04.01 - Clinical Neuroendoscopy.pdf

CLINICAL NEUROENDOSCOPY

ix

1

9

19

33

CONTENTS

PrefaceRick Abbott

History of NeuroendoscopyRick Abbott

Since the beginnings of medicine, physicians have sought minimally invasive ways topeer into body cavities. It is only in the last several decades that the promises of endos-copy have begun to be answered. What follows is a brief outline of the development ofendoscopic technology and its application to the nervous system both for diagnostic andtherapeutic procedures.

The Endoscopic Management of Arachnoidal CystsRick Abbott

There is little doubt that most arachnoidal cysts will be managed endoscopically in thefuture given the advances we have seen over the last decade in our instrumentation.Excitement to employ this new technology should be governed by the reality that weare still learning and that our current success rate is not quite as good as what can beexpected when using microneurosurgery.

Basic Principles and Equipment in NeuroendoscopyVit Siomin and Shlomi Constantini

Understanding some of the basic principles of endoscopy and awareness of available re-sources can potentially be of considerable help to experienced neurosurgeons as well asbeginners in selection of the most appropriate tools for different procedures and makingcost-effective choices when browsing through multiple commercial advertisements andpurchasing new equipment. Although numerous advantages in science and industryhave made it possible to offer a wide variety of neuroendoscopes and tools, we believethe major achievements in this field are yet to occur. This particularly refers to the devel-opment of smaller fiberoptic scopes with better image quality and three-dimensional en-doscopes and to the invention of more efficient tools for endoscopic tumor removal withthe same degree of safety as in open surgery.

The Anatomy of the Ventricular SystemDavid G. McLone

The embryology of the ventricular development of the brain assists in understanding thefinal relations between structures forming these cavities. An accurate concept of thisanatomy allows the endoscopist to maneuver within the ventricular system.

VOLUME 15 Æ NUMBER 1 Æ JANUARY 2004 v

39

51

61

67

Selecting Patients for Endoscopic Third VentriculostomyHarold L. Rekate

Endoscopic third ventriculostomy has been used for about 70 years in the treatment ofhydrocephalus but was generally abandoned with the development of valve-regulatedshunts. With improvements in the understanding of the pathophysiology of hydro-cephalus and technical equipment improvements for endoscopy, there has been a resur-gence of interest in the procedure. Late-onset aqueductal stenosis is the ideal pathologiccondition responding to this treatment, but there are multiple other conditions that arepotentially responsive to internal bypass. All patients in whom the ventricles expand atthe time of shunt failure should be considered as candidates.

Techniques of Endoscopic Third VentriculostomyDouglas Brockmeyer

Modern techniques of endoscopic third ventriculostomy (ETV) are based on the conceptof establishing a natural conduit for cerebral spinal fluid (CSF) flow through the floor ofthe third ventricle. Through the years, a wide variety of techniques have been used as ameans to this end and have included both open and closed approaches. However, therelatively recent application of endoscopic technology to intraventricular surgery hasallowed neurosurgeons to perform third ventriculostomies in a minimally invasivefashion. Advances in third ventriculostomy technique have been based on a detailedunderstanding of third ventricular anatomy, surgical trajectories, and improved instru-mentation. The goal of this article is to discuss these issues in detail and to point outthe relevant risks and known complications associated with them.

Complications of Third VentriculostomyMarion L. Walker

As experience with ETV grows, the procedure will be performed by an increasing num-ber of neurosurgeons. Although the technique has been greatly refined since its adventalmost a century ago, today’s neurosurgeon must never forget that this seemingly simpleprocedure holds the potential for a number of devastating complications. Appropriatetraining and experience are important to the success of ETV and for avoiding complica-tions. It is imperative that surgeons continue to report their experience with the compli-cations of ETV so that the procedure can continue to be made as safe as possible.

Results of Endoscopic Third VentriculostomyMark R. Iantosca, Walter J. Hader, and James M. Drake

Endoscopic third ventriculostomy is emerging as the treatment of choice for aqueductalstenosis caused by anatomic, inflammatory, and selected neoplastic etiologies. The tech-nique has also proven useful in the pathologic diagnosis and treatment of these condi-tions. Long-term results of this procedure and comparison to standard shuntingprocedures are necessary to define indications for patients with pathologic findings inthe intermediate response groups. Development of new studies for preoperative assess-ment of cerebrospinal fluid absorptive capacity and quantitative postoperative measuresof ventriculostomy function would be invaluable additions to our ability to assess can-didates for this procedure and their eventual outcome. Further study and technical re-finements will, no doubt, lead to many more potential uses for these procedures inthe treatment of hydrocephalus and its associated etiologies. The challenge for neurosur-geons will be to define the operative indications and outcomes, while refining techniquesfor safely performing these useful procedures.

vi CONTENTS

77

89

105

Loculated Ventricles and Isolated Compartments in Hydrocephalus: TheirPathophysiology and the Efficacy of Neuroendoscopic SurgeryShizuo Oi and Rick Abbott

Trapped cerebrospinal fluid spaces can, on occasion, complicate the management ofhydrocephalus and present the surgeon with a treatment dilemma. This condition canbe categorized into one of two types: those arising as a complication of shunting andthose that arise as a complication of an inflammatory process within the ventricles.Whatever the cause, the result is a significant escalation in the complexity of the manage-ment of the patient. Neuroendoscopy is typically viewed as an attractive treatment alter-native in such a setting because of its minimalistic and thus seemingly simplistic nature.We have learned that nothing could be further from the truth. This article reviews thevarious entities that can arise in the hydrocephalic patient, how they can be managedendoscopically, and what sort of result can be expected.

Neuro-oncologic Applications of EndoscopyCharles Teo and Peter Nakaji

Neuro-oncology, in all its aspects, provides an ideal venue for the application of endos-copy. The main obstacle to its use has been neurosurgeons’ lack of familiarity with thetechniques and their advantages. As the neuro-oncologic surgeon uses the endoscopemore, endoscopy will take its rightful place in the surgeon’s armamentarium. The ad-vantages of improved visualization of intraventricular pathology, better managementof tumor-related hydrocephalus, less morbid biopsies, and minimally invasive removalof intraventricular tumors are invaluable adjuncts to traditional tumor management.Furthermore, endoscopy is the logical next step for surpassing the limitations of tradi-tional microsurgery. Endoscopy is still in its infancy. Rigorous application of the technol-ogy is increasingly allowing us to provide our patients the most maximally effective andminimally invasive surgery possible.

Index

CONTENTS vii

FORTHCOMING ISSUES

April 2004

Traumatic Neurovascular Surgery

J. Paul Elliot, MD, Guest Editor

July 2004

Pain Treatment

Gary Heit, MD, Guest Editor

October 2004

Metastatic Spine Tumors

Meic Schmidt, MD, Guest Editor

RECENT ISSUES

October 2003

Intraventricular Tumors

Andrew T. Parsa, MD, PhD, andMitchel S. Berger, MD, Guest Editors

July 2003

Neuroaugmentation forChronic Pain

Jaimie M. Henderson, MD, Guest Editor

April 2003

Surgery for Psychiatric Disorders

Ali R. Rezai, MD, Steven A. Rasmussen, MD,and Benjamin D. Greenberg, MD, PhDGuest Editors

THE CLINICS ARE NOW AVAILABLE ONLINE!

Access your subscription athttp://www.TheClinics.com

Neurosurg Clin N Am 15 (2004) ix

Preface

Clinical neuroendoscopy

Guest Editor

Rick Abbott, MD

Neuroendoscopy has come of age over the lastdecade and is now an invaluable tool to the prac-tice of neurosurgery. This is particularly the case

in practices with large volumes of patients withhydrocephalus. This issue of the NeurosurgeryClinics of North America is dedicated to neuroen-

doscopy with an emphasis on its use in the treat-ment of hydrocephalus. Readers will find helpfularticles on instrumentation and anatomy as well

as several articles on various aspects of use forhydrocephalus and related conditions. With a

1042-3680/04/$ - see front matter � 2004 Elsevier Inc. All ri

doi:10.1016/j.nec.2003.12.001

look toward the future there is also an article onthe use of the endoscopy in the removal of braintumors. By reading this issue, readers will gain

an appreciation of the state-of-the-art of intracra-nial neuroendoscopy.

Rick Abbott, MD

Inn, Beth Israel Medical Center170 East End Avenue

New York, NY 10128, USA

E-mail address: [email protected]

ghts reserved.

mailto:[email protected]

Neurosurg Clin N Am 15 (2004) 1–7

History of neuroendoscopyRick Abbott, MD

Clinical Neuroendoscopy, INN, Beth Israel Medical Center, 170 East End Avenue, New York, NY 10128, USA

The insertion of tubes into the body for eitherthe deliverance of a therapy or diagnosticpurposes has taken place throughout recorded

medicine. The first documented instance of the useof a tube to inspect the rectum was in Hippo-crates’ time. Natural light was used to illuminate

the field [1]. The Babylonians described usingvaginal speculums in 500 AD, with illuminationbeing provided by ambient light [2]. Abulkaism(980–1037) and Giulio Cesare Aranzi (1530–1589)

reported on the use of mirrors to reflect ambientlight down their ‘‘endoscopic’’ tubes to allow forthe inspection of deeper body cavities [3]. In 1805,

the Alert Faculty in Vienna heard PhilippeBozzini’s report on an instrument he had de-veloped, which used candle light reflected by

a concave mirror for the inspection of the bladderand rectum [4]. They were not impressed, censor-ing Bozzini for his inappropriate curiosity and

rejecting the ‘‘magic lantern.’’In 1867, Antonin Desormeaux published a de-

scription of an endoscope whose illumination hadbeen improved by using an alcohol and kerosene

burning candle with a chimney [5]. The light thiscandle generated was collected and focused downthe shaft of the scope using a lens. This was to be

the first successful design for a cystoscope, and itwas presented to the Academy of Medicine inParis. Shortly thereafter, there were several re-

ports of therapeutic uses for endoscopes. First,Bevan reported on successfully using an endo-scope to extract foreign material from theesophagus, and Pantaleoni then reported on using

a scope to inspect the uterus of a women botheredby postmenopausal bleeding, discovering an in-trauterine polyp and cauterizing it with silver



doi:10.1016/S1042-3680(03)00065-2

nitrate [4]. In 1870, Kussmaul demonstrated usinga rigid scope to inspect the stomach [5]. Theassistance of a professional sword swallower was

required for this, however.All these described systems suffered from a lack

of magnification. They were simply tubes that

directed illumination down to their distal tip. In1879, Max Nitze described the first system thatcontained a series of lens [6]. Working with severalopticians, he described a scope with an illumina-

tion source to the distal tip, a platinum wire thatglowed when current was conducted through it.The light produced was then projected through

a prism. This wire produced heat, however,necessitating a water coolant system. Shortlythereafter, Edison invented the light bulb, and

Newman then described modifying Nitze’s scopein 1883 by substituting a small light bulb at thedistal tip for the platinum wire [7]. Boisseau du

Rocher was the next to modify the system byfabricating an outer sheath to contain thetelescope, reporting on this in 1889 [4]. Thismodification allowed the surgeon to interchange

different scopes during a procedure withouthaving to renavigate through the body to theworking site.

By the turn of the century, the promise ofendoscopy had been demonstrated, but its accep-tance was slowed because of the poor illumina-

tion. Indeed, Pantaleoni’s students remarked onnot being able to appreciate his work, claiming aninability to see anything with his hysteroscope [4].The light source evolved during the first half of

the twentieth century, and by 1950, it consisted ofa tungsten bulb at the distal tip of the scope [8].Even this was inadequate, providing poor illumi-

nation and significant color distortion. In largepart, these difficulties were the result of designflaws inherent in the Nitze endoscope design (ie,

ghts reserved.


2 R. Abbott / Neurosurg Clin N Am 15 (2004) 1–7

a telescope containing a series of lens housed in anair-filled tube) [9]. The conduction of light in anendoscope is a function of the refractory index of

the conducting medium, in this case, air. The re-fractory index is defined as the difference in thespeed of light conduction in a vacuum and themedium in question. As it turns out, the refractory

index for glass is 1.5 times greater than that forair. Harold Hopkins, a British optical physicist,used this observation to construct a new endo-

scope designed to improve on light conduction.His scope interchanged glass for air and viceversa, resulting in a series of ‘‘air lens’’ housed in

a glass tube, or a series of glass rods. Not only didthis result in an increase in the system’s refractoryindex but also in an increase in the scope’s field ofview because of a decrease in spherical aberration

at the perimeter of the lens (thus increasing thelens’s light gathering capacity). The net gain inlight transmission of Hopkins’ scope design over

Nitze’s design was ninefold. Hopkins also usedcoating on the lens to minimize refraction of lightat the lens interface with the glass rods and

calculated the layout of the lens to eliminateghosting and chromatic distortion in addition tofurther increasing light transmission. With these

improvements, serious attention could be given tousing scopes within the human body, such as thebrain’s ventricles, where ambient light could notbe directed.

An additional problem with the Nitze designwas the heat generated by its illumination sourcehoused at the distal tip of the scope. The answer

seemed to be using a light generator outside ofthe body and conducting the light down to thedistal tip. The first report of use of an external

light source was by Fourestier, Bladu, andValmier in 1952, when a scope design wasdescribed where light was transmitted downa quartz rod from the scope’s external light

source [8]. This transmitted enough light to makeendoscopic photography possible. An improve-ment on this shortly followed based on work

done by Heinrich Lamm in 1932 [4]. At thattime, Lamm had demonstrated that light couldbe conducted through a bundle of glass fibers. In

1954, Harold Hopkins picked up on this ideaand with a colleague, N.S. Kapany, discussedapplying this finding to design a more effective

light conducting system [10]. They described twotypes of fiber bundles, so-called ‘‘incoherent’’ and‘‘coherent’’ bundles. Incoherent bundles referredto bundles of glass fibers whose orientation with

neighbors was chaotic or without order. This

type of bundle could be used to transmit lightfor illumination. Conversely, a coherent bundlemaintained the orientation of fibers through its

length so that any given fiber would maintain itsexact orientation to it is neighbors throughoutthe length of the bundle. This type of bundlecould be used to transmit an image from one end

of the bundle to the other. Hopkins and Kapanyobserved that this type of bundle could be usedin the design of a ‘‘flexible’’ endoscope whose

shaft could be bent to a degree and still allow forthe conduction of an image from one end of thescope to the other. In the same issue of Nature,

there was a paper describing a technique forcoating the fibers to minimize loss of light anddegradation of image quality [5].

Hirschowitz visited Hopkins’ laboratory short-

ly after the publication of his paper in 1954 andwas impressed with the promise of the observa-tions made by Hopkins and Kapany [5]. He

returned to the University of Michigan, where hecollaborated with Wilbur Peters and LawrenceCurtiss to develop a fiberoptic gastroscope. By the

end of 1956, Curtiss had developed a glass coatingfor the optical fibers that was permanentlyadherent and capable of conducting an image

for over a meter. In January 1957, they completedconstruction of their first fiberoptic gastroscope.The next month, it was used to inspect a gastriculcer in the wife of a dental student. Within 3

years, a commercial fiberoptic endoscope, Amer-ican Cystoscope Makers’ No. 4990 (AmericanCystoscope Makers, New York, NY), became

available. By 1962, Hirschowitz was able to pub-lish a series of 500 gastroesophageal endoscopies.

In 1963, Guiot described an endoscope de-

veloped for intracranial work that had a powerfulexternal light source whose illumination wasconducted via a quartz rod to the scope’s distaltip [9]. This allowed for color photography of the

ventricle. That same year Scarff described im-proving his ventriculoscope by substituting a ‘‘fi-ber lighting’’ system with an external light source

for the previously used incandescent bulb at thescope’s distal tip [9]. In 1983, the Welch AllynCompany (Skaneateles Falls, NY) released the

first endoscope with the charged couple device,allowing for conduction of a high-quality imagefrom the scope to a television screen [5]. Other

companies quickly followed with miniaturizationof the scopes, resulting in an improved ability todocument and teach. The ability to introduce rigidand flexible endoscopes into the body and to

obtain good-quality images was thus established.

3R. Abbott /Neurosurg Clin N Am 15 (2004) 1–7

As the use of endoscopes for diagnosticpurposes increased, not surprisingly, so too didthe desire to render therapy. In 1887, Felix M.Oberlander described the first scope designed to

treat postgonorrheal strictures [6]. The scopeallowed for direct visualization of the urethraand contained an instrument channel allowing for

the introduction of various knives for cutting thestrictures under direct vision. In 1910, L’Espinassereported to a local medical society in Chicago on

the use a cystoscope to fulgurate the choroidplexus in two infants with hydrocephalus [11].One child died immediately after surgery, but the

other lived for 5 years; thus was reported thefirst therapeutic neuroendoscopic case. ErichWossidlo, a German urologist, reported on usinga galvanocaustic hook to treat urethral strictures

[6]. In 1937, Ruddock introduced a scope thatcontained an electrocautery unit and biopsyforceps as well as ancillary biopsy instruments

for use in the peritoneal cavity [12]. In 1939,Crafoord and Frenckner described using sclero-therapy to treat esophageal varices [4]. In the

1970s, lasers were introduced into the armamen-tarium of medicine. In 1973, Nath and hiscolleagues [13] experimentally used an Nd:YAG

laser fiber with an endoscope to demonstrate itsfeasibility. Two years later, Fruhmorgan et al [14]reported on its use on a patient.

Although not as aggressive as some specialists,

neurosurgeons have been actively engaged in thedevelopment of endoscopic applications for thecentral nervous system. As mentioned earlier,

L’Espinasse first described use of the endoscopein the central nervous system in 1910. In 1922,Dandy [15] reported performing an endoscopic

choroid plexectomy after a previously reportedexperience in performing open choroid plexec-tomy on four patients. The attempt to perform theendoscopic choroid plexectomy was unsuccessful,

and he did not attempt any further such cases.The next year, Fay and Grant [16] reported onsuccessfully photographing the interior of the

ventricles of a hydrocephalic child. The fact thata 40-second exposure was required speaks to therather poor illumination available at that time. In

the same year, Mixter [17] performed the firstsuccessful endoscopic third ventriculostomy usinga cystoscope. This did not gain acceptance, pre-

sumably because of the poor visualization. In1932, Dandy [18] reported on using a cystoscopeto remove the choroid plexus. His results weresimilar to those experienced when he did the

surgery via a formal craniotomy. Shortly there-

after, Putnam [19] reported on the effectiveness ofsimply cauterizing the choroid plexus using a scopeof his design. Scarff [20] then wrote a paper onhis experience in developing a ventriculoscope

that allowed cauterization of the choroid plexus.He went on to develop a scope that containeda movable electrode for cauterization of the

plexus. In 1943, Putnam [21] reported on perform-ing endoscopic choroid plexectomy on 42 pa-tients. There were 10 (25%) perioperative deaths.

Fifteen of the patients failed to respond to thetreatment, but 17 experienced success in reliefof their intracranial hypertension. In 1970, Scarff

[22] reviewed all available series of endoscopicchoroid plexus cauterization for the treatmentof hydrocephalus. Of 95 patients so treated, 14(15%) had died, whereas 52 (60%) had initial

successful results. Scarff also looked at hispatients (39 of the 95 patients) at more than 5years after their surgery and found that 7 had died

of causes unrelated to their hydrocephalus and therest required no further treatment. More recently,Bucholz and Pittman [23] reported on using the

Nd:YAG laser to cauterize the choroid plexus ofa shunted infant with ascites, effectively decreas-ing cerebrospinal fluid (CSF) production by

50%. Pople and Griffith [24] commented on their20-year experience treating 156 individuals withchoroid plexus cauterization, finding a 35% long-term success rate.

As stated earlier, Guiot reported on anendoscope with a powerful external light sourcesufficient to allow for color photography of the

ventricle. Guiot reported on using this systemto perform third ventriculostomy. The difficultywith his system was its diameter, 9.1 mm. This

prevented a wider use of his scope, and Guiotultimately ceased performing third ventriculos-tomy with his instrument. Vries [25] described hisexperience with five hydrocephalic patients on

whom he performed endoscopic third ventriculos-tomies in 1978. Although he showed that theprocedure was technically feasible, none of his

patients remained without a shunt over the longterm. Jones et al [26] reported a different experi-ence in 1990, describing a 50% success rate in

long-term management of hydrocephalus in 24patients who underwent a third ventriculostomy.Modern series have improved on this figure, with

most now reporting 60% to 90% long-termsuccess with the technique [27–33].

Neurosurgeons have used endoscopes for otherindications in the central nervous system. In 1938,

Pool [34] reported on using an endoscope, termed


a myeloscope, to visualize dorsal nerve roots of thecauda equina. Several years later, he describedusing the instrument to view the spinal cord and its

conus. In 1990, the group at Hopital Neckerin Paris described their evolution in the treatmentof suprasellar arachnoidal cysts [35]. From firstperforming stereotactically guided fenestration of

the cyst, they moved to performing the procedureendoscopically. Others have reported on endoscop-ically fenestrating other arachnoidal cysts [36–41].

Although technically more challenging, theendoscope can be used to fenestrate other in-traventricular cysts such as arise after severe

ventriculitis. In 1986, Powers [42] described twoinfants with postinfectious multicystic ventricleswho he managed using a flexible endoscope andargon laser. He was able to fenestrate the cysts

into the ventricles successfully and cure thelingering infections. In 1992, Zamorano et aldescribed using a stereotactically guided endo-

scope to treat cystic ventricles and other entities[64]. The endoscope has also been used to assist inthe evacuation of intracranial hematomas [43–45].

The greatest success has been with chronichematomas, but surgeons have described attack-ing even acute clots [11,46,47].

There are now numerous reports in theliterature of the successful biopsying or removalof intracranial mass lesions endoscopically. In1983, Powell et al [48] first reported on the

removal of a colloid cyst using the endoscope.There have been many reports since [40,49–51,53],and in 1994, Lewis et al [52] showed that this type

of resection required less surgical time andpostoperative convalescence for their patients.Fukushima [54] reported on using a flexible

endoscope to biopsy tumors in 1978. Many havereported similar success over the past severaldecades [55–60]. Endoscopic surgery has also beenused to manage cystic tumors, such as cranio-

pharyngiomas, fenestrating the cysts into ven-tricles or cisternal spaces [61–65]. One particularsuccessful application for the neuroendoscope

has been to biopsy pineal region tumors whenperforming third ventriculostomies to treat theassociated hydrocephalus [66,67]. There has been

one report, however, of secondary seeding of theendoscope’s tract after such a biopsy [68]. As earlyas 1979, surgeons reported on using the endoscope

to resect pituitary lesions, and this has becomeextremely popular over the last decade [65,69–81].Finally, there are several surgeons investigatingthe utility of using an endoscope as an assisting set

of eyes when performing microneurosurgery

[60,61]. This has allowed for a much narrowersurgical corridor and for the surgeon to ‘‘lookaround corners’’ or on the back side of structures,

such as arteries. Fries and Perneczky [11] havealso spoken of improved appreciation of micro-anatomy not apparent with the microscope.

Other applications are finding their way into

the literature as neurosurgeons gain comfort inusing endoscopic equipment. Jimenez and Barone[82,83] have reported on using the endoscope to

perform sagittal strip craniectomies in conjunc-tion with molding helmets for the treatment ofscaphocephaly. As early as 1993, neurosurgeons

reported on using endoscopes to perform thoracicsympathectomies in a minimally invasive fashion[84–89]. In 1994, Schaffer [90] reported on re-section of an intervertebral disk using an endo-

scope, and several surgeons have described usingthe endoscope to perform various spinal surgeriessince [91–95].

It seems clear that the endoscope is a uniquetool and that it has a place in the armamentariumof the modern neurosurgeon. Its applications will

only broaden as we gain instrumentation andexperience in using this system.

References

[1] Rosin D. History. In: Rosin D, editor. Minimal

access medicine and surgery. Oxford: Radcliffe

Medical Press; 1993. p. 1–9.

[2] Gorden A. The history and development of

endoscopic surgery. In: Sutton C, Diamond M,

editors. Endoscopic surgery for gynaecologists.

London: WB Saunders; 1993. p. 3–7.

[3] Gotz F, Pier A. The history of laparoscopy. In:

Gotz F, et al, editors. Color atlas of laparoscopic

surgery. New York: Thieme; 1993. p. 3–5.

[4] Lau W, Leow C, Li A. History of endoscopic and

laparoscopic surgery. World J Surg 1997;21:444–53.

[5] Hirschowitz B. Development and application of

endoscopy. Gastroenterology 1993;104:337–42.

[6] Schultheiss D, Truss M, Jonas U. History of direct

vision internal urethrotomy. Urology 1987;52(4):

729–34.

[7] Dameword M. History of the development of

gynecologic endoscopic surgery. In: Azziz R,

Murphy A, editors. Practical manual of operative

laparoscopic andhysteroscopy.NewYork: Springer-

Verlag; 1992. p. 7–14.

[8] Bordelon B, Hunter J. Endoscopic technology. In:

Green F, Ponsky J, editors. Endoscopic surgery.

Philadelphia: WB Saunders; 1994. p. 6–17.

[9] Griffith H. Endoneurosurgery: endoscopic intracra-

nial surgery. In: Symon L, editor. Advances and


technical standards in neurosurgery. New York:

Springer-Verlag; 1986. p. 2–24.

[10] Hopkins H, Kapany N. A flexible light transmitting

tube. Nature 1954;173:39–41.

[11] Fries G, Perneczky A. Intracranial endoscopy. In:

Symon L, editor. Advances and technical standards

in neurosurgery. New York: Springer-Verlag; 1999.

p. 21–60.

[12] Haubrich W. History of endoscopy. In: Sivak M,

editor. Gastroenterologic endoscopy. Philadelphia:

WB Saunders; 1987. p. 2–18.

[13] Nath G, Gorish W, Kiefhaber P. The first

endoscopy via a fiber-optic transmission sytem.

Endoscopy 1973;5:208–12.

[14] Fruhmorgan P, et al. The first endoscopic laser

coagulation in the human GI tract. Endoscopy

1975;7:156–60.

[15] Dandy W. Cerebral ventriculostomy. Johns Hop-

kins Hospital Bull 1922;33:189.

[16] Fay T, Grant F. Ventriculostomy and intraventric-

ular photography in internal hydrocephalus. JAMA

1923;80:461–3.

[17] Mixter W. Ventriculoscopy and puncture of floor of

third ventricle. Boston Med Surg J 1923;188:277–8.

[18] Dandy W. The brain. In: Lewis D, editor. Practice

of surgery. Hagerstown: WF Prior Company; 1932.

p. 247–52.

[19] Putnam T. Treatment of hydrocephalus by endo-

scopic coagulation of the choroid plexus. Descrip-

tion of a new instrument and preliminary report of

results. N Engl J Med 1934;210:1373–6.

[20] Scarff J. Endoscopic treatment of hydrocephalus.

Description of ventriculoscope and preliminary

report of cases. Arch Neurol Psychiatry 1936;35:

853–60.

[21] Putnam T. The surgical treatment of infantile hy-

drocephalus. Surg Gynecol Obstet 1943;76:171–82.

[22] Scarff J. The treatment of nonobstructive (commu-

nicating) hydrocephalus by endoscopic cauteriza-

tion of the choroid plexus. J Neurosurg 1970;

33:1–18.

[23] Bucholz R, Pittman T. Endoscopic coagulation of

the choroid plexus using the Nd: YAG laser: initial

experience and proposal for management. Neuro-

surgery 1991;28:421–6.

[24] Pople I, Griffith H. Control of hydrocephalus by

endoscopic choroid plexus coagulation—long-term

results and complications. Eur J Pediatr Surg 1993;

3(Suppl 1):17–8.

[25] Vries J. An endoscopic technique for third ventri-

culostomy. Surg Neurol 1978;9:165–8.

[26] Jones RF, Stening WA, Brydon M. Endoscopic

third ventriculostomy. Neurosurgery 1990;26(1):

86–92.

[27] Choi JU, Kim DS, Kim SH. Endoscopic surgery

for obstructive hydrocephalus. Yonsei Med J 1999;

40(6):600–7.

[28] Elbabaa SK, Steinmetz, Ross J, Moon D, Luciano

MG. Endoscopic third ventriculostomy for obstruc-

tive hydrocephalus in the pediatric population:

evaluation of outcome. Eur J Pediatr Surg 2001;

11(Suppl 1):S52–54.

[29] Jones RF, Kwok BC, Stening WA, Vonau M.

Neuroendoscopic third ventriculostomy. A practi-

cal alternative to extracranial shunts in non-

communicating hydrocephalus. Acta Neurochir

Suppl (Wien) 1994;61:79–83.

[30] Murshid WR. Endoscopic third ventriculostomy:

towards more indications for the treatment of non-

communicating hydrocephalus. Minim Invasive

Neurosurg 2000;43(2):75–82.

[31] Panigrahi M, Vinas FC, Guthikonda M. Endo-

scopic third ventriculostomy: outcome analysis of

100 consecutive procedures. Neurosurgery 2000;

46(6):1534–5.

[32] Rieger A, Rainov NG, Brucke M, Marx T, Holz C.

Endoscopic third ventriculostomy is the treatment

of choice for obstructive hydrocephalus due to

pediatric pineal tumors. Minim Invasive Neurosurg

2000;43(2):83–6.

[33] Tuli S, Alshail E, Drake J. Third ventriculostomy

versus cerebrospinal fluid shunt as a first procedure in

pediatric hydrocephalus. Pediatr Neurosurg 1999;

30(1):11–5.

[34] Pool J. Direct visualization of dorsal nerve roots of

the cauda equina by means of a myeloscope. Arch

Neurol Psychiatry 1938;39:1308–12.

[35] Pierre-Kahn A, Capelle L, Brauner R, Sainte-Rose

C, Renier D, Rappaport R, et al. Presentation and

management of suprasellar arachnoid cysts. Review

of 20 cases. J Neurosurg 1990;73:355–9.

[36] Santamarta D, Morales F, Sierra JM, de Campos

JM. Arachnoid cysts: entrapped collections of

cerebrospinal fluid variably communicating with

the subarachnoid space. Minim Invasive Neurosurg

2001;44(3):128–34.

[37] Hopf NJ, Perneczky A. Endoscopic neurosurgery

and endoscope-assisted microneurosurgery for the

treatment of intracranial cysts. Neurosurgery 1998;

43(6):1330–7.

[38] Paladino J, Rotim K, Heinrich Z. Neuroendoscopic

fenestration of arachnoid cysts. Minim Invasive

Neurosurg 1998;41(3):137–40.

[39] Decq P, Brugieres P, Le Guerinel C, Djindjian M,

Keravel Y, Nguyen JP. Percutaneous endoscopic

treatment of suprasellar arachnoid cysts: ventric-

ulocystostomy or ventriculocystocisternostomy?

Technical note. J Neurosurg 1996;84(4):696–701.

[40] Hellwig D, Bauer BL, List-Hellwig E. Stereotactic

endoscopic interventions in cystic brain lesions.

Acta Neurochir Suppl (Wien) 1995;64:59–63.

[41] Caemaert J, Abdullah J, Calliauw L, Carton D,

Dhooge C, van Coster R. Endoscopic treatment of

suprasellar arachnoid cysts. Acta Neurochir (Wien)

1992;119(1–4):68–73.

[42] Powers SK. Fenestration of intraventricular cysts

using a flexible, steerable endoscope and the argon

laser. Neurosurgery 1986;18(5):637–41.


[43] Auer LM. Ultrasound stereotaxic endoscopy in

neurosurgery. Acta Neurochir Suppl (Wien) 1992;

54:34–41.

[44] Bauer BL, Hellwig D. Minimally invasive endo-

scopic neurosurgery—a survey. Acta Neurochir

Suppl (Wien) 1994;61:1–12.

[45] Karakhan VB, Khodnevich AA. Endoscopic sur-

gery of traumatic intracranial haemorrhages. Acta

Neurochir Suppl (Wien) 1994;61:84–91.

[46] Auer LM, Holzer P, Ascher PW, Heppner F. Endo-

scopic neurosurgery. Acta Neurochir (Wien) 1988;

90(1–2):1–14.

[47] Hellwig D, Bauer BL. Minimally invasive neuro-

surgery by means of ultrathin endoscopes. Acta


[48] Powell MP, Torrens MJ, Thomson JL, Horgan JG.

Isodense colloid cysts of the third ventricle: a di-

agnostic and therapeutic problem resolved by

ventriculoscopy. Neurosurgery 1983;13(3):234–7.

[49] Rodziewicz GS, SmithMV,Hodge CJ Jr. Endoscop-

ic colloid cyst surgery. Neurosurgery 2000;46(3):

655–62.

[50] King WA, Ullman JS, Frazee JG, Post KD,

Bergsneider M. Endoscopic resection of colloid

cysts: surgical considerations using the rigid endo-

scope. Neurosurgery 1999;44(5):1103–11.

[51] Deinsberger W, Boker DK, Samii M. Flexible

endoscopes in treatment of colloid cysts of the

third ventricle. Minim Invasive Neurosurg 1994;

37(1):12–6.

[52] Lewis AI, Crone KR, Taha J, van Loveren HR,

Yeh HS, Tew JM Jr. Surgical resection of third

ventricle colloid cysts. Preliminary results com-

paring transcallosal microsurgery with endoscopy.

J Neurosurg 1994;81(2):174–8.

[53] Heikkinen ER, Heikkinen MI. New diagnostic and

therapeutic tools in stereotaxy. Appl Neurophysiol

1987;50(1–6):136–42.

[54] Fukushima T. Endoscopic biopsy of intraventricu-

lar tumors with the use of a ventriculofiberscope.

Neurosurgery 1978;2(2):110–3.

[55] Greenberg IM. Intraoperative ultrasonography with

a cystoscope for the biopsy of a deep-seated brain

lesion: case study. Neurosurgery 1986;19(1):49–58.

[56] Otsuki T, Yoshimoto T, Jokura H, Katakura R.

Stereotactic laser surgery for deep-seated brain

tumors by open-system endoscopy. Stereotact

Funct Neurosurg 1990;54–55:404–8.

[57] Zamorano L, Chavantes C, Moure F. Endoscopic

stereotactic interventions in the treatment of brain

lesions. Acta Neurochir Suppl (Wien) 1994;61:92–7.

[58] Cohen AR. Ventriculoscopic surgery. Clin Neuro-

surg 1994;41:546–62.

[59] Jallo GI, Morota N, Abbott R. Introduction of a

second working portal for neuroendoscopy. A tech-

nical note. Pediatr Neurosurg 1996;24(2):56–60.

[60] Fries G, Perneczky A. Endoscope-assisted brain

surgery: part 2—analysis of 380 procedures. Neu-

rosurgery 1998;42(2):226–32.

[61] Cheng WY, Chang CS, Shen CC, Wang YC, Sun

MH, Hsieh PP. Endoscope-assisted microsurgery

for treatment of a suprasellar craniopharyngioma

presenting precocious puberty. Pediatr Neurosurg

2001;34(5):247–51.

[62] Abdullah J, Caemaert J. Endoscopic management

of craniopharyngiomas: a review of 3 cases. Minim

Invasive Neurosurg 1995;38(2):79–84.

[63] Caemaert J, Abdullah J, Calliauw L. Endoscopic

diagnosis and treatment of para- and intra-ventric-

ular cystic lesions. Acta Neurochir Suppl (Wien)

1994;61:69–75.

[64] Zamorano L, Chavantes C, Dujovny M, Malik G,

Ausman J. Stereotactic endoscopic interventions in

cystic and intraventricular brain lesions. Acta


[65] Halves E, Bushe KA. Transsphenoidal operation

on craniopharyngiomas with extrasellar extensions.

The advantage of the operating endoscope [proceed-

ings]. Acta Neurochir Suppl (Wien) 1979;28(2):362.

[66] Ferrer E, Santamarta D, Garcia-Fructuoso G,

Caral L, Rumia J. Neuroendoscopic management

of pineal region tumours. Acta Neurochir (Wien)

1997;139(1):12–21.

[67] Ellenbogen RG, Moores LE. Endoscopic manage-

ment of a pineal and suprasellar germinoma with

associated hydrocephalus: technical case report.

Minim Invasive Neurosurg 1997;40(1):13–6.

[68] Haw C, Steinbok P. Ventriculoscope tract recur-

rence after endoscopic biopsy of pineal germinoma.

Pediatr Neurosurg 2001;34(4):215–7.

[69] Liston SL, Siegel LG, Thienprasit P, Gregory R.

Nasal endoscopes in hypophysectomy. J Neurosurg

1987;66(1):155.

[70] Gamea A, Fathi M, el-Guindy A. The use of the

rigid endoscope in trans-sphenoidal pituitary sur-

gery. J Laryngol Otol 1994;108(1):19–22.

[71] Rodziewicz GS, Kelley RT, Kellman RM, Smith

MV. Transnasal endoscopic surgery of the pituitary

gland: technical note. Neurosurgery 1996;39(1):

189–99.

[72] Jho HD, Carrau RL. Endoscopic endonasal trans-

sphenoidal surgery: experience with 50 patients.

J Neurosurg 1997;87(1):44–51.

[73] Heilman CB, Shucart WA, Rebeiz EE. Endoscopic

sphenoidotomy approach to the sella. Neurosurgery

1997;41(3):602–7.

[74] Aust MR, McCaffrey TV, Atkinson J. Transnasal

endoscopic approach to the sella turcica. Am J

Rhinol 1998;12(4):283–7.

[75] Moreland DB, Diaz-Ordaz E, Czajka GA, Zugger

CM. Endoscopic resection of pituitary lesions

through the nostril. Semin Perioper Nurs 1998;

7(3):193–9.

[76] Moses RL, Keane WM, Andrews DW, Goel R,

Simeone F. Endoscopic transseptal transsphenoidal

hypophysectomy with three-dimensional intra-

operative localization technology. Laryngoscope

1999;109(3):509–12.


[77] Shen CC, Wang YC, Hua WS, Chang CS, Sun MH.

Endoscopic endonasal transsphenoidal surgery for

pituitary tumors. Zhonghua Yi Xue Za Zhi (Taipei)

2000;63(4):301–10.

[78] Moreland DB, Diaz-Ordaz E, Czajka GA. Endo-

scopic endonasal hemisphenoidotomy for resection

of pituitary lesions confined to the sella: report of 3

cases and technical note. Minim Invasive Neurosurg

2000;43(2):57–61.

[79] Jho HD. Endoscopic pituitary surgery. Pituitary

1999;2(2):139–54.

[80] Jarrahy R, Berci G, Shahinian HK. Assessment of

the efficacy of endoscopy in pituitary adenoma

resection. Arch Otolaryngol Head Neck Surg

2000;126(12):1487–90.

[81] Ogawa T, Matsumoto K, Nakashima T, Okano M,

Ono Y, Fukushima K, et al. Hypophysis surgery

with or without endoscopy. Auris Nasus Larynx

2001;28(2):143–9.

[82] Barone CM, Jimenez DF. Endoscopic craniectomy

for early correction of craniosynostosis. Plast

Reconstr Surg 1999;104(7):1965–75.

[83] Jimenez DF, Barone CM. Endoscopic craniectomy

for early surgical correction of sagittal craniosynos-

tosis. J Neurosurg 1998;88(1):77–81.

[84] Wang YC, Sun MH, Lin CW, Chen YJ. Anatom-

ical location of T2–3 sympathetic trunk and Kuntz

nerve determined by transthoracic endoscopy.

J Neurosurg 2002;96(1 Suppl):68–72.

[85] Kim BY, Oh BS, Park YK, Jang WC, Suh HJ,

Im YH. Microinvasive video-assisted thoracoscopic

sympathicotomy for primary palmar hyperhidrosis.

Am J Surg 2001;181(6):540–2.

[86] Wahlig JB Jr, Welch WC, Weigel TL, Luketich JD.

Microinvasive transaxillary thoracoscopic sympa-

thectomy: technical note. Neurosurgery 2000;

46(5):1254–8.

[87] Vanaclocha V, Saiz-Sapena N, Panta F. Uniportal

endoscopic superior thoracic sympathectomy. Neu-

rosurgery 2000;46(4):924–8.

[88] Chung HY, Seo CG, Lee SG. Video endoscopic

sympathectomy using a fiberoptic CO2 laser to

treat palmar hyperhidrosis. Neurosurgery 1993;

32(2):327–9.

[89] Kao MC. Percutaneous radiofrequency upper

thoracic sympathectomy. Neurosurgery 1997;40(1):

216–7.

[90] Schaffer JL. Endoscopic discectomy. J Neurosurg

1994;81(6):965–6.

[91] Adamson TE. Microendoscopic posterior cervical

laminoforaminotomy for unilateral radiculopathy:

results of a new technique in 100 cases. J Neurosurg

2001;95(1 Suppl):51–7.

[92] Jho HD. Endoscopic transpedicular thoracic dis-

cectomy. J Neurosurg 1999;91(2 Suppl):151–6.

[93] Karahalios DG, Apostolides PJ, Vishteh AG,

Dickman CA. Thoracoscopic spinal surgery. Treat-

ment of thoracic instability. Neurosurg Clin North

Am 1997;8(4):555–73.

[94] Frank E. Removal of a lateral disc herniation with

malleable endoscopic forceps: technical note. Neu-

rosurgery 1997;41(1):311–3.

[95] Onik G, Richardson D, Amaral J, Jennings W,

Sholes A. Percutaneous anterior discectomy under

ultrasound guidance. Minim Invasive Neurosurg

1995;38(2):90–5.


The endoscopic management of arachnoidal cystsRick Abbott, MDa,b,*

aDepartment of Neurosurgery, INN, Beth Israel Medical Center, 170 East End Avenue, New York, NY 10128, USAbDepartment of Neurosurgery, Albert Einstein School of Medicine, New York, NY, USA

Arachnoidal cysts are a not uncommon lesionfor a neurosurgeon, particularly a pediatric neu-rosurgeon, to be called on to treat. In one autopsy

series, the incidence of these cysts was 0.1% [1]. Oflesions arising intracranially, arachnoidal cystscomprise approximately 1% [2]. There has been

active debate as to how to manage these cysts best,with the controversy driven by their benignbehavior, subtle clinical sequelae, and the potentialfor treatment failure or complications. Not sur-

prisingly, the introduction of the neuroendoscopeinto a neurosurgeon’s armamentarium typicallyleads to the consideration of its use to treat these

entities in the hope of accomplishing what a crani-otomy can while avoiding the associated morbid-ity. In the following article, the wisdom of such an

approach is explored.

Pathology and pathogenesis

In 1831, Bright [3] first described an arachnoidcyst as ‘‘. . .a serous cyst forming in connectionwiththe arachnoid, and apparently lying between its

layers. . .’’ In his book, he discussed two cases,stating that these cysts seemed to be chronic, tohave a low potential for growth, and to be of sizes

varying from that of a pea to as large as or largerthan an orange. During the past century, theseobservations were confirmed with the introductionof microscopic neuropathology, and in 1978,

Rengachary et al [4] published a photomicrographillustrating Bright’s observation. This photomicro-

* Department of Neurosurgery, INN, Beth Israel

Medical Center, 170 East End Avenue, New York, NY

10128, USA.



doi:10.1016/S1042-3680(03)00071-8

graph demonstrated splitting of arachnoidal mem-brane at the margin of the cyst and a lack oftrabecula within the cyst, showing the cyst to arise

within the arachnoidal membrane and not withinthe subarachnoid space. The cyst membranescontained hyperplasic arachnoidal cells and a thick

layer of collagen.The pathogenesis, however, has been more

controversial. Until the 1970s, authors arguedover arachnoid cysts being either secondary

phenomena occurring in regions of agenesis ofthe brain or primary events of dysgenesis of thearachnoid investing the brain. In 1955, Robinson

[5] published a series of 15 patients with middlefossa arachnoidal cysts, hypothesizing that theywere cerebrospinal fluid (CSF) collections pas-

sively filling a space left by an agenesis of a portionof the temporal lobe. Starkman et al [6] put fortha countertheory in 1958 in a report of three

autopsies done on individuals with middle fossacysts, finding these cysts to be surrounded byarachnoidal membrane, which led him to con-clude that the cysts had arisen as a result of

splitting or duplication of the arachnoid duringdevelopment. With the introduction of CT scan-ning and the ability to image the brain before and

after treatment of these cysts, it became apparentthat there was a capacity for expansion of thetemporal lobe into space provided by decompres-

sion of the cyst. In 1971, Robinson [2] withdrewhis hypothesis of agenesis, stating that the cystswere caused by maldevelopment of the arachnoid.Robinson [2], the primary proponent of these

cysts being secondary phenomena, abandoned hisposition after observing that neurologic sequelaeto these cysts were not in proportion to the cyst’s

size and that brain re-expansion could be seen onCT after the cysts were treated.

ghts reserved.



Arachnoidal cysts typically arise within normalarachnoidal cisterns. They are believed to arisearound week 15 of gestation after the rupture of

the roof of the closed fourth ventricle and creationof the foramina of Luschka and Magendie. Thisallows for the escape of CSF into the subarachnoidspace, where it replaces intracellular ground sub-

stance filling the subarachnoid space. Duplicationor splitting of the arachnoid at the time of cisternformation provides the anlage for the arachnoidal

cyst. Similarly, splitting or duplication of theependymal lining of the lateral ventricles resultsin the formation of intraventricular or ependymal

cysts. Arachnoidal cysts most commonly arise inthe middle fossa (30%–50%), but 10% arise on thehemisphere convexity, 10% in the suprasellarcistern, 10% in the quadrigeminal cistern, 10%

in the cerebellopontine (CP) angle, and 10% in themidline of the posterior fossa [7].

Intraventricular cysts typically arise in or near

the atria of the lateral ventricles. Their walls can beformed by arachnoidal cells (intraventriculararachnoidal cysts), ependymal cells (ependymal

cysts more correctly viewed as being periventricu-lar), neuroepithelial cells, or choroidal cells. Mostof these cysts, excluding the choroidal cysts, pro-

bably represent a dysgenic process and are there-fore not uncommonly associated with dysfunctionof adjacent brain parenchyma as manifested byfocal seizures and cognitive disabilities [8].

Presentation

Intracranial CSF cysts typically become symp-tomatic in patients before the age of 20 years, withmost doing so within the first decade [9]. There is

some variation in age at presentation according tothe location of the cyst, however. Arachnoidalcysts of the middle fossa present in patients before

the age of 16 years [10]. Cysts of the quadrigemi-nal cistern present earlier, usually by the age of 12months because of their compression of theaqueduct of Sylvius [8], whereas arachnoidal cysts

of the suprasellar cistern present later in child-hood, with more than 60% presenting in patientsbetween 1 and 20 years of age [8]. Only 14% of

suprasellar cysts present after the age of 20 years.The signs and symptoms at presentation arereferable to the location and size of the cyst. With

the exception of cysts of the CP angle, the mostcommon symptoms present at presentation arethose of a slowly growing mass or of hydroceph-alus (ie, headache). The headache can be gener-

alized, or it may localize to the site of the cyst.

Other symptoms at presentation point to thelocation of the cyst. Cysts of the middle fossa canbe associated with partial or secondarily general-

ized seizures. Arai et al [11] found a 39% incidencein 77 patients having cysts of the middle fossa,whereas Ciricillo et al [12] found the same in-cidence in 39 patients. Boop and Teo [13] have

reported that more than 80% of their patients withsuch cysts manifest behavioral abnormalities,attention deficit disorder, or other problems in

school. Cysts of the quadrigeminal cistern can beassociated with nystagmus, Parinaud’s syndrome,hearing disturbances, and, rarely, motor deficits.

Such symptoms are rare because of the relativelyearly onset of hydrocephalus associated with suchlesions. Suprasellar cysts can be associated withvisual disturbances (papilledema, optic atrophy, or

bitemporal field cuts), endocrinopathies, and the‘‘bobble head doll’’ sign that is pathognomic forlesions in this location. Cysts of the CP angle

typically have a long history of slowly evolvingsymptoms referable to stretching of cranial nervesand distortion of the cerebellum. Vertigo, hearing

disorders (eg, tinnitus, hearing loss), hemifacialspasm, facial paresis, trigeminal neuralgia, de-creased corneal sensation, nystagmus, intention

tremor, ataxia, and dysmetria can be features ofthese cysts. If ignored too long, obstruction of theoutlets of the fourth ventricle can result in anobstructive hydrocephalus. Clival arachnoidal

cysts distort the brain stem, causing compressionof the corticospinal tracts (paresis in the extrem-ities, hyperreflexia, and Babinski sign) and stretch-

ing of cranial nerves (diplopia). When these extendinto the supratentorial space, endocrinopathiescan evolve because of compression of the pituitary

apparatus. Other than signs of increased intracra-nial pressure, intraventricular cysts can be associ-ated with focal seizures, ataxia and other gaitdisturbances, blurred vision, or frank diplopia.

Diagnosis

The imaging tool of choice with these cystic

CSF lesions is MRI (Fig. 1). Cysts of the middlefossa, clivus, and CP angle must be differentiatedfrom epidermoid cysts. Diffusion-weighted MRI

is useful in such cases, because the two types ofcysts have different signal characteristics in theimaging sequence. Suprasellar arachnoid cysts

need to be differentiated from craniopharyngio-mas and Rathke cleft cysts. The suprasellararachnoid cysts are typically larger; grow sym-

metrically and upward, resulting in the ‘‘Mickey


Fig. 1. (A) Coronal T1-weighted MRI scan of middle fossa arachnoid cyst. (B) Axial T2-weighted scan of suprasellar

arachnoid cyst’s base with flow void seen at point of fenestration (arrow). (C) T2-weighted sagittal MRI scan of

quadrigeminal arachnoidal cyst. (D) T1-weighted coronal MRI scan of cerebellopontine angle arachnoidal cyst.

Mouse’’ sign; and do not contain calcificationcommonly seen in craniopharyngiomas. In thissituation, CT may be warranted to excludea craniopharyngioma given this modality’s sensi-

tivity for calcium.

Intraventricular CSF cysts can be differentiat-ed from epidermoids, dermoids, and parasiticcysts using MRI. The differential diagnosis shouldbe discussed with your neuroradiologist when

ordering imaging of the lesion.


Treatment

The first question to be answered is that ofwhether any treatment should be offered. Implied

in this question iswhether further growthof the cystis expected and whether symptomatic hydroceph-alus is present. The latter question is easilyaddressed by careful history taking and examina-

tion of the patient. The former canbemore difficult.There are few reports available that offer rules fortreatment based on the observed natural history of

these cysts. One based on a retrospective study ofadults with arachnoidal cysts showed that cyststended to grow over time if there was distortion of

neural structures adjacent to the lobe containingthe cyst or to bony structures [14]. This does notoffer assistance in decision making for children,however, given its retrospective nature and the fact

that because the study was conducted in a group ofadults, it represented a more benign subgroup ofpatients with this condition. Most reports dealing

with the pediatric population recommend treat-ment at the time of discovery of the cyst unless it isof a small size withminimal distortion of surround-

ing tissues and has been discovered incidentally[13,15,16]. Cysts that distort surrounding neuraltissues have been shown to alter cerebral blood flow

[10,17]. Presumably, this explains the atrophy thatcanoccur over time, as shownwhen there is a failureof parenchymal re-expansion when cysts aretreated in older individuals.

Next to be answered when treatment has beenelected is how the cyst should be treated. In the1980s, most articles discussing treatment of these

cysts favored the use of shunts. Presumably, thiswas the result of a relatively high incidence ofpostoperative complications after craniotomies in

the 1960s and 1970s. As operative techniques andthe use of microneurosurgery have evolved, a re-duction in the incidence of postoperative compli-cations has occurred. With this and a greater

appreciation of the life history of a shunted patient,there has been a natural shift in preference fromtreating these cysts with shunts to surgical fenes-

tration [18]. Theoretically, to fenestrate a cyst is tocure it; thus, a lifelong dependence on a shunt isavoided. This shift in preference toward fenestra-

tion is only being accelerated with the introductionof the endoscope.

Endoscopic treatment of intracranial CSF cysts

is technically challenging and should only beconsidered by experienced neuroendoscopists.The anatomy can be obscure, and the numberand complexity of the instruments used are greater

than those used in the more standard neuro-endoscopic cases, such as third ventriculostomy.To start, careful planning is required. Thought

should be given to the trajectory taken to thetarget(s) for fenestration. This is especially criticalwhen multiple fenestrations are being considered,as is the case with a suprasellar cyst, or when a third

ventriculostomy is needed in addition to cystfenestration, as is case with a quadrigeminal cyst.An incorrectly placed burr hole because of a poorly

planned trajectory results in an injury to thecortical tissues surrounding the scope as it isrocked back and forth to reach fenestration

targets. When grossly off, reinsertion of the guidecannula to reach a second target may not even bepossible and a new burr hole might be required.Ideally, when the target abuts or is in a ventricle, it

is best to approach it via the ventricle so thatnormal anatomic structures can be used forguidance. It can be difficult to visualize structures

on the other side of a cyst’s wall, and it is a commonoccurrence to punch holes into brain tissue whenattempting to fenestrate a cyst to an adjacent CSF

space. Thought should also be given to structuresthat are to be avoided during the introduction andadvancement of the scope. Ideally, the trajectory

should be established to avoid these structures. Ifthis is impossible, thought should then be given asto how to avoid their injury, such as establishingearly visualization with the scope before encoun-

tering the structure so as to avoid injury (eg,visualization of the fornix at the foramen ofMonroso as to avoid it as one traverses the foramen).

Next to be considered is whether a self-retaining holder for the scope is to be used. Theuse of a self-retaining system adds time and

complexity to the case. In many cases, an assistantcan navigate the scope while the surgeon workswith instrumentation through the scope. In thissetting, the assistant should be experienced in the

use of the scope and familiar with the workinghabits of the surgeon so that the two work asa fluid team. The surgeon cannot do an adequate

job if he or she is worrying about or fighting withthe assistant. In such a setting, the surgeon mightcorrectly elect to use a self-retaining system. The

other setting where the use of such a systemshould seriously be considered is when oneanticipates being at the target site for more than

a few minutes because of the need for a delicateor extensive surgical dissection. Also, if one isconsidering the use of a second instrumentchannel, it is best to have the scope held by

a self-retaining system.


With regard to the use of a second instrumentchannel, this refers to the introduction of a secondcatheter that approaches the target via a differenttrajectory than that of the endoscope [19]. It can be

useful when one anticipates the need for instru-ments that are larger than the working channel ofthe scope or when there is a need to view the

working of the instruments from an angle. This canbe useful when one anticipates pulling material outof the surgical field, because such a maneuver not

uncommonly blinds the endoscopist as the mate-rial reaches the head of the scope. The surgeon alsogains a greater appreciation of the amount of

distortion that occurs during the maneuvering ofthe instrumentation. Use of the second channeldoes add another level of complexity to the case,however. First, there is the issue of capturing and

maintaining visualization of the channel. Thisbecomes increasingly difficult as the distancebetween the burr holes of the scope and instrument

channel increases and as the depth of the targetincreases. Second, there can be difficulty in co-ordinating the manipulation of the two channels.

This takes some practice, and my advice is to makesmall movements that do not result in the completeloss of visualization of the instrument channel. The

instrument channel is first moved partially out ofthe field of view in a desired direction, and thescope is then moved to recenter the instrumentchannel in its field of view. Computer-assisted

guidance systems can be of great use in overcomingthese problems, with trackers being placed on thescope as well as on the instrument channel.

At every step in the surgery, absolute attentionmust be paid toward homeostasis. It has beenestablished that less than a teaspoon of blood

within the intraventricular fluid space blinds theendoscopist. Consequently, after the burr hole hasbeen drilled and before the dura is opened,homeostasis must be complete. This is also the

case after the dura is opened before penetrating thearachnoid. Midline cysts, such as suprasellar andquadrigeminal cysts, are typically approached via

the ventricles. A guide sleeve is typically introducedinto the lateral ventricle, its stylet is removed, andthe endoscope is then advanced into the lateral

ventricle. During this approach, it is wise to remainoriented to the trajectory of the scope so as toappreciate the general location of the scope’s tip.

When the anatomy is disorienting, the first stepshould be to check that one’s conceptualization ofthe scope’s trajectory is correct. More than once, Ihave mistakenly thought that I was in the region of

the foramen of Monro only to discover that the

scope was pointing posterior with the tip in theventricle’s atrium. This is especially common witha burr hole placed too anteriorly. Normal ana-tomic structures are searched for to further

establish orientation. Once the choroid plexus isfound, it can be followed anteriorly to find theforamen of Monro or posteriorly to point to where

a quadrigeminal cyst is herniating through thefloor of the lateral ventricle. If the scope feelsawkward in its movement, the most common

reason is that the camera is not oriented to thescope (ie, the camera’s 12-o’clock position is notthe scope’s 12-o’clock position). The scope should

be withdrawn from the head and its camera’sorientation checked.

Laterally projecting cysts, such as those of themiddle fossa and CP angle, are approached

through the overlying parenchyma laterally. Thisparenchyma can be present to a varying degree.Draped over the surface of the cysts are cortical

veins, especially in the case of cysts of the middlefossa. Schroeder et al [20] have suggested notdisturbing the lateral wall of such cysts so as to

avoid traction of cortical veins stretched over theirsurface. Once within these cysts, arterial vessels ofthe Sylvian fissure can be seen coursing the medial

wall. They can be followed down to the basilarcisterns, where the fenestrations can be made. Themembrane of these cysts has been shown to be richin collagen; consequently, it can be difficult to

penetrate. Scissors or a knife can be used to cut anopening. It can then be enlarged with sharpdissection or by repeated inflation of a balloon. It

has not been established what constitutes anadequate opening. Until such time as this has beenestablished, one should attempt to duplicate what

one would accomplish using microneurosurgeryvia a small craniotomy. If difficulty is encounteredin doing this, judgment is used, but there shouldnot be any hesitancy to convert to a craniotomy to

accomplish one’s surgical goals. On more than oneoccasion, we have done this on our service; thefamilies of our patients are always told that this is

a possibility, and consent is obtained for just sucha possibility.

Earlier mention was made of the use of

computer-assisted surgical guidance systems orframeless stereotaxis. These systems can be of greatuse when working within cysts containing no

anatomic landmarks for guidance. By using suchsystems, one moves from navigating by deadreckoning based on surface landmarks to image-based navigation, where the tip of scope and or its

channel or a second instrument channel can be seen


onMRIorCT.Before performing the surgery, suchsystems can be used for surgical planning. Struc-tures to be avoided can be incorporated in planning

the trajectory, and this information is, in turn, usedto establish the optimum location for the burr hole.Depending on the guidance system being used,there are several options that can be employed for

image guidance. Some systems require the use ofdedicated pointing devices. In such settings, a spe-cial pointer can be manufactured to replace either

the endoscope’s guidance sleeve or its stylet.Another alternative is to have the companymodifya rigid endoscope so that its system can track it

directly. I have preferred tracking the stylet ofa peel-away catheter used to establish a channel forthe endoscope because it affords the greatestflexibility for scope selection and is the simplest

and least expensive to manufacture. At least onesystem has removable tracking devices that can beattached to an instrument or catheter and then

registered to the system to allow for tracking by thesystem. Some accuracy is sacrificed when usingsuch a device, and it can take several minutes to

establish registration. Also, more than once, I havebeen unable to accomplish registration when at-tempting to use these attachable trackers.

After registering the patient and pointingdevices to the system, the burr hole is locatedusing the system and the adequacy of the trajectoryis confirmed before making a skin incision. It is

wise to take several moments to confirm theappropriateness of instrument setup at this pointso as to prevent discovering that movement of the

scope is limited by something in the surgical field ortracking is lost when the scope is in a certainposition within the head. Once this is done, the

burr hole is made and the dura opened using thesame care as mentioned previously. The appropri-ateness of the planned trajectory to the target isconfirmed one final time using dead reckoning; the

scope’s channel is then advanced toward the target.Advancement toward the target is done slowlywhile retaining the appropriate trajectory by main-

taining it on at least two image planes (Fig. 2).Some systems give a bird’s eye view or so-called‘‘target view’’ showing the tip being tracked as well

Fig. 2. Image showing approach to target (marked with +) by outer sleeve of endoscopy (its tip is the point of the cross

hairs). For this cyst, we were targeting the point where the cyst abutted the ambient cistern.


as the target; this is quite helpful, because the goalis simply to overlay the two cross hairs. Side to sideand upward or downward corrections in route tothe target cause distortion in the brain’s parenchy-

ma, because the catheter cannot slice through braintissue. This is critical to appreciate when a flexiblecatheter for the scope is being used, because once

the stylet is removed to allow for the introductionof the scope, the catheter deflects as the parenchy-ma restores itself to its resting configuration. The

result is that the catheter tip will no longer be ontarget. Consequently, the surgeonmust judge whenthere has been too great a need for correction of

trajectory. When there is concern about this, thecatheter should be removed and a second attemptmade to advance it to the target. Much has beenmade of the potential for brain shift imparting

inaccuracy to the guidance data set as the surgeryproceeds. Obviously, if one is to be working at onlyone target, this should not be of great concern if

things move along at a pace such that little CSFescapes before the catheter is on target. At thepoint that the scope is looking at the target, the

accuracy of the guidance data set is no longer ofconcern. Alternatively, if there is more than onetarget, thought should be given to the problem. In

my experience, structures that are close to thecenter of the head and vertically inferior to mypointer seem to maintain accurate registration tothe data set, whereas those near the surface or

lateral to a deflating structure or lesion tend to loseregistration as the case proceeds. This should bekept in mind when determining the order of

targeting. It is also important to remember thatthere is always a tomorrow and to discuss witha family member the possibility of needingmultiple

sessions to accomplish the surgical goal.There will be cases in which shunting of a cyst is

required. We had one case of a quadrigeminal cystthat underwent three endoscopic fenestrations into

the ventricles (twice into the lateral ventricle andonce into the third) and two open fenestrations,only to fail all procedures and become symptom-

atic because of an increasing mass effect from thecyst (Fig. 3). It was then decided to inserta cystoperitoneal shunt. The placement of a cath-

eter into an arachnoid cyst can be extremelydifficult, because the cyst wall is extremely proneto deflecting the catheter to the side when insertion

is attempted. Use of the endoscope to confirmentry into the cyst can avoid postoperativefrustration. One can use a larger scope to cuta fenestration in the cyst wall. The scope and its

guide sleeve are then advanced into the cyst, the

scope is removed, and a proximal catheter of

a cystoperitoneal system is then fed down the guidesleeve into the cyst. It is wise to determine the depthof scope insertion to establish the length of catheter

needed before removal of the scope in such a settingand to use a peel-away guide sleeve to simplifycatheter placement. Alternatively, a small intra-

luminal scope can be used as the catheter’s stintduring attempted placement of the proximalcatheter. Once within the cyst, the scope can beused to confirm cyst penetration. If the wall has not

been penetrated, unipolar cautery can be applied tothe scope to complete penetration unless there isconcern about adjacent vessels, as might be the

case for a quadrigeminal cyst. Before doing this, itis wise to back the scope out partially to visualizethe wall better and confirm that it has indeed not

been penetrated. On occasion, I have been fooledinto thinking that I had not penetrated the cystwhen, in fact, I had, and I was simply up against thefar wall of the cyst. In such a case, one can

immediately see the interior of the cyst as opposedto a trailing cyst wall.

Results

There are now a few papers describing out-

comes in small series of patients who haveundergone endoscopic management of their in-tracranial CSF cysts. Paladino et al [21] described6 of their patients treated endoscopically. One had

Fig. 3. Second recurrence of quadrigeminal cyst after

two prior fenestrations into lateral and third ventricles.

At the third operation, the cyst was fenestrated and

a catheter placed for a cystoperitoneal shunt.


a parasagittal cyst, and the other 5 had middlefossa cysts. Four of 6 cysts were successfullymanaged endoscopically, with 2 others requiring

insertion of a cystoperitoneal shunt. Kim [22]reported on 7 patients with arachnoid cysts. Threehad cysts in the posterior fossa, 2 had cysts in thesuprasellar cistern, 1 had a cyst in the middle

fossa, and 1 had a cyst on the convexity, withall experiencing symptomatic resolution. Radio-graphically, 4 of 7 cysts diminished in size,

whereas 3 of 7 disappeared. Ruge et al [23]reported on successfully managing 2 children withquadrigeminal plate cysts endoscopically, and

Furuta et al [24] reported a similar case managedsuccessfully using an endoscope to place a cysto-peritoneal shunt after a failed attempt to place thecatheter stereotactically. Brunori et al [25] and

Hayashi et al [26] have reported experiencessimilar to those of Ruge et al [23] in successfullycommunicating quadrigeminal cysts to the ven-

tricles. Hopf and Perneckzy [27] reported on 36patients with various forms of arachnoid cyststreated endoscopically, with 26 improving symp-

tomatically. Wagner et al [28] reported on 2children with arachnoid cysts managed success-fully endoscopically, and Decq et al [29] reported

on 2 children with suprasellar arachnoid cystssuccessfully managed. Walker et al [30] reporteda 64% success rate in managing 14 children witharachnoidal cysts endoscopically. In summary, as

series numbers increased, so did failures, withlarger series reporting around a 33% failure ratein endoscopically managing arachnoid cysts. This

compares with a 30% failure rate for shuntedcysts and a 19% failure rate in cyst fenestration inthe European Cooperative Study [9]. The more

aggressive cyst resection surgery had a 7% failurerate in the same study. Fewel et al [18] reporteda 27% failure rate in 102 arachnoidal cystsmanaged with either surgical fenestration or

resection. Undoubtedly, there is a learning curveassociated with this technique, and the failure ratecurrently being seen in the endoscopically man-

aged cases will lessen over time. Until such time asthe failure rate duplicates that seen with opensurgery, thoughtful consideration should be given

to the adequacy of the fenestration accomplishedendoscopically and to whether or not the pro-cedure should be converted to an open operation

if there is concern about its adequacy. Addition-ally, a frank discussion should be undertaken withthe family as to the pros and cons of the variousapproaches so that they can make an informed

decision.

Complications

As with any surgery, there are potentialcomplications when endoscopically managing

arachnoid cysts. ‘‘Minimally invasive’’ should notbe construed to mean minimal risk by the surgeon,and the family or patient should in no way be solda bill of goods that this technique has fewer risks

than open surgery. Kim [22] reported that 1 of his 7patients treated endoscopically experienced signif-icant bleeding during her surgery, requiring aban-

donment of the endoscopic procedure withconversion to an open procedure with successfulcontrol of the hemorrhage and completion of the

surgical goal of fenestration. Hopf and Perneczky[27] reported a 14%complication rate in imaging 36patients with arachnoidal cysts. Four patientsexperienced subdural hematomas or hygromas

after their surgery, with 2 of them also developingmeningitis. This may be an important observation,because we have noted infections in some of our

patients who have experienced intraoperativehemorrhaging, which required extraventriculardrainage after surgery. This is a discussion we have

with all our families before surgery. Another of thepatients reported on by Hopf and Perneczky [27]experienced hemorrhaging after treatment of a pos-

terior fossa cyst, resulting in hydrocephalus thatrequired a subsequent third ventriculostomy. Rob-inson andCohenwarn of the risks of injuring bloodvessels and other structures by the guide sleeve

when advancing the scope because of the sleeve notbeing visualized by the scope [8]. Finally, structurescan be injured during the actual fenestration

process when visualization is poor or excessiveforce is used to penetrate the tough membrane.

Summary

I have little doubt that most arachnoidal cystswill be managed endoscopically in the future given

the advances we have seen over the last decade inour instrumentation. Our excitement to employthis new technology should be governed by the

reality that we are still learning and that ourcurrent success rate is not quite as good as whatcan be expected when using microneurosurgery.

References

[1] Shaw C, Alvord EJ. Congenital arachnoid cysts and

their differential diagnosis. In: Vinken P, Bruyn G,

editors. Handbook of clinical neurology, vol. 31.

Congenital malformations of the brain. Amster-

dam: North Holland; 1977. p. 75–136.


[2] Robinson R. Congenital cysts of the brain: arach-

noidal malformations. Prog Neurol Surg 1971;4:

133–74.

[3] Bright R. Reports of medical cases, selected with

a view of illustrating the symptoms and cure of

diseases by a reference to morbid anatomy, vol. 2.

Diseases of the brain and nervous system. London:

Lonham, Rees, Orme, Brown, Green, and Highley;

1831.

[4] Rengachary S, Watanabe I, Brackett C. Pathogen-

esis of intracranial arachnoid cysts. Surg Neurol

1978;9:139–44.

[5] Robinson R. The temporal lobe agenesis syndrome.

Brain 1964;87:87–106.

[6] Starkman S, Brown T, Linell E. Cerebral arachnoid

cysts. J Neuropathol Exp Neurol 1958;17:484–500.

[7] Hogg J, Peterson A, El-Kadi H. Imaging of cranial

and spinal cerebrospinal fluid collections. In:

Kaufmann H, editor. Cerebrospinal fluid collec-

tions. Park Ridge, IL: American Association of

Neurological Surgeons; 1998. p. 19–57.

[8] Kaufmann H, editor. Cerebrospinal fluid collec-

tions. Park Ridge, IL: American Association of

Neurological Surgeons; 1998.

[9] Oberbauer R, Haase J, Pucher R. Arachnoid cysts

in children: a European cooperative study. Childs

Nerv Syst 1992;8:281–6.

[10] Boop F, Young R, Scott R. Arachnoid cysts of the

middle cranial fossa and convexity. In: Kaufmann

H, editor. Cerebrospinal fluid collections. Park

Ridge, IL: American Association of Neurological

Surgeons; 1998. p. 67–96.

[11] Arai H, Sato K, Wachi A, Okuda O, Takeda N.

Arachnoid cysts of the middle cranial fossa:

experience with 77 patients who were treated with

cystoperitoneal shunting. Neurosurgery 1996;39(6):

1108–12.

[12] Ciricillo S, Cogen PH, Harsh GR, Edwards MS.

Intracranial arachnoid cysts in children. A compar-

ison of the effects of fenestration and shunting.

J Neurosurg 1991;74:230–5.

[13] Boop F, Teo C. Congenital cysts. In: McLone D,

editor. Pediatric neurosurgery. Philadelphia: WB

Saunders; 2001. p. 489–98.

[14] Becker T, Wagner M, Hofmann E, Warmuth-Metz

M, Nadjmi M. Do arachnoid cysts grow? A

retrospective CT volumetric study. Neuroradiology

1991;33:341–5.

[15] Artico M, Cervoni L, Salvati M, Fiorenza F,

Caruso R. Supratentorial arachnoid cysts: clinical

and therapeutic remarks on 46 cases. Acta Neuro-

chir (Wien) 1995;132:75–8.

[16] Galassi E, Gaist G, Giuliani G, Pozzati E.

Arachnoid cysts of the middle cranial fossa:

experience with 77 cases treated surgically. Acta


[17] De Volder A, Michel C, Thauvoy C, Willems G,

Ferriere G. Brain glucose utilisation in acquired

childhood aphasia associated with a Sylvian arach-

noid cyst: recovery after shunting as demonstrated

by PET. J Neurol Neurosurg Psychiatry 1994;57:

296–300.

[18] Fewel M, Levy M, McComb J. Surgical treatment

of 95 children with 102 intracranial arachnoid cysts.

Pediatr Neurosurg 1996;25:165–73.

[19] Jallo G, Morota N, Abbott R. Introduction

of a second working portal for neuroendoscopy.

A technical note. Pediatr Neurosurg 1996;24:56–60.

[20] Schroeder H, Gaab M, Niendorf W. Neuroendo-

scopic approach to arachnoid cysts. J Neurosurg

1996;85:293–8.

[21] Paladino J, Rotim K, Heinrich Z. Neuroendoscopic

fenestration of arachnoid cysts. Minim Invasive

Neurosurg 1998;41:137–40.

[22] Kim M. The role of endoscopic fenestration

procedures for cerebral arachnoid cysts. J Korean

Med Sci 1999;14:443–7.

[23] Ruge J, Johnson R, Bauer J. Burr hole neuro-

endoscopic fenestration of quadrigeminal cistern

arachnoid cyst: technical case report. Neurosurgery

1996;38:830–7.

[24] Furuta S, Hatakeyama T, Nishizaki O, Fukumoto

S. Usefulness of neuroendoscopy in treating supra-

collicular arachnoid cysts—case report. Neurol

Med Chir (Tokyo) 1998;38:107–9.

[25] Brunori AA, Chiappetta F. Endoscopy for cysts.

J Neurosurg 1999;91:1067–8.

[26] Hayashi N, Endo S, Tsukamoto E, Hohnoki S,

Masuoka T, Takaku A. Endoscopic ventriculocys-

tocisternostomy of a quadrigeminal cistern arach-

noid cyst. Case report. J Neurosurg 1999;90:1125–8.

[27] Hopf N, Perneczky A. Endoscopic neurosurgery

and endoscope-assisted microneurosurgery for the

treatment of intracranial cysts. Neurosurgery 1998;

43:1330–7.

[28] Wagner HJ, Seidel A, Reusche E, Sepehrnia A,

Kruse K, Sperner J. A craniospinal enterogenous

cyst: case report. Neuropediatrics 1998;29:212–4.

[29] Decq P, Brugieres P, Le Guerinel C, Djindjian M,

Keravel Y, Nguyen JP. Percutaneous endoscopic

treatment of suprasellar arachnoid cysts: ventric-

ulocystostomy or ventriculocystocisternostomy?

Technical note. J Neurosurg 1996;84:696–701.

[30] Walker M, Petronio J, Carey C. Ventriculostomy.

In: Cheek W, Marlin A, McLone D, editors.

Pediatric neurosurgery: surgery of the developing

nervous system. Philadelphia: WB Saunders; 1994.

p. 572–81.


Basic principles and equipment in neuroendoscopyVit Siomin, MDa, Shlomi Constantini, MD, MScb,*

aDepartment of Neurosurgery, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USAbDivision of Pediatric Neurosurgery, Dana Children’s Hospital, Tel Aviv Medical Center, 6 Weizman Street,

Tel Aviv 64239, Israel

A fool with a tool is. . . still a fool. Lars Leksell

The advent of neuroendoscopy has had a re-markable impact on the field of neurosurgery.Although the use of an endoscope to diagnose and

treat various central nervous system (CNS)conditions, particularly those confined to theventricular system, has been well recognized for

years [1–7], the story of modern neuroendoscopyis just beginning. It has been attended by a numberof remarkable breakthroughs in optical physics,

technology, and instrumentation. Needless to say,understanding the basic physics and instrumenta-tion underlying today’s endoscopes is essentialfor safe and successful work with these delicate

instruments. This article presents a basic overviewof the technology that literally brought light intothe depths of contemporary neurosurgery.

Today’s market is saturated with neurosurgicalendoscopes, with many companies producingsimilar pieces of equipment. Any article attempt-

ing to comprehensively list the advantages anddisadvantages of each endoscope on the markettoday will probably be somewhat outdated bypublication because of the rapid progress in this

field. This article therefore does not even attemptto provide a full list of all commercially availableendoscopes with their descriptions. The authors

simply intend to provide readers with a generaloverview of the selected endoscopes, includinga brief explanation of their respective advantages

and disadvantages.

* Corresponding author.


(S. Constantini).

1042-3680/04/$ - see front matter � 2004 Elsevier Inc. All rig

doi:10.1016/S1042-3680(03)00075-5

A brief history of optics

Roots of the modern rigid endoscope

Until the sixties, the optic chains in a lens

system were constructed of small glass lensesinterspersed with large air spaces. A Britishphysicist named Hopkins realized that the totalamount of light transmitted through an endoscope

is proportional to the refraction index of thematerial used and that using more glass than airwould increase the amount of light by a factor of

about 2. He restructured the lens system to consistof long glass lenses interspersed with small lens-likeair spaces [3]. These ‘‘rigid rod lens scopes,’’ with

their increased light transmission ability, form thebasis of most modern endoscopic systems (Fig. 1).

Further reduction in light loss is achieved

through a special coating of the glass surfaces tominimize light reflection. Uncoated glass surfacesusually lose about 5% of their light throughreflection. Because endoscopes consist of multiple

glass lenses, the cumulative light loss can becomesignificant. The solution is to coat the glass surfacewith an ultrathin layer of magnesium fluoride. This

layer markedly decreases the reflection and im-proves the optic characteristics of endoscopes andcameras [3,8,9].

Understanding fiberoptics

Before the sixties, endoscopists used miniaturetungsten light bulbs inside the endoscopes. Thesebulbs had two disadvantages, however: the heat

generated by the bulbs could easily burn tissues,and the bulbs could not emit blue light waves, soa red color would dominate the working field.

Fortunately, fiberoptic cables were invented in the

hts reserved.


20 V. Siomin, S. Constantini /Neurosurg Clin N Am 15 (2004) 19–31

Fig 1. Conventional and rod lens systems.

sixties, a major breakthrough for endoscopic

technology. With their development, the lightsource could be separated from the endoscopeand its emitted lighted transmitted via fiberoptic

cables to the tip of the endoscope. This avoided theproblem of tissue injury from heat and alsodelivered a more natural light. Specialized cablescould also be used to conduct images to a camera.

Fiberoptic cables are made of individually coatedflexible fibers with an inner core of silica glass.Fiberoptic bundles can be arranged in twoways [3]:

� Coherently—the proximal end portrays the

image exactly as it is perceived distally. Acoherent arrangement is used for visualiza-tion during surgery.

� Incoherently—fiber bundles transmit lightonly. An incoherent arrangement is used totransmit light to the surgical site.

Most modern fiberoptic endoscopes containa minimum of 10,000 coherent fibers. Some

contain as many as 30,000 fibers. A greaternumber of fibers is an advantage, because the re-solution of fiberoptic endoscopes is proportional

to the number of fibers in the endoscope. Morefibers provide an image with a much sharperresolution.

Although a great number of fibers in bundlesused to transmit images is an obvious advantage,it is not so gainful when it comes to transmissionof light from the source box to the scope tip.

Having many individual fibers bundled togetherin light source cables increases the total surfacearea and leads to a loss of 30% to 40% of the light

being transmitted through all those individualfibers. Thus, contrary to popular belief, the lightthat emerges from the proximal tip is somewhat

weaker than the light that enters the endoscope atthe original source [3]. Light loss is discussed ingreater detail later in this article in the section on

light sources.

Working with optic angles

The tip of the scope can be bent to different

angles, bringing different aspects of the surgicalfield into view. These angles are obviously fixed inthe rigid solid lens scopes as opposed to the flexible

fiberoptic scopes, where the tip can be manuallydeflected to change the angle of view off the scope’slong axis. For the rigid solid lens scopes, the mostcommonly used angles are: 0�(‘‘head-on’’), 30�, 70�,and 120�. The 0� deflection scopes portray onlywhat they are viewing head-on, minimizing the riskof disorientation. The angled scopes are more

versatile in that they visualize areas that wouldotherwise be out of view or difficult to illuminate.Themajor disadvantage of angled scopes is that the

indirect image may cause the surgeon to becomedisoriented. The authors encountered orientationdifficulties mostly when operating on patients with

congenital lesions, such as arachnoid and porence-phalic cysts. Similar problems may also be encoun-tered in endoscopically assisted transsphenoidalsurgery and intracranial tumor removal. Steer-

ability of fiberoptic endoscopes adds anotherelement to the problem of disorientation. Extensiveexperience is necessary to get used to oblique and

‘‘around-the-corner’’ views.

Light sources

An endoscope’s light source is extremely

important, often becoming a limiting or facilitatingfactor. Currently, most neurosurgeons use high-intensity xenon light sources. The light is trans-

mitted through incoherent fiber bundles to thesurgical field. Setting the light source to between300 and 500 W provides a superior picture quality.Other types of light sources, such as halogen, are

not used in modern endoscopy because they do notgenerate a bright enough light.

21V. Siomin, S. Constantini /Neurosurg Clin N Am 15 (2004) 19–31

Light loss in a fiberoptic system is a significantissue. As a rule, 30% of light is lost at the lamp,because up to 30% of the surface area consists ofcladding and filler, which absorb and reflect light.

Fiber mismatch (between the incoherent cablebundles and the coherent optic fiber bundles)accounts for another 20% loss. Significant loss

also occurs because of mismatch between thewider transmitting cable and the thinner scopesused in neurosurgery. Additional light is also lost

from inadvertent reflection in various surfaces.Thus, at best, only about 30% of the lightgenerated within the light source reaches the

distal tip of the scope [3,8]. This problem is stillawaiting a technologic solution.

Cameras

The chip camera, or charged-coupled device

(CCD), is a critically important part of anyendoscopic system. Invention of the chip camerawas a remarkable technologic achievement, allow-

ing a significant decrease in the size of theendoscope combined with an improvement inthe quality of the transmitted image in compar-ison with previously used tube cameras. Two

kinds of CCDs are currently used in neuro-endoscopy: the single-chip camera and the three-chip camera.

The principle used in chip camera technologyis the same as in the microprocessor industry. Thechips consist of horizontal and vertical photosen-

sitive elements that are arranged in intersectinglines. The points of intersection correspond topixels, or picture units, that appear on the display

screen. Light reflected by the object being viewedhits each pixel, and a current is generated. Eachpixel on the display appears either brighter ordimmer depending on the voltage of the current

signal sent to that pixel. The voltage is a functionof the brightness of the image being portrayed:a brighter point on the original object being

viewed generates a higher voltage current, trig-gering a brighter image on the display screen. Theexact current voltage sent to each point is not

just an approximation or estimate; it is actuallya precise numeric representation of the energygenerated when light hits the horizontal and

vertical matrix of the camera. A digital imageincorporating a complete set of picture units(pixels) is stored in the camera’s memory [8].

Most endoscopic systems now use single-chip

cameras. A good resolution for neuroendoscopy isavailable with 0.5-in cameras. When the resolu-

tion of the one-chip camera is not enough, theimage may be computer enhanced.

Three-chip cameras provide better picturequality. The authors use theDavid-3-Chip-Camera

from Aesculap (Center Valley, Pennsylvania),which features a resolution of more than 800horizontal lines, compared with 500 lines in the

standard 0.5-in micro-lens-on-chip technologycamera (Fig. 2). The improved image of thethree-chip camera comes at a price, however—a

somewhat larger camera size and higher cost.

Video monitors

The monitor is an integral part of everyendoscopic system (Fig. 3). Three things shouldbe considered in selecting a monitor: resolution,

screen size, and cost.A higher resolution monitor provides better

picture quality. There is no advantage to a monitor

resolution that is significantly better than thecamera resolution, however, because the monitoronly displays the image seen and processed by the

camera. The monitor cannot improve the camera’simage. By the same token, a top-quality imagefrom the highest resolution camera must still bedisplayed on the monitor, so a lower resolution

monitor ruins even the best camera image.When choosing the screen size, remember that

the camera output is displayed over an area

significantly larger than the cross section of theoptic cable. As an image is enlarged, the imageresolution and quality decrease. Screens larger

than 13 in display poorer quality images becauseof reduced resolution. This becomes even a greaterproblem with images generated with a fiberopticscope. The pixels that are only slightly visible

Fig 2. 3-chip CCD camera (Aesculap, Center Valley,

Pennsylvania).


when fiber scopes with reasonable resolution are

transmitted to a screen smaller than 13 in undergomagnification with consequent enlargement of pix-els when projected to larger screens. Conversely,

larger screens are useful for displaying multipleimages. Finally, note that it is often most cost-effective to purchase a monitor from the manu-

facturer that supplies the rest of the system.

Endoscopic components and tools

Peel-away catheter

Because the concept of minimally invasivesurgery lies at the heart of neuroendoscopy, everystage of the procedure, from cannulation of the

fluid space to exit, should minimize interferencewith the patient’s anatomic structures. During anendoscopic procedure, the surgeon may need toinsert and remove the scope many times. It is

therefore logical to create a single ‘‘safe passage-way’’ that allows multiple reinsertions of the scopealong the same tract without jeopardizing sur-

rounding brain.Most surgeons today use disposable peel-away

catheter introducers (eg, those manufactured by

Fig 3. 100Hz monitor (Sony, New York, New York)

and 3-chip CCD camera (Aesculap, Center Valley,

Pennsylvania).

Cook [Cook Critical Care, Bloomington, IN] orNeuroview [Integra NeuroSciences, Plainsboro,NJ]). These may be attached to the scalp or drapes

with Steri-strips (3M Health Care, St. Paul, MN)or staples. Both tapered and blunt-tip cathetersare available (Fig. 4). The catheter is supplied in

various diameters ranging from number 3 tonumber 49 French. The most popular sizes inneuroendoscopy fall between number 10 and

number 20 French. Number 10 French cathetersare appropriate for 2- to 4-mm diameter endo-scopes, whereas number 20 French catheters

are good for the larger 6- to 8-mm endoscopes.Larger diameter introducers are not used inneuroendoscopy.

The disadvantage of using a peel-away in-

troducer is that it can cause bleeding because ofinjury to the choroid plexus or intraventricularveins on insertion. Unfortunately, only Neuro-

care’s (Integra NeuroSciences, Plainsboro, NJ)number 9 French catheter has centimeter mark-ings. Therefore, the authors have found it helpful

to apply some bone wax 5 cm from the tip of theintroducer to prevent unnecessarily deep insertionof the catheter and damage to intraventricularstructures.

Alternatively, one can use an obturator/oper-ating sheath. This is a reusable metal pipe that canonly be attached with a scope holder, which can

be a slight inconvenience. The introducers and thesheath enable multiple reinsertions of an endo-scope without repeated recannulations.

Rigid endoscopes

Most neurosurgeons today still use rigid glassendoscopes that are neither flexible nor steerable.An advantage of rigid rod lens scopes is that

Fig 4. 15F ‘‘peel-away’’ blunt and tapered tip introducers.


smaller diameter lenses may be used, allowing theendoscope designer to either decrease the diameterof the whole scope, producing a smaller andmore delicate instrument that can reach more

inaccessible areas, or add more space to theworking channel, producing an instrument thatcan accomplish many tasks for the surgeon. The

basic rigid glass endoscope consists of three ele-ments: an optic system, a working channel, and anirrigation port (Fig. 5). Table 1 summarizes the

features of the most common rigid endoscopes.The remainder of this section discusses theadvantages and disadvantages of typical endo-

scopes available from various manufacturers.

� The Gaab Neuro-Endoscope (K. Storz,Tuttlingen, Germany) has a 6.5-mm operatingsheath, with a 3-mm working channel. Fiber-optic light transmission is incorporated. The

Gaab endoscope with a 2-mm operativeHopkins telescope is a popular choice forsurgery because it allows visualization during

surgical instrumentation with specially de-signed rigid instruments. The 4-mm 0�, 30�,70�, and 120� diagnostic scopes are popular

for diagnostic orientation in the ventricularsystem, basal cisterns, arachnoid cysts, orcystic tumors. Because these diagnosticscopes do not have a working channel, they

have room for larger diameter lenses and aretherefore able to provide exceptionally clearvisualization and orientation. Note that

Codman & Shurtleff (Johnson & Johnson,New Brunswick, NJ) also manufactures theGaab Neuro-Endoscope. The outer diameter

of the Codman & Shurtleff endoscope is only

Fig 5. 3.2 mm rigid endoscope (K. Storz, Tuttlingen,

Germany).

5.8 mm, however, and the diameter of theworking channel is 1.6 mm.

� The Chavantes–Zamorano Neuro-Endoscope(K. Storz) has a relatively large outer di-

ameter of 8 mm. It has one central 3-mmworking channel and two separate suctionand irrigation channels, which can also be

used for flexible instruments. Better visuali-zation can be achieved with the smallerdiameter 30� and 70� diagnostic scopes. This

endoscope also has a manometer for constantmonitoring of intracranial pressure duringsurgery. The disadvantage of the Chavantes-

Zamorano system is that the scope hasa relatively large diameter and the overallsize impedes maneuverability. Nevertheless,many neurosurgeons use this endoscope,

especially for evacuation of intracerebraland, particularly, intraventricular hemato-mas, in conjunction with stereotaxy. Note

that Codman & Shurtleff also manufacturesthe Chavantes-Zamorano endoscope. Thedifference is that the diameter of that endo-

scope’s working channel is 1.9 mm.� The Auer Neuro-Endoscope (K. Storz) hasan outer diameter of 6.6 mm. This endo-

scope features two working channels: astraightforward 0� telescope that is 2.9 mmin diameter and an adapter for a Greenbergretractor. The larger size of the Auer en-

doscope is a major disadvantage; however,the excellent optics and specially positionedirrigation and suction apertures at the tip of

the endoscope facilitate a clear view of thefield.

� The Decq Neuro-Endoscope (K. Storz) comes

in twomodels. The 3.5-mm� 5.2-mmmodel issuitable for rigid or flexible instruments of1.7 mm in diameter. The 4.0-mm� 7.0-mmmodel is suitable for instruments of up to

3 mm in diameter. The telescope uses theHopkins lens system with a 30� angle ofvision.

� Aesculap produces an endoscope with a6.2-mm operating cannula (Fig. 6). Thismodel includes a 2.2-mm working channel,

one diagnostic endoscope channel, one irriga-tion channel, and one overflow channel.This endoscope is available in two lengths:

250 mm, which is compatible with stereo-tactic frames, and 160 mm. The major ad-vantage of this scope is excellent imagequality; however, the large size is a limiting

factor.


Table 1

Basic features of rigid endoscopes

Endoscope

Outer

diameter Channels

Diameter of

working

channels

Angles of

diagnostic

scopes Remark/clinical use

Neuroview 700R

(NeuroNavigational)

5.6 mm 1 working

1 irrigation

1 overflow

1 scope

2.0 mm

2.8 mm

0�, 30�, 70� Glass rod optics,

Good image,

Quite thick and

heavy

Aesculap 6.2 mm 1 working

1 irrigation

1 overflow

1 scope

2.2 mm

2.7 mm

0�, 30� Glass rod optics,

Good image,

Quite thick and

heavy, Good for

stereotaxy

MINOP

(Aesculap)

3.2 mm

4.6 mm

6.0

1 working/scope

1 working/scope

1 irrigation

1 overflow

1 working

2.8 mm

2.8 mm

2.2 mm

0�, 30� Glass rod optics,

Good image,

Short,

maneuverable

1 irrigation

1 overflow

1 scope 2.8 mm

Gaab endoscope

(Storz/Codman & Shurtleff)

6.5 mm

(5.8 mm)

1 working

1 irrigation

1 overflow

1 scope

3.0 mm

(4.0 mm)

0�, 30�, 70�,

120�Glass rod optics,

Good image,

Quite thick and

heavy

Chavantes-Zamorano

(Storz/Codman)

8.0 mm 1 working/scope

1 irrigation

1 overflow

3.0 mm

(1.9 mm)

0�, 30�, 70� Glass rod optics,

Good image,

Quite thick and

heavy, Features

manometer for

ice intracranial

measurement,

Good for

evacuation of

hematomas

Auer endoscope

(Storz)

6.6 mm 2 working

1 irrigation

1 overflow

1 scope 3.0 mm

0� Glass rod optics,

Good image,

Quite thick and

heavy, Adapter

for Greenberg

retractor

Decq endoscope

(Storz)

Oval

3.5� 5.2 mm

1/2 working

1 irrigation

1 overflow

1 scope

1.7 mm

3.0

30� Glass rod optics,

Good image

Oval

4.0� 7.0 mm

1/2 working

1 irrigation

1 overflow

1 scope

Glass rod optics,

Good image,

Quite thick and

heavy

� The MINOP Neuroendoscopy System (Aes-culap) includes the following (Fig. 7):

� Trocars with three different shaft diameteroptions. The 3.2-mm model has one

optic/working channel. The 4.6-mm mod-el has three optic/working, irrigation, and

overflow channels. The 6.0-mm model hasfour optic/working, additional working,irrigation, and outflow channels.

� 2.7-mmendoscopes that are angled to0� and30�. These are suitable for freehand surgeryas well as for fixation by the scope holder.


Fig 6. 6.2 mm rigid endoscope (Aesculap, Center Valley, Pennsylvania).

� Rigid instruments and electrodes� Good image quality; they are suitable for

‘‘pure’’ ventriculostomy, endoscopic sur-

gery, or endoscope-assisted microsurgery:

� Rigid wide-angled endoscopes for endoscope-assisted cranial neurosurgery; 0�, 30�, and 70�

angles of view are also available from

Aesculap.� The ‘‘classic’’ model Neuroview rigid scope,produced by NeuroNavigational (Integra

NeuroSciences, Plainsboro, NJ), offers supe-rior imaging characteristics. The outer di-ameter of the scope is 5.6 mm, with a 2.0-mmworking channel, irrigation and aspiration

ports, and a 2.8-mm scope channel. Thescopes are 2.7 mm in diameter and areavailable with 0�, 30�, and 70� offset angles,

providing 80� of view.

Despite higher image quality, the classic designmultiple-use scopes have some disadvantages.First, because the camera is directly attached to

the scope, they are quite heavy and cumbersome.Second, the classic scopes are thicker, leavinglarger holes in the brain. Third, there is a significantrisk of cross-contamination. Many surgeons, un-

fortunately, are familiar with bouts of postopera-tive infections secondary to cross-contamination

Fig 7. 4.6 mm rigid endoscope (Minop, Aesculap,

Center Valley, Pennsylvania).

occurring during endoscopic procedures. Finally,because of natural wear and tear of the opticalcomponents, the image quality deteriorates after

prolonged use. It was thus inevitable that dispos-able endoscopes would be introduced into practice.Note, however, that disposable rigid endoscopes

do not use a rigid lens optic system, because rigidlenses require a larger diameter. Instead, a 10,000-to 30,000-pixel fiberoptic system is used. The

quality of an image obtained through multipixelfiberoptic bundles is still significantly lower thanthat obtained with rigid lenses. Conversely, withdisposable scopes, the surgeon receives a brand

new set of optical components each time, whicheliminates the problem of wear and tear.

Fiberoptic endoscopes

This section discusses the flexible and rigidfiberoptic neuroendoscopes produced by a fewdifferent companies. Table 2 compares the advan-

tages and disadvantages of rigid versus flexibleendoscopes.

Several types of flexible scopes are availablefrom Aesculap:

� A steerable scope with an outer diameter of

3.9 mm. This scope has a 1.1-mm workingchannel and a 0.5-mm suction channel.Unidirectional steerability allows 160� turns.

� A steerable scope with an outer diameter of2.9 mm. This scope does not have a workingchannel. Although this scope is mainly used

for inspection purposes, a steerability of 140�

up and down makes this a reasonable choicefor dissection.

� A series of nonsteerable flexible endoscopes

with outer diameters ranging from 0.7 to 2.3mm. None of them, except the 2.3-mm scope(see section on syringomyeloscope), have

working channels. This group of endoscopesis usually used for endoscopic inspection,endoscopy-assisted surgery, or passing

through a shunt. These nonsteerable endo-scopes are superior to disposable flexiblescopes in image quality because they usehigher pixel optical fibers.


Table 2

Comparison of rigid and flexible scopes

Rigid scopes Flexible scopes

Advantage’s Better image Steerability

Higher resolution Application in spine

Wider view For some light weight, because camera not attached to scope

Better color

Better light transmission

Light weight of disposable scopes

Disadvantages Less maneuverable Poor image

Not applicable in spine Pixel granules

Narrower view

Less true color

Worse light

Smaller working channel

Limited selection of scopes and instruments

� Aesculap produces another endoscope origi-

nally designed by Perneczky. This instrumentis able to ‘‘look around the corner’’ whileoperating with the microscope. The Per-

neczky endoscope consists of a fiberopticscope encased in a rigid shaft with an 80�

curved tip 1.4 mm in diameter. Using this

curved tip, structures in the operating fieldthat are hidden from direct microscopic viewcan be inspected. The Perneczky endoscopemay be set into a desired position with a scope

holder (see section on scope holders).

Two types of flexible scopes are available fromCodman & Shurtleff:

� Codman & Shurtleff produces a steerableneuroendoscope that is similar to the 3.9-mm

endoscope from Aesculap. Its outer diameteris 4 mm, with a 1-mm working channel thataccepts several types of instruments and

accessories. The distal viewing tip can berotated 100� to 160� via a thumb control unit.

� Codman & Shurtleff also produces the Epicmicrovision. The Epic microvision is a semi-

flexible endoscope 1.8 mm in diameter witha bayonet-like design. The angle of the scopecan be changed by inserting it into more rigid

cannulas with straight, 20�, or 45� angles.Such small scopes can be used for:

� Better intraoperative visualization, partic-ularly in conjunction with an operative

microscope, when an ‘‘around-the-corner’’view is required

� Blunt perforation of membranes when thescopes themselves are used as dissecting

and fenestrating tools

� Endoscope-assisted shunt placement when

they are easily passed through the ventric-ular catheter as described in the followingsection.

Other fiberoptic scopes options include:

� NeuroNavigational produces a few dispos-able flexible endoscopes with outer diametersof 1.2, 2.3, and 4.6 mm. These scopes feature

a rigid design with 10,000 pixel optical fibersfor fine image resolution. The camera iselongated and not directly connected to the

endoscope body, which significantly decreasesthe weight of the instrument and makes itspencil-like body easy to handle. The 0.9-mm

Neuroview Fiberscope is one of the thinnestendoscopes available today. The 1.2-mmscope has only irrigation channels and is

designed to fit into the lumen of shuntcatheters. The 1.5-mm scope has a 0.49-mmworking channel, and the 2.3-mm scope hasa 1-mm working channel and a deflected tip,

which improves maneuverability and accessin confined areas. The 4.6-mm scope has twoworking channels.

� A few lightweight endoscopes are offeredby Clarus (Minneapolis, MN), including therigid fiberoptic Channel endoscope. It is quite

light and can be held as a pen (Fig. 8). TheChannel endoscope is available in twolengths: 13.0 and 21.6 cm. The working

channels may be either 3.15 or 2.15 mm indiameter. Irrigation is available through theirrigation port, with outflow through theworking channel. Other fiberoptic endoscopes

from Clarus include the MurphyScope and


the NeuroPen (designed for shunt insertion).Three variants of the MurphyScope are

available: 2.3-mm scopes with a bayonet-likemalleable 40� angle, a straight 50� curvedendoscope, and a 1.4-mm curved malleable

scope.

Comparison of rigid and flexible endoscopes

Rigid and flexible endoscopes each haveadvantages and disadvantages. Table 2 highlightssome of the pros and cons of each.

Endoscopes for shunt placement

It has been hypothesized that the use of a small-

diameter endoscope as a stylet within the ventric-ular catheter may allow more accurate catheterpositioning within the ventricle, which shoulddecrease the number of proximal shunt malfunc-

tions and revisions [5,10]. This theory has neverbeen proven. Nevertheless, several types of fiber-optic endoscopes are commercially available for

ventricular catheter placement. NeuroNaviga-tional, Clarus, Cordis (Miami, FL), and Codman(Johnson & Johnson, New Brunswick, NJ) all

produce semirigid lightweight scopes that have10,000 optic fibers, providing adequate visualiza-tion of intraventricular structures. The Neuro-Navigational 1.2-mm Neuroview endoscope,

a disposable unit, can be used in conjunction withseveral ventricular catheter modifications, such asthe commonly used Innervision catheter from PS

Medical (Goleta, CA). The Neuroview endoscopecan be used as a stylet to insert the catheter. Notethat although using this technique to insert the

catheter may decrease the possibility of completelymissing the ventricle, it does not necessarily

Fig 8. Light-weight fiberoptic endoscope (Clarus,

Minneapolis, MN).

improve the surgical outcome once the surgeonis within the ventricle, because no specific placewithin the ventricle has been proven superior toany other [5].

Syringomyeloscope

Endoscopes can be used in surgical treatmentof syringomyelia. Aesculap produces a flexible

syringomyeloscope with an outer diameter of 2.3mm. This scope has a working channel 1 mm indiameter, which may be suitable for 400-lm laser

fibers. This endoscope is mainly used for intra-cavity inspection and dissection after the lesion isaccessed via a laminectomy. A syringomyeloscopemay also be useful in the treatment of chronic

subdural hematomas.

Coagulation

Good hemostasis is essential in neuroendos-

copy for adequate visualization of structures aswell as safety. Hemostasis is achieved either withcautery systems (monopolar or bipolar) or with

lasers (discussed in the next section).Monopolar cautery is the most direct and

commonly used method to achieve hemostasis(Fig. 9). The Bugbee Wire (USA) is probably the

simplest monopolar cautery, commonly used byurologists as well as neuroendoscopists. TheBugbee Wire without applied cautery current

can also be used as a probe to ‘‘feel’’ tissue ormembrane (eg, ‘‘palpation’’ of the floor of thethird ventricle before fenestration) or to pierce

a membrane (eg, third ventriculostomy, fenestra-tion of cyst).

Fig 9. Various coagulation probes.


Slightly more sophisticated monopolar cauter-ies are available from Cook and Codman &Shurtleff. The Cook system is a number 3 French

retractable instrument that is kept within a pro-tective sheath while passing through the workingchannel. The surgeon squeezes the loop handle toexpose the tip as needed. Both round ball and

pencil-point tips are commercially available. TheCook electrodes may be used through rigid andflexible endoscopes. Codman & Shurtleff designed

the Micro Endoscopic Electrode (ME2), whichrelies on a retraction mechanism similar to theCook cautery.

Because there can be a problem with standardmonopolar cauteries of burnt tissue adhering tothe tip of the instrument, Heilman and Cohen [2]invented the ‘‘saline torch’’ monopolar cautery.

With this tool, the electric current is directedthrough a jet of saline, allowing the surgeon todissect and coagulate structures without direct

contact, even when completely immersed incerebrospinal fluid (CSF).

Bipolar coagulation is a more precise way of

achieving hemostasis with less scatter of currentcompared with a monopolar cautery. The simplestbipolar electrode is a fork electrode (eg, Aesculap

2.1-mm fork electrode). When using a forkelectrode, however, the surgeon cannot pick uptissue, which makes its use somewhat limited.Codman & Shurtleff therefore introduced grasp-

ing bipolar forceps 2.5-mm in diameter. Theseforceps are suitable for coagulation of vessels ofno more than 2 mm in diameter. NeuroNaviga-

tional offers both 1.0- and 2.4-mm flexible bipolarelectrodes. They can be used through rigid andflexible scopes. An advantage of these electrodes is

a built-in irrigating channel to decrease adherenceof the tissue to the wires. Finally, Clarusmanufactures a bipolar cautery that looks likea pencil and has a 30� tip angle. This allows the

surgeon better visibility of the cauterized object.

Instruments

Various grasping forceps are commercially

available from different manufacturers:

� Cook produces flexible forceps designed foruse through rigid and flexible scopes.

� Rat tooth, alligator, and mouse tooth forceps

can be used for dissection, enlargement of anopening, and pulling tissues. On a few occa-sions, the authors have used them with

endoscopic scissors to cut and remove an

occluded ventricular catheter overgrown bychoroid plexus (Fig. 10).

� Cup forceps, two- or three-pronged forceps,and flexible loop retrievals are designed to

obtain tissue for biopsy. The design of theseinstruments facilitates removal or retrieval oftissue fragments. Some of these instruments

are intended for single use. Aesculap manu-factures similar instruments in flexible andrigid variants. The malleable steering forceps

are available from Clarus.� Another helpful instrument is the Fogartyballoon (Fig. 11). Usually surgeons use

number 2 or 3 French catheters, dependingon the diameter of the working channel. TheFogarty balloon is used by many surgeons todilate the opening in the floor of the third

ventricle during an endoscopic third ventric-ulostomy. The advantage of this techniqueis that the inflated balloon compresses

Fig 10. Endoscopicmicrosurgical instruments (Aesculap,

Center Valley, Pennsylvania).

Fig 11. Balloon catheters.


surrounding tissues, constricting bleedingvessels.

� In the past few years, there have been casereports of accidental, often fatal, injuries to

the vessels of the basilar artery bifurcationcomplex [6]. Recently, a new device forperforation of the floor of the third ventricle

(‘‘no through perforator’’) was developed byGrotenhuis and became commercially avail-able from Synergetics (USA). It consists of

a 3-mm thick hollow rod with a sharp cuttingedge. After suction is applied, the floor of thethird ventricle ‘‘jumps up’’ and sticks to the

tip of the perforator, leaving the interpedun-cular vessels below. The device is turnedgently to create a round 3-mm diameter hole.The device seems to improve safety and lower

the risks involved in perforation of the floorof the third ventricle.

Lasers

The laser beam is a form of energy used insurgery to cut, coagulate, vaporize, and dissecttissues. Its applications are quite similar to those

of electricity; however, the nature of the laser isuniquely different from that of electricity.

Laser is an acronym for light amplification bystimulated emission of radiation. Laser energy can

be generated in a plasma tube by a powerfulelectromagnetic field. Application of this field tosome type of gas molecules, such as carbon

dioxide or argon, leads to their excitation.Electrons from the excited molecules start movingfrom one energy state to another. When these

molecules drop from an excited to a resting state,a photon (a unit of energy that has a characteristicwavelength) is released. The release of a photon

energy unit is called fluorescence. Once a photonhits a neighboring molecule, another photon isreleased, the wavelength of which is in phase withthe first photon. This process, called stimulated

emission, is self-perpetuating, because neighbor-ing molecules continue to bombard each otherwith photons. The mirrors on both tips of the

laser tube reflect the emitted photons, increas-ing the level of movement and energy withinthe tube until, finally, the beam emerges. On

emergence, the beam has the following threefeatures:

� The beam is coherent, because all photons are

released in the same phase.� The beam is monochromatic, because thematerial emits only a single wavelength.

� The beam is collimated, because the wavesemerge parallel to each other.

As a result of these features, a lens is able to

focus the light to a fine point, with an extremelyhigh concentration of power density [4,8].

Biologic effects of laserApplication of laser to tissue causes an in-

stantaneous increase in temperature, leading to

vaporization of the extracellular fluid and explo-sion of the cell. Four concentric zones of impactappear as a result of the interaction between laser

and tissue:

� Central zone—no cells remain, only burntdebris.

� Vacuolated cells zone—the tissue is composedof nonviable vacuolated cells that retain somestructure.

� Edematous zone—some cells in this area aredead, but most of the cells are viable withincreased intracellular water content.

� Reversible ultramicroscopic changes zone—although no alterations are seen underlight microscopy, some minor histochemicalchanges are present.

The first two zones are irreversibly dead. Thelatter two zones are viable and recoverable [4].

Types of medical lasersMany lasers are available for clinical use, but

only two are commonly used in neuroendoscopy

because of their ability to work through waterand transmit through the miniature fiberopticcables: neodymium:yttrium-aluminum-garnet

(Nd:YAG) and potassium-tetanyl-phosphate(KTP) [4,9,11,12].

Initially, lasers were popular in neuroendo-scopy, especially to make the opening in an

endoscopic third ventriculostomy [12]. Aftera few instances of injury to the vessels in theinterpeduncular fossa [6], however, laser use

significantly declined for safety reasons. Neverthe-less, the Nd:YAG laser and, especially, the KTPlaser may be useful in endoscopic tumor removal

and cyst fenestration. The tissue penetration ofthe KTP laser is less than with the Nd:YAG laser;its visible light makes it easier to handle, and it isless dependent on the tissue pigmentation, which

makes it suitable for dissection of colloid cysts.

Scope holders

Neuroendoscopy can be done either freehandby one or two surgeons (one navigating the scope


Table 3

Advantages and disadvantages of a scope holder

Freehand With rigid holder

Advantages More freedom of movement,

particularly when configuration

needs to be frequently or

continuously changes (eg, tumor

removal, inspection of

subarachnoid cisterns)

Surgeon has the use of both hands

Minimizes accidental movements

and tremor

Disadvantages More fatigue for surgeon More static (inflexible)

Greater risk of accidental

movements

Inconvenient when frequent

repositioning is needed

and the other working with instruments) or usinga rigid holder. Table 3 describes the advantagesand disadvantages of each technique.

Today, there are many scope holders available,

such as the Leyla self-retaining retractor, theflexible Greenberg retractor, the Neuroview ScopeHolder, and the Unitrac pneumatic holder pro-

duced by Aesculap (Fig. 12). The authors havefound the latter system particularly helpful, evenwhen operating freehand, because the holder can

be used as an easily adjustable hand rest. The

Fig 12. Unitrac scope holder (Aesculap, Center Valley,

Pennsylvania).

device uses the same pressurized gas that powerspneumatic drills.

Summary

In sum, understanding some of the basicprinciples of endoscopy and awareness of available

resources can potentially be of considerable help toexperienced neurosurgeons as well as beginners inselection of the most appropriate tools for different

procedures and making cost-effective choices whenbrowsing through multiple commercial advertise-ments and purchasing new equipment.

Although numerous advantages in science and

industry have made it possible to offer a widevariety of neuroendoscopes and tools, we believethe major achievements in this field are yet to

occur. This particularly refers to the developmentof smaller fiberoptic scopes with better imagequality and three-dimensional endoscopes and to

the invention of more efficient tools for endo-scopic tumor removal with the same degree ofsafety as in open surgery.

References

[1] Cohen AR. Endoscopic ventricular surgery. Pediatr

Neurosurg 1993;19:127–34.

[2] Heilman CB, Cohen AR. Endoscopic ventricular

fenestration using a ‘‘saline torch.’’. J Neurosurg

1991;74:224–9.

[3] Hulka JF, Reich H. Light: optics and television.

In: Textbook of laparoscopy. Philadelphia: WB

Saunders; 1994. p. 9–21.

[4] Hulka JF, Reich H. Power: electricity and laser.

In: Textbook of laparoscopy. Philadelphia: WB

Saunders; 1994. p. 39–43.

[5] Levy ML, Lavine SD, Mendel E, McComb JG. The

endoscopic stylet: technical notes. Neurosurgery

1994;35:335–6.


[6] McLaughlin MR, Wahlig JB, Kaufmann AM,

Albright AL. Traumatic basilar aneurysm after

endoscopic third ventriculostomy: case report [see

comments]. Neurosurgery 1997;41:1400–4.

[7] Mixter WJ. Ventriculoscopy and puncture of the

floor of the third ventricle. Boston Med Surg J

1923;188:277–8.

[8] Nobles A. The physics of neuroendoscopic systems

and the instrumentation. In: Jimenez DF, editor.

Intracranial endoscopic neurosurgery. Park Ridge,

IL: American Association of Neurological Sur-

geons; 1998. p. 1–12.

[9] Shiau JSC, King WA. Neuroendoscopes and instru-

ments. In: Jimenez DF, editor. Intracranial endo-

scopic neurosurgery. Park Ridge, IL: American

Association ofNeurological Surgeons; 1998. p. 13–27.

[10] Theodosopoulos PV, Abosch A, McDermott MW.

Intraoperative fiber-optic endoscopy for ventricular

catheter insertion. Can J Neurol Sci 2001;28:

56–60.

[11] Wharen REJ, Anderson RE, Scheithauer B, Sundt

TM. The Nd:YAG laser in neurosurgery. Part 1.

Laboratory investigations: dose-related biological

response of neural tissue. J Neurosurg 1984;60:

531–9.

[12] Wood FA. Endoscopic laser third ventriculostomy

[letter, comment]. N Engl J Med 1993;329:

207–8.


The anatomy of the ventricular systemDavid G. McLone, MD, PhD

Children’s Memorial Hospital, 2300 Children’s Plaza, #28, Chicago, IL 60614, USA

An understanding of the form and detail of theventricular system is an important point ofdeparture when beginning to learn neuroendos-

copy. Knowledge of the form and detail allowsplanning of the optimal entry point to enableviewing desired structures. This information also

acts as a road map to direct the endoscopist totargets. It is essential to know what is just beyondthe structure at the tip of the scope. Pathologic

processes, such as hemorrhage, ventriculitis, andhydrocephalus, may alter the appearance of theependymal surface of the ventricle but usually donot alter structures needed to navigate the

ventricular system. Obviously, knowledge of thefunction of structures within the field of interest isimportant to you and the patient.

Detailed function of the structures forming theventricular systems and the consequences ofinjuring them are beyond the scope of this article.

The author assumes the reader understands thefunctional anatomy.

What is seen on the monitor depends on theorientation at entry to the ventricle. Structures at

the immediate penetration point are difficult to seebecause they remain behind the lens. Noting theorientation of the patient’s head before draping is

critical for the endoscopist’s interpretation of thestructures and direction within the ventricularsystem. Initially, it may be helpful to have the

patient’s head straight brow up or lateral to limitthe mental gymnastics. As one develops a three-dimensional concept of the anatomy, orientation

becomes easier.As with all neurosurgical procedures, compli-

cations are always a possibility. Only with the useof this information and practice with the scopes



doi:10.1016/S1042-3680(03)00073-1

within the ventricular system will this becomea safe and useful procedure.

Development

The cerebral ventricular system at first seemscomplex; however, when understood from thepoint of the developmental anatomy, it is muchsimpler. Telencephalic vesicles bud from the

prosencephalon as symmetric bilateral spheres atthe extreme rostral end of the embryo. Theopenings into the diencephalic vesicle, the future

third ventricle, become the foramen of Monro.Choroid plexus develops along the dorsal raphaeof the diencephalic vesicle and splits at the

foramen, extending into each telencephalic vesicle.From this point, the rapidly developing cerebralhemispheres draw a group of structures thatoriginate near the foramen out into the ventricular

system. These structures become the endoscopicroad maps for the lateral ventricles. Structuresthat originate near the foramen and end up in the

temporal horn include the caudate nucleus,hippocampus, fornix, choroidal fissure, and cho-roid plexus. In lower mammals, the hippocampal

structures move posteriorly along the paramedianwith advancing phylogeny to end up in primitiveforms near the foramen and, in rodents, over the

posterior third ventricle.As the hippocampal structures move posteri-

orly, the single large bundle of fibers is drawn outand trails behind as the fornix. Columns of the

fornix fuse together as they pass over the foramenof Monro. At the anterior commissure, they againseparate and each side splits into a precommissural

tract and postcommissural tract. These two tractspass ventrally and laterally, one to the septal areaand the other to the mamillary body, thalamus,

and midbrain.

ghts reserved.


34 D.G. McLone /Neurosurg Clin N Am 15 (2004) 33–38

In primates, especially man, the parietal lobehas evolved from a structure principally forsensory function of the face to a structure serving

intellectual functions. This has caused a largeexpansion of the parietal lobe driving the tempo-ral lobe with its contents inferiorly and anteriorly.Rapid enlargement of the cortex of the frontal,

parietal, and temporal lobes opercularizes theinsular cortex.

The choroidal fissure with the choroid plexus

wraps around the thalamus, passing into thetemporal horn along its medial surface. Thehippocampus moves with the choroidal fissure

into the medial temporal lobe. This processcreates a lateral ventricular system that has theform of a ram’s horn. The head of the caudatenucleus begins anteriorly and laterally in the

anterior horn of the lateral ventricle. The bodyand tail of the caudate wrap around the lateralportion of the thalamus, again like a ram’s horn.

The venous drainage comes out of the caudatenucleus and thalamus in the groove between thecaudate and thalamus, the thalamostriate vein,

and ends at the foramen of Monro, where it joinsthe internal cerebral veins in the roof of the thirdventricle.

The corpus callosum originates as a commis-sural bundle anterior to the foramen of Monroand then expands superiorly and posteriorly asthe cerebral hemispheres blossom. Obviously, this

large midline commissure cannot continue intothe temporal lobe because of diencephalic struc-tures occupying the midline. Thus, the caudal end

of the corpus callosum, the splenium, comes torest at the posterior end of the third ventricleabove the habenular commissure, the pineal

gland, and the Galenic venous system.The columns of the fornix are connected to the

corpus callosum by paired thin membranes. Asthe hippocampus moves posteriorly with the

corpus callosum, these membranes are drawnout as the septum pellucidum. This creates twopotential cavities. One, located between the leaves

of the septum, is the cavum septum pellucidumand its posterior extension, the cavum verge.Below the fornix, the arachnoid and choroidal

fissures are folded back over the roof of the thirdventricle, creating the second potential space, thecavum vallum interpositum. In children with

poorly developed brains, these potential cavitiescan be large spaces that can either obstruct theventricular system and cause hydrocephalus orparticipate with the ventricles in the expansion

caused by hydrocephalus. These cavities are also

the sites where missed directed shunts or endo-scopes can end up, which presents a confusingpicture to the endoscopist.

The diencephalic vesicle becomes the thirdventricle, and the mesencephalic vesicle, which isthe largest portion of the embryonic ventricularsystem, becomes the relatively small aqueduct.

The fourth ventricle is created by the de-velopment of the cerebellum from the lip of therhombencephalon. This lip runs from side to side

in the coronal plane. Also developing along theedge of this lip is the choroid plexus of the fourthventricle. The expansion of the cerebellum rolls

the choroid plexus under the caudal edge. Thus,the choroid plexus runs from side to side from oneforamen of Luschka to the other in the inferiorroof of the fourth ventricle.

Anatomy of the lateral ventricles

In considering the lateral ventricular system, itis helpful to use the designation of the portions of

the ventricle that was employed by neurosurgeonswhen pneumoencephalography was in commonuse.

Portion 1: the frontal tip of the lateral ventricle

to the anterior edge of the foramen ofMonro

Portion 2: the anterior edge of the foramen ofMonro to its posterior edge

Portion 3: from the posterior edge of theforamen of Monro to the posterior edge ofthe thalamus

Portion 4: the trigone of the lateral ventriclefrom the posterior edge of the thalamus tothe beginnings of the occipital and temporal

hornsPortion 5: the occipital hornPortion 6: the temporal horn

Portion 1 of the lateral ventricle, with theoccipital horn and the tip of the temporal horn,

are portions of the ventricle without choroidplexus. The absence of choroid plexus and itsimmediate proximity to the foramen of Monro

are the major identifying characteristics of thisportion. Medially is the septum with the septalvein. Inferiorly is the septal area, and laterally is

the head of the caudate nucleus. The anterior limitand the roof are formed by the anterior fibers ofthe genu of the corpus callosum. These fibers passobliquely anterior and lateral to the frontal lobe;

thus, the anterior horn slopes anterior lateral.

35D.G. McLone /Neurosurg Clin N Am 15 (2004) 33–38

Portion 2 of the lateral ventricle is essentiallythe foramen of Monro and is the structure fororientation within the anterior lateral ventricle.The anterior and superior edge of the foramen is

the fornix. Above the fornix is the septumpellucidum, which forms the medial wall of thisportion of the ventricle and extends superior to

the corpus callosum. The corpus callosum formsthe roof, which joins laterally with the head and thebeginning of the body of the caudate nucleus.

The anterior thalamus forms the floor. Theanterior inferior edge of the foramen is the septalarea. Turning in the groove between the caudate

and thalamus is the thalamostriate vein, whichenters the foramen of Monro to join the internalcerebral vein. With the thalamostriate vein, thechoroids plexus passes into the foramen and

forms the posterior boarder.Portion 3 of the lateral ventricle is that portion

of the ventricle located above the thalamus. Its

main identifying characteristics are that it isthe only portion with the choroid plexus onthe medial floor and it is immediately behind the

foramen of Monro. The presence of the choroidplexus on the medial floor orients the endoscopistwhen the approach is posterior from the trigone.

The body of the caudate and the roof form thelateral wall by the corpus callosum. The thala-mostriate vein is usually a prominent vesselrunning in the groove between the thalamus and

caudate. The septum diminishes rapidly posteri-orly so that the most caudal part of the medialsurface of portion 3 is formed by the choroid

plexus, which covers the fornix. When approach-ing posteriorly, the choroid plexus and thethalamostriate vein are the road map to portions

1 and 2. The anterior thalamus is at the foramenof Monro, and the fornix and choroid plexus areclosely applied to the thalamic surface in portion3. The distance to the contralateral thalamus

through the septum pellucidum is short when thecontralateral ventricle is not dilated. Septostomydirected posterior to the foramen and at a small

contralateral ventricle risks injury to the contra-lateral thalamus and internal capsule, especiallythe genu.

Portion 4 of the lateral ventricle is theconfluence of portions 3, 5, and 6, often referredto as the trigone. It lies behind the thalamus and

contains the glomus of the choroid plexus.Medially, the visual radiations pass to occipitalpole, whereas ventrally and laterally is the whitematter of the occipital and temporal lobes. It is

important to be familiar with the characteristics of

the entry into portions 3 and 6 from this portionbecause this determines whether you enter into thefrontal or temporal horns.

Portion 5 of the lateral ventricle is the occipital

pole and the most variable in size of all theventricular portions. Again, it is one of theportions without a choroid plexus. Medially and

inferiorly are the visual radiations passing to themedial occipital lobe. Superiorly and laterally arethe fibers from the splenium of the corpus cal-

losum passing to the lateral and inferior occipi-tal lobe.

Portion 6, the final portion of the lateral

ventricular system, is located within the temporallobe. When approaching the trigone posteriorly,the temporal horn is slightly more lateral thanportion 3 and the choroid plexus is on the medial

roof rather than on the floor. Medially andsuperiorly are the choroid plexus and hippocam-pal structures. The deep white matter tracts of the

temporal lobe form the lateral wall.

Malformations of the lateral ventricles

Many of the malformations of the cerebralhemispheres are associated with hydrocephalus,

and ventriculostomy may be a treatment option.Hydrocephalus is a sign that the malformation hascreated an underlying obstruction causing prob-lems with the circulation of cerebrospinal fluid.

As in all good medical practice, a throughreview of the patient’s history and physicalexamination and study of the neuroimages before

the procedure are essential.Holoprosencephaly results from a lack of di-

vision of the prosencephalon into the two telence-

phalic vesicles. Hydrocephalus with an expandinglarge dorsal cyst may be present, requiring man-agement of the hydrocephalus. Abnormalities of

the face, such as extreme hypotelorism, reveal theunderlying brain malformation. Children with thismalformation lack intellectual development. Holo-prosencephaly is common with trisomy.

Schizencephaly is a defect in the cerebralmantle. It is most common in the third and fourthportions of the ventricular system but can occur in

the other portions. The ventricular system maycommunicate directly with the subarachnoidspace, which is referred to as open-lipped schi-

zencephaly. In closed-lip schizencephaly, the lipsof the cortical defect are fused. This may be anincidental finding in normal children or associatedwith seizures or intellectual delay. Hydrocephalus

is occasionally present.


Hydranencephaly is an extreme form ofschizencephaly in which most of the cerebralhemispheres are absent bilaterally, precluding

intellectual development. Hydrocephalus is oftenpresent. It is essential to distinguish this entityfrom maximal hydrocephalus, because the out-look is quite different.

Porencephaly is a cavity within the paraven-tricular white matter. Most communicate with theventricular system. They may be congenital or

occur secondary to injury, either vascular ortraumatic. Premature infants with intraventricularhemorrhage often have such a cavity located in

the frontal lobe, but it may be located anywherealong the germinal matrix. Poorly controlledhydrocephalus can cause a porencephalic cyst todevelop along the shunt tract.

Diverticula of the ventricular system are rare,but it is essential that they be recognized for whatthey are. Most often, they occur in cases of

hydrocephalus in which the supra- and infraten-torial compartments are isolated from each other.Differential pressures allow a portion of the

medial lateral ventricle to herniate into the para-sellar or quadrigeminal cisterns. These ‘‘tics’’usually arise along the choroidal fissure from

portions 3 and 4.Agenesis of the corpus callosum results in

a major decrease in the amount of white matterin each cerebral hemisphere. Because the third

ventricle is not constrained by the corpuscallosum, the fornix moves laterally, causingthe frontal tips of the lateral ventricles to

diverge. Portions 3, 4, and 5 are usually large,so-called ‘‘copocephaly,’’ expanding to fill thespace left by the absent white matter. Only rarely

is agenesis of the corpus callosum accompaniedby hydrocephalus.

Ependymal and choroidal cysts are not un-common within the lateral ventricles. They are

usually small and can spontaneously disappear.Occasionally, they are large and obstruct theventricular pathway, causing hydrocephalus prox-

imal to the obstruction. These cysts can bemanaged by endoscopic fenestration.

Loculations of the ventricular system are rare.

What is common is for paraventricular cysts tocoalesce and expand, producing what seems to bea loculated portion of the ventricle. When entering

these cavities, the absence of ventricular elements,choroid plexus for intense, reveals their nature.Biopsy of thewall reveals it to be composedofwhitematter containing myelinated axons. Hydrocepha-

lus is almost invariably present. The hydrocephalus

has usually been complicated by ventriculitis orventricular hemorrhage.

Anatomy of the third ventricle

A clear understanding of the structures thatcomprise the limits of this ventricle is extremely

important to the neuroendoscopist. Even moreimportantly, the structures just beyond the wallsof this ventricle must be clearly understood.

Anteriorly, the two fornices come close to-gether in the midline with only a small groove be-tween them, forming the anterior and superior

margin of the third ventricle and of the foramenof Monro. As the columns pass ventrally, eachsplits into an anterior and posterior limb over

the anterior commissure and moves laterally, withthe anterior limb ending in the septal area and theposterior limb ending in the mamillary bodies.Ventral to the anterior commissure is the lamina

terminalis, which ends in the optic chiasm. The sup-erior optic chiasm actually forms a portion of theanterior third ventricle.

The roof of the third ventricle is composed ofthe choroid plexus, internal cerebral veins, andcolumns of the fornix. As the most caudal portion

of the fornix, the columns move laterally, and aninterconnecting band of tissue, the fimbria, formsthe posterior superior roof of the third ventricle.Below these structures are the choroid plexus,

internal cerebral veins, and velum interpositum.The pineal recess and the pineal gland form the

posterior limits of the third ventricle superiorly.

Below the recess are the posterior commissure andthe aqueduct of Sylvius.

The paired mamillary bodies and the hypo-

thalamus form the floor of the third ventricle.The floor slopes downward from the mam-illary bodies into the infundibulum and pituitary

recess. The mamillary bodies are usually prom-inent structures that are easily seen. The floor ofthe ventricle immediately in front of the mamillarybodies is often not transparent, and structures

under it cannot be seen. Usually, when hydro-cephalus is present, the floor is thinned and thevascular structures are visible. More anteriorly,

the floor of the hypothalamus may be slightlypigmented. Below, exterior to the third ventricleis the posterior clinoid, and between the clinoid

and the interpeduncular fossa runs the mem-brane of Liliequist. If one passes through thefloor of the third ventricle in the midline anteriorto the mamillary bodies and the basilar artery

tip, the prepontine cistern is entered. As one

37D.G. McLone /Neurosurg Clin N Am 15 (2004) 33–38

move more rostrally and passes through the floorof the hypothalamus, the suprasellar cisternsanterior to the membrane of Liliequist are entered.The floor of the third ventricle is the hypothal-

amus, and injury can lead to endocrinopathies.Moving too far forward may cause the scope topin the floor against the posterior clinoid.

There are three recesses that can increasegreatly in size during hydrocephalus. They arethe optic, pituitary, and pineal recesses. When

distended, the optic recess leaves the anteriorcommissure as an obvious structure in theanterior third ventricle. The pituitary recess can

become large and thinned to the point of beingalmost transparent, revealing the structures in theparasellar cisterns. The pineal recess expands intoand often fills the quadrigeminal cistern.

The lateral walls of the third ventricle areformed by the medial surfaces of the thalami.Passing posteriorly along the superior margin of

the medial thalamus is the stria terminalis, whichends in the habenular nucleus and commissure.Because this structure is often obscured by the

choroid plexus of the third ventricle, it is usuallynot visible through the ventriculoscope. Themassa intermedia usually occupies the middle

portion of the third ventricle. This is a non-commissural structure of variable size connectingthe medial portions of the thalami. In the Chiari IImalformation, the massa intermedia occupies

a major portion of the third ventricle.Some important vascular structures are located

external to the third ventricle in the subarachnoid

space. Below the posterior floor of the thirdventricle in the interpeduncular cistern are the tipof the basilar artery and the two mesencephalic

portions of the posterior cerebral arteries. Thereare also a number of smaller penetrating arteriesthat branch from these main trunks. Anteriorly,just above the optic chiasm are the anterior cerebral

arteries and the anterior communicating artery.Running up the anterior surface of the laminaterminalis are the paired anterior cerebral arteries.

Malformations of the third ventricle

Severe hydrocephalus can significantly alterthe appearance of the structures at the foramen ofMonro. Even in these cases, the basic structures of

the road maps are available to locate the area forventriculostomy.

In the Chiari II malformation, the thirdventricle is usually remarkably different from

normal. A large portion is occupied by the massa

intermedia, but, more importantly, the floor isquite different. The distance between the mamil-lary bodies and the pituitary recess is extremelyshort, and the floor is thickened. This malforma-

tion often precludes any attempt at fenestration ofthe floor, especially in infants.

Agenesis of the corpus callosum allows the

third ventricle to expand superiorly between thecerebral hemispheres. The foramina of Monro aredisplaced laterally.

Anatomy of the cerebral aqueduct

The aqueduct is the small portion of theventricular system that connects the third andfourth ventricles. Superior to the aqueduct is the

quadrigeminal plate. Inferiorly is the tegmentumof the midbrain. Nuclei of the third cranial nerveare located bilaterally near the anterior portion of

the aqueduct.

Anatomy of the fourth ventricle

The normal fourth ventricle is a difficult targetfor endoscopy. Its size and orientation do not

allow easy access. More importantly, the floor ofthis ventricle is the brain stem. The sixth cranialnerve nucleus, the tract of the seventh cranialnerve, and the medial longitudinal fasciculus are

located superficially just below this floor. Al-though the cerebellum tolerates penetration,minor trauma to the floor inflicts significant neu-

rologic deficits. In pathologic conditions in whichthe fourth ventricle is enlarged, however, access ispossible.

The floor of the fourth ventricle is diamondshaped. The anterior point begins at the aqueduct,the lateral points hook slightly posterior to form thelateral recesses, and the posterior point ends at

the obex. The normal floor is relatively flat withsome identifying structures in it. Running in themidline is the median sulcus, and near the mid-

point, the stria medullaris runs from the sulcustoward each lateral sulcus. The facial colliculusis located on either side near midline just rostral

to the stria medullaris.The walls of the fourth ventricle are formed by

the superior and inferior cerebellar peduncles. The

fourth ventricle has a pyramid shape, with the apextilted slightly posterior. The anterior and posteriormedullary velum forms the roof. The posterior sideof the pyramid is convex because of the cerebellar

vermis bulging inward. The vermis also splits theapex into two recesses: the posterior superior


recesses. The midline apex is the vestigium. Thechoroid plexus is located on the posterior surface ofthe fourth ventricle and runs in the coronal plane

from one lateral recess to the other. Foramina thatdrain the fourth ventricle are located bilaterally inlateral recesses: the foramina of Luschka and, inthe midline, the foramen of Magendie. The

choroidal fissure runs with the choroid plexus fromone lateral recess to the other, separating theinferior cerebellum from the brain stem.

Malformations of the fourth ventricle

Dandy Walker cysts represent a group of

anomalies in which a portion of or the entirecerebellar vermis is absent and the fourth ventriclecommunicates with or becomes a large cyst. When

it occurs early in development, the torcula islocated high, often near the midparietal region.Hydrocephalus may or may not be present. In itsmost severe form, it is associated with other brain

malformations, such as agenesis of the corpuscallosum, and intellectual delay is common in thisform.

Loculation of the fourth ventricle, or a trappedfourth ventricle, like loculations in the lateralventricles, is usually a complication of hydro-

cephalus. Ventriculitis is the most common cause.The Chiari II malformation is a malformation

that affects the entire brain. It has a major impact

on the contents of the posterior fossa. Because thesize of the embryonic posterior fossa is much too

small to contain the rhombencephalic and meten-cephalic structures that will develop, derivatives ofthese embryonic portions of the brain extrude out

of the posterior fossa. A portion of the (occasion-ally the entire) cerebellum and fourth ventricle isdisplaced into the cervical canal. In other individ-uals, the deficient tentorium allows a major

portion of the cerebellum and fourth ventricle toherniate upward into the middle fossa between thecerebral hemispheres. When the ventricular shunt

fails or the fourth ventricle becomes trapped orisolated, the fourth ventricle expands, increasingthe pressure on the surrounding hindbrain. This is

usually manifested as a dramatic increase in thepatient’s symptoms. When the isolated fourthventricle in the Chiari II malformation expandssuperiorly, endoscopic access is available from

above in a plane parallel to the floor of the fourthventricle. This affords the opportunity to fenes-trate or shunt the fourth ventricle endoscopically

without endangering the floor of the fourthventricle.

Summary

The embryology of the ventricular develop-

ment of the brain assists in understanding the finalrelations between structures forming these cavi-ties. An accurate concept of this anatomy allows

the endoscopist to maneuver within the ventricu-lar system.


Selecting patients for endoscopic third ventriculostomyHarold L. Rekate, MD

Pediatric Neurosurgery, Barrow Neurological Institute, 2910 North Third Avenue, Phoenix, AZ 85013, USA

At the annual meeting of the American

Association of Neurological Surgeons Meetingin 1995, Professor Fred Epstein was called todiscuss our paper on the use of endoscopic third

ventriculostomy (ETV) as the ultimate treatmentfor the ‘‘slit ventricle syndrome’’ (SVS) as pre-sented by Dr Jonathan Baskin [1]. Dr Epstein

was surprised to learn that we were viewing allpreviously shunted patients with shunt problemsas potential candidates for the procedure andwondered aloud whether patients with normal

pressure hydrocephalus (NPH) might not beconsidered candidates for ETV.

The thought process leading to the following

presentation regarding who is and who is nota candidate for ETV was largely generated by thatdiscussion. Can we know beforehand who will

and who will not benefit from an ETV? What arethe absolute contraindications for performing theprocedure? What are the relative contraindica-

tions? How can we analyze the risk-benefit ratiofor ETV in various clinical settings? For the mostpart, defining the role of ETV in the overallmanagement of hydrocephalus must await large,

multicenter, randomized, prospective trials. Inthe interim, I will attempt to define a rationalapproach to the problem in partial answer to the

posed questions.

Anatomy and biophysics of the cerebrospinal fluid

as it relates to third ventriculostomy

The credit for defining the basic pathophysio-logic mechanisms in hydrocephalus goes to Walter

Dandy and his colleagues at Johns HopkinsUniversity for work done between the secondand fourth decades of the twentieth century.



doi:10.1016/S1042-3680(03)00074-3

There had been many previous pathologic de-

scriptions of the effects of hydrocephalus oncadaveric material. Dandy and Blackfan [2] stud-ied laboratory animals and patients with hydro-

cephalus by injecting supravital dyes into thelateral ventricles of the subjects. They then per-formed lumbar punctures to determine whether

the dye could be recovered in the spinal sub-arachnoid spaces (SSASs) [3]. Based on thisinformation, Dandy and Blackfan classified hy-drocephalus into communicating (the dye was

recoverable) and obstructive or noncommunicat-ing (the dye was not recoverable). As a result ofthese findings, they recommended attempting to

create a communication between the third ventri-cle and the subarachnoid spaces by performing anopen craniotomy and initially resecting one of the

optic nerves.By applying Dandy’s classification, patients

with noncommunicating hydrocephalus would be

considered candidates for ETV to create aninternal bypass for the treatment of hydrocepha-lus. Over the years, there has been a tendency toequate noncommunicating hydrocephalus with

triventricular hydrocephalus, which could bediagnosed by air studies and, subsequently, byCT and MRI. In 1960, Ransohoff and Epstein [4]

voiced their objection to the Dandy classification.They believed that all hydrocephalus was obstruc-tive and preferred to classify the condition into

intraventricular obstructive hydrocephalus andextraventricular obstructive hydrocephalus. Thelatter would be consistent with Dandy’s commu-nicating hydrocephalus. Based on this discussion,

many patients with communicating hydrocepha-lus are good candidates for ETV.

With current imaging technologies, it is usually

possible to determine the actual site of theobstruction and to plan treatment to address thespecific pathophysiologic mechanisms involved. In

ghts reserved.


40 H.L. Rekate /Neurosurg Clin N Am 15 (2004) 39–49

our laboratory, we have measured the amount ofresistance that can be anticipated at each point ofpotential obstruction. Because of the presence

of multiple ventricular catheters, we have alsoattempted to measure intraventricular pressures invarious components in human beings. Fig. 1 isa schematic diagram of the anatomy of the

cerebrospinal fluid (CSF) pathways in a patientwith aqueductal stenosis as a hydraulic analogue toan electrical circuit. In normal animals, pressure

differentials can be measured only at the level offlow of CSF from the cortical subarachnoid spaces(CSASs) into the superior sagittal sinus (SSS) [5,6].

For CSF to be absorbed at the level of the SSS,intracranial pressure (ICP) must be 5 to 7 mm Hghigher than the pressure in the SSS. Fig. 2 isa schematic diagram of the physical effect of an

ETV. Essentially, the procedure is an internalbypass between the third ventricle and the inter-peduncular cistern in the CSASs.

Given this paradigm, all patients with ob-structions between the third ventricle and the

CSASs are potential candidates for ETV. Notonly can ETV treat hydrocephalus caused byaqueductal stenosis, but it is at least of theoretic

benefit in obstruction of the outlet foramina ofthe fourth ventricle and to blockage of CSF flowat the level of the basal cisterns between theSSASs and the CSASs. Only when CSF flow is

obstructed at the level of the arachnoid villi orvenous flow from the sagittal sinus is restricteddo patients fail to benefit from ETV. These forms

of distal obstruction represent absolute contra-indications to ETV.

With contemporary neuroimaging techniques,

it may or may not be possible to determine theactual site of obstruction to the flow of CSF. Themore definitive the diagnosis of the site ofobstruction, the better is the physician’s ability

to predict the outcome of ETV when used to treatpatients with hydrocephalus. Some patients withhydrocephalus may have an obstruction at more

than one point along the CSF pathway. In spinabifida, hydrocephalus caused by a Chiari II

Fig. 1. Schematic hydraulics of the cerebrospinal fluid system with aqueductal stenosis. (Courtesy of Barrow

Neurological Institute.)

41H.L. Rekate /Neurosurg Clin N Am 15 (2004) 39–49

Fig. 2. Schematic diagram of the effect of performing an endoscopic third ventriculostomy on the circuit diagram of

cerebrospinal fluid flow. (Courtesy of Barrow Neurological Institute.)

malformation is associated with as many as fourdifferent sites of potential obstruction to the flowof CSF [7]. Based on patients’ clinical manifes-

tations at shunt failure, all these potential sites ofobstruction actually occur and any patient mayhave more than one site of obstruction.

Patients with hydrocephalus caused by intra-ventricular or subarachnoid hemorrhage (SAH)for whatever cause (prematurity, trauma, or

aneurysmal bleeding) usually have a CSF absorp-tive difficulty at the level of the basal cisterns(SSASs to CSASs). They could have an obstruc-

tion at the level of the arachnoid granulations,however, which would not respond to ETV. Pa-tients who suffer significant infections may develophydrocephalus. The most common location is also

the basal cisterns. Patients with an infection mayhave an obstruction at the level of the outletforamina of the fourth ventricle or at the level of

the arachnoid granulations, however.

Absolute contraindications for endoscopic third

ventriculostomy

‘‘Communicating hydrocephalus’’: obstruction to

the absorption of cerebrospinal fluid

ETV should not be performed if CSF is alreadyflowing unimpeded between the ventricles and theinterpeduncular cisterns. It also should not be done

if it is technically impossible to manipulate theendoscope within the lateral and third ventricles toa position that allows the floor of the third ventricle

to be visualized. These are the only two absolutecontraindications for performing the procedure. Inthe first instance, flow between the ventricles and

the CSASs is not impeded. Bypassing the aqueductof Sylvius, outlet foramina of the fourth ventricle,and the basal cisterns offers no advantage, becausethe point of obstruction is ‘‘downstream’’ from the

internal bypass. As can be seen from a review ofFig. 2, this condition occurs only if the arachnoid


granulations are obstructed or the pressure of theSSS is increased markedly.

As an isolated event, obstruction of the

arachnoid granulations is quite rare. Pathologi-cally, congenital absence of these structures is thecause of at least some cases of benign familialmegalencephaly or external hydrocephalus. Ra-

diologically, this type of obstruction shows mildto moderate ventriculomegaly and distention ofthe CSASs [8–10]. The terminal absorptive mech-

anisms can be clogged by blood or by inflamma-tory or tumor cells. This clogging is usuallyaccompanied by thickening and scarring of the

arachnoid in the basal cisterns. Although the CSFabsorptive failures in acute SAH and meningitismay be related to an obstruction at this level, thehydrocephalus usually results from an obstruction

between the SSASs and CSASs, a condition that isamenable to ETV.

Hydrocephalus caused by obstruction of

absorption to CSF at the level of the arachnoidvilli is never an all-or-none phenomenon. It ispressure dependent in that the presence of the

particulate matter changes the pressure flowcharacteristics of the valvular mechanism alreadypresent. Under normal conditions, the arachnoid

villi (microscopic structures) contained within thearachnoid granulations (macroscopically visible)function as a differential pressure valve betweenthe CSASs and SSS with a closing pressure of 5 to

7 mm Hg (70–100 mm H2O). In adults withclosed sutures and fontanels, clogging of thearachnoid villi usually increases ICP, with a min-

imal increase in ventricular size. In the context oftraumatic or aneurysmal SAH, the usual in-dication for shunting is the failure to be able to

wean the patient from an external ventriculardrain in the presence of minimal ventriculome-galy. The shunt can usually be removed later inthe course of a patient’s life, but ventricular size

may increase significantly with symptoms at thetime of shunt failure [11]. In this case, the point ofobstruction usually is at the level of the basal

cisterns and no longer at the level of thearachnoid granulations.

In infants with neonatal intraventricular hem-

orrhage or meningitis, clogging of the arachnoidvilli increases both the CSAS and ventricularvolume. In this situation, the open fontanels and

distensibility of the sutures allow CSF to accu-mulate, because ICP cannot increase to the pointthat it will overcome the valvular mechanisms inthe arachnoid villi. This condition responds to

increasing ICP by wrapping the head with an ace

bandage. ICP then increases to a point where CSFcan be absorbed [12].

The second point of obstruction leading to the

type of ‘‘communicating hydrocephalus’’ thatcannot be affected by ETV involves obstructionto outflow of venous blood from the SSS. Inadults, marked increases in pressure in the SSS

lead to pseudotumor cerebri rather than tohydrocephalus [13,14]. In infants with openfontanels and distensible sutures, marked increases

in pressure in the SSS lead to marked ventriculo-megaly associated with concomitant increases inthe volume of the CSASs. At presentation, it may

be difficult, if not impossible, to discern that thehydrocephalus is caused by increased venouspressure. This form of hydrocephalus can lead tosecondary obstruction of the aqueduct as the brain

stem is displaced inward and the temporal horns ofthe lateral ventricles displace the midbrain medi-ally [15]. This form of hydrocephalus has been well

studied in the context of achondroplasia andcraniofacial syndromes [16–20]. In the case ofCrouzon’s and Pfeiffer’s syndromes, the hydro-

cephalus may not be evident until a formal cranialremodeling operation has been performed. Thesepatients share the same syndrome in that their

ventricles do not dilate to abnormally largevolumes at the time of shunt failure. They showsigns of overtly increased ICP without ventriculo-megaly, a condition that has been referred to as

‘‘normal volume hydrocephalus’’ [21].

Unresponsive ventricles: ventricles too small ortoo distorted to manipulate an endoscope safely

Performing an ETV requires that the endoscope

be placed in the lateral ventricle and manipulatedthrough the foramen of Monro to the floor of thethird ventricle, where the hole can be made. Thesmaller the ventricles are, the more difficult it

becomes to perform the manipulations needed tofenestrate the floor of the third ventricle. If the sizeof the ventricles is normal or smaller than normal

after chronic shunting, it may be impossible tomanipulate the endoscope safely to the appropriatelocation to perform the fenestration. There are two

important circumstances in which third ventricu-lostomy is contraindicated in this context. Markedenlargement of the massa intermedia of the third

ventricle is a well-recognized concomitant of theChiari II malformation in the context of spinabifida. In these situations, the walls of the thirdventricle are often closely opposed because of

the size of the massa intermedia despite the


Fig. 3. MRI of a patient with spina bifida showing enlargement of the massa intermedia and a small third ventricle

despite marked enlargement of the lateral ventricles. (A) Sagittal T2-weighted MRI demonstrating the enlarged massa

intermedia in a small third ventricle despite massive enlargement of the lateral ventricles. (B) Axial T2-weighted MRI

demonstrating the difference in size between the lateral and third ventricles.

enlargement of the lateral ventricles (Fig. 3). Thethird ventricle can be markedly distorted inpatients with a Chiari II malformation. Even when

the third ventricle can be cannulated, the anatomiclandmarks can be so distorted that it may beimpossible to find a point on the floor of the third

ventricle to make the fenestration.The second context in which ETV is contra-

indicated because it would be unsafe to manipulate

the endoscope occurs in patients with normalvolume hydrocephalus as described previously.These patients have smaller than normal ventricles,and the ventricles do not expand at the time of

shunt failure. These patients should not undergoETV for two reasons. First, the procedure isassociated with a high risk of damaging the

columns of the fornix, mamillary bodies, cerebralpeduncles, or hypothalamus. Second, ETV in thiscontext is unlikely to be of benefit to the patient.

We have performed iohexol ventriculography in 31such patients, and 28 showed free communicationfrom the ventricle to the interpeduncular cistern

[22]. In the 3 patients who did not show suchcommunication, treatment of their hydrocephaluswas complicated by a significant infectious processafter the original shunt had been placed.

Some nonresponsive ventricles have resulted insurprising observations. In one patient whosehydrocephalus was associated with a pineal tumor

originally treated in infancy, overt shunt failureoccurred despite no change in ventricular volume.Ventriculography revealed that the dye injected

through the shunt into the lateral ventricle rapidlyflowed into the interpeduncular cisternarea (Fig. 4).The hydrocephalus of patients with normal

volume hydrocephalus is related to increasedvenous pressures, and they are not candidatesfor ETV for both reasons.

Selection of candidates for endoscopic third

ventriculostomy as an initial procedure in the

treatment of hydrocephalus

Late occlusion of the aqueduct of Sylvius

Hydrocephalus caused by occlusion of theaqueduct of Sylvius is most common in infantswho manifest overt hydrocephalus as newborns or

whose head circumference rapidly increases afterbirth. This indication would seem to be ideal forperforming an ETV, because occlusion of theaqueduct obstructs CSF flow between the third


Fig. 4. (A) Sagittal MRI of a child with hydrocephalus associated with a pineal region tumor showing communication

between the lateral ventricles and spinal subarachnoid spaces despite the presence of the tumor. (B) CT of the basal

cisterns after iohexol has been injected into the lateral ventricle showing the flow of contrast into the basal cisterns, thus

confirming communication in the cerebrospinal fluid pathways.

and fourth ventricles. Theoretically, CSF shouldbe left in the interpeduncular cistern in commu-

nication with the arachnoid granulations, whichshould be normal. Creating this internal bypassbetween the third ventricle and the cistern shouldnormalize CSF dynamics.

Aqueductal stenosis is often diagnosed byMRI only when triventricular hydrocephalus ispresent and the size of the fourth ventricle is

normal. Triventricular hydrocephalus does notalways mean that the underlying cause of thehydrocephalus is aqueductal stenosis. The phe-

nomenon of communicating hydrocephalus caus-ing dilatation of the temporal horns of the lateralventricle and compression of the midbrain fromboth sides leading to a secondary occlusion of the

aqueduct of Sylvius has been discussed previously.In these patients, shunting opens the aqueduct ofSylvius to CSF flow again [15]. Depending on the

initial site of obstruction, the patient may or maynot be a candidate for ETV.

Regardless of the site of obstruction to CSF

flow, infants are usually poor candidates fora third ventriculostomy, presumably because theventricles fail to respond after third ventriculos-

tomy despite the blockage at the aqueduct. Inthese cases, the ventricles are quite large; unlikeshunting, which literally sucks CSF from thebrain, ETV normalizes CSF absorption. The

latter process requires intraventricular pressureto increase 5 to 7 mm Hg greater than atmo-

spheric pressure, which may be impossible inbabies with open anterior fontanels. Some authorssuggest that these infants should not be candidatesfor ETV because of their poor response rate

[23,24]. Other authors quote low rates of shuntindependence after ETV but believe that thebenefit is great enough and the risks of the

procedure are low enough to justify ETV ininfants in an attempt to avoid placing a shunt [25].

Aqueductal stenosis is diagnosed in older

children and adults in two contexts. It is difficultto establish treatment criteria for and to assess theoutcome of treatment in patients who developsymptomatic hydrocephalus later in life and have

extremely large heads. Obviously, these patientshave had severe ventriculomegaly from infancy.Oi et al [26] have termed this condition long-

standing overt ventriculomegaly of the adult(LOVA). These patients usually have no overtsymptoms after undergoing an imaging study for

a seizure or minor head injury. Frequently, theyhave no symptoms related to their hydrocephalus.When asked, some patients identify daily head-

aches, clumsiness, or poor memory that may ormay not have changed recently.

How to treat these patients and, indeed,whether to treat these patients are controversial


issues. Some authors strongly believe that thesepatients will benefit from intervention, usuallyventricular shunting [27]. Data showing improve-ment in these patients are hard to find, and such

patients seldom have high ICP. My approach tothese patients is to obtain formal neuropsycho-logic studies to determine whether they exhibit

problems with higher cognitive function. If so, it isimportant to determine whether the deteriorationis recent or long standing. I then provide patients

with the pertinent information and attempt tohelp them reach a treatment decision. A signifi-cant decrease in the size of the lateral ventricles

after shunting or ETV is unusual. Shunts can beshown to be working manometrically, but it maybe difficult to determine whether an ETV hasproduced the best outcome. Imaging studies, at

least cine MRI through the floor of the thirdventricle and preferably the injection of iodinatedcontrast material into the lateral ventricles to

trace its flow into the basal cisterns, is needed toconfirm that CSF dynamics have been maximized.

There is one important caveat if the decision is

made to attempt to treat LOVA with ETV.Occasionally, the head is so large that someneuroendoscopes are too short to reach the floor

of the third ventricle. The distance from the skullto the floor of the third ventricle can be measuredon the console of the CT or MRI scanner. Aselection of endoscopes should be available to

obviate this problem.In the second clinical syndrome, older patients

develop hydrocephalus caused by aqueductal

stenosis from the presence of severely increasedICP. Patients often seek treatment from anophthalmologist for some form of ophthalmo-

plegia, such as bilateral sixth nerve palsies orParinaud’s syndrome. On examination, the patientis found to have bilateral papilledema. Thisphenomenon can occur as an isolated event with

obstruction caused by fusion of the walls of theaqueduct of Sylvius as occurs in some cases ofneurofibromatosis 1. This overt form of late

aqueductal stenosis is more likely to result fromsmall benign tumors (tectal gliomas) of theaqueduct. These are true tumors, but their prog-

nosis in terms of propensity to grow or disseminateis usually good [28]. A direct attack on thesetumors, even for a biopsy, is seldom warranted.

This acute adult form of hydrocephalus fromaqueductal stenosis is perhaps the most compel-ling and most satisfying condition to treat withETV. The success rate for this particular subset of

patients is 75% to 80%. The rate of significant

morbidity, including endocrinopathy, memorydeficits, and hemiparesis, has been reported tobe 3%. After 1 year, less than 1% are stilltroubled by complications of ETV [24]. In this

situation, the risk-benefit ratio favors ETV overinternal shunting.

Hydrocephalus associated with newly diagnosedbrain tumors, particularly of the posterior fossa

This area of study is evolving, and the use ofETV before tumor removal is discussed more than

it is reported in the literature. Posterior fossatumors usually manifest with hydrocephalus andacutely increased ICP. The blockage of the CSF

flow meets the criteria for ETV in that CSF flowbetween the third ventricle and interpeduncularcistern is obstructed. Conceptually, performing an

ETV in this context is similar to placing aninternal shunt before tumor resection. It relievessymptoms in most patients and thus allows a directattack on the tumor at a later stage. It shares with

the placement of an internalized shunt thepossibility, or even the probability, that it willnot be needed after the tumor has been removed.

It shares with shunts the real possibility of leadingto an upward herniation syndrome—the hernia-tion of the superior cerebellar vermis through the

tentorium, leading to compression of the mid-brain. Typically, contemporary pediatric neuro-surgeons begin patients on high doses of

dexamethasone at the time of diagnosis tostabilize their condition. When the tumor isattacked directly, an external ventricular drain isplaced before the lesion is removed.

The role of ETV in the overall management ofchildren with brain tumors affecting the CSFpathways is evolving. Soon a series of publications

will support its more or less routine use. We awaitthe results of those studies before applying thistechnique generally. If neurosurgeons elect to

perform an ETV before removing a posteriorfossa brain tumor, they must carefully review theanatomy as it appears on sagittal MRI scans. In

this situation, the brain stem is often pushedanteriorly against the clivus. The distance betweenthe clivus and basilar artery is short, which makesthe procedure more risky than it would be

otherwise. The procedure should be performedby an experienced neuroendoscopist and canbe made significantly safer by using frameless

stereotaxy.If tumor resection is not essential to the

patient’s outcome, the neuroendoscope becomes


a much more exciting tool. Adolescent malepatients with a pineal tumor that enhancesdiffusely, hydrocephalus, and no markers of

malignancy (elevated levels of a-fetal protein orhuman chorionic gonadotrophin) are likely tohave a pineal germinoma. Ideally, these patientsshould be managed by using an endoscope to

perform a third ventriculostomy to obtain a biopsyof the tumor. Definitive treatment can then bemanaged using conformal radiation therapy. This

approach is associated with minimal rates ofmorbidity and high rates of success.

Hydrocephalus associated with intraventricularand subarachnoid hemorrhage

Whether SAH results from diffuse craniocere-

bral trauma or from the rupture of an aneurysmor arteriovenous malformation (AVM), patientsare often left with a significant CSF absorptionproblem. Treatment is usually improved sub-

stantially by the widespread use of externalventricular drains. In these contexts, the mostcommon indication for shunting is failure to wean

the patient from the ventriculostomy without highlevels of ICP. The ventricles do not dilate. ETV iscontraindicated based on the two reasons stated

previously. The ventricles are small enough tomake performing ETV problematic. The pre-sumed point of obstruction is probably thearachnoidal granulations distal to the interpedun-

cular cisterns. Therefore, these patients are un-likely to benefit from the procedure. They respondto treatment using either ventricular shunting or

lumbar shunting. If it is possible to temporize longenough, they may not need a shunt at all.

If the inflammatory process associated with

SAH is allowed to continue long enough, theprimary point of obstruction becomes the basalcisterns. CSF flow is then occluded between the

SSASs and CSASs. This process also followsbacterial or fungal meningitis. Under the Dandyclassification system, these forms of hydrocepha-lus would be considered ‘‘communicating hydro-

cephalus.’’ Consequently, patients would beconsidered candidates for ventricular shunting orlumboperitoneal shunting. The obstruction to

flow is upstream from the interpeduncular cistern,however, and patients may indeed respond toETV. In fact, the success of the procedure has

been highest in this group.This point of obstruction causes most cases of

NPH [29]. In a small series, three of four patients

with NPH had excellent outcomes after treatment

[30]. This approach represents an exciting area forfuture research.

Special case: hydrocephalus in infancy

ETV is much less likely to be successful whenused to treat infants than when used to treathydrocephalus later in life. Consequently, a num-

ber of authors have concluded that an age lessthan 6 or 12 months should be considereda contraindication to the use of ETV in the

management of hydrocephalus. A number ofreasons may underlie this failure. The first relatesto the ability to define the point of obstruction in

these babies with severe hydrocephalus. Becauseof the presence of open fontanels and unfusedsutures, the degree of hydrocephalus is greater

than when hydrocephalus occurs later in life.Furthermore, it is difficult to interpret the scans todetermine the actual point of obstruction. Thesame patient also may have multiple points of

obstruction, especially when hydrocephalus isassociated with the Chiari II malformation (spinabifida cystica). In the latter case, there are four

different points of obstruction, and more than onesite can be obstructed at once [7]. When hydro-cephalus occurs later in life, obstruction is less

likely to occur at more than one site. It is alsoeasier to determine where the actual point ofobstruction is based on imaging studies.

The second reason relates to the actual physics

of the system. When the anterior fontanel isopened widely and the sutures are splayed, theintracranial compartment is essentially in com-

munication with and has the same pressure asatmospheric pressure. Natural absorption of CSFdepends on ICP being at least 5 mm Hg greater

than sagittal sinus pressure [5,31,32]. It is ofteneasier for the size of the head to expand than tomaintain ICP greater than 5 mm Hg. In addition,

Laplace’s law implies that the larger ventricles are,the less distending force is needed to maintainthem at that size.

Although ETV is less likely to be successful in

these children than in older children, goodoutcomes are obtained in some babies. Theprocedure also may be safer to perform in infants

than in older children or adults. The floor of thethird ventricle is more diaphanous in infants thanit is likely to be in older children. The basilar

artery can be identified with certainty and thusprotected. The ventricular membrane is so fragilethat it can usually be opened with normal

irrigation.


Theoretically, it is better to have communicat-ing than noncommunicating hydrocephalus.Whena patient with hydrocephalus related to aqueductalstenosis experiences a shunt failure, no mecha-

nisms are available to compensate for the increasedICP that accompanies failure of the shunt. If thepoint of obstruction is distal, a significantly longer

time is likely to elapse before the situation becomescritical.

Endoscopic third ventriculostomy for programmed

shunt removal

Chronically shunted patients may suffer from

severe and incapacitating headaches caused bySVS. The percentage of patients suffering from thisproblem is controversial, and some investigators

doubt its existence. If the threshold for making thisdiagnosis includes not only the classic triad ofsevere intermittent headaches lasting 10 to 90minutes, smaller than normal ventricles on imaging

studies, and a slowly refilling flushing mechanismbut the need to discontinue activity or be broughthome from school at least two times per month,about 15% of chronically shunted older children

and adults suffer from this condition [33]. Based onchronic monitoring of ICP, we have identified fourdifferent pathophysiologic mechanisms responsi-

ble for SVS: intermittent proximal obstruction,intracranial hypertension with smaller than nor-mal ventricles and a failed shunt (normal volume

hydrocephalus), intracranial hypertension witha working shunt (cephalocranial disproportion),and migraine in shunted patients [34].

Originally, these patients were managed withshunt revision using higher resistance valves anddevices that retarded siphoning. Recently, wehave offered these patients the opportunity to

assess the possibility of becoming shunt indepen-dent [1]. These patients are admitted to thehospital for externalization of their shunt, or their

shunt is removed and replaced with an external

Fig. 5. Algorithm for managing shunt-related difficulties with a shunt removal protocol. (Courtesy of Barrow

Neurological Institute.)


Box 1. Contraindications to endoscopic third ventriculostomy

Absolute contraindications1. Third and lateral ventricles too small to allow safe manipulation of the endoscope within

the ventricular system2. Tumor or other mass lesion obstructing the surgeon’s access to the floor of the third

ventricle3. Proven communication between the CSF that is in the third ventricle with CSF within the

interpeduncular cistern

Relative contraindications1. Multicompartment hydrocephalus2. Thickened floor of third ventricle3. Chiari II malformation4. Age less than 1 year5. Distortion of third ventricular anatomy

CSF = cerebrospinal fluid.

ventricular drain. Under controlled conditions inthe intensive care unit, their shunt is occluded orraised to a higher level and a scan is obtained.About two thirds of these patients become

symptomatic with marked ventricular dilatation.These patients undergo ETV, and 70% tolerateshunt removal. Patients who develop marked

intracranial hypertension without ventricular dis-tention undergo cisternographic assessment. Inmy practice, 28 of 31 patients have been shown to

have communicating hydrocephalus and havebeen treated with lumboperitoneal shunts employ-ing valve systems. In chronically shunted patients,

ventricular distention at the time of shunt failureis probably sufficient to recommend ETV (Fig. 5).

Summary

ETV using contemporary instrumentation has

been used for more than 50 years, but its use hasbecome widespread only in the last 10 to 15 years.Randomized prospective trials comparing ETVwith shunts are needed before definitive state-

ments can be made about the role of the former inmanaging the many forms of hydrocephalus. Theabsolute and relative contraindications for the use

of ETV in the management of hydrocephalus areshown in the Box 1 on this page. It is importantnot to presume that a specific radiographic or

clinical feature would prevent a patient fromresponding to this rather new procedure withouttesting the hypothesis. Patients should be given asmuch information as possible regarding the risks

and benefits of ETV so they can participate in thedecision-making process.

When should the role of ETV in the manage-ment of hydrocephalus be discussed with a pa-

tient? At the initial diagnosis of hydrocephalus,the patient or family should be informed of thispotential alternative to shunting for the manage-

ment of hydrocephalus. I also believe that patientswith working shunts who are being followedchronically should be informed about ETV as

a potential treatment option when their shuntfails. Every shunt failure or infection should beviewed as an opportunity to explore the possibility

that the patient could become shunt independent.

References

[1] Baskin JJ, Manwaring KH, Rekate HL. Ventricular

shunt removal: the ultimate treatment of the slit

ventricle syndrome. J Neurosurg 1998;88:478–84.

[2] Dandy W, Blackfan K. An experimental and

clinical study of internal hydrocephalus. JAMA

1913;61:2216–7.

[3] Dandy W, Blackfan K. Internal hydrocephalus. An

experimental, clinical and pathological study. Am J

Dis Child 1914;8:406–82.

[4] Ransohoff J, Epstein F. Proceedings: avoidance of

shunt dependency. J Neurol Neurosurg Psychiatry

1975;38:410–1.

[5] Olivero WC, Rekate HL, Chizeck HJ, et al.

Relationship between intracranial and sagittal sinus

pressure in normal and hydrocephalic dogs. Pediatr

Neurosci 1988;14:196–201.


[6] Rekate HL. Circuit diagram of the circulation of

cerebrospinal fluid. 1989. Pediatr Neurosurg 1994;

21:248–52.

[7] Rekate H. Neurosurgical management of the

newborn with spina bifida. In: Rekate H, editor.

Comprehensive management of spina bifida. Boca

Raton, FL: CRC Publishers; 1991. p. 1–26.

[8] Gilles F, Davidson R. Communicating hydroceph-

alus with deficient dysplastic parasagittal arach-

noidal granulations. J Neurosurg 1971;35:421–6.

[9] Barlow CF. CSF dynamics in hydrocephalus—with

special attention to external hydrocephalus. Brain

Dev 1984;6:119–27.

[10] Akaboshi I, Ikeda T, Yoshioka S. Benign external

hydrocephalus in a boy with autosomal dominant

microcephaly. Clin Genet 1996;49:160–2.

[11] Rekate H, Nulsen FE, Mack H, et al. Establishing

the diagnosis of shunt independence. Monogr

Neural Sci 1982;8:223–6.

[12] Epstein F, Hochwald GM, Ransohoff J. Neonatal

hydrocephalus treated by compressive head wrap-

ping. Lancet 1973;1:634–6.

[13] Karahalios DG, Rekate HL, Khayata MH, et al.

Elevated intracranial venous pressure as a universal

mechanism in pseudotumor cerebri of varying

etiologies. Neurology 1996;46:198–202.

[14] Pare LS, Batzdorf U. Syringomyelia persistence

after Chiari decompression as a result of pseudo-

meningocele formation: Implications for syrinx

pathogenesis: report of three cases. Neurosurgery

1998;43:945–8.

[15] Nugent G, Al-Mefty O, Chou S. Communicating

hydrocephalus as a cause of aqueductal stenosis.

J Neurosurg 1979;51:812–8.

[16] Francis PM, Beals S, Rekate HL, et al. Chronic

tonsillar herniation and Crouzon’s syndrome.


[17] Saint-Rose C, LaCombe J, Pierre-Kahn T, et al.

Intracranial venous sinus hypertension: cause or

consequence of hydrocephalus in infants? J Neuro-

surg 1984;60:727–31.

[18] Pierre-Kahn A, Hirsch JF, Renier D, et al.

Hydrocephalus and achondroplasia. A study of 25

observations. Childs Brain 1980;7:205–19.

[19] Cinalli G, Renier D, Sebag G, et al. Chronic

tonsillar herniation in Crouzon’s and Apert’s

syndromes: the role of premature synostosis of the

lambdoid suture. J Neurosurg 1995;83:575–82.

[20] Nishihara T, Hara T, Suzuki I, et al. Third

ventriculostomy for symptomatic syringomyelia

using flexible endoscope: case report. Minim In-

vasive Neurosurg 1996;39:130–2.

[21] Engel M, Carmel P, Chutorian A. Increased

intraventricular pressure without ventriculomegaly

in children with shunts: ‘‘normal volume’’ hydro-

cephalus. Neurosurgery 1979;5:549–52.

[22] Rekate HL, Wallace D. Lumboperitoneal shunts in

children. Pediatr Neurosurg 2003;38:41–6.

[23] Jones RF, Kwok BC, Stening WA, et al. Neuro-

endoscopic third ventriculostomy. A practical

alternative to extracranial shunts in non-communi-

cating hydrocephalus. Acta Neurochir Suppl

(Wien) 1994;61:79–83.

[24] Teo C, Jones R. Management of hydrocephalus by

endoscopic third ventriculostomy in patients with

myelomeningocele. Pediatr Neurosurg 1996;25:

57–63.

[25] Buxton N, Macarthur D, Mallucci C, et al. Neuro-

endoscopic third ventriculostomy in patients less

than 1 year old. Pediatr Neurosurg 1998;29:73–6.

[26] Oi S, Shimoda M, Shibata M, et al. Pathophysiol-

ogy of long-standing overt ventriculomegaly in

adults. J Neurosurg 2000;92:933–40.

[27] Larsson A, Stephensen H, Wikkelso C. Adult

patients with ‘‘asymptomatic’’ and ‘‘compensated’’

hydrocephalus benefit from surgery. Acta Neurol

Scand 1999;99:81–90.

[28] Pollack IF, Pang D, Albright AL. The long-term

outcome in children with late-onset aqueductal

stenosis resulting from benign intrinsic tectal

tumors. J Neurosurg 1994;80:681–8.

[29] Di Rocco C, Di Trapani G, Maira G, et al.

Anatomo-clinical correlations in normotensive hy-

drocephalus. Reports on three cases. J Neurol Sci

1977;33:437–52.

[30] Mitchell P, Mathew B. Third ventriculostomy in

normal pressure hydrocephalus. Br J Neurosurg

1999;13:382–5.

[31] Ransohoff J, Shulman K, Fishman R. Hydroceph-

alus: a review of etiology and treatment. J Pediatr

1960;56:399–411.

[32] Cutler R, Page L, Galicich J. Formation and

absorption of cerebrospinal fluid in man. Brain

1968;91:707–20.

[33] Hyde-Rowan MD, Rekate HL, Nulsen FE. Re-

expansion of previously collapsed ventricles: the slit


[34] Rekate HL. Classification of slit-ventricle syn-

dromes using intracranial pressure monitoring.



Techniques of endoscopic third ventriculostomyDouglas Brockmeyer, MD

Primary Children’s Medical Center, 100 North Medical Drive, Suite 2400,

Salt Lake City, UT 84113–1100, USA

Modern techniques of endoscopic third ven-

triculostomy (ETV) are based on the concept ofestablishing a natural conduit for cerebral spinalfluid (CSF) flow through the floor of the third

ventricle. Through the years, a wide variety oftechniques have been used as a means to thisend and have included both open and closed

approaches. The relatively recent application ofendoscopic technology to intraventricular surgeryhas allowed neurosurgeons to perform third

ventriculostomies in a minimally invasive fashion,however. Advances in third ventriculostomy tech-nique have been based on a detailed under-standing of third ventricular anatomy, surgical

trajectories, and improved instrumentation. Thegoal of this article is to discuss these issues indetail and to point out the relevant risks and

known complications associated with them.

History

An understanding of the current state of ETVis not complete without an appreciation of itshistory. Briefly, in 1922, Walter Dandy originated

the concept of third ventriculostomy by perform-ing a craniotomy and fenestrating the laminaterminalis for the treatment of hydrocephalus [1].

This was quickly followed by Mixter [2], who, in1923, used a urologic endoscope to puncture thefloor of the third ventricle, thus ushering in the era

of ETV at an early stage. In 1936, Stokey andScarrf [3] and, in 1951, Scarrf [4] described theirown procedures for third ventriculostomy. In the1970s and 1980s, a host of authors described

various techniques for third ventriculostomy, both


1042-3680/04/$ - see front matter � 2004 Elsevier Inc. All r

doi:10.1016/S1042-3680(03)00066-4

open and closed [5–12]. Notable among these

series is the Toronto experience as reported byHoffman et al [13], which describes a percutaneousthird ventriculostomy technique in the man-

agement of noncommunicating hydrocephalus.Significant experience with closed stereotactictechniques was reported by Kelly in 1991 [14].

During this period, neurosurgeons began takingadvantage of smaller and smaller endoscopes;eventually, the routine use of fiberoptic or rod-

lens ETV was accepted. Multiple studies werepublished in the 1990s reflecting considerablesuccess with endoscopic techniques [14–27]. Thecurrent state of the art in ETV obviously reflects

significant new concepts brought out by thesepioneering neurosurgeons.

Relevant anatomy

A brief review regarding the anatomy relevant

to ETV is useful. The recognition of criticallandmarks and structures is vital to the overallsuccess of ETV. In addition, complications may

be avoided when important anatomy is identifiedearly and respected throughout the procedure.

Choroid plexus

The choroid plexus lies along the floor of thelateral ventricle in the choroidal fissure and isoriented in an anterior/posterior direction. Earlyrecognition of the choroid plexus within the

lateral ventricle is a powerful navigational toolbecause its anterior extent leads to the foramen ofMonro and the third ventricle. Thus, if one’s

initial endoscopic trajectory does not lead directlyto the foramen of Monro, the choroid plexus canact as a ‘‘road map’’ to lead one to the proper site

quickly and efficiently. Even in patients with

ights reserved.


52 D. Brockmeyer /Neurosurg Clin N Am 15 (2004) 51–59

distorted ventricular anatomy, such as those withspina bifida, the choroid plexus remains in thechoroidal fissure and still leads to the third

ventricle. In patients with large ventricles, thispoint is not much of an issue, but when theventricular size is small, early recognition of thechoroid plexus is sometimes the only navigational

guide a neuroendoscopist has.

Fornix

The fornix forms the superior and anterior

margin of the foramen of Monro. It is importantthat every effort is made to avoid its injury duringendoscopic neurosurgery. Because of its location,however, it can easily be injured during passage of

the endoscope from the lateral ventricle into thethird ventricle. The risk of injury is multiplied whenmultiple passes through the foramen of Monro are

made. For that reason, the number of passes of theendoscope through the foramen of Monro shouldbe kept to an absolute minimum. In addition,

it is important to recognize that the endoscope’scamera port is only a small percentage of the cross-sectional area of the endoscope tip. For that

reason, it is easy to assume that if one passes theendoscope easily through the foramen of Monro,the fornix is not injured. This may not be the case,however, because the edge of the endoscope tip

near the light source or working chamber mayinadvertently cause injury. Potential solutions for

this problem are to be aware of the orientation ofyour endoscope’s camera in relation to the cross-sectional area of the tip and to make appropriate

adjustments in the approach trajectory. In addi-tion, if one guides the scope in a handheld fashion,exquisite tactile feedback can be obtained and thesurgeon could be alerted to ‘‘hanging up’’ the edge

of the endoscope on the fornix.

Hypothalamus

The paired hypothalami form the lateral wallsof the third ventricle. The supraoptic and para-ventricular arcuate nuclei are the structures thatare at the most risk during third ventriculostomy

(Fig. 1). Injury to these structures may havesignificant endocrinologic consequences. Althoughsome evidence suggests that the supraoptic

nucleus is related mainly to vasopressin (antidiu-retic hormone) and the paraventricular nucleusto oxytocin, both hormones are found in each

nucleus. Therefore, surgical trajectories to thethird ventricular floor must be planned with theidea that hypothalamic damage must not occur.

Although surgical trajectories are presented ina later section, this concept is important whendiscussing hypothalamic anatomy. These issuesare especially important when spina bifida pa-

tients with distorted hypothalamic anatomy un-dergo ETV.

Fig. 1. Hypothalamic nuclear anatomy adjacent to the third ventricle. Note that the supraoptic and paraventricular

nuclei are in close proximity to the third ventricle.

53D. Brockmeyer /Neurosurg Clin N Am 15 (2004) 51–59

Third ventricular floor

The floor of the third ventricle is essentiallya thinned out portion of the hypothalamus andcan potentially have functioning hypothalamic

nuclei (eg, supraoptic and paraventricular nuclei)within it. As noted previously, this consideration isespecially important in patients with spina bifida.The traditional boundaries of the third ventricular

floor include the mamillary bodies posteriorly, thewalls of the hypothalamus laterally, and the opticchiasm/infundibular recess anteriorly. Within

this area is a relative ‘‘safe zone’’ where the thirdventricular floor may be entered (Fig. 2). Thisconsists of the area just anterior to the midway

point between the mamillary bodies and the in-fundibular recess. If one penetrates the third ven-tricular floor posterior to this point, the basilarartery tip or proximal portion of the posterior

communicating artery may be encountered. If thefloor is penetrated anterior to this point, theclivus is encountered . Obviously, if a choice has to

be made, entering slightly more anterior to themidpoint is the safer than entering more posteriorto the midpoint. Usually, within the midportion

of a thinned-out third ventricular floor, there isa translucent bluish-appearing area that corre-sponds to a safe zone in which to make the initial

opening. How the opening is actually made isdiscussed in later in this article.

Fig. 2. Line drawing depicting the floor of the third

ventricle and the ‘‘endoscopic safe zone,’’ where initial

dissection during endoscopic third ventriculostomy

should begin.

Lillequist’s membrane

Lillequist’s membrane is an arachnoid planethat contains within it the basilar artery complexand separates the posterior fossa arachnoid

cisterns from the suprasellar cisterns. Once a neu-roendoscopist penetrates through the floor of thethird ventricle, Lillequist’s membrane is frequentlyencountered and hides the basilar artery complex

and prepontine cistern from view. It is importantthat Lillequist’s membrane be opened to havea successful third ventriculostomy. The goal is to

communicate the fluid of the third ventricle to theprepontine cistern area. A clear view of the basilarartery complex, pons, pontine perforators, and

clivus must be obtained before an ETV can beconsidered successful. Failure to recognize thisanatomic point may lead to endoscopic failures.

Endoscopic approaches to the third ventricle

Trajectories

For a typical patient with enlarged ventriclesundergoing ETV for the first time, a standardizedtrajectory to the third ventricle should be used.One such trajectory that yields excellent results

consists of an entry site at the intersection of thecoronal suture and the midpupillary line, with thetrajectory of the endoscope slightly medial and

oriented in line with the external auditory meatusin an anterior/posterior direction (Fig. 3A). Thisapproach yields an endoscopic trajectory toward

the foramen of Monro and into the floor ofthe third ventricle. The consistency and excellentresults of this approach cannot be overempha-sized.

A standardized approach is particularly im-portant when previously shunted patients becomecandidates for ETV. The existing shunt entry site

or ventricular catheter trajectory is sometimespositioned so that a safe approach to the thirdventricular floor is not feasible. In this situation,

a separate burr hole, and sometimes a separateincision, must be made in the standard location,and the standard approach should be employed. It

is also possible in this situation to use a flexibleendoscope to navigate into the third ventricle. Inthis author’s experience, however, orientationmay be difficult, and the acute curve of the scope

may injure brain structures or preclude passinginstruments safely down the working chamber.

The endoscope trajectory can also be changed

based on the preoperative ventricular size. If the


ventricles are generous, the traditional trajectorydescribed previously may be used. If the ventriclesare small, however, the burr hole placement

should be slightly medial to the midpupillary line,facilitating an easier approach into a narrow thirdventricle (see Fig. 3B).

Fig. 3. Line diagrams depicting trajectories to third

ventricle. (A) Traditional trajectory toward the third

ventricle floor beginning at the midpupillary line. (B)

Modified trajectory because of the small ventricle size.

Note that the entry point is moved slightly medial

compared with the standard trajectory.

Type of endoscope to use

There are two basic types of neuroendoscopesthat are available: rod-lens and fiberoptic. Thesescopes may be further subdivided into rigid,

semirigid, and flexible. Understandably, there isno single endoscope that is adaptable enoughto be superior in all applications and situations.Therefore, the neuroendoscopist must be flexible

in his or her choice of endoscopes. For instance, ifone encounters an adolescent patient with late-onset aqueductal stenosis and large ventricles,

a good argument could be made for being asminimally invasive as possible, and a small, 1-mm,semirigid endoscope could be used for the entire

procedure. In that case, the tip of the scope is usedto perform the fenestration. The optics of the1-mm endoscope are sufficient to allow positiveidentification of critical structures and rapid and

safe performance of the third ventriculostomy.A school-aged spina bifida patient with dis-

torted anatomy in the third ventricular floor

requires maximum visibility and adaptability,however. This situation requires either a rod-lenssystem or one of the larger fiberoptic scopes. In

either case, anticipating the anatomy of thepatient and maximizing visibility must be weighedagainst potential risks. For the experienced endo-

scopist, the proper endoscope choice may varyfrom patient to patient.

Techniques of endoscopic third ventriculostomy

In discussing techniques of endoscopic thirdventriculostomy, it is logical to begin with the

simplest technique available and then to proceedtoward more complex or sophisticated proce-dures. The list that is included here is by no meansmeant to be all-inclusive; however, it is meant to

cover most of the more commonly used tech-niques of ETV and serves primarily as a guide orreference for further reading.

Endoscope tip

Perhaps the easiest and most straightforwardway to perform the fenestration for third ven-

triculostomy is using the tip of the endoscopeitself (Fig. 4). Multiple case series have docu-mented its safety and efficacy over time [15,19–21,26,28,29]. In essence, the tip of the endoscope is

used as a dissecting tool to penetrate the floor ofthe third ventricle and Lillequist’s membrane.Obviously, the smaller the tip of the endoscope,

the easier the dissection becomes. This technique’s


Fig. 4. (A, B) Line drawing depicting the use of the endoscope tip to perform the fenestration through the floor of the

third ventricle.

critical drawback is that during the moments ofdissection, one’s visualization is completely ob-scured by the tissue in front of the scope. The lackof visualization is made up for by tactile feedback

during manual opening of the third ventricularfloor. Sometimes, a ‘‘windshield wiper’’ techniquemay be used to gently dissect the floor of the third

ventricle and Lillequist’s membrane, much likearachnoid dissection in other parts of the brain.With practice, this technique is a safe and effective

way of performing a large opening of the floor ofthe third ventricle. In addition, the scope tip canbe placed into the prepontine cistern to inspect thesurrounding anatomy. Obviously, this technique

has a rather steep learning curve, and the risk ofbasilar artery injury is a very real threat duringthis procedure. Only endoscopists confident with

their dissecting technique using the tip of theendoscope should employ it.

A variation of this technique was reported by

Wellins et al [30]. They describe preloadinga ventricular catheter over the endoscope. Oncethe ETV is performed, the ventricular catheter is

used to dilate the opening.

Dissection/balloon dilatation

Perhaps the most widespread and safesttechnique to open the floor of the third ventricle

is to pass some type of semiblunt dissector downthe working channel of the endoscope first andthen to open a small hole in the floor of the third

ventricle (Fig. 5A, B). Next, the dissector iswithdrawn, and a 2- or 3-French balloon-tippedcatheter (or similar device, such as a wire stoneextractor [30]) is passed through the endoscope

and through the initial fenestration in the floor ofthe third ventricle (see Fig. 5C). The balloon is

slowly inflated at the site of the fenestration toenlarge the opening gradually and gently (see Fig.5D) [15,20,25]. Once the balloon has been openedto its widest diameter, it is deflated and removed

and the endoscope is placed through the openingto inspect and confirm that Lillequist’s membranehas been opened (see Fig. 5E). If not, the pro-

cedure is repeated until Lillequist’s membrane isopen. Several endoscope manufacturers supplysmall dissecting tools that are ideal for making

the initial opening. The author’s preferred tech-nique is to use the tip of an endoscopic bipolarcoagulator.

This technique, used either with a rod-lens or

fiberoptic endoscope, is safe, simple, and effective.The floor of the third ventricle is under directvisualization during the entire act of fenestration.

For these reasons, the dissector/balloon dilatationtechnique and its variations deserve to be atthe top of the list for recommended fenestration

techniques. One notable exception is placinga fully inflated balloon-tipped catheter all theway through the third ventricle and pulling it back

through the membrane in a fully inflated fashion.It is quite easy to injure the hypothalamicnuclei or to cause significant bleeding with thistechnique.

Saline jet

Saline jet irrigation has been used in perform-ing ETV and cyst wall fenestration. The technique

involves deploying and then directing a strong jetof saline irrigation in the region of the proposeddissection. The fluid forced from the saline jetatraumatically performs the dissection, at least

theoretically. Although this technique has itsmerits and probably deserves more widespread


Fig. 5. (A–E) A series of line drawings depicting the use of an endoscope, blunt dissector, and Fogarty catheter to

perform endoscopic third ventriculostomy (see text for details).

use, a saline jet apparatus is not readily availableto most operating rooms and would need to bespecifically requested. Further work and investi-

gation of this technique are probably warrantedbefore more widespread use is recommended.

Saline torch

Manwaring [31] has previously described theuse of the saline torch. As a result of the fact thata potentially dangerous form of active energy is

used during the fenestration, most neuroendo-scopic surgeons have moved away from thistechnique toward blunt dissection of the third

ventricular floor.

Laser energy

The application of laser energy or some other

type of active energy source to perform ETV ismentioned here only to discourage its use [32].When using a laser or other type of thermal


energy source to perform the fenestration of thethird ventricular floor, absolutely no tactile feed-back is available and spread of the energy mayeasily injure or rupture critical neurovascular

structures. Several significant complications, somereported and some unreported, have occurredusing this type of technique [33]. Until safer

sources of energy are developed to provide forthird ventricular fenestration, these types oftechniques should be avoided.

Stereotactic approaches

Over the years, several authors have reportedclosed and endoscopic stereotactic approachesto third ventriculostomy [11,13,14,19]. One such

approach describes guiding a rigid rod-lens scopethrough the lateral and third ventricle using apredetermined stereotactic trajectory [19]. Al-

though this type of technique has its merits, itdoes not allow much room for flexibility, a keyfactor that may represent a critical margin of

safety in ETV. In addition, because of the extraexpense of the preoperative stereotactic scan,extra operative time for stereotactic registration,

and extra expense for the stereotactic apparatus,using a stereotactic technique is most likely nota cost-effective way to perform this procedure.Further cost-benefit analysis of this procedure

would be important.

Doppler ultrasound

In 1998, Schmidt [34] described the use of

a micro-Doppler technique to identify the basilarapex before performing the fenestration in thethird ventricular floor. In his report, two third

ventricular floors were mapped out before fenes-tration with a high degree of accuracy ofpredicting the position of the basilar artery.Although perhaps not necessary in all cases, this

type of technique may be extremely useful inperforming third ventriculostomies in patientswith thick opaque floors or in those patients with

distorted anatomy (eg, spina bifida).

Complications

Vascular

The most important vascular complication toavoid during ETV is injury to the basilar arteryand its nearby branches. Two reports have

documented injury to the basilar artery or the

posterior cerebral artery by the tip of the neuro-endoscope or a laser [33,35]. In one case,a pseudoaneurysm developed at the site of arterialinjury, which later required further surgery and

trapping [35]. Obviously, a careful look at thepreoperative MRI to identify the position of thebasilar artery in the prepontine cistern is extreme-

ly helpful in potentially avoiding vascular injury.In addition, placing the initial third ventricularfloor fenestration in the safe zone also may

prevent an arterial catastrophe.If major arterial injury does occur, several

techniques may help in avoiding disaster. One is

the preplacement of a Sheath dilator in the lateralventricle, through which the endoscope is passed.If arterial injury does occur, expression of bloodthrough the Sheath dilator may help to decom-

press the intracranial compartment during theepisode of rapid bleeding before arterial spasm. Inaddition, the ventricular cavity should be irrigated

copiously with saline irrigation for perhaps aslong as 20 or 30 minutes. Placement of an externalventricular drain and performance of a postoper-

ative CT scan are mandatory. It is important toremember that although major arterial injuriesmay occur, most can be avoided by careful

planning and meticulous technique.This type of major arterial bleeding should be

distinguished from the small amount of bleedingthat frequently occurs from the floor of the third

ventricle once the fenestration is performed. Thisbleeding is typically capillary in nature andsubsides after 1 to 2 minutes of continuous irriga-

tion. Further bleeding almost always requiresonly more patience and irrigation. If the CSFis not clear enough to see with an endoscope, the

procedure should be abandoned, an externalventricular drain placed, and an emergency headCT scan obtained.

Occlusive hydrocephalus

The occurrence of acute occlusive hydroceph-alus by blockage of the foramen of Monro with

the endoscope during continuous irrigation hasbeen described previously [36]. Recognition andprevention of this potentially lethal complicationare important. The size of the endoscope versus

the size of the foramen of Monro should bejudged during surgery; if the endoscope plugs theforamen of Monro, the irrigation should be shut

off. This problem should be in the endoscopist’smind if sudden vasomotor instability occursduring ETV.


Endocrine

Because the third ventricular floor is in directcontinuity with the hypothalamus, endocrinedysfunction may occur after ETV. In 1996, Teo

et al [37] reported four endocrine complications in55 ETVs. Their complications included diabetesinsipidus, increase in appetite, loss of thirst, andamenorrhea. The first two complications resolved

completely over time. In a separate report, Teoand Jones [38] described transient endocrinecomplications in 2 of 69 spina bifida patients

who underwent ETV. In contrast, reports of otherlarge ETV case series [15,25] have reported noendocrine complications.

Obviously, any patientwhoundergoesETV is atpotential risk for damaging critical neuroendocrinestructures, such as the supraoptic or paraventric-ular nuclei. Although it is difficult to prove with the

existing data, it makes sense that patients who areat themost risk for endocrine complications duringETV are probably those with spina bifida or a thick

third ventricular floor. As expected, the treatmentof an endoscope-caused endocrine complicationis supportive and hopefully transient. Therefore,

careful attention to the regional anatomy as well aswillingness to stop a procedure if the anatomy isunfavorable may help to avoid this complication.

References

[1] Dandy WE. An operative procedure for hydroceph-

alus. Johns Hopkins Hosp Bull 1922;33:189–90.

[2] Mixter WJ. Ventriculoscopy and puncture of the

floor of the third ventricle. Boston Med Surg J

1923;188:277–8.

[3] Stokey B, Scarrf JE. Occlusion of aqueduct of

Sylvius by neoplastic processes with a rational

surgical treatment for relief of resultant obstructive

hydrocephalus. Bull Neurol Inst 1936;5:348–77.

[4] Scarrf JE. Treatment of obstructive hydrocephalus

by puncture of the lamina terminalis and floor of

the third ventricle. J Neurosurg 1951;8:204–13.

[5] Avman N, Kanoplat Y. Third ventriculostomy by

microtechnique. Acta Neurochir Suppl (Wien) 1979;

28:588–95.

[6] Brocklehurst G. Trans-callosal third ventriculo-

chiasmatic cisternostomy: a new approach to

hydrocephalus. Surg Neurol 1974;2:109–14.

[7] Guiot G. Ventriculocisternostomy for stenosis of

the aqueduct of Sylvius. Puncture of the floor of the

third ventricle with a leucotome under television

control. Acta Neurochir (Wien) 1973;28:264–89.

[8] Jaksche H, Lowe F. Burrhole third ventriculocister-

nostomy. An unpopular but effective procedure for

the treatment of certain forms of occlusive hydro-

cephalus. Acta Neurochir (Wien) 1986;79:48–51.

[9] Natelson SE. Early third ventriculostomy in myelo-

meningocele infants—shunt independence? Childs

Brain 1981;188:277–8.

[10] Reddy K, Fewer HD, West M, Hill NC. Slit ventricle

syndromewithaqueduct stenosis: thirdventriculostomy

as definitive treatment. Neurosurgery 1988;23:756–9.

[11] Sayers M, Kosnik E. Percutaneous third ventricu-

lostomy: experience and technique. Childs Brain

1976;2:24–6.

[12] Vries J. An endoscopic technique for third ventri-

culostomy. Surg Neurol 1978;9:165–8.

[13] Hoffman HJ, Harwood-Nash D, Gilday DL.

Percutaneous third ventriculostomy in the manage-

ment of non-communicating hydrocephalus. Neu-

rosurgery 1980;7:313–21.

[14] Kelly PJ. Stereotactic third ventriculostomy in

patients with nontumoral adolescent/adult onset

aqueductal stenosis and symptomatic hydrocepha-

lus. J Neurosurg 1991;75:865–73.

[15] Brockmeyer DB, Abtin K, Carey L, Walker ML.

Endoscopic third ventriculostomy: an outcome

analysis. Pediatr Neurosurg 1998;28:236–40.

[16] Buxton N, Macarther D, Mallucci C, Punt J,

Vloeberghs M. Neuroendoscopic third ventriculos-

tomy in patients less than 1 year old. Pediatr

Neurosurg 1998;29:73–6.

[17] Dalrymple S, Kelly P. Computer-assisted stereotac-

tic third ventriculostomy in the management of

noncommunicating hydrocephalus. Stereotact Funct

Neurosurg 1992;59:105–10.

[18] Drake J. Ventriculostomy for treatment of hydro-

cephalus. Neurosurgery 1993;4:657–66.

[19] Goodman RR. Magnetic resonance imaging-

directed stereotactic endoscopic third ventriculos-

tomy. Neurosurgery 1993;32(6):1043–7.

[20] Goumnerova L, Frim D. Treatment of hydroceph-

alus with third ventriculocisternostomy: outcome

and CSF flow patterns. Pediatr Neurosurg 1997;27:

149–52.


third ventriculostomy. Neurosurgery 1990;26:86–91.

[22] Kunz U, Goldman A, Bader C, Waldbaur H,

Oldenkott P. Endoscopic fenestration of the 3rd

ventricular floor in aqueductal stenosis. Minim

Invasive Neurosurg 1994;37:42–7.

[23] Oka K, Yamamoto M, Ikeda K, Tomonaga M.

Flexible endoneurosurgical therapy for aqueductal

stenosis. Neurosurgery 1993;33:236–43.

[24] Rieger A, Rainov NG, Sanchin L, Schopp G,

Burkett W. Ultrasound-guided endoscopic fenestra-

tion of the third ventricular floor for non-commu-

nicating hydrocephalus. Minim Invasive Neurosurg

1996;39:17–20.

[25] Sainte-Rose C, Chumas P. Endoscopic third ven-

triculostomy. Tech Neurosurg 1996;1:176–84.




57–63.



versus cerebrospinal fluid shunt as a first procedure

in pediatric hydrocephalus. Pediatr Neurosurg 1999;

30:11–5.

[28] Grant JA, McLone DG. Third ventriculostomy:

a review. Surg Neurol 1997;47:210–2.

[29] Wellins JCI, Bagley CA, George TM. A simple and

safe technique for endoscopic third ventriculocis-

ternostomy. Pediatr Neurosurg 1999;30:219–23.

[30] Wong TT, Lee LS. A method of enlarging the

opening of the third ventricular floor for flexible

endoscopic third ventriculostomy. Childs Nerv Syst

1996;12:396–8.

[31] Manwaring K. Endoscopic ventricular fenestration.

In: Manwaring K, Crone K, editors. Neuroendo-

scopy.NewYork:MaryAnnLiebert; 1992. p. 79–90.

[32] Wood FA. Endoscopic laser third ventriculostomy.

N Engl J Med 1993;329:207–8.

[33] McLaughlin MR, Wahlig JB, Kaufmann AM,

Albright AL. Traumatic basilar aneurysm after

endoscopic third ventriculostomy: a case report.

Neurosurgery 1997;41:1400–3.

[34] Schmidt R. Use of microvascular Doppler probe to

avoid basilar artery injury during endoscopic third

ventriculostomy. Technical note. J Neurosurg 1999;

90:156–7.

[35] Abtin K, Thompson BG, Walker ML. Basilar

artery perforation as a complication of endoscopic

third ventriculostomy. Pediatr Neurosurg 1998;28:

35–41.

[36] Handler MH, Abbott R, Lee M. A near-fatal

complication of endoscopic third ventriculostomy:

case report. Neurosurgery 1994;35:525–7.

[37] Teo C, Rahman S, Boop FA, Cherny B. Compli-

cations of endoscopic neurosurgery. Childs Nerv

Syst 1996;12(5):248–53.



myelomeningocele. Pediatr Neurosurg 1996;25(2):

57–63.



Complications of third ventriculostomyMarion L. Walker, MD

Primary Children’s Medical Center, 100 North Medical Drive, Salt Lake City, UT 84113, USA

Improvements in instrumentation and growingexperience with neuroendoscopy have made it

possible to perform endoscopic third ventriculos-tomy (ETV) safely in a selected group of patients.The complication rate for ETV now approachesthe rate for ventriculoscopy in general, approxi-

mately 6% to 8%, and is similar to the expectedinfection rate of shunts [1,2]. Although theexposure to significant neurologic complications

is higher than the risk of a single shunt operation,the possibility of achieving long-term shunt in-dependence with ETV is preferable in good-risk

patients when compared with the cumulativemorbidity of multiple shunt procedures [3–6].The collective morbidity for a ventriculoperitonealshunt is difficult to quantify; however, a recent

prospective multicenter study of children under-going their initial shunt insertion found that 31%had shunt obstruction, 3% had overdrainage, and

8% had infection within 1 year [1]. There is a 60%failure rate at 2 years.

Many of the complications of ETV are

transient in nature. Skilled management of intra-operative problems can keep these complicationslimited to the intraoperative or perioperative

period. At least one author has suggested catego-rizing complications as ‘‘clinically insignificant’’and ‘‘clinically significant’’ [2]. Although thissystem may reflect the long-term outcome of the

patient, the most important consideration, it doesnot emphasize the potentially devastating con-sequences of seemingly minor intraoperative

events. This article seeks to define the mostcommon potential complications of ETV and toidentify factors in the pre-, intra-, and post-

operative periods that can minimize their occur-rence or effects.


doi:10.1016/S1042-3680(03)00070-6

Complications of endoscopic third ventriculostomy

Cerebrospinal fluid leak

Although rarely reported, cerebrospinal fluid

(CSF) leakage can delay wound healing andincrease the risk of infection. The possibility ofa CSF leak may be minimized by using thesmallest appropriate endoscope (especially in

patients with large ventricles) and by minimizingthe dural opening. A layered closure of the scalp isnecessary. More importantly, however, a CSF

leak is frequently a sign of failure of the thirdventriculostomy, and close observation for thispossibility is required.

Pneumocephalus

Often considered a minor or insignificantcomplication, pneumocephalus can delay post-operative recovery and can be associated with

headache, nausea, and vomiting. Entrapment ofair at the time of surgery can interfere with directvisualization of the anatomic landmarks that are

essential to performing a third ventriculostomysafely. This can be minimized by keeping thepatient’s head in a midline anatomic position withthe burr hole at or near the most superior point,

by carefully flushing all irrigation lines of airbubbles, and by irrigating gently while introduc-ing the endoscope. Attention should be paid to

minimizing CSF loss, especially in the early stagesof the procedure. Decompression of the ventric-ular system too rapidly can contribute to the

formation of a subdural hematoma. Nitrous oxideshould not be used for anesthesia during ETVbecause of the potential for formation of tension

pneumocephalus.

Ventriculitis

Fever after ETV is not uncommon. In most

cases, it is caused by residual blood in the

ghts reserved.


62 M.L. Walker /Neurosurg Clin N Am 15 (2004) 61–66

ventricular system. In this setting, fever is self-limited and lasts 24 to 48 hours. In some cases,postoperative fever has been attributed to raising

the temperature of the CSF in the third ventriclewhen employing cautery or laser energy, indirectlyheating the hypothalamus. Fever may also heraldinfection and should be monitored closely with

analysis and culture of the CSF [7,8]. Fukuharaet al [9] report multiple cases in which preexistingventricular shunt hardware was found to harbor

infection after ETV. We believe that it is im-portant to remove all shunt hardware after ETVwhenever possible.

Careful and thorough sterilization of theendoscopic equipment and the use of perioper-ative antibiotics can minimize the occurrence ofinfection. Prompt diagnosis and treatment of

ventriculitis are essential.

Subdural hematoma

Subdural hematoma has been reported afterETV [9]. Significant ventricular dilatation witha thin cortical mantle is a risk factor for subduralhemorrhage. Efforts should be made to avoid

rapid drainage of large quantities of CSF, and lostCSF should be replaced with lactated Ringer’ssolution.

Injury to periventricular structures

The floor of the third ventricle is not a mem-brane but a part of the hypothalamus. It isapparently safe to puncture the floor of the third

ventricle but only when it is thinned as a result ofthe pressure of hydrocephalus. Amenorrhea, di-abetes insipidus, loss of thirst, and increasedappetite have been reported after ETV [2,10,11].

Irrigation solution can also be the cause ofcomplications. Generous irrigation with normalsaline solutions has been associated with distur-

bance of electrolytes and hypothalamic dysfunc-tion. Late arousal and postoperative confusionhave also been noted [2]. These complications are

caused by trauma to sensitive hypothalamicstructures comprising the walls and floor of thethird ventricle.

Several strategies can minimize injury to sensi-

tive brain structures in the periventricular region.Careful preoperative evaluation may exclude pa-tients with a narrow third ventricle; some authors

have suggested a minimum third ventricular widthof 7 to 10 mm [12,13]. Third ventriculostomy canbe performed in small third ventricles, but this

increases the risk of injury and should be

attempted only in unusual circumstances and byexperienced endoscopists. As experience hasgrown, the range of patients suitable for third

ventriculostomy has widened. A number ofsurgeons have performed third ventriculostomyin patients with slit ventricle syndrome, with goodresults [14,15]. Modern stereotactic systems and

small endoscopes have added a significant marginof safety to the procedure when it is performed ina narrow ventricular system.

Positioning of the patient is crucial; we advocatekeeping the head in a midline position to simplifythe surgeon’s visualization of the anatomic land-

marks. In a small space, such as the third ventricle,disorientation can quickly lead to injury of thehypothalamic structures or the fornix.

Planning of the incision and placement of the

burr hole also play important roles in avoidingparenchymal injury; a narrow third ventricledemands a more medially placed burr hole, with

a slightly more vertical trajectory (Fig. 1).Knowledge of the relevant anatomy is crucialbecause it allows the trajectory of the endoscope

to be visualized before surgery and the procedureto be mentally ‘‘rehearsed’’ before making theincision. Once the patient is draped, the surgeon

loses access to most of the external landmarks thatdefine the trajectory of the endoscope andreorientation may be difficult.

During surgery, emphasis must be placed on

controlled efficient movement of the endoscopewith constant identification of anatomically rele-vant structures. When using small endoscopes

without a working channel, we place a peel-away

Fig. 1. An artist’s illustration of the proper approach

angle for endoscopic third ventriculostomy. The burr

hole is made at the coronal suture in the midpupillary

line. This provides an appropriate angle of approach

through the foramen of Monro to the floor of the third

ventricle.

63M.L. Walker /Neurosurg Clin N Am 15 (2004) 61–66

sheath through the foramen of Monro underdirect visualization; this prevents injury to thestructures surrounding the foramen during re-moval and reinsertion of the endoscope and

allows passage of instruments into the operativearea without risk of injury to adjacent structures.

Familiarity with the particular endoscope in

use is essential. When conditions conspire to makevisualization or perforation of the floor of thethird ventricle difficult, there should be a low

threshold for abandoning the procedure. This isconsidered minimally invasive surgery; anotherattempt can be made at a later time, or a shunt

can be placed.

Bradycardia/asystole

Several authors have reported transient brady-

cardia or asystole when perforating the floor ofthe third ventricle or while enlarging the ventric-ulostomy [2,7,9,16]. This is presumably caused by

a direct mechanical effect on the hypothalamus oras the effect of irrigation. Irrigation fluid can betrapped in the third ventricle if the endoscope fillsthe foramen of Monro, and the addition of

a seemingly insignificant volume may causea dramatic increase in pressure. Handler et al[16] reported a ventricular arrhythmia progressing

to asystole while irrigating with the endoscope inthe third ventricle. This was believed to beconsistent with expansion of the third ventricle

while the endoscope prevented egress of fluidthrough the foramen of Monro. Prevention ofthese intraoperative complications lies in limiting

the rate of irrigation, especially in the thirdventricle, and in leaving one port of the endoscopeopen for the egress of fluid. Management consistsof ceasing irrigation and withdrawing the endo-

scope until the patient is hemodynamically stable.

Vascular complications

Minor bleeding that is easily controlled withirrigation is common during ETV. Teo et al [2]

reported ‘‘clinically insignificant’’ hemorrhage in 6of 55 endoscopic ventriculostomies. Of 98 cases,Hopf et al [8] reported venous bleeding in 3 cases,arterial bleeding in 1 case, and 1 case in which

there was bleeding from a bridging vein during theopening. Fukuhara et al [9] reported intraven-tricular hemorrhage necessitating placement of

External Ventricular Drainage in 2 of 89 cases.Small vessels may be encountered within the

floor of the third ventricle. This type of bleeding

can often be controlled with patience and

continued irrigation. Brisk bleeding may also beencountered after injury to intraventricular veinsand to perforating arteries and bridging veins inthe interpeduncular cistern. Brisk bleeding re-

quires significant irrigation and may requireabandonment of the procedure. A pathway forCSF to exit is crucial in this circumstance.

Without a doubt, the most important andpotentially devastating complication of ETV isinjury to the basilar, posterior cerebral, or

posterior communicating arteries. After arterialinjury, major hemorrhage, vasospasm, pseudo-aneurysm, and delayed subarachnoid hemorrhage

have been reported, with significant morbidityand mortality (Fig. 2) [17–19]. Although thereported incidence is low, this author is aware ofother as yet unreported cases. Because the

consequences of arterial injury are so profound,it is essential to inform patients and their parentsthat major arterial injury is a real possibility.

Avoidance of hemorrhage is of utmost impor-tance. Careful patient evaluation and selectionmay exclude patients at high risk or at least alert

the surgeon to the specific risk factors. MRI andMR angiography are useful not only to establishthe third ventricular anatomy but may demon-

strate the location of the basilar artery in relationto the floor. During the procedure, the anatomy ofthe floor of the third ventricle must be carefullyanalyzed. The floor may appear opaque, and in

Fig. 2. Anteroposterior view of a basilar artery angio-

gram demonstrating a pseudoaneurysm of the posterior

communicating artery at the junction with the basilar

artery. This aneurysm created by injury to the basilar

artery during endoscopic third ventriculostomy was

successfully treated.


some cases, especially in myelomeningocele pa-tients, the walls of the hypothalamus may bepartially or entirely fused. If an area of thinning of

the floor can be identified, the procedure can beperformed; the risk, however, is increased withabnormal anatomy, and under such conditions,only an experienced endoscopist should attempt

the procedure. A microvascular Doppler probehas been employed successfully in some cases topinpoint the location of the basilar artery through

an opaque floor [20,21].The management of most intraoperative hem-

orrhage consists of irrigation and patience, as

previously noted. Electrocautery is rarely helpfuland is certainly dangerous in cramped ventricularspaces. In certain instances where the source ofbleeding is in the floor of the third ventricle at the

site of fenestration, a balloon catheter can be usedto apply hemostatic pressure to the bleeding tissue.

In our reported case of basilar artery injury,

the patient made a full recovery [17]. We believethis was in large part a result of the fact that wehad positioned the tip of the peel-away sheath

within the third ventricle, allowing the arterialblood to exit the sheath instead of filling theventricular system. An external ventricular cath-

eter should be placed if any significant intraven-tricular hemorrhage has occurred. A cerebralangiogram should be obtained within 1 to 3 daysof any significant hemorrhage to rule out a pseu-

doaneurysm.Intraoperative bleeding is difficult to quantify

as to its significance, and there seem to be differing

opinions in the literature regarding what con-stitutes ‘‘reportable’’ bleeding. It is essential forthe further refinement of this technique that

hemorrhagic complications be reported.

Failure of third ventriculostomy

The success of third ventriculostomy lies in

large part in selecting patients whose CSFphysiology can respond favorably to the pro-cedure. Initially, the procedure was performed

primarily for late-onset aqueductal stenosis, withgood results [22]. As familiarity with the techniquehas grown, the indications have broadened and

patients with noncommunicating hydrocephalusresulting from early-onset aqueductal stenosis,posterior fossa masses, tectal plate tumors, andmyelomeningocele have been included as well as

younger patients [7,23–27].A recent review found that a history of shunt

infection and postoperative meningitis were in-

dependent risk factors for ventriculostomy failureand that nearly half ofETV failures occurredwithin2 weeks of the procedure, with only 3 of 89 cases

failing more than 10 months after ETV [7]. It hasbeen suggested that delayed failures are most oftena result of closure of the ventriculostomyand canbemanaged by repeating the procedure [23].

Basic principles of complication avoidance

Patient selection

As experience has grown and the technique hasbeen refined, third ventriculostomy has beenperformed successfully and safely in patients withobstructive hydrocephalus associated with aque-

ductal stenosis, posterior fossa masses, tectal platetumors, myelomeningocele, and slit ventricle syn-drome (Fig. 3). Relative contraindications, how-

ever, such as postinfectious and posthemorrhagichydrocephalus, still exist. It is up to the surgeon tomatch his or her level of experience, skill, and

confidence with the patient’s anatomy and toinform the patient and family of the risks accord-ingly. For example, it may be appropriate in certain

circumstances to offer ETV to a patient withhydrocephalus secondary to prematurity andintraventricular hemorrhage. The family shouldbe aware that the procedure has a significantly

Fig. 3. An endoscopic view of the floor of the third

ventricle in a patient with myelomeningocele. Almost

complete fusion of the floor of the third ventricle is

present. The mamillary bodies are not identifiable. There

is thinning of the floor in two areas. Successful

endoscopic third ventriculostomy was accomplished

through the most posterior area of thinning.

65M.L. Walker /Neurosurg Clin N Am 15 (2004) 61–66

less chance of success, however. In these patients,preoperative MRI should play a large role indetermining the patient’s suitability for theprocedure.

Proper training

Manipulating the endoscope is a learned skill

and one for which there can be a steep learningcurve. Familiarity with the endoscope, knowledgeof the relevant anatomy, and preoperative visual-

ization of the trajectory are important factors inavoiding complications. We are now in an erawhen computer modeling of surgical procedures ispossible, and this resource will play an increas-

ingly important role in training for ETV.

Positioning

Proper positioning of the patient is crucial tomaintain orientation during the procedure. The

head should be in the midline position and slightlyelevated. The burr hole should be placed with thefinal trajectory in mind, and the anatomy andtrajectory should be visualized before the patient

is covered by surgical drapes (see Fig. 1).

Endoscopic equipment

The endoscope with the smallest diameterappropriate to the patient should be chosen. The

method of perforation of the floor is a subject onwhich there are many differing opinions. The mostpopular method is to use a pointed (but not sharp)instrument to open the floor and then to expand

the opening with a balloon. We firmly believe thatlaser, cautery, and other forms of energy shouldnot be employed because they significantly in-

crease the risk of arterial injury [17,18].

Intraoperative management of complications

Irrigation and patience are the two mostimportant methods for achieving hemostasisduring ETV. The surgeon should also remember

that ETV is a minimally invasive procedure andshould be abandoned if conditions (eg, bleeding,unfavorable anatomy) conspire to make fenestrat-

ing the floor too dangerous. If significant bleedingoccurs, an external ventricular drain should be leftin place at the close of the procedure.

Postoperative management

In our experience, intracranial pressure re-

mains elevated for 24 to 48 hours after ETV,

especially in previously shunted patients whoseshunts have been removed. With the exception ofinfants, we prefer to leave an external ventriculardrain in place and monitor the patient in the

intensive care unit after surgery.

Summary

As experience with ETV grows, the procedure

will be performed by an increasing number ofneurosurgeons. Although the technique has beengreatly refined since its advent almost a century

ago, today’s neurosurgeon must never forget thatthis seemingly simple procedure holds the poten-tial for a number of devastating complications.

Appropriate training and experience are impor-tant to the success of ETV and for avoidingcomplications It is imperative that surgeonscontinue to report their experience with the

complications of ETV so that the procedure cancontinue to be made as safe as possible.

References

[1] Drake J, Kestle J, Milner R, Cinalli G, Boop F,

Piatt J, et al. Randomized trial of cerebrospinal

fluid shunt valve design in pediatric hydrocephalus.

Neurosurgery 1998;43:294–303.

[2] Teo C, Rahman S, Boop FA, Cherny B. Compli-

cations of endoscopic neurosurgery. Childs Nerv

Syst 1996;12:248–53.

[3] Guzelbag E, Ersahin Y, Mutluer S. Cerebrospinal

fluid shunt complications. Turk J Pediatr 1997;39:

363–71.

[4] Iskandar B, Tubbs S, Mapstone T, Grabb P,

Bartolucci A, Oakes W. Death in shunted hydro-

cephalic children in the 1990s. Pediatr Neurosurg

1998;28:173–6.

[5] Lazareff J, Peacock W, Holly L, Halen JV, Wong

A, Olmstead C. Multiple shunt failures: an analysis

of relevant factors. Childs Nerv Syst 1998;14:271–5.

[6] Lee JY, Wang KC, Cho BK. Functioning periods

and complications of 246 cerebrospinal fluid shunt-

ing procedures in 208 children. J Korean Med Sci

1995;10:275–80.

[7] Brockmeyer D, Abtin K, Carey L, Walker M.

Endoscopic third ventriculostomy: an outcome

analysis. Pediatr Neurosurg 1998;28:236–40.

[8] Hopf NJ, Grunert P, Fries G, Resch KD,

Perneczky A. Endoscopic third ventriculostomy:

outcome analysis of 100 consecutive procedures [see

comments]. Neurosurgery 1999;44:795–806.

[9] Fukuhara T, Vorster S, LucianoM. Risk factors for

failure of endoscopic third ventriculostomy for

obstructive hydrocephalus. Neurosurgery 2000;46:

1100–9.


[10] Drake JM. Ventriculostomy for treatment of hydro-

cephalus. Neurosurg Clin North Am 1993;4:657–66.

[11] Lowry D, Lowry D, Berga S, Adelson P, Roberts

M. Secondary amenorrhea due to hydrocephalus

treated with endoscopic ventriculocisternostomy.

Case report. J Neurosurg 1996;85:1148–52.



ment of noncommunicating hydrocephalus. Neuro-

surgery 1980;7:313–21.

[13] Jones R, Stening W, Brydon M. Endoscopic third

ventriculostomy. Neurosurgery 1990;26:86–91.




[15] Reddy K, Fewer HD, West M, Hill NC. Slit

ventricle syndrome with aqueduct stenosis: third

ventriculostomy as definitive treatment [see com-

ments]. Neurosurgery 1988;23:756–9.

[16] Handler MH, Abbott R, Lee M. A near-fatal

complication of endoscopic third ventriculostomy:

case report. Neurosurgery 1994;35:525–8.

[17] Abtin K, Thompson B, Walker M. Basilar artery

perforation as a complication of endoscopic third

ventriculostomy. Pediatr Neurosurg 1998;28:35–41.

[18] McLaughlin M, Wahlig J, Kaufman A, Albright A.

Traumatic basilar aneurysm after endoscopic third

ventriculostomy: case report. Neurosurgery 1997;

41:1400–3.

[19] Schroeder HW, Warzok RW, Assaf JA, Gaab MR.

Fatal subarachnoid hemorrhage after endoscopic

third ventriculostomy. Case report. J Neurosurg

1999;90:153–5.

[20] Cartmill M, Vloeberghs M. The use of trans-

endoscopic Doppler ultrasound as a safety-enhanc-

ing measure during neuroendoscopic third

ventriculostomy. Eur J Pediatr Surg Suppl 1999;1:

50–1.

[21] Schmidt R. Use of microvascular Doppler probe to

avoid basilar artery injury during endoscopic third

ventriculostomy. J Neurosurg 1999;90:156–8.

[22] Hirsch J, Hirsch E, Sainte-Rose C, Renier D,

Pierre-Kahn A. Stenosis of the aqueduct of Sylvius.

Etiology and treatment. J Neurosurg Sci 1986;30:

29–39.

[23] Cinalli G, Sainte-Rose C, Chumas P, Zerah M,

Brunelle F, Lot G, et al. Failure of third ventricu-

lostomy in the treatment of aqueductal stenosis in

children. J Neurosurg 1999;90:448–54.

[24] Cinalli G, Salazar C, Mallucci C, Yada JZ, Zerah

M, Sainte-Rose C. The role of endoscopic third

ventriculostomy in the management of shunt

malfunction. Neurosurgery 1998;43:1323–9.

[25] Jones R, Kwok B, Stening W, Vonau M. The

current status of endoscopic third ventriculostomy

in the management of non-communicating hydro-

cephalus. Minim Invasive Neurosurg 1994;37:

28–36.

[26] Jones R, Kwok B, Stening W, Vonau M. Third

ventriculostomy for hydrocephalus associated with

spinal dysraphism: indications and contraindica-

tions. Eur J Pediatr Surg 1996;6:5–6.




57–63.

1

d


Results of endoscopic third ventriculostomyMark R. Iantosca, MDa,*, Walter J. Hader, MD, FRCS(C)b,

James M. Drake, FRCS(C)caConnecticut Children’s Medical Center, 100 Retreat Avenue, Suite 705,

Hartford, CT 06106–2565, USAbDivision of Neurosurgery, University of Calgary, Alberta Children’s Hospital,

Calgary, Alberta, CanadacHospital for Sick Children, 555 University Avenue, Toronto,

Ontario M5G 1X8 Canada

The recent resurgence of interest in ventricu-lostomy has arisen out of dissatisfaction with the

complications and long-term outcomes of con-ventional cerebrospinal fluid (CSF) shuntingsystems [1,2]. Initial attempts at ventriculostomy

via open craniotomy and subsequent percutane-ous fluoroscopic and CT-guided techniques havelargely been replaced by modern endoscopicprocedures with reduced morbidity [3,4]. Un-

fortunately, sufficient long-term follow-up datanecessary for direct comparison with outcomes ofconventional shunting (CS) procedures are cur-

rently unavailable. Clinical decisions regardinguse of endoscopic third ventriculostomy (ETV)are based on the results of studies with mean

follow-up intervals of less than 5 years. Results ofETV are most closely associated with the etiologyof hydrocephalus encountered as well as with the

clinical and radiographic features of the individ-ual patient.

Factors affecting outcome

Hydrocephalus etiology

Third ventriculostomy is designed to treatnoncommunicating hydrocephalus with patent

subarachnoid spaces and adequate CSF absorp-tion. It is not surprising, therefore, that the results



(M.R. Iantosca).


oi:10.1016/S1042-3680(03)00067-6

of this procedure are most strongly influenced bythe etiology of hydrocephalus. Box 1 depicts

hydrocephalus etiologies divided according toreported success rates of ETV. Patients withacquired aqueductal stenosis or tumors obstruct-

ing third ventricular outflow have demonstratedthe highest success rates, exceeding 75% incarefully selected series of patients [4–10]. Pre-viously shunted patients with or without myelo-

meningocele, tumors, or cystic abnormalitiesleading to fourth ventricular outflow obstruction(ie, arachnoid cyst, Dandy-Walker malformations)

and patients with congenital aqueductal stenosishave shown an intermediate response [6,7,9,11–13]. Further study of this intermediate group is

likely to identify subgroups with higher successrates. Infants developing hydrocephalus afterhemorrhage or infection or with associated myelo-

meningocele (without prior shunting procedure)have demonstrated a poor response to ventricu-lostomy [3,5,13–16]; despite limited reports ofsuccess in such patients [7,11,12,17–20], they are

more controversial candidates for this procedure.The procedure is not advisable in patients whohave undergone prior radiation therapy because

of the extremely poor response rates, alteredanatomy (ie, thickened third ventricular floor),and increased risk of bleeding [3,7,11].

Aqueductal obstruction by stenosis or tumor

Blockage to CSF flow at the level of the cere-bral aqueduct by anatomic or postinflammatoryseptations/membranes or by extrinsic compression

ghts reserved.


68 M.R. Iantosca et al /Neurosurg Clin N Am 15 (2004) 67–75

secondary to tumor, benign cyst, or vascularlesions in a surrounding structure (tectum, pineal,posterior thalamus, and superior cerebellum) iscurrently the best-studied and most successful

indication for ETV. Success rates generally exceed75% for this select group of patients [4–10]. Ina recent large clinical series, including a wide age

range of patients (3 weeks to 77 years, n = 95)Hopf et al [9] reported an ETV success rategreater than 80% for postinflammatory and

idiopathic aqueductal stenosis and a 95% successrate for aqueductal stenosis caused by benign

Box 1. Endoscopic thirdventriculostomy success rates byhydrocephalus etiology

High success rates (>75%)Acquired aqueductal stenosisTumor, cyst, or infectious lesion

obstructing third ventricular outflowTectal, pineal, thalamic, or

intraventricular tumorArachnoid cystToxoplasmosisShunt malfunction in patient with

obstructive etiology

Intermediate success rates(approximately 50%)Tumor obstructing fourth ventricular

outflowMyelomeningocele (previously shunted,

older patients)Congenital aqueductal stenosis‘‘Cystic’’ abnormalities obstructing fourth

ventricular outflowArachnoid cystsDandy-Walker malformationPreviously shunted patients with

‘‘difficulties’’Slit ventricle syndromeRecurrent or intractable shunt infectionsRecurrent or intractable shunt

malfunctions

Low success rates (<50%)Myelomeningocele (previously

unshunted neonatal patients)Posthemorrhagic hydrocephalusPostinfectious hydrocephalus (excluding

aqueductal stenosis of infectiousetiology)

space-occupying lesions. ETV was less successfulin patients with progressive neoplastic lesions,primarily posterior fossa tumors and posthemor-

rhagic (intraventricular or subarachnoid) cases(64% and 63%, respectively) [9]. Similar resultshave recently been reported in a series analyzing213 children with similar etiologies by Cinalli et al

[4] encompassing stereotactic ventriculostomy andETV. The less favorable outcomes associated withposthemorrhagic and posterior fossa tumors and

cysts likely reflect the multifactorial causes ofhydrocephalus in these patients [4,9]. Studies ofETV for other hydrocephalic etiologies are limited

to smaller series and require further investigationto establish overall success rates.

Clinical features

Box 2 depicts preoperative clinical and radio-

graphic criteria frequently cited for improvingoutcome and limiting morbidity of ETV. A recentstudy by Gangemi et al [10] offers the only

detailed assessment of the contribution of pre-senting symptoms and signs to procedure success.Clinical presentations indicative of elevated in-

tracranial pressure (ICP) were associated with thehighest rate of ETV success (83%), followed byParinaud’s syndrome (77%) and macrocephaly(71%). Other presentations (memory disturban-

ces, incontinence, and focal deficits with asso-ciated tumors) exhibited success rates ofapproximately 50% or less. Patient age and

history of prior shunting procedure are the mostfrequently cited clinical features associated withoutcome.

Age

There seems to be a significant association

between increasing patient age and a more favor-able outcome of ETV in multiple pediatric studies[6–9,11,12,15,16,21]. These studies frequently im-plicate a decreased pressure gradient across the

arachnoid granulations (secondary to open su-tures) or age-related immaturity of patient CSFabsorptive capacity as reasons for this phenome-

non. Evidence suggests that reduced success ratesare associated with the age at which hydroceph-alus initially developed as well as with the age at

the time of ventriculostomy [7,21]. The seriesby Jones et al [21] reports that only 8 of 25ventriculostomies were successful in patients lessthan 6 months of age, with 8 of 17 successful

procedures in patients developing hydrocephalus

69M.R. Iantosca et al / Neurosurg Clin N Am 15 (2004) 67–75

Box 2. Favorable clinical and radiographic features for endoscopic third ventriculostomy

ClinicalEtiology of hydrocephalus in high or intermediate success group (see Box 1.)Age greater than 6 months at time of hydrocephalus diagnosisAge greater than 1 year at time of procedure (controversial)No prior radiation therapyNo history of hemorrhage or meningitisPreviously shunted, particularly for high success group etiology (see Box 1.)

RadiographicClear evidence of ventricular noncommunicationObstructive pattern of HCPQAqueductal anatomic obstructionAbsence of aqueductal flow on T2-weighted MRI or cine study

Favorable third ventricular anatomyWidth and foramen of Monro sufficient to accommodate endoscopeRigid >6 mmFlexible > 4 mm

Thinned floor of third ventricleDownward bulging floor draped over clivusBasilar posterior to mamillary bodies

Absence of structural anomalies impeding procedureArteriovenous malformation (AVM) or tumor obscuring third ventricular floorEnlarged massa intermediaInsufficient space between mamillary bodies/basilar and clivusBasilar artery ectasia

before this age. Most of the successes in thesegroups were in previously shunted patients [21].

Several studies show success rates at or below50% in patients less than 2 [6,7,9] or 1 [8,15,16]year of age, regardless of etiology. Results are

even poorer in patients less than 6 months of age[12,21]. Teo and Jones [12] demonstrated statisti-cally significant (P = 0.005) lower success rates inpatients less than 6 months of age versus older

patients (12.5% vs. 80%). Open third ventricu-lostomy has also been shown to have a dramati-cally decreased success rate in patients less than 6

months of age [18,22]. The results of these studiesare seemingly contradicted by the recent report ofCinalli et al [4], who demonstrated no difference in

failure rates of ETV between patient groups olderor younger than 6 months, concluding that patientage should no longer be considered a contraindi-

cation to the procedure.Despite the lower success rates and increased

morbidity that they report in this younger agegroup, several authors still advocate an initial

attempt at ventriculostomy in these patients so as

to avoid the lifelong complications of shunting[8,15,16]. These authors argue that the ‘‘successes’’

are spared the long-term morbidity, mortality,and economic burden of repeated evaluations andhospitalizations for shunt evaluation. Barlow and

Ching [23] recently published an analysis ofanticipated cost reduction comparing ETV andCS for treatment of obstructive hydrocephalus attheir institution, estimating a reduction of 74

hospital days and nine surgeries annually. Thisstudy, however, assumes a success rate of 80%,with no subsequent admissions or procedures for

complications or failures among the successfulprocedures. Insufficient long-term data currentlyexist to support this argument in younger patient

groups, particularly given the often-cited reducedsuccess rates and increased rates of complications.

ETV in this younger age group has also been

advocated as a primary treatment to removeintraventricular hemorrhage or purulent CSF,assuming that this will result in a more suitableenvironment for shunt insertion [15,24,25].

Comparison of this procedure with standard


‘‘temporizing measures,’’ such as ventriculostomy,ommaya placement, and ventriculosubgaleal shunt-ing is needed to determine comparative efficacy

and safety.Interestingly, a recent report by Tisell et al [26]

demonstrates an increased failure rate of ETVin older adults with aqueductal stenosis. These

delayed failures were hypothetically attributed todecreasing CSF absorptive capacity as a result ofage-related loss of arachnoid villi and decreasing

resistance to transependymal CSF resorption overtime. Only 50% of ETVs were successful in thisseries, with almost all failures occurring after

an initially favorable response lasting between1 month and 1 year [26]. An earlier report byKelly [5] demonstrated a 94% success rate ina population of adults with the same etiology

treated with stereotactic CT-guided ventriculos-tomy. Interestingly, most patients in Kelly’s serieswere previously shunted (11 of 16), whereas only 3

of Tissell et al’s 18 patients had existing shunts.Clearly, further delineation of the pathophysiol-ogy of age-related changes in CSF absorption,

and development of studies to quantify thiscapacity, would greatly benefit clinicians attempt-ing to identify appropriate patients for ETV.

Previous shunting

Multiple studies have demonstrated a trendtoward more successful ventriculostomy outcome

in patients with existing shunt systems [6,12,21].Teo and Jones [12] demonstrated a statisticallysignificant difference in success rates for pre-viously shunted patients with myelomeningocele

(84%) versus those never shunted (29%;P = 0.002). In fact, third ventriculostomy hasbeen found to be useful in the treatment of

intractable shunt infections and malfunctions andeven slit ventricle syndrome refractory to othertreatments [5,11,24,27–30]. Other series of ventri-

culostomy in previously shunted patients havebeen less promising [13,31]. These contradictoryresults may reflect the mix of patients in small

series. High success rates in patients with aque-ductal obstructive etiologies present a strongargument in favor of ETV for patients in thisgroup who present with shunt malfunction. Im-

proved outcome in previously shunted patientswith other etiologies is commonly attributed toincreased CSF absorptive capacity; however, the

effect of the shunt itself on CSF absorption isdifficult to distinguish from the effect of increasedage in these patients.

Preoperative assessment of CSF absorptivecapacity has been advocated by some authors[32,33], but these techniques have not been widely

accepted [7,21]. Because the patency of CSFpathways is one of the assumptions on whichthe procedure is based, a preoperative CSFabsorption study would be invaluable in identify-

ing patients who would likely benefit fromventriculostomy. The observed delay betweenoperation and decrease in ventricular size (see

section on assessing outcome) suggests that CSFabsorption may increase slowly after ventriculos-tomy, however. Therefore, preoperative assess-

ment of CSF absorptive capacity may fail toidentify all appropriate patients [7]. Until anaccurate functional predictor of this capacity ispractical, radiographic assessment of the obstruc-

tive pattern of hydrocephalus by MRI is likely toremain the preoperative diagnostic procedure ofchoice.

Anatomy

Preoperative MRI optimally demonstrates allrelevant anatomic features and should be ob-

tained for all proposed third ventriculostomypatients. Initially, confirmation of noncommuni-cating hydrocephalus of favorable etiology should

be established by the pattern of ventriculardilatation. Anatomic obstruction of CSF path-ways between the aqueduct and the fourth

ventricular outflow foramina may be visible onT2- or T1-weighted contrast images. Additionally,T2-weighted and cine images should reveal noevidence of the normally present aqueductal CSF

flow pattern.Once the patient’s suitability for the procedure

has been established, the neurosurgeon must

clarify the details of third ventricular anatomythat are likely to affect morbidity. First, the widthof the third ventricle and diameter of the foramen

of Monro must be sufficient to accommodatethe endoscope of choice. Additionally, the thick-ness of the third ventricular floor and anatomy of

the proposed puncture site must be assessed inrelation to vital structures, particularly the basilarartery and its branches. A downward bulgingthird ventricular floor draped over the clivus has

been cited as a prerequisite for success of thisprocedure in the past, but others have not foundthis to be necessary [3,7]. Ultimately, the surgeon

must be satisfied that there is no structural lesion(ie, tumor, AVM) or anatomic variation thatwould render the procedure unduly difficult or


hazardous. In cases of doubt, it is reasonable tovisualize the floor of the third ventricle and toabandon the procedure if the floor is unsuitable.Reported procedure abort rates as a result of

unfavorable anatomy or intraoperative bleedingrange from 0% to 26% (Table 1).

Assessing outcome

ETV has yielded a higher success rate with

lower morbidity and mortality than earliermethods of third ventriculostomy. Mortality ratesfor open ventriculostomy procedures varied be-

tween 5% and 27%, with success rates from 37%to 75% [3,5,18,31,34,35]. Percutaneous radio-graphic and, later, CT-guided techniques reduced

this mortality rate to 2% to 7%, with a 44% to75% rate of shunt independence [3,5,14,19,31,36].Most recent studies using modern endoscopictechniques and equipment, with or without

stereotactic CT or MRI guidance, have reportedlow morbidity (3%–12%) and extremely raremortality, with success rates greater than 75%

for carefully selected patient groups. Table 1depicts the results of recent ETV studies withsuccess rates by etiology where these data are

available. The current challenge is to defineappropriate indications and devise objectivemeasures for pre- and postoperative assessmentof these patients.

The goal of ETV and, to date, the bestobjectively quantifiable measure of a successfuloutcome is shunt independence. Vague outcome

measures, such as ‘‘successful,’’ ‘‘improved,’’ or‘‘favorable,’’ are used in a number of studies[12,21]. Only two reports have critically compared

modern ETV techniques with CS [37,38]. Sainte-Rose [38] reported no statistically significantdifferences in measures of neurologic, endocrino-

logic, social, or behavioral outcomes in a group of68 patients treated for aqueductal stenosis witheither ETV (n = 30) or CS (n = 38). Tuli et al [37]demonstrated no difference in failure rate between

ETV (n = 32, 44% failure rate) and CS (n = 210,45% failure rate) in a prospectively followed groupof patients with aqueductal stenosis or tumor. The

long-term outcomes of CS, including multipleprocedures for obstruction, infection, and over-drainage, have been well documented. Failure

rates range from 30% to 47% over 1 year to63% to 70% over 10 years [39–42]. Additionallong-term studies of ETV and analytic comparisonto established CS outcomes are necessary to clarify

clinical decisions in these patients.

One of the most confusing aspects of outcomeevaluation in ETV patients is the failure of theventricles to return to normal size. Many reportsdetail a gradual decrease in the ventricular size

over months to years after surgery, with resolutionof periventricular edema and increased extracere-bral spaces coinciding with clinical improvement

[5–8,31,38]. One series of patients treated witheither CSF shunts or ETV showed no difference inintellectual outcome despite enlarged ventricles in

the ETV group [38]. Radiographic evaluation ofsuspected closure has been confounded by thepresence of persistently enlarged ventricles.

Studies detailing the serial measurements ofmultiple radiographic indices of ventricular sizeafter surgery show that the third ventricular sizeresponds more quickly (usually within 3 months)

than the lateral ventricular size (up to 2 years)[43,44]. Additionally, third ventricular size seemsto correlate most closely with outcome in these

patients [8,43,44]. A recent study limited to adultpatients showed no relation between ventricularsize reduction and outcome [26]. Kulkarni et al

[45] have demonstrated a greater reduction inventricular size after successful procedures versustreatment failures in 29 children.

MRI and Doppler ultrasound studies docu-menting CSF flow through a patent ventriculos-tomy are currently the most useful and widelyused postoperative radiographic indices. MRI

detection of T2-weighted flow void around theventriculostomy has been correlated with clinicaloutcome in ETV [46–48]. This observation has

proven most helpful in confirming ventriculos-tomy patency after surgery, particularly inpatients with persistent ventriculomegaly [47,48].

Kulkarni et al [45] have recently demonstrateda statistically significant relation between evidenceof postoperative aqueductal CSF flow on MRIand clinical success. Multiple studies have re-

ported long-term patency of third ventriculos-tomy by means of cine phase contrast (PC) MRIstudies. A report by Cinalli et al [4] included 15

patients with confirmed long-term (>10 years)patency of their ventriculostomy on cine PC MRI(median = 14.3 years, range: 10–17 years). Un-

fortunately, several recent studies in adults andchildren have demonstrated a lack of correlationbetween postoperative MRI findings and outcome

[15,16,26,37]. Quantification of flow velocitythrough ventriculostomies has also been achievedby PC MRI and Doppler ultrasound [49,50].Hopefully, methods of quantifying ventriculos-

tomy function will allow neurosurgeons to refine

Table 1

Outcome of e

pSuccess rate (%) by

etiology (n)

Author/Year Tumor/AS Dysraphic PHH/PIH Other (n-etiology)

Jones et al / 1 59% (17) 40% (5)

Jones et al / 1 M–7 Y) 68% (19) 40% (10) 0% (2)

Jones et al / 1 80% (25) 52% (21)

Sainte-Rose a 81% (82)

Teo and Jone l–17 Y) 72% (69)

Goumnerova l–44 M) 70% (20) 0% (1) 100% (2-IDO)

Baskin et al / 64% (22-SVS)

Brockmeyer e 5 l–69 M) 59% (34) 50% (16) 0% (7)

Buxton et al / –42 M) 57% (7) 0% (2) 10% (10) 20% (5-IDO)

Buxton et al / ) 43% (7) 30% (10)

Tuli et al / 19 –11 Y) 66% (32)

Gangemi et a l–54 M) 91% (110)

Cinalli et al / D–9 Y) 86% (119)

Hopf et al / 1 l–71 l) 83% (82) 63% (8)

Tisell et al / 2 M–5 Y) 50% (18)

Abbreviati /R, not reported; MMC, myelomeningocele, PHH post hemorrhagic

hydrocephalu ications (intractable shunt infection/malfunction); SVS, slit-ventricle

syndrome; W

* All patie

72

M.R.Iantosca

etal/Neurosurg

Clin

NAm

15(2004)67–75

ndoscopic third ventriculostomy by etiology

Age mean

Procedure

abort

Follow-u

mean

(range) Morbidity rate (range)

990 [11] (4 M–17 Y) 8% 16% N/R

992 [6] (PNB–78 Y) 7% 9% 27 M (3

994 [7] N/R 5% 6% N/R

nd Chumass / 1996 [8] 5.3 Y N/R N/R 1.8 Y

s/ 1996 [12] 11 Y (1 W–32 Y) 3% 9% 32 M (1

and Frim / 1997 [47] 11.2 Y (2 D–36 Y) 9% N/R 17 M (7

1998 [30] 17.3 Y (1.5–49 Y) 12% N/R (21.4 M)

t al / 1998 [13] 8.1 Y (1 D–29.5Y ) 6% 26% 24.2M (1

1998 [16] 3.7 M (0 M–10 M) 15% 0.4% N/R (6 l1998 [15] 2 M (0–11 M)* All PNB 21% 11% (6 l–42 M

98 [37] 8.1 Y (0 M–18 Y) N/R N/R N/R (1 Y

l / 1999 [10] 31 Y (7 D–81 Y) 12% 28 M (12

1999 [4] NR (1 M–18 Y) N/R 6% 2.1 Y (4

999 [9] 36 Y (3 W–77 Y) 6% 2% 26 M (3

000 [26] 48 Y (17–80 Y) N/R N/R 37 M (3

ons: AS aqueductal stenosis; D, days; IDO, idivpathie hydrocephalues; m, months; N

s; PIH post infectious hydrocephalus; PNB, Premature newborn, SC; shunt compl

, weeks; Y, year’s.

nts born prematurely: mean age at birth was 31.9 weeks (range: 26–36 weeks).


the techniques necessary to improve outcomefurther. Radiographic confirmation may also helpto define indications with more subjective out-comes, for example, the observation of less

fulminant shunt malfunctions in patients afterventriculostomy [11].

The authors use postoperative MRI imaging

with sagittal T2-weighted images or cine MRI toconfirm ventriculostomy flow and assess ventric-ular size 3 to 6 months after surgery. The absence

of a flow void, although occasionally present inpatients with good clinical outcome, was presentin 12 of 13 patients presenting with delayed

treatment failure in the series by Cinalli et al [4].Increasing head circumference, signs and symp-toms of elevated ICP, or evidence of CSF egress atthe ventriculostomy incision site generally heralds

early failures. Because of the risk of delayedfailure of ETV, and the potential mortality ifthis is unrecognized, we recommend that these

patients use medical alert bracelets or walletidentification cards. A low threshold of suspicionfor reimaging these patients should be followed,

especially for the first 5 years after ETV. Familiesand caregivers are given the same education asfor patients with shunts regarding symptoms of

failure. Education and regular follow-up areimportant to counteract the false sense of securitythat may arise in the absence of a shunt.

Failure of endoscopic third ventriculostomy

Studies of ETV are currently limited to studieswith an average of less than 5 years of follow-up.These studies have documented early (<6 months)

and late (>1 year) postoperative failures. Earlyfailures are more frequently reported in patientsless than 6 months of age [4,15,16]. Buxton et al[15,16] have reported a mean time to failure in

patients younger than 1 year of age of 1.36months. Multiple authors have reported ‘‘latefailures,’’ where the ventriculostomy closes, some-

times years after surgery [4,7,9,24,26]. Obstructionof the ventriculostomy by a gliotic scar or arach-noid membranes has been described [31,51]. Rare

cases of sudden death after late failure of ETVhave been reported [52].

Although the long-term results of ETV arelargely unknown, studies that include long-term

follow-up on stereotactic ventriculostomy casesare likely applicable to ETV. Cinalli et al’s recentreport [4] included 94 cases of stereotactic ven-

triculostomy with an average of 6.32 years offollow-up (range: 20 days to 17.4 years) and 119

case of ETV with an average of 2.1 years offollow-up (range: 4 days to 9 years). The overallfunctional ventriculostomy rate for the study was72% at 6 years, without significant differences in

the long-term results between the stereotactic andendoscopic groups [4]. Cine PCMRI studies on 15patients up to 17 years after stereotactic ventricu-

lostomy confirmed long-term patency (range: 10–17 years, median = 14.3 years) [4]. No failureswere reported in either group after 5 years of

follow-up. Additionally, 7 of 9 patients failingventriculostomy with radiographically provenobstruction of the stoma had successful repeat

ventriculostomies. Long-term data on ETV out-comes are only now beginning to emerge, andcomparison to known outcomes of standardshunting procedures will be instrumental in de-

fining the indications for initial treatment withETV and the management of treatment failures.

Summary

ETV is emerging as the treatment of choice foraqueductal stenosis caused by anatomic, inflam-

matory, and selected neoplastic etiologies. Thetechnique has also proven useful in the pathologicdiagnosis and treatment of these conditions[53–56]. Long-term results of this procedure and

comparison to standard shunting procedures arenecessary to define indications for patients withpathologic findings in the intermediate response

groups. Development of new studies for pre-operative assessment of CSF absorptive capacityand quantitative postoperative measures of

ventriculostomy function would be invaluableadditions to our ability to assess candidates forthis procedure and their eventual outcome. Fur-

ther study and technical refinements will, nodoubt, lead to many more potential uses for theseprocedures in the treatment of hydrocephalus andits associated etiologies. The challenge for neuro-

surgeons will be to define the operative indicationsand outcomes, while refining techniques for safelyperforming these useful procedures.

References

[1] Sainte-Rose C, Hoffman HJ, Hirsch JF. Shunt

failure. Concepts Pediatr Neurosurg 1989;9:7–20.

[2] Hoppe-Hirsch E, et al. Late outcome of the surgical

treatment of hydrocephalus. Childs Nerv Syst 1998;

14(3):97–9.

[3] Drake JM. Ventriculostomy for treatment of

hydrocephalus. Neurosurg Clin North Am 1993;

4(4):657–66.


[4] Cinalli G, et al. Failure of third ventriculostomy in

the treatment of aqueductal stenosis in children.

J Neurosurg 1999;90(3):448–54.

[5] Kelly PJ. Stereotactic third ventriculostomy in

patients with nontumoral adolescent/adult onset

aqueductal stenosis and symptomatic hydrocepha-

lus [see comments]. J Neurosurg 1991;75(6):

865–73.

[6] Jones RF, et al. Neuroendoscopic third ventriculos-

tomy. In: Manwaring KH, Crone KR, editors.

Neuroendoscopy, vol. 1. New York: Mary Ann

Liebert; 1992. p. 63–77.

[7] Jones RF, et al. The current status of endoscopic

third ventriculostomy in the management of non-

communicating hydrocephalus. Minim Invasive

Neurosurg 1994;37(1):28–36.

[8] Sainte-Rose C, Chumas P. Endoscopic third ven-

triculostomy. Tech Neurosurg 1996;1:176–84.

[9] Hopf NJ, et al. Endoscopic third ventriculostomy:

outcome analysis of 100 consecutive procedures [see

comments]. Neurosurgery 1999;44(4):795–806.

[10] Gangemi M, et al. Endoscopic third ventriculos-

tomy for hydrocephalus. Minim Invasive Neuro-

surg 1999;42(3):128–32.


third ventriculostomy. Neurosurgery 1990;26(1):

86–92.



myelomeningocele. Pediatr Neurosurg 1996;25(2):

57–63.

[13] Brockmeyer D, et al. Endoscopic third ventriculos-

tomy: an outcome analysis. Pediatr Neurosurg

1998;28:236–40.

[14] Jaksche H, Loew F. Burr hole third ventriculo-

cisternostomy. An unpopular but effective procedure

for treatment of certain forms of occlusive hydro-

cephalus. Acta Neurochir (Wien) 1986;79:48–51.

[15] Buxton N, et al. Neuroendoscopy in the premature

population. Childs Nerv Syst 1998;14(11):649–52.

[16] Buxton N, et al. Neuroendoscopic third ventricu-

lostomy in patients less than 1 year old. Pediatr

Neurosurg 1998;29(2):73–6.

[17] Natelson SE. Early third ventriculostomy in

meningomyelocele infants—shunt independence?

Childs Brain 1981;8(5):321–5.

[18] Patterson RH Jr., Bergland RM. The selection of

patients for third ventriculostomy bases on experi-

ence with 33 operations. J Neurosurg 1968;29(3):

252–4.

[19] Sayers MP, Kosnik EJ. Percutaneous third ventri-

culostomy: experience and technique. Childs Brain

1976;2(1):24–30.

[20] Vries JK, Friedman WA. Postoperative evaluation

of third ventriculostomy patients using 111in-

DTPA. Childs Brain 1980;6(4):200–5.

[21] Jones RF, et al. Neuroendoscopic third ventricu-

lostomy. A practical alternative to extracranial

shunts in non-communicating hydrocephalus. Acta


[22] Akdemir H, et al. Failure of open third ventricu-

lostomy for shunt infections in infants. Pediatr

Neurosurg 1999;30:305–9.

[23] Barlow P, Ching HS. An economic argument in

favour of endoscopic third ventriculostomy as

a treatment for obstructive hydrocephalus. Minim


[24] Jones RF, et al. Third ventriculostomy for shunt

infections in children. Neurosurgery 1993;32(5):

855–60.

[25] Mallucci C, et al. Neuroendoscopic third ventricu-

lostomy: the first line of treatment for blocked

ventriculoperitoneal shunts? Eur J Pediatr Surg

1997;7(Suppl 3).

[26] Tisell M, et al. How effective is endoscopic third

ventriculostomy in treating adult hydrocephalus

caused by primary aqueductal stenosis? Neurosur-

gery 2000;46(1):104–11.

[27] Yamamoto M, et al. Percutaneous flexible neuro-

endoscopic ventriculostomy in patients with shunt

malfunction as an alternative procedure to shunt

revision. Surg Neurol 1994;42(3):218–23.

[28] Perlman BB. Percutaneous third ventriculostomy in

the treatment of a hydrocephalic infant with

aqueduct stenosis. Int Surg 1968;49(5):443–8.

[29] Reddy K, et al. Slit ventricle syndrome with

aqueduct stenosis: third ventriculostomy as de-

finitive treatment [see comments]. Neurosurgery

1988;23(6):756–9.



ventricle syndrome. J Neurosurg 1998;88(3):478–84.



ment of noncommunicating hydrocephalus. Neuro-

surgery 1980;7(4):313–21.

[32] Foltz EL. Treatment of hydrocephalus by ventric-

ular shunts [letter]. Childs Nerv Syst 1996;12(6):

289–90.

[33] Pudenz R, Foltz E. Hydrocephalus: overdrainage

by ventricular shunts—a review and recommenda-

tions. Surg Neurol 1991;35:200–12.

[34] Dandy WE. Diagnosis and treatment of strictures

of the aqueduct of Sylvius (causing hydrocephalus).

Arch Surg 1945;51:1–14.

[35] Scarff JE. Evaluation of treatment of hydrocepha-

lus. Results of third ventriculostomy and endoscop-

ic cauterization of choroid plexuses compared with

mechanical shunts. Arch Neurol 1966;14(4):382–91.

[36] Forjaz S, Martelli N, Latuf N. Hypothalamic

ventriculostomy with catheter. Technical note.

J Neurosurg 1968;29(6):655–9.


versus cerebrospinal fluid shunt as a first procedure

in pediatric hydrocephalus. Pediatr Neurosurg

1998;30:11–5.


[38] Sainte-Rose C. Third ventriculostomy. In: Manwar-

ing KH, Crone KR, editors. Neuroendoscopy, vol.

1. New York: Mary Ann Liebert; 1992. p. 47–62.

[39] Sainte-Rose C, et al. Mechanical complications in

shunts. Pediatr Neurosurg 1991;17:2–9.

[40] Di Rocco C, Marchese E, Verladi F. A survey of the

first complication of newly implanted CSF shunt

devices for the treatment of nontumoral hydro-

cephalus. Cooperative survey of the 1991–1992

Education Committee on the ISPN. Childs Nerv

Syst 1994;10(5):321–7.

[41] Albright AL, Haines SJ, Taylor FH. Function of

parietal and frontal shunts in childhood hydroceph-

alus. J Neurosurg 1988;69:883–6.

[42] Piatt JH Jr. Cerebrospinal fluid shunt failure: late is

different from early. Pediatr Neurosurg 1995;23:

133–9.

[43] Oka K, et al. The radiographic restoration of the

ventricular system after third ventriculostomy.


[44] Schwartz TH, et al. Third ventriculostomy: post-

operative ventricular size and outcome. Minim


[45] Kulkarni AV, et al. Imaging correlates of successful

endoscopic third ventriculostomy. J Neurosurg

2000;92(6):915–9.

[46] Jack CR Jr., Kelly PJ. Stereotactic third ventricu-

lostomy: assessment of patency with MR imaging.

AJNR Am J Neuroradiol 1989;10(3):515–22.

[47] Goumnerova LC, Frim DM. Treatment of hydro-

cephalus with third ventriculocisternostomy: out-

come and CSF flow patterns. Pediatr Neurosurg

1997;27(3):149–52.

[48] Wilcock DJ, et al. Neuro-endoscopic third ventri-

culostomy: evaluation with magnetic resonance

imaging. Clin Radiol 1997;52(1):50–4.

[49] Lev S, et al. Functional analysis of third ventri-

culostomy patency with phase-contrast MRI ve-

locity measurements. Neuroradiology 1997;39(3):

175–9.

[50] Wilcock DJ, Jaspan T, Punt J. CSF flow through

third ventriculostomy demonstrated with colour

Doppler ultrasonography. Clin Radiol 1996;51(2):

127–9.

[51] Hirsch JF, et al. Stenosis of the aqueduct of Sylvius.

Etiology and treatment. J Neurosurg Sci 1986;

30(1–2):29–39.

[52] Hader W, et al. Death following late failures of

third ventriculostomy in children: a report of 3

cases. J Neurosurg 2002, in press.

[53] Ellenbogen RG, Moores LE. Endoscopic manage-

ment of a pineal and suprasellar germinoma with

associated hydrocephalus: technical case report.


[54] Drake J. Neuroendoscopy tumour biopsy. New

York: Mary Ann Leibert; 1992. p. 103–9.

[55] Ferrer E, et al. Neuroendoscopic management of

pineal region tumours. Acta Neurochir (Wien)

1997;139(1):12–21.

[56] Oka K, et al. Endoneurosurgical treatment for

hydrocephalus caused by intraventricular tumors.

Childs Nerv Syst 1994;10(3):162–6.


Neuro-oncologic applications of endoscopyCharles Teo, MBBS, FRACSa, Peter Nakaji, MDa,b,*

aCentre for Minimally Invasive Neurosurgery, Prince of Wales Hospital/University of New South Wales,

Barker Street, Randwick, NSW 2031, Sydney, AustraliabDivision of Neurological Surgery University of California at San Diego Medical Center, San Diego, CA, USA

No other subspecialty of neurosurgery exem-plifies the advantages of endoscopy more thanneuro-oncology. From a pure technical stand-

point, visualization is the sine qua non of surgery.If a surgeon cannot see what he is doing, he isineffective and dangerous. Cranial neurosurgery

in particular is a constant struggle against poorvisualization. In an attempt to minimize operativetrauma, the surgeon aims to limit the size of the

exposure and to avoid vigorous brain retraction.Meanwhile, the tumor/brain interface is oftenhard to distinguish, and the tumor often insin-

uates itself behind or between structures thatcannot be sacrificed. Critical structures, such ascranial nerves, vessels, and eloquent brain tissue,may prohibit direct visual access to some parts of

the tumor. At the same time, in the case of themore benign tumors that occur in the cranialcavity, total excision is vital to the patient’s

survival [1,2]. Complete removal of an acousticneuroma is nearly tantamount to cure. Otherextra-axial tumors, such as craniopharyngiomas

and meningiomas, also must be removed ascompletely as possible to reduce the incidence oftumor recurrence. Because low-grade gliomasrespond poorly to radiotherapy and chemother-

apy, prognosis is directly related to the degreeof tumor resection. Ependymomas are also his-torically insensitive to adjuvant therapies, and

total macroscopic removal is paramount [3]. Anytool that improves visualization, thereby offering



(C. Teo).


doi:10.1016/S1042-3680(03)00068-8

patients a better surgical outcome, should beembraced by the neuro-oncologic surgeon.

The endoscope is such a tool. It enhances the

surgeon’s view by increasing illumination andmagnification [4,5]. It allows the surgeon to viewtumor remnants, such as those hidden behind the

tentorial edge, a cranial nerve, or eloquent braintissue. Once the tumor is removed, the surgeoncan use the endoscope to assess the degree of

resection. Often, the same surgery can beperformed through a smaller craniotomy by usingthe endoscope, in keeping with the concept of

minimally invasive yet maximally effective surgery[6]. Adjunctive procedures, such as third ventric-ulostomy and septostomy, can be performedthrough the same access to manage related

problems, such as secondary hydrocephalus.Endoscopic tumor removal or cerebrospinal fluid(CSF) diversion may allow patients to avoid shunt

placement. Finally, the endoscope is an excellentteaching tool. The anatomic definition and uniqueangles of view available to the endoscope help

residents with their understanding of operationsand help to illuminate anatomicopathologic con-cepts underlying neuro-oncologic surgery.

Applications

Applications for the endoscope are limitedonly by one’s imagination. Initially used forintraventricular surgery alone, the endoscope has

also become an invaluable tool in the surgicalmanagement of extra-axial pathologic processes.It is being increasingly applied as an adjunct to the

removal of masses in the subarachnoid space.Within the ventricle, the endoscope is a versatiletool for diagnosing and treating tumors and

associated problems of CSF circulation.

ghts reserved.


90 C. Teo, P. Nakaji / Neurosurg Clin N Am 15 (2004) 89–103

We discuss the neuro-oncologic applications ofendoscopy under the following headings:

1. Ventriculoscopy2. Management of secondary hydrocephalus3. Endoscopic tumor biopsy

4. Endoscopic intraventricular tumor resection5. Endoscope-assisted microsurgery

Ventriculoscopy

The prognosis of some primary intracranialtumors is dependent on the presence or absence ofependymal spread of tumor. Patients with prim-

itive neuroectodermal tumors, for example, fallinto the high-risk group rather than the low-riskgroup if there is evidence of spinal or ventricular

ependymal tumor involvement. Although MRI isreliable in the detection of ependymal tumorspread in most cases, some patients may have

ependymal spread without radiologic evidence [7].Ventriculoscopy can be more sensitive than MRIwith little added morbidity. Through a frontal or

parietal burr hole, one can access the lateralventricle, examine the surface, document anyfindings with color photography, and even biopsysuspicious areas. This can be performed with a 10-

minute general anesthetic, which is substantiallyless time than is needed for an MRI examination.Furthermore, if present, definitive treatment of

CSF obstruction can be achieved by either thirdventriculostomy or tumor resection at the samesitting.

Fig. 1 shows endoscopic pictures of the lateralventricle of a patient with a pineal primitiveneuroectodermal tumor who had no imaging

evidence of ependymal tumor involvement. Thepatient had an endoscopic third ventriculostomy(ETV) and, thereafter, chemotherapy. The re-sponse to chemotherapy was well documented by

ventriculoscopy, thereby allowing the patient tobe a candidate for high-dose chemotherapy andautologous bone marrow transplantation.

Management of secondary hydrocephalus

The management of secondary hydrocephalus

is a controversial area of pediatric neuro-oncology. There are several different treatment

paradigms. A popular option would be to placea temporary external CSF drain as a primaryprocedure, remove the obstruction, and then, as

a secondary procedure, shunt those patients whocontinue to require CSF drainage. Anotheroption is to insert a permanent shunt asa primary procedure before tumor resection. A

third option would be to perform an ETV eitherprimarily or secondarily. The arguments in favorof the early management of the hydrocephalus

are as follows:

1. There are some circumstances in which thesurgeon cannot definitively treat the primarycause of obstruction, and CSF diversion is the

only option. These patients may present withsymptoms of raised intracranial pressure andimpending brain herniation.

2. The surgeon may need more time for pre-

operative assessment (eg, the patient witha bleeding diathesis, the patient who needsstaging of his or her tumor).

3. The primary CSF diversionary proceduremay be definitive. An example of this is thepatient with a tectal glioma, in whom

symptoms are usually secondary to hydro-cephalus and not the primary tumor.

4. On rare occasions, MRI reveals pathologypreviously ‘‘hidden’’ by the hydrocephalus

(Fig. 2). The patient in Fig. 2 presented withsevere intracranial hypertension and wasthought to have hydrocephalus from aque-

ductal stenosis. After ETV, it became clearthat the obstruction was secondary to a pinealregion tumor.

5. Some surgeons believe that operative con-ditions for the definitive surgery are betterafter prior drainage of CSF.

The arguments against initial treatment of thehydrocephalus are as follows:

1. Many patients do not have a permanent

CSF flow problem; therefore, some shuntsare placed unnecessarily. Of course, this isnot a concern when temporary CSF diver-

sion is undertaken using external ventriculardrainage.

2. If the obstruction is caused by a posterior

fossa tumor, there is potential for upward

Fig. 1. (A) The ependyma is frosted with plaque-like tumor that was not seen on MRI. (B) After conventional

chemotherapy, the clinical response is documented ventriculoscopically. (C) After high-dose chemotherapy and

autologous bone marrow transplantation, the ependymal surface is clear of visible tumor.

c

91C. Teo, P. Nakaji /Neurosurg Clin N Am 15 (2004) 89–103


Fig. 2. (A) The patient is an 11-year-old boy with profound hydrocephalus of unclear etiology. (B) After an endoscopic

third ventriculostomy and resolution of the hydrocephalus, an occult pineal region tumor is now apparent.

herniation with decompression of the supra-tentorial CSF spaces.

3. There is a risk of infection with external

drainage.4. There is a risk of spreading tumor cells to the

peritoneum with extracranial CSF diversion.

ETV is a reasonable option if the surgeondecides to treat the secondary hydrocephalus

before definitive treatment of the primary tumor.This operation can be done through a standardfrontal burr hole similar to that which would be

made for insertion of an external ventricular drain(Fig. 3). Similarly, CSF samples can be taken atthe time of surgery, and one may also explore the

third ventricle, taking biopsies if necessary. Thereis no risk of ongoing infection as can be seen withexternal drainage; no risk of seeding as has beendocumented with ventriculoperitoneal shunting;

and no risk of upward herniation, which canpotentially happen with any extracranial diver-sionary procedure. Finally, ETV may be the

definitive treatment if the obstruction is causedby a tumor that does not require removal, such asa tectal plate tumor.

Endoscopic septum pellucidotomy may also bethe definitive treatment for some types of hydro-cephalus. Fig. 4 shows the MRI scan of an elderly

lady who presented with headaches and disorien-tation. She was found to have a cystic tumor ofthe third ventricle causing obstruction of onlyone of the lateral ventricles. The provisional diag-

nosis was a craniopharyngioma, and she under-went endoscopic cyst fenestration and drainage,followed by septum pellucidotomy. The post-

operative scans showed complete resolution ofthe unilateral hydrocephalus and shrinkage ofthe cystic component of the biopsy-proven

craniopharyngioma.

Technical notesThe technique of ETV is described in else-

where in this issue. There are a few precautions

when performing ETV for hydrocephalus second-ary to a tumor. First, the anatomy may bealtered. A tumor like a pontine glioma may

distort the floor of the third ventricle and displacethe basilar artery forward so that the safe zone topenetrate the floor is a submillimetric area just


Fig 3. (A) The endoscopic third ventriculostomy is performed in the third ventricular floor just behind the mamillary

bodies. The close relation to the basilar apex is illustrated. (B) In chronic hydrocephalus, the third ventricular floor is

thin and visualization of the mamillary bodies is easy. (C) In acute hydrocephalus, it may not be possible to see the usual

landmarks. The floor is thicker, and a sharp technique may be necessary. In this instance, more experience is required to

identify the proper location for the ventriculostomy safely.

behind the dorsum sella and just in front of thebasilar bifurcation. Using a blunt technique tocreate the stoma would be prudent in such a case.

Second, hydrocephalus resulting from tumorobstruction may be relatively acute in onset.When this is the case, the floor of the third

ventricle is often opaque and nonattenuated. Thismakes penetration more difficult and invariablyrequires a sharper technique without visualization

of the underlying neurovascular structures. This‘‘blind’’ fenestration through a thick floor can behazardous and should not be undertaken by thenovice endoscopist. Third, ETV combined with

biopsy of posterior third ventricular tumorscannot be done through the standard ETV burrhole. The ideal trajectory for the anterior third

ventricular floor is just anterior to the coronalsuture, whereas that for the posterior thirdventricle is just posterior to the hairline or 8 cm

behind the nasion and 3 cm from the midline.The foramen of Monro is sometimes expandedenough to allow access to both parts of the third

ventricle through a single burr hole placedbetween the coronal suture and the hairline. Thismay be assessed before surgery by close in-

spection of the MRI.A comment should be made about the use of

flexible fiberscopes. It is true that with a rigid

endoscope, it is difficult to perform an ETV andbiopsy of a posterior third ventricular tumorthrough the same burr hole. Nonetheless, webelieve that a rigid endoscope is preferable to

using a flexible scope insofar as it can beimpossible to be sure where the back of theflexible endoscope is. The risk of complications in

most hands is consequently higher with flexibleendoscopes. It is our personal practice not to useflexible endoscopes in the head.


Endoscopic tumor biopsy

There are no studies showing that endoscopictumor biopsy is any better than stereotactic needlebiopsy. Anecdotally, however, there seem to be

definite advantages to this technique. Endoscopicbiopsy should be considered when the tumor iseither within the ventricle or at least presenting to

the ventricular surface. The advantages are asfollows:

1. Direct visualization of the tumor allows moreaccurate and safer sampling. A region for bi-opsy can be chosen under endoscopic vision,

and vessels can be avoided.2. The specimen obtained is larger and not

subjected to as much mechanical artifact. It

does not need to be sucked up a needle ormanipulated.

3. Any resultant bleeding can be stopped by

either coagulation or packing under directvisualization.

4. If the tumor is relatively avascular, it may beremoved totally by endoscopic techniques.

5. Other procedures can be performed at thesame operation (eg, ETV, septum pellucidot-omy).

Examples of tumors that are typically

approachable endoscopically are colloid cysts,subependymal giant cell astrocytomas, other

Fig. 4. Cystic craniopharyngioma in an elderly woman

who presented with high intracranial pressure. The cyst

was fenestrated endoscopically, and a septum pellucid-

otomy was performed. The patient’s symptoms re-

solved, and she did not require shunting.

low- (including tectal gliomas) and high-grade

gliomas [Fig. 5], central neurocytomas, subepen-dymomas, and choroid plexus tumors and cysts[8]. Most of these tumors are relatively avascular,

and hemorrhage is rarely a problem. The burrhole is made so that the scope enters the ventricleas far from the tumor as possible and so that thescope is directly viewing the tumor rather than

peering from around a corner. The distal ap-proach allows the surgeon to orient himself byidentifying normal anatomic structures before

encountering the abnormal anatomy. Becausemost of the distal part of the scope is within theventricle, it also allows the surgeon to move the

scope in multiple directions more freely withoutdamaging the surrounding normal brain. Startingfrom farther away means that only relatively smallexcursions are necessary within the normal brain

to visualize the entire target. Many surgeons arehesitant to employ endoscopy for tumors becauseof the fear of bleeding. Almost all bleeding can be

controlled with irrigation alone until the bleedingstops. If bipolar and monopolar instrumentsare available, they can be used, but they are

often ineffective. Patience and copious irrigationare the keys.

For tumors that are primarily intra-axial, there

are few advantages of endoscopic biopsy overstandard stereotactic techniques. Some extra-axialtumors that require biopsy, such as those thatmight be seen around the circle of Willis and

suprasellar region, are well suited to this tech-nique, however. For these tumors, stereotactictechniques are too dangerous and standard open

Fig. 5. Biopsy of a tumor in the posterior wall of the

third ventricle with grasping forceps. Bleeding after such

a biopsy can usually be managed with irrigation alone.


craniotomy is extremely invasive. Endoscopicbiopsy allows a minimally invasive approach withexcellent visualization and safety. The opening ismade through a supraorbital brow incision and

a small frontal craniotomy. The suprasellar regionis then approached using standard microsurgicaltechnique, and the scope is introduced when the

cisterns have been opened. Biopsies should betaken with straight or angled cup forceps passedalong the scope and not down a working channel.

The instruments should always be passed ahead ofthe endoscope so that their progress can beobserved. Blind passage of the instruments into

the endoscope’s field of view risks damagingstructures located behind the tip.

Endoscopic intraventricular tumor resection

Not all intraventricular tumors should be

approached endoscopically. The ideal tumor forendoscopic consideration has the following char-acteristics:

1. Moderate to low vascularity

2. Soft consistency

3. Less than 2 cm in diameter [9]4. Associated secondary hydrocephalus5. Histologically low grade6. Situated in the lateral ventricle

Clearly, from this list of desirable features,

almost all patients with colloid cysts are appro-priate candidates for this technique. The resultswith these tumors are often excellent, and in at

least one study, endoscopic resection was clearlysuperior to microsurgery (Figs. 6 and 7) [10].Other tumors that are well suited to total

endoscopic removal are subependymal giant cellastrocytomas around the frontal horn of thelateral ventricle, other low-grade gliomas that

are exophytic into the ventricles, central neuro-cytoma, small choroid plexus tumors, and thepurely intraventricular craniopharyngioma.

There are few articles in the neurosurgical

literature on the application of endoscopy for theremoval of intraventricular tumors. Most of theexperience with endoscopic tumor resection has

specifically addressed the removal of colloid cysts[10–12]. Whichever technique is used, patient

Fig. 6. Pre-endoscopic (A) and postendoscopic (B) removal of a colloid cyst. Note resolution of the hydrocephalus and

preservation of the normal anatomy.


Fig. 7. (A) Initially narrow view of the foramen of Monro. The choroid plexus is to the left of the image, with the septal

vein located superiorly. (B) The choroid plexus is coagulated with monopolar cautery. The cyst is now clearly visible as

a gray structure through the foramen. (C) View of the foramen of Monro after removal of the colloid cyst. The fornix is

seen to the right of the image, the septal vein superiorly, the thalamostriate vein to the lower left, and the choroid plexus

to the upper left.

selection is controversial. For tumors in the

vicinity of the foramen of Monro, the presenceof hydrocephalus is an absolute indication forsome form of surgical intervention. Headache inthe absence of hydrocephalus can be a sign of

intermittent obstruction of the foramen of Monroor some other part of the ventricular system, andpatients should be informed of the possibility of

insidious onset of hydrocephalus and suddendeath. Another indication for surgery would beradiologic progression either in size or enhancing

characteristics. This is particularly true for giantcell subependymal astrocytomas that are locatedin the frontal horn of the lateral ventricle.Whatever tumor is approached, some basic

principles apply. An approach trajectory that

avoids eloquent structures but allows a good viewof the tumor is essential. Most intraventriculartumors are dealt with in the same fashion. Oncegood visualization is achieved, the outside of the

tumor is coagulated with either monopolarelectrocautery or a laser. Copious irrigation isused to clear blood and debris and to prevent too

much heat from building up inside the ventricle. Ifthere is a cyst, it is opened and drained. Thecontents are removed via suction or piecemeal.

The remaining wall is coagulated and removedpiecemeal. Hemostasis is obtained with copiousirrigation. The scope is withdrawn while inspect-ing the tract for intraparenchymal bleeding. A


small wick of Gelfoam is placed in the corticalopening to prevent CSF leakage from the ventricle.

Endoscopic colloid cyst removalThe colloid cyst is the prototypical intraven-

tricular tumor; as such, we discuss its removal insome detail. The colloid cyst represents the idealintraventricular tumor that can be completely

resected via the endoscopic approach. The goalsof surgery are to extirpate totally the cyst wall, toavoid spillage of the cyst contents, to minimizetrauma to the adjacent structures (ie, fornices,

walls of the third ventricle), to spare venousstructures, and to minimize cortical damage alongthe approach route. The first two goals are more

difficult with the endoscope than with opensurgery, whereas the last three are actually easierto achieve. The lower morbidity and shorter

operating time form the rationale for the endo-scopic approach. In practice, complete cystdrainage and removal can be better accomplished

endoscopically than microsurgically once the skillis mastered. The choice of endoscopy as thetreatment modality should depend on (1) properpatient selection, (2) adequacy of the equipment,

and (3) adequacy of the surgeon’s experience andpreparation.

Patient selectionAlmost all patients with colloid cysts are good

candidates for the endoscopic approach. Weencourage endoscopic removal as the preferredfirst procedure, with the option of craniotomy if

endoscopy is unsuccessful. In fact, a wide range ofmanagement strategies and therapeutic optionsare available to treat patients with colloid cysts.These include observation, shunting of the lateral

ventricles, stereotactic aspiration, open transcor-tical or transcallosal resection, and endoscopicremoval. For a patient who presents nonemer-

gently, such as with an incidentally discoveredcolloid cyst, it is important to discuss all options.Nonetheless, given the risk of acute CSF obstruc-

tion, herniation, and death associated with thesemasses, we recommend definitive treatment by theleast morbid route, which is via the endoscope.

The main technical factor that affects colloid

cyst removal is the density of the cyst contents.This can best be estimated using noncontrast CT.Hypodense or isodense contents are usually fairly

liquid and can be removed through the smallapertures of the endoscope without difficulty [13].A small suction catheter attached to a 10- or 20-

mL syringe may help in this regard. Hyperdense

cysts may have more tenacious contents, whichcan prove more difficult to remove with theendoscope. Cup forceps may need to be used toremove the contents of these denser cysts piece-

meal. On MRI, high signal on T1-weightedimaging correlates with higher cholesterol con-tent [14], thicker consistency, and more difficult

removal.

TechniqueThe patient is positioned straight supine with

the head up 45�. A single burr hole is placed on

the nondominant side 8 cm behind the nasion andapproximately 7 cm lateral to the midline. This isfarther lateral than previously published by the

senior author (C.T.), a modification that hasallowed better removal of the component of thecyst in the roof of the third ventricle. A strip shave

is made for the incision, which is 4 cm long andcoronally oriented over the site of the burr hole.Preparations should be made to preserve the

option of converting to an open craniotomy, witha craniotomy tray and microscope available. If theventricles are small, frameless stereotactic guid-ance is used to guide the placement of the

endoscope sheath and trocar. Otherwise, freehandcannulation of the ventricle with a ventricularneedle precedes placement of the sheath.

A 2- to 5-mm 30� rigid endoscope with at leastone working channel is placed down into theventricle. The boundaries of the foramen of

Monro are appreciated: the fornix above andanteriorly, the choroid plexus and the septaland thalamostriate veins exiting posteroinferiorly,and the anterior nucleus of the thalamus located

below. The foramen of Monro is normally 0.3 to0.8 mm in diameter. In the presence of a colloidcyst, it is variably expanded. The cyst is attached

to the roof of the anterior third ventricle, andits anterolateral surface usually bulges into orthrough the visible part of the foramen. Every

effort is made to preserve all normal structures inremoving a colloid cyst. Generally, this can beaccomplished; however, three structures may be

sacrificed, if necessary, to provide widened accessto the tumor, in descending order of preference:choroid plexus, the thalamostriate vein, and theipsilateral fornix. The first structure is sacrificed

with impunity, but sacrifice of the latter twoshould only be considered if all other options havebeen exhausted. If choroid plexus obstructs the

view of the cyst, it can be coagulated witha monopolar probe and dissected free from theforamen without difficulty. No morbidity attends


this maneuver. The thalamostriate vein may belarge and present an obstacle. Every effort shouldbe made to preserve the vein. Past authors have

warned of potentially devastating consequencesfrom sacrificing a thalamostriate vein, but this hasnot been borne out in subsequent reports by otherauthors or in our own experience. The primary

reason not to sacrifice the vein is that it is far moredifficult to do this safely using the endoscope thanmicrosurgically. The bleeding from the vein can be

daunting. Fortunately, sacrifice of the vein rarelybecomes necessary. Last, although it is notpreferable, a thinned-out ipsilateral fornix may

need to be sacrificed to provide adequate access toa large colloid cyst. Permanent short-term mem-ory loss is sometimes a consequence of thismaneuver. Other structures that may be pitfalls

in this case are the thalamus, the internal cap-sule, and the head of the caudate, all of whichmust not be harmed. The internal capsule runs

in the groove between these other two struc-tures; damage to this structure during endo-scopic procedures has resulted in contralateral

hemiparesis.Once exposed, the surface of the cyst is gently

devascularized with electrocautery. Although a

laser may also be used for this role, the results arenot superior to monopolar electrocautery despitethe greater expense and complexity. The wall ofthe cyst is then pierced with the monopolar probe,

and the contents are removed piecemeal or withsuction attached to a hand syringe once inside thecyst. After it is emptied, the cyst wall is coagulated

and removed piecemeal. Care is taken not toprovide excessive traction on the roof of the thirdventricle, because severe bleeding from the in-

ternal cerebral veins can result. Other authorshave written that a small amount of the cyst wallmay have to be left on the third ventricle roof andthe columns of the fornix, but this has at times

been associated with recurrence. In the seniorauthor’s personal series, complete removal waspossible in 24 of 26 cases; the remaining two

patients underwent transcortical removal of theresidual. From the anterolateral approach de-scribed here, the 30� endoscope affords a view of

the roof of the third ventricle that can help toconfirm complete removal of the cyst wall.

It has not been necessary to place drains at the

time of surgery or to maintain them after surgery.Most authors with experience in this area haveused drains in their early cases and, over time,moved away from using them. A drain may be

considered in a patient in whom the risk of sudden

decompensation because of postoperative hydro-cephalus is believed to be high. Ultimately, it is upto the surgeon to make an individual assessment

on a case-by-case basis. In the senior author’spersonal series, no patient has required placementof a permanent ventricular shunt.

Variations in techniqueEndoscopic removal of colloid cysts using

a flexible fiberscope has been described. Although

this approach is possible, it offers no advantageover rigid endoscopy, has a substantially higherlearning curve, requires more coordination with

an assistant, and is associated with greater risk.This approach is absolutely contraindicated forany surgeon not already familiar with the

fiberscope.The technique described previously is a one-

portal technique. We have found this approachadequate in most cases. From the anterolateral

approach, both fornices and the roof of the thirdventricle can be seen from one side. A two-portalipsilateral technique has been described by

Jimenez [15], but we have not found this to addmuch to the procedure and believe that the addi-tion of more portals goes against the goals of a

minimally invasive approach.

Complication avoidanceCopious irrigation is important throughout the

procedure to maintain good visualization anddissipate heat generated by the electrocautery. Weuse 20-mL syringes to provide pulsed hand-

injected irrigation. This allows good feedback asto how much pressure is being generated as well asrapid adjustment of irrigation depending on local

conditions. This method does necessitate frequentchanging of syringes throughout the case. It isimportant that a good path for egress of CSF bemaintained so that fluid does not build up under

pressure during the procedure. Herniation hasbeen described as a result of induced hydroceph-alus in this setting. A tube attached to the outlet

port, which can be raised or lowered, permits the‘‘pop-off’’ pressure to be varied, allowing theamount of ventricular dilatation to be controlled.

Lactated Ringer’s solution is used as the irrigationsolution. It is slightly hypo-osmolar (276 mOsm)but is physiologically more like CSF than normal

saline. Some authors have described postoperativeconfusion for 24 to 48 hours after procedures inwhich normal saline (308 mOsm) is used as theirrigant. There is no widely commercially avail-

able CSF substitute at present.


Bleeding from the colloid cyst or choroidplexus should be addressed with monopolarcoagulation. All other bleeding should be ad-dressed with irrigation. Almost all bleeding stops

with irrigation alone if the irrigation is continuedlong enough; this may mean continuous irrigationfor as long as 20 minutes. The value of copious

irrigation cannot be overemphasized. With pa-tience, almost any degree of bleeding can con-trolled in this manner. If this is not adequate, in

some cases, bleeding vessels may have to besacrificed with the use of electrocautery. In casesof generalized ooze, a balloon may be inflated

gently in the foramen of Monro or the endoscopeitself may be used to provide focal compression.Other techniques for controlling difficult bleedinginclude irrigation with cool saline (which causes

vasoconstriction), draining the CSF to perform‘‘dry field’’ coagulation, and raising the height ofthe drain until the intracranial pressure exceeds

venous pressure. If nothing stops the bleeding, theprocedure may be need to be converted to an opencraniotomy. In our experience, this last step has

never actually been necessary.

ConclusionsThe endoscopic removal of colloid cysts is one

of the most satisfying of neuroendoscopic pro-cedures. Once some experience is gained, thetechnique is faster and less morbid than the open

approach. Skills obtained working on thesetumors provide the foundation for the techniquesthat allow the removal of other intraventriculartumors.

The principles discussed here can be appliedto any mass that meets the indications for endo-scopic removal. The following points reiterate

the techniques applicable to the removal of intra-ventricular tumors in general:

1. Have the correct instrumentation available.Essential tools are a pair of grabbing forcepsand scissors, a good assistant or a scope-

holding device so you can work with twohands, a coagulation device (either mono-polar or bipolar), a means of irrigating, and

straight and 30� angled scopes.2. Make the ventricular entry as far away from

the tumor as possible. The further the tip of

the scope is from the burr hole, the smalleris the angle required to displace the end ofthe scope a given distance. This allows thesurgeon to angle the rigid scope in different

directions without causing as much brain

injury as would otherwise occur. It also givesthe surgeon a chance to view the tumor froma distance, an important point when there isexcessive bleeding. A good amount of ventri-

cle between the entry point and the tumormeans the scope can be pulled back from thebleeding site without exiting the ventricle.

3. Maintain the ventricular volume to ensurea workable operative field. This is achievedwith continuous irrigation. Of course, it is

imperative to make sure there is an un-impeded egress of fluid so as to preventiatrogenic intracranial hypertension. A chan-

nel on the endoscope leading to an open port,serving as a pop-off valve, may be useful inthis regard.

4. Manage hemorrhage with the following

techniques: copious irrigation usually clearsthe operative field of blood and allow thesurgeon to continue without using the other

techniques. It is important to keep thetemperature of the fluid close to bodytemperature and the composition as close to

CSF as possible. Lactated Ringer’s solution ispreferred because it is the closest to isotonicof the relatively available fluids. Irrigation

settles almost all bleeding eventually, al-though it may take upward of 20 minutes ofcontinuous lavage in some cases. Using thescope or an instrument to tamponade the

bleeder can be effective, especially whencombined with head-up elevation and irriga-tion. The next technique is to attempt direct

coagulation with either a bipolar or mono-polar instrument. This technique can be quitea challenge given the obscuration of the field

with blood, the mobile vessels, the presence ofadjacent vital structures, and the difficulty offinding the exact bleeding point with a singleinstrument. Finally, if all these techniques

fail, the ventricle can be drained of CSF. Thisallows more effective bipolar and monopolarcoagulation without clouding of the operative

field. The obvious downside is the collapse ofthe ventricular walls, resulting in constrictionof the operative field. This can be partially

offset by injecting air into the ventricle, whichkeeps the walls apart if there is a good sealaround the working channels and around the

brain/scope interface.5. Finally, it is imperative to keep in mind the

patient’s best interests. Large and vascularintraventricular tumors are difficult to re-

move by pure endoscopic techniques. The


operation can become long and tedious withmore blood loss and risk to surroundingeloquent brain tissue than a routine micro-

surgical approach. Do not hesitate to aban-don the endoscope if you feel more confidentusing a more familiar technique. Alwaysremember the dictum: minimally invasive

and maximally effective.

Endoscope-assisted microsurgery

Endoscope-assisted microsurgery is the area ofneuroendoscopy that has received the least

attention by neurosurgeons but whose potentialvalue to the neuro-oncologic surgeon is thegreatest. Microsurgery itself evolved to maximize

visualization and minimize retraction; endoscopyallows the neurosurgeon to move another stepfurther toward achieving these goals. With endos-

copy, previously inaccessible or poorly accessibletumors located in the skull base, within narrowcavities, deep to key vascular or neural structures,or around corners in the intracranial space can be

clearly visualized; once visualized, they can beresected. The introduction of the endoscope hasthe potential to revolutionize the approach to

certain tumor types by allowing safe radicalremoval. These techniques are particularly appli-cable to a wide range of troublesome tumors,

including but not limited to sellar tumors, in-cluding pituitary tumors, craniopharyngioma, andRathke’s cleft cysts; clival chordomas; pineal

lesions, intraparenchymal tumors near the brainstem or cranial base; acoustic neuromas; andanteriorly or centrally located posterior fossatumors [16,17]. At the point at which dissection

must halt because of the failure of the microscopeto provide an adequate view, the endoscopeshould be brought into play.

A summary of the advantages of the endo-scope as an adjunct to microsurgery includes thefollowing:

1. Better definition of the normal and patho-

logic anatomy. The endoscope can be usedto clarify the anatomy before and duringtumor removal. Key neural or vascular struc-tures can be identified and thereby spared.

This may be particularly important whenworking around or within the brain stem,between small perforating vessels, or between

the cranial nerves.2. Identification of portions of the tumor

located behind or adherent to vital structures.

Some portions of tumor that are apparentlyinvasive into the brain have brain/tumorinterfaces that can be identified when visual-

ized at more direct angles than is possiblewith the operating microscope alone.

3. Minimization of retraction. The authorsseldom employ any fixed retractors even in

the approach to extremely deep lesions. Theendoscope allows narrow corridors to beused, reducing the need to displace sensitive

structures.4. Assessing adequacy of tumor removal. At the

conclusion of the procedure, endoscopic in-

spection allows a more accurate estimation ofthe completeness of resection or, if completeresection cannot be achieved, documentationof where and how much residual tumor

remains.5. As a teaching tool. The endoscope offers

a superior and often novel view of the

anatomy, which can be beneficial to residents’understanding of the surgical approach.Furthermore, the operating surgeon and

the student share the same view, which isnot generally true even with an operatingmicroscope.

The benefit contributed by adding endoscopy

to a traditional craniotomy cannot be overem-phasized. Tumors frequently extend at acuteangles to the cranial base or to the cortical

surfaces along which the traditional surgicalapproach is made (Fig. 8). Although theseavenues are inaccessible to the microscope, whichrequires a direct line of sight, they are ideal for

endoscopy. The degree of retraction required canfrequently be lessened substantially. Traditionally,all microsurgical approaches can be conceptual-

ized as forming the shape of a cone, with the baseon the surface of the head and the working spaceat the apex. With the use of angled endoscopes,

this limitation can be overcome. At the tip of thecone of visualization and illumination available tothe microscope, 360� of additional view can be

obtained. When working around the brain stemand cranial nerves, the corridor available to themicroscope is often narrow, because extensiveretraction is frequently not an option. The

endoscope allows the surgeon to obtain themaximal access possible via the spaces naturallypresent in the extra-axial compartment.

The endoscope has allowed a conceptualchange in the approach to low-grade or ‘‘benign’’tumors. A specific example of this point may be


Fig. 8. Preoperative (A) and postoperative (B) views of a tumor approached using a purely microsurgical approach.

Note the location of residual tumor. This residuum would still be accessible with the endoscope.

illustrated by the case of craniopharyngioma:failure to achieve a complete resection ultimately

proves disastrous for the patient. In the past, somesurgeons have discouraged radical debulking ofthe hypothalamic portion of these tumors because

of the unacceptable side effects attended bydamage to this vital structure. In fact, poorvisualization is what limits adequate and saferemoval of tumor. With the endoscope, the occult

parts of the tumor can be removed under vision,making this part of the tumor as accessible as thedirectly visualized portion (Fig. 9). Angled instru-

ments allow access to the endoscopically visual-ized parts of the tumor. Because the prognosis fora number of tumors is a function of the adequacy

of tumor removal, the use of the endoscope in thisfashion to obtain the best removal possible and todocument that removal is highly recommended.

Using this rationale, the authors encourage anaggressive approach to the management of low-grade tumors, such as pilocytic astrocytomas,pleomorphic xanthoastrocytomas, and dysem-

bryoplastic neuroepithelial tumors, as well asso-called ‘‘benign tumors,’’ such as craniopharyn-giomas and meningiomas.

Thus far, there are no studies documenting thesuperiority of the endoscopic approach to tumorremoval. Some reports have begun to document

its use in transsphenoidal surgery; surgeonsfamiliar with the improved view offered in thisvenue can easily imagine the same benefits appliedwithin the rest of the cranial cavity.

The endoscope is a powerful ally, but as is thecase with all tools, the surgeon should becomefully familiar with it before using it with patients.

Practice is required to develop the visuomotorskills necessary to guide the tip in and out ofnarrow spaces safely, to watch the video image

while still respecting superficial structures alongthe shaft, and to work with other instruments

while maintaining good visualization with theendoscope. Familiarization with the endoscopic

perspective and a review of the pertinent micro-surgical anatomy are essential before using theendoscope on patients. Used properly, it is an

invaluable adjunct to traditional microsurgery.

Technical notes

1. Guide the endoscope in and out of the fieldunder direct vision. The most dangerousaspect of using the endoscope is the risk ofimpacting structures while introducing the

endoscope. It is important to guide theendoscope by viewing it along the length ofits barrel rather than by watching the image

on the screen. After placing the endoscopeinto the working area, it is essential tocontinue to mind the shaft: if the scope is

not fixed, small barely noticed movements atthe tip can be the result of larger excursions atthe back of the scope, which potentially can

have disastrous consequences. An experi-enced assistant or a microscope with a head-up endoscope display can help in this regard.

2. Use a fixed endoscope holder so that once the

endoscope is in place, the surgeon can workwith both hands. This allows the surgeon touse more complex instruments and also

prevents the endoscope from drifting againstvital structures located superficially along theoperative corridor.

Summary

Neuro-oncology, in all its aspects, provides anideal venue for the application of endoscopy.

The main obstacle to its use has been neuro-surgeons’ lack of familiarity with the techniques


Fig. 9. (A–C) These MRI scans show a midbrain tumor in a 5-year-old girl. The tumor was removed through a small

keyhole craniotomy and a transcallosal subchoroidal approach to the third ventricle. Once the microscope demonstrated

what appeared to be complete removal, the endoscope was passed into the tumor cavity and more tumor was removed.

(D–F ) These MRI scans show complete tumor removal. The pathologic finding was a benign xanthoma. There has been

no recurrence after 4 years of follow-up.

and their advantages. As the neuro-oncologic

surgeon uses the endoscope more, endoscopy willtake its rightful place in the surgeon’s arma-mentarium. The advantages of improved visu-alization of intraventricular pathology, better

management of tumor-related hydrocephalus,less morbid biopsies, and minimally invasiveremoval of intraventricular tumors are invaluable

adjuncts to traditional tumor management.

Furthermore, endoscopy is the logical next stepfor surpassing the limitations of traditionalmicrosurgery. Endoscopy is still in its infancy.Rigorous application of the technology is in-

creasingly allowing us to provide our patients themost maximally effective and minimally invasivesurgery possible.


Fig. 9 (continued )

References

[1] Wallner KR, Gonzales M, Sheline GE. Treatment

of oligodendrogliomas with or without post-oper-

ative irradiation. J Neurosurg 1988;68:684–8.

[2] Garcia DM, Fulling KH. Juvenile pilocytic astro-

cytomas of the cerebrum in adults. A distinctive

neoplasm with favorable prognosis. J Neurosurg

1985;63:382–6.

[3] Good CD, Wade AM, Hayward RD, Phipps KP,

Michalski AJ, Harkness WF, et al. Surveillance

neuroimaging in childhood intracranial ependy-

moma: how effective, how often, and for how long?

J Neurosurg 2001;94(1):27–32.

[4] Perneczky A, Fries G. Endoscope-assisted brain

surgery:. part I—evolution, basic concept, and

current technique. Neurosurgery 1998;42:219–25.

[5] Teo C. Endoscopic-assisted tumor and neurovas-

cular procedures. Clin Neurosurg 2000;46:515–25.

[6] Perneczky A, Muller-Forell W, van Lindert E, Fries

G. Keyhole concept in neurosurgery. Stuttgart:

Thieme; 1999.

[7] Ebinger F, Bruehl K, Gutjahr P. Early diffuse

leptomeningeal primitive neuroectodermal tumors

can escape detection by magnetic resonance imag-

ing. Childs Nerv Syst 2000;16(7):398–401.

[8] Oka K, Kin Y, Go Y, et al. Neuroendoscopic

approach to tectal tumors: a consecutive series.

J Neurosurg 1999;91(6):964–70.

[9] Gaab MR, Schroeder HWS. Neuroendoscopic

approach to intraventricular lesions. J Neurosurg

1998;88:496–505.

[10] Lewis AI, Crone KR, Taha J, et al. Surgical

resection of third ventricle colloid cyst. Preliminary

results comparing transcallosal microsurgery with

endoscopy. J Neurosurg 1994;81:174–8.

[11] Abdou MS, Cohen AR. Endoscopic treatment of

colloid cysts of the third ventricle: technical note

and review of the literature. J Neurosurg 1998;89:

1062–8.

[12] Rodziewicz GS, Smith MV, Hodge CJ. Endoscopic

colloid cyst surgery. Neurosurgery 2000;46:655–

62.

[13] El Khoury C, Brugieres P, Decq P, Cosson-

Stanescu R, Combes C, Ricolfi F, et al. Colloid

cysts of the third ventricle: are MR imaging

patterns predictive of difficulty with percutaneous

treatment? AJNR Am J Neuroradiol 2000;21(3):

489–92.

[14] Maeder PP, Holtas SL, Basibuyuk LN, Salford LG,

Tapper UA, Brun A. Colloid cysts of the third

ventricle: correlation of MR and CT findings with

histology and chemical analysis. AJR Am J Roent-

genol 1990;155(1):135–41.

[15] Jimenez DF, editor. Intracranial endoscopic neuro-

surgery. Park Ridge, IL: American Association of

Neurological Surgeons; 1998. p. 132.

[16] Abdullah J, Caemart J. Endoscopic management of

craniopharyngioma: a review of 3 cases. Minim

Invasive Neurosurg 1995;38:79–84.

[17] Cohen AR, Perneczky A, Rodziewciz GS, et al.

Endoscope-assisted craniotomy: approach to the

rostral brainstem. Neurosurgery 1995;36:1128–30.


Index

Note: Page numbers of article titles are in bold face type.

A

Aesculap neuroendoscopes, 23–26

Age factors, in third ventriculostomy outcome,74–76

Aqueduct of Sylvius, occlusion of

CSF entrapment in, 84–85third ventriculostomy for, 49–51, 73–74

Arachnoid cysts, 9–17clinical presentation of, 10

complications of, 16diagnosis of, 10–11frameless stereotactic systems for, 13–14

intraventricular, 10–11laterally projecting, 13locations of, 10

pathology and pathogenesis of, 9–10shunting of, 15suprasellar, clinical presentation of, 10treatment of, 4, 11–16

Arachnoid granulations, obstruction of, 48

Asystole, in third ventriculostomy, 69

Auer Neuro-Endoscope, 23–24

B

Basilar artery, in third ventricleanatomy of, 41

injury of, in third ventriculostomy, 63, 69–70

Biopsy, of tumor, endoscopic, 4, 100–101

Bleeding. See Hemorrhage.

Bradycardia, in third ventriculostomy, 69

Brain, tumors of. See Tumor(s), brain.

Bugbee Wire, for hemostasis, 27

C

Cameras, for endoscopy, 21

Catheter, peel-away, for endoscopy, 22

Cauterization, equipment for, 27–28


doi:10.1016/S1042-3680(03)00100-1

Cerebellumanatomy of, 42–43developmental anatomy of, 35

vermis of, agenesis of, 43

Cerebral aqueduct, anatomy of, 42

Cerebral arteries, injury of, in third

ventriculostomy, 69–70

Cerebrospinal fluidhydraulics of, 45–47

leakage of, in third ventriculostomy, 67

Channel neuroendoscope, 26

Chavantes-Zamorano Neuro-Endoscope, 23–24

Chiari II malformation, 43–44

fourth ventricle malformation in, 43–44third ventricle malformation in, 42third ventriculostomy contraindicated in, 48–49

Chip cameras, for endoscopy, 21

Choroid plexusanatomy of, 37, 57–58

endoscopy of, 3

Choroidal cysts, of lateral ventricle, 39–40

Choroidal fissure, developmental anatomy of, 34

Clarus neuroendoscopes, 26

Coagulation, endoscopic equipment for, 27–28

Codman & Shurtleff neuroendoscopes, 26

Colloid cysts, endoscopic removal of, 4, 103–106

Communicating arteries, posterior, injury of,in third ventriculostomy, 69–70

Computed tomography, in arachnoid cysts, 11

Cook & Codman, cauterization equipment of,27–28

Copocephaly, 39

Corpus callosumagenesis of, 39, 42developmental anatomy of, 34–35

ghts reserved.

106 Index / Neurosurg Clin N Am 15 (2004) 105–108

Craniopharyngioma, versus arachnoid cyst, 11–12

CSF. See Cerebrospinal fluid.

Cyst(s)arachnoid. See Arachnoid cysts.

choroidal, of lateral ventricle, 39–40colloid, endoscopic removal of, 4, 103–106Dandy Walker, 43ependymal, of lateral ventricle, 39–40

solitary, drainage of, 88–89

D

Dandy classification, of hydrocephalus, 45

Dandy Walker cysts, 43

Decq Neuro-Endoscope, 23–24

Diverticula, of lateral ventricle, 39

E

Electrocauterization, equipment for, 27–28

Embryology, of ventricular system, 33–36

Endocrine disorders, in third ventriculostomy, 64

Endoscopyequipment for, 19–31

for arachnoid cysts, 9–17for cancer, 95–109for hydrocephalus, with loculated ventricles,

83–93

history of, 1–7principles of, 19–31third ventriculostomy. See Third

ventriculostomy, endoscopic.ventricular anatomy for, 33–44

Ependymal cysts, of lateral ventricle, 39–40

Equipment, endoscopic, 19–31cameras, 21fiberoptic, 19–20, 25–26

for arachnoid cyst endoscopy, 12–15for coagulation, 27–28for CSF entrapment, 87–88for shunt placement, 27

for third ventriculostomy, 60–61, 71history of, 19–20instruments, 28–29

lasers, 29, 62–63light sources for, 20–21optic angles in, 20

peel-away catheter, 22rigid, 19, 22–26

scope holders, 29

syringomyeloscope, 27video monitors, 21–22

F

Fenestration

of arachnoid cyst, 12–16of multiloculated ventricles, 89–90

Fiberoptic endoscopes, 19–20, 25–26

Fogarty balloon, for endoscopy, 28

Foramen of Monro, anatomy of, 35–36

Forceps, for endoscopy, 28

Fornix, anatomy of, 58

Fourth ventricleanatomy of, 42–43CSF entrapment in, 84–86, 89

malformations of, 43–44

G

Gaab Neuro-Endoscope, 23–24

H

Headache

in arachnoid cyst, 10in slit ventricle syndrome, shunt removal in,

third ventriculostomy for, 53–54

Hematoma, subdural, in third ventriculostomy, 68

Hemorrhagein arachnoid cyst endoscopy, 16

in third ventriculostomy, 63, 69–70intraventricular, hydrocephalus in, third

ventriculostomy for, 52subarachnoid, hydrocephalus in, third

ventriculostomy for, 52ventricular, CSF entrapment in, 85–86

Hippocampus, developmental anatomy of, 33

Holoneural canal dilatation, in hydrocephalus, 85

Holoprosencephaly, 38

Hydranencephaly, 38–39

Hydrocephaluscommunicating, third ventriculostomy

contraindicated in, 47–48

Dandy classification of, 45double compartment, 85etiology of, third ventriculostomy outcome

and, 73, 78

hydromyelic, 85

107Index / Neurosurg Clin N Am 15 (2004) 105–108

in aqueduct of Sylvius occlusion, 49–51

in arachnoid cyst, 10in brain tumors, 51–52in fourth ventricle malformations, 43–44

in infants, third ventriculostomy for, 52–53in intracranial hemorrhage, third

ventriculostomy for, 52in lateral ventricle malformations, 38–40

in third ventricle malformations, 41–42isolated compartments in, 83–93

multiloculated, 89–90

postinflammatory, 85–86shunt-related, 83–85solitary cysts in, 88–89

treatment of, 86–91occlusive, in third ventriculostomy, 63secondary, endoscopic treatment of, 96–99

third ventriculostomy for. See Thirdventriculostomy.

Hypertension, intracranial, endoscopic treatmentof, 3

Hypothalamusanatomy of, 40–41, 58injury of, in third ventriculostomy, 64, 68

I

Infant(s), hydrocephalus in, third ventriculostomy

for, 52–53

Infections, CSF entrapment in, 85–86

Inflammation, CSF entrapment in, 85–86

Instruments, for endoscopy, 28–29

Intracranial hypertension, endoscopic treatmentof, 3

Intraventricular hemorrhage, hydrocephalus in,third ventriculostomy for, 52

L

Lasers, for endoscopy, 29, 62–63

Lateral ventricle

anatomy of, 36–38CSF entrapment in, 84malformations of, 38–40

Light sources, for endoscopy, 20–21

LoculationsCSF entrapment in, 83–90of fourth ventricle, 43

of lateral ventricle, 40

Longstanding overt ventriculomegaly of the adult,50–51

M

Magnetic resonance imagingin aqueduct of Sylvius occlusion, 50in arachnoid cysts, 10–11

in third ventriculostomy outcome assessment,77, 79

Mamillary bodies, anatomy of, 40

Microsurgery, endoscopic-assisted, for tumors,106–107

MINOP Neuroendoscopy System, 23–25

MurphyScope, 26

Myeloscopy, 4

N

Neuroendoscopy. See Endoscopy.

NeuroNavigational endoscopes, 26

NeuroPen endoscope, 26

Neuroview endoscope, 25–26

O

Occipital pole, anatomy of, 38

P

Pneumocephalus, in third ventriculostomy, 67

Porencephaly, 39

R

Rigid endoscopes, 19, 22–26

S

Schizencephaly, 38

Seizures, in arachnoid cyst, 10

ShuntCSF entrapment due to, 83–85

endoscopic equipment for, 27existing, third ventriculostomy outcome

with, 76for arachnoid cyst, 15

removal of, third ventriculostomy for, 53–54

Slit ventricle syndromeentrapped CSF in, 83–85

shunt removal in, third ventriculostomy for,53–54

Subarachnoid hemorrhage, hydrocephalus in,

third ventriculostomy for, 52

108 Index / Neurosurg Clin N Am 15 (2004) 105–108

Subdural hematoma, in third ventriculostomy, 68

Syringomyeloscope, 27

T

Temporal horn, anatomy of, 38

Thalamostriate vein, anatomy of, 37

Thalamus, anatomy of, 41

Third ventricleanatomy of, 40–41, 59

malformations of, 41–42endoscopy contraindicated in, 48–49

small, endoscopy contraindicated in, 48–49

Third ventriculostomyendoscopic

anatomic considerations in, 57–59, 76–78

approaches to, 59–60balloon dilatation in, 61complications of, 63–64, 67–72

asystole, 69

avoidance of, 70–71bradycardia, 69cerebrospinal fluid leak, 67

failure, 70intraoperative management of, 71periventricular injury, 68–69

pneumocephalus, 67subdural hematoma, 68vascular, 69–70ventriculitis, 67–68

dissection in, 61equipment for, 60, 71failure of, 70, 79

for arachnoid cyst, 12for secondary hydrocephalus, 96–99goals of, 77

history of, 3–4, 57patient selection for, 45–55, 70–71

anatomic considerations in, 45–47

biophysical considerations in, 45–47contraindications, 47–49in aqueduct of Sylvius occlusion, 49–51in hemorrhage, 52

in infant hydrocephalus, 52–53in programmed shunt removal, 53–54in tumors, 51–52

positioning for, 68, 71postoperative management in, 71results of, 73–81

anatomic considerations in, 76–78assessment of, 77, 79

clinical features and, 74–78

factors affecting, 73–74failure, 79

saline jet irrigation in, 61–62

saline torch use in, 62stereotactic, 63, 79technique for, 60–63, 68–69, 98–99training for, 71

trajectories for, 59–60with narrow ventricle, 68–69

stereotactic, versus endoscopic procedure, 79

Trigone, anatomy of, 38

Tumor(s), brain, endoscopy in, 95–109history of, 4

in biopsy, 100–101in intraventricular tumor resection, 100–106in microsurgery, 106–107

in secondary hydrocephalus, 51–52, 96–99in ventriculoscopy, 96

U

Ultrasonography, in third ventriculostomy, 63

V

Ventricle(s). See also specific ventricles.anatomy of, 33–44

cerebral aqueduct, 42

developmental, 33–36fourth ventricle, 42–44lateral ventricle, 36–40

third ventricle, 40–41, 59loculated

CSF entrapment in, 83–93

multiloculated, 89–90postinflammatory, 85–86shunt-related, 83–85

solitary cysts in, 88–89treatment of, 86–91

fourth, 43lateral, 40

Ventriculitis, in third ventriculostomy, 67–68

Ventriculomegaly, longstanding, of the adult,50–51

Ventriculoscopy, for tumor evaluation, 4, 96

Ventriculostomy, third. See Thirdventriculostomy.

Video monitors, for endoscopy, 21–22

Date post:	19-Dec-2015
Category:	Documents
Upload:	bivolaru-alin
View:	230 times
Download:	0 times

04.01 - Clinical Neuroendoscopy.pdf

Documents