Purpose: To assess the trade-offs between survival and breast preservation of currently accepted approaches for ductal carcinoma-in-situ (DCIS) of the breast.
Patients and methods: Decision analysis was performed using the Markov model of hypothetical cohorts of 55-year-old white women with nonpalpable mammographic abnormalities found to be DCIS. Strategies were breast-conserving surgery (BCS), BCS with 50-Gy radiation (RT) or initial mastectomy. Recurrence rates were derived from the published literature. Main outcomes were overall, breast cancer-free, and event-free survival plus years of both breasts preserved.
Results: Using the conditions defined in this model, the actuarial survival rates at 10 and 20 years were 91.7% and 74.1% for the initial mastectomy strategy, 91.0% and 72.1% for BCS plus RT, and 89.6% and 68.2% for BCS alone. At 20 years, the initial mastectomy strategy also had a greater breast cancer-free survival rate of 74.5%, compared with 63.3% for BCS plus RT, or 46.8% for BCS alone. However, BCS alone had the highest survival rate with both breasts preserved (64.2%) compared with BCS plus RT (56.0%) or initial mastectomy (0%). Of the breast-conserving strategies at 20 years, the breast event-free survival rate (no invasive cancer or DCIS) was greater for BCS plus RT (47.2%) compared with BCS alone (28.4%). Using just survival as the primary end point, mastectomy is the optimal strategy by a small margin. However, if quality-adjusted survival is at issue, mastectomy is the choice only if the yearly reduction in quality of life due to mastectomy is less than 1%.
Conclusion: BCS with or without radiation compared with mastectomy as initial management of DCIS of the breast trades a slight decrease in survival rates for the value of breast preservation. This model should aid clinicians in matching treatments to their patients' preferences.