Changes in the Volumes of the Brain and Cerebrospinal Fluid Spaces After Shunt

8/12/2019 Changes in the Volumes of the Brain and Cerebrospinal Fluid Spaces After Shunt

1/6

Changes in the volumes of the brain and cerebrospinal uid spaces after shunt

surgery in idiopathic normal-pressure hydrocephalus

Kotaro Hiraoka a,, Hiroshi Yamasaki a,Masahito Takagi a, Makoto Saito a, Yoshiyuki Nishio a, Osamu Iizuka a,Shigenori Kanno a, Hirokazu Kikuchi a, Takeo Kondo b, Etsuro Mori a

a Department of Behavioral Neurology and Cognitive Neuroscience, Tohoku University Graduate School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai 980-8575, Japanb Department of Physical Medicine and Rehabilitation, Tohoku University Graduate School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai 980-8575, Japan

a b s t r a c ta r t i c l e i n f o

Article history:

Received 7 December 2009

Received in revised form 22 June 2010

Accepted 23 June 2010

Keywords:

Idiopathic normal-pressure hydrocephalus

Cerebrospinaluid shunt

Magnetic resonance imaging

Volumetry

Objectives:To investigate volumetric changes of the brain and cerebrospinal uid (CSF) spaces after shunt

surgery in shunt-responsive idiopathic normal-pressure hydrocephalus (iNPH), and correlations between

the changes and postoperative clinical improvements.

Methods: Twenty-one patients with shunt-responsive iNPH were studied. Magnetic resonance imaging

(MRI) of the brain was performed before and 1 year after surgery, and clinical symptoms were assessed by

the iNPH Grading Scale, a validated assessment tool of the triad of iNPH, the Modied Rankin Scale, the

Timed Up and Go Test, and neuropsychological tests including the Mini-Mental State Examination. The

volumes of the left cerebral hemisphere, infratentorial brain, ventricles, and suprasylvian and infrasylvian

subarachnoid CSF spaces were measured using an MRI-based volumetric technique.

Results: The volumes of the cerebral hemisphere and infratentorial brain did not change signi cantly after

shunt surgery (p =0.231, 0.109, respectively). The volumes of the ventricles and infrasylvian subarachnoid

CSF spaces were signicantly decreased (pb0.0001, b 0.05, respectively), with a mean change rate of26.1%

and 4.5%, respectively. The volumes of the suprasylvian subarachnoid CSF spaces increased signicantly

(pb0.0001), with a mean change rate of 43.5%. The decrease in ventricular volumes was signicantly

correlated with clinical improvement.

2010 Elsevier B.V. All rights reserved.

1. Introduction

Idiopathic normal-pressure hydrocephalus (iNPH) is a syndrome

manifesting as a triad of dementia, gait disturbance, and urinary

impairment, with ventricular enlargement and normal cerebrospinal

uid (CSF) pressure, and without preceding events such as subarach-

noid hemorrhage and meningitis. It is characterized by clinical

improvement following shunt placement[14]. Although the cause

and pathophysiology of iNPH remain unclear, the conventional view is

that it is due to obstruction of CSF circulation and/or absorption. The

effectiveness of shunt placement for the treatment of iNPH has been

established[14], but the mechanism of its effect has not yet been

claried.

Radiological studies have led to clarication of the morphological

features of iNPH, and dilation of the ventricles and sylvianssures as

well as narrowing of the subarachnoid CSF spaces on the high

convexity and interhemispheric ssure is seen on magnetic

resonance imaging (MRI)[5]. However, there have been few studies

investigating morphometric changes of the intracranial components

after shunt surgery. Kitagaki et al. showed that the mean postop-

erative CSF volume of ve patients with iNPH was signicantly

decreased in the sylvian space and in the ventricle, marginally

decreased in the basal cistern, and signicantly increased in the

suprasylvian space as compared with the preoperative volume [5]. In

the study by Anderson et al. [6], the ventricular volumes of 10 (91%)

out of 11 iNPH patients who underwent shunt surgery were

decreased, with a mean change rate of 39%. Volume changes of the

brain and the association between changes in volume of the

intracranial components and clinical outcomes after shunt surgery

remain to be claried.

In thisstudy, we investigated volumetric changesof the brain and

CSF spaces and their association with clinical outcomes after shunt

surgery. The CSF spaces were segmented into the ventricles and

suprasylvian and infrasylvian subarachnoid spaces, as previous

study[5] showed volume decrease of the ventricles, sylvian space,

and basal cistern, and volume increase of the suprasylvian

subarachnoid spaces. We assumed that cerebral volume increases

postoperatively along with the clinical improvements and as the

offset of ventricular volume decreases after shunt surgery. In

addition, we expected to nd a correlation between the volumetric

changes of the intracranial components and clinical improvement

after shunt surgery. We aimed to elucidate the pathophysiology of

Journal of the Neurological Sciences 296 (2010) 712

Corresponding author. Tel.: +81 22 717 7358; fax: +81 22 717 7360.

E-mail address:[email protected](K. Hiraoka).

0022-510X/$ see front matter 2010 Elsevier B.V. All rights reserved.

doi:10.1016/j.jns.2010.06.021

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s
http://dx.doi.org/10.1016/j.jns.2010.06.021http://dx.doi.org/10.1016/j.jns.2010.06.021http://dx.doi.org/10.1016/j.jns.2010.06.021mailto:[email protected]://dx.doi.org/10.1016/j.jns.2010.06.021http://www.sciencedirect.com/science/journal/0022510Xhttp://www.sciencedirect.com/science/journal/0022510Xhttp://localhost/var/www/apps/conversion/tmp/scratch_9/Unlabelled%20imagehttp://dx.doi.org/10.1016/j.jns.2010.06.021http://localhost/var/www/apps/conversion/tmp/scratch_9/Unlabelled%20imagemailto:[email protected]://dx.doi.org/10.1016/j.jns.2010.06.021


2/6

iNPH and the mechanism of shunt effect by precisely measuring

morphometric changes after shunt surgery in patients with denite

iNPH.

2. Methods

2.1. Subjects and study design

Patients who had been diagnosed as having iNPH and who hadundergone shunt surgery between March 2005 and November 2007

were enrolled in a prospective follow-up program. The patients were

diagnosed by neurologists as probable iNPH based on the diagnostic

criteria published by the Guidelines Committee of Idiopathic Normal

Pressure Hydrocephalus, the Japanese Society of Normal Pressure

Hydrocephalus [7], i.e., (1) individuals who develop symptoms in

their 60 s or older, (2) the presence of more than one of the triad of

gait disturbance, cognitive impairment, and urinary incontinence, (3)

MRI featuresof iNPH (i.e., both ventricular dilation (Evans IndexN0.3)

and narrowing of the subarachnoid CSF spaces on the high convexity

and interhemispheric ssure), (4) CSFpressureof 200 mm H2O orless

and normal CSF content, (5) positive CSF tap test, (6) clinical

symptoms that cannot be completely explained by other neurological

or non-neurological diseases, and (7) no obvious preceding diseasesthat could possibly cause ventricular dilation, including subarachnoid

hemorrhage, meningitis, head injury, congenital hydrocephalus, and

aqueductal stenosis. Routine laboratory tests and single photon

emission computed tomography (SPECT) of the brain were performed

for differential diagnosis, and MRI and clinical evaluations were

performed as a baseline assessment. After conrmation that there was

no contraindication to the surgical procedure, patients who fullled

the criteria underwent ventriculoperitoneal (VP) or lumboperitoneal

(LP) shunt surgery. After surgery, the patients were followed up for

1 year, and the pressure settings of their programmable valves were

adjusted in the outpatient clinic. One year after surgery, the patients

were re-admitted to the hospital for re-assessment of their clinical

symptoms and for radiological investigations.

In this study, only baseline and 1-year follow-up data were used.Only data from those patients who completed the 1-year follow-up

program and who had denite iNPH (i.e., those who showed clinical

improvements as described below) were included in order to avoid

contamination of the results with data from patients with unspecied

diagnosis. Clinical improvement after shunt surgery was dened as a

decrement of more than 1 point in total score on the iNPH Grading

Scale (iNPHGS)[8] compared to the baseline score. The iNPHGS is a

validated tool for assessment of the clinical triad. It is rated according

to theclinician'sobservation andinterview with thepatient and hisor

her caregiver in order separately to assess the severity of each

component of the triad. The score for each domain ranges from 0 to 4,

and higher scores indicate worse symptoms. Patients who dropped

out from the 1-year follow-up program and patients who did not

show clinical improvements 1 year after surgery were excluded from

volumetry.

The procedure followed the clinical study guidelines of the Ethics

Committee of Tohoku University Hospital and was approved by the

Internal Review Board. Written informed consent was obtained from

the patients or their families.

2.2. Clinical assessment

Clinical symptoms were assessedat baseline and 1 year after shunt

surgery. In addition to the iNPHGS, the modied Rankin Scale (mRS)

[9], Timed Up and Go (TUG) test[10], Mini-Mental State Examination

(MMSE) [11], Frontal Assessment Battery (FAB) [12], and the

modied version of the Neuropsychiatric Inventory (NPI) [1315]

were administered. In the TUG test, patients are timed while they rise

from an arm chair, walk 3m, turn, walk back, and sit down again, and

the number of steps is counted.

2.3. Shunt surgery and programmable valve adjustments

The operative procedure either VPshunt orLP shunt was selected

according to thecondition of the patient'scervicaland lumbarvertebrae

when investigated by spinal MRI and the preference of the patient and

his or her family. In cases of VP shunt, a catheter was placed in theanterior horn of the right lateral ventricle, and in cases of LP shunt

surgery, a catheter was placed in the lumbar subarachnoid space. The

abdominal-side catheter was guided to the abdomen subcutaneously

and inserted in the peritoneal cavity. We used the Codman-Hakim

programmable valve system in all shunt surgery. The initialpressure for

the shunt system was set before surgery, based on the patient's height

and weight [16,17]. Postoperatively, patients were followed up in

outpatient clinics; the pressure settings of their programmable valves

were adjusted step by step and clinical improvements and adverse

effects were noted. If clinical improvement was absent or insufcient,

the pressure setting was lowered by 1030 mm H2O over a period of1

2 weeks. If orthostatic headache or subdural effusion/hematoma was

observed by computed tomography (CT), the pressure setting was

increased by 30 mm H2O. Pressure adjustments were repeated until

optimal pressure for each patient was attained.

2.4. MR volumetry

Brain MRI was performed at baseline and 1 year after surgery.

Axial, three-dimensional spoiled gradient echo (SPGR) images were

obtained for volumetry. The images were generated with a 1.5-T MRI

unit (Signa Horizon LX CV/i; GE Healthcare, Milwaukee, WI).

Operating parameters were as follows: eld of view, 250 mm; matrix,

256256; contiguous sections, 1081.5 mm; repetition time, 20 ms;

echo time, 4.1 ms; and ip angle, 30. Axial T2-weighted images and

uid-attenuated inversion recovery (FLAIR) images were also

obtained for diagnosis. The volumes of the left cerebral hemisphere,

infratentorial brain, ventricles, and subarachnoid CSF spaces on the

left side were measured on the MR image (Fig. 1). The reason for notmeasuring the right cerebral hemisphere and subarachnoid space on

the right side was that there were artefacts caused by the shunt valve

on the postoperative images in the cases of VP shunt surgery. The

subarachnoid CSF spaces were segmented into two parts supra-

sylvian and infrasylvian as mentioned below.

The MRI data sets of all images were transmitted to a personal

computer from the MRI unit. Measurements were performed by one

investigator(K.H.), whowas blinded to (1)all clinicalinformation, (2)

the order (pre-surgery and post-surgery) of MRIs, and (3) the time of

enrolment. The right side of cerebral convexity and circumferential

tissues such as dura mater, cranium, subcutaneous fat, and scalp were

masked in both the pre-surgery and post-surgery images, sparing the

right lateral ventricle for volumetry, so that the investigator could not

discriminate by means of artefacts between the pre-surgery and post-surgery images in cases of VP shunt surgery. The MRI data sets were

analyzed using ImageJ 1.37 (NIH, Washington, USA), based on built-in

functions[18]. For volumetry, we used a combination of semiauto-

matic segmentation technique through density thresholding and

manual tracing, thereby avoiding partial voluming and observer bias.

The volume of each structure was obtained by automatically counting

the number of pixels within the segmented regions and then

multiplying the number by voxel size (0.9821.50= 1.44 mm3).

The reliability and validity of this method have been established and

described elsewhere[19].

The left cerebral hemisphere and infratentorial brain were

outlined and the ventricles were extracted with surrounding brain

parenchyma by tracing with a manually driven mouse cursor. The

boundary between the cerebral hemispheres and infratentorial brain

8 K. Hiraoka et al. / Journal of the Neurological Sciences 296 (2010) 712


3/6

was the plane of the intercollicular level of the midbrain. The

boundaries between the right and left hemispheres were traced

manually in the middle of the corpus callosum, septum pellucidum,

body of fornix, and anterior and posterior commissures. The ventral

boundary of the infratentorial brain was the plane including the apex

of the odontoid process. The ventricles included the bilateral lateral

ventricles, third ventricle, cerebral aqueduct of Sylvius, and fourth

ventricle. For the segmentation of the subarachnoid CSF spaces, the

ventricles were blacked out from the original MR images. The external

boundary of the subarachnoid CSF spaces was outlined by tracing the

dura mater with a manually driven mouse cursor. The boundary

between the right and left halves of the subarachnoid spaces was

determined on the basis of landmarks such as the falx cerebri and

cerebral aqueduct of Sylvius. The subarachnoid CSF spaces were

segmented into upper and lower parts along the intermediate plane

between the plane through the nasion and the external occipital

protuberance and the cranial top point of the dura mater, so that theupper part included the subarachnoid CSF spaces above the sylvian

ssure, and the lower part included the subarachnoid CSF spaces in

the sylvian ssure and below it.

Subsequently, each intracranial component was automatically

segmented on the extracted images using density threshold set at a

range between minimum and maximum pixel values. For volumetry

of the cerebral hemisphere and infratentorial brain, the maximum

value was the largest pixel value of the brain, and the minimum

value was half the mean pixel value of the gray matter (the caudate

head) and the mean value of the CSF, since the gray matter and CSF

are predominant constituents of the brainextrabrain interface. For

volumetry of the ventricles and subarachnoid CSF spaces, the

maximum value was half the value of the mean pixel value of the

gray matter (the caudate head) and the mean value of the CSF, andthe minimum value was the minimum pixel value of the CSF (lateral

ventricles).

The testretest reliabilities of the volumetry were expressed as

intraclass correlation coefcients (ICCs), which werecalculated from

MRIs repeated after a 1-week interval and volumetry by a single

rater (K.H.) under blind conditions for 10 subjects. The ICCs for the

cerebral hemisphere, infratentorial brain, ventricles, and suprasyl-

vian and infrasylvian subarachnoid CSF spaces were 0.987, 0.992,

0.995, 0.987, and 0.984, respectively. Coefcients of variation

(standard deviation/mean, where standard deviation indicates the

square-root value of the arithmetic mean of 10 variance estimates)

were 0.662%, 0.775%, 1.593%, 1.720%, and 1.661% for the cerebral

hemisphere, infratentorial brain, ventricles, and suprasylvian and

infrasylvian subarachnoid CSF spaces, respectively.

2.5. Statistical analysis

The volumetric change rate of each intracranial component was

dened as the percentage change, which was calculated as postop-

erative volume minus baseline volume divided by baseline volume

(100). The changes of iNPHGS, mRS, MMSE, FAB, and modied NPI

were calculated by subtracting baseline scores from postoperative

scores. The change rates of time and steps in TUG test were calculated

as postoperative value minus baseline value divided by baseline value

(100). Two-tailed Student's t test was used for comparison of

volume change rates of each intracranial component between VP

shunt cases and LP shunt cases. Paired, two-tailed Student'sttest was

used for comparison of baseline and postoperative volumes of each

intracranial component. Spearman's correlation was used for explor-

atory correlation analysis between volumetric change rates of

intracranial components and change (rate) of clinical assessments.

All statistical analyses were performed on the statistical softwarepackage SPSS, version 11.0.1 J (SPSS Inc., Chicago, IL). The statistically

signicant level was set atpb0.05. The signicance level for multiple

comparisons was not corrected because of the explorative nature of

this study.

3. Results

During the study period, 38 patients with probable iNPH

underwent shunt surgery. Of these 38 patients, 12 dropped out due

to: ischemic stroke (n =3), malignancy (n =2), severe infectious

disease (n =2), chronic subdural hematoma (n =2), withdrawal of

consent (n = 1), institutionalization (n =1), and femoral fracture

(n =1). Therefore 26 patients completed the 1-year follow-up

program and were re-assessed 1 year after surgery. Of these 26patients, 21 showed clinical improvements and were selected for

Fig. 1. Segmentation of the intracranial components. The intracranial components were segmented into the left cerebral hemisphere, infratentorial brain (both sides), bilateral

ventricles, and suprasylvian and infrasylvian subarachnoid CSF spaces (left side).

Table 1

Demographic characteristics of the patients (n =21).

Sex (male/female) 8/13

Operative procedure (VP/LP) 15/6

Mean SD

Age at baseline (years) 76.2 3.6

Education (years) 9.7 2.9

Duration of disease (years) 3.1 1.4

Interval between operation and postoperative MRI (months) 12.8 0.8

VP, ventriculoperitoneal; LP, lumboperitoneal; MRI, magnetic resonance imaging; SD,

standard deviation.

9K. Hiraoka et al. / Journal of the Neurological Sciences 296 (2010) 712
http://localhost/var/www/apps/conversion/tmp/scratch_9/image%20of%20Fig.%E0%B1%80


4/6

volumetry. All these 21 patients were right-handed. Their demo-

graphic characteristics are summarized inTable 1,and the results of

their clinical assessments at baseline and 1 year after shunt surgery

are shown inTable 2.

Initially, we analyzed the volumetric changes of VP cases and LP

cases separately (Table 3). As both cases showed similar volumetric

changes, VP cases and LP cases were brought together for further

analysis. Volumetric results for the patients are shown in Table 4. The

volumes of the cerebral hemisphere and infratentorial brain did not

change signicantly after shunt surgery (p =0.231 and 0.109,

respectively). The ventricular volumes decreased signicantly

(pb0.0001) after shunt surgery, with a mean change rate of26.1%

(range 5.9% to 82.0%). There was a signicant increase in

suprasylvian subarachnoid CSF spaces (pb0.0001), with a mean

change rate of 43.5%, and the infrasylvian subarachnoid CSF spaces

showed a signicant decrease (pb0.05), with a mean change rate

of4.5%. The total volumes of the cerebral hemisphere, infratentorial

brain, ventricles, and suprasylvian and infrasylvian subarachnoid CSF

spaces showed a signicant decrease (pb0.001), with a mean change

rate of

3.4%.The results of the correlation analysis between volumetric changes

and changes in clinical assessments are shown inTable 5. The change

rates of the ventricular volume showed a signicant correlation with

changes in scores of the MMSE, FAB, scores for the gait domain in the

iNPHGS, and the mRS, and change rates related to time in the TUG test

(pb0.05).The correlations indicated that the patients whose ventricular

volume decreased most showed more clinical improvements than the

patients with less substantial ventricular volume decrease. Change rates

in the volume of the cerebral hemisphere, infratentorial brain, and

suprasylvian and infrasylvian subarachnoid CSF spaces did not show a

signicant correlation with changes in clinical parameters.

4. Discussion

In this study, we did not include subjects in whom no improvement

was noted after shunt surgery. While the exclusion of patients with

iNPH who showed no improvement because of treatment failure or

other problems possibly causes a selection bias, the exclusion of those

with an ambiguous diagnosis (including misdiagnosis and comorbidity

of other diseases)makesit possible to obtainuncontaminated datafrom

patients with a denite diagnosis, which is suitable for the pathophys-

iological study of iNPH.

Initially, we hypothesized that the volume of the brain increases

after shunt surgery as the symptoms of iNPH improve and

ventricular volume decreases. However, the results revealed that

the volumes of the cerebral hemisphere and infratentorial brain did

not change signicantly after shunt surgery. Previous longitudinal

MRI studies of Alzheimer's disease have indicated that the 1-year

whole-brain volume decrease was 0.98% to 2.8% [20,21], whereas the

1-year brain volume decrease in normal aging was 0.4% to 0.45%

[20,22].In this study, the rate of hemispheric volume change after

shunt surgery was compatible with the 1-year brain volumedecrease in normal aging. The results suggest that the volumes of

the cerebral hemisphere and infratentorial brain do not increase

after shunt surgery.

The volume of suprasylvian subarachnoid CSF increased, whereas

the volumes of the infrasylvian subarachnoid CSF spaces and

ventricles decreased signicantly. If the CSF in each part is

proportionally drained by shunt surgery, the CSF volumes in both

parts should have been reduced. Thecontrastingchanges maysuggest

the preoperative existence of a pressure gradient, i.e., the pressure of

the ventricles and infrasylvian subarachnoid CSF spaces is likely to be

higher than that of the suprasylvian subarachnoid CSF spaces. Shunt

surgery may relieve the pressure gradient, evacuate CSF from the

ventricles and the infrasylvian subarachnoid CSF spaces, and

consequently increase the CSF in the suprasylvian subarachnoidspaces. The existence of a transmantle pressure gradient in iNPH,

which is a pressure gradient between the ventricles and subarachnoid

CSF spaces, has not yet been established, as it has been found in some

studies[23,24]and not in others [25]. However, the ndings of this

study support the existence of such a pressure gradient. The pressure

gradient is against Pascal's principle, which states that pressure

exerted anywhere in a conned uid is transmitted equally in all

directions throughout the uid. However, the principle may not be

applicable in cases in which CSF ows through the complex

subarachnoid spaces synchronized with heart beats.

The total volume of cerebral hemisphere, infratentorial brain,

ventricles, and subarachnoid CSF spaces decreased signicantly after

shunt surgery. This may seem anomalous considering that the total

intracranial volume is unalterable in the elderly. The decrease of the

Table 2

Results of clinical assessment at baseline and 1 year after surgery (n =21).

Baseline Post-op p

valueMean SD Mean SD

TUG test Time (s) 20.8 11.0 14.3 6.8 0.001

Steps 27.5 12.9 22.3 8.6 0.030

MMSE (/30) 20.5 5.3 23.0 6.2 0.022

FAB (/18) 8.9 2.9 11.5 3.7 0.002

iNPHGS Gait (/4) 2.5 0.6 1.7 0.9 b0.001

Cognition (/4) 2.6 0.8 1.8 1.0 b0.001

Urination (/4) 1.9 1.1 0.8 0.9 b0.001

Total (/12) 6.9 1.9 4.1 2.2 b0.001

mRS (/6) 3.0 0.9 2.0 0.9 b0.001

Modied NPI (/144) 9.1 6.6 4.3 3.9 b0.001

Post-op, postoperative results; SD, standard deviation; TUG test, Timed Up and Go test;

MMSE, Mini-MentalState Examination;FAB, Frontal Assessment Battery;iNPHGS, iNPH

Grading Scale; mRS, modied Rankin Scale; NPI, Neuropsychiatric Inventory.

Table 3

Volumetric change rates for patients following VP shunt surgery (n =15) and LP shunt

surgery (n =6).

Change rate (%) p

VP shunt

surgery

(n =15)

LP shunt

surgery

(n =6)

Mean SD Mean SD

Cerebral hemisphere 0.9 2.3 0.1 2.0 0.35

Infratentorial brain 1.0 3.7 1.9 3.2 0.63

Ventricles 27.5 21.4 22.5 15.6 0.61

Subarachnoid CSF spaces Suprasylvian 41.2 44.2 49.4 21.8 0.67

Infrasylvian 4.4 10.6 5.0 7.7 0.90

Two-tailed Student'sttest.

VP, ventriculoperitoneal; LP, lumboperitoneal; SD, standard deviation; CSF,

cerebrospinaluid.

Table 4

Volumetric results for the patients (n =21).

Baseline Post-op pvalue Change

rate (%)

Mean SD Mean SD Mean SD

Cerebral

hemisphere

446.3 54.4 443.8 57.0 0.231 0.6 2.2

Infratentorial

brain

131.6 12.5 133.2 13.4 0.109 1.3 3.5

Ventricles 124.1 24.1 90.8 25.9 b0.0001 26.1 19.7

Subarachnoid

CSF spaces

Suprasylvian 33.5 11.8 44.8 9.5 b0.0001 43.5 38.8

Infrasylvian 103.1 18.2 97.5 13.6 b0.05 4.5 9.7

Total 838.5 89.9 810.2 89.0 b0.001 3.4 3.2

Unit: cm3.

SD, standard deviation; post-op, postoperative results; CSF, cerebrospinal uid.

The total indicates the sum of volumes of the cerebral hemisphere, infratentorial brain,

ventricles, and subarachnoid CSF spaces.



5/6

total volume may be explained by volume increase of the venous

sinuses in the cranium. In most of our cases, expansion of the cross-

section of thesagittal sinus after shunt surgery was observed by visual

assessment of MRIs, although the volume of the venous sinuses was

not measured in this study.

The results were similar whether the catheters were insertedin the lateral ventricles (VP shunt surgery) or in the lumbar

subarachnoid space (LP shunt surgery), which indicates that the

shunt effect on CSF dynamics is the same whether CSF is drained

from the lateral ventricles or lumbar subarachnoid space. Relief of

the abnormal pressure gradient by shunt surgery altered the

distribution of CSF in the cranium, and deformation of the

cerebral hemisphere caused by the pressure of CSF in the

ventricles was reduced, which may contribute to the improve-

ment in symptoms.

The clinical improvements after shunt surgery correlated with the

decreaseof ventricular volumes. As postoperative clinical assessments

and MRI were performed in a single time point for each patient, it

cannot be concluded that the clinical symptoms improved in

proportion to the ventricular volume decrease in individual patients.However, the correlation may suggest an association between the

ventricular volume decrease and improvement of the symptoms.

With reference to the pathophysiology of iNPH, the correlation may

also suggest some association between the ventricular enlargement

and the manifestation of symptoms. Some previous studies have

raised concerns about the role of frontal lobe dysfunction in the

pathophysiology of iNPH[2632]. A study of iNPH by Momjian et al.

showed autoregulation disturbance of cerebral blood ow in para-

ventricular white matter, which may be caused by the abnormal

pressure of theCSF in theventricles [33]. Theabnormalpressureof the

CSF in the ventricles towards the cerebrum deforms the cerebrum,

which probably leads to neural dysfunction, especially of the frontal

lobe.

In conclusion, the ndings in this study suggest the existence of atransmantle pressure gradient, which enlarges the ventricles, deforms

the cerebral hemispheres, and causes symptoms in iNPH. Shunt

surgery may relieve the abnormal pressure gradient and reduce the

deformation of the cerebrum, which may contribute to the improve-

ment in symptoms.

5. Conclusions

Shunt surgery changed the distribution of CSF in the cranium,

but did not change brain volume. It also reduced the deformation of

the cerebral hemisphere, which may result from the pressure of CSF

in the ventricles and may contribute to the development of

symptoms.

Acknowledgment

This work was partly supported by a Research Grant from the

Ministry of Health, Labour and Welfare of Japan (2008-Nanchi-17).

References

[1] Boon AJ, Tans JT, Delwel EJ, Egeler-Peerdeman SM, Hanlo PW, Wurzer HA, et al.The Dutch normal-pressure hydrocephalus study. How to select patients forshunting? An analysis of four diagnostic criteria. Surg Neurol 2000;53:2017.

[2] Raftopoulos C, Deleval J, Chaskis C, Leonard A, Cantraine F, Desmyttere F, et al.Cognitive recovery in idiopathic normal pressure hydrocephalus: a prospectivestudy. Neurosurgery 1994;35:397404.

[3] WeinerHL, Constantini S, Cohen H, Wisoff JH.Current treatmentof normal-pressurehydrocephalus:comparisonofow-regulatedand differential-pressureshunt valves.Neurosurgery 1995;37:87784.

[4] Krauss JK, Droste DW, Vach W, Regel JP, Orszagh M, Borremans JJ, et al.Cerebrospinaluid shunting in idiopathic normal-pressure hydrocephalus of theelderly: effect of periventricular and deep white matter lesions. Neurosurgery1996;39:2929.

[5] Kitagaki H, Mori E, Ishii K, Yamaji S, Hirono N, Imamura T. CSF spaces in idiopathicnormal pressure hydrocephalus: morphology and volumetry. AJNR Am J

Neuroradiol 1998;19:127784.[6] Anderson RC, Grant JJ, de la PazR, FruchtS, Goodman RR. Volumetric measurements

in the detection of reduced ventricular volume in patients with normal-pressurehydrocephalus whose clinical condition improved after ventriculoperitoneal shuntplacement. J Neurosurg 2002;97:739.

[7] IshikawaM, HashimotoM, Kuwana N, Mori E, Miyake H, WachiA, etal . Guidelinesfor management of idiopathic normal pressure hydrocephalus. Neurol Med Chir(Tokyo) 2008;48:S1S23 Suppl.

[8] Kubo Y, Kazui H, Yoshida T, Kito Y, KimuraN, Tokunaga H, et al.Validation of gradingscalefor evaluatingsymptomsof idiopathicnormal-pressurehydrocephalus.DementGeriatr Cogn Disord 2008;25:3745.

[9] Rankin J. Cerebral vascular accidents in patients over the age of 60. III Diagnosisand treatment Scott Med J 1957;2:25468.

[10] Podsiadlo D, Richardson S. ThetimedUp & Go: a testof basic functional mobilityfor frail elderly persons. J Am Geriatr Soc 1991;39:1428.

[11] Folstein MF, Folstein SE, McHugh PR.Mini-mental state. A practical method forgrading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:18998.

[12] Dubois B, Slachevsky A, Litvan I, Pillon B. The FAB, a Frontal Assessment Battery atbedside. Neurology 2000;55:16216.

[13] Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. TheNeuropsychiatric Inventory: comprehensive assessment of psychopathology indementia. Neurology 1994;44:230814.

[14] Hirono N, Mori E, Ikejiri Y, Imamura T, Shimomura T, Hashimoto M, et al. Japaneseversion of the Neuropsychiatric Inventorya scoring system for neuropsychiatricdisturbance in dementia patients. No To Shinkei 1997;49:26671.

[15] Mori S, Mori E, Iseki E, Kosaka K. Efcacy and safety of donepezil in patients withdementia with Lewy bodies: preliminary ndings from an open-label study.Psychiatry Clin Neurosci 2006;60:1905.

[16] Miyake H, Ohta T, Kajimoto Y, Nagao K. New concept for the pressure setting of aprogrammable pressure valve and measurement of in vivo shuntow performedusing a microowmeter. Technical note J Neurosurg 2000;92:1817.

[17] Miyake H, Kajimoto Y, Tsuji M, Ukita T, Tucker A, Ohmura T. Development of aquick reference table for setting programmable pressure valves in patients withidiopathic normal pressure hydrocephalus. Neurol Med Chir (Tokyo) 2008;48:42732 discussion 32.

[18] Rasband W. ImageJ. [http://rsb.info.nih.gov/ij/] website National Institutes of

Health, Bethesda, Maryland, USA; 1997.

Table 5

Correlation matrix contrasting volumetric change rates against clinical changes and change rate ( n =21).

( Chan ge r ate) (Change)

TUG test MMSE FAB iNPHGS mRS Modied

NPITime Steps Gait Cognition Urination Total

(Change rate)

Cerebral hemisphere 0.04 0.12 0.10 0.24 0.02 0.20 0.25 0.02 0.09 0.06

Infratentorial brain 0.12 0.00 0.11 0.13 0.03 0.12 0.08 0.01 0.05 0.04

Ventricles 0.45

0.47

0.49

0.58

0.48

0.35 0.32 0.50

0.46

0.26

Subarachnoid CSF spaces Suprasylvian 0.03 0.25 0.02 0.10 0.16 0.20 0.13 0.05 0.02 0.24

Infrasylvian 0.03 0.15 0.04 0.31 0.01 0.16 0.15 0.02 0.18 0.27

Spearman's correlation, pb0.05.

TUG test, Timed Up and Go test; MMSE, Mini-Mental State Examination; FAB, Frontal Assessment Battery; iNPHGS, iNPH Grading Scale; mRS, modied Rankin Scale; NPI,

Neuropsychiatric Inventory.

Change rate (%)=(postoperative valuepreoperative value)/preoperative value100.

Change=postoperative valuepreoperative value.

11K. Hiraoka et al. / Journal of the Neurological Sciences 296 (2010) 712


6/6

[19] Mori E, Hirono N, Yamashita H, Imamura T, Ikejiri Y, Ikeda M, et al. Premorbidbrain size as a determinant of reserve capacity against intellectual decline inAlzheimer's disease. Am J Psychiatry 1997;154:1824.

[20] Fotenos AF, Snyder AZ, Girton LE, Morris JC, Buckner RL. Normative estimates ofcross-sectional and longitudinal brain volume decline in aging and AD. Neurology2005;64:10329.

[21] Chan D, Janssen JC,Whitwell JL,WattHC, Jenkins R,FrostC, etal. Change inratesofcerebral atrophy over time in early-onset Alzheimer's disease: longitudinal MRIstudy. Lancet 2003;362:11212.

[22] Enzinger C, Fazekas F, Matthews PM, Ropele S, Schmidt H, Smith S, et al. Riskfactors for progression of brain atrophy in aging: six-year follow-up of normal

subjects. Neurology 2005;64:1704

11.[23] Conner ES, Foley L, Black PM. Experimental normal-pressure hydrocephalus isaccompanied by increased transmantle pressure. J Neurosurg 1984;61:3227.

[24] Hoff J, Barber R. Transcerebral mantle pressure in normal pressure hydrocephalus.Arch Neurol 1974;31:1015.

[25] Stephensen H, Tisell M, Wikkelso C. There is no transmantle pressure gradient incommunicating or noncommunicating hydrocephalus. Neurosurgery 2002;50:76371.

[26] Sakakibara R, Kanda T, Sekido T, Uchiyama T, Awa Y, Ito T, et al. Mechanism ofbladder dysfunction in idiopathic normal pressure hydrocephalus. NeurourolUrodyn 2008;27:50710.

[27] Iddon JL, Pickard JD, Cross JJ, Grifths PD, Czosnyka M, Sahakian BJ. Specicpatterns of cognitive impairment in patients with idiopathic normal pressurehydrocephalus and Alzheimer's disease: a pilot study. J Neurol NeurosurgPsychiatry 1999;67:72332.

[28] Stolze H, Kuhtz-Buschbeck JP, Drucke H, Johnk K, Illert M, Deuschl G. Comparativeanalysis of the gait disorder of normal pressure hydrocephalus and Parkinson'sdisease. J Neurol Neurosurg Psychiatry 2001;70:28997.

[29] Miyoshi N, Kazui H, Ogino A, Ishikawa M, Miyake H, Tokunaga H, et al. Associationbetween cognitive impairment and gait disturbance in patients with idiopathicnormal pressure hydrocephalus. Dement Geriatr Cogn Disord 2005;20:716.

[30] Lenfeldt N, Larsson A, Nyberg L, Andersson M, Birgander R, Eklund A, et al.

Idiopathic normal pressure hydrocephalus: increased supplementary motoractivity accounts for improvement after CSF drainage. Brain 2008;131:290412.[31] OginoA, Kazui H, Miyoshi N, HashimotoM, Ohkawa S, TokunagaH, et al.Cognitive

impairment in patients with idiopathic normal pressure hydrocephalus. DementGeriatr Cogn Disord 2006;21:1139.

[32] Sakakibara R, Hattori T, Yasuda K, Yamanishi T. Micturitional disturbance afteracute hemispheric stroke: analysis of the lesion site by CT and MRI. J Neurol Sci1996;137:4756.

[33] Momjian S, Czosnyka Z, Owler BK, Czosnyka M, Pena A, Pickard JD. Pattern ofwhite matter regional cerebral blood ow and autoregulation in normal pressurehydrocephalus. Brain 2004;127:96572.


Documents

Changes in the Volumes of the Brain and Cerebrospinal Fluid Spaces After Shunt