7/26/2019 2822.full
1/9
BRAINA JOURNAL OF NEUROLOGY
Chronic temporal lobe epilepsy:a neurodevelopmental or progressivelydementing disease?C. Helmstaedter and C. E. Elger
Department of Epileptology, University Hospital of Bonn, Bonn, Germany
Correspondence to: Prof. Dr Christoph Helmstaedter,
Neuropsychology,
Department of Epileptology,
University Hospital of Bonn,Sigmund Freud Str. 25, 53105 Bonn,
Germany
E-mail: [email protected]
To what degree does the so-called initial hit of the brain versus chronic epilepsy contribute towards the memory impairment
observed in chronic temporal lobe epilepsy (TLE) patients? We examined cross-sectional comparisons of age-related regressions
of verbal learning and memory in 1156 patients with chronic TLE (age range 668 years, mean epilepsy onset 14 11 years)
versus 1000 healthy control subjects (age range 680 years) and tested the hypothesis that deviations of age regressions (i.e.
slowed rise, accelerated decline) will reveal critical phases during which epilepsy interferes with cognitive development. Patients
were recruited over a 20-year period at the Department of Epileptology, University of Bonn. Healthy subjects were drawn from
an updated normative population of the Verbaler Lern- und Merkfahigkeitstest, the German pendant to the Rey Auditory Verbal
learning Test. A significant divergence of age regressions indicates that patients fail to build up adequate learning and memoryperformance during childhood and particularly during adolescence. The learning peak (i.e. crossover into decline) is seen earlier
in patients (at about the age of 1617 years) than for controls (at about the age of 2324 years). Decline in performance with
ageing in patients and controls runs in parallel, but due to the initial distance between the groups, patients reach very poor
performance levels much earlier than controls. Patients with left and right TLEs performed worse in verbal memory than controls.
In addition, patients with left TLE performed worse than those with right TLE. However, laterality differences were evident only
in adolescent and adult patients, and not (or less so) in children and older patients. Independent of age, hippocampal sclerosis
was associated with poorer performance than other pathologies. The results indicate developmental hindrance plus a negative
interaction of cognitive impairment with mental ageing, rather than a progressively dementing decline in chronic TLE patients.
During childhood, and even more so during the decade following puberty, the critical phases for establishing episodic memory
deficits appear. This increases the risk of premature dementia later on, even in the absence of an accelerated decline. Material
specific verbal memory impairment in left TLE is a characteristic of the mature brain and seems to disappear at an older age. The
findings suggest that increased attention is to be paid to the time of epilepsy onset and thereafter. Early control of epilepsy isdemanded to counteract developmental hindrance and damage at a younger age.
Keywords: temporal lobe epilepsy; verbal memory; ageing; development; lateralization
doi:10.1093/brain/awp182 Brain 2009: 132; 28222830 | 2822
Received February 13, 2009. Revised May 18, 2009. Accepted May 23, 2009. Advance Access publication July 27, 2009
The Author (2009). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: [email protected]
7/26/2019 2822.full
2/9
IntroductionTemporal lobe epilepsy (TLE) is the most common form of chronic
focal epilepsy. Due to the localization of neuropathology in
memory-relevant brain structures, problems in episodic memory
represent the major cognitive co-morbidity of TLE. Episodic
memory is defined as the ability to acquire and later retrieve
new space- and time-related information. It is, therefore, an
essential component in establishing continuity and biographic
identity. Memory impairment in TLE reflects loss of structural
morphological integrity of the temporal lobe structures, as well
as the dynamic and potentially reversible influence of seizures
and treatment on brain function. However, we must also consider
whether epilepsy and seizures interfere with brain maturation, and
if chronic uncontrolled epilepsy causes a progressive mental
decline with ageing.
In 2002, an international multidisciplinary meeting was held to
discuss whether seizures damage the brain. The results were pub-
lished in Progress in Brain Research (Sutula and Pitkanen, 2002).
The collected evidence showed that seizures represent only a part
of the problem and that complex molecular, cellular, synaptic andsystem level dysfunctions must also be considered when question-
ing the chronic cognitive and behavioural impairments in epilepsy.
At first glance, it would appear almost certain that uncontrolled
epilepsy causes a mental decline. However, the surprising
conclusion of Tom Sutula and Asla Pitkanen was that seizures
may damage the brain only in some individuals, in some
conditions, . . . (Chapter 47, p. 513).
Lifetime-spanning studies on cognition in patients with chronic
epilepsy are not available. At best, longitudinal studies cover an
interval of up to 10 years. Studies addressing TLE were mostly
conducted on surgically treated patients (as opposed to pharma-
cologically treated patients), and although they indicate some
decline and a slight relevance of uncontrolled seizures, the results
are inconsistent and do not particularly favour the idea that
epilepsy is a progressively dementing disease (Elger et al., 2004).
A review of 20 longitudinal studies examined the cognitive out-
comes in patients with epilepsy and concluded that there is a
mild but definite relationship between seizures and mental
decline (Dodrill, 2004). However, these studies, and particularly
the older reports, comprise patients with very different epilepsy
conditions and they may well have included patients with mental
decline due to progressive aetiology. Particularly, for those patients
who today would be considered as suffering from progressive
mitochondrial or auto-immunological inflammatory brain diseases,
it is very difficult to separate the negative cognitive effects ofpathology from those of the severe seizures that are an expression
of the progressive pathology (Helmstaedter, 2007).
In contrast to longitudinal studies, cross-sectional studies can
span a wider age range. Interestingly, such studies clearly suggest
a slow cognitive decline with increased duration of epilepsy (Jokeit
and Ebner, 1999, 2002). However, there is a major methodolog-
ical problem with studies that examine the effect of epilepsy
duration on cognition. Since most epilepsies start early in life, a
longer epilepsy duration is heavily confounded by ageing, which
itself is associated with cognitive decline.
A reasonable way out of this dilemma is to compare age regres-
sions in patients with TLE against those in healthy subjects. The
concept behind this comparison states that if epilepsy systemati-
cally changes age-related cognitive development, then this should
become evident with significant deviations of the respective
regression curves from each other. If, for example, long-lasting
chronic epilepsy systematically adds to a decline, then an accel-
eration of such a decline should show a significantly steeper
regression for the patients than for the healthy subjects. First find-
ings in a relatively small sample indicated that, contrary to previ-
ous assumptions, the age regressions for healthy controls and
patients ran in parallel (Helmstaedter and Elger, 1999). At each
age, the patients performed worse than the healthy subjects and
the distance between groups remained stable over time. Thus, it
did not seem that progressive mental decline was caused by
chronic TLE. Instead, the stable distance between patients and
controls indicated that the cognitive problems in epilepsy must
evolve around the time of epilepsy onset or perhaps even earlier.
In keeping with this finding, there is indeed a long history docu-
menting the negative effects of an early onset of epilepsy on
cognition (Fitzhugh et al., 1965; Dikmen et al., 1975; Dikmenand Matthews, 1977; OLeary et al., 1983; Strauss et al., 1995;
Cormack et al., 2007).
MethodsWe extended our previous findings with adults and focused on learn-
ing and memory performance in both children and adolescents.
We evaluated memory data from a very large sample of 1156 patients
with pharmaco-resistant TLE covering a wide age range. Patients were
collected over a 20-year time period at the Department of
Epileptology, University of Bonn. From the total neuropsychological
database with approximately 6300 first presentations, we selected
patients based on neuropsychological assessment of verbal learningand memory and the diagnosis of TLE defined by MRI (temporal
lobe lesions) and/or ictal/interictal EEG. Of the 789 with available
imaging data, the MRI indicated hippocampal sclerosis/atrophy in
62% of the patients; 19% showed developmental lesions; 9%
tumours; 4% vascular lesions; and 6% had no findings. For the
remaining patients, no MRI data were available. According to the
MRI and/or EEG data, in 595 (52%) patients a left-sided TLE was
indicated; in 538 (46%) a right-sided TLE; and uncertain bilateral find-
ings were present in 23 (2%) patients. The mean age at the onset of
epilepsy was 14 11 years, with 85% of the onsets before 25 years of
age (26%, age 05; 33%, age 614; 26%, age 1524; 15%, age
25+). The mean duration of epilepsy was 1812 years.
All subjects underwent verbal memory testing using the Verbaler
Lern-und Merkfahigkeitstest (VLMT) (Helmstaedter et al., 2001),a German adaptation of the well-known Rey Auditory Verbal
Learning Test (RAVLT). Word-list learning is commonly used for
memory assessment in epilepsy patients. When applied before and
after surgery, it reliably reflects temporal lobe functioning, temporal
lobe pathology and the cognitive effects of TLE (Helmstaedter et al.,
1997a, b; Lee et al., 2002; Loring et al., 2008).
Since we standardized and published the VLMT in Germany
(Helmstaedter et al., 2001), we also have the normative data for
this memory test with, at present, 1000 healthy subjects tested
between the ages 6 and 80 years. This allowed us to compare the
age-related memory performance of patients from 6 to 70 years
Memory course in chronic TLE Brain 2009: 132; 28222830 | 2823
7/26/2019 2822.full
3/9
(51% males, mean age 33 12 years) against the age-related memory
performance of healthy subjects from 6 to 80 years (55% males, mean
age 3521 years) (Table 1). The VLMT requires consecutive learning
and recall of a list of 15 words (list A) over five trials; free recall of this
list after each trial is requested. After the five learning trials are com-
plete, a second word list (list B) is presented for free recall to assess the
effects of interference, Finally, unannounced recall of list A is
requested after a delay of half an hour, and a recognition trial is
performed. Recognition requires identification of the words of list A
out of an orally presented list of 50 words which comprises the words
from lists A and B as well as new distractor words. Structural analysis
of this test demonstrated that it reliably assesses verbal short-term or
working memory as well as long-term memory (Muller et al., 1997;
Helmstaedteret al., 2001). Based on the structuralfunctional correla-
tion studies, out of all possible test parameters, the total number ofcorrectly recalled words in immediate recall over the five learning
trials (learning: total trials 15) was selected because it is the most
representative of the more neocortical aspects of short-term and work-
ing memory (Elgeret al., 1997; Helmstaedteret al., 1997a, b). Of the
delayed recall scores, loss of learned words over time was chosen
because this measure provides less redundant information to learning
than absolute free delayed recall (correlation of r= 0.3 instead of
r= 0.8). This measure is the representative of the more medial
temporal-dependent aspects of long-term consolidation and retrieval
(memory: loss in delayed free recall) (Helmstaedter et al., 1997a, b;
Muller et al., 1997; Helmstaedteret al., 2001, 2002).
ResultsBy taking both linear and non-linear dependencies into account,
we performed a combined linear and non-linear regression analysis
by a modelling method called LOESS or LOWESS analysis (locally
weighted scatterplot smoothing) with a smoothing parameter
q = 0.3 (Cleveland and Devlin, 1988). This model more adequately
described the course of memory performance over the time than
standard regression methods. Instead of fitting a specification of a
function to all data in the sample, this analysis fits the regression
to localized subsets of data to build-up a function point-by-point.
The smoothing parameter q can be flexibly chosen (typically
between 0.25 and 0.5) and is the proportion of data used in
each fit. The smaller the q is, the closer the regression conforms
to the data, the larger the q is, the lesser the regression responds
to fluctuations in the data.
[Linear analyses for controls had indicated that the development
of verbal learning memory can be best explained by a quadratic
function (learning: r2 = 0.15, F= 87.1, P50.001; memory:
r2 = 0.03, F= 16.4, P50.001); whereas in patients, learning and
memory curves were best approximated by a linear and logarith-
mic function, respectively (learning: r2 = 0.031,F= 37.1,P50.001;
memory: r2 = 0.03, F= 40.1, P50.001).]
The cross-sectional development of verbal learning in relation to
age (i.e. the comparison of the age regressions of learning in
healthy subjects and TLE patients) is displayed in Fig. 1A. The
performance of both groups in 5-year increments until the age
of 30 years and 10-year increments until the age of 80 years is
displayed in Fig. 1B. The respective courses of verbal memory are
displayed in Fig. 2A and B.
Healthy controls show a course of verbal learning in relation to
age as expected from developmental psychological research(Grady and Craik, 2000; Salthouse, 2009) (Fig. 1A). There
clearly is a linear increase in the performance until about the
age of 2223 years, at which point a slow, but steady, linear
decline is present. Comparing the course of memory in healthy
subjects and patients beyond the age of 25 years shows that the
decline in learning capacity in patients runs largely parallel to that
in the controls, but at a much lower level. After the age of 50
years, the patients and controls come closer, but because of the
decreasing number of older patients, their development is difficult
to determine via the scatter plot. The results for the age groups
(Fig. 1B) show that patients and controls are closer to each other
between 60 and 70 years, mainly because of the large variance in
patient performance.The course of verbal memory (loss of learned words over time)
shows how verbal memory performance becomes optimized in
healthy subjects until young adulthood, in that retrieval efficiency
steadily improves along with the increase observed for learning
(Fig. 2A and B). The best retrieval performance (i.e. the best learn-
ing and least loss of what has been learned) is observed at the age
of 2223 years, where the mean loss approaches the zero line.
From this age onward, a decline becomes evident until age 40,
and then plateaus until the age of 70 years. After 70 years of age,
some further decline is indicated. Again, the memory courses of
healthy subjects and TLE patients over the age of 25 years run
largely parallel and the results on ageing are in keeping with our
previous findings in a selected group of patients with left mesialTLE (Helmstaedter and Elger, 1999). The results on ageing are also
in line with recent longitudinal findings on quantitative MRI in
patients with TLE and healthy subjects, which again fail to provide
evidence that chronic TLE progressively adds damage to the brain
(Helmstaedter and Elger, 1999; Liu et al., 2003, 2005).
The major finding of this study becomes evident when the
course of learning and memory development in childhood and
adolescence is compared among the groups. Here, the slopes
diverge significantly, in that a steep increase of the learning per-
formance in healthy subjects contrasts with a flattened increase in
Table 1 Distribution of patients and control subjects overage groups between 6 and 80 years
Age categories Groups Total
Controls TLE
610 134 28 162
1115 104 68 172
1620 83 107 190
2125 104 139 243
2630 92 152 244
3140 94 355 449
4150 103 208 311
5160 98 82 180
6170 136 17 153
7180 52 0 52
Total 1000 1156 2156
Until the age of 30 years, 5-year increments and in the time after 30 years,
10-year increments were choosen.
2824 | Brain 2009: 132; 28222830 C. Helmstaedter and C. E. Elger
7/26/2019 2822.full
4/9
patients (Fig. 1A and B). When compared with healthy subjects
who increased their learning and memory performance until their
early mid-twenties, the TLE patients completely failed to build-up
their respective performance. They demonstrated only a very
moderate increase in learning capacity which came to a halt at
the age of 1617 years. Instead of further improving their perfor-
mance, the patients had already reached the point at which verbal
learning turns over into a steady decline. When the groups are
Figure 1 (A) Age regressions of verbal learning (total learning over five learning trials) in patients with chronic TLE and healthy controlsubjects (CONTR) indicate developmental hindrance rather than progressive accelerated cognitive decline in the patient group.
The figure displays a scatter plot representing all individual subjects. The two vertical lines indicate the turn over when learning starts
to decline. (B) The results on verbal learning with subjects grouped by age increments of 5 and later of 10 years. Bars represent 90% CI
for the group means. Note how the gap between patients and controls increases with the age of 1620 years and how it reaches
a maximum difference with the age of 2125 years where patients already turned into decline.
Memory course in chronic TLE Brain 2009: 132; 28222830 | 2825
7/26/2019 2822.full
5/9
broken down into 5- and 10-year increments, the patients and
control subjects are closest to each other at a young age
(Fig. 1B). Thereafter, the gap opens at 1620 years of age, and
then reaches the maximum difference for the 2125 years of age
groups. At that stage, healthy subjects are still improving their
performance, whereas the patients are already in significant decline.
Unlike their performance in learning, the patients performance
in memory appears to decline linearly from the very beginning
until the age of 2627 years. Memory in healthy subjects increases
linearly and peaks at age 2122 years (Fig. 2A and B). As with
learning, the age group comparison shows that the gap opens at
age 1620 years and continues until 2125 years. Accordingly, a
Figure 2 (A) Age regressions of verbal memory (loss of learned words in half hour delayed free recall) in patients with chronic TLE andhealthy control subjects (CONTR) indicate that this performance decreases in patients from the very beginning (positive values mean
greater losses). The two vertical lines indicate the time from which on memory goes over into a plateau. (B) The results on verbal
memory with subjects grouped by age increments of 5 and later of 10 years. Bars represent 90% CI for the group means. Note the
steady decline in patients, how the gap between patients and controls increases with the age of 1620 years and how it reaches a
maximum difference with the age of 2125 years.
2826 | Brain 2009: 132; 28222830 C. Helmstaedter and C. E. Elger
7/26/2019 2822.full
6/9
slow development of verbal learning and deteriorating verbal
memory is implied, but there is no hard evidence suggesting an
accelerated decline with a longer duration of epilepsy at an older
age. Finally, it should be noted that at an older age, learning and
memory performances show a different age dependence
(i.e. learning is more significantly modulated by ageing than
memory).
With regard to the findings on learning, it might be interesting
to note that separate analyses performed for memory span at the
first learning trial on list A and the single learning trial on list B
showed very similar results as observed with total learning. The
courses of memory span performance across the ages largely
mirrored those seen with learning (figures are not shown). The
rationale for adding this analysis was that memory span depends
on frontally mediated processes of working memory rather than
on medialtemporal, long-term consolidation. However, the com-
parable results are not surprising since total learning and memory
spans are highly correlated with each other (list A, r= 0.76; list B,
r= 0.56) (Muller et al., 1997).
The underlying pathology and the hemispheric lateralization ofTLE are major determinants of the verbal memory impairment in
TLE (Wieser, 2004). Since the respective subgroups were suffi-
ciently large, we also calculated subsequent analyses by taking
the presence versus absence of the hippocampal sclerosis and
the lateralizaton of the epilepsy in the left versus right hemisphere
into account.
As for the pathology, 491 (62%) of 789 patients with MRI were
diagnosed with a hippocampal sclerosis/atrophy (AHS), 298 (38%)
patients had another pathology. Comparison of the groups
indicated that verbal learning and memory performances were
worse in patients with hippocampal sclerosis than for those
patients with another pathology (learning F= 16.1, P50.000
memory F=6.9, P50.01). However, this difference was seen
only when the total groups were taken into consideration. When
the different age groups were calculated separately, the differ-
ences between patients with AHS and other pathologies were
not significant.
We compared patients with left (N =595) and right TLE
(N = 538). Both groups were impaired in verbal learning and
memory in relation to controls and, in addition, the patients
with left TLE performed worse than those with right TLE
(F=140 and 141 with P50.000 for learning and memory).
However, as indicated in Table 2, the leftright difference in
verbal learning and memory was confined to adolescence and
adulthood and the occurrence of leftright differences across the
ages was different for verbal learning and memory. In verbal
learning (total: trials 15), significant leftright differences were
observed between the age of 3150 years. In verbal memory
(loss in delayed free recall), the leftright differences were seen
between the ages of 1150 years. With an older age (450 years),the leftright differences became less pronounced and failed to
reach statistical significance.
DiscussionAge regressions of verbal learning and memory in patients with
TLE and healthy subjects were compared with the test hypothesis
that significant deviances of the age regressions would reveal
critical phases whereby epilepsy interferes with cognitive
Table 2 Differences in verbal learning (total learning) and memory (loss of learned words after delay) across the ages asdependent on the lateralization of temporal lobe epilepsy in the left versus right hemisphere
Agecategories(years)
N mean (SD) F/sig nif . Agecategories(years)
N mean (SD) F/signif.
510 VLMT learning (trials 15) R-TLE 12 40.33 (11.98) 0.001 NS 3140 R-TLE 154 46.18 (10.06) 4.45
L-TLE 16 40.19 (11.16) L-TLE 196 44.04 (8.93)
VLMT memory(loss over time) R-TLE 12 1.08 (1.56) 0.06 NS R-TLE 154 2.35 (2.19) 26.22
L-TLE 16 1.25 (1.77) L-TLE 196 3.56 (2.19)
1115 VLMT learning(trials 15) R-TLE 36 48.33 (9.36) 0.44 NS 4150 R-TLE 101 45.22 (9.37) 13.01
L-TLE 31 46.58 (12.19) L-TLE 100 40.13 (10.59)
VLMT memory (loss over t ime) R-TLE 36 1.47 (1.64) 5.40 R-TLE 101 2.39 (2.42) 10.41
L-TLE 31 2.35 (1.42) L-TLE 100 3.45 (2.20)
1620 VLMT learning(trials 15) R-TLE 47 49.06 (8.42) 0.039 5160 R-TLE 41 42.85 (9.18) 1.84 NS
L-TLE 58 49.41 (9.47) L-TLE 38 40.11 (8.73)
VLMT memory (loss over t ime) R-TLE 47 1.55 (2.09) 4.50 R-TLE 41 2.92 (2.25) 0.36 NS
L-TLE 58 2.43 (2.12) L-TLE 38 3.23 (2.30)
2125 VLMT learning (t rials 15) R-TLE 71 48.18 (9.22) 1.81 NS 6170 R-TLE 12 42.75 (9.59) 0.21 NS
L-TLE 67 46.06 (9.27) L-TLE 3 40.00 (6.55)
VLMT memory (loss over t ime) R-TLE 71 2.21 (1.97) 5.125 R-TLE 12 2.33 (2.70) 3.93 NS
L-TLE 67 3.01 (2.19) L-TLE 3 6.00 (3.60)
2630 VLMT learning (trials 15) R-TLE 64 46.30 (10.1) 0.273 NS
L-TLE 86 45.48 (9.016)VLMT Memory [loss over time] R-TLE 64 2.09 (2.39) 13.74
L-TLE 12 40.33 (11.980)
Note that left-right differences mainly appear as a characteristic of the mature and adult brain.P50.05; P50.01; P50.001; NS= not significant.
Memory course in chronic TLE Brain 2009: 132; 28222830 | 2827
7/26/2019 2822.full
7/9
development. Since most epilepsies start early in life, with
increasing age of patients at the time of evaluation, the impact
of the onset of epilepsy will have a decreasing contribution to
the resulting performance, whereas the contribution of the
duration of epilepsy will increase. Therefore, if deviances exist
at a young age, then one should be able to confirm either
developmental hindrance or retardation. If there are deviances at
an older age, this should be indicative of an accelerated mental
decline.
If we use this as our basis, the present results provide strong
evidence of a neurodevelopmental hindrance in chronic TLE
patients, which becomes most evident in the decade after puberty.
While the gap between patients and controls increasingly widens
at a younger age, the course of learning and memory at an older
age runs largely in parallel. However, patients at every age score
are at a much lower level than healthy subjects. Thus, no progres-
sive degenerative decline can be confirmed from these findings.
So far, this is in keeping with the results that we published in 1999
involving a selected group of patients with pure left mesial TLE
(Helmstaedter and Elger, 1999).
The age at which development comes to a halt indicates thatthe problems with learning arise in later brain maturation, which
is no longer intrinsically and ontogenetically determined, but rather
mainly extrinsically driven by experience (Segalowitz and Hiscock,
2002). Learning (i.e. the short-term and working memory aspects
of memory) highly depends on attention, executive functions and
language. Healthy subjects are able to improve their learning
capacity by increasing these more extra-temporally and mostly
frontally-associated functions. Learning, as well as memory in
terms of the access to learned materials, improves with better
behavioural organization and the extension of semantic networks
and knowledge systems. One may hypothesize that patients fail in
this particular respect.
It is remarkable that the age at which patients becomeincreasingly impaired marks the approximate endpoint of func-
tional cerebral plasticity, which helps to preserve language and
language-related memory aspects in early onset epilepsy
(Helmstaedter et al., 1997a, b, Helmstaedter, 1999). As pointed
out elsewhere in more detail (Helmstaedter and Elger, 1998;
Helmstaedter, 1999), the neocortical and mesio-limbic aspects of
verbal learning and memory seem to have different functional
plasticity. Mesial encoding functions are more bilaterally disposed,
and unilateral damage can, in part, be compensated by homolo-
gous contralateral structures quite independent of age. Unilateral
selective amygdalo-hippocampectomy demonstrates that surgical
damage often causes additional memory impairment, but will not
lead to global amnesia as long as no contralateral damage exists.Plasticity for material-specific neocortical functions is, in contrast,
time-limited, in that neocortical functions become increasingly
dependent on a single hemisphere. The critical period for contra-
lateral restitution and compensation in the presence of unilateral
damage extends to puberty (Vargha-Khadem et al., 1997;
Grunwald et al., 1998; Helmstaedter and Elger, 1998;
Helmstaedter, 1999; Liegeois et al., 2008). Accordingly, the time
course of the more medial-temporal-dependent memory aspect
among the age groups is comparably stable at an older age,
whereas learning experiences a significant modulation with
neocortical development. In the cross-sectional perspective, brain
maturation and development also seem to be important for what
is referred to as lateralization-dependent material-specific memory
impairment in TLE (Saling, 2009). First, verbal learning and
memory impairment are not confined to left TLE, i.e. patients
with right TLE perform poorly as well, although to a lesser
degree than left TLE patients. Of major interest, with regard to
the lateralization of TLE is the finding that the leftright differ-
ences in verbal learning and memory are observed in adolescent
and adult patients, but not (or less so) in children and older
patients. Thus, the impairment of verbal memory and its
material-specific affection in left TLE emerge in the mature brain
when hemispheric specialization becomes complete. At an older
age, the material specificity of the memory impairment in lateral-
ized TLE interestingly seems to disappear. Whether this reflects a
converging memory impairment pattern independent on laterali-
zation because of generally decreasing memory resources and
mental reserve capacities at an older age is an interesting
hypothesis which deserves further attention.
Hippocampal sclerosis, which was the most frequent pathology
in the evaluated group of TLE patients, had a differential impacton verbal learning and memory. The patients with hippocampal
sclerosis performed worse than those with another pathology
independent of age. Due to the cross-sectional study design, we
cannot provide answers regarding the aetiology of the observed
developmental hindrance. Do the data show additive damage,
particularly in childhood and adolescence? Do they indicate early
damage which subsequently grows into the impairment when the
affected areas should have their developmental spurts? Does
spreading epileptic activity have a distant effect on the develop-
ment of brain areas not directly involved in epilepsy? Learning and
memory are closely associated with temporal lobe structures and
are thus most critical for the study of the effects of chronic TLE on
cognition. Indeed, it is likely that the time-related changes inlearning and memory are linked to fluctuations in IQ as a function
of age. As already mentioned in the introduction, there is ample
evidence from the literature to suggest that an earlier onset of epi-
lepsy is associated with a poorer intellectual and educational out-
come (Strauss et al., 1995; Dikmen et al., 1977; Helmstaedter,
2005; Cormack et al., 2007). Evidence for the neurodevelopmen-
tal hypothesis originates from studies using quantitative brain vol-
umetric measures. These studies show that an early onset TLE, in
contrast to a late onset TLE, has substantial negative effects on
extratemporal cortical development (Hermann et al., 2002, 2003;
Weberet al., 2007; Kaaden et al., 2008).
As for the concerns about dementia in TLE, it is important to
note that although the data indicate a neurodevelopmental originof memory problems, patients are nevertheless at an increased risk
of having dementiathat is, if we use the term dementia in its
broadest sense. Even without an accelerated decline, the negative
interaction of the initially established damage with normal ageing
significantly reduces the patients reserve capacities at an older age
when processes of normal or even pathological ageing take place.
In this regard, the present data parallel those from the extraordi-
nary longitudinal Scottish Mental Survey, which indicated that
childhood IQ is a significant predictor for later dementia and mor-
tality (Whalley et al., 2000; Starret al., 2008). How later acquired
2828 | Brain 2009: 132; 28222830 C. Helmstaedter and C. E. Elger
7/26/2019 2822.full
8/9
brain damage can change the patients cognitive perspective with
ageing has been demonstrated with the effects of epilepsy surgery
on the age regression of verbal learning and memory performance
(Helmstaedteret al., 2002).
At this stage of the evaluation, the data cannot provide final
answers to the questions concerning the basic mechanisms behind
the developmental hindrance and the detailed impact of different
aetiologies, pathologies, seizure activity or medication on memory
in the lifetime cycle. The main message of this study was to pro-
vide a macro view of verbal learning and memory in TLE patients
of various ages, and to reveal that our perspective towards
patients with chronic epilepsy should change; it must shift away
from viewing epilepsy as a progressively dementing disease and
back towards showing how epilepsy, at its origin, interferes with
brain maturation and cognitive development. In the long run, this
increases the risk of premature dementia. Additional efforts are
thus needed to identify patients who are at risk and to counteract
negative cognitive development in TLE at the very beginning.
ReferencesCleveland WS, Devlin SJ. Locally weighted regression: an approach to
regression analysis by local fitting. J. Am. Stat. Assoc 1988; 83:
596610.
Cormack F, Cross JH, Isaacs E, Harkness W, Wright I, Vargha-Khadem F,
et al. The development of intellectual abilities in pediatric temporal
lobe epilepsy. Epilepsia 2007; 48: 2014.
Dikmen S, Matthews CG. Effect of major motor seizure frequency upon
cognitive-intellectual functions in adults. Epilepsia 1977; 18: 219.
Dikmen S, Matthews CG, Harley JP. The effect of early versus late onset
of major motor epilepsy upon cognitive-intellectual performance.
Epilepsia 1975; 16: 7381.
Dikmen S, Matthews CG, Harley JP. Effect of early versus late onset of
major motor epilepsy on cognitive-intellectual performance: further
considerations. Epilepsia 1977; 18: 316.
Dodrill CB. Neuropsychological effects of seizures. Epilepsy Behav 2004;5 (Suppl 1): 214.
Elger CE, Helmstaedter C, Kurthen M. Chronic epilepsy and cognition.
Lancet Neurol 2004; 3: 66372.
Elger CE, Grunwald T, Lehnertz K, Kutas M, Helmstaedter C,
Brockhaus A, et al. Human temporal lobe potentials in verbal learning
and memory processes. Neuropsychologia 1997; 35: 65767.
Fitzhugh LC, Fitzhugh KB, Reitan RM. Adaptive abilities and intellectual
functioning of hospitalized alcoholics: further considerations. Q J Stud
Alcohol 1965; 26: 40211.
Grady CL, Craik FI. Changes in memory processing with age. Curr Opin
Neurobiol 2000; 10: 22431.
Grunwald T, Lehnertz K, Helmstaedter C, Kutas M, Pezer N, Kurthen M,
et al. Limbic ERPs predict verbal memory after left-sided hippocam-
pectomy. Neuroreport 1998; 9: 33758.
Helmstaedter CA. Prediction of memory reserve capacity. Adv Neurol1999; 81: 2719.
Helmstaedter C. Effects of chronic epilepsy on declarative memory
systems. Prog Brain Res 2002; 135: 43953.
Helmstaedter C. Effects of chronic temporal lobe epilepsy on memory
function. In: Arzimanoglou A, Aldenkamp AP, Cross H, Lassonde M,
Moshe SL, Schmitz B, editors.. Cognitive dysfunction in children with
temporal lobe epilepsy. France: John Libbey; 2005. p. 1330.
Helmstaedter C. Cognitive outcome of status epilepticus in adults.
Epilepsia 2007; 48 (Suppl 8): 8590.
Helmstaedter C, Elger CE. Functional plasticity after left anterior temporal
lobectomy: reconstitution and compensation of verbal memory func-
tions. Epilepsia 1998; 39: 399406.
Helmstaedter C, Elger CE. The phantom of progressive dementia in
epilepsy. Lancet 1999; 354: 21334.
Helmstaedter C, Grunwald T, Lehnertz K, Gleissner U, Elger CE.
Differential involvement of left temporolateral and temporomesial
structures in verbal declarative learning and memory: evidence from
temporal lobe epilepsy. Brain Cogn 1997a; 35: 11031.
Helmstaedter C, Kurthen M, Linke DB, Elger CE. Patterns of language
dominance in focal left and right hemisphere epilepsies: Relation to
MRI findings, EEG, sex, and age at onset of epilepsy. Brain Cognit
1997b; 33: 13550.Helmstaedter C, Lendt M, Lux S. VMLT Verbaler Lern- und
Merkfahigkeitstest. Gottingen: Beltz Test GmbH; 2001.
Helmstaedter C, Reuber M, Elger CC. Interaction of cognitive aging and
memory deficits related to epilepsy surgery. Ann Neurol 2002; 52:
8994.
Hermann B, Hansen R, Seidenberg M, Magnotta V, OLeary D.
Neurodevelopmental vulnerability of the corpus callosum to childhood
onset localization-related epilepsy. Neuroimage 2003; 18: 28492.
Hermann B, Seidenberg M, Bell B, Rutecki P, Sheth R, Ruggles K, et al.
The neurodevelopmental impact of childhood-onset temporal lobe
epilepsy on brain structure and function. Epilepsia 2002; 43:
106271.
Jokeit H, Ebner A. Long term effects of refractory temporal lobe epilepsy
on cognitive abilities: a cross sectional study. J Neurol Neurosurg
Psychiat 1999; 67: 4450.Jokeit H, Ebner A. Effects of chronic epilepsy on intellectual functions.
Prog Brain Res 2002; 135: 45563.
Kaaden S, Quesada CM, Schramm J, Urbach H, Helmstaedter C. Voxel-
based morphometry (VBM) supports neurodevelopmental hypothesis
regarding early onset epilepsy in temporal lobe epilepsy.
In: Proceedings of the 8th European Congress of Epileptology.
8th ECE: 234, September 2125, 2008. Berlin. p. 234.
Lee TM, Yip JT, Jones-Gotman M. Memory deficits after resection from
left or right anterior temporal lobe in humans: a meta-analytic review.
Epilepsia 2002; 43: 28391.
Liegeois F, Cross JH, Polkey C, Harkness W, Vargha-Khadem F.
Language after hemispherectomy in childhood: contributions from
memory and intelligence. Neuropsychologia 2008; 46: 31017.
Liu RS, Lemieux L, Bell GS, Sisodiya SM, Shorvon SD, Sander JW, et al.
A longitudinal study of brain morphometrics using quantitative mag-
netic resonance imaging and difference image analysis. Neuroimage
2003; 20: 2233.
Liu RS, Lemieux L, Bell GS, Sisodiya SM, Bartlett PA, Shorvon SD, et al.
Cerebral damage in epilepsy: a population-based longitudinal quanti-
tative MRI study. Epilepsia 2005; 46: 148294.
Loring DW, Strauss E, Hermann BP, Barr WB, Perrine K, Trenerry MR,
et al. Differential neuropsychological test sensitivity to left temporal
lobe epilepsy. J Int Neuropsychol Soc 2008; 14: 394400.
Muller H, Hasse-Sander I, Horn R, Helmstaedter C, Elger CE. Rey
Auditory-Verbal Learning Test: Structure of a modified German ver-
sion. J Clin Psychol 1997; 53: 66371.
OLeary DS, Lovell MR, Sackellares JC, Berent S, Giordani B,
Seidenberg M, et al. Effects of age of onset of partial and generalized
seizures on neuropsychological performance in children. J Nerv Ment
Dis 1983; 171: 6249.
Saling MM. Verbal memory in mesial temporal lobe epilepsy: beyondmaterial specificity. Brain 2009; 132: 57082.
Salthouse TA. When does age-related cognitive decline begin? Neurobiol
Aging 2009; 30: 50714.
Segalowitz S, Hiscock M. The neuropsychology of normal develop-
ment: developmental neuroscience and a new constructivism.
In: Segalowitz SJ, Rapin I, editors. Child neuropsychology. Part 1
2nd edn., Amsterdam: Elsevier; 2002. p. 727.
Starr JM, Deary IJ, Whalley LJ. All-cause mortality in the Aberdeen 1921
birth cohort: effects of socio-demographic, physical and cognitive
factors. BMC Public Health 2008; 8: 307.
Strauss E, Loring D, Chelune G, Hunter M, Hermann B, Perrine K, et al.
Predicting cognitive impairment in epilepsy: findings from the
Memory course in chronic TLE Brain 2009: 132; 28222830 | 2829
7/26/2019 2822.full
9/9
Bozeman epilepsy consortium. J Clin Exp Neuropsychol 1995; 17:
90917.
Sutula T, Pitkanen A, editors. Do seizures damage the brain. Amsterdam:
Elsevier; 2002.
Vargha-Khadem F, Carr LJ, Isaacs E, Brett E, Adams C, Mishkin M. Onset
of speech after left hemispherectomy in a nine-year-old boy. Brain
1997; 120 (Pt 1): 15982.
Weber B, Luders E, Faber J, Richter S, Quesada CM, Urbach H, et al.
Distinct regional atrophy in the corpus callosum of patients with tem-
poral lobe epilepsy. Brain 2007; 130 (Pt 12): 314954.
Whalley LJ, Starr JM, Athawes R, Hunter D, Pattie A, Deary IJ.
Childhood mental ability and dementia. Neurology 2000; 55: 14559.
Wieser HG. Mesial temporal lobe epilepsy with hippocampal sclerosis.
Epilepsia 2004; 45: 695714.
2830 | Brain 2009: 132; 28222830 C. Helmstaedter and C. E. Elger