Andrew Wilner, MD (Host): Scientists have recently found a highly accurate way to predict the best treatment for patients with meningioma based on patterns of gene expression in their tumors. Dr. Stephen Magill, Assistant Professor of Neurologic Surgery at Northwestern Medicine, was co-senior author of the study published in Nature Medicine. He joins this episode of Better Edge to discuss this innovative test that could change treatment for 1 in 3 patients with meningioma.

Welcome to Better Edge, a Northwestern Medicine podcast for physicians. I'm your host, Dr. Andrew Willner. Dr. Magill, please give us some background on the study you recently published in Nature Medicine. What motivated the research team to explore meningioma risk stratification and indications for postoperative radiotherapy?

Stephen Magill, MD: Thank you very much for having me. The motivation for this study was to understand when we need to escalate care for meningioma patients. Meningioma is the most common primary brain tumor, over 40,000 a year in the U.S. Of those, 80% are grade 1, about 20% are grade 2, which is about 8,000 patients a year, which is almost as many as there are glioblastomas in the US. If you have a grade 2 meningioma and we take it out, we don't have any marker to know whether you'd benefit from radiotherapy. The motivation of the study was to try to create a biomarker that could predict who responds to radiotherapy. We also know that some grade 1 tumors that should be benign, you take it out, they don't come back, are more likely to come back, and we don't know why that happens. So, we wanted something better than the World Health Organization grading to predict who is going to come back, which tumors will come back, and then also to figure out who should receive adjuvant radiotherapy after a meningioma is taken out.

Right now, there are two randomized trials that randomize patients to six weeks of radiation versus observation for meningioma. Very few cancers actually have a randomized trial like that going on, because we don't have any biomarker that will predict who would benefit from radiation. We can say, "Hey, your tumor looks a little bad. There's residual," but do we do all the radiation or not? And that was really the motivation for the study to try to answer that question, which tumors will come back and who benefits from radiation.

Host: Well, from the patient's point of view, that's really essential information. It's like, "I don't want radiation therapy if I don't need it," right? On the other hand, "If I need it, I want it." And you don't have an answer, doc. So, okay, now, back in the 20th century, when I went to medical school, there was a grading system for meningiomas based on what it looks like under the microscope. So, are you telling me that that's insufficient?

Stephen Magill, MD: Exactly. Because as a surgeon and someone who cares for patients with meningioma, we see patients who have a grade 1 tumor, shouldn't come back, come back every once in a while within a year. We don't know who that's going to be. I also have patients who are a grade 2 tumor, you take it all out and it never comes back in eight to ten years. So, the grading scale says, "Hey, this is a worse tumor," the World Health Organization grade, but it actually behaves much more benignly. And then, we also have ones where it says, "Hey, this is a benign tumor," and then it keeps coming back, or is very invasive. So, I think there's more that we can add to the World Health Organization grading.

Host: You know, it used to be when that would happen, patients say, "Why did that happen? You told me it was a benign tumor and it's not." I would say, "Well, it's bad luck." It sounds like we now have a more sophisticated answer to that, that it has something to do with the gene expression. So, tell us about your study. What did you look at and how did you make some sense out of this?

Stephen Magill, MD: Yeah. Those are the hardest conversations to have when the patient says, "Wait, this was benign. Why did it come back?" And I always say it's bad biology. And so by studying the biology, the mechanisms that are driving tumor growth, that's what gives us insights into actually differentiating not just based on how it looks under the microscope, but what is the biology of the tumor. And so, we did a large study that was published in Nature Genetics a couple years ago, looking at DNA methylation. And we found some of the mechanisms that predict bad behavior.

When we were looking at the molecular profiling on those patients, in many of them, we had done gene expression profiling. And we looked at that and we actually found there's a 34-gene signature that could predict even better than the methylation profile, better than the World Health Organization grade, better than any of the other grading scales that have been proposed incorporating other molecular features like copy number changes or things like that. When we looked at it, the gene expression panel we found was an independent predictor. So, it's actually adding new information, not just another way of looking at the same information that you get from methylation or from copy number changes or other markers of the tumor biology.

Host: So, this sounds a little bit out of the wheelhouse of neurosurgery. How did you come to sort of get involved with the-- I mean, you need to be a geneticist and understand what's happening in the cell. How did you sort of incorporate that into your training?

Stephen Magill, MD: Yeah. It's actually what led me to meningioma, because I trained as an MD PhD. I did a PhD in molecular biology and-- well, it was neuroscience, but my actual work was molecular biology, gene regulation, gene transcription factors, microRNA. So, I was looking at all of those things that regulate gene expression.

So, I was interested in that. And then, I got drawn to neurosurgery as a clinician. And when I did my residency, I spent some time in the lab, did a postdoc fellowship for one year studying cancer biology. So, all that came around as a physician scientist to help me focus on brain tumors. So, my whole clinical practice is just brain tumors. I don't do spine or vascular or some of the other neurosurgical things. And nowadays, I think to be a modern tumor neurosurgeon, you really have to have a handle on these molecular biology, the features of the tumor, what's going on biologically, because the biology of the tumor actually guides what I do in the operating room.

And I think that's why these advances are so important. If I know a patient has a high grade meningioma that's recurrent, I'm going to be much more aggressive surgically. Because if I can remove that, do more of an oncologic resection, get clean margins, that risk is acceptable and is worth it for the patient because taking on that additional risk is worthwhile. Because if they get long-term disease control because we actually remove all of the tumor and don't leave a little cell here or there, they can do much, much better. Whereas if this is a grade 1 tumor and it's benign, then any deficit is not worth it because we could treat with adjuvant radiotherapy and get long-term tumor control. But if you don't know that, and if we don't know the tumor biology, and we don't understand the biology of the disease that we're treating, then actually, I don't think we provide as good surgical care. So, to me, those two interests really blended together and it's really enjoyable as a career because we're making discoveries and then, I can tell the patients, "Hey, this is why I think we should be more aggressive." And, in the flip side, once I do the surgery itself and I follow and care for these patients, this is why we should do radiation or we shouldn't.

And that's really what this gene panel adds for us because I can then sit down with the patient and say, "Hey, they said it's a grade 2 tumor, but I got it all out. And the gene expression profile says you're actually low risk." So, those patients that we treated with radiation had no extended survival. They did great regardless of the radiation. So, let's not do radiation and we'll just follow you with serial imaging. So, I think understanding and pushing these things to me, it directly informs what I do with my patients.

Host: Wow. That's fantastic. You are the guy for the job based on your training and your skills. Now, I'm going to ask you a few questions. One, is this gene expression assessment of a tumor, is this available everywhere, or is this just something that you guys are doing at Northwestern? Or in other words, if I go to the doctor tomorrow with my meningioma, is that something they're going to do? Is that part of state-of-the-art, or is that a new thing?

Stephen Magill, MD: We are in the process of commercializing it. So, it is not yet available. Patients have started emailing us after the study saying, "Hey, can I get this? I'm having this conversation with my doctor. Can you run the gene expression profile?" Right now, it's in the lab. Two of the NRG clinical trials, the large network for clinical trials, are being planned now and are actually incorporating this as a biomarker. So when we start deciding for these recurrent meningiomas, are we going to do radiation or not? Are we going to start testing for the gene expression panel? And we're working to bring it to commercialization, but it is not there yet. So, it's not available.

In the interim, we have to use the other molecular information that's been published. So, what are the copy number changes in a meningioma? There are different grading scales proposed where we incorporate that, and I use that to inform what I do for my patients now, because it's the best that we have clinically. And it is step up from the WHO grade. And it's changed what I do. I have some patients that have grade 2 meningiomas I'm just watching because their biology, the tumor, the methylation, the copy number profile, those things that are available commercially say you're a low risk tumor. So, you had a great surgery, you had a low risk tumor, let's not do radiation. But it's really needed, and we're working hard to get it out commercially.

Host: Let me ask a very pragmatic question. So, it sounds like this is important, this gene expression test. Now, a patient presents with a meningioma, you don't know what the gene expression is because you don't have a piece of tissue, right? So, would it be worthwhile to biopsy that tumor before the resection to get an answer, or do you just do the resection as best you can and then get the answer later? How do you approach that?

Stephen Magill, MD: That is actually a conversation I have with patients, because the radiation dose we treat a meningioma with also differs from a grade 1 tumor that's treated at 50 gray versus a grade 2 tumor that we go up to 54 gray if you were treating with fractionated radiotherapy. Many meningiomas are treated primarily, and I have several patients in my practice who we've had that discussion. You could treat it primarily with radiation, but we don't know what the grade is. And finding out that grade would actually change the radiation dose that we would prescribe. And we have gone forward and done a biopsy of that, where, you know, it's in the cavernous sinus, it's in an area where resection would be more morbid than radiation. And we want to treat with radiation, but we'll go in and do a biopsy to get that information.

It also is the foundation of our future research, because now that we have these tools, what we are starting to do is to look for radiomic features on the patients where we've taken their tumors and say, "Hey, we know this is the radiomic features of a gene expression profile, low risk, medium risk, or high risk." And we can then start to have a better idea, and we can look at the tumor and say, "Hey, it looks like a meningioma. You want to treat with radiation, but this actually looks based on the radiomics like it may be a higher risk. And therefore, that might push me more to get a biopsy," particularly it comes up with the asymptomatic patient.

You know, someone who has symptoms, big tumor, like yesterday I took out a seven centimeter one. The lady was hemiparetic. And so last night, we took out this large tumor. Today, she can move her arm. She couldn't open her hand before and now she's like this, I walked in the room just before I came here and, you know, her strength is back. That patient, she needs the tumor out. There's no question there.

But if you come in, you had a car wreck, you find an incidental small meningioma, what do we do about this? That's really, I think, the exciting part moving forward for the next steps is, as we have these biomarkers, can we then correlate it with either radiomic, like non-invasive tests, or a blood sample where we could sample something from your blood that might tell us what's going on with the tumor? So, that's what we're pushing now in the lab and, you know, sort of where we're at on the research front.

Host: One last question. I remember about six months ago, I think, the hot new thing was methylation. Was it a methylated cell, tumor cell or not? Is that still important? And where does this fit into that?

Stephen Magill, MD: Yeah. You know, that's a great question. DNA methylation is very good at predicting clinical outcome, like recurrence. But when we looked at that in our samples, we could not find a methylation signature that predicted radiotherapy responses. And that's why the gene expression panel was really exciting and I think that's why it got into Nature Medicine and the press that it's gotten because people have looked for this and we haven't found it. But we found a modality that actually does predict the radiation responses.

So at this point, DNA methylation is also a little on the expensive side. It's not available widely and there's variations in the processing pipeline. That, I think, will have a usefulness. But widespread clinical, like you have to do this on every meningioma. I'm a little less bullish on that taking over as the standard of care for meningioma/ something like the gene expression panel, which is 34 genes, it is relatively cheap, very simple analysis. There's not as much post processing and variability that can come as you get with DNA methylation. So, I think it will actually have a chance to be something widely used. And we've seen that in breast cancer and prostate cancer where you get a gene expression panel that says you're this type and then that guides your management. And it's much more affordable and easier to disseminate.

I think methylation is interesting. I think it's helpful. It's a great tool for discovery because it's a genome profile. And it's also very good if you don't know it's a meningioma. Like, methylation is good telling meningioma from something else. If you don't know what it is in brain tumors, it's very useful in the pathology gray areas. But I think, within meningioma specifically, I don't see it being a huge, widely must-have. But at this point, you know, I think the gene expression panel really could be that.

Host: Okay. Well, that was so interesting. I'm going to ask you one last question, which is what are the next steps for your research?

Stephen Magill, MD: The next steps for our research are really applying this in a prospective manner, number one, so we can validate it prospectively. That takes years. We're starting some clinical observational studies to validate that. And then, the things that we began talking about, like how can we do this now, how can we get the answer to this without having to open your head and get access to the tumor? So, either radiomic or blood based correlates of what we're seeing on those tumors. And I think that will alleviate the need for a biopsy, provide really informed patient care upfront. And as I had talked at the beginning, knowing whether it's a grade 1 or grade 2 tumor can affect your surgical approach, your surgical aggressiveness. Whereas if it's a very low risk tumor, leaving a little bit in a high risk area is not a big deal. Save the patient a deficit. They are going to live for years with their meningioma. But I think if I know that ahead of time, that really gives me power as a surgeon to make the best decision intraoperatively for my patients. And so, I think that's where we'll see that translated. And I'm excited to see that happen.

Host: Dr. Magill, thanks very much for this great discussion on advances in meningioma treatment.

Stephen Magill, MD: Thank you, Dr. Wilner.

Host: To refer your patient or for more information, head on over to our website at breakthroughforphysicians.nm.org/neuro to get connected with one of our providers. And that wraps up this episode of Better Edge, a Northwestern Medicine podcast for physicians. Thanks for listening.