LAS VEGAS—Achieving negative margins in breast-conserving surgery remains a challenge to date, with 15% to 20% of patients needing reoperations to remove more tissue. But the use of intraoperative optical coherence tomography with artificial intelligence may significantly increase the likelihood that patients will not need to return to the OR.
“I have the privilege today to unveil to you a technology that I believe makes a huge difference in this arena,” said Alastair Thompson, MD, a professor of surgery at Baylor College of Medicine, in Houston, at the 2025 annual meeting of the American Society of Breast Surgeons.
“At the time of lumpectomy, one can place the specimen on this device in the OR. The entire surface of the specimen is then scanned using optical coherence tomography. OCT measures the light scatter. It penetrates to a depth of 2 mm, and will scan the entire specimen, unlike pathology, which will only look at about 2% of the entire margin,” Dr. Thompson said.
The technology uses AI to highlight what the surgeons need to examine and interpret, as well as what they will need to do to remove additional tissue. “The OCT will pick up invasive disease, and it will pick up ductal carcinoma in situ [DCIS]—it doesn’t need microcalcifications in the DCIS,” he said.
The whole process takes 10 to 15 minutes and can be performed by trained personnel before the results are interpreted by the surgeon. There are two CPT codes: one for scanning and one for interpretation.
Dr. Thompson and his colleagues worked with the FDA to design a trial to assess the safety and effectiveness of OCT/AI, defining effectiveness as the detection, interpretation, and subsequent actions, with end points of identifying and removing positive margins.
Patients had stage 0 to III invasive ductal carcinoma and/or DCIS, and positive margins were defined as per standard in the United States. “Standard-of-care margin assessment could be performed per the surgeons’ preference. Once that standard of care had been performed to the satisfaction of the surgeon, the patient was randomized to either no further action/control arm or use of the OCT/AI device,” Dr. Thompson said.
The trial included 315 patients: 109 in the control arm and 206 who had undergone lumpectomy and OCT/AI. “Two other things to point out: We were not doing further analysis on the control arm, and the pathologists were unaware which arm the patients were in,” Dr. Thompson said.
Use of OCT/AI reduced the positive margin rate from 17% to 13.6%. In the 35 patients who had residual disease after standard of care, seven (20%) were cleared of all residual disease (P=0.005). An ad hoc analysis identified further clinical benefit: In 19 patients, surgeons found additional disease that would have been missed by standard of care.
“Looking at this in a slightly different way, of the 35 patients with 56 diseased margins, which standard of care had failed to identify, we removed additional diseased margins in 14, and in seven of those patients, we had complete clearance of all residual disease,” Dr. Thompson said.
Margin level accuracy was 88%, reflecting a sensitivity of 92% and specificity of 98%. “And in contrast to technologies that use radiation, inject drugs with the consequence of allergic reactions, or have patient contact—which this device doesn’t—it was perfectly safe.”
Dr. Thompson noted that when more tissue needed to be removed, it wasn’t much, “about the same as the tip of your finger—much less than other devices have had to remove.”
The trial significantly demonstrated the removal of margins in patients who would otherwise have had to return to surgery, he said. “You could argue that with seven patients completely cleared of disease, plus the 19 additional patients, three quarters of the patients benefited from this technology, perhaps reflective of what will happen in the future.”
Mediget Teshome, MD, MPH, a breast surgical oncologist and an associate professor at UCLA Health, in Los Angeles, said any technology or approach that helps surgeons increase the rate of negative margins is highly desirable.
“Consistently achieving negative margins is somewhat of a holy grail. The current standard of care in most institutions is to perform shave cavity margins, which has been shown to reduce the rate of positive margins. But this technology—using a device in the OR to help determine if a margin is positive and which margin is positive—is very promising,” Dr. Teshome said.
However, she would like to see more data before drawing any conclusions about the device. “It does appear to have been a multicenter trial, used in a variety of settings. And Dr. Thompson stated there were no false positives, which is reassuring. Hopefully we’re moving toward a future with fewer positive margins and fewer reoperations for breast cancer. I welcome that day for sure.”
This article is from the December 2025 print issue.
Please log in to post a comment