Radiation Therapy of Epilepsy

Published on 13/03/2015 by admin

Filed under Neurosurgery

Last modified 13/03/2015

Print this page

This article have been viewed 1745 times

Chapter 106 Radiation Therapy of Epilepsy

Ellen Air, Nicholas Barbaro

History

Experience with stereotactic radiosurgery (SRS) over the 60 years since its development by Lars Leksell ¹ has demonstrated its efficacy in the treatment of various central nervous system conditions. Although Leksell initially developed this technique to treat functional disorders such as pain and movement disorders, the earliest widespread uses of SRS focused on deep-seated tumors or arteriovenous malformations (AVMs) located in eloquent regions of the brain, using it as a means of treating these lesions while avoiding the risks associated with surgical resection.² Within these contexts, seizures were treated as secondary manifestations of the primary disorder, rather than as specific conditions. Subsequent to radiosurgical treatment, significant reduction, or even resolution, of the associated seizures has been documented.

The efficacy of SRS in secondary epilepsy has been most evident following the treatment of AVMs.³^–⁹ Several large series of patients with AVMs have found complete seizure remission in 50% to 80% of patients following SRS.^5,⁷^–⁹ Small lesion size,⁷ shorter duration of epilepsy,⁵ and absence of prior hemorrhage⁹ were associated with higher seizure-free outcomes. In contrast, seizure outcome appears to be independent of AVM obliteration.^3,^4,^6,⁸ Steiner et al. found 3 of 11 patients who became seizure free following SRS did so despite persistence of the AVM,⁸ while Lim et al. found 40% of patients with partial AVM obliteration to be seizure free.⁶ A similar phenomenon was described by Schrottner et al. in the treatment of 24 patients with tumor-related intractable epilepsy. Prolonged seizure control was achieved in 54% of patients in a dose-dependent manner, while tumor control was achieved in all patients.¹⁰ These studies indicate that radiosurgery exerts an independent antiepileptic effect.

Animal studies have provided additional support for the use of radiosurgery in the treatment of seizures, as well as insight into the potential mechanism underlying its antiepileptic effect. One hypothesis has been that radiation causes destruction of epileptogenic tissue. Dose-dependent cell death and radiation necrosis develop in both normal and epileptic rat models in response to increasing doses of radiation.¹¹^–¹⁵ Similarly, a dose-dependent effect of radiation on seizure control has been demonstrated in several rat models of epilepsy. However, seizure reduction occurred in the absence of necrosis,^11,^13,^14,¹⁶ suggesting necrosis may not be required for antiepileptic effects.

Another hypothesis has been that radiation inhibits seizure-induced neurogenesis and mitosis. In electrically kindled rats, radiation prevented kindling-associated neuroblast proliferation.^17,¹⁸ Similarly, radiation halted seizure-induced mitosis in flurothyl-kindled mice.¹⁹ Despite the clear effect on neurogenesis and mitosis in these models, radiation did not inhibit mossy fiber synaptic reorganization,²⁰ kindling progression, or seizure threshold.^17,¹⁹ Alternatively, radiation may induce vascular and inflammatory changes that lead to seizure reduction.²¹ Further animal studies will be critical to understanding the mechanisms underlying radiation-induced seizure control.

Modern Indications

Hypothalamic Hamartomas

Hypothalamic hamartomas (HHs), though rare, are an important cause of debilitating epilepsy in childhood. HHs can cause a progressive epileptic encephalopathy characterized by seizures, endocrine dysfunction, cognitive and behavioral impairment, and developmental delay. Classically, HHs are associated with gelastic seizures, which appear as recurrent bouts of emotionless laughter or grimacing. In addition, most patients develop multiple seizure types, including tonic–clonic, partial complex, and drop attacks.^22,²³ As devastating is the progressive cognitive decline that accompanies the medically refractory epilepsy.²⁴

These heterotopic masses of mixed neuronal and glial cells arise from the floor of the third ventricle or mammillary bodies and present a significant therapeutic challenge due to their deep location and critical surrounding structures. Both open and endoscopic surgical approaches have been used in the resection of these lesions. While good seizure outcomes have been achieved, resection is associated with high morbidity and mortality.²⁵^–²⁷ SRS has therefore emerged as a primary treatment for HHs due to its ability to precisely target and treat HHs without the morbidity of surgical resection.

Arita et al. first applied gamma knife radiosurgery (GKS) to the treatment of a 25-year-old male with HH-associated epilepsy who remained seizure free 21 months after treatment.²⁸ Regis et al. have the largest experience with GKS treatment of HH. They reported seizure cessation in 40% of patients, with an additional 20% experiencing only rare, nondisabling seizures after 3 years.²⁹ Benefits extend to improved sleep, behavior, and learning; these have been documented independent of seizure remission.²⁹^–³⁴ No surgical mortality and limited morbidity have been reported in these patients. For these reasons, we advocate SRS as a valid option in the treatment in most patients with HHs, especially when the lesion is confined to the third ventricle.

Mesial Temporal Lobe Epilepsy

The success with secondary epilepsies and HHs has prompted interest in the use of SRS for mesial temporal lobe epilepsy (MTLE) associated with mesial temporal sclerosis. In 1999, Regis et al. reported the first seven cases of amygdalohippocampectomy by GKS. All patients exhibited a reduction in seizure frequency, and all but one patient remained seizure free 2 years after the operation. In addition, magnetic resonance imaging (MRI) changes were noted, specifically in the amygdalohippocampal target, indicating focused destruction.³⁵ Since Regis et al.’s initial study, two large multicenter prospective trials have been completed. In the European trial a 24-Gy marginal dose was used, while in the U.S. trial 20- and 24-Gy marginal doses were compared. In both series, 51 patients in total, a significant reduction in seizure frequency was observed by 1 year and 65% of patients were seizure free at 2 years.^36,³⁷ A multicenter prospective trial directly comparing GKS to open surgery (Radiosurgery or Open Surgery for Epilepsy, or the ROSE trial) is under way in the United States.

GKS has also shown efficacy in the treatment of recurrent seizures after incomplete temporal lobectomy. Yen et al. treated four such patients using GKS. This resulted in significant, persistent seizure reduction in all four patients without morbidity.³⁸

Secondary Epilepsies

As discussed earlier, SRS has shown efficacy in the treatment of seizures secondary to AVMs and tumors. SRS has also been applied in the treatment of cavernous malformations (CMs). Seizure control following GKS is less than that found in the treatment of AVMs or tumors, with remission rates ranging from 25% to 53%.³⁹^–⁴¹ In addition, GKS for CM is associated with an increased risk of radiation-induced complications, including hemorrhage.⁴² Shih and Pan retrospectively compared surgical resection and GKS of CMs and noted seizure control and complication rates were better with open surgery.⁴⁰ We therefore feel strongly that the decision to treat seizures due to CMs, as well as due to AVMs or tumors, should be guided primarily by the natural history and the risk–benefit ratio of surgical resection in each particular case.

Nonlesional Epilepsy

Currently, SRS does not offer any advantages over surgical resection in the treatment of nonlesional, cortical epilepsies because of the common need to perform invasive monitoring for seizure localization. Corpus callosotomy (CC), however, reduces the frequency or severity of generalized or multifocal seizures independent of seizure focus and may be a target for SRS. Eder et al. treated three children with SRS callosotomy, one with Lennox-Gastaut syndrome, and two following functional hemispherotomy for cortical dysplasia. The two children with persistent seizures after hemispherotomy became seizure free after GKS.⁴³

Preoperative Evaluation

Patient selection and preoperative evaluation are the same as for patients considered for open resective surgery.⁴⁴ A complete medical and neurologic history, as well as a physical examination, is critical for understanding the scope of the patient’s disease and identifying the risk factors that may affect treatment. Patients should be refractory to medical therapy. Patients with significant medical disorders (e.g., cardiac, pulmonary, or oncologic), progressive neurologic disorders (e.g., multiple sclerosis), psychiatric disorders, or a history of substance abuse or noncompliance should be excluded. Women of child-bearing age should have a negative pregnancy test and documented use of a reliable birth control method.

Complete clinical evaluation should include video-electroencephalogram recording sufficient for localization of the seizure focus, neuropsychological testing, and high-resolution (1.5 T) MRI. A thorough endocrine evaluation should be performed prior to treatment of patients with HH. For patients with MTLE, electrographic recordings should indicate the seizures, which may be simple and/or complex partial seizures with or without secondary generalization, arise from a single temporal lobe. Radiographic evidence of mesial temporal sclerosis (hippocampal atrophy or increased T2 signal) should be present in the same temporal lobe. Language and memory dominance should be determined using the intracarotid sodium Amytal test (Wada test) or functional MRI as guided by the clinical situation. A normal MRI, bilateral hippocampal pathology, or multifocal seizure onset is a contraindication for SRS treatment of MTLE.

Surgical Approaches

Radiosurgery can be performed under either general or local anesthesia, though typically we reserve general anesthesia for children. We recommend administering a small dose of a benzodiazepine sedative (e.g., lorazepam) to reduce the risk of seizures during the procedure. Pin sites are prepared with antiseptic solution, and local anesthetic is injected. The stereotactic frame must be affixed to the skull such that the base sits below the target. We then obtain a high-resolution MRI to include sequences that highlight the target (see specifics later). While the practices at individual centers vary, we find consultation among the neurosurgeon, radiation oncologist, and physicist beneficial in treatment planning. Advances in the gamma knife system and planning software (Leksell GammaPlan, Elekta, Stockholm, Sweden) allow more conformal radiation delivery using multiple isocenters. There is no predefined number of isocenters recommended for treatment. Treatment duration ranges from 1 to 2 hours, during which the patient’s blood pressure, oxygen saturation, and neurologic status should be monitored. After treatment, the frame is removed. Most patients can be discharged to home within 2 hours of treatment completion, depending on their baseline neurologic function and the amount of sedatives given during the procedure.

Hypothalamic Hamartomas

T2 and gadolinium-enhanced T1 coronal and axial magnetic resonance images must be carefully studied to delineate lesion boundaries and adjacent critical structures. HHs tend to be hyperintense on T2 and hypointense on T1 as compared to normal gray matter, and they do not enhance.⁴⁵ The planning software can be used to minimize radiation to the optic nerves and tracts, as well as normal hypothalamus, using a beam-blocking approach (Fig. 106-1). For most patients we set the 50% isodose line to be the marginal isodose line, corresponding to the tumor margin. As seizure remission is greater among patients receiving a marginal dose of at least 17 Gy,³⁰ every attempt should be made to achieve this marginal dose while preventing dosing greater than 8 Gy to the optic pathways. We typically use 18 or 19 Gy as our marginal dose.