2005 and 2012 at MD Anderson (n = 2363) were staged using the Neo-Bioscore and the AJCC 8th edition: (1) clinical anatomic stage; (2) pathological anatomic stage; (3) clinical prognostic stage; and (4) pathological prognostic stage. Five-year disease-specific survival (DSS) and overall survival (OS) rates, along with Harrell's concordance index (C-index), were estimated. A National Cancer Database (NCDB) cohort (n = 12,887) treated with NAC between 2010 and 2013 was used for validation.
Results: In the MD Anderson cohort, staging systems incorporating biologic factors better predicted DSS (bias-corrected C-index: pathological prognostic stage = 0.8026; Neo-Bioscore = 0.7483) and OS (bias-corrected C-index: pathological prognostic stage = 0.7780; Neo-Bioscore = 0.7260) than those using anatomic factors only. Similar results were seen in the NCDB cohort. In pairwise comparisons, the pathological prognostic stage was significantly better (p < 0.0001) than other staging systems in all comparisons except for OS in the NCDB cohort, where it was not significantly different than the Neo-Bioscore (p = 0.2).
Conclusion: Biologic factors are important for determining prognosis in patients receiving NAC. These data indicate that the 8th edition AJCC pathological prognostic stage is applicable in these patients.