Brain cancer is one of the most fatal forms of cancer, with a five-year survival rate of just 36%, according to the National Cancer Institute. Accurate diagnoses could improve outcomes, but current diagnostic methods, such as tissue biopsy and spinal tap, remain highly invasive and/or low resolution, making the detection and location of such tumors difficult and painful. Now, researchers from Toronto Metropolitan University, St. Michael’s Hospital and the University of Toronto have developed a ultrasensitive biosensor that, combined with surface-enhanced Raman spectroscopy (SERS) and artificial intelligence, can detect and determine the location of primary and secondary brain tumors using just 5 microliters of blood.
The biosensor consists of 3D nickel-nickel oxide nanolayers on a nickel chip and is fabricated using high-intensity lasers. The sensor allows for the detection of minute amounts of tumor-derived materials, such as nucleic acids, proteins and lipids, that have passed through the blood-brain barrier into circulation. SERS is used to detect these components and generate a molecular “fingerprint” of each sample, which is then analyzed using a DEEP neural network to find evidence of a brain tumor, define its type and predict its location within the brain.
With just a 5 μL sample of blood serum, the sensor system was able to detect brain cancer and distinguish it from breast, lung and colorectal cancer with 100% specificity and sensitivity, the researchers wrote. Additionally, the team had similar success distinguishing primary brain tumors from secondary tumors that had metastasized to the brain from the lung or breast. The system enabled the researchers to determine which of the nine brain compartments the tumor was located in with 96% accuracy. Due to its noninvasive nature, this method could allow brain tumor development to be more effectively monitored over time and help drive better treatment decisions, according to the researchers. This study was published in ACS Nano.
“This modality of using patient sera for deep surveillance is a promising noninvasive liquid biopsy tool with the potential to complement current brain cancer diagnostic methodologies,” the authors wrote.
This research was supported by the Toronto Metropolitan University Faculty of Engineering and Architectural Science Dean’s Research Fund, the Natural Science and Engineering Research Council of Canada (NSERC) Discovery Grants and NSERC Discovery Grant, Idea to Innovation and Alliance Grants.