Our next speaker is Dr. Steve Cho, and we're going to keep on with pituitaries, and he'll discuss comparison of intraoperative imaging, a receptor-specific versus passive delivery of near-infrared dyes in human pituitary surgery. Good afternoon, everyone. My name is Steve Cho. I'm a third-year medical student at the Paramount School of Medicine at UPenn, and today I'll be talking to you about recent progress in intraoperative near-infrared fluorescence guided neurosurgery, specifically in pituitary adenomas. So for neurosurgeons that deal with tumor, the resect tumor, the number one goal in tumor resection is to maximize the resection while minimizing itrogenic neurologic damage to the patients, and the best way to accomplish that is to distinguish tumor from non-tumor. The most common tool, obviously, is a surgeon's eyes, most often aided by loops or microscopes, and the surgeon in the operating room relies on their visual and tactile sensations to make this distinction between tumors and non-tumor. And even though this has been used for centuries now with pretty good results, the disadvantage is that in many cases it can be difficult to make an accurate distinction between tumor and non-tumor. So more recent technology is neuronavigation, as most of you are familiar, using preoperative MRI to correlate the tumor location with what the surgeon sees in the operating room. The main disadvantage, however, is that because the images are acquired preoperatively, it can't account for the brain shifts and changes in anatomic landmarks that occur during the surgery, and can cause discrepancy between what you see in the neuronavigation and what the surgeon actually sees in the brain. So another technology that's not as widely used yet that tries to get around that problem is intraoperative MRI, where the surgeon gets a real-time update on the MRI imaging to try to get around the problem of brain shifts. However, the main reason it's not commonly used yet is it's very expensive, time-consuming, and multiple studies have shown high false positive rates with intraoperative MRI, causing unnecessary resection of non-brain parenchyma. So today, I want to just present to you an alternative, which is intraoperative fluorescence-guided surgery, which has many of the advantages of the other options, but very few of the disadvantages. So many of you are probably familiar with endocyanin green, or ICG, which has been used for decades for vasculature endography by injecting low doses of ICG and then imaging about 10 to 15 seconds later to delineate the vasculature. However, at UPenn, we developed a novel technology that we have termed second-window ICG, or SWIG for short, and it's basically injecting patients with a much higher dose of ICG, over 100 times higher than the normal dose, 24 hours prior to surgery, rather than 10 to 15 seconds. And this allows us to image the tumor that has retained ICG in the surrounding tissue by something called enhanced permeability and retention effect, or the EPR effect. And what that says, in short, is that normal tissues, normal brain, will have intact endothelium, so when we inject intravenously ICG into the patients, the ICG is quickly washed away from the tumor and there's no retention at 24 hours after the injection. But in tumors with permeable endothelium, the ICG leaks into the surrounding tissue and is retained by the tumor and the surrounding tissue for 24 hours or greater, such that when we image in the OR, we can see the tumor that has retained ICG contrasted against the brain parenchyma that has no ICG retained in it. So to show an example, here's one of our patients. She has breast cancer metastases to her brain. As you can see, even with high-definition microscopy, it's very hard to tell where the tumor is with a cortex view. But with near-infrared imaging, with SWIG, we can see exactly where the tumor is. It's clearly delineated with high signal and high contrast to the normal brain. This is the same patient a step earlier, prior to dural opening, which shows, again, that the tumor is impossible to locate where it is. But with near-infrared imaging, we can see exactly where the tumor is, even through the dura, and this can help guide the surgeons in dural resection and exactly where to open the dura. So to summarize some of our previous results in the past year and a half, we have published multiple papers in glioblastomas, meningiomas, and metastases. And just to summarize the results, SWIG has reproducibly shown higher sensitivity compared to microscopy when it comes to detecting tumor, but a much lower specificity, as you can see in these numbers. And that's to be expected because, as I explained, SWIG is hypothesized to work via the EPR effect, and enhanced vascular permeability is not specific to tumors, as you know. Inflammation, necrosis, many other things can cause leaky vessels. So it's expected the specificity is low. So to try to get around this problem of low specificity, we decided to look at pituitary adenomas. The main reason being, as we have heard in previous presentations, gross total resection rates hover around 60% to 80% in literature. 15% to 20% or greater of pituitary adenomas eventually recur, with non-functioning adenomas having a higher recurrence rate than functioning adenomas. So this tells us that there's a need to improve resection in pituitary adenomas, and the main reason we chose pituitary adenomas for our project was because in the early 2000s, Dr. Oya Sifu and his group showed that the majority of non-functioning pituitary adenomas overexpress something called folate receptor alpha on the cell membrane. And the reason that's important is as follows. One of our collaborators at Purdue University developed a diet called OTL38, which basically simply puts ICG linked to a folate analog. And what that allows us to do is in normal cells, the normal brain and the rest of our body that have very minimal folate receptor expression, OTL38, even if present in the blood vessel, will not bind the cells, so there should be no uptake. Whereas non-functioning pituitary adenomas that significantly overexpress folate receptors, OTL38 can bind the cells and undergo resultant-mediated endocytosis, such that after injection, we can easily visualize the tumor that has uptaken OTL38. So here's just another demonstration of endoscopy this time, a non-functioning pituitary adenoma with a visible light only and the same view with near-infrared, showing strong signal in the tumor. And after resection in the same patient, we see a very strong area of fluorescence here, which was not initially resected because the surgeon was not sure if that was tumor or not. We biopsied the area, and histology confirmed that this was indeed tumor. So to summarize some of our most important results, across the top we have visible light sensitivity, specificity, positive predictive value, and negative predictive value. These are all based on biopsies and pathology. The histology results are gold standard. As you can see, SWIG, similar to our previous results, showed a much higher sensitivity, but again, a much lower specificity. But the targeted diet, OTL38, in non-functional adenoma showed a slightly reduced sensitivity, but a significantly increased specificity. And if we limited our results to just those cells that post-operatively were confirmed to actually overexpress folate receptors, because not all non-functional adenomas do, the sensitivity and specificity were 100%, as well as a positive predictive value and negative predictive value. So in conclusion, intraoperative near-infrared imaging can easily visualize tumor in otherwise impossible situations, such as through the dura, and is highly sensitive for residual tumors with nearly 100% negative predictive value. And targeted near-infrared diets, such as OTL38, furthermore, improve upon the low specificity of non-targeted diets, such as ICG, while preserving this high sensitivity. And the positive predictive value and the negative predictive value of near-infrared diets both approach 100% for margin specimens, but in appropriately selected tumors, such as non-functional adenomas. Because for OTL38, it wouldn't work in functional adenomas that do not express folate receptors. So some of our further work that we're planning right now, we currently are developing a mice intracranial tumor models, and are planning on comparing ICG and targeted near-infrared diets against 5-ALA, which is a visible light fluorophore that is currently FDA-approved for gliomas. We're also currently designing a prospective outcome study in patients using near-infrared diets, but so far our studies have not changed the scope of surgery. And finally, we're also developing a novel intraoperative imaging system, as well as new targets for novel targeted diets. So thank you to our lab, and thank you to Dr. Sean Grady, who's our chairman, and the rest of our neurosurgery department, and of course to AANS for letting me present today. Thank you.

intraoperative imaging

pituitary surgery

tumor resection

neurological damage

intraoperative fluorescence-guided surgery