Comparing an operation tomonitoring, with or without endocrine therapy (COMET): Aprospective randomized trial for low-risk DCIS (AFT-25)

ABSTRACT

Background: Approximately 50,000 women in the U.S. are diagnosed with ductal carcinoma in situ (DCIS)annually. Without treatment, it is estimated that 20-30% of DCIS will lead to invasive breast cancer. Currently, morethan 97% of women undergo surgery, with many also undergoing radiation. An alternative to surgery for low-riskDCIS is active monitoring (AM), an approach in which regularly scheduled mammography and physical exams areused to monitor breast changes and determine if, or when, surgery is needed. Trial design: COMET, a multicenterphase III prospective randomized trial, opened in the U.S. in June 2017 (clinicaltrials.gov reference NCT02926911).The hypothesis is that management of low-risk DCIS using an AM approach does not yield inferior invasive breastcancer and/or quality of life outcomes compared to surgery. Eligibility criteria Patients with a new diagnosis of unilateral, bilateral, unifocal, multifocal, or multicentric DCIS, or atypia verging on DCIS are eligible. Patients mustbe ≥40 years of age, have no contraindication for surgery, and pathologic confirmation of grade I/II DCIS. DCIS mustbe ER and/or PR≥ 10% and HER2-negative without invasion, diagnosed within 120 days of registration. Breasttissue, blood and imaging are collected at trial entry and if invasive cancer subsequently occurs, and are stored incentral repositories. Specific aims: The primary aim is to assess whether the 2-yr ipsilateral invasive breast cancerrate for AM is non-inferior to surgery. Secondary aims include comparison of 2-, 5-, and 10-yr mastectomy rate, contralateral invasive breast cancer rate, overall survival and invasive breast cancer-specific survival, as well as 5-and 10-yr ipsilateral invasive breast cancer rate between groups. Patient reported outcomes (PRO) using validatedtools are critical secondary endpoints, and will enable comparison of health-related quality of life and psychosocialoutcomes between surgery and AM groups at prespecified time points over a period of 5 years. Statisticalmethods: An accrual goal of 1200 was estimated using a 2-group test of noninferiority of proportions, with the 2-yrinvasive breast cancer rate in the surgery group assumed to be 0.10, including accounting for upstaging. Theprojected drop-out rate is 25%, for a total of 900 patients treated per allocation arm. The non-inferiority boundarywas set at 0.05. Based on a 1-sided un-pooled z-test, with alpha=0.05, a sample size of n=446 per group will have80% power to detect the specified noninferiority margin. Intention-to-treat analysis of the 2-yr invasive breast cancerrate will be conducted using all patients as randomized, and will be completed using Kaplan-Meier estimates,stratified by group, combined with Greenwood's confidence interval. Several sensitivity analyses (per protocol, as-treated, and instrumental variable) are also planned to account for loss of follow-up, rejection of randomizationallocation and withdrawals. Present and target accrual Trial accrual as of 7/1/20 is 540 randomized patients from84 activated Alliance for Clinical Trials in Oncology sites. Despite logistical challenges posed by the COVID-19crisis, patients continue to be recruited to the COMET trial. Over 80% of patients have sample sets/images stored inthe tissue and image repositories. This trial will provide definitive clinical, quality of life and biomarker evidenceregarding the trade-offs of surgery vs AM in patients with low-risk DCIS.