Epilepsy affects >68 million people worldwide.1 In 30% of cases, seizures are deemed to be resistant to medical management, defined as failure of an adequate trial of at least 2 well-tolerated antiepileptic medications.2-4 Because of its prevalence and because it is commonly drug resistant, mesial temporal lobe epilepsy (MTLE) is one of the most common forms of partial epilepsy referred for surgical management.5
Seizures in MTLE typically start at the end of the first decade of life, and patients have an increased incidence of cerebral insults or complicated febrile seizures in the first 5 years of life compared with other forms of epilepsy.6 The most common focal pathology identified in MTLE patients who undergo surgery is hippocampal (or mesial temporal) sclerosis, which consists of various distinct patterns of neuronal loss and gliosis within the hippocampus, from classic hippocampal sclerosis involving CA1 and CA4 subfields with CA2 sparing to total hippocampal sclerosis involving all subfields.7
Penfield refined the surgical procedure for temporal lobe epilepsy during the early 20th century with his integration of cortical stimulation and the use of electroencephalography.8 His initial surgical technique focused on resection of the anterolateral temporal lobe and regions of the cortex that revealed hyperactivity during intraoperative monitoring.9 This initial method of localization, however, was met with limited success: in his review of 68 patients in 1950, only 2 hippocampal resections were performed, and 20% of the initial surgeries completed were exploratory craniotomies with no cortical resection performed, as no abnormality was noted during intraoperative evaluation. Further clinical experience and animal studies pointed toward the central role of the mesial temporal structures.8 Just 2 years later, Penfield emphasized the importance of mesial temporal lobe structures in performing successful surgery for temporal lobe seizures in his landmark report. His revised technique emphasized an anterior temporal lobectomy with amygdalohippocampectomy (ATLAH).10 Further clinical case series and 2 randomized, controlled trials established craniotomy for ATLAH in cases of medically refractory MTLE to be an efficacious and cost-effective procedure.6,11,12
As the removal of mesial temporal structures is key to a successful ATLAH for MTLE, multiple targeted surgical approaches have been proposed to achieve this task, which attempts to limit resection of lateral neocortical temporal lobe tissue and potentially limit associated neurocognitive deficits.13 Paolo Niemeyer described the transcortical selective amygdalohippocampectomy (SAH) in 1958.14 In 1982, Weiser and Yasargil15 described a transylvian SAH. Subtemporal16 and zygomatic approaches to amygdalohippocampectomy have also been reported, as well as a transorbital endoscope-assisted approach.17
Minimally invasive stereotactic surgical options for MTLE have also been proposed,18 including stereotactic radiosurgery, stereotactic radiofrequency ablation,19,20 MR-guided focused ultrasound ablation, and electrical stimulation.21 These approaches have been associated with varying degrees of success that are beyond the scope of this review. Another promising minimally invasive method, MR-guided laser interstitial thermal therapy (MRg-LITT) to perform a stereotactic laser amygdalohippocampotomy (SLAH), has met with initial success leading to rapid adoption by many epilepsy surgical centers in the United States. This approach is the focus of this review.
THE DEVELOPMENT OF LASER INTERSTITIAL THERMAL THERAPY
Lasers (light amplification by stimulated emission of radiation) were developed in the late 1950s, with the first functional laser reported by Maiman22 in 1960. Potential neurosurgical applications were immediately apparent, with the first report of laser treatment of a brain tumor in 1966.23 Stellar et al,24 in 1970, reported the first use of a laser as a handheld neurosurgical instrument for tumor resection. Enthusiasm grew through the 1970s and 1980s, but the application of lasers in neurosurgery remained limited to use as a handheld tool.25 Bown first proposed the use of laser interstitial thermal therapy (LITT) for thermal ablation of tumors in 1980.26 This was followed by the first report of stereotactic laser ablation of brain tumors in 5 patients by Sugiyama et al.27
The effect of laser thermal ablation is based on the thresholds for thermal tissue damage. Temperatures of 42°C to 45°C for periods of 30 to 60 minutes inactivate key enzymes, resulting in cell death.28 Irreversible cell damage occurs more rapidly at higher temperatures and at temperatures >60°C is nearly immediate. Tissue charring and vaporization occur at temperatures >100°C; once tissue becomes charred, laser light penetration decreases, causing the laser to act only as a focal heat source, resulting in a suboptimal extent of tissue ablation. The goal of LITT, therefore, is to attain the largest region of controlled tissue ablation possible through optimal light penetration into the tissue while limiting localized tissue charring through temperature control. Another goal is to obtain a sharp demarcation between treated and untreated tissue, which is conferred by rapid heating from photon absorption.
Near-infrared light with wavelengths of 850 to 1064 nm achieves the optimal tissue penetration,29 which is necessary to achieve even heating deep into target tissue without localized charring. Initial acceptance of stereotactic laser ablation, therefore, was limited due to an inability to monitor the region of ablation in real time and poor tissue penetration. To overcome the lack of real-time guidance, in 1994, Kahn et al30 reported intraoperative magnetic resonance imaging monitoring of laser ablation (MR-guided LITT). Earlier animal studies had shown that gadolinium-enhanced T1-weighted spin-echo images acquired immediately after ablative procedures could closely predict the ablated region on pathology.31 Initially, MR guidance was provided by simultaneous T2-weighted sequences and postablation T1 image acquisition.32 Accurate real-time monitoring of the ablated region, however, remained elusive.
Noninvasive MR thermometry was developed in the 1990s. This allowed for real-time monitoring of temperature and, therefore, the ability to accurately monitor laser thermal heating to treat deep-seated lesions adjacent to critical anatomic structures.33 Although various methods were developed, DePoorter introduced the proton frequency method, which is based on the temperature dependence of water proton-resonance frequency. This method offered the best combination of temperature detection, spatial resolution, and optimal sequence acquisition time and has become the standard MR-thermometry protocol used for MRg-LITT today.34
The second limitation to wider clinical application of MRg-LITT was related to temperature control at the laser tip. Improved temperature control would lead to decreased local tissue charring and increased light penetration. Lessons were learned from the experience with radiofrequency catheter ablation in medical applications such as cardiac arrhythmias in the early 1990s, when similar challenges were faced.35 Tissue charring at the radiofrequency probe tip caused increased impedance and decreased energy delivery. Infusing saline solution through the catheter tip led to lower impedance from reduced charring and a significantly larger region of ablation. Integration of similar technology of circulating saline solution or CO2 gas through the LITT catheter probes provided a solution for improving tissue penetration and increased ablation size.36
Two commercial systems capitalized on these technological advances to build standardized devices with integrated software systems to allow MRg-LITT to become widely used. The first system for stereotactic application of MRg-LITT reported in the neurosurgical literature was the Visualase system (Visualase Inc, Houston, TX; acquired by Medtronic Inc, Minneapolis, MN in 2014); it received US Food and Drug Administration clearance in 2007 “to necrotize or coagulate soft tissue through thermal therapy. . . in neurosurgery. . . under MR guidance.” The initial report of its clinical use for metastatic brain tumors was published in 2008.37 The Visualase system uses a 980-nm diode laser. Diode-based lasers use a semiconductor as their active-based laser medium, which allows for a relatively lower power requirement and the ability to operate the laser via a 110-V power source.38 In addition, the 980-nm wavelength has a high water absorption coefficient. Because the majority of brain and tumor tissue is water, it has the theoretical advantage of creating larger ablation zones with sharper boundaries in less time.38,39 However, comparative studies within other clinical applications, such as laser lipolysis for fat reduction, found similar clinical results with laser wavelengths between 900 and 1320 nm.40 The Visualase system uses a method of fluid convection cooling at the fiber-optic laser applicator tip using irrigation with saline.41
The Visualase laser fiber assembly consists of a 1.65-mm diameter outer polycarbonate cooling catheter (through which saline solution is circulated) that contains an inner 400-μm core silica optical fiber with a cylindrical diffusing tip (3- or 10-mm length) for delivery of the laser photons37 (Figure 1). A sharpened silica tip is applied to the end of the catheter assembly to assist with tissue penetration. Two luer lock connectors located at the proximal end of the device allow for isolated in- and outflow of fluid through polyethylene terephthalate tubing. Saline coolant at room temperature is delivered by peristaltic pump with a flow rate of 15 to 30 mL/min.
The Visualase laser fiber assembly is inserted through a skull-mounted plastic or titanium anchor bolt that is threaded into the skull stereotactically over a guide rod inserted using any stereotactic system (frame based or frameless).37,39 The patient is then transferred to the MR imaging scanner for the MR-guided ablation. An alternative is to implant the assembly in the MR imaging scanner using an MR imaging–guided skull-mounted platform, obviating the need to transport the patient midprocedure.42
Once situated in the scanner, with the cooling system and laser attached to the laser fiber assembly, intersecting planning T1 MR slices are obtained to allow the ablation to be monitored in as many as 3 planes. The software interface allows the surgeon to set temperature safety points that immediately switch the laser off when a particular temperature is reached. The theory underlying the safety points is based on time-dependent reversible tissue damage occurring at 45°C to 60°C and rapid, irreversible tissue damage occurring at temperatures >60°C. Typically, 3 “low” safety points are set at 45°C to 50°C and are placed at the margins of the desired thermal ablation zone to protect off-target tissue at risk of injury, such as the thalamus.42-46 The other 3 “high” safety points are set at 90° and are placed immediately adjacent to the diffuser tip to prevent heating to levels that would create charring/carbonization, vaporization (and gas expansion), unpredictable heat spread, and possibly melting of the device tip due to inadequate dispersion of photons.
The MR images are imported continuously via Ethernet to the Visualase workstation during therapy. After initial anatomic image sets for planning are obtained, fast spoiled gradient-recalled echo images are obtained at 4-s intervals per acquisition and sent to the workstation for derivation of thermal (phase) maps. The “irreversible damage zone” is depicted as an overlay onto anatomic images in as many as 3 planes based on the Arrhenius rate process model,47 which accounts for the cumulative effects of the time-temperature history of each image voxel (Figure 2A). Postoperative imaging, including contrast-enhanced T1-, diffusion-weighted imaging and/or fluid-attenuated inversion recovery (FLAIR) images, are obtained to confirm adequate ablations (Figures 2B, 2C, 2E). Preclinical studies have established the tight relationship between the irreversible damage zone estimate and histological evidence of necrosis.48
The NeuroBlate system (Monteris Medical, Inc, Plymouth, MN) also uses a diode laser but delivers light using a 1064-nm wavelength Nd:YAG (neodymium-doped yttrium aluminum garnet) laser (Figure 3). This wavelength has a lower water absorption and theoretically higher tissue penetration, although longer treatment times may be required to ablate a similar region of tissue. The NeuroBlate system received US Food and Drug Administration clearance in 2009 “to ablate, necrotize or coagulate soft tissue through interstitial irradiation or thermal therapy in medicine and surgery in the discipline of neurosurgery with 1064 nm lasers…under MRI visualization.”
The currently available NeuroBlate system includes 4 laser probe designs: 2.2- or 3.3-mm diameter probes with either a circumferential diffusing tip, similar to the Visualase system, or a directional side-firing tip. The system also includes additional equipment designed to optimize and automate as much as possible the MRg-LITT process.49 After being fully sedated in the operating theatre, the patient may be transferred to the AtamA patient transfer board, on which the patient will remain throughout the MRI procedure. The patient may then be placed in a stereotactic frame or aligned with a frameless neuronavigation system such as the VarioGuide (Brainlab AG, Feldkirchen, Germany). Monteris Medical, Inc also developed a miniature stereotactic frame, the AXiiS device, which is fixed to the skull for precise alignment if desired.49 After proper alignment is achieved, a stereotactic drill craniostomy and durotomy are performed, followed by placement of a skull-anchored bolt for laser probe insertion, unless the AXiiS device is used to hold the probe. With the bolt or AXiiS device in place, the patient is positioned in the MRI scanner, and the laser is inserted and secured. The laser depth is controlled via a remotely controlled robotic probe driver that is integrated with the NeuroBlate software.
The NeuroBlate software, termed M-Vision, allows for aiming of the side-firing laser tip (if used), planning trajectories, and mapping the estimated region of tissue ablation.49 Once the laser is placed, a brief MRI sequence is performed to verify probe location and plan the region of ablation.39 The software is based on thermal damage threshold lines, which estimate the time that a particular region of tissue reaches a specified temperature based on the Arrhenius model.49 A system of color-coded threshold lines corresponds to (1) the innermost perimeter of tissue that has reached 43°C for 60 minutes, (2) the middle region of tissue that has reached 43°C for 10 minutes, and (3) the outermost perimeter of tissue that has reached 43°C for 2 minutes.50 Preclinical studies have shown that any tissue within the boundaries of the second line (43°C for 10 minutes) has undergone irreversible damage, whereas any areas of tissue outside the third line have undergone no irreversible damage.50
At the completion of the procedure using either system, the laser probe is removed and the patient is transported off the MRI table. The cranial fixation device is removed; the incision is irrigated and closed, usually with a single suture. The patient is then awakened and transported to the postanesthesia care unit. No intensive care unit stay is required, and discharge after 1 day of hospitalization is typical.42
LITT FOR EPILEPSY
Curry et al43 in 2012 published the first report of MRg-LITT for medically refractory epilepsy in a series of 5 pediatric patients with epileptogenic foci (Table). In this series, there was one 16-year-old female patient with MTLE due to mesial temporal sclerosis (MTS). She was noted to be seizure free on medication at 1 year post-procedure.
Willie et al42 reported the first series of MRg-LITT for MTLE in their prospective case series of 13 patients who underwent 15 LITT procedures. The authors performed a single laser application in the amygdalohippocampal region from an occipital trajectory (Figure 2). They reported a median hospitalization of 1 day, with 54% of patients achieving freedom from disabling seizures (Engel Class I) and 30.8% of patients completely seizure free (Engel class IA) at last follow-up (range, 5-26 months). The success rate increased when including only those patients with MTS, of whom 67% (6/9) achieved Engel class I at last follow-up. One long-term complication of homonymous hemianopsia due to a stereotactic error (described in the following) was noted. Improved postprocedural Engel class was noted in procedures performed later in the series, possibly suggesting an initial learning curve with the procedure as referenced in the discussion and comment by Regis and McGonigal in the article by Willie et al.42 Drane et al51 followed in 2015 with an analysis of patients at 2 institutions who underwent either open surgical ATLAH or SAH (N = 39) or SLAH (N = 19). In patients with dominant temporal lobe epilepsy, the authors noted a greater decline in patient cognitive performance for naming famous faces and common nouns in those who underwent ATLAH/SAH compared with SLAH. Likewise, patients with nondominant temporal lobe epilepsy showed a greater decline in recognition of famous faces after open resections.
Waseem et al45 reported a prospective analysis of 7 consecutive patients (5 with MTS, 2 without MTS) older than 50 years of age who underwent MRg-LITT for MTLE. The authors compared their data for 7 consecutive patients undergoing open surgical anterior-mesial temporal lobectomy with mean follow-up of 1 year. They found a seizure freedom rates of 80% after MRg-LITT (4 of 5 patients; 2 had not reached the 12-month follow-up yet) and 100% in the patients undergoing open resection, which was not a statistically significant difference. They further reported that 2 patients undergoing MRg-LITT sustained partial visual field deficits and 1 patient who underwent open surgical ATLAH sustained aseptic meningitis, which resolved with steroid therapy. Most recently, Kang et al52 reported on 20 patients 11 to 66 years of age with follow-up at 1.3 to 3.2 years (median, 13.4 months) after SLAH for MTLE. Seventeen had radiographic findings consistent with MTS, 2 had low-grade glioma, and 1 was nonlesional. Engel class I outcomes occurred as follows: after 6 months, 8 of 15 patients (53%, 95% confidence interval: 30.1-75.2%); after 1 year, 4 of 11 patients (36.4%, 95% CI: 14.9%-64.8%); and after 2 years, 3 of 5 patients (60%, 95% CI: 22.9%-88.4%). One patient experienced a hematoma in the ablation region and resultant superior quadrantanopia; 1 patient had a transient fourth nerve palsy, and 1 patient (not seizure free) with a history of depression and suicidal ideation committed suicide.
Summing up the 3 published series of SLAH for MTLE in patients with MTS with 1-year outcomes (the gold standard), including Willie et al42 (4 of 7, excluding their first 2 patients, who did not have radiological criteria for MTS), Wasseem et al45 (4 of 5, as many as 2 of these may not have had MTS), and Kang et al52 (4 of 10, excluding 1 patient with low-grade glioma) yields a mean seizure-free rate of 55% (12 of 22). The number of patients with nonlesional MTLE who underwent MRg-LITT is too small to be meaningful.
Subsequent to the initial report of Curry et al43 of MRg-LITT in 5 pediatric patients, including 1 with MTLE, very few papers report outcomes of SLAH in children. Lewis et al46 published a retrospective review of MRg-LITT procedures for lesional epilepsy in pediatric patients. They reported an overall Engel class I outcome in 41% (7 of 17 patients), most with focal cortical dysplasias. Included in their series was 1 patient with MTS, but the procedure was aborted in that patient due to inaccurate fiber placement noted on MRI, likely secondary to a reference coregistration error during frameless navigation, according to their assessment. In the Kang et al52 series of MTLE, 2 patients were pediatric (11 and 14 years old), 1 became seizure free at 6-month follow-up, and 1 was not seizure free. Thus, there is very limited experience with LITT for MTLE in the pediatric population.
SLAH: Technical Issues
Several features of mesial temporal lobe anatomy make MTLE highly amenable to MRg-LITT. When an occipital approach is used, the laser catheter traverses the length of the head and body of the hippocampus and typically enters the inferior portion of the amygdala, allowing for the ablation of the majority of these structures as well as portions of the medial parahippocampal gyrus (Figure 2). At the same time, surrounding cerebrospinal fluid (CSF) spaces consisting of the temporal horn of the lateral ventricle and the crural and ambient cisterns act as heat sinks crucial for protecting nearby areas of the brainstem and thalamus from inadvertent thermal injury. This explains why LITT yields a curved, tubular pattern of ablation that conforms to the pial boundaries of the hippocampus, although the trajectory of the laser catheter is a straight line.43-46,53 In addition, approaching the hippocampus from a posterior trajectory spares key structures, white matter tracts, and blood vessels from physical damage caused by passage of the laser catheter. This trajectory was initially applied to MTLE for use with other minimally invasive ablation techniques, such as stereotactic radiofrequency ablation, before the advent of MRg-LITT; however, all recent reports of MRg-LITT for MTLE use this approach.18
With preoperative workup and treatment planning, a high-resolution preoperative MRI of the brain is obtained with T2 and FLAIR sequences. All SLAH cases require stereotactic placement of the laser probe, whether by stereotactic head frame placement or a frameless stereotaxic system. In cases in which the stereotactic head frame is used, the frame is attached on the day of surgery, and a volumetric computed tomography scan can be performed for registration. Alternatively, an MR-compatible stereotactic platform is placed, and MRI may be obtained for both registration and surgical planning.43-46,53 An alternate protocol was recently presented by Willie et al,42 who reported 5 patients who had laser catheters placed entirely in the MRI suite with good accuracy and outcomes.
When approaching the hippocampus from the posterior, in general, only 1 laser catheter pass is required, with 3 to 5 serial ablations planned along the pass (Figure 2). At the start of the procedure, the tip of the catheter is typically located in the middle of the hippocampal head or inferior aspect of the amygdala. At this location, regions of the parahippocampal gyrus, such as the entorhinal cortex, may also be incorporated in the ablation, although the arachnoid of the hippocampal sulcus often protects the latter, and it is not a necessary portion of the ablation. The ambient cistern lies medial to the hippocampus and protects critical brainstem structures from thermal effects of hippocampal ablation. Superiorly, the hippocampus is insulated from the optic tract and the thalamus by CSF in the choroid fissure and the superior portion of the inferior horn of the lateral ventricle. However, posteriorly, care must be taken not to select a trajectory too near the thalamus, where the intervening CSF space is quite small. Similarly, posterolaterally, the lateral ventricle is very narrow and does not adequately shield the optic radiation (external sagittal stratum). Aggressive ablation in these locations can cause an inadvertent visual field deficit. Laser power in these regions, where the hippocampus is quite narrow, should be judiciously applied.
Before initial ablation, a test dose is typically performed at 30% of the maximum 15 W to visualize thermal activity on MRI. Ablations are then performed at 55% to 70% of 15 W, typically for 2 to 3 minutes per ablation cycle. Intensities can be higher (60%-70%) anteriorly in the amygdala and hippocampal head, but should be 55% to 60% posteriorly in the hippocampal body. MR thermal sequences are used to monitor the temperature during ablations. FLAIR or diffusion-weighted imaging may be used to monitor the extent of ablation between ablation cycles. Care needs to be taken throughout the procedure to avoid CSF loss, which has been shown in deep brain stimulation cases to shift deep brain structures by as much as 6 mm.54,55
The initial ablation is made at the anteriormost extent of the amygdala/hippocampus. The laser catheter is then retracted at typically 0.5- to 1.0-cm intervals to complete the total 3 to 5 serial ablations. The lateral mesencephalic sulcus is often used as a landmark on axial imaging for the posterior extent of the hippocampus that should necessarily be incorporated, but ablations ideally should reach (but not exceed) the level of the tectal plate.42 Various sequences, such as a T1 sequence with contrast enhancement (Figure 2E), diffusion-weighted imaging, T2-weighted sequences, and FLAIR (Figure 2B), may be performed after the completion of the procedure to document the final extent of ablation. The average hippocampal lengths reported during MRg-LITT are ≥2.5 cm, and typically 60% to 80% of the hippocampus is ablated (Figure 2).42
Although many surgeons place patients on a steroid after the procedure and taper over a week,42-46,51,53 it has not been proved effective in improving surgical outcomes, and use must be balanced against potential complications. Use of dexamethasone postoperatively has been attributed to a subsequent hospitalization in a pediatric patient due to gastritis.46 The typical duration of surgery is reported at 2 to 3 hours, but this does not usually include the 1 to 2 hours required for frame placement and imaging.43-45,53
For each hippocampal ablation via the occipital approach, several patient-specific features need to be considered when planning the trajectory. Because the occipital approach uses such a long trajectory, slight variations in amygdalohippocampal anatomy may greatly influence catheter placement (Figure 2). In general, a trajectory that passes through the ventricle or choroid plexus above the hippocampus should be avoided. However, in some cases, such as when significant atrophy of the hippocampus is present, avoiding the ventricle is not possible. In our experience, entering the ventricle and/or transgressing the choroid plexus can be performed in select cases without complication.
Irregularly shaped or dilated ventricles may also interfere with the standard inferomedial approach and require a more lateral transventricular trajectory. It is important, however, not to damage the temporal lobe white matter pathways, which might confer cognitive adverse effects,51 and the optic radiation lateral to the ventricle. When planning the catheter entry site for any trajectory, it is important to avoid blood vessels on the gyral surface or in the sulci to avoid blood and CSF loss and minimize irritation to the brain.42-46
Risks and Complications
Few complications have been reported from MRg-LITT of the medial temporal lobe. The greatest risk is failure in stereotactic positioning of the laser probe, resulting in damage to off-target structures. Placing the catheter too medially may damage the brainstem, whereas placing the catheter too superiorly may damage the thalamus, internal capsule, or basal ganglia. A trajectory that is too lateral or inferior, on the other hand, may damage temporal lobe areas that affect important cognitive functions. Passing the catheter too deeply may damage the middle cerebral artery. The use of MRI for guidance reduces the risk of improper positioning. Willie et al, for example, did have a patient who experienced homonymous hemianopsia after a repeat SLAH, attributable in this case to direct damage of the lateral geniculate nucleus or optic tracts due to improper placement of an anchor bolt via the stereotactic head frame, associated with upward displacement of the catheter. Superior deviation from the desired trajectory was detected by lateral fluoroscopy relative to the frame center, which allowed for correction of the trajectory before MRI confirmation and ablation. Other instances, although rare, of homonymous hemianopsia or superior quadrantanopia can be attributed to selection of superior trajectories whereupon ablation spread into the geniculate region of the thalamus or a too far posterolateral ablation damaging the optic radiation in the external sagittal stratum in the lateral wall of the lateral ventricle. Although there have been no reports of brainstem injury, medial spread of thermal energy can lead to transient third or fourth nerve damage with catheter location more medial within the hippocampus or amygdala.
Most treatment failures (lack of seizure control) attributed to technical aspects of the surgery are caused by (1) a trajectory that spares medial structures of the hippocampus, such as the subiculum or uncal apex; (2) not passing the catheter deep enough to reach the amygdala; (3) not ablating hippocampal body far enough posteriorly, to at least the level of the lateral mesencephalic sulcus; and (4) small ablation volume due to too few ablations or low power. Most treatment failures not related to technical aspects of the surgery are attributed to patient-related factors (eg, bilateral or extratemporal seizure onsets) or patient noncompliance with antiepileptic medications.43-46,53 Suboptimal patient selection is a more common cause of treatment failure than technical factors, but this problem equally besets open resections, in which, on average, ∼25% of patients do not achieve seizure freedom.
Previous open surgical procedures for epilepsy do not preclude the use of SLAH, and seizure freedom has been obtained in patients in whom ≥1 failed previous surgical treatments failed.42,45,46 Similarly, failure of SLAH does not preclude subsequent open ATLAH. Likewise, repeat MRg-LITT can also prove successful in cases in which an initial MRg-LITT procedure did not achieve satisfactory results. Indeed, 2 of the authors (JTW and REG) have performed a repeat ablation in 6 patients in whom initial SLAH failed but who then underwent repeat procedures aimed at additional ablation of retained hippocampal tissue and/or extrahippocampal structures (amygdala, entorhinal cortex, parahippocampal white matter). Preliminary observations suggest that ∼50% of patients may be rendered seizure free after further ablation of residual amygdala, hippocampus, or other mesial extrahippocampal structures (JT Willie and RE Gross, unpublished observations, 2016). The neurocognitive effects of staged ablation of these extrahippocampal structures are unclear but are currently under investigation.51
SLAH using MRg-LITT is an important emerging therapy for MTLE. At epilepsy centers with sufficient MRg-LITT experience, SLAH is replacing or is presented as an up-front alternative to ATLAH or SAH in the surgical treatment of MTLE. Future research is required to establish the comparative effectiveness of SLAH in relation to open resection and to determine its long-term efficacy and cost-effectiveness.
Drs Gross and Willie serve as consultants to and have received research grants from Medtronic and MRI Interventions and receive compensation for these services. Medtronic and MRI Interventions develop products related to the research described in this presentation. The terms of this arrangement have been reviewed and approved by Emory University in accordance with its conflict of interest policies. Dr Jagid is a paid consultant for Medtronic. Dr Laxton serves as a consultant for Monteris, Inc and receives compensation for these services. The other authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
1. Nevalainen O, Ansakorpi H, Simola M, et al. Epilepsy-related clinical characteristics and mortality: a systematic review and meta-analysis. Neurology. 2014;83(21):1968–1977.
2. Mula M, Cock HR. More than seizures: improving the lives of people with refractory epilepsy. Eur J Neurol. 2015;22(1):24–30.
3. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med. 2000;342(5):314–319.
4. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia. 2010;51(6):1069–1077.
5. Tellez-Zenteno JF, Hernandez-Ronquillo L. A review of the epidemiology of temporal lobe epilepsy. Epilepsy Res Treat. 2012;2012:630853.
6. Engel J Jr, McDermott MP, Wiebe S, et al. Early surgical therapy for drug-resistant temporal lobe epilepsy: a randomized trial. JAMA. 2012;307(9):922–930.
7. Thom M, Mathern GW, Cross JH, Bertram EH. Mesial temporal lobe epilepsy: how do we improve surgical outcome? Ann Neurol. 2010;68(4):424–434.
8. de Almeida AN, Teixeira MJ, Feindel WH. From lateral to mesial: the quest for a surgical cure for temporal lobe epilepsy. Epilepsia. 2008;49(1):98–107.
9. Penfield W, Flanigin H. Surgical therapy of temporal lobe seizures. AMA Arch Neurol Psychiatry. 1950;64(4):491–500.
10. Penfield W, Baldwin M. Temporal lobe seizures and the technic of subtotal temporal lobectomy. Ann Surg. 1952;136(4):625–634.
11. King JT Jr, Sperling MR, Justice AC, O'Connor MJ. A cost-effectiveness analysis of anterior temporal lobectomy for intractable temporal lobe epilepsy. J Neurosurg. 1997;87(1):20–28.
12. Wiebe S, Blume WT, Girvin JP, Eliasziw M.Effectiveness, Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med. 2001;345(5):311–318.
13. Al-Otaibi F, Baeesa SS, Parrent AG, Girvin JP, Steven D. Surgical techniques for the treatment of temporal lobe epilepsy. Epilepsy Res Treat. 2012;2012:374848.
14. Spencer D, Burchiel K. Selective amygdalohippocampectomy. Epilepsy Res Treat. 2012;2012:382095.
15. Wieser HG, Yasargil MG. Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy. Surg Neurol. 1982;17(6):445–457.
16. Park TS, Bourgeois BF, Silbergeld DL, Dodson WE. Subtemporal transparahippocampal amygdalohippocampectomy for surgical treatment of mesial temporal lobe epilepsy. Technical note. J Neurosurg. 1996;85(6):1172–1176.
17. Chen HI, Bohman LE, Emery L, et al. Lateral transorbital endoscopic access to the Hippocampus, amygdala, and entorhinal cortex: initial clinical experience. ORL J Otorhinolaryngol Relat Spec. 2015;77(6):321–332.
18. Quigg M, Harden C. Minimally invasive techniques for epilepsy surgery: stereotactic radiosurgery and other technologies. J Neurosurg. 2014;121(suppl):232–240.
19. Parrent AG, Blume WT. Stereotactic amygdalohippocampotomy for the treatment of medial temporal lobe epilepsy. Epilepsia. 1999;40(10):1408–1416.
20. Vojtech Z, Malikova H, Kramska L, et al. Long-term seizure outcome after stereotactic amygdalohippocampectomy. Acta Neurochir (Wien). 2014;156(8):1529–1537.
21. Gross RE, Mahmoudi B, Riley JP. Less is more: novel less-invasive surgical techniques for mesial temporal lobe epilepsy that minimize cognitive impairment. Curr Opin Neurol. 2015;28(2):182–191.
22. Maiman TH. Stimulated optical radiation in Ruby. Nature. 1960;187(4736):493–494.
23. Rosomoff HL, Carroll F. Reaction of neoplasm and brain to laser. Arch Neurol. 1966;14(2):143–148.
24. Stellar S, Polanyi TG, Bredemeier HC. Experimental studies with the carbon dioxide laser as a neurosurgical instrument. Med Biol Eng. 1970;8(6):549–558.
25. Ryan RW, Spetzler RF, Preul MC. Aura of technology and the cutting edge: a history of lasers in neurosurgery. Neurosurg Focus. 2009;27(3):E6.
26. Bown SG. Phototherapy in tumors. World J Surg. 1983;7(6):700–709.
27. Sugiyama K, Sakai T, Fujishima I, Ryu H, Uemura K, Yokoyama T. Stereotactic interstitial laser-hyperthermia using Nd-YAG laser. Stereotact Funct Neurosurg. 1990;54-55:501–505.
28. Nikfarjam M, Christophi C. Interstitial laser thermotherapy for liver tumours. Br J Surg. 2003;90(9):1033–1047.
29. Germer CT, Roggan A, Ritz JP, et al. Optical properties of native and coagulated human liver tissue and liver metastases in the near infrared range. Lasers Surg Med. 1998;23(4):194–203.
30. Kahn T, Bettag M, Ulrich F, et al. MRI-guided laser-induced interstitial thermotherapy of cerebral neoplasms. J Comput Assist Tomogr. 1994;18(4):519–532.
31. Tracz RA, Wyman DR, Little PB, et al. Magnetic resonance imaging of interstitial laser photocoagulation in brain. Lasers Surg Med. 1992;12(2):165–173.
32. Reimer P, Bremer C, Horch C, Morgenroth C, Allkemper T, Schuierer G. MR-monitored LITT as a palliative concept in patients with high grade gliomas: preliminary clinical experience. J Magn Reson Imaging. 1998;8(1):240–244.
33. Ishihara Y, Calderon A, Watanabe H, et al. A precise and fast temperature mapping using water proton chemical shift. Magn Reson Med. 1995;34(6):814–823.
34. Wlodarczyk W, Hentschel M, Wust P, et al. Comparison of four magnetic resonance methods for mapping small temperature changes. Phys Med Biol. 1999;44(2):607–624.
35. Mittleman RS, Huang SK, de Guzman WT, Cuenoud H, Wagshal AB, Pires LA. Use of the saline infusion electrode catheter for improved energy delivery and increased lesion size in radiofrequency catheter ablation. Pacing Clin Electrophysiol. 1995;18(5 Pt 1):1022–1027.
36. Snakard EP, Miller M, Berridge B, Gowda A, McNichols RJ, Fossum T. Cooled-tip diode laser catheter for improved catheter ablation of ventricular tachycardia. J Invest Surg. 2001;14(6):357–366.
37. Carpentier A, McNichols RJ, Stafford RJ, et al. Real-time magnetic resonance-guided laser thermal therapy for focal metastatic brain tumors. Neurosurgery. 2008;63(1 suppl 1):ONS21–ONS28; discussion ONS28-ONS29.
38. Ahrar K, Gowda A, Javadi S, et al. Preclinical assessment of a 980-nm diode laser ablation system in a large animal tumor model. J Vasc Interv Radiol. 2010;21(4):555–561.
39. Willie JT, Tung JK, Gross RT. MRI-guided stereotactic laser ablation. In: Golby AJ, ed. Image-Guided Neurosurgery. London, United Kingdom: Academic Press; 2015:375–403.
40. Wassmer B, Zemmouri J, Rochon P, Mordon S. Comparative study of wavelengths for laser lipolysis. Photomed Laser Surg. 2010;28(2):185–188.
41. McNichols RJ, Kangasniemi M, Gowda A, Bankson JA, Price RE, Hazle JD. Technical developments for cerebral thermal treatment: water-cooled diffusing laser fibre tips and temperature-sensitive MRI using intersecting image planes. Int J Hyperthermia. 2004;20(1):45–56.
42. Willie JT, Laxpati NG, Drane DL, et al. Real-time magnetic resonance-guided stereotactic laser amygdalohippocampotomy for mesial temporal lobe epilepsy. Neurosurgery. 2014;74(6):569–584; discussion 584-585.
43. Curry DJ, Gowda A, McNichols RJ, Wilfong AA. MR-guided stereotactic laser ablation of epileptogenic foci in children. Epilepsy Behav. 2012;24(4):408–414.
44. Tovar-Spinoza Z, Carter D, Ferrone D, Eksioglu Y, Huckins S. The use of MRI-guided laser-induced thermal ablation for epilepsy. Childs Nerv Syst. 2013;29(11):2089–2094.
45. Waseem H, Osborn KE, Schoenberg MR, et al. Laser ablation therapy: an alternative treatment for medically resistant mesial temporal lobe epilepsy after age 50. Epilepsy Behav. 2015;51:152–157.
46. Lewis EC, Weil AG, Duchowny M, Bhatia S, Ragheb J, Miller I. MR-guided laser interstitial thermal therapy for pediatric drug-resistant lesional epilepsy. Epilepsia. 2015;56(10):1590–1598.
47. Svaasand L, Welsch A, van Gemert M. Optical-Thermal Response of Laser-Irradiated Tissue. New York, NY: Plenum Press; 1995.
48. McNichols RJ, Gowda A, Kangasniemi M, Bankson JA, Price RE, Hazle JD. MR thermometry-based feedback control of laser interstitial thermal therapy at 980 nm. Lasers Surg Med. 2004;34(1):48–55.
49. Mohammadi AM, Schroeder JL. Laser interstitial thermal therapy in treatment of brain tumors–the NeuroBlate System. Expert Rev Med Devices. 2014;11(2):109–119.
50. Sloan AE, Ahluwalia MS, Valerio-Pascua J, et al. Results of the NeuroBlate System first-in-humans Phase I clinical trial for recurrent glioblastoma: clinical article. J Neurosurg. 2013;118(6):1202–1219.
51. Drane DL, Loring DW, Voets NL, et al. Better object recognition and naming outcome with MRI-guided stereotactic laser amygdalohippocampotomy for temporal lobe epilepsy. Epilepsia. 2015;56(1):101–113.
52. Kang JY, Wu C, Tracy J, et al. Laser interstitial thermal therapy for medically intractable mesial temporal lobe epilepsy. Epilepsia. 2016;57(2):325–334.
53. Medvid R, Ruiz A, Komotar RJ, et al. Current applications of MRI-guided laser interstitial thermal therapy in the treatment of brain neoplasms and epilepsy: a radiologic and neurosurgical overview. AJNR Am J Neuroradiol. 2015;36(11):1998–2006.
54. van den Munckhof P, Contarino MF, Bour LJ, Speelman JD, de Bie RM, Schuurman PR. Postoperative curving and upward displacement of deep brain stimulation electrodes caused by brain shift. Neurosurgery. 2010;67(1):49–53; discussion 53-54.
55. Khan MF, Mewes K, Gross RE, Skrinjar O. Assessment of brain shift related to deep brain stimulation surgery. Stereotact Funct Neurosurg. 2008;86(1):44–53.
The authors provide a thoughtful overview of LITT therapy for MTLE. Although laser therapy for MTS is an attractive option for patients and sounds less risky than an open standard temporal lobectomy, there are still several outstanding issues that remain before this treatment can be offered as standard of care.
Long-term seizure control is still unknown after LITT, whereas a plethora of long-term data is available to evaluate the results of open temporal lobectomy. The cost containment and potential earlier return to work resulting from a shorter length of hospital stay may be outweighed by an unknown percentage of patients who will require a second procedure to render them seizure-free as well as the unknown number of patients who may require lifelong antiepileptic medications but who may have been off medications if an open lobectomy had been done.
The potential improvements in memory outcome with LITT are more dramatic for left-sided cases, in which the potential cognitive side effects of open surgery can be truly deleterious. For right-sided surgery, the data are less clear and neuropsychological studies of nondominant open surgery have shown long-term improvement in memory. Whether LITT should be offered as first-line therapy for right-sided TLE is unclear because the surgical results are so outstanding in this group.
Finally, the risks of LITT tend to be understated. Length of hospital stay is not the only important metric. Visual loss, misplaced catheters, and inadequate lesioning are reported more and more frequently as the procedure proliferates. The latter issue of inadequately lesioned hippocampus is also not well explored. It has been established in the open-surgery literature that complete (vs incomplete) hippocampectomy and larger volume of resected tissue correlate positively with outcome. What percentage of the hippocampus is lesioned with a linear single-pass LITT probe and how this correlates with outcome are not adequately studied.
Further studies are clearly needed to justify this therapy and its role in the decision-making algorithm for patients with MTS.
Theodore H. Schwartz
New York, New York
The review presented by Wicks et al summarizes the historical development of MRI-guided laser-induced thermal therapy and its application in selective ablation of the amygdala and hippocampus in patients with mesial temporal lobe epilepsy. The use of MR-LITT to ablate the mesial temporal structures has the advantage of sparing injury or removal of neocortical areas of the temporal lobe, thereby potentially reducing neurocognitive impairment. In fact, analysis of a small number of patients undergoing stereotactic laser amygdalohippocampectomy (SLAH) compared with anterior temporal lobectomy or selective amygdalohippocampectomy suggests greater impairment in the Boston Naming Test and modified Iowa Famous Faces Test with open surgery compared with SLAH.1 These early neurocognitive results in addition to avoidance of a craniotomy and its inherent complications encourage further development of the use of SLAH in the treatment of mesial temporal lobe epilepsy. Although the clinical data reported to date indicate a treatment effect of 55% seizure freedom at 1 year, it is important to note that this seizure outcome estimate derives from an unselected small case series (N = 22) with inherent heterogeneity in patient factors for postoperative seizure freedom. Comparatively, the rate of seizure freedom (no seizures impairing consciousness) summarized here is less than that reported by Wiebe et al2 for anterior temporal lobectomy (64%, N = 36) at 1 year after surgery. Given the sparse clinical data currently available regarding the longevity of seizure control, surgical risk, and seizure outcome, it is not yet clear whether clinical equipoise exists to hold SLAH as an equal alternative option to anterior temporal lobectomy in routine clinical treatment of mesial temporal lobe epilepsy. Further experience with SLAH in carefully designed studies assessing the surgical risk, neurocognitive outcomes, and seizure outcome will provide a basis for clinical equipoise and therefore the rationale for a randomized trial comparing anterior temporal lobectomy with SLAH.
Roberto Jose Diaz
Montreal, Quebec, Canada
Ricardo J. Komotar
1. Drane DL, Loring DW, Voets NL, et al. Better object recognition and naming outcome with MRI-guided stereotactic laser amygdalohippocampotomy for temporal lobe epilepsy. Epilepsia. 2015;56(1):101–113. View Full Text | PubMed | CrossRef Cited Here... |
2. Wiebe S, Blume WT, Girvin JP, Eliasziw M.Effectiveness, Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med. 2001;345(5):311–318. View Full Text | PubMed | CrossRef Cited Here... |