A clinical study to identify the possible etiology of complex partial seizures using magnetic resonance imaging brain findings and its implications on treatment.

CONTEXT: Epilepsy is one of the common neurological disorders worldwide. Fundamentally, there are two types of epilepsy-primary generalized epilepsy and localization-related epilepsy. Partial seizures account for about 40% of childhood seizures in some series and can be classified as simple or complex.[1] Partial seizures, more so the complex partial seizures (CPSs), are presumed to have a structural etiology. AIMS: (1) To study the magnetic resonance imaging (MRI) brain findings in CPSs in children aged 1-18 years. (2) To identify treatable causes of CPSs based on MRI findings and institute appropriate treatment. STATISTICAL ANALYSIS: Statistical analysis was performed using percentages and proportions.
METHODS: Hospital based prospective study in which MRI brain was done on all newly diagnosed children with complex partial seizures, aged 1 to 18 yrs, during the study period. Final diagnosis was made correlating clinical features, radiological features and other supportive evidences, and appropriate treatment instituted. Follow up of cases was done until the completion of treatment (maximum 6 months).
RESULTS: Among the 64 children who were clinically diagnosed to have CPSs and subjected to MRI study of the brain, 40(62.5%) children were detected to have structural lesions, of which neurocysticercosis (NCC) was noted in 17 (42.5%), tuberculoma in 12 (30%), hippocampal sclerosis (HS) in 6 (15%), gliosis in 4 (10%), and tumor in 1 (2.5%) patient. Sixty-two (96.8%) children were treated medically, and 2 (3.2%) children underwent surgery.
CONCLUSIONS: Etiology of CPS based on MRI findings showed a substantial number of medically- and surgically-treatable pathologies. This study done on South Indian children showed neuro infections to be the most common cause of CPS, followed by HS, with NCC being the most common lesion noted. MRI not only identifies specific epileptogenic substrates, but also determines the specific treatment and predicts prognosis and should be the imaging modality of choice in the evaluation of CPS.

Introduction

Primary generalized epilepsy is considered to be genetic in etiology, whereas most localization-related epilepsy is presumed to be the result of a cerebral insult, even though the insult cannot be determined in about half of all epileptic patients, regardless of age.[1]

Localization-related epilepsy is said to be more common in developing countries. Imaging has led to identification of the cause of complex partial seizures (CPSs) in more children, which determines appropriate treatment (medical or surgical) in a given child with CPS.[2]

Magnetic resonance imaging (MRI) gives precise localization and histological nature of lesions and subsequently, this is of immense help to both clinicians as well as neurosurgeons in their attempt to achieve a faster and more accurate method of discovering the nature of the pathologies.[3]

Dedicated studies in pediatric population using MRI brain have been few till date in India.

The following study has been undertaken to study the etiology of CPS based on MRI findings and manage accordingly.

Methods

Source of data

All children newly diagnosed with CPSs, aged 1–18 years attending inpatient and outpatient sections of the Department of Pediatrics of our hospital.

Study design

Hospital-based prospective study

Study scans were done with 1.5 Tesla MRI scanner

Based on clinical features, MRI findings, and other supportive evidences, final diagnosis was made and appropriate treatment given.

Follow-up of cases was done until the completion of treatment or for a minimum of 6 months.

Study period

October 2011 to March 2013.

Inclusion criteria

All children newly diagnosed with CPSs, aged 1–18 years, attending inpatient and outpatient sections of the Department of Pediatrics, with available consent were included in the study.

Exclusion criteria

Children with developmental delay, cerebral palsy, and mental retardation

Children with seizures following head injury.

Method of collection of data

Methods

A detailed clinical evaluation of children was done.

The following investigations were carried out in all children:

Hemoglobin, total counts, differential count, ESR, urine routine examination

Blood urea, random blood glucose, serum creatinine, serum calcium

MRI brain.

The images of MRI were T1- weighted, T2-weighted, and subjected to fluid attenuated inversion recovery, short T1 inversion recovery, functional MRI, and the hippocampal sequence along with contrast wherever indicated. Both coronal and sagittal sections were taken to increase the yield of study. All children were given antiepileptic drugs (AEDs). Other specific treatments given were based on the MRI findings and other supportive investigations.

The following investigations were done wherever indicated:

Chest X-ray—posterior-anterior view

Mantoux test

Sputum examination for acid fast Bacilli

Cerebrospinal fluid analysis.

Electroencephalogram (EEG) was done wherever possible.

Results

The most common age group found to be affected by CPS in our study was the 10–15-year age group.

Among the 64 children diagnosed to have CPS in our study, 40 (62.5%) had abnormal MRI study.

Ring-enhancing lesions (RELs) formed the most common MRI abnormality (n = 27, 67.5%).

EEG was done in 26 of the 64 children with CPS. EEG was abnormal in 10 (38.4%) of these children.

Two cases (5%) were managed surgically and with AEDs—one was a case of hippocampal sclerosis (HS) and the other was glioma. Thirty-eight (98%) cases were managed medically.

Discussion

Sixty-four children clinically diagnosed with CPSs during the study period of 18 months and fulfilling the study criteria were included and subjected to MRI study of the brain.

Majority of children were in the age group of 10–15 years (46.8%), [Table 1] and majority (65.6%) of them were male children.

Table 1

Age wise distribution of children with complex partial seizures

MRI study revealed abnormalities in 40 (62.5%), [Table 2] children. This is in contrast to the study by Parihar et al.,[4] which showed abnormalities on MRI in 39 (90.7%) of the 43 pediatric patients included in their study. This high positivity obtained in their study was because infants (28 days to 1 year) and children with delayed development were included. Amirsalari et al.[5] and King et al.[6] had found abnormalities only in 28.5% and 17% of their subjects, respectively. This lower rate of detection could probably be attributed to the much lesser incidence of infectious diseases in these parts of the world.

Table 2

Total magnetic resonance imaging abnormality to normal magnetic resonance imaging

In the present study, among the group of 40 (62.5%) children with abnormal MRI findings, neurocysticercosis (NCC) was the most common (n = 17, 42.5%) lesion, [Table 3] followed by tuberculoma (n = 12, 30%), HS (n = 6, 15%), gliosis (n = 4, 10%), and glioma (n = 1, 2.5%).

Table 3

Etiological distribution of lesions

Parihar et al.,[4] in their study, also found NCC to be the major abnormality (n = 24, 55.8%). In another study by Swingler et al.,[7] which used computed tomography (CT) of the brain to evaluate children with CPS in Cape Town, an area with high prevalence of tuberculosis and cysticercosis like our country, also found neuro-infections to be the most common lesion.

NCC [Figure 1] is highly prevalent in this part of the world and hence it was no surprise that it was identified as the most common of MRI lesions. Similar to Patil's study, in our study, single lesions were more common than multiple lesions.[8] Three of these NCCs identified were in the vesicular/colloidal stage (one cyst with scolex, two cysts in the colloidal stage) and the other cysts were in the granular nodular and calcified stages. Anti-parasitic therapy is indicated in patients with active lesions of NCC[9] and this ability of MRI to categorize them aids in taking therapeutic decisions. Clinical and radiological response to anti-parasitic therapy was quite satisfactory in these children, whereas others were treated with AEDs only, as per standard protocols.

Figure 1

Neurocysticercosis left - T1–weighted spin echo gadolinium (GAD) magnetic resonance imaging showing a small single ring enhancing lesion in the right temporal lobe suggestive of neurocysticercosis middle – flair image of the same right – posttreatment flair image showing resolution of the lesion

Tuberculoma [Figure 2] was identified in 12 (30%) study subjects. All of these children had positive lipid lactate peak on magnetic resonance spectroscopy (MRS) and positive Mantoux test. Six of these children had h/o recent contact with pulmonary tuberculosis, and these children received anti-tubercular therapy (ATT) along with AEDs. The lesions in these children showed resolution in the repeat imaging done after 12 weeks, as recommended by Bagga et al.[10] in their article on tuberculomas. Abdul-Ghaffar et al.[11] had also noted remarkable improvement of lesions within 6 weeks and complete resolution of the lesions within 12 weeks of the start of ATT.

Figure 2

Tuberculoma. Above left-” T1-weighted spin echo gadolinium (GAD) axial magnetic resonance imaging showing small conglomerate ring enhancing lesion in the postero-superior left temporal lobe with significant perilesional edema. Above right- Resolution of the lesion after treatment. Below- MRS showing elevated lipid lactate peak suggestive of tuberculoma

HS forms an important etiology of CPS, for which MRI is the diagnostic modality.[12] The importance of detecting HS lies in its amenability to surgical therapy and hence control of seizures. In our study, six (15%) children were found to have HS, with bilateral involvement in one child. One of these children [Figure 3] had intractable seizures and developed deficits in verbal and visual memory. He was referred for surgical correction and after presurgical evaluation such as video EEG, neurophysiological testing, and positron emission tomography, he underwent right-sided anterior temporal lobectomy and amygdalohippocampectomy. He is now asymptomatic post-surgery and is on tapering therapy of AEDs. The symptoms of the other five children with HS are well-controlled with AEDs alone.

Figure 3

Hippocampal sclerosis The coronal T1-weighted (left), fluid attenuated inversion recovery (right) images showing right-sided mesial temporal sclerosis. Notice the volume loss, which indicates atrophy and causes secondary enlargement of the temporal horn of the lateral ventricle. The high signal in the hippocampus reflects gliosis. Coronal T1-weighted and fluid attenuated inversion recovery images are the most sensitive for detecting mesial temporal sclerosis

One child showed a partial REL on MRI brain, which had choline peak on MRS suggestive of low-grade glioma in the left temporoparietal lobe. He had no neurological deficits at presentation. In the following year, during which he was being followed up and evaluated for surgery, for the tumor resectability and prognosis, he developed epilepsia partialis continua. He then underwent resection of the lesion, but his condition deteriorated despite the surgery, and he succumbed to the illness.

Gliosis, which represents the end result of trauma, infection, and infarction,[7] was found in four (10%) children. This group responded well to AEDs.

Interestingly, 29 (72.5%) of the abnormal MRI scans showed the presence of RELs (single or multiple). This phenomenon of RELs (single or multiple) is unique to our country and has not been reported in Western studies, as both tuberculosis and cysticercosis are endemic in our region. Our study found a higher percentage of REL (67.5%) compared to Wadia et al.'s[13] study on focal epilepsy in North India using CT brain.

Cortical dysplasias/malformations known to be commonly associated with CPS[1] were not found in our study probably because we had excluded children with delayed development and mental retardation.

In our study, 18 (45%) children had right-sided lesions, 17 (42.5%) left-sided lesions, and five (12.5%) bilateral lesions. Frontal lobe involvement was seen in 16 (40%), temporal lobe in eight (20%), and parietal lobe in four (10%) children. One lesion was found to be in the thalamus and one in the midbrain. Parihar et al.[4] found parietal lobe involvement in 19 (44.19%), followed by temporal lobe involvement in six (13.95%). Four of our study subjects had multiple lesions, which was similar to the study by Nair et al.[14]

Two cases were managed surgically along with AEDs in our study out of the 40 MRI-diagnosed structural lesions. One was a case of HS and the other was that of glioma 38 cases were managed medically. This was in contrast to the Indian studies by Rasool et al.,[15] Parihar et al.,[4] and Sharma et al.,[16] whose studies found no cases requiring surgical management.

EEG was carried out in 26 of the 64 children with CPS. EEG was abnormal in 10 of these children (38.4%). Six children had abnormal MRI findings amongst the group of 16 children, with no abnormality in EEG. EEG was abnormal in three (23%) of the children with normal MRI study. A higher percentage of this unexpected difference that the MRI detecting epileptogenic focus in normal EEG subjects, comparable to our study, was also found in Doescher et al.'s[17] study in Minnesota on 181 children, which found 42% abnormality on MRI study in subjects with normal EEG. Our findings indicate that a normal EEG does not reliably predict the absence of epileptogenic focus in CPS and suggests that normal results on EEG should not be used to place a patient in a low-risk group of not requiring MRI scan for complete evaluation [Table 4].

Table 4

Electroencephalographic correlation with magnetic resonance imaging scans

Conclusion

Accurate diagnosis of the cause of seizures is crucial for finding an effective treatment. With its high spatial resolution, excellent inherent soft tissue contrast, multiplanar imaging capability, and lack of ionizing radiation,[3] MRI has emerged as a versatile tool in the evaluation of patients with central nervous system disorders. MRI not only identifies specific epileptogenic substrates, but also determines specific treatment, predicts prognosis, as well as helps in limiting the use of AEDs for prolonged periods of time. MRI scan is likely to identify abnormalities in majority of children with CPS and hence, it should be the primary modality of investigation in evaluating children with CPS. The factors which may prevent it from becoming the primary modality of investigation could be its high cost and lack of universal availability.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest