Multidisciplinary Care in Glioblastoma - Episode 2
Transcript:Steven A. Toms, MD: To make the diagnosis of glioblastoma, you need to get some tissue. As a surgeon, I’m involved in that all the time, either doing a biopsy or a surgical resection of the tissues. Once that tissue is taken out, it’s still diagnosed by traditional pathologic methods. A piece of that tissue is taken out, frozen or put in paraffin, and then looked at under a microscope along with some stains.
Traditionally, when you look at a glioblastoma under the microscope, there are several things that stand out. You see several different cell types in there; you see vascular proliferation—which is what bevacizumab works against. You tend to see necrosis—areas where the tumor has grown so rapidly that it’s outstripped its blood supply. Often in these areas of necrosis, which are areas where the tumor has outstripped its blood supply and the center of the tumor has died off, you see the tumor cells lining up two, three, four cells deep all around that. That’s called pseudopalisading, and it’s one of the histological hallmarks of glioblastoma. When I’m operating on a glioblastoma, the thing that tells me that I’m probably working with a glioblastoma is finding the areas of necrosis—dead, almost liquidy tissue in the center of the tumor. And very often, there are thrombose blood vessels within that I really don’t see in any other tumors that I take out.
Susan C. Pannullo, MD: Glioblastomas have as their hallmark an infiltrative nature. It’s very difficult to discern a glioblastoma cell from normal brain tissue that can be scattered widely throughout the brain—not just at the site where they occur. They are rapidly growing. In fact, a patient with a glioblastoma can have progression of their disease—doubling of their disease—within the course of 10 days, which is unusual even for the most virulent kinds of cancer. One of the challenges of that type of behavior is first getting the patient to therapy relatively quickly, which usually starts with an operation. And then moving on to other treatments for the patient, which need to encompass a relatively wide penumbra around where the tumor is in order to encompass errant cells that are occurring outside of the area, for example, where the initial surgery has occurred.
One of the characteristics of glioblastoma is that they’re able to harness abnormal blood vessels, and they use these blood vessels—it’s believed—to propagate and proliferate. The abnormal blood vessels are believed to bring in nutrients that the cells need to survive and to grow. This property is one of the aspects of glioblastomas that is thought to present a therapeutic opportunity in terms of limiting the tumor’s access to the abnormal vessels needed to help it thrive.
Steven A. Toms, MD: Over the past decade or two, we’ve started to try and develop molecular subclassifications of these tumors. Initially, back in 2006, they were classified into three sets called proneural, mesenchymal, and proliferative types. This led to different prognostic factors. Some groups that I’ve been with, and many other groups, have tried to pull out different molecular markers that seem to be prognostic for this.
However, the most important prognostic categories that we’ve seen with glioblastoma include the MGMT methylation—which predicts whether or not they will respond to the alkylating agent temozolomide. Patients with methylation tend to do better prognostically, and they tend to be those who rose from lower-grade gliomas. But the single most important prognostic markers of the last few years have been the IDHs—the inosine dehydrogenase molecules—which are involved in the Krebs cycle. And what these seem to predict is going back to a very old classification of glioblastomas.
Before people started doing molecular classifications, they were split into two major groups, which were the de novo glioblastoma, which usually occurred in elderly patients and was often associated with epidermal growth factor receptor modifications, and the progressive glioblastoma—also called secondary glioblastoma. And these older classifications have been with us for about 20 or 25 years. The progressive glioblastomas were classically shown to have retinoblastoma gene mutations as well as p53 mutations. But that group is marked these days by the IDH1 mutations and those of the secondary or progressive glioblastomas with the IDH1 mutations. They tend to have the best overall prognosis when you’re looking at a single gene.
There are many, many different ways to subclassify the genes, the messenger RNA, the proteins of these—to look for different prognostic factors. And those have been done by many, many different groups many times. But we’re still not quite to the point where we’re ready to talk about personalized medicine or selecting a therapy off-the-shelf that attacks a particular molecular marker on a particular glioblastoma yet. Hopefully we’re going to get there in the next few years, and we are working on it.
The other important thing to remember about glioblastoma and cancers in general is that glioblastoma was originally called glioblastoma multiforme, or GBM, because there are many different parts of it and there are many different molecular signatures in different parts of the tumor. In fact, if you look at the tumor itself, it’s made up of cancer stem cells, which help to repopulate the cancer: cancer stromal cells that are a bit more differentiated, endothelial cells and other vascular cells that are brought in, supporting glia from around the brain that work with cytokine networks to perpetuate these, as well as the body’s own immune modulatory cells that are coming in to try to attack the tumor. They sometimes get taken over by some of the cytokine networks or inter-talk between the tumor cells within that tumor organ that help perpetuate the tumor.
There have been a lot of things going on inside that tumor, and when we look at a lot of the data out there right now, most of the data, initially, were taken on these molecular markers from tumor samples that were homogenized and contained a mixture of all these different cells. And what we’re seeing investigators do these days is come in, take tumors out, and then microdissect out the different cell types to look for gene signatures that are specific to each cell type. And one of our investigators here is working on a similar project today.
Transcript Edited for Clarity