Clinical manifestations

Presentation and course

Familial mesial temporal lobe epilepsy is characterized by seizures with mesial temporal semiology (rising epigastric sensation, psychic and experiential phenomena). Déjà-vu and jamais-vu appear to be over-represented in this group of patients (07). Focal impaired awareness seizures with oral or manual automatisms and prominent postictal confusion (identical to those observed in nonfamilial mesial temporal lobe epilepsy) are frequent. Focal to bilateral tonic-clonic seizures can occur, usually at the beginning of the disease, before treatment is initiated. Age at seizure onset may vary from 1 to 56 years, with a mean of 10 years (56). History of febrile seizures in infancy seems to be less frequent (around 11%) (56) than in a surgical familial mesial temporal lobe epilepsy series alone (30%) (57), and much less frequent than in a random surgical temporal lobe epilepsy series not selected for familial temporal lobe epilepsy (40%) (01). Familial mesial temporal lobe epilepsy is inherited as an autosomal dominant trait with incomplete penetrance, although complex inheritance cannot be ruled out in some families.

Interictal EEG abnormalities consist of unilateral or bilateral epileptiform discharges over mesiotemporal regions, best observed with zygomatic or sphenoidal electrodes. Ictal recordings show rhythmic activity over the anterior and mesial temporal regions. Although the majority of patients have a benign clinical course, including spontaneous remission, refractory seizures may occur in up to 29% of patients (56). MRI shows varying degrees of hippocampal atrophy and of hyperintense T2 signal (56), and this may also be found in asymptomatic family members (58). However, hippocampal atrophy is more frequent and more severe in patients with refractory seizures (59; 56).

In a series by Crompton and colleagues, 20 families with familial mesial temporal lobe epilepsy have been described (30). Affected family members had a benign outcome except for one individual who required surgery (normal MRI and normal pathology). Contrary to the first description from this group, which was population-based, this new series was also composed of patients who had been referred to the investigators. It should be noted that few families in this series were identified through probands that did not have temporal lobe epilepsy.

Another clinical, MRI, and EEG study evaluated outcome predictors in 17 families from a Brazilian series over a mean of 7.6 years of follow-up (71). The authors found that patients with infrequent seizures could develop frequent seizures (17.6%) or seizure freedom (23.5%). Twenty-one percent of patients who were seizure-free had recurrence, but with infrequent seizures. Refractory patients remained uncontrolled unless they had undergone surgery, and 12% of asymptomatics presented with infrequent seizures. Predictive factors of poor outcome were hippocampal atrophy on MRI and interictal epileptiform discharges on EEG.

Prognosis and complications

Although the first description of familial mesial temporal lobe epilepsy defined it as a benign condition (15; 17), it is now well-known that some patients may be refractory to medical treatment. Overall, 81% of patients with familial mesial temporal lobe epilepsy achieved good seizure control on medication or remitted spontaneously. Refractory seizures were observed in 19%, and surgical treatment was considered in these cases (57). When unilateral or clearly asymmetric bilateral EEG-MRI abnormalities were observed, patients with familial mesial temporal lobe epilepsy had an overall likelihood of 85% to become seizure-free following surgery (57). Indeed, a positive family history of epilepsy and of febrile seizures has been associated with long-term remission following surgery in large cohorts of patients with mesial temporal lobe epilepsy (01; 33).

Memory deficits have been observed in familial mesial temporal lobe epilepsy patients, and frequent seizures potentiate the effect of hippocampal atrophy on memory performance (02). As observed in classical mesial temporal lobe epilepsy series, left hippocampal atrophy correlates with verbal memory and general memory deficit, whereas no correlation between right hippocampal atrophy and visual memory deficit was observed.

Clinical vignette

Hippocampal atrophy with hyperintense T2 signal and abnormal hippocampal internal structure was identified in these three siblings with familial mesial temporal lobe epilepsy.

Patient III-3 was a 50-year-old man who had focal impaired awareness seizures with rare focal to bilateral tonic-clonic seizures since the age of 4 and had not received medication for several years. He had sporadic focal seizures only when he drank alcohol. His MRI showed left hippocampal atrophy. Patient III-4 was a 47-year-old woman who had a simple febrile seizure when she was 2 years old. She started having focal impaired awareness seizures with occasional focal to bilateral tonic-clonic seizures at age 9 years. She experienced frequent seizures, refractory to medication, and her MRI showed bilateral hippocampal atrophy. Patient III-5 was a 47-year-old woman, the twin sister of patient III-3, who had only had a few seizures in her life. The first seizure with focal to bilateral tonic-clonic presentation was at 8 years. She never took anti-epileptic medications, and her sporadic seizures were associated with major stress or alcohol intake. Her MRI showed bilateral hippocampal atrophy with left side predominance.

Biological basis

Etiology and pathogenesis

The etiology of familial epilepsies must first be determined by the genetic basis that indicates an inherited disease and secondly by the structural and functional abnormalities that are associated with this genetic background.

No single gene molecular basis has been confirmed for familial mesial temporal lobe epilepsy to date, although several loci have been mapped. In a French family with febrile seizures and nonlesional temporal lobe epilepsy occurring in the same family members, there was linkage to chr 18qter as well as to chr 1q25-q31, although the latter did not reach a lod score of 3.0 (13). These authors suggested that this might represent digenic inheritance. Linkage to chr 12q22-q23.3 has been described in a Belgian kindred with familial mesial temporal lobe epilepsy and febrile seizures (26). A de novo 12q22-q23.3 duplication has been identified through array-CGH performed in a patient with dysmorphic features and temporal lobe epilepsy (93).

Linkage to chr 18p11.3-11.2 has been demonstrated in one large Brazilian kindred originally described by Kobayashi and colleagues (59; 67). The linkage to chromosome 18p11.31 in one Brazilian family has shown a Z(max) of 3.12 at D18S452 using the presence of hippocampal atrophy or hyperintense hippocampal T2 signal in MRI for phenotypic classification, suggesting that the structural abnormality itself segregates independently from the clinical manifestations (69). Linkage to chr 4q13.2-q21.3 was demonstrated in a large American family with four generations of affected individuals and no MRI abnormalities (46). However, confirmation of these loci in other families has not yet been reported (84), again suggesting the possibility of complex inheritance. A locus on chr 3q has been identified in two families with nonlesional mesial temporal lobe epilepsy (25; 39).

Various polymorphisms have been found in association with mesial temporal lobe epilepsy, but these could not be replicated in larger cohorts of patients (88; 22). An increased frequency of the T allele at the 511 position of the interleukin 1beta gene has been described among patients with mesial temporal lobe epilepsy with hippocampal sclerosis as compared with patients without hippocampal sclerosis and with controls; it was even higher in those with a history of prolonged febrile seizures (51; 52). However, this has not been confirmed in other series (20; 49; 73). Overrepresentation of a functional polymorphism in the prodynorphin gene promoter (possibly related to seizure suppression) has been found in patients with temporal lobe epilepsy and a positive family history of epilepsy as compared to controls (83).

A GABA B receptor 1 polymorphism (G1465A) in patients with mesial temporal lobe epilepsy was overrepresented as compared to the normal population, especially in those with refractory seizures (43; 42). However, other studies failed to replicate such findings (88; 22). Mutations and polymorphisms in other GABA receptor genes (GABRA1, GABRA5, GABRG2, and GABRD) were also excluded in patients with temporal lobe epilepsy and previous history of febrile seizures (62). Polymorphisms of the serotonin transporter gene (5-HTT) have been reported in association with temporal lobe epilepsy (63). The presence of all these polymorphisms would suggest a polygenic or multifactorial mode of inheritance, as first hypothesized by Andermann (04; 05; 06).

Mutations in the SCN1A and SCN1B genes, encoding the sodium channel alpha 1 and beta 1 subunits, respectively, have been identified in affected individuals with the mesial temporal lobe epilepsy phenotype in families with genetic epilepsy with febrile seizures plus (GEFS+) (27). However, these mutations have not yet been reported in pure familial mesial temporal lobe epilepsy kindreds. In addition, no linkage to the SCN2A gene (68) or to voltage-gated potassium channels (66) has been found in Brazilian familial mesial temporal lobe epilepsy kindreds.

No families with familial mesial temporal lobe epilepsy were found to have an LGI-1 mutation, not even those families in which one or more members had auditory features alone or in association with mesial symptoms (77; 11; 16; 84). This further supports the fact that familial mesial temporal lobe epilepsy and familial lateral temporal lobe epilepsy constitute distinct genetic syndromes. Loss-of-function mutations in the DEPDC5 gene on chromosome 22q, encoding the DEP domain containing 5 protein and associated with familial focal epilepsy with variable foci, have been identified in two kindreds with familial mesial temporal lobe epilepsy (47).

Detailed phenotyping is crucial to allow gene discovery, which has not yet been advanced in familial mesial temporal lobe epilepsy as compared to other focal familial epilepsy syndromes. Familial aggregation of focal seizure types and semiology has been systematically assessed in the Epilepsy Phenome/Genome Project cohort comprising 302 individuals with nonacquired focal epilepsy from 149 families (89). As expected, fear/anxiety and déjà vu/jamais vu showed significant familial aggregation in temporal lobe epilepsy families, representing 24% of kindreds.

The presence of clear-cut hippocampal atrophy in both affected and asymptomatic family members in familial mesial temporal lobe epilepsy suggests that the hippocampal abnormalities themselves could be inherited, and not necessarily lead to epilepsy (58). The phenotype would then be dependent on interaction with other modifying factors. Three published papers addressed the possibility of inherited hippocampal abnormality. Fernandez and colleagues studied families of patients with a previous history of febrile seizures, including asymptomatic individuals (40). They observed subtle abnormalities in hippocampal configuration, internal structure and volume in subjects with and without febrile seizures. This suggested the presence of preexisting hippocampal malformations as an associated factor leading to febrile seizures and subsequent hippocampal sclerosis. A subsequent study of children with prolonged febrile seizures and their monozygotic asymptomatic twins did not confirm these findings (48). Another study, including volumetry and T2 relaxometry, did not confirm the presence of hippocampal atrophy, although the volumes from asymptomatic family members was lower than controls, suggesting that this finding could be a predisposing abnormality (91).

Volumetric studies of the temporal lobes showed similar patterns of atrophy in patients with familial and nonfamilial mesial temporal lobe epilepsy (41). Yasuda and colleagues described differences in the extent of gray matter atrophy determined by MRI voxel-based morphometry in patients with mesial temporal lobe epilepsy with and without family history (99). Familial patients showed a less widespread pattern of atrophy and had less severe dysfunction in neuropsychological assessment and less frequent history of precipitating injury as compared to those without family history of seizures. Overall, these findings support again the influence of genetic factors determining hippocampal abnormalities in familial mesial temporal lobe epilepsy whereas more widespread atrophy in nonfamilial cases might be related to environmental factors.

A longitudinal quantitative MRI study was performed in patients with familial mesial temporal lobe epilepsy and compared to sporadic cases, showing that patients with familial mesial temporal lobe epilepsy have progressive hippocampal atrophy that occurs independently of seizure frequency (29). In a longitudinal study involving 21 patients with familial mesial temporal lobe epilepsy and nine asymptomatic relatives, the same group reported hippocampal T2 relaxometry changes in the mean time interval between MRIs of 4.4 ± 1.5 years (86). Interestingly, MRI-negative patients and asymptomatic relatives showed intermediate relaxometry values as compared to MRI-positive patients and controls. This further suggests genetically determined, and in some individuals, seizure-independent hippocampal abnormalities.

Mexican kindreds described as familial mesial temporal lobe epilepsy in a large cohort showed MRI abnormalities in 25% of affected individuals, including 15% with MRI signs of hippocampal sclerosis and 10% described as hippocampal “hypotrophy” (criteria for the latter, however, neither explained nor illustrated). As the authors described some affected family members with unusual seizure semiology features within mesial temporal lobe epilepsies, some classified as having extra temporal epilepsy foci (including parietal and frontal epilepsy), the differential diagnosis of familial focal epilepsy with variable foci should be considered (44).

Subtle reduced cortical volumes in the anteromedial temporal cortex of asymptomatic siblings of patients with sporadic mesial temporal lobe epilepsy associated with hippocampal sclerosis suggest that such localized traits are possibly heritable (03). Interestingly, the alterations of cortical morphology found in unaffected siblings of patients with left mesial temporal lobe epilepsy associated with underlying hippocampal sclerosis (left entorhinal cortex and parahippocampal gyrus) differed from unaffected siblings of right-sided patients (right entorhinal cortex and temporal pole). In 127 asymptomatic first-degree relatives of patients with mesial temporal lobe epilepsy associated with hippocampal sclerosis from three independent cohorts, no changes in cortical thickness have been confirmed, which suggests that changes in the cortical surface area may represent an inherited trait that preceded onset of seizures (03; 23).

Available qualitative pathology from surgical specimens obtained from operated familial mesial temporal lobe epilepsy patients showed the typical pattern of mesial temporal sclerosis: selective neuronal loss in CA1, CA3 and CA4 with relative preservation of CA2, and variable involvement of the amygdala and parahippocampal region (57). A more detailed study on the pathological substrate of familial mesial temporal lobe epilepsy has been presented by Andrade-Valenca and colleagues (10). They evaluated the surgical specimens from 20 familial mesial temporal lobe epilepsy and 39 sporadic mesial temporal lobe epilepsy patients. No differences could be found in cell densities in CA subfields, fascia dentata, polymorphic region, subiculum, prosubiculum, and presubiculum. However, greater intensity of mossy fibers in the fascia dentata-inner molecular layer was found in patients with sporadic mesial temporal lobe epilepsy. The authors raised the hypothesis that patients with familial mesial temporal lobe epilepsy respond differently to plastic changes induced by cell loss, neuronal deafferentation, or epileptic seizures.

mRNA expression profiles in hippocampal surgical brain specimens of patients with familial mesial temporal lobe epilepsy show a different profile compared to sporadic mesial temporal lobe epilepsy (65). In familial specimens, there is an over-representation of the biological pathways related to protein response, mRNA processing, and synaptic plasticity and function. In sporadic specimens, the gene expression profile suggests that the inflammatory response is highly activated. Although an enrichment of gene sets involved in inflammatory cytokines and mediators and chemokine receptor pathways was identified in both groups, there was also an enrichment of epidermal growth factor signaling, prostaglandin synthesis and regulation, and microglia pathogen phagocytosis pathways in sporadic mesial temporal lobe epilepsy. These molecular signatures suggest different underlying molecular mechanisms that may have implications for the development of therapeutic approaches.

The observation of mesial temporal sclerosis in operated familial mesial temporal lobe epilepsy patients who became seizure-free suggests that mesial temporal sclerosis represents the epileptogenic substrate, at least in some of these families, analogous to what is observed in nonfamilial or “sporadic” mesial temporal lobe epilepsy cases.

Most likely, familial mesial temporal lobe epilepsy will be found to have a major gene leading to hippocampal abnormalities, and the phenotype could be influenced by additional genetic and environmental modifying factors.

The identification of a positive family history of seizures in patients with mesial temporal lobe epilepsy is not sufficient for a diagnosis of familial mesial temporal lobe epilepsy. The best definition of familial mesial temporal lobe epilepsy is based on the familial recurrence of mesial temporal lobe epilepsy, which is defined by clinical-EEG criteria according to the ILAE recommendations (28), in the absence of any suggestion of other focal or generalized epilepsy syndromes in other affected family members. Thus, the observation of at least two patients with mesial temporal lobe epilepsy in one family is necessary but not sufficient for the definition of familial mesial temporal lobe epilepsy. The observation of an autosomal dominant inheritance pattern with incomplete penetrance implies the presence of asymptomatic carriers of the genetic abnormalities, who can transmit the disease to their offspring. Therefore, we should consider inclusion of families not only with affected first-degree relatives, but also with affected second- and third-degree relatives. This criterion has not been employed in some reported series, leading to exclusion of many possible familial mesial temporal lobe epilepsy kindreds (75).

The presence of family members with episodes of either generalized tonic-clonic seizures only with no witnessed onset, or with febrile seizures alone, does not exclude the diagnosis of familial mesial temporal lobe epilepsy because these do not fulfill criteria for other epilepsy syndromes. Families with individuals who have such clinical manifestations should also be considered to have familial mesial temporal lobe epilepsy, as long as at least two affected individuals fulfill the clinical and EEG criteria for mesial temporal lobe epilepsy and no other epilepsy type is found to suggest familial focal epilepsy with variable foci.

Epidemiology

There is no predominance of familial mesial temporal lobe epilepsy in any particular ethnic group to our knowledge. Families with familial mesial temporal lobe epilepsy have been described in Australia, Canada, Brazil, Italy, Belgium, France, Macedonia, and Mexico. The real prevalence of familial mesial temporal lobe epilepsy worldwide is probably underestimated.

Ascertainment of these families requires detailed questioning of patients and family members. This has only been emphasized because, in the past, mesial temporal lobe epilepsy as well as other focal epilepsies were considered to be symptomatic and due largely to environmental factors. A preliminary hospital-based study found that familial mesial temporal lobe epilepsy represented 7% of all patients with mesial temporal lobe epilepsy (59).

Among 62 families with focal epilepsy evaluated through the Epi4K Consortium, 37% (n = 23) had familial temporal lobe epilepsy, most of them with seizure semiology indicating mesial temporal focus. Two families had reported symptoms in keeping with FLTLE and three families had family members with semiologies indicative of both mesial and lateral foci. MRI, whenever available, did not show evidence of hippocampal atrophy in visual analysis (36).

An interesting report on the incidence of familial mesial temporal lobe epilepsy in patients with newly diagnosed MRI negative mesial temporal lobe epilepsy used a structured interview of first-degree relatives of patients seen in a First Seizure Clinic, and compared data obtained with pairwise age- and sex-matched controls (74). Whereas at clinic assessment only 2 of 40 patients had been recognized to have familial mesial temporal lobe epilepsy, blind review of interview transcripts in their relatives allowed diagnosis of mesial temporal lobe epilepsy in 9 of 121 relatives, with intense déjà vu being the most common feature. Physiological déjà vu was commonly and equally prevalent in relatives and controls. The differentiation of epileptic and nonepileptic déjà vu features in this study thus resulted in the diagnosis of eight familial mesial temporal lobe epilepsy kindreds (18.2%) in whole cohort.

In Macedonia, four small kindreds with possible familial mesial temporal lobe epilepsy were identified from active inquiry on family history of seizures among 52 patients with mesial temporal lobe epilepsy seen in a 2-year period (31).

Prevention

No method of prevention is known method, except in large families with proven linkage. In these families, amniocentesis and prenatal diagnosis employing haplotype analysis may be feasible. This will, of course, become more accurate and easier when the gene mutations can be determined. However, prenatal diagnosis is not likely to be practiced routinely because the usually mild phenotype of mesial temporal lobe epilepsy would probably not warrant termination of pregnancy.

Differential diagnosis

A family history of seizures is common among mesial temporal lobe epilepsy patients. Many of them have one or more relatives who have experienced a single episode compatible with either focal impaired awareness seizures or generalized tonic-clonic seizures, and a history of febrile seizures is also frequently found. However, unless there are two individuals in the family with well-defined mesial temporal lobe epilepsy, a diagnosis of familial mesial temporal lobe epilepsy cannot be made.

In addition, there are other familial epilepsy syndromes in which patients with mesial temporal lobe epilepsy are found. In familial focal epilepsy with variable foci (80; 98; 97; 21; 18), most reported families mapped to chromosome 22q. Mutations in the DEPDC5 gene have been identified in familial focal epilepsy with variable foci families, but also in other familial focal epilepsies (32; 47; 64). Affected family members may present with various forms of focal epilepsy, including frontal lobe epilepsy and mesial temporal lobe epilepsy (55).

A DEPDC5 nonsense mutation (c.918C> G; p.Tyr306*) has been identified in a familial mesial temporal lobe epilepsy family with two affected members, expanding on the possible phenotypes associated with this gene originally linked to familial focal epilepsy with variable foci (85; 12). In addition, two siblings with refractory mesial temporal lobe epilepsy had a pathological diagnosis of focal cortical dysplasia IIIa (ie, hippocampal sclerosis and focal cortical dysplasia) in brain tissue from anterior temporal resection, but no genetic diagnosis was made (38). Possibly pathogenic variants in the DEPDC5 gene, associated with FFEVF, were found in sporadic temporal lobe epilepsy patients with and without hippocampal atrophy (90).

Ricos and colleagues investigated mutations in NPRL2 and NPRL3 genes, which together with DEPDC5, are part of the GATOR1 complex regulating mTOR pathway, and they found mutations in each of these genes in small kindreds composed of only two patients with temporal lobe epilepsy that could have a mesial phenotype, but that was not clearly specified (76).

In GEFS+ related to mutations in genes SCN1A, SCN2A, SCN1B, and GABRG2 (96; 94; 95; 37; 61; 87; 45), patients may present heterogeneous epilepsy phenotypes. Febrile seizures are the most common phenotype, followed by febrile seizures plus (FS+), where individuals have seizures with fever that may persist beyond the age of 6 years and may be associated with afebrile generalized tonic-clonic seizures (78; 81). Less frequent phenotypes seen in GEFS+ involve other generalized and partial seizure types, including mesial temporal lobe epilepsy (78; 81; 95; 79).

Another focal epilepsy syndrome with some individuals presenting seizure semiology consistent with mesial temporal lobe epilepsy has been described in one large Brazilian kindred and named partial epilepsy with pericentral spikes (53). The majority of affected individuals, however, presented hemitonic or hemiclonic seizures, and molecular analysis showed linkage to chromosome 4p15.

Siren and colleagues described a kindred with heterogeneous benign epilepsy phenotypes (febrile seizures, possible temporal lobe epilepsy, and childhood absences) linked to three loci on chromosomes 5q11.2, 18p11-q11, and 17q12-q24 (82).

It is essential to evaluate the phenotype of all possibly affected individuals before classifying the family to have a specific familial syndrome. In addition, because the severity of the phenotypes in family members may vary, we can never be absolutely sure that an “isolated” or “sporadic” mesial temporal lobe epilepsy patient does not have familial mesial temporal lobe epilepsy. Molecular studies can be helpful in excluding other familial epilepsies with already identified gene mutations.

An updated review of these familial epilepsy syndromes can be found in the dedicated MedLink Neurology summaries: Genetic epilepsy with febrile seizures plus, Familial lateral temporal lobe epilepsy, Familial focal epilepsy with variable foci, and Autosomal dominant nocturnal frontal lobe epilepsy.

Diagnostic workup

As it is impossible to distinguish, so far, familial from nonfamilial mesial temporal lobe epilepsy in a single individual, it is necessary to obtain a detailed family history with the mother or the grandmother of each patient, to exclude familial recurrence (which is often hidden in families). All family members with suggestive symptoms of epilepsy should be interviewed personally or by telephone, and medical records should be obtained whenever possible.

Patients with familial mesial temporal lobe epilepsy should have the same investigations as nonfamilial patients. Routine and sleep EEG should be performed, as should MRI to rule out the presence of any treatable lesion. Refractory patients should have video-EEG monitoring and a neuropsychological work-up for pre-surgical evaluation, as well as volumetric and other post-processing MRI studies, as required.

Mutation analysis is not yet available for familial mesial temporal lobe epilepsy; however, the diagnosis can be confirmed by haplotype analysis in large families. If other familial epilepsy syndromes are suspected, including the lateral form of familial temporal lobe epilepsy, specific molecular testing can now be performed.

Management

Treatment should be based on the patient’s response to anti-epileptic drugs, and the rationale is similar to that in nonfamilial patients. Patients with familial mesial temporal lobe epilepsy may have refractory seizures, and surgical treatment should be considered, based on clinical-EEG-MRI data, despite the context of a familial epilepsy syndrome (24; 57).