8 Functional Testing for Brain Mapping
8.1 Introduction
Many of the greatest challenges in brain mapping are epistemological. In other words, our great challenge is figuring out how to gather meaningful information, how to assess its quality, and what to do with that information. This chapter addresses the nature of our information about the brain and its function, especially with regards to functional testing, both before and inside the operating room, and when relevant provides some practical information on how we do the testing presently.
8.2 What Tasks Should You Test?
As a surgeon, you by definition won’t be doing the testing during surgery, but you need to understand the task, specifically what functional systems it is (probably) testing, what task to do when, and when to ignore information versus acting on it. Also if the program is not firmly established at your hospital, you will probably have to guide the early efforts.
A common (mis)conception people have about brain mapping is that you simply have the patient do some task, stimulate the brain, and if you find something, just stay away from that part of the brain. That paradigm works reasonably well for areas with simple anatomy, such as some aspects of speech and motor function. The problem is that finding a cortical site, doesn’t fully protect you if you then go on and cut that area’s connection in the sub cortex, or remove another area in that network necessary for the task (for example math skills are poor in a densely aphasic patient). You will also be disappointed at times, as most higher brain functions require multiple interconnected brain areas working together in order to produce the function, so the test will often not provide you the black or white result you expect, say with something like arm movement.
Another problem with just testing a part of the brain in an undirected fashion is that you may not find a site, especially if you are looking for it in the wrong part of the brain. In addition, it may not be something so critical that it’s worth leaving a great part of the tumor in the head to save the function. For example, it would be nice to name colors, but I am not certain, it is more important than treating the GBM.
Finally, not all functional tests translate well to intraoperative brain mapping. Some are too complicated or long. For example, a long figure construction based neuropsychological test would require a long period of brain stimulation before you know for sure whether that function was successfully performed or not, which of course can eat up valuable time (most patients will not do testing indefinitely) and which can increase the risk for seizures especially with longer periods of cortical stimulation. Other tests don’t have a clear yes-no answer (which you need when you need to make a call of whether or not to make a cut). For example, a test which evaluated as a 1–30 score and then compared to normative data is cumbersome and cannot be continuously monitored, which limits its utility. Finally, not all tests reliably become abnormal when you approach the structure of interest.
Interestingly, we have found that most tests can be successfully performed in the operating room with the patient in the correct position with access to both arms. Obviously, you cannot test gait or balance, but most cognitive tests are possible, including some complicated tasks of real life activities. But not all that can be done in surgery is useful.
I term the process of creating an intraoperative test from an office based neuropsychologic test “miniaturization.” Restating information from Chapter 4, below are a few guidelines I use for selecting, creating and integrating new tests into my mapping cases:
8.2.1 Sughrue’s Guidelines for Attempting to Find and Save a Higher Brain Function
For a function to be amenable to a focused effort to find it and save it, a higher brain function should:
Be worth saving such that it is better to reduce your extent of resection to achieve it.
Have a reasonably well understood functional network, with some idea of what structures are essential for performing that function and their essential interconnections.
Have a test of function which can be reasonably performed in the operating room by at least some patients.
Have a test which clearly becomes abnormal in a reliable and easily observable way when the system is disturbed.
These aren’t scientific laws, but they help us think about how to expand our mapping in a way that is logical and practically achievable.
8.3 The Testing Team
Brain mapping is often done by neuropsychologists, but I have found that physical and speech therapists are often better at these cases. First, they often have more flexible schedules than neuropsychologists, which is an important issue when you do 6 awake surgeries a week which is not uncommon for us, but is equally important as you are starting to do awake surgery, as nothing can kill interest in doing awake surgery then when logistics make the cases impossible. More importantly, therapists are great at encouraging and pushing the patient to keep going. The longer they map, the more you can do safely, so the skill of motivating patients is key to successful mapping. Plus a physical therapist has a broader experience in assessing the nature and quality of movement, which increases the sensitivity of your motor system testing. Finally, in an ideal world, they can also begin the rehab process in an integrated manner from the beginning. Every center is different ultimately and the best person to do these cases is one who is dedicated to these cases.
8.3.1 Preoperative Testing
In the preoperative evaluation, the patient meets the testing team and is familiarized with the process. Most importantly they are taught the tests, as the time to learn how to do them is not in a head holder as you are coming out of light sedation. The patient is educated about the possible problems which can be caused in the short term by brain swelling, and about the long term plan for rehabilitation.
Finally, the patient’s preoperative baseline is assessed. It is important to know what tests are not available for use due to deficits, but also what parts of certain tests should be taken out of the test for that patient. For example, a non-English speaking, non-American patient unable to name an American football team is not anomic, and shouldn’t be tested on this. Similarly, a patient with language problems shouldn’t be tested in the operating room on naming objects that they can’t name in an office.
Diffusion Tensor Imaging (DTI)
During the preoperative period, we obtain a DTI image and examine the tract anatomy which together with the tumor anatomy and the patient’s preoperative function, helps us make a plan for what we are going to do (see Chapter 9).
Because DTIs (and its relative diffusion spectrum imaging (DSI)) are such a dominant feature of this book, it’s worth mentioning a few details here about what these images are, as brief understanding of these modalities is helpful for understanding their limitations (I promise that I am not about to break into an engineering discussion). Both studies are MRI images which utilize a feature termed fractional anisotropy to characterize the bulk movement of water in the brain. Water moving in an organized direction is typically following a white matter tract, and this feature is exploited to describe tract anatomy.
DTI images quantify tract direction by determining the dominant eigenvector (i.e., the basic direction most of the water is moving) for each voxel. The limitation of this method is that it oversimplifies the data at crossing points by assigning one direction to the entire voxel even if there are more than one. There exists some software which compensate for this by dense data acquisition and liberally seeding the tract data so that the crossing point issue becomes less relevant (i.e., it accurately shows tracts on both sides of the crossing point) but it does limit the ability to trace a tract from origin to target.
DSI images are able to resolve tract crossings and are obviously the major image source of images for the anatomy sections of this book. The challenge with their use in routine clinical use is their long acquisition time, which can add over an hour to the end of preoperative volumetric imaging. There also are not seamless integration programs for linking this to image guidance programs presently (as there are for DTI). Also there are not studies validating their accuracy at predicting functional anatomy (there are with DTI). It is a great research tool, but we presently use DTIs for our cases.
The key point with both tests is that brain edema can wreck the regional tract quality with either test (purportedly worse for DSI, but I have never got a satisfying answer why that is), which is a major issue with high grade tumors especially. Thus, it is important to use tractography as a guide to the basic location of the tract, as opposed to a millimetric representation of the tract anatomy.
8.4 The Tests
8.4.1 Motor Testing
Testing the motor system is unique in that you can see positive and negative phenomenon relevant to the motor system. In fact, the presence of positive phenomenon is one of the greatest impediments to progress in intraoperative motor testing, in my opinion, as it is so obvious that it discouraged neurosurgeons from considering that there is anything else worth looking for besides the location of the motor cortex. I repeatedly emphasize in this book that a disconnected motor strip is synonymous with a paralyzed contralateral side.
If the motor system is in the surgical field (i.e., it is near a disconnecting cut I wish to make), I always test it first, before testing other systems. This is for two reasons. First, it is easy to find an unambiguously positive site early, and this allows us to confidently use as low a stimulation current as possible. Second, because of the role of the motor system as the final output of most other networks (for example, speech requires mouth and tongue motor function), it is important to know where the motor system is to subtract its contribution from other findings. One common example is that stimulation of the face motor cortex can make the patient unable to talk due to positive contraction of the face or mouth muscles, and knowing where this is ahead of time makes understanding the speech areas of the posterior frontal lobe easier (they usually are slightly anterior to this site).
The Basics
Positive motor mapping typically localizes to the motor cortex and manifests as involuntary face, arm, or leg movement. Smaller movements can be seen at lower currents, and in that case I turn up the current until it is obvious because I want to ensure I am disrupting sites later in the mapping. Even in patients with some invasion of the premotor or motor cortex causing hemiplegia, it is often possible to get muscles to fire that are palpable to a physical therapist, and I have restored function in some patients by preserving a compressed but not destroyed network, though obviously, this is inconsistently possible.
Negative motor mapping involves having the patient perform spontaneous contralateral movements, and stimulating to look for arrest of that movement (negative motor sites). The simplest form is simultaneous arm-leg movement, but it can be a more complex movement in higher functioning patients like playing an instrument. Some have stated that negative motor sites are expendable, but in many cases I have found them to be coterminous with the SMA or premotor areas, and while these might remap, I try to save them to avoid motor planning problems like SMA syndrome. SMA syndrome does not always improve despite the literature claim that it does, and at minimum, a hemiplegic mute patient with a GBM is probably not getting to radiotherapy as promptly as one who does not have SMA syndrome. I feel the small improvement of extent of resection by taking out the SMA is not worth the problems, and that this syndrome is usually avoidable.
A More Nuanced View
Finally, it is important to note that quality of movement is an important thing to consider in mapping the motor system as discoordinated movement it is often a sign of mild dysfunction (a sign you are getting too close), or compromise to important proprioceptive or other related planning inputs to the motor system needed to maintain meaningful use of the limb. Coordination requires various elements or muscles of the body working together to effectively complete an action, smoothly and efficiently. This likely means the networks responsible for inhibiting and the network responsible for actively contracting muscle are not acting appropriately together, therefore presenting a decrease in control. Coordination requires a balance of activity of all involved muscles.
In our motor testing paradigm, we take into the consideration the following aspects of a movement being performed, in addition to whether the arm is or is not moving.
Concentric control. This refers to control during shortening of a muscle. A loss of concentric control is manifested as a weakness or slowing of the movement not related to attention of neglect issues.
Eccentric control. This refers to control with lengthening of a muscle, measures some form of inhibitory control of a muscle. In surgery, problems with eccentric control manifest as the patient dropping the limb or hand to the initial position as opposed to lowering it in a controlled fashion.
Kinesthetic awareness. This refers to awareness of a limb’s position as it moves in space. In surgery, this presents as the patient taking a random appearing and awkward path to the target.
Coordination. The presence of dyskinesia, or uncoordinated movements.
Attention to task. This can be affected by problems with one of any number of networks.
Speed of movement. Movements can become slower or less coordinated as a system begins to be affected by your presence near by the relevant network.
It is not entirely clear to me at present how any of these specific findings line up with a specific pathway or specific network. We have found problems with many of these features of movement well away from the motor strip, descending motor fibers, or the sensory strip, highlighting the extent of the cerebrum which contributes to normal motor function. It is not also clear how careful we need to be just because of an event like loss of eccentric control or lack of kinesthetic awareness, as not all patients who we continue pushing the resection after such an event lose useful limb control, though some clearly do, and likely we would find more subtle problems with limb control with more sensitive and rigorous measures.
Again the beauty of awake versus asleep motor testing is that you can meaningfully assess the end product, as opposed to just knowing that you are not cutting into the motor strip. It is critical, however, for the therapist doing the testing to push the patient to ensure these mild problems are not due to a lack of effort. Also, this is not to say that I always back off at the first sign of an issue, but I always consider this in the overall plan for the patient as I remove the tumor. In other words, if it keeps happening every time I touch a certain area, I consider the possibility that leaving that part of the tumor might be a better way to preserve function. That judgment is dependent on tumor grade, our available adjuvant treatment options, preoperative function levels, and how much tumor will be left. I do not think leaving a large amount of enhancing tumor in a motor area of a recurrent GBM case is a good strategy to avoid mild incoordination (the tumor will often go on to finish the entire motor system if we can’t control it, but leaving a small amount of low grade tumor might be a good trade).