Prognostic factors in patients with mesial temporal lobe epilepsy


Address correspondence to Paolo Tinuper M.D., Department of Neurological Sciences, University of Bologna, Via Ugo Foscolo 7, 40123 Bologna, Italy. E-mail:


Purpose: To disclose clinical, electrophysiologic, and neuroradiologic factors correlated to prognosis in patients with mesial temporal lobe epilepsy (MTLE).

Methods: One hundred thirty-six MTLE patients were studied for family history, clinical characteristics, instrumental data [electroencephalography (EEG), video-EEG, neuroimaging], and outcome. The population was divided into drug-resistant (DR: 108 patients, 79.4%) and non–drug-resistant (NDR: 28 patients, 20.6%) groups; all variables were analyzed in the two groups.

Results: The comparison between the two groups shows a relation between resistance to therapy and febrile seizures (FS) (DR 43.5% vs. NDR 17.8%, p = 0.008), mesial temporal sclerosis (MTS) (DR 64.8% vs. NDR 32.1%, p = 0.0025), early age at seizure onset (DR 23.1% vs. NDR 3.6% p = 0.0160), and epileptiform interictal abnormalities (DR 89.7% vs. NDR 68%, p = 0.010). FS were more frequent in patients with MTS than in patients without (46.28% vs. 26.3%, p = 0.0199). Sixty-nine patients underwent surgery and 85.3% of them had a good outcome.

Conclusion: MTLE is a heterogeneous syndrome. Establishing the factors responsible for and associated with drug resistance is important for therapeutic purposes, as prompt diagnosis of drug resistance must lead to early surgical management. This study shows that FS, MTS, early age at seizure onset, and epileptiform interictal abnormalities are negative prognostic factors and that FS are related to MTS.

Mesial temporal lobe epilepsy (MTLE) is the most common form of partial epilepsy in adults (Engel et al., 1997). The seizures are typically characterized by autonomic and/or psychic subjective symptoms (most commonly a sensation of something rising from the stomach), staring, or gestural or oral-alimentary automatisms. Aura may be isolated or may be followed by staring or secondary generalization. The seizures are often resistant to antiepileptic drugs (Engel et al., 1997), even though benign cases have been reported (Labate et al., 2006). Establishing the etiologic and physiologic mechanisms underlying this form of epilepsy, namely the factors responsible for and associated with drug resistance, is important for therapeutic purposes. Drug resistance must be diagnosed as early as possible because the ensuing seizures can be eliminated surgically by anterior temporal lobectomy in a high percentage of patients (70–90%) (Thadani et al., 1995). Although mesial temporal sclerosis (MTS) is the most common lesion identified in these patients, the role of this lesion in the development of epilepsy refractory to medical treatment remains unsettled (Labate et al., 2006). Clinically, MTLE is often associated with a personal history of febrile convulsions and the onset of drug-resistant seizures during adolescence or adulthood. Prolonged febrile seizures (FS) during infancy have been associated with severe damage to the temporomesial structures (Cendes et al., 1993). Nonetheless, the role of FS in the development of MTLE remains controversial (Tarkka et al., 2003). The purpose of our study was to research clinical, semiological, electrophysiologic, and neuroradiologic features related to prognosis in a large group of patients with MTLE.


Among a population of patients with temporal lobe epilepsy enrolled at the epilepsy centers participating in the study, we selected those responding to the following inclusion criteria:

  • 1 A clinical history of seizures with autonomic features (i.e., epigastric sensation) as the first ictal symptom in a patient able or not to verbally refer the beginning of a seizure, followed or not by motor arrest or staring with mainly oro-alimentary automatisms;
  • 2 Interictal or ictal electroencephalography (EEG) epileptiform abnormalities over temporal or frontotemporal regions;
  • 3 High-resolution magnetic resonance (MR) scanning of the brain, including three spatial planes with T1-weighted, IR (inversion-recovery) and possibly fFLAIR (fast fluid-attenuated inversion-recovery) coronal views.

Possible candidates were identified by both (1) a prospective observation of patients admitted for the first time to epilepsy centers participating in the study from January 2004 to January 2008, and (2) retrospectively reviewing all clinical records from the databases of the centers. All candidates in the second group with long-lasting disease followed for many years had a seizure diary and clinical evaluation visits at least every 6 months. The data from each patient were tabulated with full details of seizure type, auras, seizure frequency, response to therapy, antiepileptic drug (AED) used, ictal and interictal EEG recording, neuroradiologic findings, and presurgical and surgical evaluation with histopathologic data. Other clinical items tabulated were sex, dominance, family history, personal antecedents, age at seizure onset, delay between seizure onset and diagnosis, follow-up duration, delay between onset of partial seizures with autonomic symptoms and secondarily generalized tonic or tonic–clonic seizures, family history (for FS and/or epilepsy), and potential personal antecedent events such as severe asphyxia, severe head injury, viral encephalitis, bacterial meningitis or cerebral abscess, and FS. FS were differentiated according to Hirtz et al. (1997) into simple (lasting <10 min), complex (lasting from 10–29 min), and febrile status epilepticus (seizures lasting more than 30 min). All patients underwent repeated EEG recordings during wakefulness and, when possible, during diurnal sleep induced by sleep deprivation. Clinical, neuroradiologic, and electrophysiologic data were tabulated with full details in a database developed ad hoc for this study. On the basis of the outcome after medical treatment, patients were divided into drug-resistant (DR) and resistant (NDR). We defined drug resistance according to Perucca (1998) as the persistence of seizures after adequate trials with at least three antiepileptic drugs used in monotherapy or in polytherapy at the maximally tolerated dosage (excluding situations where a meaningful dose escalation is prevented by allergic or idiosyncratic adverse reactions) and when, by their nature and frequency, the seizures compromise quality of life (Perucca, 1998). We considered responsive (NDR) those patients who were seizure free for at least 2 years. We studied the relationship with drug resistance and all the clinical and instrumental variables examined. Patients with a brain lesion detectable on MRI were defined as lesional (LES); patients without clear-cut neuroradiologic abnormalities, including patients with aspecific alterations (i.e., mild cortical atrophy or leukoaraiosis), were deemed nonlesional (NLES). The DR patients were differentiated into operated and not-operated, depending on surgical treatment. We defined as familiar (F) those patients with at least a first- and second-degree relative with a history of epilepsy or FS, and sporadic (S) cases without first- and second-degree relatives with epilepsy or FS. As to the frequency of the secondarily generalized seizures (SG), we considered “frequent SG” as daily, weekly, monthly; “rare” SG as annual, sporadic, unique. Surgical outcome was based on Engel’s classification using all 13 subclasses (Engel, 1987).

Statistical analysis

A descriptive analysis was made of the occurrence of the main clinical factors in the two groups (DR and NDR cases). The Fisher’s exact test was used to test the differences for dichotomous data.


We collected data on 230 cases out of 8,000 epileptic patients admitted to the centers involved in the study. Among them, 136 fulfilled the inclusion criteria (63 males, 73 females). Age at time of evaluation ranged between 5 and 80 years (mean ± SD: 31 ± 11.8); mean duration of follow-up was 6.9 ± 6.3 years (range 1–40); mean duration of disease was 22.9 ± 12.6 years (range 1–55). One hundred and eight patients (79.4%) were DR and 28 (20.6%) NDR. Comparing DR and NDR groups, the DR group had a higher frequency of the following variables: FS history (43.5% vs. 17.8%, p = 0.008), MTS (64.8% vs. 32.1%, p = 0.0025), age at seizure onset before 3 years (23.1% vs. 3.6% p = 0.0160), and persistence of epileptiform interictal abnormalities after medical treatment (89.7% vs. 68%, p = 0.010). Among the DR group, 85 patients showed a lesion on neuroimaging. In the LES group 79 had MTS (80.6%). MTS was more frequent in the DR group than in NDR (64.8% vs. 32.1%). Table 1 shows the main features in the DR and NDR groups. Table 2 reports findings on FS and positive family history for epilepsy and FS in patients with MTS.

Table 1.   Main features of the DR (108 patients, 79.4%) and NDR (28 patients, 20.6%) groups
 TotalDRDR (%)NDRNDR (%)p-value
  1. FS, febrile seizures; DR, drug-resistant; MTS, mesial temporal sclerosis; NDR, responsive; SE, status epilepticus; SG, generalized seizures. Bold values are statistically significant (p<0.05).

Dominant hand
 Right 1239890.72589.30.596
 Both hands432.314.2
Family history for epilepsy191614.8310.70.763
Family history for FS161110.3517.90.323
 FS simple302655.3480 
 FS complex121123.4120 
Pregnancy abnormal1076.5310.70.429
Delivery dystocic242119.4310.70.406
Age at onset
 0–3 years272623.113.60.0160
 4–79 years1098275.92796.4
SG frequency at onset
Status epilepticus1098.313.60.687
EEG interictal pathologic (epil + aspecific)13210799.12589.30.027
Epileptiform abnormalities1139689.717680.010
 Bilateral 111.2
 Right 484249.4633.3
 Left 302731.8316.7
 Dual pathology11912.9222.2
Table 2.   Relation between MTS, FS, and family history for epilepsy or FS
 MTS+MTS+ (%) MTS−MTS− (%)p-value
  1. FS, febrile seizures; MTS, mesial temporal sclerosis; SE, status epilepticus. Bold value is statistically significant (p<0.05).

 FS simple2362.2746.7 
 FS complex924.33200.760
Family history for epilepsy1215.2712.30.8029
Family history for FS78.9915.80.282

Surgery and histopathologic data

Sixty-nine patients underwent surgery. Histopathologic findings were available in 63 operated patients and disclosed MTS in 45 cases (71.4%), of which 38 had been previously detected by MRI (84.4%). MTS was absent in 18 operated patients (28.6%), and in half of these MTS had been previously reported after MRI study. Postsurgical follow-up was 58 ± 45.58 months (range: 5–227 months). Among 69 patients operated, 59 were in Engel class I (85.5%), 10 patients were in classes II–IV (14.5%). Among 45 patients with histopathologically proven MTS, 41 had outcome class I (91%), whereas four patients (9%) were in the other classes.


We presented a population of 136 cases with MTLE, selected on the bases of semiological and neurophysiologic features, to research factors related to prognosis. A large percentage of patients in our series were drug resistant (79.4% DR vs. 20.6% NDR), slightly more than reported in the literature (Currie et al., 1971). This difference could be related to the fact that we included many patients referring to third level centers specialized in presurgical or surgical evaluation in severe epilepsies. We did not find any differences between DR and NDR patients regarding general population characteristics like sex and dominant hand. A slight prevalence of family history of epilepsy in DR was evident, even if not significant. We found a higher frequency of FS in the personal history of DR patients than in NDR ones (p = 0.008). Hitiris et al. (2007) observed that FS were associated with drug-resistance in a general population of 780 newly diagnosed epileptic patients. The percentage of patients with FS rose to 67% in a study on patients with MTLE (French et al., 1993). Our study found a more frequent personal history of FS in patients with MTS than in patients without MTS (46.8% vs 26.3%, p = 0.0199). The strong association between FS and MTS is well documented in clinical (Harvey et al., 1995) and surgical (Thadani et al., 1995) reports. Cendes observed that the amygdala and hippocampus were smaller in patients with antecedents of prolonged FS compared with patients without. This suggests that MTS could be the consequence of prolonged FS over a particularly vulnerable period of infancy or that hippocampal structures could be sensitized to generate prolonged seizures by the action of some prenatal or perinatal insults (Cendes et al., 1993). Nevertheless other authors disagree on the role played by FS in the etiopathogenesis of MTS. A prospective study (follow up: 12 years) found that the development of MTS is not common in children with a history of simple or prolonged FS (Tarkka et al., 2003). Children who underwent an MRI study 48 h after a prolonged febrile convulsion had hippocampal abnormalities consistent with hippocampal edema (Scott et al., 2003). However, follow-up MRI disclosed only that an asymmetry between the hippocampal bodies persisted (with smaller volume on the side affected by edema) without MR signs of MTS, suggesting that the hippocampal injury associated with prolonged FS may include a spectrum of alterations from subtle neuronal injury to MTS, perhaps depending on a selective vulnerability in some patients (Scott et al., 2003). This vulnerability could be related to a genetic predisposition, that is, the involvement of γ-aminobutyric acid (GABA)B receptor 1 polymorphism (Kauffman et al., 2008). Another controversial point is the causative relationship between MTS and DR found in our study (p = 0.0025). Many reports emphasize the role of MTS as the main factor responsible for poor outcome in MTLE (Harvey et al., 1995), either alone or associated with dual pathology (Semah et al., 1998). Engel et al. (1997) defined MTLE associated with hippocampal sclerosis as one of the most medically refractory forms of human epilepsy, and histopathologically documented MTS is reported in 50–90% of patients operated for intractable MTLE (Thadani et al., 1995). Nevertheless, the role of MTS in DR is still debated, as MTS has also been found in patients with a good prognosis (Labate et al., 2006). Unlike other previous series, our study is more representative of the general population of MTLE patients observed in clinical practice, because selection was based solely on seizure semiology and not on neuroradiologic findings or the temporal evolution of disease.

Two other minor factors related to a poor outcome emerge from our study: an early age at onset of epilepsy (p = 0.0160) and interictal epileptiform EEG abnormalities (p = 0.010). Abnormal (epileptiform) tracings have frequently been associated with a higher risk of poor outcome, but the methodologic limitations of the studies, namely discrepancies in EEG recording, the use of different criteria to define EEG abnormalities, and the differing interpretations of the tracing, mean that EEG data do not serve for prognosis.

MTLE is a heterogeneous syndrome. Establishing the factors associated with DR is important for therapeutic purposes, as prompt diagnosis of drug resistance must lead to an early surgical approach. Our study on a large and heterogeneous population with either benign or intractable MTLE identified MTS, FS, early age at seizure onset, and persistent epileptiform interictal abnormalities as negative prognostic factors. One of the limitations of our study is that a large number of patients came from third-level epilepsy centers, making the DR population larger. A prospective longitudinal study is needed to verify the role of these findings in the natural history of MTLE.


We wish to thank Mrs Alessandra Laffi for assistance. This study was supported by “Ministry of University and Research grant prot. 2003068749_003.”

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Disclosure: None of the authors has any conflict of interest to disclose.