Incorporating margin status information in treatment decisions for women with ductal carcinoma in situ: a decision analysis.

To integrate margin status information into the decision to undergo radiation therapy (RT) following breast-conserving surgery (BCS) for women with ductal carcinoma in situ (DCIS). We developed a decision-analytic Markov model to project quality-adjusted life years (QALYs) for a hypothetical cohort of 55-year-old women with DCIS over a lifetime horizon treated with or without RT following BCS. We estimated the transition probabilities of local DCIS and invasive recurrences based on the margin status (free, close, or positive) from a systematic literature review. Other probability estimates and utilities were collected from the published literature. Using the conditions defined in this model, expected QALYs after BCS alone were better than those after BCS with RT under the free-margin scenario (15.72 vs. 15.58) and worse in the close-margin (15.44 vs. 15.50) and positive-margin scenarios (15.20 vs. 15.33). The probability of receiving a salvage mastectomy varied from 10 to 28%, depending on margin status and treatment. One-way sensitivity analyses showed that the optimal treatment was sensitive to patients' preferences and RT side effects. Probabilistic sensitivity analyses revealed that BCS alone would be the best strategy in 54% of the cases under the free-margin scenario, 48% under the close-margin scenario, and 44% under the positive-margin scenario. This study illustrates that margin status is able to provide supplementary information on the decision of DCIS treatment. Our analyses also highlight the importance of patients' preferences in decision making. Our findings suggest that RT is not necessary for all patients with DCIS undergoing BCS.