Neurosurgical forum

Page 1

1170

Neurosurgical forum

L e t t e r s t o t h e e d i t o r

Drug Resistance

DITOR

: I read with interest the article by Rel-

lecke, et al: (Rellecke P, Kuchelmeister K, Schachenmayr

W, Schlegel J: Mismatch repair protein hMSH2 in primary

drug resistance in in vitro human malignant gliomas. J Neu-

rosurg 101:653–658, October, 2004).

Abstract

Object. The mismatch repair (MMR) system has previously been

implicated in acquired chemoresistance in malignant gliomas in hu-

mans. Its impact on the primary chemoresistance in glioblastoma

multiforme (GBM) has not been determined in detail, however.

Methods. The authors investigated the expression of both the

MMR genes (hMSH2, hMLH1, hPMS1, hPMS2, and hMSH6) at the

transcriptional level through reverse transcription–polymerase chain

reaction and the hMSH2 protein through Western blot and immuno-

histochemical analysis of tumor tissue and primary cell cultures of

25 in vitro human de novo GBMs without prior experimental treat-

ment. Results of these analyses were compared with data on in vi-

tro chemoresistance to nine drugs that are in general use in glioma

therapy.

All MMR genes were expressed in the GBMs, with no significant

difference among the individual tumors except in one respect; that

is, GBMs showed either relatively high levels or barely detectable

levels of hMSH2 messenger (m)RNA and protein expression. All

multiresistant tumors demonstrated high hMSH2 expression, and

all but two of the sensitive tumors exhibited low hMSH2 mRNA

levels.

Conclusions. Analysis of the data indicates that functional alter-

ations of the MMR system are involved in the primary drug resis-

tance in in vitro human malignant gliomas. Analysis of hMSH2 ex-

pression might therefore predict therapeutic responses in humans

with GBMs.

The authors demonstrate “All multiresistant tumors

showed high hMSH2 expression” and conclude “that high

hMSH2 expression might be a predictor of primary chemo-

resistance.”

These data and conclusions seem to differ from other re-

ports suggesting that an intact mismatch repair (MMR) sys-

tem is required for methylator-induced apoptosis and low or

absent expression of hMSH2 and hMLH1 MMR proteins

confers resistance.

1–4

I would be interested in the authors’ thoughts regarding

this discrepancy.

ICHARD

A. R

OVIN

, M.D.

Marquette, Michigan

References

1. D’Atri S, Tentori L, Lacal PM, Graziani G, Pagani E, Benincasa

E, et al: Involvement of the mismatch repair system in Temozolo-

mide-induced apoptosis. Mol Pharmacol 54:334–341, 1988

2. Friedman HS, Johnson SP, Dong Q, Schold SC, Rasheed BK, Big-

ner SH, et al: Methylator resistance mediated by mismatch repair

deficiency in a glioblastoma multiforme xenograft. Cancer Res

57:2933–2936, 1997

3. Son BH, Ahn SH, Ko CD Ka IW, Gong GY, Kim JC: Significance

of mismatch repair protein expression in the chemotherapeutic re-

sponse of sporadic invasive ductal carcinoma of the breast. Breast

J 10:20–26, 2004

4. Taverna P, Liu L, Hanson AJ, Monks A, Gerson SL: Charac-

terization of MLH1 and MSH2 DNA mismatch repair proteins in

cell lines of the NCI anticancer drug screen. Cancer Chemother

Pharmacol 46:507–516, 2000

ESPONSE

: We thank Dr. Rovin for his interest in our ar-

ticle on hMSH2 expression in primary glioblastoma cell

lines. To our surprise multiresistant glioblastomas showed

high MSH2 expression, a finding that is in contrast to re-

ports on other tumors.

The four papers cited by Dr. Rovin involve different tu-

mor entities and samples and used different methods. In

their manuscript, D’Atri and coworkers

investigated a lym-

phoma cell line and its repair-deficient subline. Son, et al.,

investigated 71 sporadic breast cancer specimens. Taver-

na and coworkers

investigated a leukemia cell line, two

colon cancer cell lines, and two cell lines derived from ovar-

ian cancers. They all found that low MMR expression

correlated with resistance against different chemotherapeu-

tic agents. In another approach Friedman, et al.,

demon-

strated the induction of a resistant phenotype in a glioblas-

toma xenograft.

The experimental data cannot be directly compared with

the results of our study. We investigated the constitutive

expression without previous drug treatment in primary hu-

man glioblastoma cell cultures. D’Atri, et al.,

Friedman, et

al.,

and Taverna, et al.,

used established cell lines. During

the establishment of a permanent line, different biological

changes occur which are the consequence of genetic alter-

ations and clonal expansion. Therefore, established tumor

cell lines are not representative for the primary tumor ma-

terial.

Taverna and colleagues

investigated cell lines from tu-

mor entities belonging to the hereditary nonpolyposis colo-

rectal cancer (HNPCC) spectrum. In the tumorigenesis of

these tissues, hMSH2 is constitutively downregulated either

by genetic mutations or by transcriptional silencing. The re-

sult is microsatellite instability (MSI). Both mutations and

MSI are absolutely rare in gliomas, indicating that an intact

MMR system is of different importance in these tissues and

glial cells.

The paper by Son, et al.,

shows a potential role of

hMLH1 in chemoresistance of primary breast cancer sam-

ples by immunohistochemistry. We found no significant dif-

ference in the expression of hMLH1 in gliomas, indicating

that the regulation of primary chemotherapy resistance is

different in breast tissue and glial cells.

What, in fact, is the functional role of MMR genes in

malignant gliomas prior to therapy? In accordance with

Wei and coworkers

we detected low expression levels

of hMSH2 in human gliomas, which is not equivalent to

the loss of expression due to mutations in tumors of the

HNPCC spectrum. Then, one would expect MSI. More

likely, these low levels could indicate a functional status of

low repair capacity. The high hMSH2 expression in the ma-

jority of the resistant tumors in our study could correspond

to an upregulation of the MMR system, probably resulting

in an enhanced repair capacity and a more chemoresistant

phenotype.

J. Neurosurg. / Volume 102 / June, 2005

06_05_JNS.3 6/1/05 3:52 PM Page 1170

Page 2

ÜRGEN

CHLEGEL

, M.D., P

.D.

Munich, Germany

References

1. D’Atri S, Tentori L, Lacal PM, Graziani G, Pagani E, Benincasa

E, et al: Involvement of the mismatch repair system in Temozolo-

mide-induced apoptosis. Mol Pharmacol 54:334–341, 1998

2. Friedman HS, Johnson SP, Dong Q, Schold SC, Rasheed BK, Big-

ner SH, et al: Methylator resistance mediated by mismatch repair

deficiency in a glioblastoma multiforme xenograft. Cancer Res

57:2933–2936, 1997

3. Son BH, Ahn SH, Ko CD, Ka IW, Gong GY, Kim JC: Sig-

nificance of mismatch repair protein expression in the chemo-

therapeutic response of sporadic invasive ductal carcinoma of the

breast. Breast J 10:20–26, 2004

4. Taverna P, Liu L, Hanson AJ, Monks A, Gerson SL: Character-

ization of MLH1 and MSH2 DNA mismatch repair proteins in

cell lines of the NCI drug screen. Cancer Chemother Pharma-

col 46:507–516, 2000

5. Wei Q, Bondy ML, Mao L, Gaun Y, Cunningham J, Fan Y, et al:

Reduced expression of mismatch repair genes measured by mul-

tiplex reverse transcription-polymerase chain reaction in human

gliomas. Cancer Res 57:1673–1677, 1997

Alagille Syndrome

DITOR

: We read with great interest the article by

Cowan and colleagues (Cowan JA, Barkhoudarian G, Yang

LJS, Thompson BG: Progression of a posterior commu-

nicating artery infundibulum into an aneurysm in a patient

with Alagille syndrome. J Neurosurg 101:694–696, Oc-

tober, 2004), which introduces Alagille syndrome for the

first time into the neurosurgical literature.

Abstract

The authors present a case in which a posterior communicating

artery (PCoA) infundibulum progressed into an aneurysm in a pa-

tient with Alagille syndrome (arteriohepatic dysplasia). The 3-mm

PCoA infundibulum had been noted on angiography studies ob-

tained 5 years earlier, prior to clip occlusion of a basilar tip aneu-

rysm. Recently, the patient presented to the emergency department

with the sudden onset of headache and decreased mental status. A

computerized tomography scan of the head with three-dimensional

angiography revealed no gross subarachnoid hemorrhage, but did

demonstrate a 5-mm PCoA aneurysm. Lumbar puncture demon-

strated xanthochromia and a large quantity of red blood cells. The

patient underwent open surgery for aneurysm clip occlusion and ob-

tained a good recovery.

This case illustrates the small but growing number of examples

of infundibulum progression. It also indicates the need for a close

follow up in patients with congenital abnormalities that may pose

an increased risk for what has traditionally been considered a benign

lesion.

As Alagille syndrome is primarily a disease of the hepat-

obiliary system, hepatologists and gastrointestinal special-

ists have long studied the many clinical implications of this

multifaceted disease process. The advances in treatment of

this syndrome since its initial description in 1975, specifi-

cally the impact of liver transplantation and diversion pro-

cedures, have resulted in children with Alagille syndrome

now living well into adulthood. As a result, a new clinical

entity of this syndrome has been unveiled that is germane to

the neurosurgeon. Cohort studies performed in the last dec-

ade have established the leading cause of mortality in these

patients to be intracranial hemorrhage.

As Cowan, et al.,

reported, the alarming rate of intracranial hemorrhage in pa-

tients with Alagille syndrome is 14%. In fact, patients with

Alagille syndrome today are more likely to die from intra-

cranial hemorrhage than from any other complication of

this syndrome. The mortality rate has been reported as high

as 25% in one of the largest series and yet the cause of in-

tracranial hemorrhage remains ill defined.

Further study of the Jagged1-Notch signaling receptor

pathway reviewed by Cowan, et al., may reveal the source

of increased rates of intracranial hemorrhage. We believe

the increased rates of intracranial hemorrhage may be

linked to aneurysmal subarachnoid hemorrhage, based on

our experience managing these patients, which we describe

in a report currently under review. A recent report by

Schlosser, et al.,

also reports an Alagille syndrome patient

with multiple intracranial aneurysms and subarachnoid

hemorrhage. The mutation in the Jagged1 gene clearly

plays a role in the development of the cardiovascular system

and angiogenesis, as Cowan and colleagues astutely identi-

fy.

Less certain is the role the Jagged1 gene plays in vessel

homeostasis and response to injury.

3,5

The interval develop-

ment of a posterior communicating artery aneurysm from

an infundibulum supports the conjecture that aberrations in

vessel homeostasis rather than an event at angiogenesis

would facilitate such a progression. If an aberrancy in ves-

sel homeostasis is the cause, the question remains whether

or not patients with Alagille syndrome are more likely to

form aneurysms and whether these aneurysms are more

likely to rupture because of this aberrancy. The young ages

of patients with Alagille syndrome and ruptured aneurysms

(reported by Cowan, et al., Schlosser, et al.,

and in our

report; median age 24) suggest, at the least, an alteration in

the natural history of intracranial aneurysms in these pa-

tients. Regardless of the cause, establishing the incidence of

intracranial aneurysms in Alagille patients and the role of

screening these patients is an important first step in address-

ing the leading cause of mortality in this syndrome.

UIS

M. T

UMIALÁN

, M.D.

ANIEL

L. B

ARROW

, M.D.

Emory University School of Medicine

Atlanta, Georgia

References

1. Crosnier C, Attié-Bitach T, Encha-Razavi F, Audollent S, Soudy

F, Hadchouel M, et al: JAGGED1 gene expression during human

embryogenesis elucidates the wide phenotypic spectrum of Ala-

gille syndrome. Hepatology 32:574–581, 2000

2. Kamath BM, Spinner NB, Emerick KM, Chudley AE, Booth

C, Piccoli DA, et al: Vascular anomalies in Alagille syndrome: a

significant cause of morbidity and mortality. Circulation 109:

1354–1358, 2004

3. Lindner V, Booth C, Prudovsky I, Small D, Maciag T, Liaw L:

Members of the Jagged/Notch gene families are expressed in in-

jured arteries and regulate cell phenotype via alterations in cell

matrix and cell-cell interaction. Am J Pathol 159:875–883, 2001

4. Schlosser HG, Kerner T, Woiciechowsky C, Benndorf G: Mul-

tiple cerebral aneurysms and subarachnoid hemorrhage in a pa-

tient with Alagille syndrome. Am J Neuroradiol 25:1366–1367,

2004

5. Villa N, Walker L, Lindsell CE, Gasson J, Iruela-Arispe ML,

Weinmaster G: Vascular expression of Notch pathway recep-

tors and ligands is restricted to arterial vessels. Mech Dev 108:

161–164, 2001

J. Neurosurg. / Volume 102 / June, 2005

Neurosurgical forum

1171

06_05_JNS.3 6/1/05 3:52 PM Page 1171

Page 3

Cortical Organization

DITOR

: I would like to welcome the article by

Nakada, et al: (Nakada T, Fujii Y, Kwee IL: Coerced train-

ing of the nondominant hand resulting in cortical reorga-

nization: a high-field functional magnetic resonance imag-

ing study. J Neurosurg 101:310–313, August, 2004), as it

contained a wealth of laterality indexed information not dis-

cussed by the authors.

Abstract

Object. The authors investigated brain strategies associated with

hand use in an attempt to clarify genetic and nongenetic factors in-

fluencing handedness by using high-field functional magnetic reso-

nance imaging.

Methods. Three groups of patients were studied. The first two

groups comprised individuals in whom handedness developed spon-

taneously (right-handed and left-handed groups). The third group

comprised individuals who were coercively trained to use the right

hand and developed mixed handedness, referred to here as trained

ambidexterity. All trained ambidextrous volunteers were certain

that they were innately left-handed, but due to social pressure had

modified their preferred hand use for certain tasks common to the

right hand.

Although right-handed and left-handed volunteers displayed vir-

tually identical cortical activation, involving homologous cortex

primarily located contralateral to the hand motion, trained ambi-

dextrous volunteers exhibited a clearly unique activation pattern.

During right-handed motion, motor areas in both hemispheres were

activated in these volunteers. During left-handed motion, the right

supplemental motor area and the right intermediate zone of the ante-

rior cerebellar lobe were activated significantly more frequently than

observed in naturally right-handed or left-handed volunteers.

Conclusions. The results provide strong evidence that cortical or-

ganization of spontaneously developed right- and left-handedness

involves homologous cortex primarily located contralateral to the

hand motion, and this organization is likely to be prenatally deter-

mined. By contrast, coerced training of the nondominant hand dur-

ing the early stages of an individual’s development results in mixed

handedness (trained ambidexterity), indicating cortical reorgani-

zation.

First, their finding of bilateral activation of motor cortex

(using high-field magnetic resonance [fMR] imaging) upon

moving the ostensibly dominant side of those who were

converted/coerced to “the right path” was a confirmation of

an earlier study by Siebner, et al.,

using positron emission

tomography (PET). The fact that moving the neurally non-

dominant hand is associated with bilateral activation of mo-

tor cortices has been known for a long time. For example,

Hoshiyama and Kakigi,

in stimulating the median nerves

at the wrist and recording movement-related cortical po-

tentials from the scalp of a group of right- and left-hand-

ed patients, showed sequential bilateral activation of both

hemispheres when members of each group wrote with their

nondominant hands. Writing with the dominant hand result-

ed in the activation of the contralateral side alone. The same

finding in the motor arena had been made even earlier by

Kristeva, et al.

documenting the impossibility of moving

fingers of both hands simultaneously.

The anatomy underpinning these findings has been de-

lineated recently, based on clinical correlation of diaschitic

weakness after callosal interruption as it relates to the hand-

edness of patients.

2,3

As seen in the data of Nakada, et al.,

however, not all avowed hand preferences are trustworthy

(more so among left-handed patients), as there are people

who were forcibly converted/coerced to the ways of the ma-

jority (right handedness), or they have done so on their own

as a young child, as moving either hand is always under our

own volition.

The data from the authors indicate that the choice of

one’s favorite hand is not entirely behavioral in nature,

which is also evident from the fact that the incidence of

right/left handedness in the congenitally blind is the same

as that in the sighted.

The incidence of self conversions

has yet to be determined precisely, but a rough estimate

shows that it occurs in no less than one in five in the gener-

al public.

Thus, the conclusion of the authors that “cortical organi-

zation for right and left handedness is virtually identical” is

not valid. Measurements of reaction time of two effectors

on each side is the most secure way for such determina-

tions.

2,3

RAJ

ERAKHSHAN

, M.D.

Charleston, West Virginia

References

1. Derakhshan I: A new collosal syndrome. Neurorehabil Neural

Repair 17:72–73, 2003 (Letter)

2. Derakhshan I: Callosum and movement control: case reports.

Neurol Res 25:538–542, 2003

3. Derakhshan I: Handedness and macular vision: laterality of motor

control underpins both. Neurol Res 26:331–337, 2004

4. Derakhshan I: In defense of the sinistrals: anatomy of handedness

and the safety of prenatal ultrasound. Ultrasound Obstet Gyne-

col 21:209–212, 2003

5. Hoshiyama M, Kakigi R: Changes of somatosensory evoked

potentials during writing with the dominant and non-dominant

hands. Brain Res 833:10–19, 1999

6. Kristeva R, Keller E, Deecke L, Kornhuber HH: Cerebral poten-

tials preceding unilateral and simultaneous bilateral finger move-

ments. Electroencephalogr Clin Neurophysiol 47:229–238,

1979

7. Siebner HR, Limmer C, Peinemann A, Drzezga A, Bloem BR,

Schwaiger M, et al: Long-term consequences of switching hand-

edness: a positron emission tomography study on handwriting in

“converted” left-handers. J Neurosci 22:2816–2825, 2002

ESPONSE

: We read with great interest the comments pro-

vided by the reader. We acknowledge that most studies are

built from the knowledge derived from those that have pre-

ceded them. We would like to underscore that our study

benefits twofold from high-field MR imaging, namely: ac-

curate neuroanatomical localization (a very desirable quali-

ty requisite for advancing our understanding of the role of

specific neuroanatomical substrates with respect to function

such as the primary motor cortex

), and subject-specific

analysis of data.

The latter contrasts to studies represent-

ing the cumulative average across multiple subjects, such

as those performed with PET, the first functional imaging

technique.

Whereas averred handedness is subject to uncertainty

for numerous reasons, including those the reader provided,

fMR imaging activation patterns in response to our para-

digm are not under the conscious control of a subject. Our

fMR imaging study identified specifically and consistently

those neuroanatomical substrates activated with movement

of the dominant hand (confirmed prior to fMR imaging

studies by a handedness inventory quotient), and demon-

strated these were anatomically homologous between left

Neurosurgical forum

1172

J. Neurosurg. / Volume 102 / June, 2005

06_05_JNS.3 6/1/05 3:52 PM Page 1172

Page 4

and right handedness. The consistency of the findings could

not have been expected if the averred handedness was

merely a whim. Whether fMR imaging can be embraced as

the conclusive test of hand dominance remains to be further

explored.

We hope that the reader finds the opportunities afforded

by high-field fMR imaging a real bonus to our continued

understanding of brain function.

SUTOMU

AKADA

, M.D., P

.D.

Brain Research Institute

University of Niigata

Niigata, Japan

References

1. Fujii Y, Nakada T: Cortical reorganization in subcortical hemipa-

resis: neural mechanisms of functional recovery and prognostic

implication. J Neurosurg 98:64–73, 2003

2. Nakada T, Suzuki K, Fujii Y, Matsuzawa H, Kwee IL: Indepen-

dent component-cross correlation-sequential epoch (ICS) analysis

of high field fMRI time series: direct visualization of dual repre-

sentation of the primary motor cortex in human. Neurosci Res 37:

237–244, 2000

Trigeminal Neuralgia

DITOR

: We wish to provide a brief update to our

article published in the March 2005 issue (Sheehan J, Pan

HC, Stroila M, Steiner L: Gamma knife surgery for trigem-

inal neuralgia: outcomes and prognostic factors. J Neuro-

surg 102:434–441, March, 2005). In that article, we not-

ed no case of facial dysesthesia after gamma knife surgery

(GKS).

Abstract

Object. Microvascular decompression (MVD) and percutaneous

ablation surgery have historically been the treatments of choice for

medically refractory trigeminal neuralgia (TN). Gamma knife sur-

gery (GKS) has been used as an alternative, minimally invasive

treatment in TN. In the present study, the authors evaluated the long-

term results of GKS in the treatment of TN.

Methods. From 1996 to 2003, 151 cases of TN were treated with

GKS. In this group, radiosurgery was performed once in 136 pa-

tients, twice in 14 patients, and three times in one patient. The types

of TN were as follows: 122 patients with typical TN, three with

atypical TN, four with multiple sclerosis–associated TN, and seven

with TN and a history of a cavernous sinus tumor. In each case, the

chosen radiosurgical target was located 2 to 4 mm anterior to the

entry of the trigeminal nerve into the pons. The maximal radiation

doses ranged from 50 to 90 Gy. The median age of the patients was

68 years (range 22–90 years), and the median time from diagnosis

to GKS was 72 months (range 1–276 months). The median follow

up was 19 months (range 2–96 months). Clinical outcomes and post-

radiosurgical magnetic resonance (MR) imaging studies were ana-

lyzed. Univariate and multivariate analyses were performed to eval-

uate factors that correlated with a favorable, pain-free outcome.

The mean time to relief of pain was 24 days (range 1–180 days).

Forty-seven, 45, and 34% of patients were pain free without med-

ication at the 1-, 2-, and 3-year follow ups, respectively. Ninety,

77, and 70% of patients experienced some improvement in pain

at the 1-, 2-, and 3-year follow ups, respectively. Thirty-three (27%)

of 122 patients with initial improvement subsequently experienced

pain recurrence a median of 12 months (range 2–34 months) post-

GKS. Among those whose symptoms recurred, 14 patients under-

went additional GKS, six MVD, four glycerol injection, and one

patient a percutaneous radiofrequency rhizotomy. Twelve patients

(9%) suffered the onset of new facial numbness post-GKS. Chang-

es on MR images post-GKS were noted in nine patients (7%). On

univariate analysis, right-sided neuralgia (p = 0.0002) and a previ-

ous neurectomy (p = 0.04) correlated with a pain-free outcome; on

multivariate analysis, both right-sided neuralgia (p = 0.032) and pa-

tient age (p = 0.05) were statistically significant. New onset of facial

numbness following GKS correlated with undergoing more than one

GKS (p = 0.002).

Conclusions. At the last follow up, GKS effected pain relief in

44% of patients. Some degree of pain improvement at 3 years post-

GKS was noted in 70% of patients with TN. Although less effective

than MVD, GKS remains a reasonable treatment option for those

unwilling or unable to undergo more invasive surgical approaches

and offers a low risk of side effects.

A 65-year-old man presented with a 2-year history of

right-sided TN. He underwent GKS in October 2001 at the

University of Virginia. He was treated with a single 4-mm

isocenter and a peripheral dose of 40 Gy to the right trigem-

inal nerve. By December 2001, the patient had complete

pain relief and was off medication. In October 2002, he not-

ed the onset of right-sided numbness that began near the

corner of his mouth. The numbness spread to include all

three distributions of the trigeminal nerve. He also devel-

oped a burning sensation in the right side of his tongue and

gums. He likened the sensation in his mouth to that when

one drinks coffee that is scalding hot. He stated that he fre-

quently bit his tongue because of the numbness.

A follow-up MR image of the brain failed to reveal any

abnormality in the pons or the region of the trigeminal

nerves. There was no evidence of abnormal enhancement or

edema. The patient was started on Neurontin and Lamictal,

which have provided limited relief of his symptoms.

The patient’s TN pain remains gone, but the dysesthesia

that began 1 year post-GKS persists. Compared with his

TN, which was more severe but intermittent in nature, the

numbness and dysesthesia are less severe but constant. Dys-

esthesias following GKS for TN are rare but have been pre-

viously noted.

1–3

ASON

HEEHAN

, M.D., P

.D.

ADISLAU

TEINER

, M.D., P

.D.

University of Virginia

Charlottesville, Virginia

References

1. Brisman R: Gamma knife surgery with a dose of 75 to 76.8 Gray

for trigeminal neuralgia. J Neurosurg 100:848–854, 2004

2. Maesawa S, Salame C, Flickinger JC, Pirris S, Kondziolka D,

Lunsford LD: Clinical outcomes after stereotactic radiosurgery

for idiopathic trigeminal neuralgia. J Neurosurg 94:14–20, 2001

3. Pollock BE, Phuong LK, Foote RL, Stafford SL, Gorman DA:

High-dose trigeminal neuralgia radiosurgery associated with in-

creased risk of trigeminal nerve dysfunction. Neurosurgery 49:

58–62, 2001

J. Neurosurg. / Volume 102 / June, 2005

Neurosurgical forum

1173

06_05_JNS.3 6/1/05 3:52 PM Page 1173