PHILADELPHIA — An intraoperative, multimodal optical cancer detection system detected cancer with high degrees of accuracy, sensitivity, and specificity in real time during surgery in a validation study among patients with brain tumors, according to results published in Cancer Research, a journal of the American Association for Cancer Research.
“With brain cancer, near-perfect detection is important so that we can remove as much cancer as possible, without removing healthy tissue,” said Kevin Petrecca, MD, PhD, a neurosurgeon at the Montreal Neurological Institute and Hospital, and associate professor of neurology and neurosurgery at McGill University in Montreal, Quebec. “Residual cancer post-surgery is associated with decreased time to recurrence and lower survival.”
“Raman spectroscopy (RS) alone can achieve 90 percent detection accuracy, but by combining RS with intrinsic fluorescence spectroscopy (IFS) and diffuse reflectance spectroscopy (DFS), we are excited that our results showed 97 percent accuracy, 100 percent sensitivity, and 93 percent specificity,” said Frédéric Leblond, PhD, director of the Laboratory for Radiological Optics, associate professor in the Department of Engineering Physics at Polytechnique Montreal and a researcher at the University of Montreal Hospital Research Centre, who developed the tool together with Petrecca.
Petrecca explained that the optical detection system consisted of a handheld probe coupled with an analytics platform on a mobile cart. He noted that the probe contains miniaturized RS, IFS, and DFS technology, and has a tip the same size surgeons use to remove brain cancers. “The probe can scan a 500 micron-diameter area of tissue, and since a cell is roughly 10 microns across, the tool is able to detect down to small numbers of cancer cells,” said Petrecca. Leblond developed the platform’s machine-learning algorithm using archived tissue samples of brain cancer and normal cells from brain surgeries classified by standard pathology.
In the validation study, Petrecca, Leblond, and colleagues investigated the use of the optical system among 15 patients with grade II-IV gliomas and metastatic brain cancers (from primary lung or colon cancer or melanoma), who were undergoing open cranium surgery at the Montreal Neurological Institute and Hospital. They interrogated 10 to15 sites in each patient, 161 sites in total, taking optical readings of normal and tumor tissue regions at each site for blinded post-analysis to determine if cancer cells were present, and compared the data to corresponding spectral data.
Image analysis demonstrated improvements, measured as area under the curve, by as much as five percent from 95 percent for RS alone compared to up to 100 percent for RS combined with IFS and DFS, approaching the threshold for complete resection. Further, the optical system detected cancer with virtually the same degree of accuracy, sensitivity, and specificity among all cancer types investigated.
“Our findings are novel since optical techniques are not standard in any surgeries at present,” said Petrecca. “The results also indicate a strong potential for this technology to be adapted to a wide range of surgical and detection applications, including laparoscopic and robotic surgeries, and colonoscopy.”
The study was supported by funds from the New Researchers program from the Fonds de recherche du Québec – Nature et technologies (FRQNT), the Discovery Grant program from Natural Sciences and Engineering Research Council of Canada (NSERC), and the Collaborative Health Research Program by the Canadian Institutes of Health Research and NSERC. Petrecca serves as chief medical officer and Leblond serves as chief technical officer at ODS Medical Inc., a diagnostic medical device company they founded in 2015.
American Association for Cancer Research