Podcast: Selecting Second-Line Therapy for Relapsing/Remitting Pediatric Low-Grade Glioma (pLGG)

Eugene Hwang, MD: Hi. I’m Eugene Hwang. I'm a pediatric neuro-oncologist at Children's National Hospital in Washington, DC and excited to talk to you today about pediatric low-grade gliomas and how to select salvage regimens in those tumors that have recurred.

Moderator: What are some important considerations when choosing a treatment regimen for pediatric low-grade gliomas?

Eugene Hwang, MD: When you're considering treatment, salvage therapy, for a pediatric low-grade glioma that has recurred, it's probably recurred after, in this day, upfront primary chemotherapy. That chemotherapy tends to be IV, whether it's vinblastine and Avastin, or vincristine-carboplatin, or TPCV, or something along those lines. But it's important to think about whether you understand the molecular phenotype of that low-grade glioma when you're thinking about recurrent therapy. The reason is because we think of these tumors as predominantly single-driver tumors, which means that you have the opportunity to consider targeted therapy in the recurrent setting.

Moderator: Are there any other considerations?

Eugene Hwang, MD: It's also important to think about when you're contrasting salvage therapies, what the cost-to-benefit ratio of those patients are. Should you go back in and re-resect a tumor that has grown, understanding that resection may cause permanent morbidity for that patient? Should you redo a chemotherapeutic approach? So, we've had a long history of understanding the benefit and risk to that kind of treatment, and there are several that you can choose from in that context. Or should you flip to a targeted therapy, which again, really requires you to know what the particular molecular phenotype of that child's low-grade glioma is.

Moderator: How and when do you decide to use a targeted therapy?

Eugene Hwang, MD: There are other components to really think about why you should switch to targeted therapy or why you should consider a targeted therapy instead of doing another surgery or another chemotherapy. I think that in the context of a tumor that we really think of as one that's going to be survived, these children will all hopefully live relatively normal lives. Then the importance of mitigation of impact to that child becomes much more important. If this were a child that had an aggressive cancer, there's no question—whether it takes that child out of school or causes side effects that's just what you have to do. But when you don't have to do that, then suddenly this whole realm of considerations need to really enter into the mathematics. I think the first would be: What's the impact to that patient and that family? First, the logistical administration of those drugs.

So, nothing about efficacy and toxicity yet, but what do you consider in your decision-making process? When we think about that, the mechanics are relatively clear. For most chemotherapies, that's IV weekly. You have to have a line. You have to be worried about fevers. You have more nausea. You have more fatigue. You have to, obviously, go in if you’re neutropenic and be admitted to the hospital. You might need transfusions. So toxicities that I think are not common in IV-based chemotherapy, low-grade glioma regimens, but exist. And that also requires the family to be coming in once a week, and I think that for those treatments, while they are certainly tolerated and important when that's the only option, still have a negative impact on the job and livelihood and logistical support needed for the family to go in and get those treatments. Now, there are some factors that we think about in terms of, okay, you've thought about chemotherapy and now you're thinking, okay, well, what are the targeted therapies that I might use?

We already talked about how you really do need to know what the molecular phenotype is. When you have a BRAF fusion, you tend to be very sensitive to both second-generation BRAF inhibitors as well as MEK inhibitors, but first-generation BRAF inhibitors can cause a paradoxical activation of the tumor. So it's something that you certainly don't want to just try without knowing because you may actually accelerate growth rather than stabilize or regress that tumor. That's one concern with neurofibromatosis, that is probably one of the most common genetic predispositions that cause low-grade gliomas in children. And in that particular patient population, there's a concern that first-generation BRAF inhibition will actually accelerate that tumor's growth rather than inhibit it. And so, there's a little bit of extra consideration that needs to be thought of when you know either that patient SNF or whether that patient is deserving of a BRAF inhibitor at that point.

It is true that efficacy and toxicity between the MEK inhibitors and the BRAF inhibitors tend to be fairly similar, meaning it doesn't seem that the MEK inhibitors are particularly or significantly different within themselves in terms of inducing regressions or causing overarching toxicities, although there are subtle differences, and so understanding what those are is very important as well. Again, because the location of many of these tumors and the way they grow back, especially in the thalamus or the optic pathway glioma, surgery just isn't a viable option in those situations. In those situations, too, if you've had chemotherapy, which is really the context that we've tested most of these MEK and BRAF inhibitors in, and so we've seen such tremendous signals of efficacy and minimal signals of toxicity that it seems quite reasonable to consider targeted therapy when that patient has received chemotherapy before. When you think about whether that patient has received prior targeted therapy before, then there's probably some more nuance.

So, in general, what you're probably suggesting is that you have a patient that had a recurrent low-grade glioma, they must have responded, or perhaps they didn't respond well, to an initial salvo of targeted therapy, and the question then is like, what do you use after that? I think that there remains some equipoise about whether you can switch between the MEK inhibitors. Again, I said that the efficacy-toxicity doesn't differ dramatically between them, but it's unclear that switching to another agent of the same class will induce a different response than the first agent would, which you have experience in utilizing, of course. There's some caveats there: If the first agent worked but had a lot of toxicity or vice versa, then perhaps there's some good rationale to trying a different in-class MEK inhibitor for instance. I think that we've also looked at the second-generation RAF inhibitors as a logical place to go.

Sometimes, when somebody has had prior targeted therapy with MEK inhibitors, for instance, the reason really being that in some ways, if you use that calculus we talked about before, how hard is it to take this medicine from a logistical standpoint, a quality-of-life standpoint, and a toxicity standpoint. Then the ability to go to once-weekly dosing, as opposed to once or twice-a-day dosing, especially when we used to think you had to be completely NPO for hours on either side of that dosing for the MEK inhibitors, that the once-weekly dosing is attractive logistically. And so, again, you think about somebody who's had targeted therapy with a MEK inhibitor, many of those would be eligible for second-generation RAF inhibitors, and that logistical component of consideration would be something that you would say, well, maybe taking it once-a-week is better. And again, these are things that really deserve good conversations with the families about the pros and cons in a partnership in moving forward from that as well.

Moderator: How do you monitor patients on targeted therapies?

Eugene Hwang, MD: In any situation where you're using targeted therapy or chemotherapy, there are often questions in low-grade glioma about how to tell if it's working or how do you know that that tumor is coming back? How do you know when to bite the bullet or pull the trigger to restart treatment? Even the oral targeted therapies are kind of a pain. You got to go into the clinic at least a cadence that you wouldn't have had to before. You have to suffer the consequences of taking medications and having side effects even if they're milder than I think the most intense side effects that we have in our practice. So, I think that there's really two arenas that we evaluate. The progression of a low-grade glioma, the biggest is, of course, radiographic, and that, of course, is with MRIs. Those MRIs tend to have to be contrast-enhanced because a lot of these tumors do enhance nicely.

But, interestingly, the metric of what we think about today under the umbrella and auspices of the RAPNO guidance is looking at the T2-FLAIR signal, actually, rather than the enhanced signal. So, when we think about high-grade gliomas, especially in the adult world, TG-FLAIR is not a huge tent pole for their assessment of whether that tumor is gone. It's really the enhancing component of the tumor that they really focus on. But in pediatric low-grade glioma, it really should be the T2-FLAIR signal, and that's something that's pretty important to think about when you think about whether a tumor is growing. The second piece of imaging that you really think about too is what do you do with cysts? Because cysts are fluid-full pockets, but they may actually be quite big and there's no question that they'll impact the functioning of the brain around them if they get big enough or if they're in just the wrong space.

So that's a complex component. We try to use imaging measurements to really assess what the solid tumor growth or stability is. It's not always easy, especially if the cysts are more medial or if they're more than one cysts. And then there are certain tumors that have a more microcystic disease where they're small cysts everywhere, and so that becomes challenging from a radiographic standpoint. In addition to which it's not like most of these tumors are little balls, they sometimes tend to have pretty esoteric outlines and so where you sort of measure the growth, or the recurrence, or the response of that tumor, becomes complex. And so there's a lot of, I think, interest in not only partnering with AI, but using radio imaging to look at volumetric analysis of these tumors, hoping that you can get a really good sense mathematically whether this tumor is really growing or not. Again, with a caveat that it's hard to know what to do with cysts when they're in certain areas knowing that that's not true tumor growth, that's just fluid growth.

The other real area that we sort of focus on in terms of assessing whether somebody is recurring or not is of course in their symptomatology. So, sometimes a patient has an MRI scan that looks really stable—not stable-ish, not stable but only between the last three months, not if you look over the last year or two—but truly stable. And yet that patient may be deteriorating visually or rhetorically or they just may have some complexities in their day-to-day life functioning, their functional scores, which really are probably attributable to the tumor even if it hasn't been looking like it's growing more. These are trickier places. It is easier to think about initiating therapy in a tumor that is clearly radiographically progressing than somebody who is having symptomatic deterioration, especially when our metrics of symptomatic assessment are kind of subjective. And so, I think the easiest place to pull that trigger is visually, if somebody is losing vision, then we'll often use targeted therapy.

Sometimes Avastin, which we think about as more rapidly inducing a response, I think, to protect vision. But even the targeted therapies, really, because if you're starting to see visual deterioration, you certainly don't want to wait until it's too late, and you also need to acknowledge that even effective therapy may take some time to reduce the size of that tumor—the pressure on the optic pathway—and so starting at an earlier time point might give you a little bit more leeway when you're thinking about visual deterioration, for instance. So, it's still difficult to use, I think, functional outcomes today to pull the trigger, but in certain situations, I think it's a really important metric to consider, and it's also not a slam dunk always about when to start treatment from radiographic progression. If you've somehow managed to, or the tumor doesn't have a lot of cysts and you're using the T2-FLAIR and you think that tumor is growing, then you have to think a little bit more about what the context of that growth is.

Is that growth really close or even already impinging upon the optic pathway and you need to really be careful about any little growth and pulling the trigger to more treatment? Is that a space in the posterior fossa that's in the resection cavity and really it could grow for a long time, it probably wouldn't do anything or really alter the kinetics of that patient's outcomes? I think those are really important considerations. Also, when and what did that patient just get treated with? Sometimes we see these situations, certainly with bevacizumab, but also with some of the targeted therapy, when towards the end of therapy or after that therapy, you can sometimes see what we consider a rebound effect. What that means is—what I've heard it described as—is the tumor sort of takes a deep breath and kind of re-expands, and you look at it on the MRI scan and you think, oh my gosh, that tumor is definitely getting bigger.

And by measurement there's no question it's more than a few millimeters larger. And then, even without any therapy, a lot of those patients will just stabilize out. And while the molecular underpinnings of that phenomenon aren't totally understood, it's important to think about, on those first few MRI scans, can you hold the line and not restart therapy for just a little while, hoping that this might be that sort of relaxation or rebound that does not portend true progressive disease? Again, those are all in the context of, well, how much time can you give a tumor to get bigger based on what it's going to hurt in the patient, or potentially hurt in the patient, if it gets much larger?

Moderator: Are there other factors to consider when selecting a second-line therapy?

Eugene Hwang, MD: We've talked a little bit about how to make some of these decisions—targeted therapy, surgery, chemotherapy—and one of the themes around the targeted therapy has remained quality of life. When we think about starting therapy or what therapy to start, in low-grade glioma world, where these kids are predominantly cured, we have to think about long-term outcomes, and we also have to think about short-term impact. We have to think about how hard is it for this child to get treatment A or treatment B; how hard is it for the family to support the child to getting treatment A or treatment B? We know that the financial toxicity of children getting treatment for any kind of tumor is extraordinarily high. Lots of our families who are doing just fine lose houses, lose family coherence, the finances really come apart. These are things that sometimes we don't see as directly as physicians.

We rely on our social work team, but it's something that is very important for us to consider as well. So, when we think about quality of life in a child that doesn't necessarily need to be admitted all the time for intensive inpatient chemotherapy, then it's really important to sort of attribute a quality of life to each kind of treatment. And while there are outliers in every kind of therapy that you might choose, it's a place where the targeted therapy has really, I think, shown some advantage over some of the more in-person IV therapies. And that's because that parent who brings that child in doesn't have to miss a day of work, and that child doesn't have to miss a day of school. And this is purely logistical from both a financial standpoint but also a time standpoint. The toxicities or side effect profiles of these agents are things that are super important to think about as well. Are these drugs or treatments that will in general kind of be okay with the child but will kind of make them feel kind of crummy and so they won't really be able to really thrive in the way that we hope they will. and in that avenue as well, again, while there's pretty significant overlap between the IV chemotherapies and the targeted therapies, in general, the targeted therapies have seemed to be more amenable and permissive of a relatively regular quality of life. We've had a few patients, and I think that those numbers will get smaller as we move a little bit away from IV chemotherapy, who were treated with IV chemotherapy and then treated with oral targeted therapy and they mostly hands down have said they would never choose IV chemotherapy having known both of those things. Even those families with kids that didn't have a lot of problems with the IV chemotherapy, I think, would still choose the oral targeted’s in this day and age if that decision was left up to them, and if that efficacy signal, which is also something that's really important to consider, was considered, essentially. So when you're talking to families about these things, I think that especially in some of these upfront randomized trials where we have both chemo and oral targeted therapy as options, it's important to talk about the pros and cons of all things, and that always includes the potential side effects, the known and the unknown.

We know a lot about 50 years of chemotherapy. We know very little about 50 years of MEK inhibition or BRAF inhibition at this point. And also the logistics…and while the families are often like, ‘Well, whatever is best for my child, I’ll do,’ I think that the proof is sort of in the pudding—that again, when those families come back after doing both of them, they will almost invariably pick the one that's easiest for them because it seems to be at least equivalently efficacious, and that tends to be the targeted therapies.

Moderator: What do you think the future holds for low-grade gliomas?

Eugene Hwang, MD: So, I think that the—really, in general, the future of low-grade glioma assessments probably revolves around three areas: One is radio imaging. One is outcome metrics that are a little bit more objective than a feeling that someone is better or worse neurologically or visually. And the third would be something in really investigational, like cell-free DNA, for instance. When you're talking about imaging, there's been a lot of work about trying to decipher the best way to image these low-grade gliomas, and we talked about some of them. We talked about the volumetric components that are really being studied closely now. Understanding that we think that that actually may give us a better sense of a tumor that is growing or not growing, although a big caveat there is going to be because we haven't had volumetric assessments previously, they're going to have to go back and look at a lot of MRIs from a lot of kids over many years to see whether the volumetric predictions will be a better or more accurate depiction of whether that tumor is getting worse or not getting worse.

Essentially, there are efforts currently to sort of standardize neurologic outcome measures. Ones that can really talk about people's ataxia, their visual motor coordination, maybe their cognition, all the components that any kind of pediatric brain tumor, but especially one that lends itself to a long life, are able to be assessed on. And those, I think, will be incorporated, and are being incorporated, into some of the trials now so that we understand best, well, if you need to pull the trigger on re-treatment or you want to see whether somebody is really not doing well, even if their tumor isn't growing, these objective outcome measures will be important, too. And then, I think that we've all sort of heard about the advent of this really incredibly interesting technology that can be extraordinarily sensitive about detecting fragments of DNA from a tumor, particularly from the CSF. That's an area that I think we're all excited about, especially in our malignant tumors. It may be a technique that really gives us an insight into the MRD of these cancers. Where their utility is for low-grade gliomas is a little less certain. It tends to be a little harder to detect some of the molecular oscillations in low-grade gliomas than it does in our more malignant and spreading tumors. But at the same point, the ability to lump low-grade gliomas into the technologies that are being developed for the much more aggressive malignancies will be an incredible step forward as well.