Activity Transcript

Eudocia Q. Lee, MD, MPH: Hello. I'm Eudocia Quant Lee. I'm an associate professor of neurology at Harvard Medical School, and director of clinical research in neuro-oncology and a senior physician at Dana-Farber Cancer Institute in Boston, Massachusetts. Welcome to this program entitled "Current and Future Directions in Neuro-Oncology."

I am pleased to be joined today by several of my colleagues. Dr Gavin Dunn is an Associate Professor of Neurosurgery, Neurology and Pathology. He works in the Immunology Center for Human Immunology and Immunotherapy Programs at Washington University School of Medicine in St. Louis, Missouri. Dr Jennifer Moliterno Gunel is an Associate Professor of Neurosurgery, the Clinical Director of the Brain Tumor Center, and Chief of Neurosurgical Oncology at the Yale School of Medicine in Yale New Haven Hospital in New Haven, Connecticut.

Welcome. Today, we are going to be discussing some of the latest developments in neuro-oncology and potential future directions. To begin, let's discuss how we treat glioblastoma.

First, some background. As many of you know, glioblastoma is the most common type of malignant primary brain tumor. It accounts for a significant portion of morbidity and mortality within patients with primary brain tumors. The overwhelming majority are IDH wild type, meaning they do not harbor an IDH mutation, and patients with IDH-wild-type glioblastomas have a worse prognosis than those with IDH mutation.

Glioblastoma increases with age at diagnosis. And we know that age relates to survival, with 5-year overall relative survival of only 6.8% for glioblastoma in general. Survival decreases with age.

With respect to molecular pathogenesis, we know that glioblastoma arises from neuroglial stem or progenitor cells and are characterized by molecular heterogeneity. Molecular profiling has identified genes in core pathways commonly mutated in sporadic glioblastoma. And these are characterized by somatic molecular defects in 3 major processes: initiating tumor growth, evading senescence, and enabling immortal growth.

Some of the pathologic hallmarks of glioblastoma include a diffusely infiltrative neoplasm with an astroglial appearance, angulated nuclei, irregular chromatin, microvascular proliferation, and pseudopalisading necrosis. There are several subtypes of glioblastoma, as demonstrated here. We will hear from Dr Moliterno about the updated WHO criteria, but at this point, if you see a patient where the tumor specimen does not show the classic histopathologic features of glioblastoma, but do contain the molecular signature of a glioblastoma, it should be treated as a glioblastoma.

The presentation varies according to the part of the brain that is involved in the patient. Most glioblastomas are diagnosed following symptomatic presentation, whether that be seizures, headaches, focal neurologic deficits, mental status changes, or sign of increased intracranial pressure. A contrast-enhanced magnetic resonance imaging (MRI) is the diagnostic tool of choice for glioblastoma. As you can see in the images, this typically manifests as an enhancing necrotic appearing mass, surrounded by nonenhancing signal abnormalities, which relate to edema and infiltrative tumor.

Now onto the guidelines based on the Society for Neuro-Oncology and the European Association of Neuro-Oncology (EANO). With respect to medical management and supportive care, we continue to use corticosteroids, preferably dexamethasone. This is used to reduce symptomatic vasogenic edema. It can help alleviate neurologic deficits and also can help alleviate signs of intracranial pressure. Because the side effects of dexamethasone worsen with increased dose and duration, we typically try to use the lowest doses that can achieve a benefit to the patient.

With respect to seizure management, SNO recently published updated guidelines to update the old American Academy of Neurology (AAN) practice parameter. Essentially, their recommendations are very similar. There is no clear benefit with continued use of antiepileptic drugs (AEDs) as prophylaxis in a patient who has never had a seizure. We do still recommend tapering those AEDs within 1 to 2 weeks after surgery. Of course, a patient who has had seizures needs to remain on their antiepileptic drugs.

With respect to venous thromboembolism, we know that there is a high risk in the glioblastoma population, particularly in the perioperative period, that even persists beyond that. The preferred anticoagulant is not well established. There is data to support low-molecular-weight heparin, which is traditionally what we have used. But there is emerging data, as well, to support the direct oral anticoagulants.

Now, to move on to some standard therapies for newly diagnosed glioblastoma. In general, we aim for maximal safe resection, and then we consider the patient's age and their functional status, in helping to guide our treatment regimens. If we look at patients aged 18 to 70 who have a good functional status, the standard treatments essentially are consideration of clinical trials. This is particularly important, since we know the limitations of the currently available standard-of-care options. There is also radiation with concurrent temozolomide, followed by 6 cycles of adjuvant temozolomide. That was established in a regimen published in The New England Journal of Medicine. And there are more recent data specifically for MGMT-methylated patients, that an alternative regimen is with 6 weeks of radiation in combination with lomustine and lomustine-temozolomide.

For patients who are older—and by older we will define as over the age of 65 to 70—or for patients who do not have a good functional status, we need to consider whether or not those patients would be able to tolerate multimodality therapy. If the answer is yes, then we would consider still, radiation with temozolomide. If the answer is no, and the patient might do better with a single modality, that will be determined, in part, by their MGMT status. If they are methylated, they are more likely to do better with temozolomide monotherapy. If they are MGMT unmethylated, they are more likely to do better with radiation. And of course, for patients who are not doing well, hospice or best supportive care is also an option.

If we look next at recurring glioblastoma, again, we can separate patients out into whether they have a good functional status or a poor functional status. Those who are not doing well should probably be referred to hospice or best supportive care. For those who are fit enough to receive additional therapy, you might consider whether they are a candidate for surgery or whether they are a candidate for radiation. In general, there is no clear standard of care. We often refer patients to clinical trials. We can consider, depending on the clinical situation, whether their MGMT is methylated or unmethylated; there are few different options.

To go into a little more detail about surgery, this should obviously be tailored to the individual patient. The goal is to achieve a maximal safe resection. If a gross total resection can be achieved, that is preferred. Obviously, we do not want to leave a permanent neurologic deficit, so there needs to be careful consideration about the extent of surgery. There is older data to support carmustine wafers, inserted at the time of surgery. It is a Food and Drug Administration (FDA)- and European Medicines Agency (EMA)-approved treatment, although most of the studies were performed before temozolomide.

Following surgery, as I mentioned, the standard of care is generally radiation with concurrent temozolomide, followed by 6 adjuvant cycles of temozolomide. Again, the MGMT promoter helps predict the responsiveness of the tumor to temozolomide, with those whose tumors are MGMT unmethylated being less likely to respond to treatment. Therefore, in a patient who has a poor functional status or has concerns about the ability to tolerate treatment, there might be discussion about withholding temozolomide. Tumor-treating fields can be added during the adjuvant phase of temozolomide. It is also important to note that within the first 3 months after receiving radiation with concurrent temozolomide, the MRI can look worse. This relates to what we call pseudoprogression, or treatment effects, and not necessarily true progression.

There have been a number of studies that have looked specifically at elderly patients. Depending on the study, elderly is defined over the age of 65 or 70 years. The median age of diagnosis of glioblastoma is actually 65 years. You will definitely encounter many patients which would fall into this elderly category. As we know, patients who are older tend to have a worse prognosis. They do not tend to tolerate treatment as well, and have more comorbidities.

For those patients who perhaps have a worse functional status, we might consider single-modality therapy. Again, what you might choose depends on the MGMT status. There have also been a number of studies that looked at hypofractionated radiation schedules. In a fit, robust, elderly patient, we tend to give hypofractionated radiation with concurrent temozolomide, followed by 6 cycles of adjuvant temozolomide.

With recurrent glioblastoma, there is no clear standard-of-care salvage therapy. There are a number of options, including tumor-treating fields . . . bevacizumab, which is approved in the US but not in the EU. This has been demonstrated to improve progression-free survival, and reduced corticosteroid use, but has not been demonstrated to improve overall survival. For patients who are MGMT methylated, you might consider a temozolomide rechallenge, especially if time has passed since their last course of temozolomide. Or you might consider a nitrosurea, like lomustine. Reirradiation is also an option. And for those patients who are not doing well, or do not have a clear option, referral to palliative care is important.

This year, we've also seen the World Health Organization update their classification of central nervous system (CNS) tumors, and we will be hearing from Jennifer about this.

Jennifer Moliterno Gunel, MD, FAANS: Thank you so much, Dr Lee, it is a real pleasure to be here to discuss the updated WHO classification of CNS tumors. I am going to have an emphasis on glial tumors given that is what we are primarily discussing today. The fifth edition of the WHO Classification of Tumors of the CNS has recently been published in 2021. This introduces major changes that advance the role of molecular diagnostics for CNS tumor classification. Everyone has been taking these molecular aspects into account, and it is nice to see that the WHO is now doing that.

This was based on a hybrid or mixed organization. Tumors are grouped according to genetic changes that enable a complete diagnosis. There are looser oncogenic associations and histological and histogenetic similarities, again, with the emphasis on molecular genetics. New types and subtypes have been defined by these molecular features. Some of the changes that had been made in the current classification system involve changes in CNS tumor taxonomy. This includes “type” being used instead of “entity,” and “subtype” being used instead of “variant.” In addition, with regards to nomenclature, this follows the recommendations of the 2019 guidelines meeting to make nomenclature more consistent and simple. Names have been simplified as much as possible, and so location, age, or genetic modifiers are only included if they have clinical utility.

Modifier terms, such as anaplastic, which we use in the care of patients, are not routinely included. Terms may reflect historical associations that have become embedded in common usage, and a great example of that is with medulloblastoma.

With regards to gene and protein nomenclature for CNS tumor classification, the sequence alteration relative to transcript reference sequence is reported using a “c.” prefix for the coding DNA sequence, followed by the nucleotide number and nucleotide change. The predicted protein sequence change then follows a “p.” prefix with the reference amino acid, the amino acid number, and the variant amino acid resulting from the mutation. Here, is a good example of BRAF.

This is an example on this slide of the key diagnostic genes, molecular pathways, and/or combinations in major primary CNS tumors. We see astrocytoma, IDH mutant with, of course, the known IDH genes and mutations, and then so on, including oligodendroglioma with the various IDH considerations, as well as the 1p/19q-codeleted glioblastoma, and so on.

With regards to seeing tumor grading, it has now moved closer to how grading is done for non-CNS neoplasms. More specifically, there are 2 aspects of CNS tumor grading that have changed. First, Arabic numerals are used instead of the Roman numerals, and neoplasms are graded within types rather than across different tumor types. With regards to grading within types, traditionally, CNS tumors had a grade assigned to each entity, and grades were applied across different entities. Grade was correlated to biological behavior, but there is a shift now to within tumor type grading extended to many categories. This is to provide more flexibility in using grade relative to the tumor type to emphasize biological similarities within tumor types, rather than approximate clinical behavior and to conform with WHO grading and non-CNS tumor types.

When we discuss clinical pathological grading, the WHO CNS5 has generally retained the ranges of grades used for tumor types and prior additions, combined histological and molecular grading. Again, there is a big emphasis on molecular incorporation into this classification. Molecular parameters have now been added as biomarkers of grading and for further estimating prognosis within multiple tumor types. In addition, a molecular parameter can sometimes add value to histological findings in assigning a grade.

For a not otherwise specified diagnosis, as well as a not elsewhere classified diagnosis, the new classification system allows for separation of standard, well-characterized WHO diagnoses and those diagnoses that result from either a lack of necessary diagnostic, such as molecular information or nondiagnostic, for instance, for a WHO diagnosis or negative results. The NOS suffix indicates that the diagnostic information, histological or molecular, necessary to assign a specific WHO diagnosis is not available or has not been performed. The NEC suffix, however, indicates that the necessary diagnostic testing has been successfully performed, but that the results do not readily allow for a WHO diagnosis. This can then lead to a more descriptive diagnosis, but it is nice to have these classifications in the updated system to really know what has been exhausted in terms of the diagnosis and the differential.

For gliomas, glioneuronal tumors, and neuronal tumors, they are now divided into 6 different families, as you can see in this slide. The first being adult-type diffuse gliomas, the second being pediatric-type diffuse low-grade gliomas, the third being pediatric type diffuse high-grade gliomas, the fourth circumscribed astrocytic gliomas, the fifth being glioneuronal and neuronal tumors, and the sixth being ependymomas. This is important, as it nicely separates adult and pediatric tumors, which we know can certainly behave much differently.

In addition, 14 newly recognized types of gliomas, glioneuronal tumors, and neuronal tumors have been added. Division of diffuse gliomas, as I mentioned, has been into adult type and pediatric type, and simplification of the classification of common, adult-type, and diffuse gliomas has also been a big part of the new classification system.

On this last slide, we can see a summary of the gliomas, glioneuronal tumors, and neuronal tumors. This shows the various names that these tumors are being classified as. It is a pleasure to review the current classification and update at the WHO system, and I will turn it back over to Dr Lee.

Dr Lee: Thank you, Jennifer. We are now starting to see more targeted therapies emerging for the treatment of brain tumors. Gavin, will you take us through some of these agents and where we are with them?

Gavin Dunn, MD, PhD: Thank you very much, Dr Lee. I am excited to talk about a couple of interesting topics today. The first one we will talk about is targeted molecular therapies in glioblastoma, where we are, and the strategies we are taking to develop new treatments that are more precise than conventional chemotherapeutic agents.

As you know, there are challenges for taking these types of approaches in brain tumors like glioblastoma. Specifically, unlike other tumor types where there are specific biological dependencies that are targetable, that has been more difficult to discern in tumors like glioblastomas. First, a lot of agents do not actually cross the blood-brain barrier, so that is an issue with any type of chemotherapy or targeted therapy. There is also a lack of easy or clear targets such that if you inhibit them, there are biological dependencies that the tumor requires to survive that therefore will be an effective treatment. That has been a biological puzzle that has been difficult to understand in these tumors. A lot of these signaling pathways, because of tumor heterogeneity, are redundant through overlapping, so it has been important to try to understand which treatments one needs to combine to try to target that redundancy.

Unlike other solid-tumor types, molecular targeted therapy in glioblastoma has been challenging. There is a lot of ongoing and exciting work in this area, specifically work targeting the DNA damage response and also targeting tumor metabolism.

With respect to the DNA damage response, these are classes of agents that are familiar for the treatment of glioblastoma, because a lot of the agents that we use are alkylating agents that induce DNA damage in these tumor types. One aspect to try to make this even more tailored and targeted is to understand the tumor-specific DNA repair vulnerabilities. What I mean by that is, in the case of DNA double-strand breaks, these are the main cytotoxic lesions that are induced by DNA-damaging agents. Single-strand breaks can also be recognized as important lesions for synthetic lethality. Synthetic lethality means that any particular treatment on its own as a monotherapy may not be effective, but in the setting of another lesion or vulnerability, is lethal to the tumor.

This is a diagram of a number of the different DNA repair pathways that converge on the cell cycle. With respect to inducing lesions that are double-strand breaks, you can see the BRCA1 pathway, BRCA2, ATM. The single-strand breaks on the right, you can see the DNA-dependent protein kinases signaling through ATR and ATM as well. Again, all converging on the cell cycle.

Now, there are several classes of agents that are designed to target several kinases that were just schematized in that cartoon. There are a number of poly(ADP-ribose) polymerase (PARP) inhibitors, for example, veliparib, olaparib, and BGB290. There are ATM kinase inhibitors, such as AZD1390. Inhibitors of the DNA-dependent protein kinase, or DNA-PK, include CC115. Also, agents targeting the Wee1 kinase are in clinical studies as well, such as an AZD1775.

Moving onto another biological area that can be targeted in glioblastoma, and one of those areas is tumor metabolism. Now, tumor metabolism, in many other tumors, is a very quickly developing field and there are many therapeutic strategies being taken. The same thing is true for brain tumors. Again, very similar to other cancers, tumor metabolism is an absolutely critical determinant of glioma progression and sustainment. What a lot of oncogenic mutations do, in addition to modifying the cell cycle, are also modulating metabolism to promote survival, proliferation, and resistance to treatment, among other things.

A lot of this actually goes back to what is called the Warburg effect, which is the classic recognized adaptation biochemically that glioblastomas and many other tumors take. Essentially, it is the changing of the metabolic state of the tumor to a road of glycolysis from a more conventional mitochondrial oxidative phosphorylation, really uncoupling the metabolic state and the metabolic requirements from ambient oxygen availability. Essentially, metabolic independence therefore creates an opportunity, therapeutically, to target tumors. In addition to oxidative metabolism, there is also cholesterol metabolism that may be another target. Interestingly, this may actually be related to epidermal growth factor receptor (EGFR)-driven biology. In these tumors, as you know, EGFR is amplified in 60% or more of glioblastoma, so this is a critical part and almost pathognomonic of glioblastomas.

I am not going to go into excruciating detail with respect to this, which is probably reminiscent of your biochemistry textbooks, but you can see there are many nodes in which targeting metabolism convergence from the tricarboxylic acid (TCA) cycle in the mitochondria, which is in the middle of this diagram, to epigenetic inhibition and nucleus. The complexity of this matrix just shows how important metabolism is for fundamental cell survival, and also all of the different areas that are potential treatment nodes, and also to merit further study from a therapeutic standpoint. This is a very exciting area and also a rapidly moving field.

In the last few slides, I want to just mention a couple of other targeted therapeutic efforts that are designed to treat tumors with specific molecularly identifiable lesions. The NTRK fusion is a really great example of that. This is the neurotrophic tyrosine receptor kinases, but there is a fusion. You can see on the right in the table under the column most frequently reported NTRK fusions, there are many different types of fusion partners with the NTRK1 gene, and they are found at differing percentages.

Glioblastoma is not a high percentage, but in nonbrainstem, high-grade glioma, which is the category right under glioblastoma, there is a range of 10% to even up to 40%, again, of different fusion partners, but all NTRK family members. You can see in the other tumor types, there is a range of different NTRK fusion frequencies. It is going to be very important even in glioblastoma where NTRK fusions are not a high percentage of alteration.

We envisioned the treatment down the road will be identifying many different types of targetable patient populations and designing specific therapies for that group. So that is the most important part. It is easy to test for these fusions by immunohistochemistry, fluorescence in situ hybridization (FISH), polymerase chain reaction (PCR), and also next-generation sequencing (NGS), which is particularly important to identify the partner of the NTRK fusion.

There are a number of inhibitors that have gained approval across cancers, just dependent on the presence of NTRK fusion, including entrectinib and larotrectinib. There are additional later-generation NTRK inhibitors that are also in progress, repotrectinib and also selitrectinib. There have been some very interesting responses to this, and this is a very interesting area.

Along the lines of identifying specific alterations and tumor subsets, another good example of this is the histone 3 K27M population. These histone mutations are frequent and actually define the diffuse midline glioma population, which as you know is a particularly aggressive tumor type and unfortunately, [has] a very poor prognosis. There are some very interesting efforts to target that H2K27M mutant population with histone deacetylase (HDAC) inhibitors, like vorinostat and panobinostat, and also ONC201, which actually works through a different mechanism. This is a drug that antagonizes the dopamine receptor D₂ and actually does show potency in the histone 3 K27M-mutated gliomas. Again, this is a particularly aggressive tumor type, and so these are exciting and ongoing areas of study.

We have touched on a number of targeted treatments in glioblastoma. There are a number of other targeted agents that have either been studied or are continuing to be studied, and you can see that in the table here. BRAF is mutated in other tumor types, so these alterations in these targets are not all glioblastoma specific. Certainly, we expect to learn something about the efficacy of these agents if they have worked in other tumor types. EGFR, again, as we talked about earlier, is very highly amplified, and actually when you see it amplified in glioblastoma, it is a tremendously high copy number. Exportin-1, fibroblast growth factor receptor (FGFR), and other targets are listed here.

Again, we are at the tip of the iceberg in defining the right targets and then developing drugs to these targets, and there are many ongoing studies, targeting MDM2, mechanistic (mammalian) target of rapamycin (mTOR) complexes, cyclin-dependent kinase (CDK) 4/6 (which are amplified in a subset of glioblastomas), other cyclin-dependent kinases, the histone 3 K27M alteration we talked about earlier. The interleukin 4 (IL-4) receptor is also an interesting target that is overexpressed in glioblastoma, and additional targets, like hypoxia-inducible factor 2 alpha (HIF2ɑ) and the proteasome as well.

Lastly, what do clinical trials targeting glioblastoma with diverse alterations look like? That is really where a number of interesting trial designs have been developed. If you are able to identify a number of targets but they are not a particularly high incidence, then it actually is prudent to develop trials where you can study a lot of different targeted therapies at once if you are able to identify them. These are 2 great examples here. In Heidelberg, at the National Center for Tumor Diseases, there's a Neuro Master Match study. Patients are molecularly matched to targeted therapies plus radiation therapy in the newly diagnosed setting without MGMT promoter methylation. And the National Cancer Institute in the MATCH trial is a very similar approach, where targets that are identified through next-generation sequencing can be targeted through a number of different number of different agents.

That gets at the idea of what platform trials are, which are clinical protocols that allow a number of different therapies to be evaluated at once, which is particularly beneficial and helpful for looking at targeted therapies, where you may identify a number of different targets across a defined a patient population. As you can see, these are very good examples of that. You can see in each row that there are a number of different targets that are being looked at across these 3 different trials, in really, both newly diagnosed and also recurrent disease. Thank you.

Dr Lee: Thank you, Gavin. What about immunotherapy? We are seeing this used across many different tumor types. What has been the experience in glioblastoma?

Dr Dunn: Thank you, Dr Lee. I am also excited to talk about the role of immunotherapy in CNS tumors. It is something we work on in the lab and have thought quite a bit about. This is an exciting area, as you know. Probably the biggest change in cancer, both in research and in care, in the last 5 to 10 years has been the successful integration of immunotherapies into many solid-tumor types. We have all been excited to see that extended into glioblastomas.

Unfortunately, there are no FDA-approved immunotherapies yet. We have seen negative phase 3 studies with nivolumab and also the lack of a signal with the pembrolizumab studies as well. This is not through lack of effort that we do not have FDA-approved compounds yet for immunotherapy in brain tumors. That is likely due for a number of reasons.

We consider glioblastoma to be what we would call an immunologically cold tumor, meaning that when you look at the tumor after it is resected, unlike other tumors like melanoma and other solid-tumor types, we do not see a lot of T cells or tumor-infiltrating lymphocytes actually present within the tumor, meaning there has not been a significant immune response mounted to that tumor, for any number of reasons. These tumors typically have what we call a low mutational burden, which in turn, leads to a lower burden of neoantigens that T cells might recognize during the endogenous immune response. Very clearly, there are many immunosuppressive mechanisms that this tumor employs to dampen or attenuate the immune response, like tumor growth factor beta (TGF-β), IL-10, and regulatory T cells as well. In addition to that, we are still learning how the immune response in the central nervous system works.

With respect to the state of the studies of immunotherapy to date, these are data from a randomized open-label phase 3 study comparing nivolumab versus bevacizumab in patients with the first recurrence of glioblastoma. This is the CheckMate-143 study in recurrent glioblastoma. The primary endpoint was overall survival, and as you can see in these data, that primary endpoint of overall survival was not met, meaning that the overall survival of the bevacizumab-treated patients was similar to patients treated with nivolumab. Similarly, that has been the same observation made in studies using another anti-programmed cell death protein 1 (anti-PD-1) agent, pembrolizumab, which was also not effective as a single agent and also with bevacizumab in recurrent glioblastoma. You can see here, these Kaplan-Meier plots with both progression-free survival and also overall survival, there is no difference between the control and the pembrolizumab arms here.

With this background, there is a clear impetus for new, different, or alternative immunotherapeutic strategies, and in the world of glioblastoma, there are 3 pretty significant efforts ongoing: CAR T cells, oncolytic viruses, and vaccines. We are going to go through each one of these and drill down a little bit more on what they are and where they are in glioblastoma treatment.

Chimeric antigen receptor (CAR) T cells have been spectacularly successful in hematologic malignancies. Essentially, you are using the recognition moiety and the extracellular domain of an antibody-like molecule with a T-cell functionality of the intracellular domain, and you transduce and express them, either in T cells or even NK cells, or natural killer cells. These are molecules that are engineered, that is where the chimeric aspect comes in. One of the very appealing aspects of CAR T cells is that they are essentially human leukocyte antigen (HLA) independent. T cells recognize peptides in the context of human leukocyte antigen. Humans have thousands of different types of HLA molecules, so it is very appealing to have something that can target a tumor cell in a way that is agnostic to the patient's HLA haplotype.

There are many strategies in this domain directed against multiple antigens, which is particularly important because glioblastoma is molecularly very heterogeneous. If you target a single antigen, being able to engineer epitopes spreading, which is the targeting of other different molecules in the tumor, this is almost a holy grail in any kind of immunotherapeutic approach. This is something that is very much being pursued in the CAR world.

Combining with other agents, both conventional, like radiation treatment, chemotherapy, and also checkpoint inhibitors. And delivery is really important. Do you deliver intravenously (IV)? Do you deliver in situ, directly into the tumor? Do you deliver intrathecally? You can see a summary of a number of ongoing trials targeting EGFRvIII and one study targeting also combining pembrolizumab. Another study where you are introducing intracerebrally, or into the brain, EGFRvIII CAR T cells. IL-13 receptor alpha-2 is another target being studied extensively, and several studies are using that with or without a checkpoint blockade, such as nivolumab, again, which is an anti-PD-1 agent, and also ipilimumab, which is an anti-cytotoxic T-lymphocyte-associated antigen 4 (anti-CTLA-4) agent. And then HER2 is another target that is also being explored. So this is a very, very exciting area.

Oncolytic viruses are also being explored in glioblastoma, and this is a field that has really been developing rapidly and almost exponentially in the last few years. Essentially, you are taking natural viral strains, and there are a number of different viral strains being used, or genetically engineered viruses that are designed to infect and replicate selectively in tumor cells. Usually, the way that these are deployed is that you are directly introducing them to the tumor and trying to achieve direct cytotoxic activity caused by the virus. What many people are learning is that this is really an oncoviral immunologic effect, in that this cytotoxicity then induces antigen presentation, then what we would potentially consider a second phase of first innate and then an integrated adaptive immune response caused by the antigens that are released. This is again, a collateral benefit of directly lysing a tumor cell. This is a tumor unleashing the tumor antigens to the immune system, to then generate a more potent secondary response.

There are many different viruses being explored. And again, this is a very exciting area. There is the Delta-24-RGD, which is an oncolytic adenovirus. There is a poliovirus, there are adenoviruses that are combined with IL-12, and also ways to regulate that IL-12, there are other adenoviruses that are combined with radiation treatment, temozolomide, and also retroviruses like the TOCA 511 approach. There were additional viruses still that either have been tried or are being explored or will be explored in this area. Again, but this really dovetails with activating new system as well.

Finally, vaccines. Vaccines, we consider a way to direct the immune system to clonally expand specific T cells that recognize particular features of glioblastoma cells. So here it is important to pick the right antigens with which to immunize patients. I want to be clear here, that these are typically not prophylactic vaccines in the way that you usually think about them. These are therapeutic vaccines. Here are 3 examples of those.

The ICT107 approach is taking dendritic cells and essentially loading them with some shared tumor antigens, like MAGE-1, HER2, and others, in the newly diagnosed setting. One vaccine trial that was completed was targeting EGFRvIII using a vaccine called rindopepimut, and that was targeted to a neoantigen that has created an EGFRvIII deletion, which is seen in about 15% to 20% of patients with newly diagnosed glioblastoma. And finally, DCVax, which is a dendritic cell vaccine, in which dendritic cells are matured in the peripheral blood of brain tumor patients and then exposed to the tumor lysate from that patient's own tumor. That combination is injected back into the patient. There are very interesting ongoing studies there. But I think what this table represents is that vaccines may take several different forms, but they are all designed to use inpatient T-cell clonal expansion specific to the patient's tumor. Thank you.

Dr Lee: Thank you, Gavin. It is really exciting to see all the progress that has been made within neuro-oncology. And we hope that this progress will continue and it will lead to new treatments for patients.

And again, I want to thank Jennifer and Gavin for this great discussion. And thank you for participating in this activity. Please continue on to answer the questions that follow, and complete the evaluation.

This is a verbatim transcript and has not been copyedited.