A Prospective, Multicenter, Randomized, Double-Arm Trial to Determine the Impact of the Perimeter B-Series Optical Coherence Tomography and Artificial Intelligence System on Positive Margin Rates in Breast Conservation Surgery

Abstract

Introduction: Optical coherence tomography (OCT) enables high-resolution, real-time tissue-imaging up to 2mm deep. The Perimeter B-Series OCT System combines wide field-OCT (WF-OCT) with artificial intelligence (AI) to help identify areas suspicious for cancer. The objective of this trial is to assess adjunctive use of WF-OCT and standard care (SOC) in identifying unaddressed positive margins (UPMs). Methods: Prospective, multicenter, randomized, double-arm trial in up to 309 adult women undergoing breast conservation surgery (BCS) for biopsy-confirmed Stage 0-III invasive ductal cancer and/or DCIS, including post-neoadjuvant therapy (NCT05113927). After BCS and SOC intraoperative margin assessment participants will be randomized (2:1) to device or SOC arm. Controls may undergo intraoperative pathology or frozen section analysis, per routine. WF-OCT will be done on all device arm specimens, with an opportunity to excise tissue from the cavity post-analysis, up to 6 total shaves and 2 in each orientation. Results: The study hypothesis is that WF-OCT & AI use will reduce the proportion of subjects with UPM. The primary endpoint is occurrence of ≥1 UPM for a subject; the secondary endpoint is the number of UPMs per subject. Safety includes Adverse Events; false-positive shaves/subject; and BREAST-Q Satisfaction with Breasts. Other outcomes are total excised tissue volume, initial and all procedures; and margin-level effectiveness (sensitivity, specificity, NPV and PPV). Conclusion: Primary endpoint was powered assuming a 15% UPM rate, 90% power, 2:1 allocation, and 10% loss of subjects. Both primary and secondary effectiveness endpoints are based on within-subject comparison of the 2 treatments in the device arm (SOC+WF-OCT).