Educationals

These must be excluded usually clinically. A detailed history of the actual event, precipitating condition, etc are usually sufficient. Among the important differentials at different ages are conditions like neonatal sleep myoclonus, breath holding spells, migraine variants, syncope and pseudoseizures. The mistakes most often made are relying on EEG and imaging findings to reach the diagnosis. Often these may be coincidentally abnormal or may be wrongly interpreted.
Seizure type

This is the second important diagnostic level helping not only in epileptic syndromic and etiological diagnosis but also in choosing the most appropriate AED. Again the history is most helpful and this should be obtained not only from the caregiver/witness but also the child. Often even very young patients can describe auras or other phenomena if simple questions are asked.

Epileptic syndrome diagnosis

This refers to collating information on seizure type and semiology, age at onset, associated handicaps/symptoms, EEG findings etc leading to a more informed decision making process. This helps in delineating etiology, need for further investigation, needs for treatment, choice of drug, ease of seizure control, length of treatment and evolution of other handicaps. For example benign rolandic epilepsy may not need treatment or investigations beyond an EEG and prognosis is always excellent for complete control and remission. Infantile spasms being a heterogeneous condition needs investigations for etiology and specific AEDs like steroids or Vigabatrin. It may also be difficult to control and may have associated handicaps like autism and mental retardation. It must be emphasized that a syndromic diagnosis may not always be possible especially at the initial visit and a diagnosis may evolve over time.

Etiological diagnosis

An etiological diagnosis must be rigorously attempted in all cases. This may be surmised only from the history as a familial syndrome like childhood absence or febrile seizures or may need sophisticated imaging and biochemical techniques for condition like neurocysticercosis, tuberous sclerosis cortical dysplasia, biotinidase deficiency etc. The etiological classification can then be symptomatic indicating a clear-cut cause or idiopathic which usually means familial. Cryptogenic refers to presumption of a cause but an inability to document it with presently available diagnostic techniques.

Acute symptomatic seizures must be differentiated from unprovoked seizures, as the long-term management would differ in most cases. The former are seizures secondary to acute brain events like trauma or hypoxic, ischemic encephalopathy (HIE), meningoencephalitis, hypocalcemia, hypoglycemia etc. Even febrile seizures and seizures secondary to neurocysticercal granulomas should probably fall into this category. Most neonatal seizures likewise would be acute symptomatic seizures. These seizures do not necessarily lead to chronic epilepsy i.e. recurrent unprovoked seizures and the overall risk of such a long-term outcome is only around 10-15%.

Goals of Therapy -
what are we trying to achieve?
The doctor must have a clear idea of the goals of treatment in different types of cases and this must be clearly explained to the patient at the outset. Unrealistic patient expectations and poor doctor communication is the major reason for `doctor shopping' so prevalent in India.

Seizure control

This is the most obvious goal AED therapy. A balance has to be achieved between seizure control and drug toxicity. Seizure control is important in those epilepsies where prolonged convulsive seizures and status epilepticus are frequent. It is also important in disabling drop attacks causing injury. Subtle and subclinical absence seizures in normal children should also warrant an attempt at maximal control as these interfere with school performance. However brief myoclonias or absences need not be totally controlled in a severely handicapped child where the seizures are just one of the many manifestations of a diffuse neurologic illness.

Ensuring normal development and learning

This aspect is not always in our control; however every attempt should be made to safeguard cognitive and higher functions for a more optimal long-term outcome. Development and learning could be affected the basic neurologic substrate, by the neuropsychologic adverse effects of AEDs, by frequent seizures, status epilepticus and when there are bilateral and/or diffuse epileptiform EEG abnormalities. Examples of the latter are infantile spasms, Lennox-Gastaut syndrome and the epileptic (Landau-Kleffner syndrome, Electrical Status in Slow wave Sleep.).

These so-called `epileptic encephalopathies` may have few obvious or subtle clinical seizures and seizure control alone without normalization of the EEG would not be enough to ensure a good cognitive outcome. Every attempt should be made to choose the most appropriate AED with the neurobehavioral side effects as monotherapy; polytherapy has been regularly shown to increase these problems. Frequent episodes of status can lead to diffuse injury or more focal injury (e.g: hippocampal sclerosis) leading to irreversible cognitive impairment.