Evaluation of magnetic resonance imaging-negative drug-resistant epilepsy.

A structural brain lesion in patients with drug-resistant epilepsy (DRE) greatly increases the likelihood of identification of the seizure focus and ultimately seizure-free outcome following resective epilepsy surgery. In contrast, surgical outcomes of true non-lesional DRE are less favorable. Therefore, discovery of an underlying lesion is paramount in the pre-surgical work-up of patients with DRE. Over the years, the surgical treatment of pharmacoresistant epilepsy has evolved from straightforward lesional cases to include cases with hippocampal sclerosis. With the advent of magnetic resonance imaging (MRI), most cases of mesial temporal sclerosis became more easily identifiable on pre-operative neuroimaging. With the widespread use of high-resolution MRI with epilepsy protocols over the last two decades, our ability to visualize subtle structural changes has been greatly enhanced. However, there are some cases of lesional epilepsy, which remain unidentified on these routine MRIs. In such "non-lesional" refractory epilepsies, further investigation with advanced neuroimaging techniques, including metabolic imaging, as well as electrophysiological studies may help to identify the previously non-visualized focal brain abnormalities. In this review, we outline the current status for evaluation of MRI-negative DRE.

Introduction

Drug-resistant epilepsy (DRE) is defined as failure of adequate trials of two (or more) tolerated, appropriately chosen and appropriately used antiepileptic drug regimens (whether administered as monotherapies or in combination) to achieve freedom from seizures.[1] Establishing a diagnosis of DRE is an important milestone in the treatment of epilepsy as it marks the transition of a patient who is taking medications to control a condition and living a relatively normal life to someone who is at risk of worsening seizures, injuries or even death as well as social stigma and economic hardship associated with uncontrolled seizures. Defining the presence of DRE also serves as an important landmark for the treating physician to take stock of the overall situation and reevaluate the diagnosis of epilepsy and treatment plan. It should prompt the physician to evaluate for possible remediable causes of DRE such as inadequate trials of antiseizure medications, use of wrong medications for the type of epilepsy or even misdiagnosis of epilepsy. On the other hand, a firm diagnosis of DRE in an adult may provide a distinct opportunity to discuss other treatment options with the patient because seizure freedom is unlikely with the addition of other antiseizure medications. Other treatment options in cases of DRE include: ketogenic diet, neuromodulation therapy and neurosurgical interventions such as resective epilepsy surgery to provide potential cure from uncontrolled seizures and palliative surgeries to help reduce the severity and frequency of disabling seizures.

Therefore, evaluation of a patient with suspected DRE fundamentally requires three initial steps. First, a detailed history to firmly establish the diagnosis of seizures, help classify epilepsy syndrome and severity of the condition. Information regarding semiology and frequency of seizures, precipitating factors for seizures, past medical history including birth and developmental history and details of data of the antiseizure medications all are important to help achieve this goal. Second, a prolonged video-electroencephalography (EEG) recording should be obtained to capture the electroclinical events to help determine the type and location of the seizure onset in case of focal epilepsy and decide whether surgical treatment would be an appropriate treatment route to pursue. Third, obtain additional investigations to help identify the underlying cause of focal seizures and also help with localization. High-resolution magnetic resonance imaging (MRI) is the ideal structural imaging modality as it provides the finest detail of the brain parenchyma. When MRI is negative in a patient with suspected or established focal epilepsy, further work-up is critical to help localize the epileptogenic zone, especially when surgical resection is being considered.

From a surgical point of view, the primary aim of additional testing in patients with medically refractory epilepsy is to delineate the epileptogenic zone and excise it with minimal or no functional neurological deficit. An epileptogenic zone is defined as “cortex that is necessary and sufficient for initiating seizures and whose removal (or disconnection) is necessary for the complete abolition of seizures”.[2] Thus defining the epileptogenic zone and surrounding eloquent cortex are the two pillars of evaluation in patients with drug resistant focal epilepsy. Overall, the success rate of surgical intervention is best in patients with an identifiable structural lesion that is responsible for their epilepsy. However, identifying the epileptogenic region is more difficult in patients with so-called cryptogenic partial epilepsy where no such structural lesion can be identified on pre-operative neuroimaging. Over the years, improved technology has diminished the pool of such patients and more and more epileptic lesions can now be successfully detected using newer imaging technologies. This field is constantly changing but the investigations to identify the epileptogenic lesion at present can be categorized into:

Structural neuroimaging,

Functional neuroimaging,

Electrophysiological evaluations,

Neuropsychological evaluation and

Laboratory and genetic testing.

Structural Neuroimaging

Though MRI of the brain is much superior to computed tomography (CT) scans, it is generally more complicated to interpret, costlier and less widely available in rural areas. However, different MRI sequences can be obtained to evaluate for a specific question in mind. All MRI scans are not performed with the same diligence and sequences, as such; a “negative MRI” has to be viewed with some degree of suspicion. A routine MRI done with 10 mm thick slices and 15 mm interslice gaps with a T1-weighted and T2-weighted axial image acquisitions and diffusion-weighted images may be adequate to evaluate for a gross cerebral abnormality such a sizable tumor, major malformation of the brain or acute stroke, but may not be adequate to pick up subtle abnormalities that may be responsible for DRE. A routine MRI with standard sequences in axial planes read by a general radiologist who is not very familiar with epilepsy investigations may fail to detect up to half of the focal epileptogenic lesions.[3] This is a common problem and therefore a properly designed MRI protocol looking for subtle structural lesions associated with DRE and review by an experienced physician is essential. Most institutions develop their own MRI protocol to choose appropriate sequences for their local machine and a reasonable time allotted to the study. In our opinion, high-resolution MRI with epilepsy protocol should include at least following sequences:

Spoiled gradient echo (SPGR), magnetization-prepared rapid acquisition with gradient echo (MPRAGE), or similar volumetric T1-weighted images with excellent gray-white differentiation obtained or reconstructed in the coronal plane perpendicular to the long axis of the hippocampus,

Axial and coronal fluid-attenuated inversion recovery (FLAIR) images along the long axis of the hippocampus and

Axial and coronal T2-weighted images along the long axis of the hippocampus. If MRI with administration of contrast was not obtained at least once previously, then pre-and post-contrast T1-weighted images should be obtained in all three planes. In addition, various other strategies can be employed to obtain finer details to evaluate for small or subtle lesions, especially to identify developmental brain lesions.

High resolution, volumetric and 3-D MRI

This can be done on a good quality 1.5T MRI machine. It requires obtaining MRI in thin slices (1-2 mm thick) without a gap between two consecutive slices. This should be performed with T1-weighted sequence and it should be either obtained or reconstructed in the coronal plane perpendicular to the long axis of the hippocampus for review. The T1-weighted images provide the best detailed information on hippocampal anatomy and cortical gyral patterns and a clear contrast between gray and white matter. Volumetric MRI can help identify and quantify the extent of hippocampal atrophy and it may also be useful in identifying other structural anomalies such as polymicrogyria, pachygyria and heterotopias. Measuring volumes of hippocampus or other brain structures by an automated method or with manually assisted computer technology are very sensitive tools and can identify hippocampal atrophy associated with hippocampal sclerosis with good accuracy.[456] Likewise, 3-D and curvilinear reconstruction of diagnostic MRI can help visualize and determine the extent of underlying abnormalities [Figure 1].

Figure 1

3-D curvilinear reconstruction of a patient with a magnetic resonance imaging abnormality suspicious for frontal pachygyria. The extent and pattern of lesion is clearly visible on the reconstructed images

MRI with high-field strength

Newer techniques and technological advances have made it possible to obtain brain MRI with higher strength magnets. Using 3T scanners improves detection rate of a potentially epileptogenic lesion.[789] Currently 3T magnets are routinely used in many centers, but higher strength machines (4T, 7T or even 9T) are being used experimentally under research protocols. The experience with very high strength magnet is limited, but in general higher strength of magnet combined with surface coil technology provides improved resolution of cerebral anatomy and gray-white differentiation. Imaging with higher magnetic field-strength provides improved signal-to-noise ratio (SNR) which in turn allows shorter imaging times for a given resolution and/or higher resolution for a given imaging time. Higher SNR allows better resolution with smaller voxel size and image thicknesses of <1 mm, which has proved especially helpful in imaging of abnormalities of cortical development (e.g. cortical dysplasia).[10]

MRI with special sequences

In addition to the high-field magnets, MR technology utilizing different sequences can further enhance gray-white matter differentiation. MRI is an exquisitely versatile tool because manipulation of multiple different parameters changes tissue imaging characteristics that can be further manipulated to help detect certain types of lesions. For example, the periodically rotated overlapping parallel lines with enhanced reconstruction sequence has been used to image the hippocampus in great detail that can rival histopathological sections at low magnification.[11] It provides excellent contrast between gray and white matter and compensates for subjects moving during the scan, so structural details are preserved. This unique sequence promises to provide further understanding of hippocampal pathology that can be detected preoperatively and thereby help improve identification of epileptogenic zone in patients with DRE. Another novel technique is susceptibility-weighted imaging which uses tissue magnetic susceptibility differences to generate a unique contrast and is exquisitely sensitive to iron and calcium. It provides an excellent tool to identify lesions that contain iron or calcium even in very small amount, such as cavernous malformations, brain tumors, previous hemorrhage due to trauma, lesions associated with Sturge-Weber syndrome [Figure 2].[12]

Figure 2

Axial T2-weighted image (left) of a young woman presenting with complex partial seizures, which is relatively unremarkable. A clear lesion was noted in the right temporal pole on susceptibility-weighted imaging (right). The lesion was histopathologically confirmed to be a cavernous malformation. Patient remains seizure-free following lesionectomy

Re-evaluation of MRI with the epileptogenic focus in mind

The MRI abnormalities in patients with DRE are often subtle. A physician experienced in caring for epilepsy patients is necessary to identify these subtle changes. The treating epileptologist and epilepsy surgeon should of course always review the MRI on a high-resolution monitor and further scrutinize the images in the areas of interest. Additional review of the initial MRI following other investigations is especially helpful [Figure 3].

Figure 3

Fluid-attenuated inversion recovery images of a young woman with a long-standing history of drug-resistant epilepsy. Magnetic resonance imaging (MRI) was initially reported to be normal. Presurgical work-up revealed left temporal seizure onset. Further review of the MRI showed blurring of gray-white matter interface primarily affecting the left temporal lobe (arrows) but also extending into left parietal region (arrowheads)

Hippocampal volume measurement

Hippocampal volume loss and increased T2 signal in hippocampus observed on MRI is associated with hippocampal sclerosis histopathologically. This can be assessed qualitatively by visual inspection and asymmetry >20% that can usually be detected by an experienced observer. However, subtle asymmetry or bilateral abnormalities can often be difficult to identify. Volumetric quantification by acquiring a volumetric sequence and measuring hippocampal volume either by a semi-manual or a fully automated measurement technique can reliably detect hippocampal volume loss with high specificity and sensitivity [Figure 4].[513]

Figure 4

Hippocampal volumetric measurement provided additional data in support of left hippocampal atrophy in this patient presenting with psychomotor seizures

MR spectroscopy

The biochemical content of brain tissue and anatomic distribution of various metabolite signals of specific compounds can be detected non-invasively using the highly sensitive proton magnetic resonance spectroscopy (1H-MRS) technique. Common metabolites used clinically include N-acetyl-L-aspartate (NAA), creatine (Cr), phosphocreatine (PCr) and choline (Cho).[14] Particularly in temporal lobe epilepsy, MRS can effectively lateralize the seizure focus by identifying a reduction of NAA peak, which is usually associated with neuronal dysfunction [Figure 5].[15]

Figure 5

Magnetic resonance spectroscopy showed decreased N-acetyl-L-aspartate peak in the left hippocampus in this patient with normal magnetic resonance imaging. Further electrophysiological studies confirmed left hippocampal onset of seizures

Functional Neuroimaging-Positron Emission Tomography (PET)

Fluorodeoxyglucose PET scan

Fluorodeoxyglucose (FDG)-PET is a test where [18F]-labeled 2-deoxyglucose is administered intravenously. The FDG is transported by the same mechanism as glucose into the brain, but it cannot be utilized as a fuel and is trapped intracellularly. The amount of glucose uptake is proportional to the regional metabolic needs; hence PET imaging of the brain following FDG administration reveals a snapshot of metabolic activity in the brain. In cases of focal epilepsy, the area of seizure onset and at times surrounding areas are hypometabolic interictally and pick-up smaller amounts of FDG compared to surrounding normal brain. The precise mechanism for this reduced metabolic demand is unclear but thought to be due to excess inhibition. Hence, an interictal FDG-PET scan showing an area of hypometabolism (relatively cold area) roughly indicative of epileptogenic zone.[161718]

If the scan is performed during a seizure (ictal PET scan) or in an individual with frequent interictal spiking, it shows increased metabolism (hot spot).[19] Hence, EEG monitoring during FDG infusion is necessary to look for electrographic seizure or frequent spiking.

Correlation of FDG-PET abnormality and final determination of epileptogenic zone is well established in temporal lobe epilepsy, especially in patients where MRIs and EEG monitoring are non-localizing.[20] In patients with temporal lobe epilepsy and negative MRI but positive findings on FDG-PET scan surgical outcome is excellent and similar to patients with hippocampal sclerosis [Figure 6].[21]

Figure 6

A fluorodeoxyglucose (FDG) positron emission tomography (PET) scan (right) of a 41 year-old woman with refractory epilepsy provided findings in support of a diagnosis of right temporal lobe epilepsy. Here the image shows unremarkable magnetic resonance imaging T1-weighted image (left) and Fluid-attenuated inversion recovery sequence (middle) with FDG-PET scan showing right temporal medial and anterior temporal hypometabolism compared to the left side (arrows)

Over the years several investigators have tried to use various ligands to study neurotransmitter or receptor distribution. These are primarily used for research and commonly use a radiolabeled carbon molecule [11C] in place of one of the native carbon molecules in a compound to be studied. One can use ligands to noninvasively image brain chemistry and distribution of various receptors and molecules by labeling a neurotransmitter or its precursor or a ligand with high receptor affinity.

Flumazenil (FMZ) PET scan

FMZ, a gamma-aminobutyric acid B (GABA-B) receptor agonist that can reveal benzodiazepine receptor distribution was one of the early compounds used in patients with epilepsy with good success.[22232425] Role of GABA in neuronal inhibition and seizures is well known and extensively studied. Therefore, evaluating GABA receptor distribution in patients with epilepsy, especially focal epilepsy was quite natural. The FMZ-PET images show decreased FMZ binding in the epileptogenic area. FMZ scans often show more restricted abnormality when compared to FDG-PET scan and is better aligned to the epileptogenic zone [Figure 7].[2627]

Figure 7

Fused images with overlay of fluorodeoxyglucose-positron emission tomography (PET) scan (left), flumazenil (FMZ)-PET scan (middle), and FMZ-PET plus electrophysiological data (right) on 3-D magnetic resonance imaging reconstruction. The red areas correspond to regions of decreased tracer uptake or binding (≥10% compared to the contralateral area). Note that the volume of FMZ-PET abnormality is smaller and is more concordant with the seizure onset zone determined by intracranial recording (each small circle superimposed on the brain surface represents a subdural grid electrode)

Alpha-11C-methyl-L-tryptophan (AMT) PET scan

AMT is an amino acid PET tracer which is not incorporated into proteins; instead, tryptophan (and also AMT) is transported in brain tissue via the large neutral amino acid transporter and can be metabolized via the kynurenine or serotonin pathways. In patients with focal epilepsy due to certain causes, such as tuberous sclerosis, cortical dysplasia and certain epileptogenic brain tumors, AMT-PET scan typically shows increased uptake in the epileptogenic area [Figure 8].[28293031]

Figure 8

Axial T1-weighted magnetic resonance imaging showing subtle cortical thickening in the left middle frontal gyrus in this patient presenting with new-onset focal motor seizures. Alpha-methyl-tryptophan-positron emission tomography imaging revealed intense uptake in the involved region, which was histopathologically-confirmed to be focal cortical dysplasia

18F-trans-4-fluoro-N-(2-[4-(2-methoxyphenyl)piperazin-1-yl)ethyl]-N-(2-pyridyl) cyclohexane carboxamide (FCWAY) PET scan

FCWAY is a selective 5-HT1A-receptor antagonist that has been studied in patients with temporal lobe epilepsy. 5-HT1A-receptor binding is reduced ipsilateral to the temporal lobe seizure focus, hence the PET scan using this ligand shows decreased binding on the side of seizure onset.[32]

PET scan using other tracers such as 11C-diprenorphine, 18F-cyclofoxy and 11C-carfentanil to investigate binding to opiate receptors; and 11C-deprenyl to measure monoamine oxidase B activity are performed at various institutes as part of ongoing research programs.[333435] Many of them show promising results but have not shown to surpass the clinical utility of FDG-PET scan. Therefore, they are not routinely used clinically, but remain a great tool to study neurochemistry in various disease states including epilepsy noninvasively.

Functional Neuroimaging-Single-Photon Emission Computed Tomography (SPECT)

SPECT scan is a tomographic imaging technique using gamma rays and is performed acquiring multiple 2-D images from multiple angles. A computer reconstructs a tomographic 3-D dataset that can be viewed as thin slices in desired plane. It measures gamma rays emitted by the tracer directly, whereas PET scanning uses positron emission. Tracers used for SPECT scan are more stable and making it much cheaper to obtain compared with PET scan but does not provide as good of resolution when compared with PET scan. For brain SPECT scan, a commonly used tracer is technetium-99m-hexamethylpropylene amine oxime (99mTc-HMPAO) where 99mTc is attached to HMPAO.[36] When injected in blood, this compound is taken up by brain in the first-pass and is in proportion to blood flow at the time. The brain can be imaged using scanning camera at a later time and will generate an image of the brain blood flow at the time of injection. Hence, if focal increase or decrease in blood flow is present at the time of injection one can see it as “hot” or “cold” when compared to the surrounding brain.[37] The main downside is that its resolution is poor compared with PET scan and other structural imaging techniques such as CT or MRI. Thus brain images are very grainy and small focal change in tracer activity is not recognizable.

Ictal and interictal SPECT scan

Seizures are metabolically demanding events and blood flow increases at the time of seizure to satisfy the increased local metabolic need. In case of a focal seizure, blood flow increase will only be in the brain tissue involved in seizure activity. Hence, injection of the tracer at the time of seizure (ictal scan) will show increased blood flow in the area of seizure and can be seen as “hot spot.” Interictally, epileptogenic area is metabolically less active and displays as a “cold spot.” Thus, ictal and interictal SPECT scan be used to image the seizure focus.[38]

The difficulty with ictal SPECT is that it depends upon the time of injection to give a snapshot of brain blood flow at that moment. Focal seizure is a dynamic event and it can start at one location and quickly or gradually involve other areas as it spreads. At some point during a seizure, the electrical activity may abate where it originated and become more active at another location. Hence, an injection later in the seizure will provide a snapshot of seizure spread and will not capture the origin of it. This can be very misleading depending upon the speed and pattern of seizure spread as well as timing of the injection. Ictal scan can best be obtained in a video-EEG monitoring unit where trained personnel can monitor the patient for a seizure clinically as well as electrographically and quickly inject the radiotracer. However, this requires storage of the radiotracer in the video-EEG room or nearby and trained knowledgeable personnel available at bedside 24/7 to recognize the seizure very quickly and promptly inject the tracer. This makes this test very expensive.

Subtraction Ictal Single-Photon Emission Computed Tomography Co-registered to Magnetic Resonance Imaging (SISCOM)

A special technique that takes advantage of the difference between ictal and interictal SPECT scan by subtracting interictal images from the ictal image and then co-registering the data with patient's own structural MRI. Theoretically speaking, the interictal scan should show decrease blood flow in the epileptogenic area and ictal scan should show increased blood flow precisely in the same location during a seizure. Subtracting enhances this difference as all other data should be eliminated as blood flow to other parts of the brain should have remained relatively similar between the ictal and interictal images. This technique is called SISCOM. It is proven to be quite sensitive and specific in identifying seizure focus that can lead to successful surgical outcome.[394041] This technique can be very useful in patients with MRI-negative DRE [Figure 9].

Figure 9

Patient with drug-resistant epilepsy with negative magnetic resonance imaging and non-lateralizing electroencephalography. Likewise, fluorodeoxyglucose-positron emission tomography scan showed bilateral areas of hypometabolism. Patient underwent subtraction ictal single-photon emission computed tomography scan with co-registration with the diagnostic magnetic resonance imaging which showed clear right temporal neocortical increased blood flow at onset of the seizure. This was confirmed with intracranial monitoring

Functional Magnetic Resonance Imaging (fMRI) for Defining Eloquent Cortex

fMRI is a noninvasive technique to image task-related increased blood flow in a specific brain region. The technique commonly used is blood-oxygen-level-dependent (BOLD) contrast, which utilizes increased regional blood flow in response to metabolic demand associated with a task, e.g., finger tapping will increase blood flow in contralateral motor hand area. Individual movements lead to a very small change that is below detection sensitivity, however, if the scan is repeated multiple times when task is being performed, the task-related change can be detected. Normally, the BOLD scan obtained at the resting level is subtracted from the values obtained during the task to enhance visualization of the cortex involved in processing of the particular task. Thus, fMRI can be utilized for various brain regions by obtaining task related scans for different modality. Commonly used tasks include motor tasks to locate the motor homunculus, various language tasks to locate Broca's and Wernicke's areas. The findings are only as good as the task is, e.g., motor tasks are simple but language paradigm used to identify Broca's area is good only if the tasks provided require activation of only that area. This is difficult as paradigms used to check motor language processing may simultaneously activate primary auditory cortex and Wernicke's area when instructions are given and the subject is listening and interpreting it as well as it will activate the phonation mechanism used to produce the sound.

However, with increasing experience, robust validated protocols have been established to test for commonly used tasks to locate the language areas as well as motor, visual and auditory cortex. The non-invasive nature of the technique and its co-registration with structural MRI gives a very precise understanding of relationship of a lesion is present or epileptogenic zone if defined prior to surgery. In addition, neuronavigation methods can identify the location of eloquent cortex in the operating room if fMRI information is available and co-registered with structural imaging. This leads to precise presurgical planning of resection margins and meaningful discussion regarding potential benefits versus surgical risks can be held with the patient and family members prior to epilepsy surgery.

Combining fMRI and EEG in the study of interictal epileptiform discharges

The fMRI technology can be used to image interictal spike origin. Spike triggered averaging of the BOLD signal can image blood flow changes associated with spike and reveal the origin of the spike.[4243] Similar technique can be used in seizures but it is difficult to capture a seizure while fMRI is being performed, so it is limited to study interictal spikes.

However, the location of the seizure onset and interictal spike origin may not be completely congruent and always requires some assumption. It also remains difficult to use if interictal spike are of multifocal origin, but the seizures arise from a single focus.

Electrophysiological Evaluation

Magnetoencephalogram (MEG)

MEG is a technique that studies magnetic field associated with interictal spikes instead of the commonly studied electrical field by EEG. The main advantage of the technique is that the magnetic field does not attenuate due to intervening tissue (meninges, skull and scalp) as EEG signal does. Recording magnetic field with high density array over the scalp one can determine the origin of the magnetic source of the interictal spike in 3D space.[44] If the MEG information is co-registered with MRI one can view the origin of the spike on brain structural imaging again in 3D space [Figure 10]. Again the problem of assumption that there is a single, identical source for interictal spike and seizure remains as MEG also studies interictal phenomenon.[4546]

Figure 10

Magnetoencephalogram in a patient with intractable epilepsy with normal magnetic resonance imaging and non-localizing scalp electroencephalography, interictal fluorodeoxyglucose-positron emission tomography, ictal single-photon emission computed tomography. Note the interictal spike detection and source localization in the left temporal lobe

Dense array EEG and EEG source localization

Recent advances in computer processing of the large amount of data from EEG signal has led to the creation of dense array EEG recording and from that information to plot electrical source of that activity. This is similar to MEG but here traditional EEG signal is used.[4748]

Neuropsychological Evaluation

When a standardized series of neuropsychological testing is administered by an individual well versed in the evaluation of patients with DRE may detect subtle impairments of neurocognitive functioning. As different brain regions specialize in different cognitive function, these differences can point to the epileptogenic focus as it will interfere with normal function of that region. This provides indirect but additional evidence for localization of seizure onset. When neuropsychological findings are in agreement with the clinical, electrophysiological and functional neuroimaging data in patients with MRI-negative DRE, decision to surgically intervene is supported. Conversely, when there is incongruence between the neuropsychological findings and other data, or there is suspicion of multifocal or generalized dysfunction, further investigations is usually required to determine whether the patient has a surgically-remediable epileptic focus.

Laboratory and Genetic Testing

Laboratory testing

In patients with DRE, drug compliance may be a factor and monitoring antiseizure medication levels help to establish compliance and adjust dosing as necessary. It can also help to monitor for potentially serious side-effects of these medications and may help to make a case for surgical treatment if they occur.

It has been increasingly recognized that some patients with intractable focal epilepsy may have underlying immunological disorder with antibodies directed against various central nervous system targets. It may be prudent to looks for such derangements in appropriate clinical setting. Screening for antibodies such as anti-nuclear antibody, anti-deoxyribonucleic acid, extractable nuclear antigen screen, anti-thyroid antibody, anti-glutamic acid decarboxylase, anti-N-methyl-D-aspartate receptor antibody, anti-voltage-gated potassium channel, etc. can be helpful. If such disorder is established, treatment with immunological therapy such as pulse or long-term corticosteroids, intravenous immunoglobulins, plasma exchange or even immunosuppressive therapy with cyclophosphamide, rituximab, etc. have shown to provide benefit in some instances.[49505152]

Genetic testing

Increasingly various genetic underpinning of focal epilepsy is being understood. Most of the single gene or chromosomal disorders where DRE is part of the phenotype present in childhood and seen by pediatric neurologists or geneticists. Some syndromes may present in adulthood or continue into adulthood and may present as MRI-negative DRE. In such a patient characteristic presentation of a known genetic syndrome or a strong family history, genetic testing might help to establish proper diagnosis and end search for a cause. More importantly, it may also prevent unnecessary and unsuccessful surgical intervention. Some of the genetic causes of MRI-negative DRE include autosomal dominant partial epilepsy with auditory features due to mutation in LGI1 gene, autosomal dominant nocturnal frontal lobe epilepsy with various mutations in nicotinic acetylcholine receptor α and β subunits CHRNA4, CHRNB2 and CHRNA2, focal epilepsy with variable foci with various mutations in DEPDC5 gene, reflex epilepsies such as hot water epilepsy, primary reading epilepsy, idiopathic photosensitive occipital lobe epilepsy, etc. Generalized epilepsy with febrile seizures plus syndrome with mutation in various sodium channel genes SCN1A, SCN1B, SCN2A or GABA receptor (GABRG2) genes is being increasingly recognized syndrome in children but semiology is variable and remains not completely understood story.[53545556]

Conclusion

As suggested above, there are varied reasons for MRI-negative DRE. The primary aim of the evaluation in such individual is to establish epileptogenic zone and determine if surgical remedy is feasible or not and at what cost. General guiding principles are to establish that stereotypical focal seizures are the cause of habitual events. The second step is to determine the location of seizure focus. Focal seizures lead to various functional impairments even in the absence of clearly visible structural abnormality on MRI. Such can be determined by functional neuroimaging, electrophysiological and neuropsychological testing. The more concordance between the various modalities of tests, the higher the confidence in location of seizure onset and higher the success rate of surgical treatment. If there are some discordant results, one may need to evaluate the strength of such discordance and preponderance of the evidence before making a judgment to proceed with surgical resection. One also has to decide whether one-stage surgery (usually means a temporal lobe resection in a patient with MRI-negative DRE) or there is a need for two-staged approach to define the epileptogenic zone more precisely. Some of the other tests help to establish the location of the eloquent cerebral cortex in relationship to the epileptogenic zone, which in turn helps to determine resection margins and risk of post-operative deficits and/or seizure recurrence.