The review, authored by Aaditya Sinha and Arnie Purushotham of King’s College London, synthesizes evidence from five key clinical studies of intraoperative specimen imaging technologies. The unifying principle is elegant: patients receive an intravenous injection of 18F-fluorodeoxyglucose, or 18F-FDG, a radiolabelled glucose analogue that accumulates preferentially in malignant tissue. Because breast cancer cells, like many tumors, exhibit the Warburg effect and rely heavily on aerobic glycolysis, they import glucose at rates far exceeding those of healthy tissue. The tracer enters cells through glucose transporters, chiefly GLUT1, and is phosphorylated by hexokinase into 18F-FDG-6-phosphate, which cannot proceed through glycolysis and becomes metabolically trapped inside the cell. The result is a molecular beacon that marks tumor tissue with a brightness proportional to its metabolic activity.

Three distinct imaging modalities have emerged to read that signal from excised lumpectomy specimens. The first, Cerenkov luminescence imaging, or CLI, capitalizes on a phenomenon first documented by the Russian scientist Pavel Cerenkov in 1934. When positrons emitted by decaying 18F travel through tissue faster than light travels in that medium, they polarize surrounding molecules, which then relax and emit faint blue optical photons. Sensitive cameras can capture this intrinsic glow without any external light source, revealing where FDG-avid tumor cells sit relative to the cut edge of the specimen. In a first-in-human feasibility study, Grootendorst and colleagues injected patients with 150 to 230 MBq of 18F-FDG roughly 45 to 60 minutes before surgery and imaged 15 margins from 12 patients. CLI achieved 89 percent sensitivity and 95 percent specificity, correctly identified elevated tumor radiance in 10 of 12 patients, and showed strong agreement with histopathology, with a kappa statistic of 0.81 and an interrater reliability among surgeons of 0.912.

The second technique, flexible autoradiography, or FAR, takes a different physical approach. Instead of detecting Cerenkov photons, it places a thin scintillating film in direct contact with the excised tissue, capturing beta particles emitted by FDG-labelled cancer cells and converting them into scintillations that yield high-resolution surface maps of tracer activity. Because the detector sits millimeters from the tissue, FAR achieves millimetre to sub-millimetre spatial resolution, far exceeding standard gamma-based approaches. A multicentre first-in-human study by Jurrius and colleagues enrolled 66 patients who received about 3 MBq/kg of 18F-FDG preoperatively, with imaging completed within 10 minutes of excision. The technique delivered an overall diagnostic accuracy of 80.5 percent, with 46.2 percent sensitivity and 81.7 percent specificity. Notably, in a high-activity subgroup, sensitivity rose to 71.4 percent and the negative predictive value reached 98.4 percent, suggesting the method is most reliable when tracer uptake is strong.

The third and most technically ambitious approach combines both optical modalities. In the CLI-FAR technique, evaluated prospectively at Guy’s Hospital in the United Kingdom by Sinha and colleagues, patients received a standardized 250 MBq dose of 18F-FDG approximately 145 minutes before intraoperative imaging. Excised specimens were first imaged with three-dimensional intraoperative X-ray, then CLI-FAR was applied to margins suspected of involvement, or to the two closest margins when no suspicion existed. Seven trained consultant surgeons reviewed every intraoperative finding and acted on them immediately, excising additional tissue when imaging suggested residual disease. Histopathologists, blinded to the imaging results, then independently assessed margin status, tumor type and grade, receptor status, and lymphovascular invasion. The dual-modality approach achieved 76.9 percent sensitivity and a striking 97.8 percent specificity, and reduced re-excision rates by 69 percent, with an average image interpretation time of just six minutes.

All three optical techniques rely on the LightPath imaging system, an in vitro diagnostic device developed by Lightpoint Medical in the United Kingdom. The system centers on an ultrasensitive electron-multiplying charge-coupled device camera, paired with a white-light reference camera, housed in a fully light-tight specimen chamber. For CLI, the specimen is imaged uncovered, with the camera detecting Cerenkov photons in the 550 to 850 nanometre wavelength range. For FAR, a 12-micrometre scintillator is affixed to the specimen, separated by a 3-micrometre Mylar sheet to prevent contamination, and a band-pass filter isolates the scintillation signal. Each acquisition takes 300 seconds with 8-by-8 pixel binning, and each margin must be imaged individually because the camera captures only a two-dimensional field at any moment.

The fourth technology abandons optical detection altogether in favor of true three-dimensional tomography. The XEOS AURA 10 is the first and currently the only mobile PET/CT specimen imager designed for the operating room, combining a high-sensitivity PET detector with a built-in CT scanner. Its small field of view and dedicated design yield roughly five-fold higher spatial resolution than conventional whole-body PET, allowing sub-centimetre foci of residual FDG-avid tumor to be localized in three dimensions within minutes of excision. A proof-of-concept study by Göker and colleagues imaged 20 patients using the MOLECUBES micro-PET and micro-CT scanners, with a supervised automated algorithm segmenting regions of elevated uptake. Automated image analysis achieved 90 percent sensitivity but only 60 percent specificity, with an area under the curve of 0.86, while blinded surgeon interpretation reached 79 percent sensitivity and 72 percent specificity. The subsequent BIMAP study, conducted in Belgium by De Crem and colleagues with 41 patients, using the integrated AURA 10 device, achieved 91 percent sensitivity and 94 percent specificity, particularly in invasive ductal carcinoma, and enabled four patients to avoid additional surgery. The estimated number needed to treat to prevent one reoperation was just four.

Beyond diagnostic accuracy, the practical advantages of these techniques are considerable. Radiation exposure to surgical staff across all studies remained low, ranging from approximately 15 to 38 microsieverts per procedure, well within safe occupational limits, and imaging added minimal time to operative workflows. Unlike frozen section analysis or specimen radiography, which rely on anatomical landmarks or slow laboratory processing, radionuclide imaging detects metabolic activity directly, offering a molecular perspective that can reveal microscopic residual disease invisible to palpation or conventional imaging. The technologies are also portable: they have been deployed in general hospital settings without on-site PET/CT infrastructure, as Lambert and colleagues demonstrated in a feasibility study covering 32 surgical procedures across multiple malignancies, including breast, thyroid, renal, prostate, and skin cancers, all completed intraoperatively with a median staff dose of about 15 microsieverts.

Significant challenges remain, however. CLI is vulnerable to chemiluminescence artifacts that require substantial training to distinguish from true radioactivity, and the standardized scintillator used for FAR can only accommodate specimens up to 4 centimetres in diameter, a limitation that becomes acute as oncoplastic procedures increasingly produce larger specimens, though larger scintillators are becoming available. FDG-PET also performs poorly in tumors with low intrinsic glucose metabolism, such as invasive lobular carcinoma and in situ lesions, which can produce false negatives. Timing is another constraint: 18F-FDG has a physical half-life of about 109.7 minutes, and while acceptable imaging results can be achieved up to roughly 180 minutes post-injection, a target interval of 60 to 120 minutes is generally recommended, demanding careful coordination between radiopharmacy, surgery, and imaging teams. PET interpretation also carries a learning curve, and the review’s authors suggest that involving radiologists or nuclear medicine physicians in real-time image reading could further improve accuracy, at the cost of additional coordination.

Despite these hurdles, the accumulated evidence paints a compelling picture. Across CLI, FAR, CLI-FAR, and intraoperative PET/CT, imaging findings correlate robustly with histopathological gold standards, re-excision rates fall substantially when results are acted upon immediately, and the techniques integrate into existing surgical workflows without requiring operating theatres to be rebuilt around nuclear medicine infrastructure. The review’s authors conclude that radionuclide-based intraoperative specimen imaging represents a transformative opportunity for precision-guided oncologic surgery, one that could spare thousands of breast cancer patients the physical and psychological burden of a second operation. As larger scintillators, refined interpretive algorithms, and tailored imaging protocols mature, the prospect of a single, definitive cancer surgery, guided by the faint blue glow of a decaying isotope, moves steadily closer to routine clinical practice.

Subject of Research: Radionuclide imaging techniques for intraoperative margin assessment in breast-conserving surgery

Article Title: A review on the use of radionuclide imaging techniques to detect margin positivity in intraoperative specimens during breast-conserving surgery

Article References: Sinha, A., & Purushotham, A. (2025). A review on the use of radionuclide imaging techniques to detect margin positivity in intraoperative specimens during breast-conserving surgery. Frontiers in Medical Engineering, 3, Article 1628589. https://doi.org/10.3389/fmede.2025.1628589

Image Credits: AI Generated

DOI: 10.3389/fmede.2025.1628589

Keywords: breast cancer, breast-conserving surgery, surgical margins, Cerenkov luminescence imaging, flexible autoradiography, PET/CT, 18F-FDG, re-excision, intraoperative imaging, radiotracers, molecular imaging, oncological surgery