• Mesial temporal sclerosis;
  • Two-hit hypothesis;
  • Microdysgenesis;
  • Genetic predisposition;
  • Initial precipitating injury;
  • Prolonged Febrile seizures;
  • Temporal Lobe epilepsy


  1. Top of page
  2. Abstract

Summary:  Mesial temporal sclerosis (MTS) is found in about two-thirds of patients with refractory temporal lobe epilepsy (TLE), and surgical removal of the sclerotic structures eliminates seizures in the majority of cases undergoing surgical resection. Although multiple factors have been implicated in the genesis of MTS, it is still unclear why some individuals are more likely to develop hippocampal sclerosis than others. Epileptologists have proposed that there must be at least two factors involved—an initial precipitating injury (IPI), such as a prolonged febrile seizure, CNS infection, or head trauma, and a second factor that increases vulnerability to neuronal injury. This has been termed the “two-hit hypothesis.” Three of the many factors that could possibly heighten susceptibility to neuronal injury and MTS are discussed here. These are microdysgenesis, hippocampal dysgenesis, prior seizures, and genetic predisposition. We conclude that there is currently no compelling evidence to support a role for microdysgenesis in MTS. Hippocampal dysgenesis, on the other hand, may account for febrile seizures and possibly MTS in a small subpopulation of patients with TLE. Additional larger studies are needed to confirm these findings. Experimental evidence indicates that an epileptogenic hippocampus can result from prolonged febrile seizures in infant rats, even though these seizures do not cause MTS in the rat. It is not known if this pathophysiological sequence occurs in humans. Lastly, there appears to be a strong genetic component that predisposes some individuals to MTS, regardless of whether they experience an IPI.

Mesial temporal sclerosis (MTS) is the most common lesion observed in patients with refractory temporal lobe epilepsy (TLE) across all age groups and is consistently found in 60–70% of the cases referred for surgical resection. Neuropathologically, MTS is characterized by the loss of hippocampal pyramidal and dentate hilar neurons with extensive gliosis with varying involvement of the amygdala and other mesial temporal structures (1). The observation that removal of these sclerotic mesial temporal structures eliminates seizures in up to 80% of the cases argues that MTS may be necessary for the generation and clinical expression of the seizures. At this time, however, we do not know what causes MTS to develop in a child's brain. Radiological diagnosis of MTS is based on the presence of three magnetic resonance image (MRI) findings:

  • • 
    Atrophy of the hippocampus, reflecting the loss of neurons
  • • 
    Increased hippocampal T2 signal intensity, reflecting an increase in gliosis due to some prior insult
  • • 
    Loss of hippocampal internal architecture normally discernable on MRI

Retrospective studies indicate that patients with MTS frequently have a history of an initial precipitating injury (IPI), such as febrile seizures, central nervous system (CNS) infection, head trauma, or birth trauma (2,3). As shown in Fig. 1, nearly all of the 60 patients with intractable TLE described by French et al. (2) had a history of an IPI. Febrile seizures were most common and occurred in 53% of the patients. Similar findings from other studies have stimulated much discussion about the possible role of febrile seizures in the genesis of MTS (4,5).


Figure 1. The frequencies of initial precipitating injuries associated with intractable TLE in 60 patients [from French et al., 1993 (2)].

Download figure to PowerPoint

Based on the results from epidemiological studies, we know that the majority of children who have a prolonged febrile seizure never develop MTS or TLE (6,7). Only a small percentage of children are susceptible to these events. The two-hit hypothesis proposes that there are two factors involved in the etiology of MTS, an IPI plus some other factor that increases the child's vulnerability to hippocampal injury. Fig. 2 shows the MRI-documented sequence of events that led to the development of MTS in a 2.3-year-old girl who had partial complex febrile status epilepticus lasting 3.5 h. The scan on the left documents the swelling and increased T2 signal that occurred in the right hippocampus 24 h after status epilepticus, suggesting edema due to an acute injury. Nine months later, the second scan revealed persistent increase in T2 signal and mild atrophy of the right hippocampus, consistent with a radiological diagnosis of MTS. The critical question is why this particular child was prone to a prolonged febrile seizure and subsequent MTS. Did she have some subtle preexisting structural or molecular abnormality in her brain, or was her brain perfectly normal prior to the seizure? Was she genetically predisposed to both prolonged seizures and subsequent neuronal injury? Is the resultant sclerotic hippocampus now capable of generating limbic seizures? Three of the many factors that may help to explain increased vulnerability of a hippocampus to injury and the subsequent development of MTS are dysgenesis, prior seizures (particularly febrile seizures), and genetic predisposition.


Figure 2. MR images demonstrating hippocampal swelling (24 h) and mild atrophy of the right hippocampus (9 months) after prolonged febrile status epilepticus in a 2.3-year-old girl.

Download figure to PowerPoint


  1. Top of page
  2. Abstract

The dysgenesis hypothesis purports that there is some preexisting structural abnormality in the brain that predisposes a child to febrile seizures, neuronal injury, and MTS. In addition to lowering the seizure threshold, dysgenesis is also thought to increase the susceptibility of the brain to the injurious effects of prolonged seizure activity and hence, MTS.

This discussion will be limited to microdysgenesis of the temporal lobe and hippocampal dysgenesis. Gross dysgenesis, such as hemimegalencephaly, gyral malformations, and gross heterotopias, is rarely a cause of mesial temporal lobe epilepsy, is uncommonly associated with MTS, and will not be discussed further. As the name implies, hippocampal dysgenesis involves structural abnormalities of one or both hippocampi. Microdysgenesis, as described in resected temporal lobe specimens of subjects with TLE, consists of tiny microscopic areas with varied structural abnormalities scattered throughout the temporal cortex and white matter. Multiple investigators have described microdysgenesis as a common accompaniment of TLE with MTS.

Four of the more common histological characteristics of microdysgenesis in the TLE literature include perivascular clustering of oligodendroglia in white matter, clustering of neurons, often with abnormal morphology, in cortical layers II–VI, an excess of heterotopic neurons (>10) in the deep white matter, and the presence of glioneuronal hamartia (8,9). The significance of these changes is mysterious and the causes are unknown. Although there are many published reports of microdysgenesis in patients with TLE, very few included blinded observers and randomized control tissue specimens for comparison. In one carefully executed study, Kasper et al. (9) examined tissue specimens from 29 control and 47 TLE cases for 15 separate features of microdysgenesis. The results revealed that only the four features named above were significantly increased in TLE specimens. Each of the four features was present in 17–30% of the TLE specimens and 30 of 47 patients with TLE had one or more. These data indicate that there was a strong association between microdysgenesis and TLE in this subpopulation of patients with MTS.

If microdysgenesis were playing a decisive role in increasing susceptibility to febrile seizures, neuronal injury, and MTS, one would expect to see microdysgenesis more frequently in TLE patients with a history of febrile seizures than in those without such a history. This, however, has not proven to be the case. Pathological reports from 33 children with TLE, 15 with and 18 without a history of febrile convulsions, demonstrated no significant difference in the incidence of microdysgenesis between the two groups (see Fig. 3) (10).


Figure 3. The frequency of microdysgenesis observed in TLE patients with (HxFC) and without (No FC) a history of febrile seizures [from Porter et al. 2003 (10)]. TLE = temporal lobe epilepsy.

Download figure to PowerPoint

Finally, the pathological significance of several of these abnormalities could be questioned based on recent neuropathological findings. Careful stereological study of surgical specimens has revealed an increase in the size and packing density of neurons in the temporal neocortex and white matter of patients with TLE and MTS (11). It seems possible that these changes, which accompany temporal lobe atrophy in TLE, could give the false impression of abnormal cortical neuronal clustering and increased white matter neuronal counts. Glioneuronal hamartias have also recently been reported in the amygdala and temporal horn of the lateral ventricle in a high percentage of nonepileptic controls (8).

Taken together, the available evidence suggests that of the role of temporal lobe microdysgenesis in febrile seizures and MTS must be viewed with some skepticism. Future studies will need to employ rigid controls to judge the significance of these features of microdysgenesis in TLE.

Hippocampal Dysgenesis

Dysgenesis limited to the hippocampus, usually noted on MRI, has been reported sporadically in subjects with TLE and MTS (12). However, in 1998, Fernandez and coworkers described two families in which multiple members had experienced at least one simple febrile seizure. All family members with febrile seizures, and some without, had characteristic MRI abnormalities of the left hippocampi. MRI volumetric measurements revealed smaller left hippocampi in all affected family members when compared with a control group. Affected members of Family A often had, in addition to smaller left hippocampi, loss of internal architecture and flattened hippocampal formations without abnormal T2 signal intensity. Only one member of each family had MTS and TLE (13). Each of these two subjects had experienced either 25 (Family B) or >50 (Family A) febrile seizures before 4 and 10 years of age, respectively, and had documented hippocampal sclerosis. These findings suggested that subtle, preexisting familial hippocampal dysgenesis may predispose individuals to febrile seizures and, in rare instances, to the subsequent development of MTS if febrile seizures are repetitive and numerous.

Careful and extensive pathological analysis of hippocampi removed at surgery may reveal that hippocampal dysgenesis is more common than thought. Sloviter et al. (14) recently described so-called “tectonic” hippocampal malformations involving CA1 and the subiculum consisting of nests of atypical neurons bulging out from these cell layers, compressing and displacing the dentate granule cell layer, hence the term tectonic. These areas were found scattered in hippocampi resected en bloc from subjects with temporal lobe epilepsy. However, the hippocampi with “tectonic” malformations either had mild cell loss or no apparent cell loss, that is, they did not have accompanying MTS either pathologically or by MRI. The role of the malformations in the epilepsy remains to be determined. Interestingly, a history of febrile seizures was no more common in the subjects with “tectonic” malformations than in a comparison group with MTS.

Data from animal studies might suggest a role for hippocampal dysgenesis in the etiology of MTS. In a series of experiments, Germano and colleagues (15,16) demonstrated that rat pups exposed to methylaxzoymethanol (MAM) during gestation display extensive cortical and hippocampal dysgenesis and a heightened vulnerability to hyperthermia-induced seizures. As shown in Fig. 4, the columnar organization of the cerebral cortex and the pyramidal cell layer of the hippocampus were significantly disrupted in the rats exposed to gestational MAM. Hyperthermia, produced by elevating core body temperature to >42°C for 90–150 s, induced cortical electrographic seizures in both MAM and control pups. However, MAM-exposed pups had a higher incidence of behavioral (clinical) seizures, probably due to increased spreading of seizure activity. Neuronal counts indicated that there was significant pyramidal cell loss in the CA1 and CA3/4 hippocampal subfields of MAM rats 4 weeks after exposure to hyperthermia and that the cell loss was independent of seizure activity. It was concluded that extensive dysgenesis lowers the threshold for hyperthermia-induced behavioral seizures and increases the vulnerability of the immature rat brain to irreversible neuronal injury.


Figure 4. Cortical (left panels) and hippocampal (right panels) dysgenesis observed in 14-day-old rat pups that were exposed to gestational MAM (14,15) [from Germano et al. 1996 (14) and Germono and Sperber, 1998 (15)]. Copyright © 1998 Wiley-Liss, Inc., A Wiley Company. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint

Although it seems likely that hippocampal dysgenesis could increase susceptibility to febrile seizures and MTS in humans, the reported instances are limited as yet and additional studies are needed to firmly establish a link. Data from animal models indicate that severe cortical and hippocampal dysgenesis can lower seizure thresholds and enhance hyperthermic neuronal injury (15,16). However, hyperthermia causes neuronal injury in rats with severe dysgenesis regardless of whether they have behavioral seizures (15). More studies are clearly needed to clarify the role of hippocampal dysgenesis in the etiology of TLE and MTS.

Febrile Seizures Produce Transient MRI Abnormalities and Lower Limbic Seizure Threshold

It has been suggested that prolonged febrile seizures may produce hippocampal injury and thereby increase hippocampal excitability leading to the gradual development of MTS and TLE. Evidence supporting this theory is derived from a sequence of animal studies by Baram and colleagues using an immature rat model of prolonged febrile seizures (17–20). The investigators argue that this is an age-appropriate model because 10-day-old rat pups (P10) are developmentally similar to a 1 to 2-year-old child, the age at which febrile seizures are most frequently seen (19,20). In this model, 98% of all rat pups exposed to hyperthermia (core body temperature = 42–43°C) on P10 develop prolonged limbic seizures lasting about 20 min. As shown in Fig. 5, febrile seizures produced transient hippocampal CA1 injury 24 h later, as indicated by increased pyramidal cell uptake of silver staining, but no permanent cell loss 3 months after the insult (17). Hence, prolonged febrile seizures did not produce MTS in these animals. They did, however, appear to permanently lower seizure thresholds in adulthood. Using the same model, Dube and colleagues demonstrated that adult rats with a history of prolonged febrile seizures displayed normal baseline hippocampal electrical activity and no evidence of spontaneous seizures. However, when challenged with a subthreshold dose of kainic acid, these rats displayed full-blown limbic behavioral seizures that progressed to status epilepticus in most cases (Fig. 6). In contrast, only a few control rats had brief seizures following kainic acid injection and none progressed to status epilepticus (18).


Figure 5. Prolonged experimental febrile seizures on P10 produce transient hippocampal CA1 injury but no permanent cell loss in immature rats. Left panel shows increased silver staining of pyramidal neurons, suggestive of acute injury, 24 h after prolonged seizures; Right panel demonstrates that total, pyramidal, and interneuron cell counts were similar between experimental and control rats 3 months later indicating that there was no permanent injury (16,18) [from Toth et al. 1998 (16) and Bender et al. 2003 (18)]. Copyright © 2003 Wiley-Liss, Inc., A Wiley Company. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint


Figure 6. EEG recordings from adult rats that were subjected to prolonged febrile seizures on P10 in comparison to controls. Left: Baseline recordings demonstrate normal hippocampal electrical activity and no evidence of spontaneous seizures in rats with a history of hyperthermia-induced seizures, rats with a history of hyperthermia or untreated control rats. Right: EEG recording demonstrating full-blown limbic seizure activity in experimental rats following a subthreshold dose of kainic acid; control rats never progressed to status epilepticus following kainic acid injection (17) [from Dube et al., 2000 (17)]. Copyright © 2000 Wiley-Liss, Inc., A Wiley Company. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint

Using small animal MRI techniques, Baram and colleagues have recently also documented transiently increased T2 signal between 24 h and 8 days after prolonged hyperthermic seizures in the amygdala, piriform cortex, and dorsal hippocampus in their P10 rat model (20). These findings provide a very interesting parallel to those observed in humans. Scott et al. have published similar results showing that febrile seizures, lasting 30–90 min (mean = 45 min), also produced transient MRI changes in the hippocampi of very young children (21). Pooled data from 21 infants indicated that hippocampal volume was significantly increased and T2 relaxation time significantly prolonged within 2 days after seizures in comparison with controls. Follow-up MRIs, 4–8 month later, revealed that both the volume and signal changes completely reversed in 14 infants. Two infants had volumetric asymmetry greater than control limits, but normal T2 relaxation times at follow-up (22).

Taken together, the results from human infant and P10 rat pup studies suggest that febrile seizures are associated with acute hippocampal injury and/or transient MRI abnormalities that resolve within several months of the insult. In animals, prolonged febrile seizures during the early postnatal period can also result in a permanent lowering of seizure threshold in adulthood, thus suggesting that it is possible to create an epileptogenic hippocampus in an apparently structurally normal brain. Could a similar enhancement of hippocampal excitability develop following prolonged febrile seizures in humans and lead to MTS, and TLE? This hypothesis is currently being tested in a multicenter clinical trial designed to assess the incidence, outcome, and risk factors for acute hippocampal injury, MTS, and TLE following febrile status epilepticus. Children will be recruited following febrile status epilepticus and followed long term with acute and chronic MRI imaging studies to determine risk factors for development of MTS and TLE.

Genetic Predisposition May Enhance Susceptibility to MTS

Using clinical criteria of mesial temporal lobe epilepsy, Cendes and coworkers identified 45 families that had at least two individuals with mesial temporal lobe epilepsy and termed this syndrome familial mesial temporal lobe epilepsy (FMTLE) (23–25). Clinical criteria included a history of simple or complex partial seizures with characteristics of temporal lobe origin (e.g., rising epigastric sensation, fear, experiential phenomenon, autonomic symptoms); Electroencephalographic (EEG) criteria included either classical temporal lobe spikes or intermittent slow wave abnormalities with episodes of rhythmic localized temporal delta activity. A normal EEG did not rule out FMTLE. Some affected individuals had medically intractable frequent seizures (i.e., refractory TLE), whereas others had infrequent seizures or had remitted entirely (i.e., benign TLE). MRIs performed on family members documented hippocampal atrophy in 99 of 142 subjects and increased T2 signal in 67 of 97 subjects (25). Analyses revealed that hippocampal atrophy and T2 signal changes correlated with the severity of epilepsy. Hippocampal atrophy was seen in 88% of cases with refractory TLE in comparison with 65% of benign cases. Increased T2 signal was also observed more frequently in cases with refractory TLE. However, atrophy and increased T2 signal were also observed in some family members who had benign TLE and occasionally even in members with no history of epilepsy at all (23,24). These results indicate that MTS in FMTLE does not develop as a consequence of repeated seizures although frequent seizures may cause progression of MTS. Furthermore, there was no evidence that febrile seizures played a critical role in the genesis of MTS in these families. Only 11% of the subjects with FMTLE had a history of febrile seizures, and the vast majority of these were simple febrile seizures. Although there was a trend for febrile seizures to be more frequent in family members with refractory TLE or hippocampal atrophy, it was not statistically significant (23).

These results suggest that there can be strong genetic factors that predispose some individuals to MTS such that, given the appropriate genotype, MTS can develop with no apparent IPI.


  1. Top of page
  2. Abstract

Current thinking posits that multiple factors are involved in the genesis of MTS as characterized by the “two-hit hypothesis.” At this time, however, it is unclear how each of these factors affects the vulnerability of an individual to insult. Additional well-controlled neuropathological studies are needed to clarify the role of microdysgenesis in MTS. Additional MRI and neuropathological data are needed to confirm the role of preexisting familial or sporadic hippocampal dysgenesis in MTS and TLE. Although an association between hippocampal and cortical dysgenesis and a subsequent increase in the susceptibility to hyperthermic seizures and neuronal injury has been documented in animals, no such relationship has been demonstrated in humans. Finally, even though experimental evidence indicates that prolonged febrile seizures can produce an epileptogenic but structurally normal hippocampus in a previously normal brain, this phenomenon has not been documented in man. The ongoing prospective clinical study of febrile status epilepticus described earlier in this article may help to clarify the relationship between prolonged febrile seizures, evidence of hippocampal injury or dysgenesis on MRI, genetic predisposition, and ultimate development of MTS and TLE.


  1. Top of page
  2. Abstract
  • 1
    Cendes F, Andermann F, Dubeau F, et al. Early childhood prolonged febrile convulsions, atrophy and sclerosis of mesial structures, and temporal lobe epilepsy: an MRI volumetric study. Neurology 1993;43: 10837.
  • 2
    French JA, Williamson PD, Thadani VM, et al. Characteristics of medial temporal lobe epilepsy: I. Results of history and physical examination. Ann Neurol 1993;34: 77480.
  • 3
    Salvanova V, Markand ON, Worth R. Clinical characteristics and predictive factors in 98 patients with complex partial seizures treated with temporal resection. Arch Neurol 1994;51: 100813.
  • 4
    Lewis DV. Febrile convulsions and mesial temporal lobe sclerosis. Curr Opin Neurol 1999;12: 197201.
  • 5
    Cendes F, Andermann F. Do febrile seizures promote temporal lobe epilepsy? In: BaramRZ, ShinnarS, eds. Febrile seizures. San Diego , CA : Academic Press, 2002: 7886.
  • 6
    Hesdorffer DC, Hauser WA. Febrile seizures and the risk of epilepsy. In: BaramTZ, ShinnarS, eds. Febrile seizures. San Diego : Academic Press, 2002;6376.
  • 7
    Nelson KB, Ellenberg JH. Predictors of epilepsy in children who have experienced febrile seizures. N Engl J Med 1976;295: 102933.
  • 8
    Yachnis AT, Roper SN, Love A, et al. Bci-2 immunoreactive cells with immature neuronal phenotype exist in the nonepileptic adult human brain. J Neuropathol Exp Neurol 2000;59: 1139.
  • 9
    Kasper BS, Stefan H, Buchfelder M, et al. Temporal lobe microdysgenesis in epilepsy versus control brains. J Neuropathol Exp Neurol 1999;58: 228.
  • 10
    Porter BE, Judkins AR, Clancy RR, et al. Dysplasia: a common finding in intractable pediatric temporal lobe epilepsy. Neurology 2003;61: 3658.
  • 11
    Bothwell S, Meredith GE, Phillips J, et al. Neuronal hypertrophy in the neocortex of patients with temporal lobe epilepsy. J Neurosci 2001;21: 4789800.
  • 12
    Lehericy S, Dormont D, Semah F, et al. Developmental abnormalities of the medial temporal lobe in patients with temporal lobe epilepsy. AJNR Am J Neuroradiol 1995;16: 61726.
  • 13
    Fernandez G, Effenberger O, Vinz B, et al. Hippocampal malformation as a cause of familial febrile convulsions and subsequent hippocampal sclerosis. Neurology 1998;50: 90917.
  • 14
    Sloviter RS, Kudrimoti HS, Laxer KD, et al. “Tectonic” hippocampal malformations in patients with temporal lobe epilepsy. Epilepsy Res 2004;59: 12353.
  • 15
    Germano IM, Zhang YF, Sperber EF, et al. Neuronal migration disorders increase susceptibility to hyperthermia-induced seizures in developing rats. Epilepsia 1996;37: 90210.
  • 16
    Germano IM, Sperber EF. Transplacentally induced neuronal migration disorders: an animal model for the study of the epilepsies. J Neurosci Res 1998;51: 47388.
  • 17
    Toth Z, Yan XX, Haftoglou S, et al. Seizure-induced neuronal injury: vulnerability to febrile seizures in an immature rat model. J Neurosci 1998: 18: 428594.
  • 18
    Dube C, Chen K, Eghbal-Ahmadi M, et al. Prolonged febrile seizures in the immature rat model enhance hippocampal excitability long term. Ann Neurol 2000;47: 33644.
  • 19
    Bender RA, Dube C, Gonzalez-Vega R, et al. Mossy fiber plasticity and enhanced hippocampal excitability, without hippocampal cell loss or altered neurogenesis, in an animal model of prolonged febrile seizures. Hippocampus 2003;13: 399412.
  • 20
    Dube C, Yu H, Nalcioglu O, et al. Serial MRI after experimental febrile seizures: altered T2 signal without neuronal death. Ann Neurol 2004;56: 70914.
  • 21
    Scott RC, Gadian DG, King MD, et al. Magnetic resonance imaging findings within 5 days of status epilepticus in childhood. Brain 2002;125: 19519.
  • 22
    Scott RC, King MD, Gadian DG, et al. Hippocampal abnormalities after prolonged febrile convulsion: a longitudinal MRI study. Brain 2003;126: 25517.
  • 23
    Kobayashi E, Lopes-Cendes I, Guerreiro CA, et al. Seizure outcome and hippocampal atrophy in familial mesial temporal lobe epilepsy. Neurology 2001;56: 16672.
  • 24
    Kobayashi E, Li LM, Lopes-Cendes I, et al. Magnetic resonance imaging evidence of hippocampal sclerosis in asymptomatic, first-degree relatives of patients with familial mesial temporal lobe epilepsy. Arch Neurol 2002;59: 18914.
  • 25
    Kobayashi E, D'Agostino MD, Lopes-Cendes I, et al. Hippocampal atrophy and T2-weighted signal changes in familial mesial temporal lobe epilepsy. Neurology 2003;60: 4059.