22.1 Introduction

The most important surgical factor in the prognosis of gliomas is the extent of resection. In the case of high-grade gliomas, gross total removal may improve overall prognosis (from a median survival of 12 to 18 months). ^{2 ,​ 3} In low-grade gliomas, curative resection is possible. ⁴ However, this can be jeopardized when the tumor is close to or even contains functional brain areas. Most gliomas infiltrate the surrounding brain, making it difficult to decide where the tumor ends and where functional brain tissue starts. In the case of gut tumors, for example, the surgeon can remove the tumor and adjoining segments until pathology has confirmed negative margins in what is referred to as extended resection. This is not the case in brain tumors because extended resection could lead to removal of key functional structures such as those dedicated to the so-called eloquent functions (motor or speech functions). When this is the case, the surgeon, according to the patient’s preferences, must decide where to stop the tumor resection in order to preserve function, thus impacting disease prognosis. Accordingly, tenets of glioma surgery are to achieve maximal resection to improve survival, while respecting functional areas to maintain quality of life.

But what would happen if we could be able to move these eloquent functions away from the tumor, allowing increased extent of resection by letting other cortical areas, distant from the original ones, take charge of these key functions? In principle, this would be possible because of the plastic nature of cerebral connections. However, this property needs to be harnessed and induced artificially in order to apply it to the needs of our particular patients.

22.2 The Concept of Brain Plasticity

Plasticity is a property of the brain that allows it to adapt continuously to the environment. It can occur by several mechanisms, from changing the strength with which neurons communicate to the possibility of increasing the size and number of dendritic branches in order to make new connections between neurons. The concept of brain plasticity was initially conceptualized as early as 1904 by Santiago Ramón y Cajal ⁵ and since then it has been identified under healthy and pathological conditions, both in animals and humans.

One of the properties of plasticity is topographic plasticity. This means that the location of functions in the brain cortex is not fixed and can vary due to different mechanisms. It that sense, when Paul Broca described the area bearing his name at the left inferior frontal cortex, ⁶ it was widely accepted that the expression of speech was always located at this site. The same occurred with Wernicke’s area at the left upper superior temporal cortex, the motor areas on the primary motor cortex, or the sensory areas along the sensory motor cortex as Penfield and others have described. ⁷ Later on, researchers including Michael Merzenich have demonstrated that certain areas in the brain are able to be displaced or to increase its size in response to several challenges from the environment. For example, this group showed that the size of the area responding to a particular frequency in the auditory cortex of the owl monkey may suffer modifications depending on the stimuli received. ⁸ In humans, Álvaro Pascual-Leone’s group, for example, showed an effect of intermodal plasticity in normal volunteers that, after being blinded for 6 days, processed auditory information in the occipital lobe. ⁹

These plasticity mechanisms are not immediate, but require long-term training. In a study evaluating London taxi drivers, Maguire et al showed that there were changes in the size of the drivers’ hippocampus after years of daily navigation training. ¹⁰ Accordingly, patients who undergo a sudden loss, such as in stroke or trauma, are not able to spontaneously recover the lost function located within the affected areas. However, when the damage occurs steadily during a long period, plasticity mechanisms may compensate for the damage. It is known that slowly growing brain tumors (such as low-grade ones) may induce changes in the cortical representation of the functions. Robles et al observed that in patients operated on for low-grade gliomas in whom a complete resection of the tumor was not possible due to the vicinity of eloquent areas, the location of the affected functions had been displaced when evaluated during a second operation between 4 and 5 years later. ⁴

The rationale of our studies is based on the idea of artificially accelerating this natural plasticity process of the brain tissue in areas containing eloquent functions at danger due to tumor presence and infiltration. According to the available literature, this could be done by a progressive inhibition of the key areas, which provokes an artificial dysfunction susceptible of being recovered through personalized training. That is what we called “prehabilitation.”

22.3 First Case: Functional Inhibition through Repetitive Transcranial Magnetic Stimulation

Our first attempt was done trying to use a noninvasive method, repetitive transcranial magnetic stimulation (rTMS), to produce inhibition on an eloquent area located at the tumor region. The first case was a 59-year-old woman who presented with a left inferior frontal gyrus tumor. She was operated awake and the resection was limited because of the presence of speech functional areas (the pathological diagnosis was anaplastic oligodendroglioma). Experimentally, we tried to provoke a virtual lesion using rTMS in “theta burst” (which was supposed to produce an inhibition in the area which lasted for about 20 minutes), at the frontal end where Broca’s area is supposed to be located. ¹¹ The experiments included 12 sessions followed by intense speech rehabilitation. A functional language evaluation was performed immediately before TMS, immediately after TMS, and 20 minutes after intense speech. The assessment was conducted using the Boston test for aphasia (BDAE) that provides scores of repetition, nomination, auditory language comprehension, oral expression, and writing. ¹² While results indicated expected absence of effect of rTMS on comprehension and identification, it appeared that rTMS stimulation had gradually less impairing effect on nomination, indicating that the procedure changed the characteristics of production of speech. ¹³ However, the lack of changes in topographic location of the function did not allow us to increase the resection of the tumor, reducing the chances of function prevention during surgery. We hypothesized that the lack of topographic changes during that procedure was because the rehabilitation was done at different times as the inhibition, so we were reinforcing the function at the very same location. A new protocol needed to be designed.

22.4 Second Case: Functional Inhibition through Continuous Stimulation

The previous experience, where the rehabilitation had strengthened the patient’s function, led us to hypothesize that a continuous inhibition of the functionally affected area was necessary, together with an intense and personalized rehabilitation program for promoting compensatory mechanisms in other brain areas, and, thus, reallocating the affected function. We proposed to use continuous electrical current delivered through a subdural electrode grid implanted over the cortex. That solution would solve the time stimulation limitation of TMS, which cannot be delivered 24/7 to the patient. Following that rationale, a 27-year-old male with an anaplastic astrocytoma in the inferior left temporal gyrus was operated for a second time. During the first surgery, 5 years earlier, no resection was conducted due to the location and finding of speech functions within the tumor. Since then, the patient underwent radiotherapy and chemotherapy, but temozolomide produced aplastic anemia. Then, chemotherapy was interrupted and the patient developed impairments in speech production. The surgical team then decided to reoperate on the patient. In that occasion, a complete neuropsychological assessment was conducted before and after the surgery, including Mini-Mental State examination (MMSE), ¹⁴ letter-cancellation task, ¹⁵ verbal fluency, ¹⁶ Token Test, ¹⁷ and Boston Naming Test (BNT) ¹⁸ that demonstrated impairments in attention, comprehension, and fluency.

Functional magnetic resonance imaging (fMRI) evaluating fluency and comprehension activation patterns confirmed that those functions were at risk since they were located within the tumor. The patient underwent an awake surgery with cortical stimulation and electrophysiological and speech function monitoring, using the standard criteria of resecting exclusively noneloquent areas (Fig. 22‑3a), followed by the implantation of a grid for electrocorticography over the unresected tumor and surrounding cortex (Fig. 22‑3b). One week later, a mapping recording was performed using stimulation through the grid to test whether there were still active functions linked to the affected areas. Results indicated that most of the tumor still contained speech areas (Fig. 22‑3c). Consequently, we initiated the “prehabilitation” protocol. It consisted of continuously stimulating at 130 Hz and 1 ms of pulse width. The stimulation was bipolar and the threshold was based on causing a disabling but affordable defect in the patient’s performance that could be counterbalanced with an intense rehabilitation plan conducted by a neuropsychologist. The next day, once the speech impairment had been overcome through the rehabilitation, the threshold of the stimulation was increased following the same criteria. After 7 days, the increased stimulation was exclusively accompanied by motor side effects and did not produce any effects on speech. After a second fMRI, it was verified that the function had shifted to the contralateral hemisphere. Finally, in a repeated operation, we verified the clearance of functional activity in the stimulated cortex (Fig. 22‑3d), and, using the same surgical criteria as earlier, we were able to perform a more extensive resection. ¹⁹