Abstract P6-09-06: Balancing the harms and benefits of radiation therapy for DCIS: A decision analysis examining the risk of radiation-associated sarcoma

Abstract

Abstract INTRODUCTION: More than 60,000 women are diagnosed with ductal carcinoma in situ (DCIS) annually and offered the option of breast conserving surgery (BCS), often including radiation (RT) to reduce local recurrence. Although the incidence of radiation-associated sarcoma (RAS) is low (0.05-0.25% at 10 years), the low mortality associated with DCIS and large number of DCIS diagnoses means that an increasingly large number of women are at risk of RAS. This study sought to weigh the risk of RAS with the benefits of BCS+RT for DCIS. METHODS: A second-order Monte Carlo micro-simulation model of women ages 35 and older with DCIS was constructed. The decision analysis compared harm-benefit ratios of sarcoma-related deaths per breast cancer deaths averted within 20 years of treatment with BCS+RT versus BCS alone. Stratified analyses were performed by age group to account for differential life expectancy. To generate parameter estimates for model inputs, Bayesian network meta-analysis was used to synthesize rates of DCIS and invasive recurrence from clinical trials of BCS+RT and BCS alone using a Weibull specification. Sarcoma incidence was estimated non-parametrically using SEER. Constant hazard rates for breast cancer mortality after invasive recurrence and RAS mortality were estimated from clinical trials. To account for uncertainty, probabilistic sensitivity analysis was conducted using 10,000 Monte Carlo samples and 95% credible intervals (CrI) were constructed for event rates and harm-benefit ratios. RESULTS: The micro-simulation model of an age-distributed cohort demonstrated that 1 in 840 women with DCIS (95%CrI 1:648 to 1:3522) would develop RAS within 20 years after treatment with BCS+RT. Overall, there would be 1 RAS-related death for every 12 breast cancer deaths averted (95%CrI 1:7 to 1:19) by the addition of RT to BCS. Stratified analysis demonstrated that the harm-benefit ratio was higher in women <75 years of age, with more RAS-related deaths caused per breast cancer deaths averted. The model was most impacted by parameter estimates for rates of invasive recurrence, breast cancer mortality after invasive recurrence, and RAS incidence rates. CONCLUSIONS: The risk of developing a RAS following BCS+RT for DCIS should not be overlooked. This may be especially true for women at low risk of recurrence and younger women (<75 years in our model). These findings contribute to the ongoing conversation about consequences of overtreatment of DCIS, and should be incorporated into shared-decision making discussions regarding the optimal management of DCIS for a given patient. Age-Stratified Incremental Harm-Benefit Ratios for BCS+RT versus BCS AloneAge GroupRAS Deaths: Breast Cancer Deaths Averted* Posterior Median Ratio (95% CrI)Overall1:12 (1:7 to 1:19)35 to 541:10 (1:6 to 1:14 )55 to 741:11 (1:7 to 1:15)75+1:17 (1:9 to 1:24)*Probabilistic sensitivity analysis using 10,000 second-order parameter samples with a 20 year time horizon Citation Format: Marquita R Decker, Joseph F Levy, Lee G Wilke, David J Vanness, Heather B Neuman. Balancing the harms and benefits of radiation therapy for DCIS: A decision analysis examining the risk of radiation-associated sarcoma [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-09-06.