Dr. Fine: Watching that many patients and, and helping that many patients go through the death and dying process has made me focus my science on something that has a chance of making a difference, as opposed to just being academically successful. And i- if we're doing science that ultimately doesn't really have anything to do with patients, then I'm not interested.
Erin Welsh: Dr. Howard Fine is a neuro-oncologist at NewYork-Presbyterian and director of the Brain Tumor Center at Weill Cornell Medicine. He has dedicated his decades-long career to studying one of the most lethal types of cancer, glioblastoma.
Dr. Fine: Glioblastoma, it's not a lump in your head. It's a tumor made up of billions of individual tumor cells, each with a mind of its own, that infiltrates diffusely into the brain.
Erin Welsh: Glioblastoma is the most common and aggressive primary brain tumor in adults. When Dr. Fine began working in this field a few decades ago, the median survival for glioblastoma patients was about 12 to 13 months.
Dr. Fine: Now, 40 years later, you know, 2 to 3,000 clinical trials later, several billions of dollars of research, and the median survival now is about 15 months.
Erin Welsh: Despite major advances in understanding the biology of glioblastoma, outcomes for patients haven't significantly improved.
Dr. Fine: Because how do you pick m- millions or billions of tumor cells out of a functioning brain without scrambling the brain? The field is still, for the most part, still trying to treat this like a, a normal cancer. So I think there needs to be at least some people in the field who are saying, "Eh, maybe we need to look at a different approach." And that's what we do.
Erin Welsh: Dr. Fine and his team are pioneering a groundbreaking research platform called GLICO. It harnesses the unique biology of the disease and utilizes cutting-edge research methods to create cerebral organoids, miniature replicas of the brain that are specific to each patient's tumor, to find individualized therapies for glioblastoma patients. Their approach could meaningfully improve outcomes for this disease for the first time.
I'm Erin Welsh, and this is Advances in Care.
Today, I'll speak with Dr. Howard Fine about how he and his team are using advanced bioengineering techniques and machine learning to revolutionize treatments for glioblastoma.
Erin Welsh: Dr. Fine, thank you so much for joining us today. It's great to have you here.
Dr. Fine: It's a pleasure to be here.
Erin Welsh: You have spent nearly four decades working in glioblastoma, and I understand that you spent many years at the NIH, where you led research programs into studying this disease. How did that research background influence the way that you approach both science and clinical care at NewYork-Presbyterian and Weill Cornell Medicine?
Dr. Fine: When I was recruited to build a trans-institute program at the National Institutes of Health, the head of, um, the NIH at that time, Harold Varmus, who's here at Cornell now, had the forethought to say, you know, maybe one of the problems with studying glioblastoma is that everyone's in their own silos, and that this is a disease that crosses many disease types and many different physiologies from neuroscience to cancer. And so they asked me to build a- the first program in the intramural NIH campus. And instead of placing me in research buildings where there were all the cancer researchers, they actually placed me and my laboratory in a neuroscience building, and that really changed my career.
Erin Welsh: That's so interesting. Tell me more about how that exposure to neuroscience set you on this different path.
Dr. Fine: When I was studying research in glioblastoma, I was approaching it like my colleagues in, you know, lung cancer and GI cancer and so forth, looking at oncogenes and signaling pathways. When I got to the NIH and I got surrounded by all these neuroscience people, there was this laboratory of this world-renowned guy who was studying these things called neuro stem cells. And one day, one of my senior postdoctorate scientists came to me and he said, "You know, in that laboratory, they're studying these neuro stem cells. And you know, I was talking to those guys, and those cells look a lot like some of these glioblastoma cells." And we began talking about it and we said, "You know, do you think maybe…glioblastoma could be a tumor of neuro stem cells? And if that's true, do you think maybe when we take tumors out of patients, we could actually find these stem-like tumor cells? What happens if we take tumors out of patients and then grow them in conditions that are conducive to the growth of normal neuro stem cells?" And sure enough, the minute we started doing that, we started seeing this outburst of these little round conglomerates of cells that looked just like how normal neuro cells grow, but when we looked at them genetically and so forth, they had all the characteristics of a tumor. And so we became one of the two groups in the world that first learned how to identify and grow outside of patients these things subsequently called glioma stem cells.
Erin Welsh: I mean, that's an incredible, uh, accomplishment and really opened the door to, to better understanding this condition. Can you explain how this was such a pivotal moment in uncovering why glioblastoma doesn't respond to treatment like other types of cancer?
Dr. Fine: We've known always, we've known for 100 years that the bugaboo about this tumor is its diffuse infiltration and it spreading into the brain. But what we now know is they're not just anatomically spread out. The cancer cells, they're functionally integrated into the brain. They come from something that's pretending to be a neuro stem cell. And so as this tumor develops, all the tumor progeny of that stem cell wants to be like a normal progeny of neuro stem cell, wants to be a neuron, a nerve cell, an astrocyte, an ependymal cells. And what do those normal cell types do? They make connections with each other. They, they make neural networks. Well, it turns out what we've learned in the last five to 10 years is that's exactly what glioblastomas do. It's basically hijacking the normal functioning brain by interdigitating not just anatomically, but functionally, and there is no other tumor that does that. I actually wrote a, a prospective paper in Cancer Discovery last year where I proposed that what GBM actually is, is a developmental aberrant organ system that is integrated in the normal organ, which is the brain. And so it's trying to do its developmental thing like a normal brain, but it's screwed up. And in doing so, it's gonna take the patient with it.
Erin Welsh: And I imagine that means that research and modeling methods to find those treatments have to shift, too. Can you talk a little bit more about how traditional study models like mouse brains, for instance, or culturing neural cells, how that doesn't quite work as well for glioblastoma?
Dr. Fine: You know, it was quite clear after 15 years or 20 years in the field that what I was seeing in the clinic, the way the tumors were behaving, what they look like under the microscope, wasn't matching what we were studying in the laboratory. You know, I, I was doing very well, you know, academically, and people say, "Oh, Howard, you know, does translational research." And at some point I, you know, I had to wake up with myself and said, "You're not really doing translational research because what you're publishing about and what you're studying in the laboratory has no realistic applications to the patients you're caring for. And if you have to be truthful with yourself, you're probably never gonna develop anything in the laboratory that has a realistic chance of helping these patients." So you either continue to, you know, academically be successful and clinically be a good doctor and take care of patients, or you're gonna try to make a difference with your work in the laboratory that can translate. And if you're gonna do that, then you better study something that looks more like what's happening in patients. We need to understand the microenvironment that this tumor not only grows in, but as we've been talking about, interdigitates and functions within. What that essentially means is we need a model of the human brain.
Erin Welsh: Right. You need to actually culture tissue, differentiated cells, not just one cell type, not just with one environment, but recreate a bunch of different environments to see how a tumor cell will act. And that's what led you to create the GLICO platform, which is this patient-specific organoid model for glioblastoma. Could you take me through the basics of this platform and its different components?
Dr. Fine: Yeah. So the “aha moment” happened about six or seven years ago when a paper was published in Nature that showed you could take totally potent stem cells and through a series of coaxing them with certain types of cytokines and differentiating agents, not genetically, and different bioengineering techniques, you can coax those embryonic stem cells to form basically a mini brain, something we call cerebral organoid. It had many of the same characteristics of a young developing human brain. We said, "Wait a minute. This is what we've been waiting for, for the tumor." So we already had our glioma stem cells. Now all we had to do is learn to retro-engineer our glioma stem cells into these mini brains, into these cerebral organoids. All you have to do is within these little bioreactors where we grow these cerebral organoids, put these glioma stem cells, and within a few hours, they immediately attach onto the cerebral organoid and start invading. And for all the world look just like the surgical specimens that we see with human glioma. These tumor cells home to their home, and their home is the human brain, and they see these cerebral organoids as the human brains. Now we have immune cerebral organoids, and we co-culture the patient's own glioma stem cells that we get from the operating room. We make their own cerebral organoids, so now we have fully genetically matched cerebral organoids, fully genetically matched microglial immune cells with the glioma stem cells growing within the patient's own genetically matched environment.
Erin Welsh: And so you, you are able to create for each patient a model of how a glioma might progress in that particular patient.
Dr. Fine: Exactly. We can model the patient's own tumor within their own brain, within their own immune system. They aren't just diffusely infiltrated into the cerebral organoids. They're functionally integrated. And just to give you an example of what that means potentially just therapeutically, if we take a given patient's glioma stem cell line and grow it in a Petri dish and expose it to, say, a concentration of 1 millimolar of a chemotherapy agent, let's say temozolomide, the standard treatment for glioblastoma, and at 10 millimolar let's say it kills all, all the tumor cells. If you then take those same glioma stem cell lines and put them in GLICO, they are 10 to 100 times more resistant to that same dose of temozolomide as those cells when they're grown by itself. And that's why we believe that if you're gonna screen for drugs for this, this disease, it has to be within this type of context.
Erin Welsh: That, I mean, that makes complete sense, and this seems like it's really filling this gap where the preclinical studies are actually going to completely recapitulate what will happen in a clinical trial because these models can match what's happening in a real-life setting in a glioblastoma. And so with that in mind, how are you using GLICO to investigate potential therapeutics?
Dr. Fine: Because of this new biology, it offers the opportunity to begin to look at new potential therapeutic targets that other cancer biologists and oncologists never thought about. Instead of, you know, targeting these oncogenic pathways and looking at DNA-damaging agents, we're looking at, well, if you can disconnect these tumor cells from their connections with neurons, maybe we should be looking at drugs that allow us to inhibit those connections. And now maybe some of even the traditional therapies might be significantly more effective. So, and you know, well, do we have drugs like that? Well, we sure do, but they're not in the cancer realm, right? As far as these malignant synapsis, do we have any drugs that can inhibit them? Yeah, they're called anti-seizure medicines. They're, there's all the anti-psychotic medicines that are neurotropic that attack ion channels. So the point is, we have the opportunity, and that's one of the, our major focuses, is using the model to identify very novel types of therapeutic targets that people studying standard cancer don't think about because it's not relevant. But this is not another cancer, right?
Erin Welsh: You mentioned this being a paradigm shift in the way that we treat glioblastoma, and I think there's also seems like a paradigm shift in the way that we conceptualize cure or what the overall goals are. So what does that mean for you? Why might eliminating all cancer cells not be the right goal for glioblastoma?
Dr. Fine: About 80% of patients have what we call a complete radiographic resection when they go to surgery, so there's no tumor cells left. And after they recover from surgery, most of these patients are totally fine because all they have is a few tumor cells scattered around and we can't see them. And if that was all there was, these patients could live normal lives, both functionally and probably, if their cells didn't grow, longevity, right? So if we can just prevent those tumor cells from beginning to interdigitate into the network, which gives them the ability to thrive, to grow, to proliferate, we don't have to kill every last tumor cell. We just need to disconnect them.
Erin Welsh: So it's not only offering this personalized approach to which sequence or combination of drugs might work best, but also rethinking the way that we are treating this cancer overall. And I'm curious, what's the next step to really shift the paradigm for this disease and impact patient lives?
Dr. Fine: This GLICO generation and this high throughput drug screening, you know, it takes me six to 12 months to train a PhD geneticist to make these GLICOs. It's incredibly resource-intensive. This is not something that's gonna be done at most hospitals or almost any hospital. So how does that ultimately change the field? We have a couple hundred or more patient-derived glioma stem cell lines. Each has been extensively characterized as far as their genomics, their transcriptomics, their methylation profiles, and other molecular features. We're planning on taking these 200 retrospective glioma stem cell lines, doing high throughput drug screens, finding the drugs that are most effective for that particular glioma stem cell line, and then taking all this molecular genetic data for those stem cells, all the corollary clinical data, and we're developing AI-predictive models so that the hope and dream is five or six years from now, we don't have to make GLICOs to predict what's going to be the best patient. The GLICO gives us what's called the ground truth data. We have the ability to do thousands of drug screens in a clinically relevant way on a patient-by-patient basis and plug that into the algorithm.
Erin Welsh: Right, really u- upscaling...it's not even precision medicine at that point. It's, it's upscaling-
Dr. Fine: It's the next level of precision, exactly.
Erin Welsh: Exactly.
Dr. Fine: That's exactly, that's the next level. And so the hope would be then, then Mrs. Smith comes in with a glioblastoma at any hospital, you know, 10 years from now or less. Tumor's taken out, they do the standard, you know, genetic screening, and they plug it into the algorithm, and they say, "Here are the five drugs that have the best chance of helping her."
Erin Welsh: I mean, that is, that would be transformative entirely for this field that up to this point, like you said, has not received very much hope. And over your career, you have interacted with thousands of patients with glioblastoma, and I'm, I'm curious to hear your thoughts on how that experience has shaped the way that you view treatment goals for your patients.
Dr. Fine: I hope the legacy I leave is not just the science, but maybe the people I train. You know, most of my laboratories have always been mostly PhD scientists, so the, the clinic drives the laboratory, and what the laboratory is doing is ultimately trying to get back to the clinic, and that's what drives me. Watching that many patients and, and helping that many patients go through the death and dying process has made me focus my science on something that has a chance of making a difference as opposed to just being academically successful. And I- if we're doing science that ultimately doesn't really have anything to do with patients, then I'm not interested.
Erin Welsh: I really appreciate you taking the time to chat with me, Dr. Fine. This has been such a thought-provoking and really inspiring conversation, and I'm so excited to see where this research leads.
Dr. Fine: It was great speaking with you. Thank you.
Erin Welsh: Thanks so much to Dr. Howard Fine for talking with me today about the extraordinary advancements he's pushing forward to transform the way that glioblastoma is understood, studied, and treated. I'm Erin Welsh. Advances in Care is a production of NewYork-Presbyterian Hospital. As a reminder, the views shared on this podcast solely reflect the expertise and experience of our guests. To listen to more episodes of Advances in Care, be sure to follow and subscribe on Apple Podcasts, Spotify, or wherever you get your podcasts. And to learn more about the latest medical innovations from the pioneering physicians at NewYork-Presbyterian, go to nyp.org/advances
