Febrile seizures and mesial temporal sclerosis
No association in a long-term follow-up study
Citation Manager Formats
Make Comment
See Comments

Abstract
Objective: To determine whether febrile seizures cause mesial temporal sclerosis (MTS), the occurrence of MTS was evaluated in an unselected series of patients with febrile seizures.
Methods: Twenty-four patients with a prolonged first febrile seizure, 8 with an unprovoked seizure after the first febrile seizure, and 32 age-, sex-, and handedness-matched control subjects with a single simple febrile seizure without later unprovoked seizures were selected from 329 febrile seizure patients followed up prospectively. The occurrence of MTS was evaluated after a mean follow-up time of 12.3 years by MR volumetry of amygdala and hippocampal formation and qualitative analysis of mesial temporal structures.
Results: None of the patients had MTS. The mean total volumes of the right and left hippocampal formations and amygdala did not differ significantly between any of the three groups. The qualitative analysis revealed no sclerotic changes in the mesial temporal area. The patients with a prolonged initial febrile seizure had a lower mean right–left volume difference in hippocampal formations than the control subjects, but this had no effect on the outcome.
Conclusion: The occurrence of MTS following even prolonged febrile seizures is an uncommon event, confirming the good clinical outcome of febrile seizures.
Mesial temporal sclerosis (MTS) is the most common lesion found in patients with temporal lobe epilepsy (TLE).1 Approximately one third of patients with drug-resistant TLE have experienced seizures during fever, which were interpreted to have been febrile seizures (often complicated). These patients have a higher incidence of not only hippocampal sclerosis but also more severe changes in the mesial temporal area than the TLE patients with no feverish convulsive events during childhood.2,3⇓ These reports nevertheless concern highly selected patients with drug-resistant TLE who have been analyzed retrospectively, and thus the finding of a relationship of febrile seizures to MTS may be biased. It may well be that the observed association is a manifestation of increased seizure susceptibility linked to previous mesial temporal pathology.
Van Landingham et al.4 observed a high incidence of hippocampal intensity or volume changes in MRI immediately after a complicated febrile seizure (10/27 patients). Nine of the 10 patients with focal features in the prolonged febrile seizure had changes in at least one hippocampal formation; 4 of these were signs of acute edema in the hippocampal formation of the hemisphere in which the seizure originated. Their patients had a high incidence of abnormalities in developmental status (22%), and some had had very long focal febrile seizures lasting up to 216 minutes, which may have contributed to their findings.
To clarify the relationship of febrile seizures to MTS further, we evaluated it in patients who were followed up prospectively from the time of their first febrile seizure.
Patients and methods.
Patients.
The series consisted of 329 unselected patients with febrile seizure who had participated in our clinical studies evaluating factors triggering the first febrile seizure, risk factors, or prophylactic treatment of recurrences of febrile seizures at the Department of Pediatrics, University of Oulu, Finland, during the years 1984 to 1990. The detailed monitoring and follow-up schedules have been reported earlier.5-7⇓⇓ All the patients were followed prospectively after their first febrile seizure. In short, the criteria for eligibility were no previous seizure disorder or progressive neurologic disease and no signs of intracranial infection. The parents of the patients were interviewed for background information, including a description of the characteristics of the initial febrile seizure, and a baseline neurologic evaluation was performed.
We hypothesized that the 8 patients who had had an unprovoked seizure after the initial febrile seizure and the 30 patients with a prolonged first febrile seizure (duration of ≥30 minutes) would have the greatest risk of developing MTS. It was therefore these patients who were offered an opportunity to participate in the current follow-up survey. One patient met both criteria. Children whose first febrile seizure lasted <30 minutes were not included even if the seizure was focal (11 patients) or multiple (45 patients). Of the eight patients with later unprovoked seizures, three had complex partial seizures, two had rolandic epilepsy, one had myoclonic seizures, and one experienced a single unprovoked focal seizure with secondary generalization. Three patients with a prolonged initial febrile seizure could not be reached, and three others chose not to participate. The final number of patients was 32. An age-, sex-, and handedness-matched control with a simple febrile seizure (duration of ≤15 minutes, one seizure during the fever episode, and no focal features) with no recurrences or unprovoked attacks was selected for each patient.
The mean age of the patients with a prolonged initial febrile seizure at the time of the MRI examination was 14.4 years (range 9.9 to 20.2 years), that of the patients with a later unprovoked seizure being 12.5 years (10.4 to 14.2 years), and that of the control patients 14.2 years (10.3 to 20.4 years). The mean follow-up times in these groups were 12.5 (8.5 to 14.7), 11.2 (8.9 to 12.6), and 12.5 years (9.6 to 14.7 years) (table 1).
Table 1 Descriptive data on patients with initial prolonged or simple febrile seizure (FS) and those with later unprovoked seizures
In the neurologic evaluation, data on the patient’s previous medical history and school performance (special school attendance, personal assistant teacher, or repeating a school grade) were collected from the medical records and by interviewing the patients. A thorough neurologic assessment was conducted by one of the authors.
The protocol was accepted by the ethical committee of the Medical Faculty, University of Oulu, and informed consent was obtained from the parents or the patients as appropriate.
MRI protocol.
MRI was performed using a 1.5 T scanner (Signa EchoSpeed; General Electric Medical Systems, Milwaukee, WI). T1-weighted sagittal images (repetition time [TR] 500 milliseconds, echo time [TE] 8 milliseconds, 23-cm field of view [FOV], 5-mm slice, 2-mm gap, 256 × 224 matrix, 2 excitations) were obtained together with double fast spin echo T2-weighted axial (TR 3,500 milliseconds, TE 14/98 milliseconds, echo train length [ET] 8, 23 × 17-cm FOV, 5-mm slice, 1-mm gap, 256 × 224 matrix, 2 excitations) and coronal (TR 4,000 milliseconds, TE 24/96 milliseconds, ET 8, 22 × 16-cm FOV, 3-mm slice, 0-mm gap, 256 × 256 matrix, 1 excitation) slices. T2-weighted axial images were obtained parallel to the temporal lobes and coronal images perpendicular to them. A three-dimensional (3D) coronal spoiled gradient echo (SPGR) series was also obtained (TR 40 milliseconds, TE 7 milliseconds, 23 × 17-cm FOV, 2-mm slice, 256 × 192 matrix, 1 excitation).
The 3D-SPGR images, which provide high gray matter and white matter contrast, were transferred to a workstation for volumetry. First, 2-mm-thick reformatted images were generated perpendicular to the hippocampal formations, and then the volumes of both amygdala and hippocampal formations were measured on the reformatted images by one radiologist who was blinded to the clinical history of the subjects. The boundaries of these structures were defined according to previous reports.8 The house-made software uses a semiautomated technique, combining tracing and threshold approaches.
Two radiologists visually evaluated all MRIs. The analyses were first made separately, after which a consensus was reached. Special attention was paid to the size, shape, and signal intensity of the hippocampal formations.
An average right–left absolute hippocampal volume difference of 0.2 cm3 has been reported in healthy adults.9 With use of standard deviations from this series, lower and upper threshold values of −0.2 and 0.6 cm3 were calculated and then used to classify the patients into groups with right-sided hippocampal atrophy, nonlateralizing volume difference, and left-sided hippocampal atrophy.10 As there are no normal values for adolescent patients, we used our control patient group with a single simple febrile seizure as a source for reference values. Our finding of a mean right–left hippocampal volume difference of 0.18 cm3 is quite close to the above value of 0.2 cm3. The right–left hippocampal volume difference threshold values in our adolescent control patient group (2-SD range) were −0.13 and 0.49 cm3.
Statistics.
The data were stored in a microcomputer and analyzed with SPSSWin (version 9; SPSS, Chicago, IL). The mean absolute hippocampal formation and amygdala volumes and the mean right–left volume difference in hippocampal formations were compared between the groups with a one-way analysis of variance. In the case of a significant result, the analyses were continued using Bonferroni post hoc multiple comparison. The Kruskal–Wallis test was used to compare the numbers of patients with problems in school performance or abnormal neurologic findings between the groups. The significances of the differences between pairs of parametric variables were assessed with the Student’s t-test.
Results.
No sclerosis was found in the qualitative evaluations of the mesial temporal area, and there were no significant differences in the absolute right or left hippocampus or amygdala volumes between the patients with a prolonged initial febrile seizure or an unprovoked seizure after the initial febrile seizure and those with a single simple febrile seizure. The mean of the right–left hippocampal volume difference in the patients with later unprovoked seizures did not differ from that of the control subjects (table 2). The mean difference was lower in the patients with a prolonged initial febrile seizure, however, that is, 0.039 cm3 as compared with 0.18 cm3 (p < 0.01) in the control subjects. When the initial prolonged seizure was focal, the mean difference was even lower, that is, −0.081 vs 0.18 cm3 (p < 0.001), but the other parameters were similar as in children with symmetric prolonged seizures. The results of a comparison of the mean greater–smaller hippocampal volume difference between the groups were similar to the analysis using the mean right–left hippocampal volume difference. According to our threshold values, three patients from the prolonged febrile seizure group had a smaller right hippocampal formation (right–left volume difference ranging from −0.47 to −0.15 cm3). One left-handed control patient had a volume difference of 0.52 cm3. All these four patients were neurologically normal without seizures.
Table 2 Results of MRI volumetry among children with initial prolonged or simple febrile seizure and those with later unprovoked seizures
The school performance and neurologic findings of the patients with a prolonged initial febrile seizure were as good as those of the control subjects, and the children with problems in school performance or abnormal neurologic findings had similar hippocampal formation and amygdala volumes and right–left volume differences in the hippocampal formations to the patients with a normal neurologic status and no problems in school performance. Three patients with a prolonged initial febrile seizure, three patients with a later unprovoked seizure, and eight control subjects had mild abnormalities in motor function. Three patients in the prolonged febrile seizure group, two in the unprovoked seizure group, and one control had a history of problems in school performance, but there were no significant differences in school performance between the patients with prolonged primary febrile seizure, those with a later unprovoked seizure, and the control subjects.
The initial febrile seizure was multiple in three children. Their hippocampal and amygdala volumes and right–left hippocampal volume differences did not differ from the control values. Similarly, these measurements did not differ in four children who had up to 10 recurrent febrile seizures during the follow-up.
In the children with a prolonged initial seizure, the seizure lasted 30 to 59 minutes in 18 children, 60 minutes in 5 children, and 150 minutes in 1 child. The measured volumetric parameters did not differ between these children. Even the child with the longest initial febrile seizure had normal hippocampal volumes, and his right–left hippocampal volume difference of −0.03 cm3 was within our normal limits. The duration of focal prolonged seizures was 30 to 59 minutes in seven children and 60 minutes in one child.
Discussion.
We found no evidence of MTS in these patients with different forms of febrile seizures, indicating that the occurrence of MTS following even prolonged febrile seizures is an uncommon event. Owing to the relatively small sample size of children with very prolonged febrile seizures, their role as a risk for MTS could not be excluded with certainty. The mean absolute volumes of amygdala and hippocampal formation and the right–left volume difference in hippocampal formation were compared with values for age-, sex-, and handedness-matched patients with a single simple febrile seizure. These patients were chosen as controls because a simple febrile seizure is unlikely to cause permanent structural brain damage, and we had detailed follow-up data on the children recorded in a similar manner to that applying to the patients.
We compared the absolute hippocampal formation and amygdala volumes of the patients with the results for age- and sex-matched control subjects to exclude the possibility of bias caused by the effect of head size and variation in the developmental phase of the patients on absolute structure volumes. The right–left hippocampal volume difference is independent of total intracranial volume and thus comparable between subjects.9 Using our control patient group with a single simple febrile seizure as the source of reference values for the right–left hippocampal volume difference, we found four patients with data outside these values, all of whom were neurologically normal without seizures. The clinical significance of these findings is not known at the moment, and a follow-up of these patients is warranted to find out if the changes will be indicative for unilateral hippocampal atrophy, as pointed out by Jack et al.9 Although MRI volume studies are the most sensitive noninvasive detectors of moderate or severe MTS, it is not known if subtle forms of MTS can be detected by these studies.9
Right–left volume asymmetry in patients with severe TLE and unilateral hippocampal sclerosis has been linked to hippocampal atrophy.11-13⇓⇓ Thus, our sole finding of a significantly lower hippocampal right–left volume difference without any decrease in the mean absolute hippocampal volumes among the patients with a prolonged initial febrile seizure compared with the matched control subjects cannot be taken as a sign of structural damage.
Multiple initial febrile seizures and recurrent febrile seizures had no effect on hippocampus in our series. However, the small sample size of children with these features prevents us from drawing any definitive conclusions about their role.
MTS is found in only about 1% of children with newly diagnosed epilepsy,14,15⇓ but no long-term follow-up studies about its occurrence in an unselected material of children with febrile seizures have been previously performed. Although our sample size was relatively small, our results support the conclusion that there is no causal relation between febrile seizures and MTS in an unselected series of febrile seizure patients. We had a population-based series followed up prospectively from the first febrile seizure, from which we selected the cases most susceptible to the possible adverse sequelae of febrile seizure, namely, those with a prolonged initial febrile seizure or unprovoked recurrences. It is improbable that new cases with epilepsy will arise later among these febrile seizure patients because usually the epilepsy begins during the first 9 years after the febrile seizure and our follow-up time was over 12 years.16
Our negative result and the previous finding of a linkage between a history of febrile seizures and MTS in patients with severe TLE are not necessarily in contradiction, even though we think that there is no causal relationship between febrile seizures and MTS. Conclusions concerning the outcome of febrile seizures per se cannot be reached from highly selected TLE patient series. In patients with severe TLE, the fever may have been the first trigger leading to the seizures, as the structural changes in the mesial temporal area involved in MTS may well have already developed earlier.17
As we found that febrile seizures do not cause either neurologic sequelae or brain damage, the current approach and recommendations for the treatment and follow-up of febrile seizures may be regarded as sufficient. Yet, it is important to prepare the families for possible recurrences and to give them sufficient first aid instructions on how and when to give medication to avoid the prolongation of seizures combined with the valuable information that febrile seizures have an excellent clinical outcome.
Acknowledgments
Supported by the Arvo and Lea Ylppö Foundation, the Research Foundation of the Orion Corporation, the Foundation for Pediatric Research, and the Alma and K.A. Snellman Foundation.
- Received May 10, 2002.
- Accepted August 21, 2002.
References
- ↵
- ↵
Cendes F, Andermann F, Gloor P, et al. MRI volumetric measurement of amygdala and hippocampus in temporal lobe epilepsy. Neurology . 1993; 43: 719–725.
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
Watson C, Andermann F, Gloor P, et al. Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic resonance imaging. Neurology . 1992; 42: 1743–1750.
- ↵
- ↵
- ↵
- ↵
Barr WB, Ashtari M, Schaul N. Bilateral reductions in hippocampal volume in adults with epilepsy and a history of febrile seizures. J Neurol Neurosurg Psychiatry . 1997; 63: 461–467.
- ↵
- ↵
- ↵
Berg AT, Testa FM, Levy SR, Shinnar S. Neuroimaging in children with newly diagnosed epilepsy: a community-based study. Pediatrics . 2000; 106: 527–532.
- ↵
Annegers JF, Hauser WA, Elveback LR, Kurland LT. The risk of epilepsy following febrile convulsions. Neurology . 1979; 29: 297–303.
- ↵
Fernández G, Effenberger O, Vinz B, et al. Hippocampal malformation as a cause of familial febrile convulsions and subsequent hippocampal sclerosis. Neurology . 1998; 50: 909–917.
Disputes & Debates: Rapid online correspondence
- Febrile seizures and mesial temporal sclerosis: No association in a long-term follow-up study
- Rod Scott, Great Ormond Street Hospital United Kingdomr.scott@ich.ucl.ac.uk
Submitted March 07, 2003 - Reply to Letter to the Editor
- Heikki Rantala, University of Oulu Finlandheikki.rantala@oulu.fi
- Matti Uhari
Submitted March 07, 2003
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Related Articles
- No related articles found.