Advances in the surgical treatment of speech-eloquent gliomas

Advances in the surgical treatment of speech-eloquent gliomas

Chapter 6 Advances in the surgical treatment of speech-eloquent gliomas Nathan Konga and Matthew Tatea,b,c a Feinberg School of Medicine, Northweste...

1MB Sizes 0 Downloads 98 Views

Chapter 6

Advances in the surgical treatment of speech-eloquent gliomas Nathan Konga and Matthew Tatea,b,c a

Feinberg School of Medicine, Northwestern University, Chicago, IL, United States, bDepartment of Neurology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States, c Department of Neurological Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL, United States

Introduction: Language network in glioma patients Contemporary glioma surgery follows the maxim of maximal safe resection, with the goal of cytoreduction balanced by maintenance of neurologic function. Consistent data has shown that gross total resection, defined as complete removal of overt tumor identified on MRI (magnetic resonance imaging), results in decreased 1- and 2-year mortality and 6- and 12-month disease-free progression as compared to subtotal resections.1 On the other hand, sparing function is of high priority for providers and patients as preservation of motor, visual, and language functions significantly improves the quality of life and even extensive surgery is not curative. Brain mapping serves as the tool by which neurosurgeons can optimize this onco-functional balance. One of the most difficult scenarios for glioma surgeons involves resection of tumors in or near cortical and subcortical language hubs. Recent data from glioma patients undergoing awake mapping for resection of gliomas has defined critical cortical sites of language function within the superior temporal gyrus, posterior part of the middle temporal gyrus, the supramarginal gyrus, and the posterior middle and inferior frontal gyri adjacent to the precentral sulcus.2 Most notably, these data argue that the final common output for speech resides in the inferior portion of the precentral gyrus, not Broca’s area, which is more relevant for higher order speech processing and can often be resected if required oncologically. Equal in importance to understanding the critical cortical regions in a given patient is an appreciation for the major white matter pathways mediating language function, such as the arcuate fasciculus (AF)/superior New Techniques for Management of ‘Inoperable’ Gliomas. https://doi.org/10.1016/B978-0-12-813633-1.00006-2 © 2019 Elsevier Inc. All rights reserved.

47

48

New techniques for management of ‘inoperable’ gliomas

longitudinal fasciculus (SLF), the inferior fronto-occipital fasciculus (IFOF), and the inferior longitudinal fasciculus (ILF).3 Taken together these cortical and subcortical data in glioma patients are consistent with the dorsal and ventral stream hypothesis,4 whereby phonological and semantic information are transmitted dorsally and ventrally, respectively. In this context, recent technological advances have armed surgeons with an improved ability to pinpoint the relationship between the tumor and critical language network centers and their connections, leading to improvements in extent of resection, thereby prolonging survival, without sacrificing language function. In this review, we highlight recent advances in the understanding of the functional anatomy of language as it pertains to intrinsic tumor surgery, with particular emphasis on contemporary tools that can be used to efficiently study real-time language organization in an individual patient harboring a brain tumor and translate that personalized information into a plan for safe surgical resection. Preoperative techniques that will be discussed in this chapter include functional magnetic resonance imaging (fMRI), diffusion tensor imaging (DTI), electroencephalography (EEG), magnetoencephalography (MEG), transcranial magnetic stimulation (TMS), and neuropsychological testing (NPT). Intraoperative methods include direct electrical stimulation (DES), electrocorticography (ECoG), intraoperative magnetic resonance imaging (IoMRI), and real-time NPT.

Functional magnetic resonance imaging One common preoperative technique for surgical planning of speech-eloquent gliomas is fMRI. This modality has gained popularity in the last 25 years and is now the most utilized noninvasive method for functional brain mapping.5 The basic premise of fMRI is that increased neural activity leads to increased local concentrations of deoxygenated hemoglobin. The fMRI processing algorithms are able to create a blood oxygen level-dependent (BOLD) image by inferring the concentration of deoxyhemoglobin while the patient is being instructed to perform tasks such as speaking and listening and compared to baseline nontask conditions. This BOLD image can be co-registered on an anatomic MRI image to create a map of where the neural activity lies during these language tasks and this can also be incorporated into the intraoperative neuronavigation suite during surgical resection. Fig. 1 demonstrates a typical example of presurgical fMRI language data of a right-handed patient with a left insular low-grade glioma. Thus, fMRI gives insight into the functional anatomy of a given function as well as the relationship with the tumor in a noninvasive way and with high spatial precision. It can be used to study the activation patterns of the brain in patients who cannot undergo more traditional methods of functional analysis such as direct electrical stimulation (DES, discussed later). fMRI has the advantage of allowing simultaneous study of all regions of the brain and does not have the spatial constraints of other methods covered in this chapter. Along these lines, one very useful situation for fMRI in planning surgery is a left-handed

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

49

FIG. 1 Language-based fMRI. A standard set of language task-based fMRI (word generation, rhyming, reading comprehension, picture naming, antonyms) for a patient with a left insular low-grade glioma. Note the overlapping activation regions of multiple tasks (red dashed circle) within expected language areas in the superior temporal (A) and inferior frontal (B) gyri.

patient. For a right-hemisphere tumor, if fMRI shows the normal dramatic left hemisphere lateralization with no major activation sites on the right side, particularly if preoperative NPT reveals no significant baseline language dysfunction, the surgeon can consider performing the surgery under asleep conditions given the low likelihood of surgery-induced primary language dysfunction. Similarly if a left-handed patient has a left-sided tumor but language is predominantly localized to the right side (although it should be noted that this is uncommon) and no overt language deficits are present at baseline, again an asleep surgery can be considered, at least with respect to preserving major language networks. Fig. 2 illustrates practical examples of these two scenarios. One major drawback of fMRI is that it is not a direct measurement of neural function or activity. By using deoxyhemoglobin as a proxy for neuronal activity, the specificity and sensitivity of the test has been controversial, with some studies showing good correlation between fMRI and functional maps created in the operating room with DES,6, 7 while other studies report less reliability. In a practical sense, the consensus among glioma surgeons is that fMRI may be used to guide surgical planning or to identify possible language sites, but that intraoperative DES should be used to confirm the criticality of a given site.8, 9 In order to understand this reluctance to incorporate fMRI data into decisionmaking for an individual patient, we must consider the fundamental limitations of fMRI. For example, fMRI data is inherently noisy due to variability from heart rate and patient movement, particularly in tumor patients.10 This can lead to a low-resolution image where it is difficult to determine if changes in neuronal activity have indeed occurred or if the results are simply a product of slight shifts in perfusion.11 It has been estimated that isotropic spatial

50

New techniques for management of ‘inoperable’ gliomas

FIG. 2 Language laterality using fMRI. Two left-handed patients with high-grade gliomas that underwent language fMRI to better understand language lateralization. In panel A, the patient had a left-sided tumor but fMRI showed strongly lateralized activation on the right side (red dashed circle). In contrast, the patient in panel B had a right-sided tumor and fMRI demonstrated the more typical strongly left-lateralized language function. Both patients underwent asleep craniotomies with no preoperative or postoperative language deficits noted after gross total resection.

resolution on the order of 2  2  2 mm3 is required for unambiguous mapping for surgical purposes.11 This is much higher than the standard isotropic resolution of 4  4  4 mm3 that is currently used. Perhaps more importantly, standard fMRI has limitations in temporal resolution, as one can only study features at a timescale equal to or slower than blood flow changes, which are typically slower than electrical (i.e., primary) changes. Also, local blood flow can be altered in and around brain tumors, making fMRI analysis more difficult. Finally, even if we assume perfect acquisition and analysis of fMRI data, at best the data represents the areas that are involved in a given function, which is not the same as understanding which brain regions are critical for a given function. In other words fMRI may define several areas that receive increased blood flow but perhaps most of these regions could be compensated for if resected, while the key for neurosurgeons is to determine which are the truly incompensable areas. Due to its controversy surrounding reliability, there is no formal recommendation of using fMRI for preoperative neurosurgical planning. Even so, some neurosurgical centers use fMRI as an adjuvant modality for surgical planning. For example, the neurosurgery department at Brigham and Women’s Hospital integrates fMRI information formally into their surgical decision-making process.12 However, for most centers, fMRI is limited to identifying regions of interest around the resection location that are confirmed in the operating room with DES, or as mentioned previously, to determine global language laterality in cases where this may change surgical management.

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

51

In addition to traditional task-based fMRI discussed above, one emerging application of fMRI is in the use of resting state fMRI (rsfMRI). This technique relies on local blood flow correlations while the patient is not engaging in any overt behavioral task, thereby capturing intrinsic network nodes and their connections. There has been reasonable reproducibility of core resting state networks among individuals, suggesting that the method is reliable,13 and these data may aid surgeons in identifying areas that are part of local and global language networks that should be spared during surgical resection. In addition, given that efficient task performance is not a prerequisite for rsfMRI, information could be derived even if the patient is unable to perform language tasks, as can be seen in a subset of glioma patients. Shortcomings of rsfMRI include difficulty in analyzing data in patients with intrinsic brain damage due to noisy data acquisition and more fundamentally the exact meaning of these rsfMRIderived networks with regard to patient function are unknown given that by definition there is no behavioral correlate to these data. Nonetheless, as rsfMRI is adapted by more centers and more data published regarding reliability of rsfMRI-derived metrics in predicting functionally important networks, it is likely that rsfMRI will become a major preoperative surgical adjunct in the future.

Diffusion tensor imaging The DTI is another noninvasive preoperative imaging modality that can guide surgical planning. It takes advantage of water displacement along white matter fiber tracts, where it was noted that diffusion is increased parallel to tracts compared to that along a perpendicular axis.14 These DTI data can be acquired during traditional anatomic MRI sequences and can provide insight into the major subcortical connections that exist within the brain. Thus, DTI provides information about structural connectivity, as compared to functional connectivity data from rsFMRI. Perhaps most importantly, DTI can illustrate to the surgeon the relationship of a tumor to underlying white matter tracts. In a case series of 34 patients in India, DTI allowed the surgeon to determine if the tumor had only displaced the white matter tracts versus invading and disrupting the underlying tracts.15 Such information is of clinical utility, allowing neurosurgeons to estimate the position of the tumor in relation to major white matter tracts (such as AF, SLF, IFOF for language), which can aid in planning surgical trajectory and estimating the starting point for subcortical language stimulation during surgery (Fig. 3). Also, given recent data that white matter pathways demonstrate less plasticity than cortical sites, these bundles often represent the deep functional boundaries during aggressive glioma resections.3 Thus, DTI data can be helpful prior to surgery by allowing surgeons to estimate extent of resection and to even predict short- and long-term deficits after surgery. Despite these advantages of DTI in surgical planning and execution, a number of disadvantages must be

52

New techniques for management of ‘inoperable’ gliomas

FIG. 3 Diffusion tensor imaging (DTI) patterns in glioma. The most common pattern seen on DTI in glioma patients is displacement of white matter tracts, as shown in panel A, with a high-grade glioma (yellow) causing anterior/superior displacement of the inferior portion of the left arcuate fasciculus. Less common is splitting of white matter bundle by the tumor as shown in panel B where a pilocytic astrocytoma (yellow) divides the descending motor fibers (white arrows).

recognized when interpreting DTI information. First, DTI represents a mathematical model of white matter bundles that is subject to boundary conditions and seeding algorithms that introduce variability. In addition, DTI accuracy can be compromised in areas of crossing fibers and/or significant edema, which are significant and common issues for neuroradiologists charged with processing and interpreting DTI data in the vicinity of gliomas. Finally, by definition DTI does not actually say anything about the functional role of a defined white matter bundle, which at this time can only be learned by DES in the operating room. Despite these limitations, DTI remains one of the most useful surgical adjuncts for resection of tumors near eloquent language regions and can also be advantageous for neurosurgeons at any stage of training working to perfect their understanding of the major subcortical language pathways.

Electroencephalography and magnetoencephalography EEG and MEG are two additional noninvasive methods that rely on concepts similar to direct cortical recordings. EEG is a tool where electrodes are attached to the scalp surface and neuronal electrical activity measured during a given task (i.e., evoked potentials). MEG is similar except that the patient is placed in a device that detects small changes in magnetic fields produced by primary neuronal electrical activity. The advantages of EEG/MEG are similar to fMRI, namely, it can examine the entire cortical surface simultaneously, and thus can help with preoperative localization of language functions. A distinct advantage that EEG/MEG has over

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

53

fMRI is that the measurement is more physiologic as it directly captures electrical activity as opposed to the indirect measurement through blood flow. However, the main drawback of EEG is that the current from the neurons on the cortical surface must travel through several layers before reaching the scalp surface, and these layers have their own inherent distortions and resistances. As a result, the recorded activities on scalp EEG electrodes are a weighted sum of the underlying sources. This results in poor spatial resolution (typically on the order of 5– 9 cm), as well as a low signal-to-noise ratio, leading to limited clinical utility in terms of functional localization.16 Additionally, there is a time lag given the multiple resistive layers, causing poor temporal resolution and phase information.17 In contrast, MEG has much improved spatial resolution and thus can be quite useful in localizing cortical function, and MEG has been shown to correlate well with DES and navigated TMS in a series of glioma patients.18 The main limitation of MEG is the high cost, resulting in few centers acquiring these data, and thus there is a relative lack of data compared to other methods discussed in this chapter.

Transcranial magnetic stimulation TMS is a technique where electrical current is noninvasively applied to the cortical surface. The electrical current is discharged through a coil, which generates a magnetic field. The magnetic field penetrates through the skin, skull, and meninges without encountering significant distortion. This induces an electrical field in the underlying cortical surface. These fields can either provoke positive effects, such as hand muscle activation when stimulating the primary motor cortex, or inhibit activity, thereby creating a temporary lesioning effect, such as speech arrest with ventral premotor cortex stimulation.19 TMS, like the preoperative techniques described above, is noninvasive and has the ability to map functional cortical areas both near and far from the tumor location, as compared to DES which is limited to the exposed cortex during surgery. TMS also more closely approximates physiologic conditions as it relies on direct application of an electrical current to the cortical surface (as is done in DES), compared to indirect measurement such as fMRI. It is for this reason that TMS has been reported to have better spatial resolution than fMRI, especially for motor cortex mapping.20 Based on several comparison studies, it is estimated that TMS and DES differ by only an average of 4–8 mm for motor cortex mapping.21–24 TMS also has better temporal resolution compared to traditional fMRI, yielding real-time functional data, whereas fMRI requires significant postprocessing analysis and time-averaging. A recent study of various surgical adjuncts demonstrated that the use of TMS, when added to DTI and DES, improved surgical outcomes in a prospective multicenter study of 127 patients.25 Specifically, the authors found that when TMS was incorporated on top of existing preoperative planning techniques, the number of gross total resections increased by 16% and tumor resection amount improved by 10% on average.25

54

New techniques for management of ‘inoperable’ gliomas

The major disadvantage of TMS is reliability, particularly with respect to language mapping. The optimal parameters for reproducing speech disturbances have not been established and there are a high number of false positives when compared to DES mapping.26 There is also a theoretical concern of seizure induction during stimulation, although in a recent study that included 733 patients, 50% of whom had a seizure history, no seizures or adverse events were attributed to TMS. The study concluded that TMS was safe and well tolerated in patients even with epilepsy and should be considered for any patient with tumors near eloquent areas.27 Another practical issue is that TMS in the temporal region, which is often of interest in language mapping, can cause patient discomfort and/or reliability issues due to temporalis muscle contraction. As TMS becomes more reliable, in particular for language mapping, it could be very useful for decision-making and surgical planning for tumors near language regions. In particular TMS results could help to guide intraoperative mapping strategies. Finally, and perhaps most interesting from a neuroscience perspective, TMS could provide a nice noninvasive method of demonstrating brain plasticity through serial TMS sessions postoperatively. Such plasticity maps could identify if and when function has been distributed to different regions, information that could guide timing of surgery for recurrent/residual tumor. Lastly, TMS could be used to proactively guide redistribution of function away from tumor regions preoperatively or to augment recovery of function postoperatively.

Direct electrical stimulation DES is a technique that relies on the surgeon stimulating the cortical surface directly. This is most commonly done using a hand-held bipolar probe but can also be accomplished through an electrical grid. For motor mapping, this can be done in either awake or asleep conditions. For language mapping, the patient is kept awake and during language tasks stimulation is performed, thereby causing a temporary and reversible “virtual lesion” effect. The patient is typically asked to perform counting or object naming tasks during stimulation with the goal of predicting the functional outcome if that particular area were to be resected.28 Thus if a temporary and reproducible language deficit is elicited during stimulation, the surgical team considers this area as critical for language and will not resect the area even if involved by tumor. Using this method, surgeons can pinpoint the areas that can and cannot be removed, resulting in the maximal removal of affected brain tissue while minimizing permanent neurologic decline. DES data can be collected in real time and is to date the most reliable technique for language mapping. Thus intraoperative DES is the current “gold standard” technique for establishing critical functional regions as they relate to tumor anatomy and allow for maximal safe resection.29 An example of a typical resection with functional borders defined by positive cortical and subcortical language sites derived during DES is shown in Fig. 4.

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

55

FIG. 4 Cortical/subcortical DES mapping. Postresection intraoperative photograph of functional sites elicited by DES at 2 mA for a left middle frontal gyrus glioma: lip movement (B/C, yellow); tongue movement (E/F, green); speech delay within pars opercularis (K, orange). Other tags shown denote face sensory/motor sites. At the inferior/posterior border of the resection cavity (dashed yellow outline), subcortical stimulation of the inferior fronto-occipital fasciculus (asterisk) caused speech hesitancy. A, anterior; P, posterior; S, superior, I, inferior.

While DES has many advantages, most notably nearly a century of intraoperative data proving its utility, there are still a number of limitations. One major disadvantage is the invasiveness of the procedure, as it requires a craniotomy, and with the exception of primary motor pathway mapping, DES patients must be awake during the procedure.2,30 Also, DES can only give information about the region of the cortex that is exposed during surgery and thus available for stimulation. Distant ispilateral and/or contralateral functional contributions cannot be investigated. Finally, depending on the technique used to determine the appropriate current threshold, there is a potential for false positives and negatives. One technique used in DES is to stimulate the motor cortex with increasing currents until a robust response is observed. This current is then used for the remainder of cortical and subcortical mapping for other functions, such as language mapping. However, other regions or tasks may require currents higher than that established in the motor cortex which could result in a false negative result. Another technique for establishing a stimulation threshold that is often used in the setting of smaller, tailored craniotomies, termed “negative mapping,” is based on increasing the current intensity until afterdischarges are seen on intraoperative EEG.31 This current is then used to interrogate the function of the cortical surface during language tasks without the benefit of a true positive control. Such a method can lead to higher current

56

New techniques for management of ‘inoperable’ gliomas

application relative to positive-control defined intensities, which in theory could lead to additional false positives or seizures. Another disadvantage of DES relative to other mapping modalities is stimulation-induced seizures, which can lead to prolonged surgical time and/or inability to perform reliable cortical mapping. Finally, patients with overt neurologic deficits in the function of interest, for example, dysphasia or motor weakness, can make interpretation of DES data difficult. Nonetheless, as mentioned previously, from a practical perspective, DES remains the workhorse of neurosurgeons aimed at safe but effective glioma resection.

Additional intraoperative adjuncts: ECoG and intraoperative MRI Many of the noninvasive recording techniques described above have spatial and temporal limitations caused by signal analysis through the skull and scalp. Some of these limitations can be overcome by recording directly on the brain surface during surgery, leading to improvements in spatial and temporal resolution. The most common of these techniques is (ECoG) which is essentially the same concept as EEG except electrodes are applied directly onto the cortical surface following a craniotomy. This theoretically allows for greater sensitivity and precision as compared to scalp EEG. An observed phenomenon during ECoG recordings is that there is a significant power increase in the high gamma frequency band during certain body movements.32 This has also been observed for other brain activities including sensation, vision, and speech.33–35 Thus, the distribution of high gamma activity during these activities can be used to generate a functional map. The use of ECoG has only been compared to DES in the motor cortex with a sensitivity of 66.7% and specificity of 97%.32 For comparison, the sensitivity of fMRI is reported to be 52.6% in the same study.32 The major drawback of ECoG is that it has not been well studied as a mapping modality for speech/language in clinical scenarios and generally requires a significant amount of postprocessing, making real-time mapping challenging. However, such intraoperative processing of high gamma power to yield spatial maps of functional epicenters will likely become a mainstay of surgical mapping moving forward and the first practical application of ECoG in the context of glioma surgery. Another noninvasive technique that can be utilized during surgery is intraoperative MRI (IoMRI). Typically, MR images are obtained prior to surgery and displayed intraoperatively after registering to a reference frame. However, during the course of the operation, brain shift can occur as a result of intrinsic edema, mass resection, and drainage of cerebrospinal fluid. These shifts can lead to inaccurate localization as defined by standard neuro-navigation systems and may lead to inaccurate resections. IoMRI was developed to avoid this issue by updating anatomic information during surgery. In a retrospective review of 164 patients, it was found that those who had IoMRI were more likely to achieve gross total resection (49.3% vs 21.4%) and survived longer (90 months vs

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

57

39 months).36 Although there were some methodological issues with that particular study and the true usefulness of IoMRI is controversial, at least in principle IoMRI may have a role in some surgical scenarios, for example, deepseated gliomas or very large tumors with substantial brain shift, or more generally where transition between tumor and normal brain is less distinct and/ or local anatomic cues are less obvious. In practice, IoMRI is not widely used at this time, mainly due to significant cost, increased operative time, and paucity of randomized data demonstrating utility for experienced surgeons who utilize DES-defined functional boundaries. However, perhaps expanding IoMRI to include real-time updating of DTI data, which could, for example, better define white matter tracts after partial decompression from initial tumor debulking, or reprocessing of DTI using intraoperative DES-defined functional sites as seed points for tractography, may represent a useful adjunct in the future.

Neuropsychological testing NPT refers to a detailed assessment of functional domains that can be performed prior to and during surgical resection. NPT is a straightforward way to detect deficits in both basic and higher order cognitive function in patients with brain tumors. It is often a sensitive marker of clinical deficits, even in patients without overt neurologic complaints, perhaps due to gradual loss of function and or patient compensation strategies. For example, studies of NCT estimate that 80% of patients with central nervous system tumors have a physical or cognitive deficit.37 There are three main forms of NPT: subjective assessments, screening, and batteries.38 Subjective assessments are reports from the patient regarding the types of deficits and limitations they have in their everyday life. These results are often combined with the other two tests in order to create a comprehensive view of the patients’ clinical deficits. Screening is the most widely used form of NPT and can be performed at bedside, in several minutes, and by any healthcare professional. Screening tests include the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA). They test a variety of cognitive domains including language, working memory, and abstract reasoning. Results outside of the reference range are a strong indicator that neurocognitive impairment exists. However, like with many screening tests, if a patient scores within the reference range of normal, this cannot rule out that a subtle deficit does exist. For example, it was found that 30% of patients with brain tumors were impaired by the MoCA test, but that 70% of those same patients had deficits on more extensive NPT.39 The full neurocognitive battery testing is a comprehensive test usually administered by a trained clinical psychologist or psychiatrist. The test, usually lasting hours, examines a variety of domains including speed and executive function, learning and memory, visuospatial functioning, verbal fluency, mood, and in some cases social cognition. The test is also flexible, allowing the

58

New techniques for management of ‘inoperable’ gliomas

administrator to swap tests if certain deficits present themselves during testing. Full battery testing is usually indicated for patients who can tolerate the long exam and have impairments on screening exams, have concerns about neurocognitive symptoms, have tumors that are near presumed eloquent areas, or in the postoperative setting to assist with return to work appropriateness/timing. The next frontier of NPT is to translate simplified versions of the comprehensive battery into the operating room with real-time interpretation during DES. With respect to language, DES is most commonly performed in the context of counting, picture naming, or reading. Importantly, a trained neuropsychologist performs relevant tasks preoperatively to establish baseline accuracy levels which allows for choosing appropriate tasks and ensuring that the patient is functioning at an adequate level such that intraoperative testing will be fruitful. During surgery, real-time communication between the surgeon and the neuropsychologist during DES of cortical and subcortical structures allows for efficient and safe, yet aggressive resections. In particular, the ability of the neuropsychologist to convey not only an error but also the precise nature of the disorder (dysarthria, circumlocution, phonemic vs semantic paraphasia, hesitation, perseveration, etc.) can be quite helpful for the surgeon. For example, semantic paraphasias during frontal or temporal lobe surgery suggests proximity to the IFOF, while phonological or articulatory disturbances during stimulation in the frontal or parietal lobe indicate that the SLF/arcuate is nearby. In summary, NPT is becoming an increasingly important adjunct for surgical planning and aggressive glioma resections in eloquent language regions.

Conclusions In this chapter, we have discussed in some detail the armamentarium of the neurosurgical team for modern surgical treatment of gliomas near speech eloquent areas. Preoperative assessments include comprehensive neuropsychological testing, anatomic imaging including DTI, and task-based and resting state fMRI. These preoperative data are combined to decide upon a surgical plan that often includes intraoperative direct electrical cortical stimulation during administration of carefully chosen language tasks by a neuropsychologist in the operating room. After a cortical map of positive language sites is determined intraoperatively, resection is initiated and alternated with subcortical stimulation that is guided by the neurosurgeon’s three-dimensional understanding of the course of relevant white matter tracts relative to the overt tumor borders as suggested by DTI. Such a strategy of cortical and subcortical stimulation during resection with feedback from an experience neuropsychologist allows for a high extent of resection while minimizing permanent neurologic morbidity. Following surgery, patients again undergo serial language testing to determine changes from baseline and/or patterns of reorganization that can be used in medical decision-making going forward.

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

59

References 1. Brown TJ, Brennan MC, Li M, et al. Association of the extent of resection with survival in glioblastoma: a systematic review and meta-analysis. JAMA Oncol. 2016;2(11):1460–1469. 2. Tate MC, Herbet G, Moritz-Gasser S, Tate JE, Duffau H. Probabilistic map of critical functional regions of the human cerebral cortex: Broca’s area revisited. Brain. 2014;137(Pt 10):2773–2782. 3. Ius T, Angelini E, Thiebaut de Schotten M, Mandonnet E, Duffau H. Evidence for potentials and limitations of brain plasticity using an atlas of functional resectability of WHO grade II gliomas: towards a “minimal common brain” NeuroImage. 2011;56(3):992–1000. 4. Hickok G, Poeppel D. The cortical organization of speech processing. Nat Rev Neurosci. 2007;8 (5):393–402. 5. Smith K. Brain imaging: fMRI 2.0. Nature. 2012;484(7392):24–26. 6. Partovi S, Jacobi B, Rapps N, et al. Clinical standardized fMRI reveals altered language lateralization in patients with brain tumor. Am J Neuroradiol. 2012;33(11):2151–2157. 7. Roessler K, Donat M, Lanzenberger R, et al. Evaluation of preoperative high magnetic field motor functional MRI (3 Tesla) in glioma patients by navigated electrocortical stimulation and postoperative outcome. J Neurol Neurosurg Psychiatry. 2005;76(8):1152–1157. 8. Giussani C, Roux F-E, Ojemann J, Sganzerla EP, Pirillo D, Papagno C. Is preoperative functional magnetic resonance imaging reliable for language areas mapping in brain tumor surgery? Review of language functional magnetic resonance imaging and direct cortical stimulation correlation studies. Neurosurgery. 2010;66(1):113–120. 9. Roux F-E, Boulanouar K, Lotterie J-A, Mejdoubi M, LeSage JP, Berry I. Language functional magnetic resonance imaging in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery. 2003;52(6):1335–1347. 10. Hocke LM, Tong Y, Lindsey KP, de Frederick B. Comparison of peripheral near-infrared spectroscopy low-frequency oscillations to other denoising methods in resting state functional MRI with ultrahigh temporal resolution. Magn Reson Med. 2016;76(6):1697–1707. https://doi.org/ 10.1002/mrm.26038. 11. Yoo SS, Talos IF, Golby AJ, Black PM, Panych LP. Evaluating requirements for spatial resolution of fMRI for neurosurgical planning. Hum Brain Mapp. 2004;21(1):34–43. 12. DiMaio SP, Archip N, Hata N, et al. Image-guided neurosurgery at Brigham and women’s hospital. IEEE Eng Med Biol Mag. 2006;25(5):67–73. 13. Smith SM, Fox PT, Miller KL, et al. Correspondence of the brain’s functional architecture during activation and rest. Proc Natl Acad Sci USA. 2009;106(31):13040–13045 PMCID: 2722273. 14. Assaf Y, Pasternak O. Diffusion tensor imaging (DTI)-based white matter mapping in brain research: a review. J Mol Neurosci. 2008;34(1):51–61. 15. Dubey A, Kataria R, Sinha VD. Role of diffusion tensor imaging in brain tumor surgery. Asian J Neurosurg. 2018;13(2):302–306 PMCID: 5898096. 16. Nunez PL, Silberstein RB, Cadusch PJ, Wijesinghe RS, Westdorp AF, Srinivasan R. A theoretical and experimental study of high resolution EEG based on surface Laplacians and cortical imaging. Electroencephalogr Clin Neurophysiol. 1994;90(1):40–57. 17. Nunez PL, Srinivasan R, Westdorp AF, et al. EEG coherency. I: statistics, reference electrode, volume conduction, Laplacians, cortical imaging, and interpretation at multiple scales. Electroencephalogr Clin Neurophysiol. 1997;103(5):499–515. 18. Tarapore PE, Tate MC, Findlay AM, et al. Preoperative multimodal motor mapping: a comparison of magnetoencephalography imaging, navigated transcranial magnetic stimulation, and direct cortical stimulation. J Neurosurg. 2012;117(2):354–362 PMCID: 4060619.

60

New techniques for management of ‘inoperable’ gliomas

19. Auriat AM, Neva JL, Peters S, Ferris JK, Boyd LA. A review of transcranial magnetic stimulation and multimodal neuroimaging to characterize post-stroke neuroplasticity. Front Neurol. 2015;6:226 PMCID: 4625082. 20. Lefaucheur JP, Picht T. The value of preoperative functional cortical mapping using navigated TMS. Neurophysiol Clin. 2016;46(2):125–133. 21. Krieg SM, Shiban E, Buchmann N, et al. Utility of presurgical navigated transcranial magnetic brain stimulation for the resection of tumors in eloquent motor areas. J Neurosurg. 2012;116 (5):994–1001. 22. Paiva WS, Fonoff ET, Marcolin MA, Cabrera HN, Teixeira MJ. Cortical mapping with navigated transcranial magnetic stimulation in low-grade glioma surgery. Neuropsychiatr Dis Treat. 2012;8:197–201 PMCID: 3363137. 23. Picht T, Mularski S, Kuehn B, Vajkoczy P, Kombos T, Suess O. Navigated transcranial magnetic stimulation for preoperative functional diagnostics in brain tumor surgery. Neurosurgery. 2009;65(6 Suppl):93–98 [discussion 8-9]. 24. Picht T, Schmidt S, Brandt S, et al. Preoperative functional mapping for rolandic brain tumor surgery: comparison of navigated transcranial magnetic stimulation to direct cortical stimulation. Neurosurgery. 2011;69(3):581–588 [discussion 8]. 25. Picht T, Frey D, Thieme S, Kliesch S, Vajkoczy P. Presurgical navigated TMS motor cortex mapping improves outcome in glioblastoma surgery: a controlled observational study. J Neuro-Oncol. 2016;126(3):535–543. 26. Murdoch BE, Barwood CHS. Non-invasive brain stimulation: a new frontier in the treatment of neurogenic speech-language disorders. Int J Speech Lang Pathol. 2013;15(3):234–244 11 p. 27. Tarapore PE, Picht T, Bulubas L, et al. Safety and tolerability of navigated TMS for preoperative mapping in neurosurgical patients. Clin Neurophysiol. 2016;127(3):1895–1900. 28. Larsen S, Kikinis R, Talos IF, Weinstein D, Wells W, Golby A. Quantitative comparison of functional MRI and direct electrocortical stimulation for functional mapping. Int J Med Rob Comput Assisted Surg. 2007;3(3):262–270 PMCID: 3733359. 29. De Witt Hamer PC, Robles SG, Zwinderman AH, Duffau H, Berger MS. Impact of intraoperative stimulation brain mapping on glioma surgery outcome: a meta-analysis. J Clin Oncol. 2012;30(20):2559–2565. 30. Duffau H, Capelle L, Sichez N, et al. Intraoperative mapping of the subcortical language pathways using direct stimulations. an anatomo-functional study. Brain. 2002;125(Pt 1):199–214. 31. Saito T, Muragaki Y, Maruyama T, Tamura M, Nitta M, Okada Y. Intraoperative functional mapping and monitoring during glioma surgery. Neurol Med Chir. 2015;55(1):1–13. 32. Qian T, Zhou W, Ling Z, Gao S, Liu H, Hong B. Fast presurgical functional mapping using taskrelated intracranial high gamma activity. J Neurosurg. 2013;119(1):26–36. 33. Towle VL, Yoon HA, Castelle M, et al. ECoG gamma activity during a language task: differentiating expressive and receptive speech areas. Brain. 2008;131(Pt 8):2013–2027 PMCID: 2724904. 34. Crone NE, Hao L, Hart Jr J, et al. Electrocorticographic gamma activity during word production in spoken and sign language. Neurology. 2001;57(11):2045–2053. 35. Sinai A, Bowers CW, Crainiceanu CM, et al. Electrocorticographic high gamma activity versus electrical cortical stimulation mapping of naming. Brain. 2005;128(Pt 7):1556–1570. 36. Olubiyi OI, Ozdemir A, Incekara F, et al. Intraoperative magnetic resonance imaging in intracranial glioma resection: a single-center, retrospective blinded volumetric study. World Neurosurg. 2015;84(2):528–536 PMCID: 4532608.

Advances in the surgical treatment of speech-eloquent gliomas Chapter

6

61

37. Lehmann JF, DeLisa JA, Warren CG, deLateur BJ, Bryant PL, Nicholson CG. Cancer rehabilitation: assessment of need, development, and evaluation of a model of care. Arch Phys Med Rehabil. 1978;59(9):410–419. 38. Day J, Gillespie DC, Rooney AG, et al. Neurocognitive deficits and neurocognitive rehabilitation in adult brain tumors. Curr Treat Options Neurol. 2016;18(5):22. 39. Robinson GA, Biggs V, Walker DG. Cognitive screening in brain tumors: short but sensitive enough? Front Oncol. 2015;5:60.