Temporal lobe epilepsy (TLE) is a form of epilepsy in which recurrent seizures originate within networks of the temporal lobe of the brain. It can affect structures deep within the lobe, particularly the hippocampus and amygdala, or the temporal neocortex. Typical seizures may begin with an unusual sensation, emotion, or memory experience, followed by impaired awareness, behavioral arrest, or repetitive movements. Some spread to produce bilateral tonic–clonic convulsions.
TLE is one of the most extensively studied focal epilepsies and a major indication for epilepsy surgery. Its importance extends beyond seizures: temporal brain networks contribute to memory, language, emotion, and social functioning, and both the underlying disorder and its treatment can affect these capacities. TLE is not a single disease with one cause or course. It includes conditions associated with hippocampal scarring, developmental abnormalities, tumors, immune-mediated disease, genetic susceptibility, and cases without an identifiable structural lesion.
Definition, classification, and occurrence
A diagnosis of epilepsy generally requires at least two unprovoked seizures more than 24 hours apart, one unprovoked seizure with a sufficiently high predicted recurrence risk, or recognition of an epilepsy syndrome. The International League Against Epilepsy’s 2014 practical definition specified a recurrence probability of at least 60% over the following ten years for the second of these situations. An isolated seizure during an acute metabolic disturbance or brain infection does not, by itself, establish chronic TLE. The distinction concerns whether a persistent predisposition to seizures exists.
TLE belongs to the focal epilepsies, in which seizures begin in networks limited to one cerebral hemisphere. Its principal anatomical division is between mesial temporal lobe epilepsy, involving medial structures such as the hippocampus, and lateral temporal lobe epilepsy, involving temporal neocortex. Some patients have involvement of both, independent seizure onsets in the two temporal lobes, or a broader “temporal-plus” network extending into neighboring regions. These categories describe clinically useful patterns rather than perfectly bounded biological entities.
Seizure type and epilepsy type are separate levels of classification. A person with TLE can experience focal seizures with preserved awareness, seizures with impaired awareness, and focal-to-bilateral tonic–clonic seizures. Older records frequently use “simple partial,” “complex partial,” and “secondarily generalized” seizures. These terms broadly correspond to the later awareness-based categories, although classifications and terminology have continued to evolve. The expression “psychomotor epilepsy,” historically associated with temporal seizures, is now mainly of historical interest.
There is no single dependable worldwide estimate of the proportion of epilepsy attributable to TLE. Population studies differ in age distribution, access to imaging and electroencephalography, and the criteria used to assign a lobar origin. TLE is particularly prominent in adult epilepsy clinics and surgical series, but such samples overrepresent persistent and drug-resistant disease. Their figures cannot be treated as population prevalence estimates. Childhood temporal epilepsy may also be underrecognized because symptoms are less easily described and seizure behavior can be less localizing.
Onset is possible from infancy to late life. Developmental lesions and certain tumors may present during childhood, whereas vascular injury, tumors, or immune-mediated disorders can produce later-onset disease. A history of childhood seizures followed by a period of relative remission and subsequent recurrence occurs in some patients with mesial TLE, but it is not a required or universal course.
Temporal anatomy and seizure networks
The temporal lobes lie along the sides of the cerebral hemispheres, below the lateral fissures. Their outer cerebral cortex includes regions involved in auditory processing, language, and recognition of objects and faces. On the medial surface lie the hippocampus, entorhinal and parahippocampal cortices, and the amygdala. These structures interact with one another and with frontal, parietal, subcortical, and brainstem systems. Consequently, a temporal seizure is a disturbance of a connected network, not simply an electrical event confined to a single small structure.
The hippocampus supports the formation and organization of episodic memories. Its internal circuitry includes dentate granule cells and pyramidal neurons in the cornu ammonis fields, conventionally designated CA1 through CA4. Information reaches hippocampal circuitry through the entorhinal cortex and returns through several interconnected pathways. These circuits normally regulate the timing, strength, and selectivity of neural activity. Their organization also provides pathways through which abnormal activity can synchronize and recur.
The amygdala participates in emotional processing and autonomic responses. Its involvement helps explain why a temporal seizure may produce sudden fear, changes in heart rate, or a visceral sensation before awareness becomes impaired. Temporal association cortex contributes to the interpretation of complex sensory experience. Seizures involving these regions may cause auditory distortions, altered familiarity, or experiential phenomena rather than an obvious motor disturbance. Similar symptoms can nevertheless arise when seizures beginning elsewhere spread into temporal networks.
Connections help account for the evolution of a seizure. Activity starting in medial temporal structures may reach lateral cortex, the opposite temporal lobe, or frontal and subcortical circuits. Impairment of awareness is associated with distributed network dysfunction, including disruption of systems that maintain alertness and integration of information. Bilateral convulsions require further recruitment of motor networks; they do not mean that the original epilepsy was generalized from onset.
Functional asymmetry is important but not absolute. In most people, language is predominantly represented in the left hemisphere, and left temporal structures often contribute strongly to verbal learning. Right temporal structures contribute to aspects of nonverbal memory and spatial or visual processing. Individual organization varies with handedness, development, early injury, and epilepsy-related reorganization. Thus, “left equals language” and “right equals visual memory” are useful approximations, not sufficient descriptions for surgical planning.
Causes and pathological substrates
A major structural association of mesial TLE is hippocampal sclerosis, characterized by neuronal loss and gliosis within the hippocampal formation. The pattern of neuronal loss varies. An International League Against Epilepsy classification published in 2013 distinguishes major histological patterns according to the relative involvement of hippocampal subfields. On imaging, sclerosis commonly appears as hippocampal atrophy, abnormal internal architecture, and increased signal on fluid-sensitive sequences. Microscopic sclerosis and an imaging diagnosis are related but not interchangeable findings.
The relationship between hippocampal sclerosis and seizures is complex. Prolonged seizures can injure vulnerable hippocampal tissue, while an injured and reorganized hippocampus can promote recurrent seizures. Some patients report prolonged febrile seizures early in life, and prospective studies have documented acute hippocampal abnormalities after febrile status epilepticus in a subset of children. However, most children with febrile seizures do not develop TLE. Neither an ordinary febrile seizure nor a history of fever establishes the cause of later epilepsy.
Other structural substrates include focal cortical dysplasia, vascular malformations, injuries, and slowly growing tumors. Ganglioglioma and dysembryoplastic neuroepithelial tumor are well-recognized epilepsy-associated tumors, often presenting in younger people. Temporal encephaloceles—small herniations of brain tissue through skull-base defects—are another potentially overlooked substrate. Two abnormalities can coexist, such as hippocampal sclerosis and an adjacent developmental lesion, a situation often termed dual pathology.
Acquired causes include traumatic injury, stroke, and infection. Encephalitis caused by herpes simplex virus has a particular tendency to involve temporal and adjacent regions and can leave a lasting epileptogenic injury. Autoimmune encephalitis may produce temporal seizures together with memory impairment, psychiatric symptoms, or altered consciousness. Antibodies directed against LGI1 are associated with a characteristic syndrome that may include very brief faciobrachial dystonic seizures, while other immune-associated epilepsies have different clinical patterns. These disorders are not interchangeable with ordinary chronic TLE, and antibody results require clinical interpretation.
Genetic susceptibility can act alone or together with structural injury. Familial lateral temporal epilepsy may involve brief auditory symptoms or receptive language disturbance; pathogenic variants in LGI1 and other genes account for some families. Familial mesial temporal epilepsy also occurs, often without an obvious single-gene explanation. The LGI1 gene disorder is biologically distinct from acquired disease involving antibodies against the LGI1 protein. In many patients, the cause remains unknown despite detailed investigation.
Mechanisms of seizure generation and persistence
A seizure develops when neural populations enter an abnormal state of excessive or unusually synchronized activity. In TLE, this can involve altered intrinsic neuronal excitability, changes in excitatory and inhibitory synapses, and abnormal interactions among local and distant circuits. The shorthand of “too much excitation and too little inhibition” captures part of the process but misses its timing and spatial organization. Inhibitory neurons can initially restrain a seizure, participate in pathological synchrony, or fail as the event evolves.
Glutamate is a principal excitatory neurotransmitter, while gamma-aminobutyric acid, or GABA, mediates much of the brain’s inhibition. Changes in their receptors, transporters, and release mechanisms can influence seizure susceptibility. Ion channels controlling sodium, potassium, calcium, and chloride also determine how readily neurons fire and how they respond to synaptic input. Altered chloride regulation can weaken the inhibitory effect of GABA in some circumstances. These mechanisms vary across patients and experimental models rather than defining one universal temporal epilepsy pathway.
In hippocampal sclerosis, neuronal loss occurs alongside extensive circuit remodeling. Surviving dentate granule cells may develop abnormal recurrent connections through mossy fiber sprouting. Changes in interneurons, extracellular matrix, and cellular support systems can further modify network behavior. Such findings are important, but their presence does not prove that any one abnormality is necessary or sufficient for seizures. Some are possible causes of persistent excitability, some may result from repeated seizures, and others may represent compensatory responses.
Glial cells and the vascular environment also matter. Astrocytes regulate extracellular potassium and neurotransmitter clearance; disturbances can make sustained firing easier. Inflammatory signaling may alter synaptic function and neuronal thresholds. Breakdown of the blood–brain barrier can expose brain tissue to circulating substances that influence glia and neurons. These mechanisms link tissue injury to longer-term network change and are investigated as potential treatment targets, although their contribution in an individual patient is difficult to measure directly.
Epileptogenesis denotes the development of a brain state capable of generating spontaneous recurrent seizures. After an injury, this process may unfold over months or years, but TLE need not follow a simple injury–silent interval–epilepsy sequence. Repeated seizures can further alter networks, while some epilepsies remain relatively stable. Experimental work commonly uses electrically induced kindling or chemically induced status epilepticus to study these changes. Such models reproduce selected features of human disease, not its full clinical diversity.
Seizure manifestations and associated symptoms
An early subjective seizure symptom is often called an aura. In this setting, an aura is itself a focal seizure, not merely a warning that a seizure might happen. Common temporal auras include a rising sensation from the upper abdomen, sudden fear without an external cause, déjà vu, an unfamiliar feeling toward familiar surroundings, and unusual smells or tastes. An event can end at this stage, with awareness and the ability to respond preserved.
As a seizure spreads, behavior may pause and responsiveness may diminish. The person may stare, stop speaking, or continue an activity without normal interaction. Repetitive, apparently purposeful movements—automatisms—can include lip smacking, chewing, swallowing, picking at clothing, or manipulating nearby objects. These actions are generally not under ordinary voluntary control. Their presence supports a focal seizure interpretation in an appropriate context, but they are not unique to TLE.
Lateral temporal seizures may begin with elementary sounds such as buzzing or ringing, distortions of sound, or difficulty understanding spoken language. More elaborate experiences can include fragments of music or altered perceptions, although these do not by themselves identify a precise onset site. Motor signs sometimes help establish the affected hemisphere. Sustained dystonic posturing of one arm, for example, often points to seizure involvement of the opposite hemisphere. No single behavior reliably replaces combined clinical and electrical assessment.
A typical focal seizure with impaired awareness often lasts roughly one to two minutes, followed by a variable recovery period. Postictal confusion, sleepiness, headache, and difficulty recalling the event are common. Temporary aphasia after a seizure can support involvement of the language-dominant hemisphere. Some events are much shorter or longer, and prolonged or repeatedly recurring seizures without recovery can form status epilepticus. Temporal nonconvulsive status may present mainly as sustained confusion or fluctuating responsiveness rather than continuous convulsions.
Seizure symptoms must be distinguished from symptoms occurring between seizures. Persistent memory difficulty, low mood, and anxiety may coexist with TLE but are not automatically ongoing seizure activity. Likewise, fear, unusual familiarity, or a strange smell in isolation has many possible explanations. The characteristic clinical pattern is usually a recurrent, abrupt, relatively brief, and stereotyped sequence. Witness descriptions are especially important because impaired awareness and amnesia often limit the affected person’s account.
Diagnostic investigation and localization
Diagnosis begins with a detailed medical history, including descriptions from witnesses and, when available, recordings of typical events. Important details include the first symptom, sequence of behavioral changes, responsiveness, duration, recovery, sleep relationship, and whether different event types occur. The diagnostic task has several layers: determining whether events are epileptic, classifying the seizure type, locating the network of onset, and identifying an underlying cause. Evidence sufficient for one layer may be insufficient for another.
Electroencephalography records electrical activity through scalp electrodes. Temporal spikes or sharp waves between seizures can support the diagnosis, but a normal routine EEG does not exclude TLE. Deep medial temporal activity may be poorly visible at the scalp, and a brief recording may miss intermittent abnormalities. Sleep and repeated or prolonged recordings can increase the opportunity to detect relevant activity. Conversely, an EEG abnormality must be interpreted alongside the clinical events rather than treated as a diagnosis by itself.
Video-EEG monitoring combines continuous electrical recording with observation of behavior. Capturing a habitual event can clarify its nature and relationship to an electrical seizure pattern. However, scalp EEG may show a temporal pattern only after a seizure has spread from its actual origin, and some focal seizures have no clear scalp correlate. Distinguishing the onset zone from areas recruited later is particularly important before an operation intended to stop seizures.
Magnetic resonance imaging using an epilepsy-focused protocol searches for hippocampal sclerosis, developmental abnormalities, tumors, vascular lesions, and other structural causes. Thin-section, appropriately oriented sequences and expert review improve detection. An apparently normal MRI does not mean that no structural or network abnormality exists. In selected cases, positron emission tomography identifies interictal regions of reduced glucose metabolism, while ictal perfusion imaging can help identify regions activated during seizures. These findings support localization but do not independently define the tissue requiring treatment.
Neuropsychological testing assesses learning, memory, language, attention, and other functions. It provides a baseline and contributes to estimates of treatment-related cognitive risk. Functional MRI can help establish language organization; selected centers also use the intracarotid anesthetic procedure, traditionally called the Wada test. When noninvasive findings remain insufficient, stereo-electroencephalography records from implanted depth electrodes.
The differential diagnosis includes syncope, panic attacks, sleep disorders, migraine phenomena, dissociative seizures, and metabolic disturbances such as hypoglycemia. More than one condition can occur in the same person. Diagnostic confidence therefore depends on agreement among history, observed events, electrical findings, imaging, and the broader clinical context.
Antiseizure medication and medical treatment
Antiseizure medications are the usual initial treatment for recurrent temporal seizures. They reduce the probability of seizures by modifying ion channels, neurotransmitter release, or inhibitory signaling. Their effectiveness does not ordinarily demonstrate reversal of the underlying structural lesion or of epileptogenesis. Drugs used for focal seizures include lamotrigine, levetiracetam, carbamazepine, oxcarbazepine, lacosamide, and several other agents. There is no single medication that is best for all forms of TLE.
Treatment selection depends on more than seizure classification. Age, other illnesses, medication interactions, reproductive circumstances, cognitive and psychiatric effects, and the practicalities of dosing all influence prescribing. Lamotrigine affects voltage-gated sodium channels; levetiracetam binds synaptic vesicle protein SV2A; lacosamide enhances slow inactivation of sodium channels. These different mechanisms can produce similar clinical aims while differing in adverse-effect profiles. Combining drugs with different targets does not guarantee improved control or fewer adverse effects.
Dose-related adverse effects can include dizziness, sleepiness, imbalance, and slowed thinking. Particular agents may produce behavioral changes, low serum sodium, rash, or organ-specific toxicity. Some antiseizure drugs induce hepatic enzymes and thereby alter the metabolism of other medicines. Pregnancy introduces additional considerations because fetal risks vary among medications and drug concentrations can change during gestation. These issues are part of individualized medical management rather than features unique to temporal epilepsy.
A widely used International League Against Epilepsy definition, published in 2010, identifies drug-resistant epilepsy after failure of two adequately chosen and tolerated medication schedules, used alone or in combination, to achieve sustained seizure freedom. The definition requires attention to whether the diagnosis was correct, whether the drug was appropriate for the seizure type, and whether exposure was adequate. Failure caused by an unsuitable regimen is not equivalent to failure of an adequate therapeutic trial.
Some patients attain lasting seizure control with medication, while others continue to have seizures despite multiple agents. The probability of complete control generally becomes lower after repeated adequate treatment failures, although later remission remains possible. Medical care also addresses associated conditions and specific causes. For example, an active immune-mediated disorder may require immunotherapy in addition to seizure suppression, whereas a stable scar-related epilepsy has a different treatment rationale. Acute treatment of prolonged seizures is distinct from the ongoing daily regimen.
Resective surgery, ablation, and neuromodulation
TLE has an established role in neurosurgery because a well-localized temporal seizure-generating network can sometimes be removed or disconnected. Surgical assessment examines whether a target can be defined and treated without unacceptable functional harm. It integrates seizure descriptions, video-EEG, MRI, neuropsychology, and additional tests when required. Drug resistance does not automatically make an individual operable, and an abnormal scan alone is not sufficient to establish the appropriate surgical target.
Anterior temporal resection removes part of the anterior temporal lobe, often together with medial temporal structures. Selective amygdalohippocampectomy targets the amygdala and hippocampal region while preserving more lateral cortex. The optimal approach depends on the suspected network, lesion, anatomy, and functional considerations. In appropriately selected patients, approximately 60–80% may become free of disabling seizures in the early years after temporal resection, with results varying by pathology, selection, follow-up, and the definition of seizure freedom.
A landmark randomized controlled trial led by Samuel Wiebe and published in 2001 compared temporal surgery with continued medical treatment in drug-resistant TLE. At one year, 58% of the surgical group and 8% of the medical group were free of seizures impairing awareness. The study established that surgery could outperform continued medication in a carefully selected population. Its result does not apply unchanged to every temporal epilepsy, especially bilateral or poorly localized disease.
Risks include infection, hemorrhage, visual field loss, memory decline, language difficulties, and changes in mood. Injury to temporal optic radiations can produce an upper-quadrant visual field defect on the opposite side. Dominant temporal resections may impair verbal learning or naming, particularly when the tissue being removed retains substantial function. Risk assessment considers both the functional capacity of the proposed surgical region and the reserve of the remaining brain.
MRI-guided laser interstitial thermal therapy uses a stereotactically placed laser to ablate selected tissue. It offers a less extensive surgical approach, but seizure outcomes and cognitive effects depend on the target and patient group, and long-term comparative evidence is less extensive than for established resections.
Neuromodulation is another option in selected drug-resistant cases, especially when resection is unsuitable. Vagus nerve stimulation, anterior thalamic deep brain stimulation, and responsive neurostimulation modulate seizure networks without removing the temporal lobe. Responsive systems can detect selected intracranial electrical patterns and deliver stimulation. These therapies more often reduce seizure frequency than eliminate seizures altogether; they are not equivalent substitutes for a potentially curative resection in every case.
Cognitive effects, comorbidity, and long-term course
Memory complaints are prominent in TLE because the disorder involves memory-related structures, but their causes are multiple. Hippocampal damage, seizures, interictal epileptiform activity, disrupted sleep, medications, depression, and anxiety can all influence performance. Standard testing may reveal difficulty learning new material, while everyday concerns may involve remembering appointments, conversations, or autobiographical events. The subjective severity of memory difficulty does not always closely match performance on conventional tests.
Some people report accelerated long-term forgetting: information seems adequately learned and retained over conventional testing intervals but is lost disproportionately over days or weeks. Others have gaps in autobiographical memory. These patterns have been described particularly in temporal epilepsies, including transient epileptic amnesia, but they are not universal. Their mechanisms remain under investigation, including the possible effects of seizures or abnormal activity on memory consolidation during sleep.
Depression and anxiety disorders are important comorbidities. Their relationship with epilepsy is bidirectional and involves biological, psychological, and social factors. Symptoms can precede epilepsy diagnosis, arise in response to its consequences, or fluctuate around seizures. The historical idea of a single “temporal lobe personality” is not supported as a general clinical description. Behavioral or religious interests, creativity, irritability, and personality traits cannot reliably identify TLE.
Long-term prognosis ranges from sustained remission to persistent drug-resistant seizures. A localized removable lesion, a well-defined unilateral onset, and agreement among diagnostic findings can be favorable features for surgical seizure outcome. Bilateral disease, broader networks, and discordant investigations can complicate treatment. Even after years without seizures, recurrence is possible, including after surgery. Conversely, persistent seizures do not imply inevitable progressive intellectual decline.
The disorder can affect education, employment, driving eligibility, independence, and health-related quality of life. Injury and mortality risks are influenced strongly by seizure type and control. Sudden unexpected death in epilepsy is a recognized complication of epilepsy, with risk particularly associated with ongoing tonic–clonic seizures; it is not specific to temporal onset. Driving rules and occupational restrictions vary by jurisdiction. These outcomes depend not only on brain pathology but also on access to diagnosis, treatment, and social support.
Historical development and current research
Nineteenth-century neurologist John Hughlings Jackson described focal seizure phenomena and experiences he called “dreamy states,” helping connect apparently psychological events with localized brain dysfunction. His observations contributed to the recognition that seizures could involve complex perception and memory without beginning as generalized convulsions. Later clinical and pathological work increasingly associated these experiences with temporal structures, although modern network concepts are broader than the early localization models.
The introduction of clinical EEG in the twentieth century strengthened the distinction between focal and generalized epilepsies. In the 1930s and 1940s, investigators including Frederic and Erna Gibbs and William Lennox helped characterize electrical abnormalities associated with psychomotor seizures. Wilder Penfield and collaborators developed surgical mapping and operative approaches that linked temporal electrical activity, stimulation responses, and clinical symptoms. These advances made systematic temporal epilepsy surgery possible before modern structural imaging.
A closely related development transformed understanding of memory. In 1953, Henry Molaison, later widely known as H.M., underwent bilateral medial temporal surgery for severe epilepsy. William Beecher Scoville and Brenda Milner reported his profound impairment of new long-term memory in 1957. The case demonstrated the importance of medial temporal structures for memory formation. It also underscored why contemporary epilepsy surgery requires careful assessment of functional risk; the historical bilateral procedure is not a model for routine modern temporal resection.
MRI subsequently allowed hippocampal abnormalities and many epileptogenic lesions to be recognized before surgery. Functional imaging, advanced EEG analysis, intracranial recording, and image-guided procedures have further refined localization. Research now increasingly treats TLE as a family of network disorders shaped by different pathologies rather than a single hippocampal disease. Quantitative imaging and computational models investigate how local abnormalities interact with distributed systems controlling memory and awareness.
Current investigations include inflammatory pathways, genetic susceptibility, glial function, and predictors of treatment response. Candidate biomarkers include high-frequency electrical oscillations, quantitative imaging measures, and molecular signals, but no single marker reliably diagnoses all TLE or predicts every individual outcome. Machine-learning methods are being studied for lesion detection and seizure forecasting, with validation across different hospitals and patient populations remaining essential.
A major distinction in this research is between suppressing seizures and preventing or reversing the processes that sustain epilepsy. Gene-based approaches, cell therapies, and targeted modulation of pathological circuits are being explored, but they do not yet constitute established routine treatments for TLE. Translating experimental results requires demonstrating durable benefit, functional safety, and applicability to clearly defined patient groups.