Mass General Brigham investigators have shown that they can accurately distinguish between normal tissues and tumor cells using a novel imaging technique: high-speed camera that measures the fluorescence lifetime (FLT) of a dye injected into the bloodstream.
“Our lab has been studying fluorescence lifetime imaging since 2002, but this is the first time that anyone has combined it with tumor imaging and injectable dyes in humans,” says corresponding author Anand Kumar, of the Athinoula A. Martinos Center for Biomedical Imaging at Massachusetts General Hospital. “By doing so, we’ve developed a technique for accurately distinguishing tumor tissue from healthy tissue across cancer types.”
Kumar and co-authors explain in Nature Biomedical Engineering that, in the past two decades, there has been dramatic growth in the development of fluorescent probes and imaging systems that target tumor-specific biomarkers, but to date, tumor uptake of these probes has been variable and resulted in poor sensitivity and specificity.
Furthermore, the probes are used in conjunction with fluorescence intensity-based imaging, which cannot readily distinguish tumor-specific and non-specific fluorescence, is measured in arbitrary units, and depends on parameters such as the size and depth of the tumor, dye uptake, detector efficiency, illumination power, and distance from the target.
By contrast, FLT is a photophysical property that is measured in standardized units (nanoseconds), and is largely independent of system parameters, tumor probe uptake, and depth in tissue.
In 2017, Kumar and team reported that when they combined the FDA-approved near-infrared dye indocyanine green (ICG) with FLT detection they could accurately identify tumor tissue in mouse models of subcutaneous human breast and brain tumors. Specifically, the tumors in mice injected with ICG had a longer FLT than normal tissue.
They therefore approached human studies with “cautious optimism” because “although the early preclinical studies suggested a high lifetime contrast between tumor and normal tissue, many things change when you take a drug from animals to humans, including the way the dye distributes in the body and the way it is taken up in tumors,” Kumar tells Inside Precision Medicine.
The researchers initially tested their technique among 22 patients undergoing surgery for primary and metastatic liver tumors. The patients were injected with ICG between 24 and 72 hours prior to surgery and wide-field imaging was performed on the resected specimens within 30 minutes of removal using a custom-built time-domain (TD) fluorescence imaging system.
In patients with moderately differentiated hepatocellular carcinoma (HCC), the FLTs within the tumor were significantly longer than the FLTs in surrounding normal liver parenchyma, at 0.65 versus 0.52 nanoseconds. The difference was verified at the cellular level using confocal fluorescence lifetime imaging microscopy.
A receiver operating characteristic analysis showed that the average accuracy for classification between tumor and normal tissue was 98.4%with wide-field TD imaging in the HCCs samples. By comparison, the accuracy was just 51.5% when the researchers only measured fluorescence intensity.
Kumar and colleagues observed similar results when they tested samples from patients with head and neck cancers (n=18), brain tumors (n=9), and bone and soft tissue cancers (n=22), with the mean accuracy of FLT above 97% for all tumor types. The technique was also able to distinguish benign from metastatic lymph nodes.
“So far it looks like this method works for all the solid tumors we studied, so we do not yet know where it may not work,” says Kumar. He noted however that ICG uptake was limited in atypical head and neck cancers. “Also, for patients undergoing radiation or chemotherapy, there is a lot of dead tumor tissue, and the ICG uptake is poor in these areas, so the method will not work very well. We are still investigating this question using larger studies,” Kumar remarked.
Nonetheless, he said that the take home message from the study is that “fluorescence lifetime can improve tumor visualization in patients injected with the FDA approved dye.”
However, ICG it is not yet approved for clinical use as a tumor marking agent, despite its approval for other indication and further work is needed to prove its safety and efficacy in this setting.
Kumar says his team is now developing intraoperative imaging systems that soon will allow them to test the technology directly in patients during various cancer surgeries. “This will lead to larger scale clinical trials to test the efficacy of this approach,” he notes. “We are also further investigating the mechanisms for why ICG FLT is enhanced in tumors, [which] should motivate further research to develop compounds that are even brighter than ICG and can be injected right before surgery.”
Until those compounds are developed, the team is working to further optimize the timing and dosage of ICG, which could improve the performance of FLT even further.
Kumar hopes the findings “will improve surgeons’ ability to identify tumors during surgery, allowing complete tumor removal, without damaging vital healthy tissues and function. For example, in tongue cancer it is very important to preserve as much of the tongue as possible to avoid speech impediments.”
He adds: “Ultimately, we hope this technology will reduce recurrence of cancer while also avoiding repeat surgeries and preserving healthy tissue and function, thereby impacting patient lives and quality of life.”
