15 Unique Issues with Recurrent Gliomas



10.1055/b-0039-172175

15 Unique Issues with Recurrent Gliomas



15.1 Introduction


This is a relatively short chapter, as there is nothing specific to the anatomy of recurrent gliomas which make the functional anatomy different than first-time surgery. But they are harder, without a doubt, and these challenges and some advice for tackling them are provided as are some examples.



15.2 Issues Unique to Recurrent Gliomas




  1. Wound issues are a much bigger deal: This is especially true with a history of Avastin, but previous radiotherapy can also make this worse, and a poorly healing glioma wound can spell more of a disaster than the tumor was.



  2. There was a previous surgeon: If it was someone else, they may have placed an incision in a way I would never want to be forced to reuse. If it was me, there is a good chance the tumor recurred on the margin of the resection cavity, which in my case was probably eloquent (I usually leave it for a reason).



  3. The arteries may have been radiated: Radiated arteries are more prone to injury, even if you are careful.



  4. The anatomy is abnormal: If the previous resection was aggressive, and involved a complete or partial lobectomy, the remaining normal anatomy probably has shifted into the cavity. For example, I expect to find the IFOF and possibly the MCA near the floor of the middle fossa if I have done an aggressive lobectomy (Fig. 15.1).



  5. The dura is stuck to the brain: Even if you are careful, you can injure eloquent cortices opening the dura and lose the chance to map before the case has even got started. Less obviously, if the brain is stuck to the dura, every move may transmit force to the dura and cause pain making the patient less cooperative. If I did the past surgery, it is likely that the leading edge next to the cavity was eloquent.



  6. The patients are often less functional: They are often less able to map due to tumor recurrence and the long term effects of radiotherapy.



  7. It is often less clear what to take out: At the first surgery, anything enhancing is probably tumor as is more of the T2 hyperintense brain. All bets are off at a repeat surgery: enhancement can be radiation necrosis or pseudoprogression and T2 change could be tumor or radiation effect, or encephalomalacia. Often it is less certain that we should be operating, let alone doing something aggressive (Fig. 15.2).



  8. The mandate is less clear: The data are more variable for repeat surgery, especially for GBM.



  9. There is less brain than before and it’s more likely that what’s left is eloquent: This is more specific to those who do aggressive resections the first time.


One thing to note is that all of these rules primarily apply to true recurrence surgeries and less to cases where an inadequate first surgery is being followed with a more complete surgery, as these surgeries are more like first time cases.

Fig. 15.1 (a, b) These images demonstrate cases of recurrent gliomas after temporal lobectomy. Note that while the brain roughly reaches the temporal floor, this is the insula, and not the temporal lobe touching the middle fossa floor. This is critical to note in planning these cases.
Fig. 15.2 Image depicting one of the almost infinite appearances of recurrent glioma. This image shows extensive T2 changes, and it is entirely unclear what is tumor, what radiation change is, and what is edema. This highlights the challenges of decision making in these cases, as one wants to be aggressive if you are going back in; however, it is unclear what to be aggressive with exactly with our limited imaging.


15.3 Strategies for Addressing these Problems




  1. Make the surgery count: The only thing worse than operating in a radiated repeat wound for a small recurrence, is missing part of that recurrence because you were not thorough. This seems like an obvious statement, but it’s quite easy to do when the anatomy is abnormal and the tumor is small.



  2. Expose the minimal amount of brain possible: The less surface area of adherent dura you need to peel off of potentially eloquent brain to achieve your goal of surgery the better. Often I will plan surgery for recurrence in deeper areas of a resection cavity to only open a small area over the cavity so that no cortex is put at risk and will start subcortical mapping immediately after opening the dura (Fig. 15.3 , Fig. 15.4). There is no rule that one has to expose cortex in a re-do craniotomy.



  3. Make functional vs oncologic trade-offs rationally: I call this the “bang-for-your-buck” rule. If removing a tumor, or part of a tumor involves a great deal of risk to a functional system to remove 1 cc of enhancement, then this is not a good trade off compared to chemotherapy alone. Alternately, removing a large amount of tumor in a nonfunctioning system or a disconnected part of the brain is a good trade off (Fig. 15.5). Less intuitive is the idea that the system may be working, but unlikely to be salvaged for long. For example, a patient with a mostly paralyzed leg and a big tumor in the motor strip, is probably going to lose the leg motor function completely in the near future, and the decision to leave this tumor because “I don’t want to hurt his leg function,” is irrational as well-performed surgery can only improve on that natural history (Fig. 15.6).



  4. Stay out of the old wound, if possible: The best way to avoid radiated wound problems is to not reopen it if you can avoid doing so. Often, a different cut or angle can achieve a good resection safely and stay out of the wound. Using large resection cavities can also help achieve this goal. Another trick is to use an eyebrow approach to address recurrences in the inferior frontal lobe from previously operated temporoinsular tumors (Fig. 15.7). You can also decide not to open the full wound in some cases.



  5. Get a running start: Sometimes a small recurrence can look deceptively easy. It looks like a lump sitting at the bottom of a resection cavity which should just pluck out. It is worth noting that to pluck it out, you generally need to cut into some brain and separate this lump from the brain. Just trying to shave off a few layers and stop will often leave some tumor in the brain, and is largely inconsistent with our understanding of gliomas. Even small tumor recurrences should be removed with a J-shaped or C-shaped cut using anatomic and/or functional boundaries to define resection goals (Fig. 15.8).



  6. Develop a realistic goal: Some recurrent tumors are bad, and the patient is going to lose function no matter what you do. Other times you have to try to get the patient into the best situation you can and rely on adjuvant therapy to do the rest. On the other hand, you must always remember that you are treating a tumor which is highly capable of killing this patient in rapid and terrible fashion, and that you are not going to help the situation by quitting.

    Fig. 15.3 This image demonstrates an incision planned for a recurrent glioblastoma. The previous surgery was for a larger tumor, and he presents with a small recurrence. This highlights the ability of minimally invasive techniques to reduce the threshold for going back, and reducing the need to expose large amounts of cortex scarred to dura.
    Fig. 15.4 These two cases demonstrate the technique of starting a recurrent glioma in the previous resection cavity: Case 1: This is a recurrent hippocampal GBM. I did the previous case awake as evidenced by the typical cuts. (a)The tumor has recurred in the medial temporal structures, and is following the PHG and hippocampus into the fornix and into the posterior cingulate gyrus. All of these are deep structures and do not need to be exposed. The posterior cut of the lateral temporal lobe from the previous operation likely contains speech (I did the operation and I stopped my cut there for that reason) and now this is the anterior most cortex under the dura in this case. The goal here is to use subcortical mapping to try to get a lateral plane from anterior to posterior to run lateral to the enhancing area of the tumor, and to sweep this tumor out of the brain safely. I do not want to peel dura off the speech cortices as this will be stuck and may injure them. Further, I do not need to make an aggressive lateral temporal cut as this is unnecessary in this case. (b) The craniotomy is small and uses only enough of the incision to get into the cavity and in the cavity find an angle lateral to the tumor to disconnect the lateral border of the tumor from the lateral networks. Once this is clear, then we will put the patient to sleep and resect the medial structures. (c) The resection is excellent, with removal of tumor in the posterior cingulate and fornix. The speech is normal in this patient. Case 2: These films demonstrate a case of glioblastoma which has recurred in several sites in the depths of the resection cavity. Given the lack of a need to perform cortical mapping, we limited our opening to just over the cavity, and began working in the depths. This eliminated the need to peel dura off brain that we planned on leaving. (d)The tumor recurrence sites are on the deep portions of the cavity, including the basal ganglia, and the contralateral cingulate. There is no role for touching the superficial cortices to make a starting “J-cut”, so it’s best to leave them alone. We will focus our efforts to slowly teasing tumor out of the speech networks and cingulate gyrus with subcortical mapping.
    (e) The previous incision was partially reopened and we started straight to subcortical work in this case after orienting ourselves to the cavity. (f)These DTIs are fascinating and informative. Review of the IFOF on the sagittal images, where the frontal lobe has been removed, shows the IFOF terminates in the caudate as the frontal rami have been cut. The patient had mostly normal speech. When working near this portion of the tumor, we had to proceed slowly as working or stimulating in these areas repeatedly caused anomias. Eventually, we got this small piece of caudate tumor out; however, this case is the best evidence I have personally seen that the target of semantic speech networks running in the IFOF is the basal ganglia and not the frontal lobe, as the IFOF in this case localized closely to language and was cut everywhere else in this case. (g) I have seen the situation here once before. We were able to remove the cingulate tumor in this case on the other side, despite the previous removal of the cingulate. It raises the possibility that the networks can reorganize to other areas and even both cingulate gyri might be able to remove in some cases, if they are done with a break for reorganization between resections. I would not recommend this as a general matter of course, or asleep, but in this case, it was tolerated well.
    Fig. 15.5 This image demonstrates a case of a temporal glioma which has recurred in the high risk deep white matter of the TPO junction. This is a case of “low bang for your buck,” by which I mean a case where great risks are necessary to get only a slight cytoreductive benefit. These are cases where I tend to seek out alternative options, such as LITT, radiosurgery, or chemotherapy alone, as opposed to going back. In contrast, a larger tumor, or one in a safer area would give you a better return on your efforts.
    Fig. 15.6 This image depicts a GBM in the motor cortex in a patient who has become hemiplegic from this tumor. While surgery is unlikely to save the motor function, cytoreduction can be helpful in saving other functions, such as the SLF, and we should not be deterred simply because it is in a bad area, and our goals are limited.
    Fig. 15.7 This image depicts a patient with an extensive frontal GBM, which had been treated with two rounds of radiotherapy and Avastin. We performed a frontal lobe resection through the eyebrow to avoid this high risk wound.
    Fig. 15.8 This schematic shows the J-shaped cut helpful for small glioma recurrences. The cut begins by entering the edge of brain just behind the tumor and starts a new plane. It continues behind and eventually deep to it (thus the J-shape) landing in the resection cavity. Obviously, this cut is dictated by the mapping.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

May 9, 2020 | Posted by in NEUROLOGY | Comments Off on 15 Unique Issues with Recurrent Gliomas

Full access? Get Clinical Tree

Get Clinical Tree app for offline access