Background: Identification of women with DCIS who have a very low risk of local recurrence risk (LRR) after breast-conserving surgery (BCS) is needed to de-escalate therapy. We evaluated the impact of 10-year LRR estimates after BCS, calculated by the integration of a 12-gene molecular expression assay (Oncotype Breast DCIS Score®) and clinicopathological features (CPFs), on its ability to change radiation oncologists' recommendations for RT after BCS for DCIS.
Methods: Prospective cohort study of women with DCIS treated with BCS. Eligibility criteria were as follows: age > 45 years, tumor ≤ 2.5 cm, and margins ≥ 1 mm. Radiation oncologists provided 10-year LRR estimates without RT and recommendation for RT pre- and post-assay. Primary outcome was change in RT recommendation.
Results: 217 patients were evaluable, with mean age = 63 years, mean tumor size = 1.1 cm, and mean DCIS Score = 32; 140 (64%) were in the low-risk (<39), 32 (15%) were in the intermediate-risk (39-54), and 45 (21%) were in the high-risk groups (≥55). The assay led to a change in treatment recommendation in 76 (35.2%) (95%CI 29.1-41.8%) patients. RT recommendations decreased from 79% pre-assay to 50% post-assay (difference = 29%; 95%CI 22-35%) due to a significant increase in the proportion of patients with a predicted low LRR (< 10%) post-assay and recommendations to omit RT for those with a low predicted risk. The assay was associated with improved patient satisfaction and reduced decisional conflict.
Conclusion: The DCIS Score assay combined with CPFs identified more women with an estimated low (<10%) 10-yr LR risk after BCS, leading to a significant decrease in recommendations for RT compared to estimates based on CPFs alone.
Keywords: Assay; DCIS; De-escalation; Ductal carcinoma in situ; Genomic; Radiotherapy.