close

Вход

Забыли?

вход по аккаунту

?

Correlation of widespread preoperative magnetic resonance imaging changes with unsuccessful surgery for hippocampal sclerosis.

код для вставкиСкачать
Correlation of Widespread Preoperative
Magnetic
. * Resonance ‘Imaging Changes
with Unsuccessful Surgerv for
Flippocampal Sclerosis
I
U
d
S. M. Sisodiya, MRCP, PhD,” N. Moran, MRCP,” S. L. Free, PhD,* N. D. Kitchen, MD, FRCS(SN),*I
J. M. Stevens, FRCR,*$ W. F. J. Harkness, FRCS,? D. R. Fish, MD, F R O , * and S. D. Shorvon, MD, FRCP*
~
~~
Despite meticulous preoperative assessment, about 30% of patients with refractory partial epilepsy due to hippocampal
sclerosis fail to become seizure free after appropriate temporal lobe surgery. Perioperative complications, hippocampal
remnants, and bitemporal disease do not account for all failures; extrahippocampal epileptogenic tissue must persist in
some patients. Such dual pathology is detected on routine visual inspection of magnetic resonance images in about 15%
of patients with hippocampal sclerosis, but most such patients are excluded from surgery. We postulated that some
patients have occult extrahippocampal cerebral structural abnormalities (i.e., subtle dual pathology) and that the presence
of these abnormalities would be associated with a poor surgical outcome. Quantitative postprocessing of preoperative
magnetic resunance images from 27 patients subsequently proved to have hippocampal sclerosis demonstrated extrahippocampal structural abnormalities in 14, 10 of whom did not become seizure free, while 11 of 13 patients without such
changes did become seizure free (x2,p < 0.005). Such structural information may supplement clinical decision making
in some patients being evaluated for epilepsy surgery and help to explain the biological basis of poor outcome from such
surgery.
Sisodiya SM, Moran N, Free SL, Kitchen ND, Stevens JM, Harkness WFJ, Fish DR, Shorvon SD.
Correlation of widespread preoperative magnetic resonance imaging changes with
unsuccessful surgery for hippocampal sclerosis. A n n Neurol 1997;41:490-496
The most common cause of medically refractory epilepsy in adults is believed to be hippocampal sclerosis
(HS) [l]. In long-term follow-up of patients with HS
treated surgically, approximately two thirds become seizure free [l-51: The remainder continue to have seizures. These figures have improved little despite more
sophisticated preoperative assessment [6]. Surgery is expensive [7] and can lead to psychosocial and neurological morbidity. Therefore, understanding the reasons
for failure of surgery to abolish seizure activity is important from scientific, individual, and service provision perspectives.
Although in particular patients, failure of surgical
treatment can be ascribed to perioperative complications or incomplete resection, there remain patients
with histologically proven HS and apparently complete
resections who do not become seizure free, for reasons
that are not understood [4,81. Predictors of outcome
for temporal lobe surgery are imperfect; a history of
febrile convulsions [9] and the detection of hippocam-
pal volume loss, asymmetry, or signal changes on preoperative magnetic resonance imaging (MRI) [4, 10,
111 are favorable factors, but none guarantee seizure
freedom. In uncomplicated patients with a poor outcome, epileptogenic tissue must still be present [la].
Such “dual pathology”, with lesions in addition to HS,
may clearly be visible on preoperative MRIs [i3, 141.
Dual pathology is most common in patients with neuronal migration defects, or cerebral dysgenesis [ 141,
and among these the most common association is between subependymal heterotopia and HS [15]. Not all
pathologies, however, are visible on MRIs, and there
may be subtle abnormalities in addition to visible
changes of HS on preoperative MRIs. We previously
showed widespread cerebral structural disruption, not
visible on routine inspection of high-resolution MRIs,
in patients with cerebral dysgenesis (CD) [16]. We
suggested that this finding might account for the generally poor outcome after surgical removal of the CD
area that appears localized with visual inspection alone
From the *Epilepsy Research Group, Department of Clinical Neurology, ?Department of Neurosurge~y,and $Department of Neuroradiology, Institute of Ncurology, National Hospital for Neurology
and Neurosurgery, London, UK, and National Society for Epilepsy.
Chalfont St. Pcrer, Rucks, UK.
Received May 2, 1996, and in revised form Sep 9. Accepted for
publication Sep 12, 1996.
490
Address correspondence to Llr Sisodiya, Department of Neurology,
Radcliffe Infknary, Woodstock Road, Oxford OX2 GHE, UK.
Copyright 0 1997 by the American Neurological Association
[ I , 171. In this study, we investigated patients with epilepsy and known HS using the same quantitative technique, to determine whether such widespread structural
disruption might also occur in some of these patients.
We hypothesized that patients with such additional abnormalities would not become seizure free after removal of a diseased hippocampus, and tested this hypothesis in a group of patients studied retrospectively.
Materials and Methods
Subjects
Thirty-three neurologically normal control subjects were
scanned after informed consent and approval of the local
ethics committee were obtained. Twenty-seven patients eval-
uated for epilepsy surgery were studied. The inclusion criteria for patients were that all should have (a) hippocampal
changes (qualitative or quantitative) on preoperative MRIs,
(b) HS o n histological examination, (c) complete hippocampal resection o n postoperative imaging (see below), and (d)
at least 2 years' follow-up since surgery. There was no age
difference between the normal control group and the patient
group (Mann-Whitney, p > 0.1). For the patients, the following data were collected (Table): age at operation, duration of habitual epilepsy prior to surgery, neuropsychometric
data, electroencephalographic (EEG) findings, imaging findings, and histology. If present, an initial insult [9] relating to
the epilepsy was also noted. Some patients had intracarotid
amobarbital (Amytal) procedures; 2 had preoperative intracranial electrical studies. All patients underwent standardized
Clinical, Investigational, and Surgical Detaih
~
Psychometry
Patietii
No.
Age (yr) at
Operation
Sex
Ictal
EEG
Findings
VIQ
PIQ
Hippocampal
Volume Ratio1
Corrected
Volume (See
Notes)
IAP
Side of
Surgery
Follow-up
Histology
35
32
35
M
F
M
RT
rcr
RT
123
123
89
102
122
98
RfL 0.63'
RIL 0.72"
RIL 0.56"
C
C
C
R
R
R
HS
HS
HS
48
30
36
103
84
120
95
LIK 0.56"
RIL 0.60"
C
L
HS
C
R
HS
RT
RT
LT
RT
RT
81
88
86
112
74
121
RIL
RIL
LIR
RIL
RIL
LIR
C
C
C
C
C
ND
R
R
L
R
R
L
HS
HS
HS
HS
rr
99
79
97
88
82
97
F
LT
94
104
LIR 0.61-
ND
L
No HC;
TLG
HS
31
M
RT
99
73
RIL 0.37'
C
R
HS
26
14
15
16
17
18
19
18
25
27
26
23
23
M
LT
LT
RH
LT
RT
B
67
105
79
77
94
75
76
96
94
102
83
70
LIR
LIR
RIL
LIR
RIL
LIR
A
0.72'
0.71"
C
C
C
L
L
R
L
R
HS; p.
F
F
M
M
F
C
L
HS
28
36
31
36
36
29
20
28
F
R
74
75
WL 0.46"
C
R
HS
24
18
26
F
5
6
7
8
9
10
11
39
30
26
33
22
30
M
12
23
13
F
SAB
~
LT
RT
4
Surgical
Outcome
-
~
1
2
3
(mo)
42
25
F
F
NF
+
(IIIA)
F
NF
(111.4)
F
F
F
M
M
0.53"
0.84"
0.65"
0.70'
0.65"
0.72"
0.82"
O.9Oh
0.70"
0.51"
ND
HS
HS
HS
HS
HS
30
30
27
24
36
24
F
F
24
NF
(IIIA)
NF
(111.4)
21
22
48
22
M
M
RT
LT
133
104
137
127
RIL 0.75"
WL 0.8Gb
C
R
L
HS
HS
24
24
23
28
F
RH
107
112
RIL 0.95b
C
R
HS
36
24
25
26
27
22
22
32
38
F
LT
RT
RT
LT
91
99
80
103
89
92
91
122
LIR
RIL
WL
LIR
C
C
L
R
R
L
HS
24
24
25
28
F
F
F
0.47"
0.73"
0.60"
0.68"
ND
ND
ND
HS
HS
HS
F
NF (IIR)
F
F
i
F
NF (IIB)
F
F
NF (IB)
NF
WB)
NF
(IIIA)
N F (IID)
NF
(IVA)
NF
(IVN
+
+
+
+
F
F
F
F
"Hippocampus volume smaller than 2 SD from mean after correction for intracranial volume [19].
bBoth hippocampi within normal range (within 2 SD of mean) after correction.
R = right; L = left; R T or LT = right or left temporal ictal onset on videotelemerric recording; B = widespread bilateral abnormality; RH = right
hippocampal onset on intracranial recording; PIQ = performance I Q V I Q = verbal IQ; IAP = intracarotid amobarbital procedure; C = results
concordant with other findings and compatible with surgery without risk of amnestic syndrome; N D = not done; A = verbal memory deficit and
dysphasia after left-sided injection, no deficit after right-sided injection; HS = hippocampal sclerosis; No HC = hippocampal tissue not found on
histological examination, but typically associated temporal lobe gliosis (TLG) seen; p. = microdysgenesis; NF = not free (scale from [ 2 ] ) ; F = free (see
= < 2/80 abnormal values.
text); SAR = structurally abnormal brain (see text); + = > 1/80 abnormal values;
~
Sisodiya et al: Surgical Outcome and Preoperative MRI
491
temporal resections, with removal of neocortex sufficient to
allow access to mesial structures.
Scanning
Preoperative and postoperative MRI was performed on all
subjects according to the same protocol, on a 1.5-T General
Electric Signa unit (Milwaukee, W). A coronal spoiled gradient recall (SPGR) sequence was used for image analysis
(TR 35/TE 5/number of excitations [NEX] 1; flip angle 35
degrees, acquisition matrix 256 X 128, field of view 24 cm,
producing 124 contiguous slices, each 1.5-mm thick). Sagittal TI-weighted (TR 500/TE IO/NEX a), and axial proton
density (TR 2800/TE 30/NEX 1) and T2-weighted (TR
28001TE 90/NEX 1) images were also routinely acquired.
All images were reviewed by an experienced neuroradiologist
0.M. S.), who examined specifically for the presence of abnormalities in addition to either hippocampal changes or a
previous temporal lobectomy; on postoperative images, the
completeness of hippocampal resection was judged by looking for remnants of the head or body of the hippocampus,
and by determining the posterior extent of the removal; patients were selected only if resection of hippocampal tissue
extended at least to the middle level of the brainstern posteriorly. All hippocampal volumes were measured on an independent console using a previously published protocol [ 181,
with quantified interrater and intrarater reliability [ 191. In
addition, absolute hippocampal volumes were calculated in
each case by correction for intracranial volume [19].
Image Processing
Analysis of regional distribution of volume was performed on
all preoperative images using an MRI postprocessing method
previously described in detail [16]. In essence, the method
quantifies the amount of cortical gray matter or subcortical
matter (white matter and basal nuclei excluding the caudate)
within prescribed proportions of each individual hemisphere
(Fig). Normal ranges (lying within 3 standard deviations
from each mean for control subjects) for the 80 variables
thus generated can identify and locate abnormalities of the
proportional distribution of a particular tissue within the
hemispheres. No control subject had more than one abnormal value for the 80 variables. O n this basis, the presence of
more than one abnormal variable (of 80 in total) in a particular subject was considered abnormal and defined a “structurally abnormal brain” (SAB); this finding was previously
given a biological interpretation [ 161. The analysis requires
the segmentation of each image from each subject on a dedicated image-processing workstation (Allegro, ISG Technologies, Toronto, Canada), and takes a skilled operator approximately 3 hours per subject. All the images in this study were
segmented by one operator (S. M. S.), blinded to the postoperative outcome of the patients. Although for larger and
longer-term studies, more than one operator may be needed,
it is acknowledged that the ideal is for all studies to be processed by one operator [20]. Interrater and intrarater reliabilities have been shown to be high [16]: The intrarater reliability [21] for segmentation of gray and subcortical matter,
assessed on 4 subjects segmented completely five times each,
was over 0.95, while the same measure on analysis of regional distributions was also higher than 0.95.
492
Annals of Neurology
Vol 41
No 4
April 1997
Histology
The pathological diagnosis of HS was based on visual determination of gliosis and definite neuronal loss of ar least 30%
in sectors CAI and CA3, though formal hippocampal cell
counts were not performed (hippocampal volume loss on
MKI is also known to correlate with HS on histology [4, 10,
111). The presence of additional CD (e.g., microdysgenesis)
was also noted.
Outcome
All patients had a minimum duration of follow-up of 2 years
after surgery. Outcome was assessed using a strict interpretation of Engel’s scale [2], in that patients were considered to
be seizure free only if they had not had any seizures or auras
for a continuous period of at least 2 years up to the last
follow-up. Outcome was assessed by a worker (N. M.)
blinded to the MRI findings.
Statistics
Analysis was performed using SPSS (SPSS, Chicago, IL).
The statistical test used in each case is detailed. Significance
is taken at the 0.05 level.
Results
Patient Data
Patients were selected for surgery on the basis of concordance of clinical, neuropsychometric, E E G , and
routine MRI preoperative data. Five of 27 patients had
more widespread imaging or electroclinical abnormalities. T w o (Patients 16 and 23) had independent, bilateral temporal ictal onset on scalp E E G ; chronic intracerebral
recordings
showed
stereotyped
right
hippocampal icral onset of all seizures (preceding leftsided changes and clinical onset by many seconds) in
both patients. Patient 20 had widespread ictal scalp
EEG changes over the right cerebral hemisphere, b u t
had imaging evidence and histological proof of right
HS. Patients 22 a n d 23 had additional structural abnormalities (subependymal heterotopia, bilateral in Patient 23 a n d ipsilateral to the hippocampal disease in
Patient 22). Both had histologically proven HS. In
both, the heterotopic tissue was unchanged on postoperative imaging. Patient 19 had neurofibromatosis type
1; imaging revealed only hippocampal volume asymmetry (left side smaller) with no other cerebral abnormalities; clinically the seizures were left temporal
though EEG showed no clear ictal localization. An
amobarbital test showed that she would n o t suffer a
severe amnestic syndrome from a left temporal lobectomy, which she went on to have.
No patient had bilateral hippocampal volume loss
(see Table). No patient had significant perioperative
complications. T h e surgical outcome i n patients not
rendered seizure free is given in the Table. T h e percentage of all patients completely seizure free was 56%,
rising to 60% if only patients with only HS visible on
preoperative imaging are considered. The mean dura-
C
D
(A) Derivation of 80 quantitative variables: view f i o m the vertex of segmented right and lefi cerebral hemispheres of a control subject. Each hemisphere is divided independently into blocks, each of which stretches for one tenth of the anterior-posterior extent of
the hemisphere. Within each block, the volume o f cortical p a y matter (GM) and subcortical matter (SM) (white matter and basal
nuclei and thaLamus, excluding the caudate) is measured. Each volume is divided by the total segmented volume (of both hemispheres), producing 20 quantitative variables in each hemisphere. In addition, the volume o f each G M (B) or SM (c)block in the
le$ hemisphere is divided by the volume of its contrahteral homologue, producing a &rther 20 variables assessing symmety of distribution of GM or SM. Each GM block volume is also divided by the volume of its homologous ipsilateral SM block, generating a
jiuther 10 variables assessing grayhubcortical matter ratio in each block (not shown). The position of abnormal blocks can be determined; the darker blocks represent blocks with abnormal values (D)-in this patient (Patient ZI), there are abnormal values f i r
blocks containing the sclerotic hippocampus, an adjacent block, and noncontiguous blocks in the ipsilateral and contralateral hemispheres demonstrating the widespread nature of the structural changes. See Reference I G f o r &El details. (Fig C is from Sisodiya SM,
Free SL, Stevens ]M, et al. Widespread cerebral structural changes in patients with cortical dysgenesis and epilepry. Brain 1995;
118:1039-1050, by permission of Oxford University Press.)
Sisodiya et al: Surgical Outcome and Preoperative MRI
493
tion of postoperative seizure freedom for the 15
seizure-free patients was 30.9 months (range, 24-48
months). The mean duration of follow-up for all patients was 30.7 months (range, 24-48 months).
Analysis of Volume Distributions dnd Outcome
By definition, no control subjects had a SAB [16].
Fourteen patients had more than one abnormal value
for the 80 variables (median number of abnormal values, 5.5; range, 2-22), and therefore had SAB. Of
these 14 patients, 4 were seizure free (for 24, 27, 30,
and 36 months to the latest follow-up). Ten were not
seizure free; they included 6 with fully concordant preoperative data localizing the seizure focus to the excised
temporal lobe and 4 with some evidence of more widespread (but not discordant) preoperative data. Of the
13 patients without SAB, 1 1 were seizure free. There
was a significant difference between those predicted by
the quantitative analysis not to become seizure free
(ix., those with SAB) and those who might become
seizure free (i.e., those without SAB) with respect to
p < 0.005).
actual outcome
Of the 14 patients with SAB, 13 had ipsilateral abnormal values positioned outside the anterior-posterior
extent of the ipsilateral temporal lobe; 10 also had abnormal values in the contralateral hemisphere.
The positive predictive value of the test (i.e.,
whether the prediction that a patient will not be free of
seizures is in fact true) is 71%. The negative predictive
value (i.e., that a patient not predicted to be seizure
free is in fact seizure free) was 85%. The sensitivity of
the test was 83% and the specificity 73%.
Post hoc analysis of the patients grouped according
to the presence or absence of SAB showed that there
was no difference between the two groups (MannWhitney, two-tailed, p > 0.1) with respect to age at
operation, sex, duration of habitual epilepsy, presence
of hippocampal asymmetry, and preoperative performance or verbal I Q measures. In the SAB group, more
patients had had a right-sided operation (MannWhitney, p < 0.05). More patients who actually became seizure free had had an initial insult than did
patients who did not become seizure free
p <
0.05); the presence of an initial insult did not relate to
the presence of SAR
p > 0.1).
(x2,
(x2,
(x2,
Discussion
As Spencer wrote, “it is clear that MRI evidence of
hippocampal atrophy does not guarantee persistent seizure cure after temporal lobectomy” [ 8 ] . Explanations
for unsuccessful surgery have included the presence of
bitemporal structural abnormalities, extraternporal lesions, conservative removal of the diseased hippocampus, and widespread structural pathology (i.e., persisting epileptogenic zones) [4, 12, 22, 231.
494 Annals of Neurology
Vol 41
No 4
April 1997
The existence of bitemporal structural abnormalities
may be established preoperatively [ l o , 191. None of
our patients had bilateral hippocampal volume loss. Although this does not absolutely exclude bitemporal disease, as histologically proved hippocampal disease can
exist in the absence of MRI-detected volume loss [24],
it renders it unlikely. In addition, putative bilateral
sclerosis does not preclude a seizure-free outcome after
unilateral temporal lobectomy [ 101.
Incomplete excision of the diseased hippocampus is
sometimes cited as a possible cause of surgical failure.
In an extensive study quantifying resection postoperatively with MRI, Nayel and colleagues [25] showed
that seizure control could be achieved with mesiobasal
resection as far posteriorly as the anterior half of the
mesencephalon. In this study, one of the inclusion criteria was that the hippocampal excision should be complete to at least this level. Hippocampal remnant posterior to this is thought to be less relevant to seizure
outcome in most patients [26].Thus in our patients,
persistent ipsilateral epileptogenic hippocampal tissue is
unlikely to explain surgical failure.
The possibility that the epileptogenic zone is more
extensive than the hippocampus has been raised by a
number of workers [4, 12, 231. In some patients, including 2 of ours (Patients 22 and 23), more than one
structural abnormality is visible on preoperative MRIs.
The frequency of detection of such dual pathology is
increasing, HS being found in 15% of patients with
lesional epilepsy and additional CD in 15% of patients
with MRI-identified HS [13, 141. Both of our patients
with dual pathology on MRI had subependymal heterotopia; both had histologically proven HS. It is currently difficult to determine which patients with dual
pathology may become seizure free after resection of
one of the lesions, even if the lesion chosen for excision
is the one to which seizure onset appears to be localized preoperatively.
In reports published to date, both pathologies in patients with dual pathology are visible on routine inspection of MRI. The pathology, however, need not be
visible macroscopically [27]or on MRIs (as previously
suggested for failure of surgery in patients with an atrophic hippocampus [28]), or even by routine histology,
occasionally being revealed only with specialized techniques such as Golgi staining or electron microscopy
[29, 301.
We showed that the presence of additional extrahippocampal, cerebral structural abnormalities, identified
quantitatively on visually normal-appearing brain
MRIs, is significantly associated with a poor outcome
in patients with HS. Indeed, it may be that in these
patients, HS is a marker of a widespread abnormality
of cerebral structure, extrahippocampal components of
which partake of epileptogenesis after, or possibly before, temporal lobe surgery [28]. We cannot say
whether extrahippocampal changes are reorganizational
or ilysgenetic (the latter more likely if other data are
considered [13, 14, 161).
An identified initial insult (“initial precipitating injury” [9]), most commonly a childhood febrile convulsion, often associated with HS [9, 311 is held to predict
a favorable surgical outcome. The association is irnperfect: From Mathern and colleagues’ report [9], 28 to
38% of patients with an identified initial precipirating
insult did not achieve a class I outcome. In our group,
the absence of an insult was significantly associated
with surgical failure (predictive value of 58%); the
presence of SAB, on the other hand, was more highly
predictive of failure to become seizure free (predictive
value of 71%). A history of a precipitating insult does
not necessarily provide an explanation for the occurrence of epilepsy. Underlying, possibly developmental
or genetic, abnormalities could predispose to insults
becoming epileptogenic [32], as not everyone who has
febrile convulsions develops epilepsy [31]. In some patients, these underlying abnormalities may be widespread so that removal of a diseased hippocampus does
not remove the whole epileptogenic zone, and seizures,
once initiated, persist. Our technique may allow identification of such patients, permitting better prognostication within a generally more favorably responsive
group.
Two caveats need emphasis: (1) Our method depends critically on segmentation, the reliability of
which must be ensured if decisions are to be based on
its results; and (2) the absence of SAB does not imply
a patient will necessarily be free of seizures-there are
many other factors.
We identified abnormalities of cerebral configuration
(regional distribution of volume) due to extensive cerebral structural changes that may explain poor outcome following mesial temporal surgery. An advantage
of our method, though not its original intent, is the
possibility of a more precise prediction of a poor outcome than is currently possible without invasive studies. If the results are supported by the prospective study
currently underway, surgery that a priori is unlikely to
be unsuccessful may be considered more circumspectly,
with quantification supplementing clinical judgment.
We believe that the findings are likely to be independent of the method, so that development of other, less
laborious methods capable of detecting the same underlying biological changes may enable simpler evaluation of some patients for epilepsy surgery.
We gratefully acknowledge the support of the Wellcome Trust
(S. M. S.), the Patrick Berthoud Trust (N. M.), and the NSE
(S. M. S.).
We thank Dr J. Duncan for permission to study his patients.
Presented in part at a joint meeting of the Royal Society of Medicine, British Society of Clinical Neurophysiologists, and American
Academy of Clinical Neurophysiologists, London, June 1996.
References
1. Bruton CJ. The neuropathology of temporal lobe epilepsy. lnstitute of Psychiatry, Maudsley Monograph No. 3 1. Oxford:
Oxford University Press, 1988
2. Engel J Jr, Van Ness PC, Rasmussen TB, Ojeniann LM. Outcome with respect to seizures. In: Engel J Jr, ed. Surgical treatment of the epilepsies. 2nd ed. New York: Raven, 1993:60962 1
3. Engel J Jr. Surgery for seizures. N Engl J Med 1996;334:647652
4. Berkovic SF, McIntosh AM, Kalnins RM, er al. Preoperative
MRI predicts outcome of temporal lobectomy: an actuarial
analysis. Neurology 1995;45:1358-1363
5. Zentner J, Hufnagel A, Wolf HK, et al. Surgical treatment of
temporal lobe epilepsy: clinical, radiological, and histopathological findings in 178 patients. J Neurol Neurosurg Psychiatry
1995;58:666-673
6. Rasmussen TB. Surgical treatment of complex partial seizures:
results, lessons, and problems. Epilepsia 1983;24(suppl 1):S65-
S76
7 . Vickrey BG, Hays RD, Rausch R, et al. Outcomes in 248 patients who had diagnostic evaluations for epilepsy surgery. Lancet 1995;346:1445-1449
8. Spencer S S . MRI and epilepsy surgery. Neurology 1995;45:
1248-1250
9. Mathern G W , Babb TL, Vickrey BG, et al. T h e dinicalpathogenic mechanisms of hippocampal neuron loss and surgical outcomes in temporal lobe epilepsy. Brain 1995;1 18:105118
10. Jack CR, Trenerry MR, Cascino GD, et al. Bilaterally symnietric hippocampi and surgical outcome. Neurology 1995;45:
1353-1358
11. Van Paesschen W , Sisodiya SM, Connrlly A, et al. Quantitative
hippocampal MRI and intractable temporal lobe epilepsy. Neurology 1995;45:2233-2239
12. Nakasato N , Levesque MF, Babb TL. Seizure outcome following standard temporal lobectomy: correlation with hippocampal
neuron loss and extrahippocampal pathology. J Neurosurg
1992;77:194-200
13. Raymond AA, Fish DR, Stevens JM, et al. Association of hippocampal sclerosis with cortical dysgenesis in patients with epilepsy. Neurology 1994;44:1841-1845
14. Cendes F, Cook MJ, Watson C, et al. Frequency and characteristics of dual pathology in patients with lesional epilepsy.
Neurology 1995;45:2058-2064
15. Li LM, Fish DR, Sisodiya SM, et al. High resolution magnetic
resonance imaging in adults with partial or secondary generalized epilepsy attending a tertiary referral unit. J Neurol Neurosurg Psychiatry 1995;59:384-387
16. Sisodiya SM, Free SL, Stevens JM, et al. Widespread cerebral
structural changes in patients with cortical dysgenesis and epilepsy. Brain 1995;118:1039-1050
17. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37:476-487
18. Cook MJ, Fish DR, Shorvon SD, et al. Hippocampal volumetric and morphometric studies in frontal and temporal lobe epilepsy. Brain 1992;115:1001-1 01 5
19. Free SL, Bergin PS, Fish DR, et al. Methods for normalization
of hippocampal volumes measured with MR. AJNR 1995;16:
637-643
Sisodiya et al: Surgical Outcome and Preoperative MlU
495
20. Jackson EF, Narayana PA, Falconer JC. Reproducibility of nonparametric feature map segmentation for determination of normal human intracranial volumes with MR imaging data. J
Magn Reson Imaging 1994;4:692-700
21. Snedecor GW, Cochran WG. Statistical methods. 6th ed. Iowa
State University Press, 1967
22. Wingkun EC, Awad IA, Luders H, Awad C. Natural history of
recurrent seizures after resective surgery for epilepsy. Epilepsia
1991;32:851-856
23. Awad I, Rosenfeld J, Ah1 J, et al. Inrractable epilepsy and structural lesions of the brain: mapping, resection strategies, and seizure outcome. Epilepsia 1991;32:179-186
24. Jackson GD, Kuzniecky RI, Cascino GD. Hippocampal sclerosis without detectable hippocampal atrophy. Neurology 1994;
44:42-46
25. Nayel MH, Awds IA, I ~ d e r sH. Extent of rnesiobasal resection
determines oiitcome after temporal lobectomy for intractable
partial seizures. Neurosurgery 1991;29:55-61
26. Kim JH, Tien RT, Felsberg GJ, et al. Fast spin-echo MR in
496
Annals of Neurology
Vol 41
No 4
April 1997
27.
28.
29.
30.
31.
32.
hippocampal sclerosis: correlation with pathology and surgery.
AJNR 1995;16:627-636
Galaburda AM, Kemper TL. Cytoarchitectonic abnormalities in
developmental dyslexia: a case study. Ann Neurol 1979;6:94100
Fish DR, Spencer SS. Clinical correlations: MRI and EEG.
Magn Reson Imaging 1995;13:1113-1117
Huttenlocher PR. Dendritic development in neocortex of children with mental defect and infantile spasms. Neurology 1974;
24203-2 10
Purpura DP. Dendriric spine “dysgcnesis” and mental retardation. Science 1974;186:1 126-1 128
Kuks JBM, Cook MJ, Fish DR, et al. Hippocampal sclerosis in
epilepsy and childhood febrile seizures. Lancer 1393;342:13911394
Sloviter RS. The functional organization of the hippocampal
dentate gyrus and its relevance to the pathogenesis of temporal
lobe epilepsy. Ann Neurol 1994;35:640-654
Документ
Категория
Без категории
Просмотров
1
Размер файла
696 Кб
Теги
correlation, preoperative, hippocampus, change, unsuccessful, magnetic, widespread, imagine, surgery, sclerosis, resonance
1/--страниц
Пожаловаться на содержимое документа