Epilepsy that is attributed to a lesion typically develops after a prolonged latent period, although brain organization could be adversely affected much earlier. Discrete, well-localized structural epileptogenic lesions in so-called silent brain areas may cause no perceptible disturbances other than recurrent ictal events, but more extensive lesions are likely to impinge on critical cortical areas and be associated with some degree of
interictal impairment. Consequently, cognitive deficits and neurologic signs and symptoms commonly accompany epileptogenic substrates that tend to be diffuse or multifocal. Disorders such as tuberous sclerosis or cortical dysplasia can be sufficiently mild that only a single localized lesion is identified on magnetic resonance imaging (MRI), but chronic behavioral impairment could suggest, in these cases, the presence of much more widespread aberrations in neuronal migration and differentiation that are beyond the resolution of current imaging techniques. For other lesions, disturbances in cerebral function could reflect distant cerebral abnormalities of a related disorder; for instance, the high incidence of schizophrenia reported to occur in women with hamartomas and gangliogliomas⁶¹ was postulated to be caused by a karyotype abnormality associated with ectodermal dysplasias and abnormal behavior.⁴⁹ Progressive lesions that continue to destroy or disrupt neuronal function are clearly more likely to give rise to interictal behavioral problems than static lesions, the latter of which permit recovery of function through compensatory mechanisms. For the same reason, rapidly progressive disease processes cause more interictal disturbances than slowly progressive ones. It is also conceivable that some neoplastic lesions could secrete neuroactive substances that influence interictal behavior, and some vascular lesions can cause nonepileptic signs or symptoms by bleeding or a steal phenomenon. Lesions that were thought to be static, such as hippocampal sclerosis, are now reported to be common sequelae of stress in its widest sense.⁵⁶ Thus, stress would be an important factor in the epilepsies and in psychiatric disorders.

The neurologic signs and symptoms produced by focal structural lesions depend on the function of the brain area involved. Discrete neocortical lesions in primary areas produce very localized deficits, whereas diffuse bilateral abnormalities cause cognitive impairment and more global dysfunction. Lesions of the limbic system are of particular interest in epileptology because they are the most likely to produce epileptic seizures and because these lesions are associated with much more varied interictal behavioral disturbances than lesions of neocortex.³⁵

The most common deficit associated with lesions of the mesial temporal limbic system is impairment of episodic memory, which is usually material specific for the hemisphere involved.⁴¹ This is more noticeable, and therefore more disabling, when the lesion is in the language-dominant temporal lobe, and there is usually an associated verbal learning disturbance. The limbic system, however, also plays a key role in modulating mood and affect, basic drives (including sexual function), and motivation, functions that, when distorted by structural or functional lesions, could conceivably give rise to virtually every behavioral aberration attributed to epilepsy, including specific psychiatric illnesses.

It is important to note that lesions in certain locations that are epileptogenic might be more likely to cause enduring disturbances referable to the normal function of that area than lesions that are purely ablative.⁶¹ Because there is a certain degree of plasticity of brain tissue, even in the adult, destructive lesions can cause transient deficits that eventually recover as other brain areas take over the function. When the neuronal disruption is intermittent, however, as with epileptic seizures, it is possible that natural compensatory mechanisms are not activated, and recovery of function does not occur. Furthermore, epileptic discharges propagate and can disrupt function at a distance from the primary epileptogenic region, including contralateral structures. Consequently, specific temporal lobe or frontal lobe signs and symptoms that require bilateral homotopic dysfunction would not occur with a nonepileptic destructive lesion in one hemisphere but are commonly seen with unilateral epileptogenic lesions. The existence of secondary bilateral involvement and its reversibility are readily demonstrated following successful anterior temporal lobectomy, when material-specific memory referable to the contralateral temporal lobe subsequently improves and patients may experience an overall increase in IQ.^47,62

Classical neurology has an adult perspective. Its premise is of some fault or accident that disrupts a previously normally functioning central nervous system (CNS). The lesson of epilepsy has been that the faults are often early prenatal or late prenatal or early postnatal, but the problem of interest (the epilepsy or the cognitive capacities or the behavior) becomes “eloquent” at some later stage. Models of this exist in syphilis, encephalitis, rheumatic heart disease, Huntington chorea, and many other conditions. The developmental perspective is to consider the possibility that the lesion or fault could have been “eloquent” in other ways before the condition of interest arose or could become “eloquent” in those ways later in development. This removes a need for thinking that a chronic lesion that arose in embryonal life has somehow acquired new powers at age 14 (or whenever) because “it has started to cause epilepsy.” It allows consideration that the condition of interest (e.g., epilepsy) arises from a constellation of circumstances that have not existed previously, a combination of aging and maturation together with hormonal and psychosocial factors.

Ordinarily, localized destructive cortical lesions that occur early in the course of development are less likely to cause lasting deficits than lesions occurring in later life because of the increased plastic potential of the immature brain. For instance, patients with extensive damage to the left hemisphere incurred within the first few years of life usually have a pathologic compensatory shift of language dominance to the right hemisphere, with little or no disturbance in speech or language.⁴⁸ On the other hand, more diffuse bilateral lesions occurring early in life are usually accompanied by a developmental delay that persists into adulthood. Taylor,⁶² however, challenged functional shift as the only basis of handedness change, because alien tissue lesions are associated with left handedness whether they are in the left or right hemisphere. He suggested that an alien tissue lesion could act by reducing the effect of the gene for right-handedness bias. If this is confirmed in other studies, it would help our understanding of prenatal organization of cerebral functions.

As mentioned in the previous section, epileptogenic lesions may be more likely to produce a lasting effect on behavior than destructive lesions because intermittent disturbances could confound compensatory recovery processes. This appears to be particularly true during the period of cerebral development. Furthermore, the widespread electrical epileptiform electroencephalogram (EEG) abnormalities that characterize certain epileptic syndromes in infancy and early childhood, such as West syndrome, can reflect propagation of discharges from a localized epileptogenic lesion that also produces the same global behavioral disturbances usually associated with a diffuse bilateral destructive lesion. The evidence for this comes from the observation that some infants and small children with developmental delay who initially appear to have catastrophic secondary generalized epilepsy actually have a single lesion localized to one hemisphere, and surgical removal of this area not only eliminates the ictal events, but also helps to reverse the developmental delay.⁵³

One inarguable cause of an inappropriate diagnosis of “interictal” behavioral disturbances, although perhaps rare, is nonconvulsive status epilepticus. Simple partial seizures of limbic origin can cause virtually any psychiatric sign or symptom in clear consciousness, and the scalp EEG in this condition is almost always normal.⁷¹ Sensory symptoms caused by simple partial status epilepticus of neocortical origin can give rise to bizarre experiences that are acted on in strange ways to create a persistent interictal behavioral disturbance. These forms of simple partial status epilepticus (also called aura continua) are easy to recognize when they occur in patients with known epilepsy, and the ongoing signs and symptoms match those of the habitual simple partial seizures (auras) that usually precede more obvious ictal events. In patients who have never had an obvious epileptic seizure and who have a normal EEG, this diagnosis would be exceedingly difficult. Rapid reversal of signs and symptoms with intravenous benzodiazepines suggests, but in no way proves, a diagnosis of epilepsy because many nonepileptic conditions can also transiently respond to this treatment. Complex partial status epilepticus and absence status can also rarely be mistaken for a psychiatric condition when the clouding of consciousness is subtle. EEGs in these conditions, however, reveal the diagnosis, although sometimes only after extreme persistence.²

In patients with known epilepsy and well-documented, frequent interictal EEG abnormalities, the only surface EEG correlate of a simple partial seizure is, classically, disappearance of the interictal EEG spike. For instance, in temporal lobe epilepsy, a simple partial seizure is associated with an ictal discharge in mesial temporal lobe structures, which are then no longer able to generate interictal spikes that ordinarily would propagate to temporal neocortex and appear on the scalp EEG. Consequently, disappearance of the typical interictal EEG spike pattern in a patient with persistent behavioral disturbances could be used as supportive evidence for simple partial status epilepticus. Indeed, one postulated explanation for Landolt’s so-called “forced normalization”³¹ (the occasional observation that the EEGs of some patients with known epilepsy normalize during episodes of psychosis) is that the behavioral disturbance is actually simple partial status epilepticus.⁷²

Another incontestable consequence of epileptic seizures is cell death from the excitotoxic effect of excessive release of excitatory amino acids during status epilepticus.⁵⁹ It is well known that susceptible cerebral structures, particularly the hippocampus, undergo irreversible damage in patients after prolonged convulsive status epilepticus, and studies in experimental animal models show that this occurs even when cerebral perfusion, ventilation, and normothermia are preserved.³⁹ Much is now known about the cellular mechanisms of excitotoxic cell death, but it is unclear whether (and if so, to what extent) this occurs in human epilepsy apart from the severe insult associated with convulsive status epilepticus.⁵⁹ Excitotoxic effects are associated with massive calcium influx, which may be insufficient to cause cells to die but, rather, turns on genetic mechanisms that could profoundly alter cell structure or function, perhaps inducing transsynaptic changes at a distance from the epileptogenic lesion. Such a mechanism could provide a ready substrate for persistent interictal behavioral disturbances.