Background: Survival in breast cancer maybe more accurately predicted by a scoring system based on the biologic characteristics of the tumor such as diameter, number of involved nodes, and hormone receptor status than by conventional staging.
Methods: Medical records of 419 patients with a diagnosis of breast cancer from a university hospital between 1997 and 2002 were reviewed. Tumor characteristics were given a value reflecting the impact on survival. Tumor diameter was assigned 1 point per cm, specifically: 1 point for 0.1 to 1 cm, 2 points for 1.1 to 2 cm, 3 points for 2.1 to 5 cm, and 4 points for 5.1 cm and over. Nodal status was assigned 1 point if the patient had 1 to 3 positive nodes, 5 points if 4 to 10 positive nodes were found, and 10 points if more than 10 nodes were identified. Hormone receptor status was given 1 point if either estrogen or progesterone receptors were not present or 2 points if both were absent. The breast cancer severity score (BCSS) was used to analyze overall and disease-free survival by Kaplan-Meier analysis. Significance of differences in survival was determined by log-rank analysis.
Results: An increased severity score was inversely proportional to overall and disease-free survivals. A BCSS of greater than 7 was most predictive of outcome with 5-year survival of 68%; and a BCSS of less than 7 had a 5-year survival of 98% (P = 0.0028). A BCSS of greater than 7 also discriminated disease-free survival within stage II (P = 0.02) and stage III (P = 0.01).
Conclusions: The BCSS provides a better resolution of outcome than traditional staging, and works both for overall survival and within specific stages. Using the BCSS, management and follow-up can be specifically tailored for low- and high-risk patients.