Language After Hemispherectomy Language After Hemispherectomy Susan Curtiss Stella

language after hemispherectomy language after hemispherectomy susan curtiss stella de bode department of linguistics, neuro/psyc

Language after Hemispherectomy
Language after Hemispherectomy
Susan Curtiss
Stella de Bode
Department of Linguistics, Neuro/Psycholinguistics Laboratory,
University of California, Los Angeles
and
W. Donald Shields
Department of Pediatrics, UCLA School of Medicine, Los Angeles
Running head: Language after Hemispherectomy
This study was supported by NIH grant NS 28383 awarded to the third
author. Thanks to Dr. G. Mathern for his help in establishing
etiologies and comments on the manuscript. We gratefully acknowledge
assistance with transcription and coding by Jeannette Schaeffer, Todd
Masilon, Wendy Hayashi, and John Grinstead and especially acknowledge
the participation and courage of all of the children in this study and
their families.
Correspondence concerning this article should be addressed to Susan
Curtiss, Neuro/Psycholinguistics Laboratory, 405 Hilgard Ave., UCLA,
CA 90095; [email protected]
Abstract
We report on the effects of etiology, side of damage, age at surgery,
age at seizure onset and seizure duration on language development in a
large pediatric hemispherectomy population. Surprisingly, side of
damage did not predict outcome nor did early age at surgery. Moreover,
the most striking findings regarding linguistic outcomes were
demonstrated in the right hemispherectomies; specifically, many failed
to develop language at all, and for those who did, age at seizure
onset/surgery was strongly predictive of outcome. Additionally,
developmental pathology was more strongly associated with poor
language than was acquired pathology. Implications of these and other
findings are discussed.
Key words: hemispherectomy, seizure/surgery onset,
acquired/developmental pathology, linguistic outcome
1. Introduction
This paper reports on linguistic outcomes of a large series of
pediatric hemispherectomies and examines some of the factors that may
be predictive of language outcome. This work is part of a larger
project examining language development following hemispherectomy, to
better understand the linguistic capacity of each, isolated,
developing hemisphere. We have previously reported on functional
category development of a subset of our population (Curtiss &
Schaeffer, 1997a, b1). In this paper we examine the question of
hemispheric plasticity for language development and the effects of
age, seizure-related and etiological factors on language following
hemispherectomy.
Clinical, experimental, and neuroanatomical data have increasingly
converged to suggest that at or even before birth, in the vast
majority of individuals, the left hemisphere is pre-potent to support
language acquisition, that is, genetically programmed to serve as the
neural substrate for language (e.g., SUBSTITUTE REFERENCE FoR THIS ONE
AND DELETE Gallagher & Watkin, 1996; Molfese & Segalowitz, 1988;
Witelson, 1982). However, the picture is not entirely clear regarding
how this specialization of the left hemisphere (LH) unfolds in normal
development, what the role of the right hemisphere (RH) is in normal
language development, or what the capacity of the RH to support normal
linguistic development is, completely on its own. Although there are
single cases reported in the literature describing excellent
linguistic outcomes after left hemispherectomy, some (e.g., Dennis,
1980a, b) were not studied during the course of language acquisition
and thus developmental patterns in these cases remain unknown. Other
cases have also reported near-normal language after left
hemispherectomy (e.g., Ogden, 1988; Vargha-Khadem and Mishkin, 1997),
however, it is difficult to generalize from so few cases to the normal
population.
Event Related Potential (ERP) studies of normal children in early
stages of overt lexical acquisition and clinical studies of children
with unilateral lesions lend support to the view that the left
hemisphere is “pre-potent” for language (Aram & Eisele, 1992; Cohen,
1992; Mills et al, 1994; Molfese, 1989). We will refer to this view as
the “prepotency” view. Importantly, it predicts that even with early
damage to the left hemisphere, the right hemisphere will not support
full and normal language acquisition.
On a somewhat different view, it has been proposed that despite this
initial predisposition of the LH, the RH has an equal potential to
subserve language, but the continuing lateralization of language to
the LH overrides this potential (Kinsbourne, 1974; Lecours & Joanette,
1985; Locke, 1997). This view is consistent with the Kennard principle
(Kennard, 1940), which asserts that the earlier the brain damage,
the better the outcome. Studies of unilateral LH lesions in childhood
have been used to support this position as well (e.g., Stiles & Thal,
1993; Thal et al, 1991). Moreover, in addition to a view that the RH
has equal potential to subserve language, some of these data have been
used to suggest that the RH plays a key role in language development,
and if damaged, language development will be impaired. This “equipotential”
account, then, makes three predictions: 1) that with significant early
damage to the LH, the initially equivalent linguistic potential of the
RH will be realized, 2) that the earlier the damage, the greater the
potential which can be tapped, and 3) that damage to the RH may result
in linguistic impairment.2
Because so little is understood about the mechanisms responsible for
inter- vs. intra-hemispheric language transfer after LH focal lesions
(de Bode, 1998; Duchowny et al, 1996; Helmstaedter et al, 1994), the
potential of the RH to serve as the substrate for the acquisition of
grammar may not be adequately determinable by studies of focal damage
in children. It is often unknown whether impaired performance in these
cases reflects the best efforts of the damaged LH, the linguistic
performance of the RH, or some combination of both. Moreover, findings
regarding language transfer often conflict with each other regarding
age, etiology, and lesion location. Thus, it is the study of language
development subsequent to hemispherectomy -- the removal of an entire
cortical hemisphere -- that may best directly address the question of
the linguistic potential of the RH.
The specific linguistic effects of hemispherectomy remain an area of
some controversy. There have been a number of studies reporting the
greater capacity and proficiency of the LH over the right with respect
to morphosyntactic comprehension, production, and judgments (Day &
Ulatowska, 1979; Dennis & Kohn, 1975; Dennis and Whitaker, 1976;
Dennis, 1980a, b; Stark et al, 1995; Vargha-Khadem et al, 1991;
Vargha- Khadem and Polkey, 1992) and with respect to reading and
spelling (Ogden, 1996). Others, (Strauss & Verity, 1983; Riva &
Cazzaniga, 1986; Vargha-Khadem et al, 1997) have reported excellent,
even normal linguistic abilities after hemispherectomy of either side.
Given that hemispherectomy is performed in treatment of intractable
epilepsy, such variable findings may be accounted for only by
considering such factors as etiology of damage, age at seizure onset,
duration of seizures, and age at surgery in addition to side of damage
(Gordon, 1996; O'Leary, 1983; Rossi, et al. 1996). However, little
research has been carried out specifically addressing the effects of
such factors on language. For example, the relationship between
etiology of the underlying pathology and linguistic outcome in
children with catastrophic epilepsy has yet to be established.
We have attempted to begin to address these questions by studying the
language of a large series of hemispherectomied children and analyzing
which of a number of clinical factors was predictive of their spoken
language outcome. Our central questions were first, what is the
linguistic capacity of each isolated developing hemisphere and,
second, what is the effect of specific clinical factors on language
development after hemispherectomy. In particular, we were interested
in examining whether or not there is a systematic relationship between
underlying neuropathology and extent and patterns of linguistic
development following hemispherectomy. Our hypotheses are listed in
(1) below:
(1) Hypothesis 1: Children who have undergone left hemispherectomy are
expected to show marked deficits, particularly with respect to
grammatical development, compared to their right hemispherectomy
counterparts.
Hypothesis 2: Early surgery is expected to lead to better language
outcome, particularly for left hemispherectomies. However, age at
surgery will not be predictive of outcome for right hemispherectomies
Hypothesis 3: Late age at seizure onset and surgery will be directly
related to better linguistic outcome for right hemispherectomies
because the neural basis for language in the left hemisphere will have
had a longer period to have become established in instances of later
onset.
Hypothesis 4: Duration of seizures will be inversely related to
quality of language development.
Hypothesis 5: Seizure control will be directly related to linguistic
outcome because seizure control is expected to be a reflection of the
neurological status of the remaining hemisphere.
Hypothesis 6: Developmental pathology will be associated with a poorer
linguistic prognosis than acquired pathology , thus,
6a) children with Rasmussens and infarct (acquired damage) will have
better linguistic outcomes than children with cortical dysplasia (CD,
developmental damage) and
(6b) developmental pathology which has affected the entire hemisphere
(i.e. hemimegalencephaly) will lead to the poorest outcome as compared
to multilobar damage which entails that some cortex is still
functional during development.
2. Methods
Subjects
Our subject population consisted of 48 children who had undergone
hemispherectomy as part of the UCLA Pediatric Epilepsy Surgery
Research Program. (See Peacock et al, 1995 for more details3).
Patients were included in our sample if, among other criteria, they
had catastrophic childhood epilepsy, their seizures were resistant to
antiepileptic medications, they had surgery before 18 years of age,
they were monolingual speakers of Standard American English or were in
a Standard American English environment, and they were patients for
whom follow-up information was available.
A breakdown of the subject population by side of damage, age at
seizure onset, age at surgery, disease etiology, gender and etiology
is presented in Table 1. For questions regarding effects of side of
damage, age, and seizure-related factors, we used all 48 subjects in
the analysis. For questions regarding etiology, we based our analyses
on the 42 subjects for whom etiology had been firmly established.
Insert Table 1 here
Each child's etiology was diagnosed on the basis of examination of
their entire medical history, including pathology report, PET, MRI
scans, prolonged EEG monitoring, seizure history, drug history,
interictal and ictal scalp EEG, and assessment of neurodevelopment. To
firmly establish etiology for each child, each diagnosis was rated by
a neurosurgeon and a second rater (a neuropathologist) blind to the
rating of the first.
Etiology was represented according to the following breakdown:
a. Cortical dysplasia (CD) vs. non-cortical dysplasia (nonCD), further
subdivided on the basis of extent of damage in the CD group as
follows:
1) entire hemisphere damaged (hemimegalencephaly)
2) multilobar/diffuse damage
3) focal, localized damage (only 2 subjects)
b.
The nonCD group was subdivided into two groups based on etiology:
1) Rasmussens 4
2.
Infarction
As displayed in Tables 1, seizure onset ranges from birth to 12 years,
with a mean age of 51.1 months for the right and 48.56 months for the
left hemispherectomies. All of the children were old enough to be
expected to talk at the time of postoperative assessment.5
Postsurgical linguistic evaluation
Because of the difficulty in establishing a common metric of
comprehension across the full age range studied, we concentrated on
language production. In addition, at the point of assessment used in
the analyses reported here, all of our subjects were of an age to be
expected to talk in the course of normal language development (i.e., 3
years or older).
Postsurgical data on the children’s language production were collected
by means of the MacArthur Communicative Development Inventories
(Fenson et al., 1990), or via language sampling, as developmentally
appropriate or possible. A question specifically asking for speech
onset age was added to the MacArthur for our use. Language samples
were collected by means of the Story Game from the Kiddie Formal
Thought Disorder Scales (Caplan et al, 1989) or via naturalistic
conversation with the examiner interviewing the patient on topics
including family, friends, school, birthday, TV shows, and favorite
activities. These samples were recorded, transcribed and then analyzed
for their grammatical and lexical content using CALC6, a detailed
grammatical analysis focusing on functional category structures,
constituent structure and movement.
Based on these data, we assigned each child a Spoken Language Rank (SLR),
using a five-point scale, as shown in (2) below.
(2) 0 = no speech
1 = has fewer than 20 words
2 = has more than 20 words but no word combinations
3 = constructs short telegraphic utterances
a. Helping the monkey
b. Him brown
4 = is a fluent speaker, but does not yet have the target grammar
a. Because Sammy was growned up first, so he’s the biggest and I
growed up and Chris growed and Ruben was last.
5 = has the target grammar
a. I forgot to tell them what I want
b. I hope I have my iron cast off
c. I hope it’s off by Thanksgiving because I love to downhill ski
Some children’s language appeared to fall between numerical ratings,
and they received ranks reflecting these intermediate states (e.g.,
3.5).
These rankings provided a global index of linguistic outcome. As the
mean time post-surgery was 5.28 years for the left hemispherectomies
and 6.31 years for the right hemispherectomies, these rankings
represent at least a medium-term outcome.
3. Results
Statistical analyses were performed in order to examine the
relationship between SLR and side of surgery, age at seizure onset,
age at surgery, seizure control, duration of seizures, and type of
damage (developmental or acquired). Tables 1a and 1b include
post-surgical SLRs for each child.
Side of surgery
Our first hypothesis was that the children with left hemispherectomy
would be expected to have lower SLRs than their right hemispherectomy
counterparts. However, there was no significant correlation between
side of surgery and language outcome indexed by SLR (left group mean,
2.60, right mean, 2.26; F’ = 0.1777, p > 0.5162 ). Both of these are
surprising findings and run counter to our hypothesis.
Age factors
We had hypothesized that early surgery would lead to better language
outcome, particularly for the left hemispherectomies, and that late
onset of seizures would have far less impact on spoken language for
the right hemispherectomies than the left. There was a strong
correlation between both age at surgery and SLR (p<0.0004), and age at
seizure onset and SLR (p<0.0018). However this relationship did not
hold for both right and left hemispherectomies. For the left
hemispherectomies, there was no significant correlation between either
age at surgery (F=1.985, p> 0.1146) or age at seizure onset and SLR
(p>0.0958). In contrast, there was a strong, direct relationship
between both age at surgery and age at seizure onset and SLR for the
right hemispherectomies (p<0.0018 and p<0.0113, respectively).
Examining Tables 1a and 1b in detail, we can see that the right
hemispherectomies show first, an association between early seizure
onset and lower SLR and second, later seizure onset and higher SLR.
The same pattern is not displayed by the left hemispherectomy group.
A similar pattern is evidenced between age at surgery and SLR for the
right hemispherectomy group, and is again a pattern distinct from that
of the left hemispherectomy group.
These results thus only conformed to our predictions in part. As
predicted, right hemispherectomies with late seizure onset and late
surgery had higher SLRs, but the association between early surgery and
poor outcome for the right hemispherectomies was not something we had
hypothesized, nor was the lack of a relationship between age at
surgery or age at seizure onset and outcome for the left
hemispherectomy group, findings suggesting a less-than-straightforward
relationship between age factors and linguistic outcome. We will
return to this issue in the discussion.
Duration of seizures7
We hypothesized that duration of seizures would be inversely related
to SLR, and looking at the statistical results for the group as a
whole, it does appear that duration of seizures was inversely
correlated with SLR (p< 0.0262). However, upon closer examination, we
found that, contrary to hypothesis 4, this relationship was
significant only for the right hemispherectomies (p< 0.0176). Duration
of seizures was not predictive of SLR for the left hemispherectomies
(p> 0.4008).
Seizure Control
Seizure control was a significant predictor of SLR for the group as a
whole (p< 0.0023), and for both left and right hemispherectomies (LH
p< 0.0306, RH, p< 0.0498). However, this effect is largely or entirely
an effect of seizure control on language performance in the group with
developmental pathology. The group with acquired pathology shows no
significant relation between seizure control and language rank. Thus
the two groups evidence a clear disparity with respect to the effect
of seizures on linguistic outcome, and our hypothesis that seizure
control would be inversely related to linguistic outcome regardless of
side of damage was supported only for those children with
developmental damage.
Etiology of pathology8
To reiterate our predictions in hypotheses 6a and b, we hypothesized
that hemimegalencephaly would lead to the poorest outcome, and that
children with Rasmussens and infarction would have better SLRs than
children with any form of CD. These etiologically defined subgroups
had the mean SLRs displayed in (3).
(3)
CD groups:
*
hemimegalencephaly: 0.64 - reflecting either no speech or the
emergence of first words
*
multilobar: 2.73 - reflecting a mix of one-word speech and short
word combinations
NonCD groups:
*
Rasmussens - 4.11 - reflecting fluent speech, rich with
grammatical structure
*
Infarction - 2.57 - reflecting a mix of one-word speech and short
word combinations
A statistical comparison of the CD and nonCD groups' SLRs demonstrated
a significant difference between groups (p<0.016), with the nonCD
group outperforming the CD group (mean nonCD group SLR 3.17 vs. 1.93
for the CD group). This result confirms part of Hypothesis 6a and
indicates that the CD group has either developed no language or is at
the earliest stages of spoken language development, while the nonCD
group is producing sentences and on their way to developing mature
grammars. However, comparing the SLRs of the two nonCD groups, we
found a significant difference, with the Rasmussen’s group far
outperforming the Infarction group (p< 0.0144). The Rasmussen’s group
on the whole has developed fluent, almost mature speech in contrast to
the Infarction group, who fall between the stages of producing many
words in isolation and producing short word combinations. Moreover,
the multilobar group and the Infarction group had equivalent SLRs,
contrary to our predictions in Hypothesis 6a.
To test hypothesis 6b, we next compared SLRs of the "entire
hemisphere" group (hemimegalencephaly) with the multilobar group.
There was a highly significant difference in mean SLR between the two
groups p< 0.0026, with the multilobar group consistently outperforming
the other. This confirms Hypothesis 6b. Thus, despite the fact that
the multilobar group is still, several years post-surgery, at quite
early stages of language development, reflecting critical impairments
in linguistic growth, they are still performing significantly better
than children with hemimegalencephaly.
4. Summary
As summarized in Table 2 (see below) many of our predictions were not
supported by our findings. Our first hypothesis was not supported. We
failed to find a significant correlation between side of surgery and
linguistic outcome in terms of the qualitative and quantitative
characteristics of speech production indexed by SLR. Hypothesis 2 was
also not supported. Earlier surgery did not lead to higher SLR for the
left hemispherectomies as predicted. Moreover, contrary to our
predictions, age at surgery was predictive of outcome for the right
hemispherectomies, for early age at surgery led to lower SLRs in the
group. Hypothesis 3 was supported by our findings. Late seizure onset
did have less negative impact on language in the right
hemispherectomies, in that late onset age was associated with high
SLRs. Hypothesis 4 was only partly confirmed. Duration of seizures was
predictive of SLRs for the right hemispherectomies, but not left.
Hypothesis 5 was also only supported in part. Seizure control was
inversely related to language outcome only for those children with
developmental pathology. Hypotheses 6a and b were largely confirmed.
Those children with developmental pathology evidenced poorer spoken
language than those with acquired pathology, and those with
hemimegalencephaly showed the least linguistic growth of all subgroups
of children (de Bode and Curtiss, 1998).
Insert Table 2 here
5. Discussion
Our study is the largest of its kind and has provocative implications
for neurolinguistic models. To begin, three of our results were
particularly unexpected. First was our finding that left
hemispherectomy was not more predictive of poor SLR or to put it in
reverse, that right hemispherectomies was not more consistently
associated with better language outcome. In fact, almost one-third
(31.8%, 6/19) of the right hemispherectomies in our sample have no
productive language at all, even 6 years post-surgery, compared to
16.67% of the left hemispherectomies. This finding appears
inconsistent with much of the clinical literature on hemispheric
specialization for language in children (e.g., Aram et al, 1985;
Eisele and Aram, 1993; Krashen, 1972; Woods and Teuber, 1978; Woods
and Carey, 1979; regarding focal lesions, and Dennis & Whitaker, 1976,
1977; Dennis, 1980a, b; Gott, 1973; Lovett, Dennis & Newman, 1986;
Zaidel, 1973; Gadian et al 1996; Helmstaedter et al, 1997;
Vargha-Khadem and Polkey, 1992 regarding language following
hemispherectomy).
This finding is consistent with some clinical reports on
hemispherectomy, however. There are other reports in the literature
studying one or a few individual cases, each, of good language
outcomes after left-sided damage. Ogden (1988, 1996) for example,
reports fluent speech in both hemispheres after hemispherectomy;
Isaacs et al (1996) found that side of damage was not predictive of
outcome; Stark et al (1995) report that neither side showed deficits
in lexical production; Feldman et al (1992) report that by age 24
months, children with left and right-sided perinatal damage showed
comparable language to normals on several linguistic indices; and
Curtiss and Jackson (1989), Vargha-Khadem et al (1997), and Dennis
(1980a, b) report cases of fluent, complex language abilities after
left hemispherectomy. In our own sample, a number of our left
hemispherectomies have SLRs of 4 or 5 (7/30 or 23.3%), reflecting
fluent, grammatically rich and complex speech, fully comparable to
Alex reported in Vargha-Khadem et al (1997) and SM reported in Curtiss
and Jackson (1989) (e.g., GG and NN below), and a few young Left
hemispherectomies appear to be on the road to developing mature,
normal grammars (e.g., GD), although in each case, sentence
programming and phonetic fluency appear to be problematic, as
illustrated in (4) - (6).
(4) GG:
(4a) Once upon a time or this, this story took place right here at
UCLA, and this child didn't want to see a doctor, but his mother
forced him and this doctor had to give this boy a shot, and then after
that, they went to eat at this UCLA, the cafeteria, and then after
that, they , they went home and went to sleep. (I and D structures,
infinitive clauses, causative; restarts)
(4b) This ghost went to this boy's house to play and visit, and they
had a good time. (purpose clause)
(4c) What do I like about them? (object-extracted WH, do in C)
(4d) My good dream was playing in this room with the toys. (participle
clause)
(4e)...because it's my favorite thing to believe in (adjunct clause,
complement clause)
(4f) I would make them go to a hospital like here. (past modal, small
clause)
(5) NN:
(5a) I forgot to tell them what I want (object-extracted WH,
infinitival and complement embeddings; nom. and acc. pronouns)
(5b) mm, a cofee table and a fixture for over the dining room table,
and we found some, they're going to send it to us.(nominatuve and
accusative pronouns determiners, restart)
(5c) Feels like warm. My dad didn't bring any shorts. If I want shorts
I'll probably have to walk around in my underwear. (3rd sg. agreement,
aux. do, modal, complex conditional)
(5d) and there was a motorboat that was runned on batteries with a
control (expletive subject, past be, passive be, relative clause)
(5e) We went downtown. We went to the computer museum...where they
have that big computer and that big mouse, they have a big (pause) big
keyboard that you walk on and stuff. (adjunct clause,restart. relative
clause) ... Then, upstairs, they showed us a robot,(a pause) if a
robot can make a peanut butter and jelly sandwich. (reg. past, acc.
pronouns, modal, recast complement clause) And I watched the people
that were there said, put the peanut butter on the bread and he put
the whole thing on, the whole jar on the bread.(reg. past, particle
movement, relative clause, reformulation)
(5f) I know what I want to do when I get out of this place
(WH-complement c;aise. WH adjunct clause)
(6) GD
(6a) I want daddy to stay with me (small clause, nom. vs. acc. case
pronouns)
(6b) I can go home and tell my momma I have a -my -a baby (modal, det,
object CP, restarts)
(6c) I have a baby named Rachel (det, reduced OS relative)
(6d) I'm still playing with him (aux, s-v agr., nom vs. acc. case
pronouns)
Our findings therefore imply a surprising degree of plasticity and RH
potential for language development and in this respect support the
second view, the “equipotentiality” view.
In addition, consistent with reports by others of language impairment
following damage to either hemisphere in a few focal lesion cases
(Reilly et.al., 1998; Thal et al, 1991; Stiles & Thal, 1991) our
findings on such a large sample of children who remain without
language with only an isolated left hemisphere, provide even stronger
evidence of the involvement of the RH in early stages of language
acquisition. Considered along with the findings of Molfese (1989;
1990) and Mills et al’s (1993) ERP studies on lexical processing in
infants, we speculate that the RH may play a crucial role in storing
and processing prosodic and metrical structure as a means of cracking
the linguistic code and in making an initial determination of lexical
and phrasal boundaries . Perhaps as Locke (1997) suggests, the RH
plays a key role in learning and storing lexical items which the LH
utilizes for grammatical analysis and computation.
It should be made clear that these results regarding the potential of
the RH for grammatical development do not disconfirm the normal LH
specialization for grammar. The literature clearly documents a
specialization of the LH for computation of syntactic and phonological
structure (e.g., Aram et al, 1995; Eisele & Aram, 1993; Dennis &
Whitaker, 1975; Dennis, 1980a, b; Neville, 1993; Stark et al, 1995;
Stark, 1997; Vargha-Khadem et al, 1991; Vargha-Khadem & Polkey, 1992).
Indeed, examination of grammatical performance in several of our
hemisphectomy subjects reveals the expected hemispheric differences
with respect to grammar in our population as well. Investigation of
functional category development in the speech of 13 of the children in
our hemidecorticate sample revealed better performance by the Right
hemispherectomies with both I- and D-system elements (Curtiss &
Schaeffer, 1997a, b). Our overall results for children with left
hemisphere damage are therefore difficult to interpret and require
further careful investigation. The picture regarding the effect of
side of surgery on linguistic function is clearly a complex one.
Another surprising result is the lack of a significant correlation
between age factors and SLR for the Left hemispherectomies. Counter to
the results of many other studies and our own predictions, the
children with earlier LH damage did not predictably develop more
language. Other factors such as seizure control and etiology clearly
played a role in defining which children were those who would or would
not show good spoken language outcomes. As an example, child NN had a
left hemispherectomy at age 9;8. The principle that the earlier the
damage the better the developmental outcome is clearly too simplistic
here, as is a straightforward plasticity hypothesis in which early
left hemispherectomy automatically leads to successful language
transfer.
Our finding that the older the child with RH damage the better the
language outcome is consistent with our hypothesis, however. Later RH
damage is a circumstance which allows for the necessary developmental
phenomena, including hemispheric connectivity, to trigger and foster
the unfolding of LH specialization for language. Thus, the later the
RH damage, the more fully the neural substrate for language in the LH
is established, and the greater opportunity the RH will have had to
play its role.
A main question of this study was the relationship between underlying
etiology and linguistic development post hemispherectomy. In contrast
to some of our other results, our findings here were consistent and
point to an important relationship between the pathogenesis of a
disease leading to hemispherectomy and postsurgical linguistic growth
and recovery. To our knowledge, our findings here are the first of
their kind and lend support to our approach wherein it is fruitful to
examine the relationship between specific etiologies and specific
linguistic outcomes. In particular, we predicted and found that
developmental pathology which has affected the entire (resected)
hemisphere leads to the poorest outcome; namely, little or no language
development. Our assertion is that this result is to be expected as
hemimegalencephaly prevents establishment of any interhemispheric
connectivity, a neurological factor we hypothesize to be essential to
the development of a rich and mature linguistic system. Thus, a
neurolinguistic model in which developmental changes in
interhemispheric connectivity play an important role in accounting for
patterns of results shows promise in helping to account for specific
patterns of results and more generally in understanding the factors
underlying the neural instantiation of language in the developing
brain.
In conclusion, we have tantalizing evidence that the RH plays some
important role in very early language development and that the RH,
under the right circumstances, can construct a normal mental grammar,
constrained by UG. However, while our findings indicate that the RH
may play a critical role at an early point in language development,
they do not disconfirm the LH as the seat of language in most major
respects in the normal (modal) case. We continue to examine this large
population more fully regarding the details of etiology and
grammatical development in the hope that this work will contribute to
our understanding of the potential of each hemisphere to construct a
normal mental lexicon and grammar and the circumstances under which it
will do so.
References
Aram, D. and Eisele, J. (1992). Plasticity and function recovery after
brain lesion in children. In H. Benton, et al. (Eds.), Developmental
Psychology. Milan: Fronco Angeli, 171-184.
Aram, D. M., Ekelman, B. L., Rose, D. F., & Whitaker, H. A. (1985).
Verbal and cognitive sequelae following unilateral lesions acquired in
early childhood. Journal of Clinical and Experimental Neuropsychology,
7( 1), 55-78.
Caplan, R., Guthrie, D., Fish, B., Tanguay, P.E., and David-Lando, G.
1989. The Kiddie Formal Thought Disorder Rating Scale, K-FTDS).
Clinical assessment, reliability, and validity. Journal of the
American Academy of Child and Adolescent Psychiatry, 28, 208-216.
Chomsky, N. (1995). The Minimalist Program. Cambridge, Massachusetts:
The MIT Press.
Cohen, M. (1992). Auditory/verbal and visual/spatial memory in
children with complex partial epilepsy of temporal lobe origin. Brain
and Language, 20, 315-326.
Curtiss, S., de Bode, S. and Mathern, G. 2000. Spoken Language
Outcomes after Hemispherectomy: Factoring in Etiology. Submitted to
Brain and Language.
Curtiss, S., & Schaeffer, J. (1997a). Syntactic development in
children with hemispherectomy: the INFL system. Proceedings of the
Boston University Conference on Language Development 20. Somerville,
MA: Cascadilla Press, PG. NOS..
Curtiss, S., & Schaeffer, J. (1997b). Development of the I- and
D-system in children with hemispherectomy. Paper presented at the
Fourth Annual Society for Cognitive Neuroscience.
Curtiss, S. and C. Jackson. (1989). How well can the right hemisphere
acquire language? Ms.
Day, P., & Ulatowska, H. (1979). Perceptual, cognitive, and linguistic
development after early hemispherectomy: two case studies. Brain and
Language, 10, 287-317.
de Bode, S. (1998). Interhemispheric language transfer and functional
plasticity. In A. Greenhill, M. Hughes, H. Littlefield, and H. Walsh
(Eds.), Proceedings of the 22nd Annual Boston University Conference on
Language Development. Somerville, Mass: Cascadilla Press, 134-140.
de Bode, S. and Curtiss, S. (1998). Neurobiological mechanisms of
language acquisition in Sturge Weber Syndrome. Paper presented at the
23rd Boston University Conference on language Development.
de Bode, S. and Curtiss, S. (1999). The relationship between
linguistic outcomes and etiologies of lesions leading to
hemispherectomies. To appear in Brain and Cognition, January
Dennis, M. (1980a). Capacity and strategy for syntactic comprehension
after left or right hemidecortication. Brain and Language, 10,
287-317.
Dennis, M. (1980b). Language acquisition in a single hemisphere:
semantic organization. In D. Caplan (Ed.), Biological Studies of
Mental Processes (pp. 159-185). Cambridge, MA: MIT Press.
Dennis, M., & Kohn, B. (1975). Comprehension of syntax in infantile
hemiplegics after cerebral hemicortication: left hemisphere
superiority. Brain and Language, 2(4), 472-482.
Dennis, M. and Whitaker, H. (1975). Language acquisition following
hemidecortication: linguistic superiority of left over the right
hemisphere. Brain and Language, 3, 404-433
Dravet, C., Guerrini, R., Mancini, J., Saltarelli, A., Livet, M. O., &
Galland, M. C. (1996). Different outcomes of epilepsy due to cortical
dysplastic lesions. In R. Guerrini (Ed.), Dysplasias of Cerebral
Cortex and Epilepsy . Philadelphia: Lippincott-Raven Publishers.
Duchowny, M., Jayakar, P., Harvey, A. S., Resnick, T., Alvarez, L.,
Dean, P., & Levin, B. (1996). Language cortex represantation: effects
of developmentsl versus acquired pathology. Annals of Neurology, 40(1),
31-38.
Eisele, J. A., & Aram, D. M. (1993). Differential effects of early
hemisphere damage on lexical comprehension and production.
Aphasiology, 7(5), 513-523.
Feldman, H. M., Holland, A. L., Kemp, S. S., & Janosky, J. E. (1992).
Language development after unilateral brain injury. Brain and
Language, 42(1), 89-102.
Fenson, L., Dale, P., Reznick, S., Bates, E., Thal, D., Reilley, J., &
Hartung, I. (1990). MacArthur Communicative Development Inventories.
Technical Manual. San Diego: San Diego State University.
Gadian, D. G., Isaacs, E. B., Cross, J. H., Connely, A. (1996).
Lateralization of brain function in childhood revealed by magnetic
resonance spectroscopy. Neurology, 46(4), 974-977.
Gallagher, T., & Watkin, K. (1996). 3-D ultrasonic fetal neuroimaging
and familial language disorders: in utero brain development. Montreal:
McGill University, Faculty of Medicine.
Gordon, N. (1996). Rasmussen's encephalitis. Developmental Medicine
and Child Neurology., 38, 133-136.
Gott, P. (1973). Language after dominant hemispherectomy. Journal of
Neurology and Neurosurgical Psychiatry, 36, 1082-1088.
Helmstaedter, C., Kurthen, M., Linke, D. B., and Elger, C.E. (1994).
Right hemisphere restitution of language and memory functions in right
hemisphere language-dominant patients with left temporal lobe
epilepsy. Brain, 117, 729-737.
Helmstaedter, C., Kurthen, M., Linke, D. B., & Elger, C. E. (1997).
Patterns of language dominance in focal left and right hemisphere
epilepsies: relation to MRI findings, EEG, Sex, and age at onset of
epilepsy. Brain and Cognition (33), 135-150.
Isaacs, E., Chrisie, D., Vargha-Khadem, F., & Mishkin, M. (1996).
Effects of hemispheric side of injury, age at injury, and presence of
seizure disorder on functional ear and hand asymmetries in hemiplegic
children. Neuropsychologia, 34(2), 127-137.
Krashen, S. (1972). Language and the left hemisphere. UCLA Working
Papers in Phonetics, 24.
Kennard, M. (1940). Relation of age to motor impairements in humans
and subhuman primates. Archives of Neurology and Psychiatry, 44,
377-397.
Kinsbourne, M. (1974). Mechanisms of hemispheric interaction in man.
In M. Kinsbourne & W. Smith (Eds.), Hemispheric Disconnection and
Cerebral Function. Springfiels, Il: Charles C. Thomas, 34-37.
Lecours, A., & Joanette, Y. (1985). Keeping your brain in mind. In J.
Mehler & R. Fox (Eds.), Neonate Cognition . Hillsdale, NJ: Lawrence
Erlbaum.
Locke, J. L. (1997). A theory of neurolinguistic development. Brain
and Language, 58, 265-326.
Lovett, M., Dennis, M., & Newman, J. (1986). Making reference: the
cohesive use of pronouns in the narrative discourse of hemidecorticate
adolescents. Brain and Language, 29, 224-251.
Mills, D., Coffey-Corina, S. A., & Neville, H. J. (1994). Variability
in cerebral organization during primary language acquisition. In G.
Dawson & K. W. Fischer (Eds.), Human Behaviour and Developing Brain
(pp. 427-455). New York: Guilford Press.
Molfese, D. (1989). Electrophysiological correlates of word meaning in
14-month-old human infants. Developmental Neuropsychology, 5, 79-103.
Molfese, D. a. Segalowitz, S. (1988). Brain Lateralization in
Children. New York: The Guilford Press.
Neville, H. (1993). Neurobiology of cognitive and language processing:
Effects of early experiences. In M. Johnston (Ed.), Brain Development
and Cognition. Oxford: Blackwell, 424-448.
Ogden, J. A. (1988). Language and memory functions after long recovery
periods in left-hemispherectomized subjects. Neuropsychologia, 26(5),
654-659.
Ogden, J. (1996). Phonological dyslexia and phonological dysgraphia
following left and right hemispherectomy. Neuropsychologia, 34(9),
905-918.
O'Leary, D., Lovell, M., Sackellares, C. et al. (1983). Effects of age
of onset of partial and generalized seizures on neuropsychological
performance in children. The Journal of Nervous and Mental Disease, v.
171, 624-629.
Peacock, W. J., Wehby-Grant, M. C., Shields, W. D., Shewmon, D. A.,
Chugani, H. T., Sankar, R., & Vinters, H. V. (1996). Hemispherectomy
for intractable seizures in children: a report of 58 cases. Child's
Nervous System( 12), 376-384.
Reilly, J., Bates, E., and Marchman, V. (1998). Narrative discourse in
children with early focal brain injury. Brain and Language, 64(3),
335-376
Riva, D., & Gazzaniga., L. (1986). Late effects of unilateral barin
lesions before and after the first year of life. Neuropsychologia, 24,
423-428.
Rossi, P., Parmeggiani, A., Santucci, M., et al. (1996).
Neuropsychological and psychiatric findings in cerebral cortex
dysplasias. In R. Guerrini et al. (Eds.), Dyspasias of Cerebral Cortex
and Epilepsy. Philadelphia: Lippincott Raven Publ., 345-351
Stark, R. E., Bleile, K., Brandt, J., Freeman, J., & Vining, E. P.
(1995). Speech-language outcomes of hemispherectomy in children and
young adults. Brain and Language( 51), 406-421.
Stark, R. (1997). Follow-up study of a right and a left
hemispherectomized child: implications for localization and impairment
of language in children. Brain and Language, 60(2), 222-243
Stiles, J., & Thal, D. (1993). Linguistic and spatial cognitive
development following early focal brain injury: patterns of deficit
and recovery. In G. Dawson & K. W. Fischer (Eds.), Human Behaviour and
Developing Brain (pp. 643-664). New York:: Guilford Press.
Strauss, E., & Verity, C. (1983). Effects of hemispherectomy in
infantile hemiplegics. Brain and Language, 20, 1-11.
Thal, D., Marchman, V., Stiles, J., Aram, D., Trauner, D., Nass, R.,
and Bates, E. (1991). Early lexical development in children with focal
brain injury. Brain and Language, 40, 491-527.
Vargha-Khadem, F., Isaacs, E. B., Papaleloudu, H., Polkey, C. E., and
Wilson, J. (1991). Development of language in six hemispherectomized
patients. Brain, 114, 473-495.
Vargha-Khadem, F., Carr, L. J., Isaacs, E., Brett, E., Adams, C., &
Mishkin, M. (1997). Onset of speech after left hemispherectomy in a
nine-year-old boy. Brain( 120), 159-182.
Vargha-Khadem, F., & Polkey, C. E. (1992). Review of cognitive outcome
after hemidecortication in humans. In F. D. Rose & D. A. Johnson
(Eds.), Recovery From Brain Damage (pp. 137-151). New York: Plenum
Press.
Witelson, S. (1982). Bumps on the brain. In S. Segalowitz (Ed.),
Language Functions and Brain Organization . New York: Academic Press.
Woods, B. T., & Carey, S. (1979). Language deficits after apparent
clinical recovery from childhood aphasia. Annals of Neurology, 6,
405-409.
Woods, B. T., & Teuber, H. L. (1978). Changing patterns of childhood
aphasia. Annals of Neurology, 3, 273-280.
Zaidel, E. (1973). Linguistic competence and related functions in the
right cerebral hemisphere of man following commissurotomy and
hemispherectomy. , California Institute of Technology, Pasadena.
Table 1. Subject Profles
Number/Sex
Side
Age/Surg
Age/Onset
Sz control
LangRank
CD:entire hemi
1M
Left
3;4
8days
no
0
2M
Left
2;9
10 days
yes
3;5
3M
Left
0;3
birth
yes
0
4F
Right
2;5
6 mo
yes
1
5F
Right
2;2
5 days
no
0
6M
Right
1;5
2 days
no
0
7F
Right
0;5
birth
yes
0
CD:multilobar involvement
8M
Left
1;8
6 weeks
yes
3
9F
Left
1;5
birth
yes
4
10M
Left
2;10
birth
yes
3
11M
Left
1;5
birth
no
2;5
12M
Left
1;0
birth
yes
1
13M
Left
1;8
birth
yes
4;5
14F
Left
0;5
birth
yes
3
15M
Left
0;8
2 mo
yes
3
16F
Right
0;4
birth
no
1;5
17F
Right
0;9
5 mo
no
3
18F
Right
1;1
birth
no
1
CD: focal/mild
19F
Left
6;8
birth
no
3
20M
Right
3;8
9 mo
no
1
Non-CD: Rasmussen
21M
Left
4;7
3;4yrs
yes
3;5
22M
Left
5;5
2;5yrs
yes
3
23F
Left
5;11
5;7 yrs
no
4
24M
Left
5;3
2;11
no
3
25F
Right
17;3
12 yrs
no
5
26F
Right
14;1
5 yrs
no
5
27F
Right
5;11
4;6 yrs
yes
5
28M
Right
3;5
2 yrs
yes
4
Non-CD: Infarction
29M
Left
8;6
birth
yes
3
30M
Left
6;2
10 mo
yes
4
31F
Left
4;0
birth
yes
3;5
32F
Left
9;8
6 mo
yes
5
33F
Left
6;8
birth
yes
4
34M
Left
3;9
6 mo
no
0
35M
Left
2;7
3 mo
no
2
36F
Left
1;3
birth
no
0
37M
Left
9;0
11mo
no
0
38M
Right
7;9
birth
yes
4
39F
Right
5;1
2 mo
no
2;5
40F
Right
4;3
4 mo
yes
3
41M
Right
2;2
8 mo
yes
5
42F
Right
0;10
4 mo
yes
0
Unknown Etiology
43F
Left
3;4
3 weeks
no
3
44F
Left
1;8
birth
no
0
45F
Left
1;3
birth
yes
4
46F
Left
1;0
2mo
no
2
47M
Right
4;3
21/2 mo
yes
5
48M
Right
2;11
2 weeks
no
0
Table 2. Summary of findings
Hypotheses
Supported
Or Not
Right
Left
Develop-mental
Acquired
1. The L hemispherectomy group will perform poorer than the R
hemispherectomy group
1.
Early surgery will lead to better linguistic outcome
2.
Late age at seizure onset and surgery will be directly related to
better linguistic outcome for R hemi-s
3.
Duration of seizures will be inversely related to quality of
language development
4.
Seizure control will be directly related to linguistic outcome
5.
Developmental pathology will be associated with a poorer
linguistic prognosis than acquired pathology
6a. The Rasmussen’s and Infarction group will outperform children with
cortical dysplasia4
6b. The entire hemisphere pathology (hemimegalencephaly) will lead to
the poorest outcome
No
P>0.5162
No
Yes
Partially, Only for RHs
Yes
Yes
Yes
p<0.016
Yes
p<0.0026
Mean SLR=
=2.26
Sz onset
p<0.0113
Surgery:
p<0.0018
p<0.0176
p<0.0498
Mean SLR=
=2.60
p<0.0306
p<0.0001
Mean SLR=
=1.93
n/s
Mean SLR=
=3.17
Curtiss and Schaeffer report two major findings in these papers: 1)
that the children who underwent left hemispherectomy had a
significantly higher incidence of funcrional category errors than
those who had undergone right hemispherectomy, and 2) that all
children were developing grammars embodying functional category
projections, thus evidencing acquisition patterns constrained by UG.
2 We are aware of counter-evidence to the first prediction (e.g,
Dennis 1980a,b); however, this account may still hold, once all
relevant factors are taken into account (see Curtiss, de Bode and
Mathern, 2000).
3 We excluded children from bilingual or non-English monolingual
backgrounds; thus our population included only 48 subjects, while
Peacock et al report on 58).
4 Gordon et al. (1996) suggest that Rasmussen’s may actually be a
developmental pathology rather than acquired. However, the vast
majority of neurologists agree that it is an acquired disorder.
5 The ‘Seizure Control' column shows the degree of seizure control
achieved postsurgically. It should be noted that even for those
children who continue to have seizures, significant seizure relief was
achieved. Persisting seizures suggest a bilateral focus.
6 CALC (Computerized Analysis and Language Coding System) is a
detailed morphological and syntactic analysis cased broadly on a
minimalist framework of syntactic theory (Chomsky, 1995).
7 From seizure onset to surgery
8 We did not explore extent of lesion. All children had pervasive
damage (hemispherectomies). Also, it has been suggested that extent of
lesion and cognitive outcome do not correlate (Dravet, Guerrini et
al. 1996).

4
33