Modern Breast Cancer Staging Should Include Tumor Biology

February 27, 2015
Leah Lawrence
Leah Lawrence

According to a presentation at MBCC, modern breast cancer staging systems should include grade, estrogen receptor status, and possibly other biologic markers.

Kelly K. Hunt, MD, encouraged oncologists to think outside the box when it comes to an improved staging system for breast cancer.

According to Hunt, any modern breast cancer staging system should not only incorporate the traditional tumor-node-metastasis staging, but should also include grade, estrogen receptor (ER) status, and, possibly, other biologic markers.

Hunt presented information on a modern-day staging system during her presentation, “Incorporating Tumor Biology Into an Improved Staging System for Breast Cancer,” at the 32nd Annual Miami Breast Cancer Conference, held February 26–28 in Miami Beach, Florida.

Traditionally, breast cancer has been staged using the American Joint Committee on Cancer (AJCC) staging system.

“The system was first developed in 1959 and is updated periodically to bring in new knowledge but there is really very little knowledge about the biology of breast cancer that has been incorporated into the staging system,” Hunt said.

According to the 7th edition of the AJCC staging system, patients are assigned a pathologic stage after surgical evaluation of the primary tumor and regional lymph nodes based on T (the primary tumor), N (regional lymph nodes), and M (distant metastasis) stages.

However, research has shown that patient disease-specific survival within each of the assigned AJCC stages has a wide variation, begging the question, would incorporation of biologic tumor markers in the AJCC staging system result in more precise determination of patient prognosis?

Hunt and colleagues explored this question in a study of 3,728 patients with invasive breast cancer, published in 2011 in the Journal of Clinical Oncology.[1] In the study, they tested the significance of adding several factors to staging: grade, lymphovascular invasion, ER status, progesterone receptor (PR) status, combined ER and PR status, or combined ER, PR, and human epidermal growth factor receptor 2 (HER2) status. None of the survival curves separated patients by disease-specific survival when looked at on their own, Hunt said.

Therefore, the researchers assigned a value of 0 to 2 to each of the factors that were determined to be independent predictors of disease-specific survival. Variables that had a significant effect of disease-specific survival with a hazard ratio (HR) between 1.1 and 3 were given 1 point, and those with a HR between 3.1 and 6 were given 2 points. An overall staging score was then calculated by summing the scores for the individual independent predictors of disease-specific survival.

Hunt and colleagues then evaluated six different staging systems: (1) pathologic stage alone, (2) pathologic stage plus grade, (3) pathologic stage, grade, and lymphovascular invasion, (4) pathologic stage, grade, and ER status, (5) pathologic stage, grade, and combined ER and PR status, and (6) pathologic stage, grade, and combined ER, PR, and HER2 status. They compared 5-year disease-specific survival rates, Akaike information criterion (AIC), and Harrell’s concordance index (C-index) for the systems.

“Looking at the pathologic stage the curves were close together, even among patients with stage I to stage III disease, outcomes were not that different,” Hunt said. “However, when pathologic stage was added to grade and ER/PR status, the curves separated and there was a much better discrimination of outcomes of patients when you include those biologic factors.”

According to Hunt, the systems that looked at pathologic stage, grade, and ER status, and pathologic stage, grade, and combined ER/PR status had the highest C-index (0.80), with the combined ER/PR system having the lowest AIC. Unfortunately, there was a small sample size for the pathologic stage, grade, and ER/PR system, making it difficult to properly compare the top two systems.

The researchers then validated their findings in an external cohort of 26,711 patients from the Surveillance, Epidemiology, and End Results (SEER) database. Using the two best staging systems, it was obvious that the disease-specific survival curves again showed better discrimination in terms of prognosis and outcomes. In the external cohort, the model again showed a robust C-index (0.79).

Based on these data, Hunt and her colleagues recommended the incorporation of grade and ER status in addition to pathologic stage for staging of patients with breast cancer.

Finally, Hunt and colleagues looked at the incorporation of these factors into the staging of patients undergoing neoadjuvant chemotherapy at the University of Texas MD Anderson Cancer Center. In the neoadjuvant setting, patients are assigned a clinical stage initially, but pathologic stage changes because of treatment response. In the setting of neoadjuvant chemotherapy, survival curves again showed a better discrimination when grade and ER status were added to pathologic stage of patients.

“Overall, this system seems to work for patients treated with surgery first, and those treated first with neoadjuvant chemotherapy,” Hunt said.

According to Hunt, the AJCC is currently reviewing its staging system for breast cancer patients.

Reference

1. Yi M, Mittendorf EA, Cormier JN, et al. Novel staging system for predicting disease-specific survival in patients with breast cancer treated with surgery as the first intervention: time to modify the current American Joint Committee on Cancer staging system. J Clin Oncol. 2011;29:4654-61.