Pediatric epilepsy: The Indian experience.

Epilepsy is a common clinical entity in neurology clinics. The understanding of the genetics of epilepsy has undergone a sea change prompting re-classification by the International league against epilepsy recently. The prevalence rates of epilepsy in India are similar to those of developed nations. However, the large treatment gap is a major challenge to our public health system. Perinatal injuries are a major causative factor in pediatric group. We have discussed a few common etiologies such as neurocysticercosis and newer genetic epilepsy syndromes. We have also briefly touched upon the Indian experience in pediatric epilepsy surgery.

Introduction

There are very few epidemiological studies looking at the incidence of epilepsy from India. The limited data show that the incidence and prevalence rates are surprisingly similar to those in developed countries. A recent study conducted in Kolkata's urban population showed an annual incidence rate of 27.27 per 100,000 per year.[1] A recent rural epilepsy surveillance program from Uttarakhand showed a prevalence rate of two or more unprovoked seizures to be 7.5 per 1000.[2] This is slightly higher than the prevalence rate in Kerala, a state with higher literacy rates and better public health awareness (4.9/1000).[3] A pediatric study from Kashmir valley shows prevalence rates of 3.74/1000 in males and 3.13/1000 in females.[4]

Etiology

Precise determination of etiology is challenging due to the poor availability of neuroimaging studies and an even lesser access to other investigations (such as genetic studies).

Febrile seizures (FS), head injury, positive family history of epilepsy and developmental delay have been found to be the risk factors for epilepsy in Indian studies.[5] However, there are significant regional variations. Neurocysticercosis (NCC) is highly endemic in certain areas such as northern India. Southern states report a high prevalence of a specific reflex epilepsy called hot water epilepsy.

Febrile Seizures

FS do not constitute epilepsy but are a common problem in pediatric neurology/epilepsy clinics. These are seizures precipitated by fever in the absence of an intracranial infection and preceding afebrile seizures. The incidence rates in India are comparable to those in the developed world. The Yelandur survey estimated the prevalence to be 3.28-5.71/1000[6] whilst the more recent Uttarakhand survey[2] found a prevalence of 2.27 per 1000 population. An EEG is not indicated for simple FS. There is widespread consensus that regular anti-epileptic drugs are not recommended in simple FS. However, prophylactic oral or rectal diazepam or oral clobazam can be used with antipyretics. The risk factors for experiencing subsequent FS are onset of seizures at a younger age, seizures with low-grade fever, complex FS, multiple FS and positive family history of FS or epilepsy. The risk factors for experiencing subsequent epilepsy are pre-existing developmental delay, positive family history of epilepsy and complex FS. The risk from the Yelandur study was estimated to be 1.2%. Some of the febrile seizure associated epilepsy syndromes have recently described to be associated with mutations in sodium channel genes (SCN1A-related spectrum of epilepsies: Please see discussion that follows under the heading “Investigations”).

Neurocysticercosis

NCC is a major risk factor in certain regions of India, with a significant impact on the local prevalence rates of epilepsy. States like Kerala and Jammu and Kashmir report a very low prevalence, whilst northern states like Uttar Pradesh and Bihar report higher prevalence rates.[7] In a recent study from rural Uttarakhand – a newly formed state – NCC related seizures accounted for a significant proportion of the crude prevalence rate of active epilepsy.[8] 24.8% of the patients with active epilepsy had seizures secondary to NCC and 9.9% had remote symptomatic seizures related to calcified granuloma. In a community survey of 50,617 individuals from South India, the prevalence of active epilepsy was 3.83 per 1000 and computed tomography (CT) scan detected NCC in 28.4%.[7]

In children, NCC has been implicated in 0.4% of all neurological complaints.[9] The cysts can be distributed in various patterns – parenchymal, subarachnoid, intraventricular or disseminated. Live cysts and granulomas are “active” lesions, whereas residual calcification and gliosis are “inactive” lesions of NCC. In children, 60-76% of the lesions are single parenchymal granulomas. Calcified lesions are rare as compared to adults (15% vs. 55%). The commonest presentation is new onset partial seizures. Neurological examination is usually normal. Neuroimaging is diagnostic; if magnetic resonance imaging (MRI) is not feasible in the acute situation, a contrast-enhanced CT scan should be done. An important differential diagnosis is tuberculomas; the clinical scenario, location of the granulomas and certain radiological characteristics (including magnetic resonance spectroscopy if required) can help in differentiation. While there is consensus about the effectiveness of cysticidal medications (albendazole) on disappearance of live cysts, there is still a debate on whether it influences the outcome in the granuloma stage of the disease. Caution should be exercised in the use of cysticidal drugs in patients with multiple lesions, as there is risk of cerebral edema due to simultaneous degeneration of multiple lesions. Our practice is to use steroids, albendazole and anti-epileptic medications in children with active lesions. Anti-epileptic medications are given till the lesions become “inactive.” Cases with periodic “reactivation” of the lesion (re-occurrence of peri-lesional edema) and those with remote symptomatic epilepsy due to calcified lesions require long-term anti-epileptic medications. Refractory cases may certainly benefit from epilepsy surgery.

Neonatal Hypoglycemic Brain Injury

This is emerging as a major preventable cause of remote symptomatic childhood epilepsy in India. Neonatal hypoglycemic brain injury (NHBI) was found to be a prominent etiology (23% patients) of remote symptomatic epilepsy with an onset prior to 3 years of age in a cohort of 100 patients referred to our tertiary epilepsy center.[9] Low birth weight, feeding difficulties and low segment caesarian section (LSCS) were identified as risk factors. Spasms were the most common seizure type. Comorbidities included mental retardation, autism, visual impairment and poor hand use.

Classification of Epilepsy Syndromes

The classification as proposed by ILAE has recently been modified. New concepts and terminology have been introduced based on the current understanding. Previous classification of “idiopathic,” “symptomatic” and “cryptogenic” has been replaced by “genetic,” “structural/metabolic” and “unknown,” respectively. This indirectly places a large emphasis on availability of facilities such as neuroimaging, genetic studies, etc. It is no surprise, therefore, that there are very few studies that look at the application of ILAE classification in the Indian subcontinent. A pediatric study from Mumbai (using 1981/1989 classification) reported 55.3% partial, 27% generalized, 13.5% undetermined, and 4.1% specific epilepsy syndromes.[10] This is similar to data from a mixed (adult and pediatric) population of patients, which reported 62.9% partial epilepsies with a staggering majority of 62.7% being symptomatic.[11] Surprisingly, commoner epilepsy syndromes such as childhood and juvenile absence formed a very small proportion (0.4%). The paucity of the commoner pediatric epilepsy syndromes such as benign childhood epilepsy with centro-temporal spikes, etc. is surprising. It is difficult to entirely attribute this to the inherent referral bias. Our own data of syndrome classification in 156 children showed that one-fifth of all cases were benign epilepsies of childhood, while West's syndrome constituted 10%, probably secondary to referral bias (Udani, unpublished data).

Refractory Epilepsy

Refractory epilepsies constitute about 10-20% of childhood epilepsies. In infancy and early childhood, epileptic encephalopathies such as Lennox-Gastaut syndrome (LGS), West's syndrome, Dravet syndrome, etc. tend to be refractory to treatment. In a study of 123 children with “difficult-to-control” epilepsy, onset below 2 years of age, male sex, other neurological abnormalities and certain seizure types emerged as risk factors for refractoriness. Perinatal insults seem to predominate the etiological spectrum.[12] They contributed to about 50% of symptomatic epilepsies with onset in the first 3 years of life in our study.[13] In later childhood and adolescence, symptomatic epilepsies due to an underlying structural cause tend to mainly constitute the refractory group.

Risk factors for refractory epilepsy from Indian studies of adults are similar to international data, that is, partial seizures, presence of neurological deficits, history and radiological evidence of previous CNS insults, FS, high initial seizure frequency of more than one per month and nonresponse to first anti-epileptic drug.[14]

Indeed, inappropriate polytherapy at suboptimal dosages contributes to these high figures. Compliance is also compromised due to high costs of newer anti-epileptic drugs. Some therapies such as adrenocorticotropic hormone (ACTH) for West's syndrome require short-term injections and monitoring of blood pressure, urine dipstick for glucose, etc., adversely affecting their acceptance and use.

Investigations

As mentioned above, appropriate classification of epilepsy requires that the patient is fully investigated with neuroimaging (preferably MRI), genetic studies, neuropsychological evaluation, etc. Poor availability and high costs prove to be limiting factors, especially in rural areas.

Significant advances have been made in understanding the genetics of epilepsy. Epileptic encephalopathies constitute an important clinical group for genetic testing. Dravet syndrome, epilepsy limited to females with mental retardation and specific infantile spasm phenotypes are a few epilepsy syndromes now known to have an identifiable genetic mutation. Needless to add, identification of the causative mutation has significant implications for treatment, prognostication and genetic counseling.

We would like to make a special mention of the spectrum of epilepsies associated with mutations in the alpha subunit of sodium channels (SCN1A). It ranges from simple FS and generalized epilepsy with FS plus (GEFS+) at the mild end to Dravet syndrome (SMEI) and intractable childhood epilepsy with generalized tonic–clonic seizures (ICE-GTC) at the severe end. Less commonly observed phenotypes include myoclonic-astatic epilepsy (MAE or Doose syndrome), LGS, infantile spasms, and vaccine–related encephalopathy and seizures. The phenotype can vary even within the same family. The inheritance is autosomal dominant but most cases of SMEI and ICE-GTC are the result of a de novo heterozygous mutation. Preferred anti-epileptic drugs are clobazam and stiripentol. Topiramate, valproic acid and phenobarbitone are also of some benefit. We have recently studied suspected patients from our clinic and have identified more than 30 children with mutations in this gene. Several new mutations were identified and comparison to controls is underway (Udani, unpublished data).

Treatment

Pharmacological therapy is the mainstay of epilepsy treatment. There is a vast choice of anti-epileptic medications available currently. Though there is no convincing evidence that newer anti-epileptic medications are superior in efficacy, they are certainly better tolerated and have fewer interactions as compared to the older drugs.[15] The higher costs and availability limited to urban areas are important points to consider in the context of our public health system. We would like to emphasize here that this should not prove to be a hindrance in delivering effective treatment in rural areas. The Yelandur study showed that more than half (58-66%) of their patients responded to phenobarbitone and phenytoin, either as mono or polytherapy.[16] Indeed, similar results were reported from a study conducted in children, that is, 65% patients (treated with phenytoin or phenobarbitone) were seizure free at 1-year follow-up.[17] This study mainly looked at behavioral side effects as an outcome measure. It concluded that concerns about phenobarbital-related behavioral side effects may not be valid in developing countries and that this drug is an effective, acceptable anti-epileptic for Indian children, especially in the rural context.

Non-pharmacological treatment options include epilepsy surgery, ketogenic diet and vagal nerve stimulation. These are currently offered at a few tertiary care centers. We would like to discuss briefly the option of epilepsy surgery.

Patients with medically refractory and surgically remediable epilepsy constitute a very important clinical group of patients. We would just like to re-emphasize the point here that pharmacological treatment (i.e. anti-epileptic drugs) is not anti-epileptogenic and is merely symptomatic. However, epilepsy surgery can be essentially curative in carefully selected cases. Pediatric cases amenable to resective surgery are those with cortical malformations, gliotic lesions, developmental tumors, hemispheric pathologies such as hemimegaencephaly, etc. These patients tend to have medically refractory epilepsy often with episodes of prolonged seizures. Continued seizures cause cascade of detrimental neuro-plastic changes that ultimately lead to neuronal cell death. There is some controversy as to whether this is cause or effect. However, what is undisputed is that refractory epilepsy is associated with cognitive decline. Hence, it is crucial that these patients are identified early and referred for detailed pre-surgical evaluation to tertiary centers.

At our center, we analyzed 58 pediatric epilepsy surgeries between 1998 and 2007 (Udani et al., unpublished data). At a mean follow-up period of 3 years (range 1-10 years), all (100%) patients who underwent temporal resection and 55% of patients with extra-temporal resection were rendered seizure free. 64% of patients with hemispheric surgery had a favorable outcome.

Epilepsy surgery can also be palliative such as corpus callosotomy and vagal nerve stimulation. The rationale behind corpus callosotomy is that it will decrease the frequency and severity of primary and secondary generalized seizures. While axial tonic, massive myoclonic, and atonic seizures (drop attacks) and generalized tonic-clonic seizures respond best, partial seizures do not respond and may even worsen. In our study, 44% patients reported a favorable outcome (22% were seizure free and 22% reported improvement). In a case series of 17 pediatric patients who underwent callosotomy, at a median follow-up of 4.7 years, 64.6% patients had a favorable seizure outcome.[18] This study also examined the psychosocial outcome, which is quite impressive. 76.4% parents rated improved seizure frequency, 70.6% parents appreciated improvement in behavior and 64.7% felt that their child was more alert and attentive.

Challenges Ahead

It is an unfortunate truth that despite our recent economic progress, a wide treatment gap is very much a reality. This has been reported to be as high as 78% in the Yelandur study, while the recent Uttarakhand study reports it at 71%. Apart from economic backwardness, socio-cultural beliefs also contribute significantly. People often use two or more systems of healing at the same time. The negative attitudes also adversely affect schooling, marriage, employment, and other social activities. This treatment gap is a real challenge for medical personnel and public health care workers.