Hippocampal damage after prolonged febrile seizure:
One casein a consecutive prospective series
*yTakuya Tanabe,*Keita Hara,zShuichiShimakawa,zMihoFukui, andzHiroshiTamai
*Departmentof Pediatrics,HirakataCityHospital,Osaka,Japan;yDepartmentof PediatricNeurology, Tanabe-Kadobayashi
The factors that contribute to hippocampal damage as
a sequela, and its frequency, in patients experiencing feb-
rile status epilepticus, remain unknown. Of the 472
patients with febrile seizures admitted to our hospital
between February 2004 and August 2008, 77 had pro-
longed seizures. Among them, 59 underwent magnetic
resonance imaging (MRI). A 21-month-old girl showed
hippocampal changes after her first episode of febrile
status epilepticus. The seizure lasted about 35 min, with
eye deviation to the right and ictal rhythmic discharges in
high-signal intensities in T2and fluid-attenuated inversion
recovery (FLAIR) images of the left hippocampus. Left
hippocampal volume diminished over the next several
months suggesting the occurrence of neuronal cell death.
Inno othercases, not eventhose withlonger seizure dura-
tions, did significant hippocampal changes develop. The
frequency of hippocampal damage was 1.7% in this case
series. The involvement of factors other than seizure
duration merits further study.
KEY WORDS: Hippocampal atrophy, Febrile seizure,
Prolonged seizure, Magnetic resonance image.
Febrile seizures (Fs) have been widely recognized as an
important antecedent event in the development of temporal
lobe epilepsy (TLE) since the original report by Falconer
et al. (1964). Several similar reports through the 1990s
have suggested that intractable TLE is preceded by pro-
longed Fs (Abou-Khalil et al., 1993; Kuks et al., 1993).
However, all such cases have been assessed retrospectively
from the viewpoint of intractable epilepsy. Therefore,
whether or not prolonged Fs can actually cause hippocam-
pal damage, the clinical factors that may participate in this
sequela and incidence of hippocampal damage, have yet to
Methods and Subjects
Four hundred seventy-two patients with Fs were admitted
to the pediatrics department of Hirakata City Hospital
betweenFebruary 2004andAugust2008.Intracranial infec-
tions were excluded by examination of cerebrospinal fluid
(CSF) or intensive observation. The clinical symptoms of
these consecutive patients were assessed prospectively.
Cases with atypical features such as prolonged seizure dura-
tion and apparently asymmetric seizure manifestation were
examined by magnetic resonance imaging (MRI). Seventy-
seven cases had prolonged Fs lasting more than 15 min.
Informed consent was obtained from the parents after a full
explanation of the reasons for MRI examination. In 18
cases, parental consent was not obtained, and these
children were excluded from the study. Therefore, 59 cases
underwent MRI examinations.
MRI was performed within a week of seizure onset in
52 cases, including a case with hippocampal damage.
Coronal sections perpendicular to the long axis of the
hippocampal formation on T1, T2, and FLAIR images were
routinely obtained for each patient. Increased signal inten-
sity on the T2and/or FLAIR images was considered to be
clinically significant and MRI scans were followed up.
These MRI findings were evaluated visually, and indepen-
dently, by two pediatric neurologists (first and second
A case with hippocampal abnormality
This girl was 21 months of age at seizure onset. Her
mother had also experienced Fs as an infant. The patient’s
development had been normal previously, with no history of
Accepted November 29, 2010; Early View publication February 5,
Address correspondence to Takuya Tanabe, Department of Pediatric
Neurology, Tanabe-Kadobayashi Children’s clinic, 1-49-31 Higashiyama,
Status epilepticus occurred the day after she developed a
fever. Tonic extension of the extremities appeared, with
tonic deviation of both eyes to the right. The seizure lasted
about 35 min, stopping when intravenous diazepam was
administered. A brief seizure occurred unexpectedly 2 h
later, during recording of electroencephalography (EEG).
The ictal EEG findings showed very high voltage rhythmic
activity in the left hemisphere, predominantly involving the
An MRI examination was conducted 72 h after cessation
of febrile status epilepticus. There was no marked hippo-
campal laterality in volume or appearance in the T1image.
On the other hand, T2and FLAIR images revealed obvi-
ously increased signal intensities in the left hippocampus.
As to the comparison between the T2and FLAIR images,
the latter showed signal abnormalities more clearly (Fig. 1).
Changes could be seen in MRI series at 46 days and 1 year
after the episode, although those changes have not been
assessed by volumetric MRI study. The increased signal
intensity on FLAIR images appeared to be gradually
decreasing on subsequent imaging studies (Fig. 2).
Our patient is now 6 years old. She has not experienced
an afebrile seizure, to date. No epileptic discharges were
seen on interictal EEG follow-up. No obvious developmen-
tal deficits have manifested. Her mother recently showed no
hippocampal abnormalities on brain MRI.
hippocampalvolume diminished progressively.
Consecutive case series
Fifty-nine cases, including the girl described, with pro-
longed Fs were enrolled in this study. None, except this girl,
showed hippocampal changes after Fs. Ages at onset ranged
MRIfindingsat 72 hafter febrile
increasedonT2and FLAIR images.
The FLAIRimage appears to depict
signalchanges more clearly than the
T2image. A,T1image; B,T2image;
Progressive atrophy of left
Left hippocampalvolume progres-
sivelydiminished during the
examinationperiod. The increased
appears tobe gradually decreasing.
T. Tanabe et al.
from 8 months to 5 years [mean € standard deviation
(M € SD); 23 € 14 months old). Seizure duration was
47.0 € 15.5 min (median 30 min, range 15–100 min).
Thirty-nine patients showed lateralized features, such as
focal convulsion, Todd’s paralysis, and eye deviation. The
case described in detail herein, with hippocampal damage,
did not have unique clinical features in terms of seizure
duration, recurrent episodes, age and so on. Twenty-three
patients had longer seizure durations than this case.
We have described herein one patient, among our pro-
spective series of Fs cases, with hippocampal damage after
a prolonged Fs. Changes in MRI findings appeared to have
occurred after the episode of Fs in this case, because a hip-
pocampal abnormality progressed gradually for several
months after her first and only episode of febrile status epi-
lepticus. We have, therefore, documented a case clearly
showing hippocampal damage as a sequela of febrile status
Van Landingham et al. (1998) reported the first series
with MRI evidence that prolonged Fs could, in rare cases,
be associated with mesial temporal sclerosis. Scott et al.
reported increased T2signal intensity and increased volume
in the acute phase. Large hippocampal volumes and prolon-
gation of T2relaxation time within 48 h after the onset of a
prolonged Fs may reflect acute edema. Furthermore, reduc-
tions in hippocampal volume and T2relaxation time several
months after a prolonged Fs may reflect hippocampal injury
and neuronal loss (Scott et al., 2002, 2003). The MRI
abnormality in our present patient was compatible with
previously reported radiologic findings, suggesting the
same pathophysiologic process as in cases described in
The cases in our present study and those described in
other reports indicate that prolonged Fs may result in hip-
pocampal damage. However, some controversy persists.
Tarkka et al. (2003), who carried out long-term follow-up,
found no association between Fs and temporal sclerosis.
The actual frequency of hippocampal damage after Fs has
yet to be determined. In the present study, 1 of 59 patients
with prolonged Fs, among a total of 477 inpatients with
seizures, was found to have developed hippocampal dam-
age. If no such patients can be identified during the inves-
tigative period, even of a long-term study, the frequency
of hippocampal damage will be 0%, as Tarkka et al.
(2003) found in their study. Although, the true prevalence
of such a condition would obviously be very difficult to
evaluate, upper 95% confidence limit is 9% in this present
study and 13% in Tarkka’s series, respectively. It seems
likely that up to around 10% of patients will have hippo-
There has been no evolution to the onset of TLE, to
date, in our present patient. A prospective study of
children with Fs demonstrated no development of TLE in
such cases (Shinnar, 1998). Therefore, a discrepancy
remains between prospective and retrospective Fs studies.
It seems that prolonged Fs may result in hippocampal
damage, although it remains unclear whether this condi-
tion can lead to mesial TLE. In this regard, Dube et al.
(2006) recently reported that prolonged experimental Fs
lead to later-onset limbic TLE in a significant proportion
of rats, showing a direct causal relationship between pro-
longed Fs and the development of TLE. Such observations
may be useful for understanding the relationship between
Fs and human TLE.
The questionthat mustbe answered, butthat ismostdiffi-
cult to answer, is why certain patients have hippocampal
damage, whereasthe greatmajority donot.Whatare the dif-
ferences between the majority who are unaffected and the
small minority who appear to be susceptible to hippocampal
damage? Seizure duration has been suggested to be the most
significant factor (Maher & McLachlan, 1995). However,
the seizure duration in our case was only 35 min. Other
cases with relatively long seizure durations showed no
hippocampal damage. The most likely explanation for this
difference is variability in susceptibility to damage associ-
ated with prolonged seizures. A variety of predisposing
factors may increase the possibility of hippocampal damage
(Fernandez et al., 1998). There is also debate as to whether
hippocampal sclerosis is a developmental or an acquired
lesion (Riney et al., 2006). Our case had a family history of
Fs, but had no abnormal MRI findings suggesting a dual
pathology. Preexisting subtle abnormalities of mesial tem-
poral structure could not be evaluated in this study because
observation began at the time of the first episode of febrile
Our patient’s course and those of other recently reported
cases indicate that prolonged Fs can, in fact, cause hippo-
campal damage. However, what remains unknown is
whether this damage is the same as hippocampal sclerosis
associated with mesial TLE in patients with a history of
febrile status epilepticus and whether there are differences
among patients in susceptibility to such hippocampal
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
Abou-Khalil B, Andermann E, Andermann F, Oliver A, Quesney
LF.(1993) Temporal lobe epilepsy after prolonged febrile
Hippocampal Atrophy after Febrile Seizure
convulsions: excellent outcome after surgical treatment. Epilepsia
Dube C, Richichi C, Bender RA, Chung G, Litt B, Baram TZ. (2006) Tem-
poral lobe epilepsy after experimental prolonged febrile seizures: pro-
Falconer MA, Serafetinides EA, Corsellis JAN. (1964) Etiology and patho-
Fernandez G, Effenberger O, Vinz B, Steinlein O, Elger CE, Dohring W,
Heinze HJ. (1998) Hippocampal malformation as a cause of familial
febrile convulsions and subsequent hippocampal sclerosis. Neurology
Kuks JB, Cook MJ, Fish DR, Stevens JM, Shorvon SD. (1993) Hippocam-
pal sclerosis in epilepsy and childhood febrile seizures. Lancet
most important predictor of temporal lobe epilepsy? Brain 118:1521–
Riney CJ, Harding B, Harkness WJF, Scott RC, Cross JH. (2006) Hippo-
campal sclerosis in children with lesional epilepsy is influenced by age
ly A. (2002) Magnetic resonance imaging findings within 5 days of sta-
Scott RC, King MD, Gadian DG, Neville BG, Connelly A. (2003) Hippo-
campal abnormalities after prolonged febrile convulsion: a longitudinal
Shinnar S. (1998) Prolonged febrile seizures and mesial temporal sclerosis.
Tarkka R, Paakko E, Phytinen J, Uhari M, Rantala H. (2003) Febrile sei-
zures and mesial temporal sclerosis. No association in a long-term fol-
Van Landingham KE, Heinz ER, Cavazos JE, Lewis DV. (1998) Magnetic
resonance imaging evidence of hippocampal injury after prolonged
T. Tanabe et al.