Goals of Resection in Glioblastoma

Transcript:Steven A. Toms, MD: When a patient sits down with me to talk about surgery, there are three major options. You can do nothing at all if the patient is very elderly and not in good shape. These days, you can make the diagnosis radiographically. Some very elderly patients in poor prognosis decide to just go hospice once that radiographic tentative diagnosis is made. But, most commonly, we look at deciding whether we’re going to do a needle biopsy versus a resection. If we do a resection, or removal of that tumor, our goal is always to get as much as we can out safely without leaving the patient with a neurologic deficit.

Biopsy tends to be reserved for patients with deep-seated tumors, or tumors in which radiographically, you don’t have a very good idea of what’s going on. I used to do a lot more biopsies earlier in my career, but with advances in MRI and radiography, we can probably tell 90% to 95% of the time if a patient has a glioblastoma, versus a metastasis, versus a lymphoma—other brain tumors that can commonly mimic glioblastoma in the neighborhood of three-quarters to four-fifths of the time. So, it’s those rare cases where there’s some diagnostic dilemma or a deep-seated tumor, that usually gets the biopsies first.

If a biopsy is done, and it does prove that it’s a glioblastoma, and is in an area that’s resectable, or we have a pretty good idea going into this that we’re dealing with a glioblastoma just from the MRI, then we focus on the discussion of how much of the tumor we can safely get out.

These days, we use a lot of adjuncts in surgery to help get tumors out safely—most commonly using what’s called stereotactic navigation, which is a way in which we use an MRI. Often we use fiducials on the patient’s head, which are little markers to tell us where parts of the head are so that we can register, or make sort of a GPS map, of the patient’s head, that we use for surgery. And, we use a little light pen during surgery to tell us where the edges of the tumor are. So, we use that in combination with intraoperative ultrasound or other intraoperative imaging. Some places use intraoperative MRIs, although we use ultrasound most of the time at Geisinger.

And then the last adjunct to help us out with our surgical resection is one that I rely upon fairly heavily called neurophysiology—electrical stimulation of the brain itself or pathways in the brain to try to make sure that we are doing a safe resection during surgery.

In the old days, I used to do many of these operations awake. Now, I can do just about any surgery near motor activity of the brain asleep simply by mapping out the surface of the brain electrically using something called motor-evoked potentials, as well as stimulating the white matter underneath the cortex of the brain with a stimulator during surgery to try to stimulate those white matter pathways to tell me when I’m getting close to the descending motor tracts that help control the movement of the arms and legs.

I used to have to do this awake and just monitor the patient neurologically. But now, we’ve been able to do this very consistently over the last 10 to 15 years. So, patients early in my career who had tumors, in or near motor areas I would do awake, I can now do asleep quite effectively.

I still do awake operations a few dozen times a year—mostly for tumors in or near speech areas, which we still find difficult to map out while a patient is asleep. We can use radiologic adjuncts such as functional MRIs and diffusion tensor imaging, which maps the little white matter pathways in the brain, to give us an idea of where these speech pathways are. But we have no way to monitor them in real time and during the operation. During the operation, when you’re working on the brain, the brain’s about the consistency of a bowl of Jell-O. And if you think of the tumor as being a piece of fruit embedded in that bowl of Jell-O, once you make the incision or cut into that Jell-O, the Jell-O shifts a little bit. So, the mapping, or GPS system that you made may not be entirely accurate.

When I have a patient with a tumor near speech areas, especially near expressive speech areas, I like to have that patient awake. We stimulate the surface of the brain to find out exactly where their speech areas are, how they relate to the map we have of the tumor, and then while we’re taking the tumor out, continually monitor the patient with a neurologist by talking to the patient under the drapes—doing speech testing and language testing to make sure I’m not doing anything that’s interrupting with that patient’s speech.

When we do that we have a very low morbidity. We often cause some temporary worsening with the speech in the neighborhood of 15% to 20% of the time, but only about 1% or 2% of the time we have an injury that leaves the patient with a permanent deficit. As we mentioned earlier, avoiding a permanent deficit is the most important part of the brain tumor surgery, aside from getting as much of that tumor out as possible because if we hurt the patient, they’re not going to do as well in the long run.

Now, the same discussion occurs with the patients whether they’re 10 years old or 90 years old. Obviously we’ve got a different ability to recover with an older patient versus a younger patient; we may have differences in their overall survival. But, it’s that same balance of the risk of surgery and trying to get as much as we can versus the patient’s recovery ability, and what their goals and desires are in life to figure out the best path and the most accurate, complete resection we can for each patient before we move on to the next stages of therapy—which of course include radiation and drug therapy.

One of my partners, Dr. Nick Marco, did a beautiful paper in the Journal of Clinical Oncology a couple of years ago that showed, mathematically, how percent of resection varies with outcome. Essentially, what that model showed is that there’s not a single cut-off like we used to think 10 or 20 years ago—that you had to get 98% or 85% of the tumor out and the patient was magically transported to a better prognostic group. In fact, there’s a curve that’s almost a logarithmic-shaped curve: the greater the percent resection, the greater the prognosis. And that varies both with age and Karnofsky performance level, as well as percent resection. So a young patient in a good Karnofsky performance group does even better with a greater percent resection than an elderly patient who’s not doing as well. What this leads to, is a discussion with the patient, focusing on quality of life and percent resection, before the surgery begins.

If we do anything that causes a neurologic deficit during surgery, the patient’s overall survival, and their quality of life bottoms out, and they do not do very well overall.

So our goal, surgically, is to do the maximum safe resection to get as much of that tumor out as possible. It’s a goal of my team to almost always get an imaging complete resection—or a gross total resection it’s sometimes called—where we can’t see any visible tumor left on the MRI. And that does put the patient in a better prognostic category than have the tumor out or a biopsy alone.

Transcript Edited for Clarity