PD02-02: A Decision Analysis of Contralateral Prophylactic Mastectomy in Women Undergoing Treatment for Sporadic Unilateral Breast Cancer.

Abstract

Abstract Background: The intent of contralateral prophylactic mastectomy (CPM) is to improve survival after a diagnosis of unilateral breast cancer by reducing the risk of contralateral breast cancer (CBC). CPM rates are rising among women with sporadic breast cancer, despite limited evidence that its benefits outweigh its harms. Although CPM is highly effective in reducing a woman's risk of CBC, the competing mortality risk from a patient's index breast cancer may offset its benefits. Furthermore, any examination of CPM needs to consider quality of life effects. Methods: We developed a Markov decision analytic model to estimate the effect of CPM in women with newly diagnosed unilateral breast cancer. The primary outcomes examined were gains in life expectancy (LE) and quality-adjusted life expectancy (QALE) for CPM compared with no CPM in 18 hypothetical cohorts of 45-year old women. Data from the British Columbia Cancer Agency (BCCA) was used to generate AJCC stage and molecular subtype-specific estimates of the risk of developing distant metastases from an index breast cancer. A correction factor was applied to account for the omission of relevant systemic therapy (including trastuzamab) in some women in the BCCA cohort. Additional model parameters, including utilities (quality of life weights) for breast cancer and CPM health states, were identified from the published medical literature. LE and QALE estimates were not discounted in the base case. Univariate sensitivity analysis was used to examine the impact of plausible variation in the key model parameters on model results. Results: CPM improved LE in all cohorts (range: 0.06 - 0.54 years, Table 1). AJCC stage had more effect on LE than molecular subtype (stage I mean, 0.43 years, stage III mean, 0.11 years). However, after adjusting for quality of life, a strategy of no CPM was favored in all cohorts. Univariate sensitivity analysis demonstrated that the only model parameter that influenced the outcome of QALE was the utility for health after CPM. In the base case the utility after CPM was 0.81 (compared to 0.85 for No CPM). The preferred strategy did not change from No CPM to CPM unless the utility after CPM exceeded 0.83. Model results were otherwise stable across the ranges of the key parameters examined, including the risk of distant metastases resulting from a patient's index breast cancer by stage and subtype, duration of survival with metastatic breast cancer, and the risk of CBC. Conclusions: The primary drivers of survival after unilateral breast cancer are stage at diagnosis and molecular subtype. Our model demonstrates that CPM confers modest additional LE gains, even in women with early-stage, favorable-subtype breast cancer. Furthermore, this modest benefit is negated if one assumes a small reduction in quality of life due to CPM. The decision to pursue CPM as part of treatment of unilateral breast cancer should include consideration of both patient specific breast cancer characteristics and individual preferences. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr PD02-02.