Comparing an operation to monitoring, with or without endocrine therapy (COMET), for low-risk ductal carcinoma in situ (DCIS).

TPS616 Background: Approximately 50,000 women in the U.S. are diagnosed with ductal carcinoma in situ (DCIS) each year. Without treatment, it is estimated that only 20-30% of DCIS will lead to invasive breast cancer (IBC). However, over 97% of women are currently treated with surgery +/- radiation. An alternative to surgery is active monitoring (AM), a management approach in which mammograms/physical exams are used to monitor breast changes and determine when, or if, surgery is needed. The COMET Study will compare risks and benefits of AM versus surgery for low-risk DCIS in the setting of a Phase III multicenter prospective randomized trial. The study is funded by the Patient-Centered Outcomes Research Institute. The COMET trial opened in the U.S. in June 2017 (Clinicaltrials.gov reference: NCT02926911). In November 2021, the Data Safety Monitoring Board reviewed the trial and suggested that it continue as planned. Patient accrual will continue until 12/31/2022. Methods: The primary objective is to assess whether the 2-year ipsilateral IBC rate for AM is non-inferior to that for surgery. Secondary objectives include determining whether AM is non-inferior to surgery for 2-year mastectomy rate; breast conservation rate; contralateral breast cancer rate; overall and breast cancer-specific survival. Patient reported outcomes will enable comparison of health-related quality of life and psychosocial outcomes between surgery and AM groups at baseline, 6-months, and years 1-5. Eligibility criteria include: age > 40 at diagnosis; pathologic confirmation of grade I/II DCIS or atypia verging on DCIS without invasion by two pathologists; ER and/or PR ≥ 10%; no mass on physical exam or imaging. The accrual goal is 1200 randomized patients across 100 Alliance for Clinical Trials in Oncology sites. Sample size is estimated using a 2-group test of non-inferiority of proportions, with the 2-year IBC rate in the surgery group assumed to be 0.10 based on published studies and non-inferiority margin of 0.05. Based on a 1-sided un-pooled z-test, with alpha = 0.05, a sample size of n = 446 per group will have 80% power to detect the specified non-inferiority margin. Final analysis plan will include a per protocol component as well as a pragmatic component for patients who are randomized and decline participation in their assigned arm. Primary analyses will adjust for dropout, non-compliance and contamination by utilizing instrumental variable methods. Clinical trial information: NCT02926911.