Browse the corpus

EPILEPSY AND FUNCTIONAL NEUROSURGERY Nitin T andon and Konstantin V . Slavin everal aspects of the management of seizures and epilepsy are relevant to a general neurosurgical practice, which we cover in this chapter. First and foremost, all candidates should know how to manage a patient present ing with a new- onset seizure or in status epilepticus with a brain lesion or after a craniotomy. Second, candidates are expected to be able to explain how to perform funda mental epilepsy procedures such as a temporal lobectomy for hippocampal sclerosis or resection of an epileptogenic lesion— this may be combined with the need for intracra nial recordings with implanted electrodes. Third, it is useful to have a clear process in place for mapping language and motor function for the resection of tumors located in the eloquent cortex. Lastly, the thought process behind devel oping an appropriate plan for the surgical management of movement disorders and the technical nuances of managing such cases should be known by the examinee. CASE 1 CLINICAL PRESENT A TION A 34- year- old right- handed chemical engineer, originally from Denmark, and now living and working in the United States, had an episode of confusion and difficulty finding words to express himself while at work. This episode lasted about 10 minutes, during which his colleagues said that he could not speak or made only unintelligible sounds. He was taken for an evaluation at a local emergency department for a presumed stroke, which included magnetic resonance imaging (MRI). The MRI revealed a brain mass, and he was referred to your office. He has had no further seizures and denies any other significant medical history. On your examination, he is neurologically intact— specifically, he has no dysfluency, no dysnomia, and no comprehension or speech production errors. He states that his native tongue is Danish and that he also speaks German and English. DIAGNOSTIC TESTS The patient has undergone MRI testing without and with contrast (Figure 9.1). This revealed a lesion in the left anteromedial temporal lobe that measures about 3.5  cm in diameter, has ill- defined boundaries, and is hypoin tense on T1- weighted imaging and hyperintense on T2- weighted and fluid- attenuated inversion recovery (FLAIR) sequences. There is enhancement after contrast administration and mild mass effect from the mass on the temporal horn of the lateral ventricle. ANALYSIS OF THE CASE Likely Diagnosis These imaging studies, coupled with the history of a single episode most consistent with a seizure, in a young other wise healthy male, who has since made a rapid and complete recovery, raise the possibility of a neoplasm as opposed to a vascular lesion in the temporal cortex. It is not uncommon for patients to have no symptoms with a gradually enlarging mass. The probability that this represents an inflammatory or ischemic lesion is very low, given the imaging findings and the clinical presentation. Indications for Surgery This non– contrast- enhancing lesion is most likely to be a low- grade glioma. Initial management options include biopsy or resection. Given that the lesion is superficial and the patient is young (which pushes the plan strongly in favor of a maximal resection), it is most reasonable to pro ceed with plans for a resection rather than an initial biopsy. Management Options An important consideration is the relationship of the mass to the critical language cortex in the temporal lobe. The mass is

e patient is young (which pushes the plan strongly in favor of a maximal resection), it is most reasonable to pro ceed with plans for a resection rather than an initial biopsy. Management Options An important consideration is the relationship of the mass to the critical language cortex in the temporal lobe. The mass is 92 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW located in the middle temporal lobe, and it is therefore likely that it abuts or involves language cortex. Given that the patient is right handed, the likelihood is high that he is left hemisphere dominant for language. He is also a polyglot, and Danish and English are of greatest important to him. From various studies of polyglots, it is clear that primary and secondary languages occupy distinct, although usually adjacent, cortical substrates. Therefore, it is important to map at least the two languages most important to this patient. Most centers now use some type of noninvasive activation study (functional magnetic resonance imaging [fMRI] or magnetoencephalography) to localize language in such cases, and these studies would be helpful in this case to lateralize language and to try to localize the relationship between language function and the lesion. fMRI mapping of language function was carried out in this patient. Language was mapped with visual and auditory cues (Figure 9.2). Visual naming was performed in both Danish and English. The patient had a difficult time understanding the auditory cues, and performance was low on that task but good on the others. The activation map indi cates that the left hemisphere is dominant for language and that Broca’s area is well localized. However, no activation is seen around the lesion. This raises an important issue— is the sensitivity of fMRI good enough that we can assume that the patient has no useful language function in the vicinity of the mass? NO. In the best studies, the sensitivity and specificity of fMRI in detecting essential language sites localized by electrical cortical stimulation mapping (CSM) is about 80% to 85%. Therefore, fMRI should be used chiefly to lateralize language function and not to replace CSM to localize language sites. Given the proximity of language sites to the tumor in this case, the standard approach is to perform an awake craniotomy with language mapping to enable resection with minimal compromise of language. There are certain prerequisites for an awake craniotomy and language mapping. The patient must be cooperative and not claustrophobic. The functions (e.g., naming or repetition) that the surgeon wishes to map should be nearly or completely intact (trying to map a significantly impaired function is often a futile process because the patients wake up in the operating room a bit worse than they were in the clinic, given the residual effect of anesthetic agents). The surgeon and the neuroanesthesiologist should have a protocol in place for these cases. In general, the patient is not intubated, a laryngeal mask airway may be placed, and the anesthetic agents used are mostly if not completely the intravenous CD E Figure 9.1 A, B: Axial fluid- attenuated inversion recovery magnetic resonance images showing a mass in the left midtemporal lobe. C– D: Coronal T2- weighted magnetic resonance images showing the relationship to the superior temporal gyrus and the vein of Labbé.

ts used are mostly if not completely the intravenous CD E Figure 9.1 A, B: Axial fluid- attenuated inversion recovery magnetic resonance images showing a mass in the left midtemporal lobe. C– D: Coronal T2- weighted magnetic resonance images showing the relationship to the superior temporal gyrus and the vein of Labbé. E PILEPSY AND fUNCTIONAL NEUROSURGERY • 93 anesthetics. These include remifentanil or sufentanil for pain and dexmedetomidine and/ or propofol for the anesthetic. Inhalational anesthetics are used minimally or not at all. Given that electrical stimulation of the brain is likely to result in the production of seizures, a preoperative loading dose of anticonvulsants is essential in each case. It is useful to do this a day or so before the procedure so that the patient is not too groggy during the awake portion of the operation as a consequence of the loading dose. Phenytoin or fosphenytoin (18 mg/ kg load) is preferred for this purpose, although levetiracetam (15 mg/ kg) or lacosamide (3 mg/ kg) can also be used in case of intolerance or allergy to phenytoin. Steroids (dexamethasone, 8- 10 mg) should be given in each case, and mannitol can be used on an individual basis, based on the patient’s age, renal function and the degree of mass effect. Details of the Procedure The patient is positioned lateral with a large shoulder roll and is held in place with tape and restraints. The head is immobilized with pins, which minimizes movements, par ticularly during microdissection, although positioning on a doughnut is also acceptable (in this case, a head post is clamped to the skull after the craniotomy to ensure mini mal head movements). The head is always in a lateral ori entation, slightly extended, and should be such that the patient appears comfortable and also that the anesthesiologist has easy access to the airway. A field block is created using a long- acting topical anesthetic such as 0.25% bupivacaine. After the craniotomy flap is elevated, the dura is also infiltrated with local anesthetic, the patient is awoken, and the laryngeal mask airway (if used) is removed. A neurophysiologist or a neuropsycholo gist well versed in language mapping then presents stimuli to the patient while the surgeon proceeds with electrical CSM. Stimulation is carried on after the patient is familiarized with the testing process, and a balanced Faradic current, to avoid charge deposition in the brain, is used. An Ojemann stimulator (Integra) or a Grass S88X stimulator is commonly used. A train of balanced square waves is delivered at 50 Hz. Each Auditory nouns- English Visual nouns- Danish Visual nouns- English Figure 9.2 Sagittal (left column) and coronal (right column) views of a blood oxygen level– dependent (BOLD) functional magnetic resonance imaging sequence during auditory and visual cued object naming. Strongly left lateralized activity in the frontal lobe is seen. Activity is also seen just posterior to the lesion.

h Figure 9.2 Sagittal (left column) and coronal (right column) views of a blood oxygen level– dependent (BOLD) functional magnetic resonance imaging sequence during auditory and visual cued object naming. Strongly left lateralized activity in the frontal lobe is seen. Activity is also seen just posterior to the lesion. 94 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW set of waves varies from 200 to 500 μs in duration (longer duration waves are used for mapping in children), and these are delivered for 3 to 5 seconds during task performance. Stimulation currents range from 1 to 10 mA, and mapping is performed at 10 mA or at 1mA below the potential that results in persistent after- discharges or overt phenomenol ogy, whichever is lower. Concurrent electrocorticography is essential to monitor for after- discharges and seizures. It also provides visible evidence of completeness of the electrical circuit by the presence of artifacts in the recording. The tasks used for CSM vary across institutions, but the following are commonly used: • Object naming using pictorial stimuli (e.g., Boston Naming T est) • Repetition of spoken phrases • Spontaneous speech (e.g., counting, reciting the alphabet, nursery rhymes); this is useful in localizing sites of speech arrest • Auditory naming— naming driven by a phrase or a sentence (e.g., question, “what a king wears on his head”; answer, “crown”). Stimulation sites where the task results in a delay, paraphrasia, anomia, or speech arrest are noted, as are those locations where stimulation produces overt sensorimotor phenomenology (Figure 9.3). CSM carries a small risk for producing seizures, and such seizures may severely limit the mapping process or lead to a postponement of the planned resection. Therefore, patients should receive therapeutic doses of anticonvulsants before mapping. Despite this, stimulation of perilesional cortex can result in seizure induction. This can be controlled by irrigation of the brain surface with ice- cold saline. In the patient in this case, English and Danish were both mapped. This facilitated a resection of the tumor, part of which was located posterior to the vein of Labb é. The patient recovered from surgery well and was discharged home on postoperative day 2. COMPLICA TIONS In addition to complications that are generic to any crani otomy, such as infections, cerebrospinal fluid (CSF) leak, hemorrhage, and stroke, particular complications related to this case are as follows: 1. Postoperative seizures. These are managed by maintaining high doses of anticonvulsants for 2 to 4 weeks after surgery in any patient who presented with a gliomas and a seizure. 2. Postoperative language deficit. This is minimized by the intraoperative mapping and monitoring of language function. 3. Delayed postoperative language deficit . This is not uncommon and could be related to edema or ischemia around the resection cavity. Language function is often worst on days #2 and 3, particularly if resection is very proximate to eloquent language sites. An MRI to rule out a stroke may be obtained (Figure 9.4), as should an electroencephalogram (EEG), if these deficits are profound or occur in the immediate postoperative period. Figure 9.3 Intraoperative photographs showing the left temporal lobe (surgeon’s orientation) and the results of the language mapping. A: Single alphabets are for the maps in English, and double letters are for mapping in Danish. The vein of Labbé is situated near the posterior boundary of the tumor. B: The postresection images show preservation of the language sites as well as cortisectomy, both anterior and posterior to the vein of Labbé, that skeletonize yet preserve the vein. C = comprehension; L = N + C; N, auditory cued naming; T, tongue deviation.

f Labbé is situated near the posterior boundary of the tumor. B: The postresection images show preservation of the language sites as well as cortisectomy, both anterior and posterior to the vein of Labbé, that skeletonize yet preserve the vein. C = comprehension; L = N + C; N, auditory cued naming; T, tongue deviation. E PILEPSY AND fUNCTIONAL NEUROSURGERY • 95 CASE 2 CLINICAL PRESENT A TION A 22- year- old right- handed woman is referred by her neurologist for definitive management of her epilepsy that is not well controlled with anticonvulsant medica tions. Her current seizures began at the age of 15 years. She has an aura of a rising sensation in her abdomen fol lowed by a loss of awareness and inability to respond to questions, and she is told that she is confused for a few seconds after each event. She has three or four seizure events a month. Additionally, she has had a total of five grand mal seizures over the past 7 years. She has previ ously been on and has failed adequate drug treatment with oxcarbazepine, levetiracetam, and valproic acid. She is currently taking a combination therapy of lamotrigine and lacosamide. She is accompanied at the clinic visit by her mother, who states that the patient is a product of a full- term normal vaginal delivery, and also that none of her siblings has ever had seizures. The patient did have febrile convulsions at 2 and 3 years of age. A total of five separate events occurred, all in the context of body tem peratures higher than 102 ° F . The patient was placed on phenobarbital at that time, and this was stopped when she turned 6 years old. She was seizure free and off medi cations for 9 years until the seizures began again at age 15. DIAGNOSTIC TESTS The patient’s video EEG reveals interictal EEG spikes at F8- T4 and right hemispheric seizure onsets, maximum at the same electrodes. The video shows stereotypic seizure semiology for all recorded seizures with a sudden cessation of activity, loss of responsiveness, right- hand automatisms, and tonic posturing of the left hand. The seizure lasts about 2 minutes, and the patient is able to speak immediately thereafter and regains full awareness and orientation soon after the termination of the ictal rhythm on the scalp EEG recordings. The MRI reveals volume loss of the right hip pocampus along with increased signal in the hippocampus on FLAIR sequences (Figure 9.5). No other abnormalities are detected in the brain. The patient’s neuropsychologi cal testing reveals a full- scale intelligence quotient (IQ) of 110, with a verbal IQ of 114 and a performance IQ of 104. The testing reveals deficiency in visuospatial memory, consistent with a right mesial temporal abnormality. Positron emission tomography reveals hypometabolism in the right mesial temporal lobe (Figure 9.6). ANALYSIS OF THE CASE Likely Diagnosis This patient has a classical history of mesial temporal or hippocampal sclerosis causing medically intractable complex Pre-op Post-op Figure 9.4 Coronal magnetic resonance imaging show the resection cavity and the extent of resection.

in the right mesial temporal lobe (Figure 9.6). ANALYSIS OF THE CASE Likely Diagnosis This patient has a classical history of mesial temporal or hippocampal sclerosis causing medically intractable complex Pre-op Post-op Figure 9.4 Coronal magnetic resonance imaging show the resection cavity and the extent of resection. 96 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW partial seizures. The history of febrile convulsions is typical, as is the latency between the seizures in early childhood and the onset of intractable epilepsy in the teenage years. The seizure semiology is characteristic for mesial temporal seizures, further evidenced by the EEG recordings. In summary, the patient has classical imaging and electrophysiologic features of right mesial temporal lobar epilepsy (MTLE). Indications for Surgery After failing adequate treatment with three anticonvulsants, the probability of an additional anticonvulsant leading to a durable cessation of seizures is less than 3%. This patient has failed treatment with four anticonvulsants; therefore, the probability of an additional anticonvulsant leading to seizure control is miniscule. Surgical treatment— anterior temporal lobectomy for MTLE— has been shown in a prospective randomized trial to be much more effective than treatment with anticonvulsants. This should therefore be the management plan in this case. Given that the patient is right- handed, the probability of meaningful or memory function being present in the right hemisphere is very small, and this is borne out by the neuropsychological tests that are concordant with this being a disease of the language- nondominant hemisphere. This may be confirmed by fMRI to lateralize language, but an invasive test such as an intracarotid amytal injection (W ada test) is unnecessary in this case. Therefore, no additional testing is necessary before definitive surgical inter vention for the epilepsy. Management Options Several approaches are described for a temporal lobectomy. These include anterior temporal lobectomy with amygdalohippocampectomy, a selective amygdalo- hippocampectomy Figure 9.5 Coronal fluid- attenuated inversion recovery (A) and T2- weighted (B) magnetic resonance images showing increased signal and volume loss in the right hippocampus. AB C Figure 9.6 Coronal (A, B, C) fluorodeoxyglucose positron emission tomography images showing hypometabolism in the right temporal lobe.

ectomy Figure 9.5 Coronal fluid- attenuated inversion recovery (A) and T2- weighted (B) magnetic resonance images showing increased signal and volume loss in the right hippocampus. AB C Figure 9.6 Coronal (A, B, C) fluorodeoxyglucose positron emission tomography images showing hypometabolism in the right temporal lobe. E PILEPSY AND fUNCTIONAL NEUROSURGERY • 97 or a laser interstitial thermal ablation of the hippocampus. Here, we describe a classical anterior temporal lobectomy plus amygdalo- hippocampectomy. Details of the Procedure The patient is positioned with a shoulder roll under the right shoulder with the head in a lateral orientation, right side uppermost. It is useful to tilt the vertex slightly toward the floor and to extend the neck. A small inverse questionmark incision is made to encompass most of the temporalis muscle, and a myocutaneous flap is elevated. A craniotomy flap is elevated extending as low in the middle cranial fossa as possible, the superior aspect of the mastoid air cells may need to be drilled down, and this region is then closed off with bone wax. The dura is then opened in a C- shaped fashion. Electrocorticography is then carried out in conjunction with the neurologists, and the diagnosis of anteromesial temporal epilepsy is confirmed. The approximate distance from the tip of the temporal pole to a spot 4–5 cm from it along the middle temporal gyrus is marked as the posterior extent of the lateral resection. In case the vein of Labbe is located anteriorly, care is taken to preserve this vein by tai loring the lateral resection if needed. A corticectomy is then made in the superior temporal gyrus (STG) (or in the middle temporal gyrus if on the left side) and taken forward to the temporal pole. The piaarachnoid is coagulated and sharply divided. A  subpial approach is then used to elevate the STG away from the sylvian fissure and then more inferiorly, from the inferior limb of the circular sulcus of the insula (Figure 9.7). Next, a vertical corticectomy, extending downward from the posterior edge of the STG corticectomy and staying at or anterior to the 5- cm mark, is performed. The inferior edge of this is extended medially, and the basal cortex is aspirated in subpial fashion till the collateral sulcus is identified. The collateral sulcus is then exposed, with care taken to pre serve the pia within it, and is traced superiorly until the white matter of the temporal lobe is exposed. Dissection is carried further superiorly along the line of the collateral sulcus until the temporal horn of the lateral ventricle is exposed. After the temporal horn is exposed, the cut in the STG and the superior circular sulcus of the insula is con nected to this by cutting across the temporal stem. Lastly, a cut is made in the floor of the temporal horn, along a groove called the lateral ventricular sulcus, down to the basal pia. In this fashion, the temporal pole is removed en bloc, and attention is directed to the hippocampus and amygdala. The choroid plexus is then identified, and the amygdala is resected below an imaginary line connecting the choroidal point— a pale- blue translucent portion of the choroidal fissure— to the genu of the middle cerebral artery. The subpial technique is used to resect the medial structures, preserving the contents of the ambient cistern— the posterior cerebral artery and the third nerve. Next, the hippocampus is retracted slightly laterally away from the choroidal point. Using microsurgical techniques, the fimbria of the hippocampus is aspirated to expose the ves sels in the hippocampal hilus. The hilus is further exposed along its length, and the vessels are coagulated and cut as far distal as possible to prevent inadvertent damage to “en passage” vessels supplying the brainstem.

int. Using microsurgical techniques, the fimbria of the hippocampus is aspirated to expose the ves sels in the hippocampal hilus. The hilus is further exposed along its length, and the vessels are coagulated and cut as far distal as possible to prevent inadvertent damage to “en passage” vessels supplying the brainstem. The hippocampus is rotated laterally, and the tail is then cut, after which the hippocampus is removed and submitted for subsequent histopathologic analysis. Remnants of the uncus and the tail of the hippocampus are aspirated using subpial resection techniques. The surgical bed is copiously irrigated, and the dura is closed in a watertight fashion. AB C Figure 9.7 Coronal views showing the landmarks and the stages in a temporal lobectomy. A: The superior temporal sulcus and the collateral sulcus are highlighted. B: The neocortical resection extends to the ventricle and stops at the collateral sulcus. C: The medial temporal resection includes the amygdala and hippocampus.

ure 9.7 Coronal views showing the landmarks and the stages in a temporal lobectomy. A: The superior temporal sulcus and the collateral sulcus are highlighted. B: The neocortical resection extends to the ventricle and stops at the collateral sulcus. C: The medial temporal resection includes the amygdala and hippocampus. 98 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW The “selective” amygdalo- hippocampectomy targets removal of the medial structures through either a trans sylvian approach or a selective removal of the inferior or middle temporal gyri. Meta- analyses comparing the out comes of these minimalist approaches with a traditional resection suggest similar or only slightly better neuropsy chological outcomes combined with somewhat worse sei zure- free outcomes. This paradox, of a smaller resection resulting in a similar neuropsychological outcome, may be explained as a consequence of the collateral white matter damage associated with the selective approaches. A more recently introduced approach— the MRI- guided laser interstitial thermal therapy, or Visualase (Medtronic), approach— avoids the white matter injury by using a ste reotactically placed laser fiber to deliver thermal energy, monitored with real- time MRI to ablate the medial tem poral lobe structures. The entry tract is through an occip ital entry point and should be within the hippocampus and amygdala for the longest possible component of the trajectory. Early studies suggest that the neuropsycho logical outcomes, particularly with regard to preventing naming deficits, are better with this approach than with traditional or selective resections, though the seizure free outcomes at 1 year, are not as good. In this case, the patient underwent an uneventful resection and had no discernible neuropsychological compro mise at follow- up neuropsychological testing carried out 1 year later. COMPLICA TIONS In addition to complications that are generic to any cra niotomy, such as infections, CSF fistula, hemorrhage, and stroke, particular complications related to this case are as follows: 1. Anterior choroidal artery stroke (Figure 9.8). This is the most common vascular injury that can occur and is usually related to excessive manipulation in the choroidal fissure or overzealous resection of the mesial structures. The patient will emerge from anesthesia hemiparetic or hemiplegic, with a contralateral hemianopsia. Speech is affected in the dominant hemisphere. 2. CSF rhinorrheas can be due to a leak through the dura into the mastoid air cells if these are not well sealed with bone wax. 3. Damage to visual pathways can occur by excessive retraction of the roof of the temporal horn where the optic tract is located. 4. Diplopia from damage to cranial nerve III. The third nerve is located in the ambient cistern and is routinely visualized during the resection of the uncus. If the subpial technique is not used, or excessive bipolar coagulation is used near the ambient cistern, injury to this nerve may occur. Figure 9.8 Diffusion weighted image (left) and T2 weighed image showing an acute infarction of the posterior limb of the left internal capsule - in the territory of the left anterior choroidal artery.

chnique is not used, or excessive bipolar coagulation is used near the ambient cistern, injury to this nerve may occur. Figure 9.8 Diffusion weighted image (left) and T2 weighed image showing an acute infarction of the posterior limb of the left internal capsule - in the territory of the left anterior choroidal artery. E PILEPSY AND fUNCTIONAL NEUROSURGERY • 99 CASE 3 CLINICAL PRESENT A TION This 63- year- old right- handed man presents with com plaints of severe tremor that occurs only during active movements in his hands. The tremor started more than 20 years ago; it has been getting progressively worse to the point that the patient now cannot write or even hold a cup of water or a utensil, and this results in a significant disability. The patient mentions that his father had similar symptoms but much later in his life. He states that the symptomatic improvement with oral medications was temporary; right now, his tremor does not seem to respond to medications at all. The only thing that reliably controls the tremor is alcohol, but after the alcohol effects wear off, the tremor tends to get even worse. The patient is referred by his general neurologist for the surgical treatment of his tremor. DIAGNOSTIC TESTS The first question is, obviously, about the exact diagnosis— answering it helps with decision making regarding surgical intervention. T wo features that enable making the diagnosis are the tremor type and the associated symptoms. The tremor that occurs during the movement is characteristic of essential tremor (ET) rather than Parkinson’s disease (PD). Holding objects with a stretched hand will provoke the tremor. The tremor becomes worse in terms of its amplitude as the movement gets closer to the target, such as with bringing a cup closer to the lips or the pen closer to the paper. The tremor in ET cases is fairly regular and reproducible. The rest of the neurological examination in ET patients is usually normal; presence of other findings suggests a concomitant neurological condition. Responsiveness of the tremor to alcohol intake is typical for ET but is not required for making the diagnosis. ET tends to run in families. However, a negative family history is not an exclusion criterion because nonfamilial cases of ET are not uncommon. PD- associated tremor tends to be rest tremor— the patient presents with rhythmic movements in the extremities at rest. Also, parkinsonian tremor tends to disappear when voluntary movement is initiated. It is very regular and rhythmic and is frequently described as a “pill- rolling” phenomenon. T remor in PD is part of a larger clinical syndrome that includes other motor (rigidity, bradykinesia, gait and posture disturbances) and nonmotor (constipa tion, sleep disorders) manifestations. The intention tremor and otherwise normal neurological examination observed in our patient would be inconsistent with the diagnosis of PD. Parkinsonian symptoms improve with levodopa and dopamine agonists rather than alcohol. The tremor observed in multiple sclerosis, various neu rodegenerative conditions, or after head injuries tends to be intentional as well, but it may resemble a severe ataxia with an irregular pattern and larger amplitude of movements. Commonly, other neurological manifestations are present, and the brain imaging may reveal specific abnormalities such as demyelination or tissue loss. ANALYSIS OF THE CASE Indications for Surgery In most cases, definitive diagnosis and failure to respond to medical treatment are sufficient to consider surgical intervention. Whereas PD patients should get a formal neuropsychological evaluation to ascertain their cogni tive status, this is generally not needed in ET cases unless there is a particular concern about concomitant demen tia.

e diagnosis and failure to respond to medical treatment are sufficient to consider surgical intervention. Whereas PD patients should get a formal neuropsychological evaluation to ascertain their cogni tive status, this is generally not needed in ET cases unless there is a particular concern about concomitant demen tia. It is, however, strongly recommended to get an opin ion from a movement disorder neurologist, especially in less- than- straightforward cases. Failure of medical management is usually defined as persistence of symptoms despite using therapeutic doses of established medications or when there is an inability to tolerate such medications because of side effects. Another consideration for surgery is the functional impact of the disease— the potential risks of surgical intervention are not justified if the symptoms do not affect the patient’s func tional status. Choice of Intervention In surgery for movement disorders, there are two main issues— the target for intervention and the surgical modal ity. For tremor control in both ET and PD, the ventral intermedius nucleus (VIM) of the thalamus is an estab lished target. For control of most PD symptoms, both the subthalamic nucleus (STN) and ventroposterior seg ment of the internal part of the globus pallidus (GPi) are accepted targets. For treatment of dystonia, GPi is the most commonly used target. As to the intervention, in the past the lesional approach was commonly used (e.g., stereotactic radiofrequency thalamotomy, pallidotomy), but this is now almost always replaced by stereotactic placement of chronic stimulation electrodes and implantable pulse generators. The deep brain stimulation (DBS) approach is now considered a “gold standard” in movement disorder surgery; the devices

used (e.g., stereotactic radiofrequency thalamotomy, pallidotomy), but this is now almost always replaced by stereotactic placement of chronic stimulation electrodes and implantable pulse generators. The deep brain stimulation (DBS) approach is now considered a “gold standard” in movement disorder surgery; the devices 100 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW used for DBS are fully approved by the U .S. Food and Drug Administration for use in ET and PD and on a humanitarian device exemption basis for use in dystonia. Both lesions and DBS produce immediate and lasting symptomatic improvement. The rate of surgical procedural complications is comparable between two modalities, but there are important distinctions. Lesions are definitely cheaper and do not require general anesthesia. There is no concern about malfunction of an implanted device, no need for adjustments of stimulation settings, and no need to replace the generator because of eventual battery depletion. However, the side effects of lesioning cannot be resolved by reprogramming or turning the stimulator off, and the rate of complications from bilateral lesioning is unaccept ably high. Moreover, in contrast with interventions on the VIM nucleus of the thalamus and GPi, where the lesional approach has been used for decades, clinical experience with STN lesioning is limited, and the concern about pos sible development of hemiballism due to STN destruction remains open. The last question to answer is whether to proceed with unilateral or staged bilateral or a single- stage bilateral DBS procedure. For ET, we and many others start with unilateral intervention on either the more symptomatic or the dominant side. If needed, the second side can be treated a few months later— most patients eventually require second- side surgery. For PD and generalized dystonia, unilateral inter vention is rarely sufficient; therefore, most centers choose to intervene on both sides either in the same sitting or a few months apart. Doing both sides concurrently is more efficient and cheaper. However, prolonged bilateral surgery may increase the risk for procedure- related confusion and, at least theoretically, increases the possibility of brain shift on the second operated side due to the development of pneumocephalus. Details of the Procedure There are many ways one can find the way to the desired deep targets— and most of them are equally accurate and precise. Despite widespread acceptance of frameless imageguided navigation and intraoperative three- dimensional imaging (computed tomography [CT] or MRI), most centers continue using traditional stereotactic frames and intraoperative fluoroscopy. The stereotactic frame is applied on the morning of the procedure. Stereotactic imaging is obtained with the frame and a proper localizer to calculate the target coordinates, and high- resolution brain MRI or high- resolution frame- based CT is obtained (the latter is fused with the preoperative frameless MRI). For STN and GPi targeting, the atlas- based coordinates are usually crosschecked against those obtained from direct visualization of the target. For VIM targeting, direct visualization is not an option; therefore, the atlas coordinates that are based on established intracranial landmarks are routinely used. Either way, the radiographic coordinates are then confirmed with neurophysiologic intraoperative testing. A standard VIM thalamic target is placed in the plane of anterior and posterior commissures, at one fourth of the intercommissural distance behind the midcommis sural point and at about 11 mm lateral to the wall of third ventricle.

adiographic coordinates are then confirmed with neurophysiologic intraoperative testing. A standard VIM thalamic target is placed in the plane of anterior and posterior commissures, at one fourth of the intercommissural distance behind the midcommis sural point and at about 11 mm lateral to the wall of third ventricle. This puts the target medial to the internal cap sule and anterior to the sensory thalamus; proximity of these structures is subsequently tested with intraoperative stimulation. The planning also calculates the angles for the optimal trajectory of the approach— a precoronal entry is recommended to place the trajectory in front of the motor pathways, and a lateral angle is chosen to avoid ventricular penetration because transventricular trajectories may carry a higher risk for hemorrhagic complications. After the imaging and planning are completed, the patient is brought to the operating room, where a bur hole is placed according to the surgical plan. A standard perforator attached to a high- speed drill is used to create an opening that is matched by the electrode- holding device. The elec trode insertion starts with insertion of a guiding cannula that stops few millimeters short of the target. Use of microelectrode recording (MER) is an optional stage to refine targeting— thus far, no studies have shown that the use of MER results in better clinical outcomes. In particular, MER is more useful in STN and GPi targeting where it helps to define ventral and dorsal borders of the targeted structures; in VIM targeting, MER may help to detect tremor cells and define somatotopic representation of the nucleus. With or without MER, the best physiologic confirma tion of precise targeting is obtained with so- called macrostimulation. Such stimulation allows one to check beneficial effects of stimulation on specific symptoms (e.g., tremor, rigidity) and defines thresholds for motor and sensory side effects. For VIM targeting, the proximity of the internal capsule is tested by determining the threshold of motor contractions at a low frequency of stimulation. Low thresholds indicate that the electrode is located too laterally and has to be repositioned in a medial direction, usually by 2 mm. Higher frequency stimulation defines proximity of the sensory thalamus that lies immediately behind the VIM nucleus. Eliciting paresthesias in the contralateral side of the face and hand indicates good electrode position for control of hand tremor (the goal of intervention), but very low threshold of sensory effects suggests too posterior position

defines proximity of the sensory thalamus that lies immediately behind the VIM nucleus. Eliciting paresthesias in the contralateral side of the face and hand indicates good electrode position for control of hand tremor (the goal of intervention), but very low threshold of sensory effects suggests too posterior position E PILEPSY AND fUNCTIONAL NEUROSURGERY • 101 of the electrode, necessitating its reposition in a more anterior location. Lack of tremor control and absence of par esthesias suggest that the electrode is too anterior, whereas lack of tremor control and absence of motor response, even at high amplitudes of stimulation, may indicate that the electrode is too medial. After the optimal target location is confirmed with physiologic testing, the DBS electrode is inserted and locked in place with a bur hole– based holding device. Fluoroscopy or other means of intraoperative imaging are used to confirm the electrode position. The electrode is then connected to a special plug that is buried under the skin for easier electrode identification during the second stage of surgery, and the incision is then closed and the frame removed. Usually, the patient undergoes postopera tive imaging to check for any hemorrhagic complications and to document the electrode location (Figure 9.9) and is kept in the hospital overnight for observation. The second stage of surgery, which includes implantation of a pulse generator, is usually done 5 to 10 days later under general anesthesia on an outpatient basis. Device programming may be initiated on the day of implantation or may be delayed by a few weeks to allow the patient to recover from the microlesioning effect that is sometimes observed after electrode insertion. The programming sessions may have to be repeated on multiple occasions until optimal stimula tion settings are established. TROUBLESHOOTING If the patient presents with stimulation- induced side effects, the stimulation settings may have to be adjusted accordingly— changing electrode polarity, stimulation amplitude, and pulse width would usually resolve most problems. The frequency of stimulation and the mode (monopolar vs. bipolar) may have to be changed as well if needed. If, on the other hand, the patient loses beneficial effects of stimulation, the first step is obviously to check the gen erator; battery depletion in non- rechargeable devices would necessitate generator replacement, but in rechargeable devices this issue may be solved by changing the recharge technique and recharge frequency. The impedance check allows for the detection of a short circuit in the system, indicating the fracture of insulation, or an open circuit situation due to either a disconnect between the system components or the fracture of either the electrode or the extension cable. Normal- range readings should prompt an attempt to regain stimulation benefits by changing the param eters of stimulation and polarity of the electrodes, but if this does not work, the next step would be to image the patient’s brain looking for electrode migration. All of these circumstances— disconnection, fracture, and migration— necessitate reoperation. Similarly, reoperation is needed in the case of skin erosion over the implanted hardware or an infection of the entire system or its components. Figure 9.9 Postoperative computed tomography scan (A) and magnetic resonance image (B) of bilateral thalamic deep brain stimulation electrodes.

ation— necessitate reoperation. Similarly, reoperation is needed in the case of skin erosion over the implanted hardware or an infection of the entire system or its components. Figure 9.9 Postoperative computed tomography scan (A) and magnetic resonance image (B) of bilateral thalamic deep brain stimulation electrodes. 102 • G OODMAN ’S N EUROSURGERY O RAL B OARD R E v IEW SURGICAL NUANCES DBS procedures are performed differently in every cen ter. Some additional nuances are worth mentioning here. For example, an alternative to the frame is a frameless approach— this may include a custom- made targeting device that is individually created for each specific proce dure based on the patient’s anatomy, or a bur hole– based miniature stereotactic frame that is integrated with an intraoperative navigation system. Performing the procedure with the aid of intraoperative CT allows one to keep the patient awake, but use of intraoperative MRI usually necessitates general anesthesia and patient immobilization. This translates into pure anatomic targeting because physiologic testing of an anesthetized patient in an MRI environment is all but impossible. A new alternative to traditional lesioning and to DBS is the use of MRI- guided focused ultrasound. Here, a test lesion with lower temperature in the focal point may allow one to see beneficial effects and side effects associated with this particular location; the target may be adjusted based on the patient’s feedback before the higher temperature (permanent) lesion is created. USEFUL HINTS Being able to match the right target with a common clini cal indication is the first step in addressing each movement disorder case. Knowing the anatomic relationship of the stimulation target with surrounding structures is the key to functional targeting. For the purposes of the Oral Board Examination, the ability to determine the direction of tar geting error is more important than memorizing exact coordinates for each common DBS or lesioning target.