Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Epilepsy affects 1% of the world’s population and incurs an enormous burden of disease. One-third of patients with epilepsy fail to achieve adequate seizure control with antiepileptic drugs or cannot tolerate their side effects. In temporal lobe epilepsy, patients may experience dyscognitive or complex partial seizures, with auras, and possibly tonic-clonic seizures. Auras are generally characteristic in semiology, including epigastric sensation, automatisms, and taste or smell sensation, among others. There are significant alterations in daily living, disability, loss of work and overall quality of life in poorly controlled temporal lobe seizures. The timely and appropriate selection for patients to undergo surgical intervention following medically refractory epilepsy has proven superior to continued medical therapy alone in regard to seizure freedom. Surgical technique for anterior temporal lobectomy (ATL) has evolved over many years, with evolution of variable approaches. In this chapter, the foundational principles behind temporal lobe resection surgery are discussed as well as a detailed technical explanation on temporal lobectomy and its variances.
The ATL is one of the most common operations for treatment of heavily epileptogenic temporal lobes. Mesial temporal lobe epilepsy (MTLE), defined by hippocampal sclerosis, is of the most pharmacoresistant forms of epilepsy. For a subset of patients with intractable MTLE, ATL offers the possibility of a seizure-free life. With appropriate selection, validated Class I evidence has demonstrated seizure freedom rates ranging from 70% to 80% with surgical intervention. Seizure freedom has paramount implications in both the adult and pediatric populations.
While the modern ATL has evolved over the years, the first applications and descriptions of the technique have existed for over a century. Jackson and Colman described seizures associated with “tasting movements” and a “dreamy state” related to a lesion in the mesial temporal lobe in 1898. Dr. Penfield at the Montreal Neurological Institute was one of the first to describe the methodical approach to temporal lobectomy. With better refinements in electroencephalograms (EEGs) and electrocorticography (ECoG), Penfield demonstrated that mesial limbic structures played a key role in what was described as “psychomotor epilepsy.” Advances in electrophysiology as well as anatomical and functional neuroimaging techniques have allowed for more focal localization of seizure foci. This has led to more “selective” resections, with sparing of the temporal lobe neocortex and parts of the amygdalohippocampal complex. Falconer and Morris described early versions of such procedures and their initial experience that would form the basis for the modern ATL. , Niemeyer soon proposed a mesial resection technique that spared the lateral cortex ; this was later adapted by Wieser, Yasargil, and colleagues and it was the forerunner of the contemporary selective amygdalohippocampectomy (SAH).
The evaluation of a patient with intractable epilepsy begins with a thorough history. Assessment of seizure onset, semiology, timing, and duration of epilepsy are recorded. All prior treatments with antiepileptic drugs, including duration, dosages, and responses are noted. A family history of epilepsy, along with identification of any other etiologic factors (e.g., history of head trauma, infection of the nervous system, exposure to neurotoxins, and so on) are queried.
Patient selection hinges on the concordance of data from several modalities, including high-resolution magnetic resonance imaging (MRI) with special attention to the mesial temporal lobe structures, EEG, single-photon emission computed tomography (SPECT), SPECT co-registered to MRI (SISCOM), and magnetic encephalography (MEG). Additional testing including neuropsychological assessment, WADA, and functional MRI studies may be needed.
In 25% of patients, additional data may be needed for seizure focus localization. Examples include cases involving possible bitemporal involvement and cases where the extent of neocortical temporal lobe involvement is not clearly defined. In such cases, invasive monitoring is indicated. Placement of epidural, subdural, and intraparenchymal electrodes can help elucidate and further refine the seizure focus in these patients. Stereo-EEG, a technique where intraparenchymal electrodes are inserted through 3-to-5-millimeter bony anchors placed through the skull openings have gained significant popularity recently, owing to advances in stereotactic techniques and intraoperative imaging that allow for precise localization and broad sampling of brain regions.
It is the concordance of these tests that are used to (1) identify the patient as one with the “surgically remediable syndrome” of MTLE, (2) localize a seizure focus suitable for resection, and (3) determine that the benefit obtained from resecting the elucidated epileptogenic zone would outweigh any language or memory deficit likely to be experienced.
The “standard” ATL, in which the lateral neocortex and mesial hippocampal structures are removed in a single specimen, was developed in the early 1950s. More recently, a two-part resection, in which the lateral and mesial portions are resected separately, has been favored by most surgeons. ATL refers to resection of all the neocortical temporal lobe structures, including the superior temporal gyrus (STG), middle temporal gyrus (MTG), inferior temporal gyrus (ITG), fusiform or occipitotemporal gyrus, parahippocampal gyrus and the uncus. However, more often a tailored approach is used to minimize removal of normal tissue. With the use of ECoG, functional imaging, and vascular anatomy, the extent of lateral neocortical resection is determined on a patient-by-patient basis. Thus preoperative patient data as well as intraoperative electrophysiology determines resection. This is often quantitated by the number of centimeters of the temporal lobe gyri resected as measured posteriorly from the temporal pole.
The procedure is usually performed under general anesthesia, unless ECoG or language mapping are planned. In the former case, nitrous oxide should be discontinued during ECoG, but the patient should remain paralyzed with depolarizing muscle relaxants; in the latter, the procedure may be performed while the patient is awake. , When ECoG is planned, perioperative use of anxiolytic medications (diazepam, midazolam, etc.) that might suppress EEG activity should be avoided.
In temporal lobectomy surgery, the patient is positioned laterally in head fixation with the nose elevated 20 degrees above the horizontal sagittal plane. This allows for better long-axis visualization of the hippocampus. To assist in venous drainage, the head of the table is raised 20 degrees ( Fig. 100.1A ). At the author’s institution, a microscope is used for the mesial structure resection. Tissue aspiration and subpial dissection is performed with the use of an ultrasonic aspirator, although bipolar electrocautery is also acceptable. Use of neuronavigation is optional.
After proper positioning, the surgical area is shaved with clippers, prepared, and degreased utilizing alcohol solution. A frontotemporal curvilinear incisional mark is made (see Fig. 100.1B ) being mindful not to extend past the posterior margin of the pinna to prevent necrosis. The surgical area that is marked off is prepared and draped in a sterile fashion. Skin incision is performed and soft tissue dissection is carried down to the root of the zygoma utilizing monopolar cautery. A myofascial flap with the temporalis muscle is reflected anteriorly.
Three burr holes are made, one at the key hole, the second at the root of the zygoma, and one at the posteriormost extent of the squamous portion of the temporal bone, just below the superior temporal line. A craniotome is used to complete the frontotemporal craniotomy. Further exposure, if necessary, is gained by using a rongeur to trim down the greater sphenoid wing, and inferiorly, to access the middle fossa. A durotomy is made starting at the posterior edge of the craniotomy in a C-shaped fashion and reflected anteriorly, anchored at the greater sphenoid wing. The cortical surface is inspected for any abnormalities, as well as for the presence of large draining veins such as the Labbé vein or the middle cerebral vein. At this time baseline ECoG activity can be recorded over the frontotemporal surface of the neocortical structures. A subdural strip may be inserted at the base of the temporal lobe in the coronal plane, placing the tip on the medial parahippocampal gyrus. Depth electrodes may alternatively be utilized to record the hippocampus through the MTG. The depth electrode is oriented perpendicular to the surface of the brain at the level of the MTG, inserted to a depth of 4 cm. This may be assisted by stereotactic image guidance.
Neocortical resection may extend between 3 and 4.5 cm from the temporal tip along the sylvian fissure on the dominant side, and 3 to 5 cm on the nondominant hemispheres. However, there are critical considerations in tailoring the extent of anteroposterior resection. Alternations may need to be made with the information obtained from cortical stimulation mapping. Further, attention must also be paid to the neurovasculature. Particular note should be made of the vein of Labbé, which is identified on preoperative MRI and confirmed intraoperatively. Secondly, attention should be paid to the cortical M3 branches overlying the superior and middle temporal gyri. Some of these vessels may arise anteriorly, then curve posteriorly, beyond the limits of the resection and should be preserved.
In nondominant cases, further posterior extension to 5 or 6 cm may possibly be achievable without significant impact on memory, language, or cognitive function. More generous lateral resections are typically reserved for situations in which there is suspicion of epileptogenicity in the neocortex. In these cases, ECoG and language mapping may be useful to tailor the resection to match the pathophysiology and to avoid postoperative language deficit. In most cases of standard MTLE, however, ECoG and large lateral resections are not necessary. , A more sparing approach to the lateral resection was developed by Spencer and colleagues, in which only 3 to 3.5 cm of the temporal tip is removed, leaving the STG intact, regardless of laterality. ,
Parenchymal resection is done in the subpial plane, emptying the cortical and subcortical structures starting with the STG or MTG ( Fig. 100.2 ). In tailored resections, often the STG is spared. The superior temporal sulcus is coagulated using bipolar electrocautery after identification and sparing of arterial end-branches. Dissection and aspiration is continued inferomedially through the white matter utilizing a bipolar or ultrasonic aspirator until the collateral sulcus is identified ( Figs. 100.3 and 100.4 ). Once the collateral sulcus is identified, the next step is to establish the cortical-pial incision line. The subpial cortical resection continues with the emptying of the middle and interior temporal gyri, as well as the fusiform gyrus (see Fig. 100.3 ). Anatomically, the medial wall of the fusiform gyrus is the previously identified collateral sulcus. Thus, the posterior and medial margins of the lateral resection are complete. To complete the resection of the neocortex, further dissection is undertaken along the white matter temporal stem at a 45-degree angle, in an effort to avoid entrance into the temporal horn prematurely ( Fig. 100.5 ). The collateral sulcus may be used again as a useful landmark, carrying out the dissection laterally and above the structure. This completes the en bloc resection of the neocortex.
Next, using the ultrasonic aspirator (30% of maximum suction and intensity) or bipolar cautery and suction under microscopic guidance, the uncal gyrus is emptied. The resection starts anteriorly at the temporal pole and ensues in a posterior direction. It is imperative that the mesial pia is identified and preserved during this process ( Fig. 100.6 ). Once identified, attention is turned to the semilunar gyrus, which marks the antero-superior limit of the uncus and runs parallel to the endorhinal sulcus. These anatomical structures may be difficult to identify initially, resulting in surgeon uncertainly on the extent of uncal resection, especially the superior border. Relevant vascular anatomy can be of great help at this stage, since it is often constant and aids in identifying a safe line of superior and posterior resection of the amygdala. As the uncal resection progresses, the supraclinoid carotid artery (ICA) is visualized through the mesial pia. Resection follows the ICA line until the takeoff of the posterior communicating artery (PCoA) and anterior choroidal (AChA) arteries are identified. The AChA runs in the endorhinal sulcus through the pia and can be used to identify this sulcus. NB: It is important to maintain the resection below the anatomic line of the endorhinal sulcus, as the anterior fibers of the optic radiations lie directly above .
The ICA is identified again through the pia along with the third cranial nerve as the mesial dissection continues. Posteriorly, the free edge of the tentorium is identified. Keeping the mesial pia intact, the ICA is followed until the bifurcation. The M1 is then used as the superior margin of aspiration, allowing for safe radical resection of the uncus. Resection of the amygdala is completed after the identification of the posterosuperior point, the proximal M1 (see Fig. 100.6B ).
Attention is now turned away from the mesial pia. The posterior margin of the previous neocortical resection is identified under the microscope. The white matter immediately superior to the collateral sulcus (which corresponds to the subcortical white matter of the MTG) is identified and aspirated medially until the temporal horn of the lateral ventricle is entered. With the aid of a nerve hook, the lateral ventricular wall is opened posteriorly. Retractors are utilized to improve visualization of the ventricular structures. One retractor is placed on the roof of the temporal horn, elevating and protecting fibers of the optic radiation just superiorly. The second retractor, typically thinner, is placed in the temporal horn, retracting posteriorly in the long axis, parallel to the hippocampus. In order to preserve fibers of the Meyer loop, which project laterally and anteriorly from the lateral geniculate body, a low and anterior opening into the ventricle is preferred (see Fig. 100.6 ). Preservation of Meyer loop fibers is ideal; however, the resultant superior quadrontopsia from disruption of these fibers is well tolerated by patients and seldom interferes with reading, working, or driving.
Once inside the temporal horn, the choroid plexus is visible. It is followed to the choroidal point, its anterior origin (see Figs. 100.4B and 100.6 ). This is an important intraventricular landmark, allowing the surgeon to navigate the surrounding anatomy. Anterior to the choroidal point is a triangular structure, the velum terminale. Anterior to the velum terminale lies the intralimbic gyrus of the uncus, identified as a white rectangle. The intralimbic gyrus is the most posterior and superior gyrus of the uncus, and has a close association with the cerebral peduncle mesially. The intralimbic gyrus is safely resectable using the intraventricular approach as follows: with great care, ultrasonic aspiration of the intralimbic gyrus is undertaken, identifying the mesial pia that separates it from the cerebral peduncle. This should be done first, to provide the mesial landmark of safe resection. The mesial pia of the intralimbic gyrus marks the posterosuperior margin of safe resection for the amygdala. An imaginary line is drawn from the mesial pia of the intralimbic gyrus, to the pia of the proximal M1, which was previously identified. Connection of this imaginary line by subpial resection, proceeding caudally, allows for resection of the uncus and the caudal amygdala (see Fig. 100.6 ).
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here