Journal of Neuro-Oncology 69: 209–219, 2004.

Ó 2004 Kluwer Academic Publishers. Printed in the Netherlands.

Endoscopic and minimally invasive microsurgical approaches for treating

brain tumor patients

Behnam Badie1, Nathaniel Brooks1 and Mark M Souweidane2

1
Department of Neurological Surgery, University of Wisconsin, Madison, WI; 2Department of Neurological
Surgery, The Weill Medical College of Cornell University and Memorial Sloan-Kettering Cancer Center, New
York, NY, USA

Key words: endoscope, intraventricular, keyhole, neurosurgery, pituitary

Summary

Recent developments in neuroendoscopy and minimally invasive procedures have greatly impacted the
diagnosis and treatment of brain tumors. In this paper, we will review these innovations and discuss how
they have influenced our approach to the treatment of intraventricular and pituitary tumors. Finally, the
concept of keyhole neurosurgery is illustrated by discussing ‘eyebrow orbitotomy’ approach as an example.
As noninvasive therapeutic alternative become available, future neurosurgeons will be challenged to develop
effective and less invasive surgical approaches for the diagnosis and treatment of patients will brain tumors.

Introduction neoplasms while minimizing injury to normal tissue.

To qualify as a minimally invasive surgery, first
Over the past two decades the practice of neuro- the procedure has to be considered invasive. Thus,
surgery has been heavily influenced by the tech- radiosurgery, which may provide an effective
niques of minimally invasive surgery. In order to treatment for some brain tumors, is not included
compete with nonsurgical treatment modalities in this category. Second, minimally invasive
(such as radiosurgery) neurosurgical techniques procedures have to diminish iatrogenic injury to
are being modified to minimize patient discom- both CNS and nonCNS tissue. Although skull
fort, shorten hospital stay, improve cosmetic re- base techniques have been designed to minimize
sults and allow patients to return to their normal CNS injury by reducing brain retraction and
activities as soon as possible. The interest in this manipulations, such techniques still require sig-
field has been so overwhelming that even an entire nificant soft tissue dissection and are not consid-
journal (i.e. Minimally Invasive Neurosurgery) ered to be minimally invasive based on our
has been devoted to the publication of such definition. Finally, the procedure has to be as
advancements. Despite this enthusiasm, a precise effective as conventional techniques in the treat-
definition of minimally invasive neurosurgery does ment of brain tumors. This efficacy should not
not seem to exist. Grouped with endoscopic, en- only be judged by surgical cure rates (for benign
dovascular and stereotactic surgery, even radio- tumors), but also by the complications and mor-
therapy has been considered to represent bidity of the procedure. Since most techniques
minimally invasive neurosurgery by some authors. described here are still at their infancy, long term
To better focus our discussion on the impact of class I outcome results are not yet available to
such techniques on the field of neuro-oncology, a critically evaluate their efficacy. Thus, this paper
precise definition of minimally invasive neurosur- will focus on minimally invasive and neuroen-
gery is essential. For the purpose of this paper doscopy techniques that have been shown to im-
minimally invasive neurosurgery is defined as: pact the care of patients with brain tumors based
Invasive procedures that effectively treat brain on class II and III data.
210

The objective of this chapter is not to document surgical field provided by this technology has im-
every minimally invasive technique in neurosurgi- proved the neurosurgeon’s ability to dissect and
cal practice, but rather to outline surgical princi- preserve normal CNS structures. Since micro-
ples and tools that have been applied to the scopes provide an image along a linear corridor,
management of select brain tumors. Specifically, the angle of surgical view can be adjusted only to a
we will emphasize the role of neuroendoscopy in certain extent. This may be a limitation when
the treatment of intraventricular and pituitary tu- visualizing deep structures along a long surgical
mors. Finally we will discuss the concept of ‘key- corridor. In such cases, seeing around ‘surgical
hole’ craniotomy in neurosurgical practice and corners’ is only possible by retracting normal
provide an example of such an approach fre- structures. Endoscopes, on the other hand, have
quently used by the authors. lens systems which allow for wide-angle (fish-eye)
viewing. When endoscopes are advanced towards
an object, there is improvement of magnification
Historical background without loss of the depth of field. Furthermore, the
angle of surgical view can be adjusted by selecting
The field of neurosurgery as a specialty was not the appropriate lens system. These characteristics
defined until the end of the 19th century. In the allow visualization of structures around ‘surgical
early days, surgeries had to be tailored to poor corners’ and through small openings.
lighting and lack of magnification. The develop- The essential components of an endoscopic
ment and implementation of the operating micro- armamentarium include an endoscope, a high-
scope in the 1960s allowed the surgeon to operate resolution camera, a xenon light source, a video
through a narrow tunnel to resect deep-seated le- monitor, and compatible instrumentation. A wide
sions. This significant advancement in technology, variety of endoscopes are currently available
however, was not developed or first implemented which adequately satisfy the essential requirements
within the field of neurosurgery. Otolaryngologists of intracranial endoscopy. Two major categories
were first to adopt this technology in their field [1]. of endoscopes are currently employed; the solid
The initial description of the microscope for use in lens endoscope and the fiberoptic endoscope. The
neurosurgery was made in 1968, and despite the former offers a superior image resolution, is reus-
demonstrated benefits of this technology, micro- able, and is nonmalleable. These endoscopes typ-
neurosurgery was not introduced in some depart- ically are available having varying angles of view
ments until the 1980s [2]. with reference to the scope axis (0°, 30°, 70°, and
Application of endoscopy to surgical practice 120°). Alternatively, the fiberoptic endoscope, has
followed a somewhat similar pattern. The majority less image clarity, is disposable, and can be de-
of the technological advancements in the develop- signed to have stearable capacity. The image res-
ment of endoscopes fell into the hands of urologists olution of fiberoptic endoscopes is directly
and gynecologists. In fact both Dandy [4] and dependent upon the number of fiber optic com-
Mixter [3] reported using cystoscopes and ure- ponents within the scope.
thrascopes in their early foray into ventriculoscopy As mentioned, the critical technological
[3,4]. This initial interest in endoscopy, however, advancement allowing for neurosurgical applica-
was hampered by the difficulties encountered with tions of endoscopes has been the reduction in
the use of relatively cumbersome instruments. With equipment size. The diameter of the endoscope can
refinements in endoscope design and advances in vary widely. Currently, the smallest fiberoptic
digital imaging, these limitations have now been diagnostic scope available measures 1.1 mm in
addressed and endoscopes have become an essen- diameter. This device is ideally suited for intralu-
tial tool in minimally invasive neurosurgery. minal catheter techniques including catheter posi-
tioning or diagnostic ventriculography. Most rigid
lens systems measure between 3 and 6.5 mm in
Basic principles of Endoscopy diameter. This variance is usually a function of the
required versatility of compatible instrumentation.
The contribution of microscope to neurosurgery in Ideally, a dedicated neuroendoscopic arsenal
undisputed. Illumination and magnification of would include a variety of endoscopes, each being
 211

applied for a specific procedure. In actuality, most intraventricular tumors, commonly employed
surgeons favor one type of endoscope given the open surgical approaches have a relative increase
types of procedures being performed, budgetary in potential morbidity. However, the location of
constraints, and familiarity with the equipment. intraventricular tumors being within a CSF con-
Proficiency in microscopic techniques is an taining interface affords excellent light and image
essential component of performing neuroendos- transmission, a requisite for endoscopic surgery.
copy. The surgeon has to be prepared to complete
the operation using conventional methods when Patient selection
surgical objectives are not met. Furthermore, grad-
ual introduction of endoscopy as an adjunct to While endoscopic tumor biopsy or resection
conventional techniques will help inexperienced should always be considered as a therapeutic
surgeons gain familiarity with the technology and alternative for patients with intraventricular brain
better understand its limitations. Certain limitations tumors, not all patients are adequate candidates.
of neuroendoscopy have to be carefully considered The potential of using neuroendoscopy for an
prior to planning these surgeries [5,6]. Because most intraventricular brain tumor is predicated upon
endoscopic procedures are performed through small the pathologic considerations, the intended goal of
openings, use of other rigid instruments along the surgery, and the expected need for further surgery.
endoscope may be difficult. Furthermore, in certain Incumbent in this consideration is the attempt to
procedures like endonasal surgery, repeated clean- reasonably predict tumor histology given the
ing of the endoscope is necessary, and the instru- clinical scenario and the radiographic images.
ment has to be held by a surgeon for easy Thus, in patients in whom the disease will be
manipulation. This will limit the number of surgical treated primarily with nonsurgical means (primary
instruments that can be concomitantly used by the CNS germ cell tumors, primary CNS lymphoma,
surgeon’s other hand. Repeated passage and disseminated metastatic disease, and hypotha-
manipulation of the endoscope through the opera- lamic/chiasmatic astrocytomas) endoscopic tumor
tive field may also result in damage to important biopsy can offer a distinct benefit by avoiding a
surrounding structures, such as the fornices in third more extensive intracranial procedure. In addition
ventricular surgery. Finally, lack of stereoscopic to diagnostic sampling, patients with primary
visualization may result in errors in depth percep- cystic tumors (colloid cyst, epidermoid cyst,
tion. Because most of these limitations are due to Rathke’s cleft cyst) are optimal candidates for
poor instrumentation, they can be easily overcome endoscopic procedures given the ease of cyst
with better technology. In the case of intraventric- aspiration and cyst wall ablation or resection.
ular surgery, for example, small tools used for tumor Endoscopic resection of small solid tumors should
manipulation and aspiration are being made to pass also be considered as a therapeutic option, the
through endoscope side ports. Other multifunc- success of which is dependent upon the tumor
tional instruments, such as the SCANLANR consistency, a feature that may be difficult to
Advantageä Badieä Bipolar (Scanlan Interna- predict preoperatively. In addition to tumor
tional, Minneapolis, MN), will be beneficial in biopsy or tumor resection, decompression of tu-
endoscopic procedures by allowing the surgeons to mor cysts is easily accomplished with endoscopic
use only one hand. Despite these limitations, endo- means. Endoscopic cyst fenestration can also be
scopes have so far transformed the art of diagnosis coupled with the accurate placement of catheters
and treatment of certain brain neoplasms such as for intracavitary therapeutic purposes or sequen-
intraventricular and pituitary tumors. tial aspiration. Relative contraindications for a
primary endoscopic procedure includes solid tu-
mors in which the size or composition prohibits
complete removal, and in situations where tumor
Intraventricular tumors biopsy would not predictably obviate the need for
an open microsurgical resection.
Endoscopic surgery for intraventricular brain tu- In addition to procedures directed at the tumor
mors is a logical application of endoscopic tech- mass, endoscopic approaches can offer significant
nology. Because of the central location of advantages in patients with concomitant hydro-
212

cephalus. Many patients will benefit by treating

the tumor manifestations of noncommunicating
and compartmentalized hydrocephalus through
the use of endoscopic third ventriculostomy and
endoscopic septostomy, respectively.

Surgical technique

Surgical planning is critical for successful endo-

scopic tumor procedures. Most critical is selecting
an entry site that offers the most direct intraven-
tricular, linear route to the target. This principle
avoids undue torque on the cortical and intra-
ventricular surfaces. A relatively anterior entry
site with reference to the coronal suture is most
important for lesions situated in the posterior Figure 1. Intraoperative view through a solid lens endoscope.
third ventricle or pineal region [7–9]. Because this A right-sided intraventricular dysembryoplastic neuroepithelial
relative anterior entry site usually lies anterior to tumor (T) is clearly visualized. The surrounding structures,
including the septal vein (sv), the fornix (f), and choroid plexus
the hairline, a forehead crease can be used for the
(cp) are used for orientation.
approach, minimizing any cosmetic concern by
using autologous cranioplasty with harvested bone
chips and a meticulous skin closure using 6–0 Tumor biopsy
monofilament suture. Once the burr hole is
created, it is helpful to create a conical shape to In a series of endoscopic surgery performed for
the outside of the burr hole using a high-speed intraventricular tumors by one of the authors
drill. This simple maneuver allows greater vari- (MMS), 47 of the 65 were performed for tissue
ability in the trajectory and maneuverability of sampling. Tumor location varied and consisted of
the scope. 10 lateral ventricular tumors and 37 third ven-
Normal anatomical landmarks including the tricular tumors. In these patients, the overall sur-
choroid plexus, venous tributaries, foramen of gical objective of tumor sampling was successfully
Monro and the septum pellucidum are used to achieved in 98% (46/47) of cases. Comparing this
establish appropriate lateralization and orienta- result with other large series (n > 20), this repre-
tion (Figure 1). Continuous irrigation is begun to sents the highest diagnostic yield to date for
maintain a clear medium of image transmission, endoscopic tumor biopsy. This result represents a
for enhancing hemostasis, and to maintain ven- significant improvement from most other large
tricular patency. Lactated Ringers solution studies (n ¼ 21–87) with a diagnostic yield ranging
warmed to 37 °C is preferred given the closer from 52% to 63% [10–12]. In their series of 34
approximation in osmolality with cerebrospinal patients with pineal region tumors, Pople et al. [13]
fluid. A simple manual hand-operated syringe also achieved a better than 90% diagnostic yield
system or an automated pump can provide an using an endoscopic technique for tumor biopsy.
irrigation system. Commonly employed is an irri-
gating bipolar apparatus. Regardless of the irri-
gation method, it is critical that a constant purge Tumor resection
be employed to avoid over-insufflating the ven-
tricular system and elevating the intracranial Solid tumor removal, although a logical applica-
pressure. The simplest method for accomplishing tion of endoscopic techniques, is somewhat limited
this involves maintaining patency of a portal sep- due to the inadequacy of compatible instrumen-
arate from the working channel or using a system tation and the small caliber of current endoscopic
in which the diameter of the sheath is greater than portals. The use of piecemeal removal using cup-
the endoscope. ped biopsy forceps entails prolonged operative
 213

times even for relatively small tumors. Instead, the use of multiple endoscopes becomes technically
resection of the solid tumor is principally achieved burdensome and begins to negate the concept of
through the use of aspiration with a variable suc- minimally invasive. The feasibility of endoscopic
tion catheter. Obviously, the success of such a removal of solid tumors within the ventricular
procedure is dependent upon the tumor size and system would logically be expected to improve
composition. The degree of resection relies on the with the advent of compatible instrumentation
ability to define an interface between the tumor designed for tissue ablation such as an ultrasonic
and subjacent white matter. This is defined pri- aspirator.
marily upon visual inspection and the different
tissue densities. In our series of 65 patients, three
underwent total removal of a solid tumor, a third Colloid cyst
ventricular ependymoma, a third ventricular low-
grade glioneuronal tumor, and a lateral ventricular Considering their cystic composition, colloid cysts
dysembryoplastic neuroepithelial tumor. All tu- are ideal lesions for endoscopic excision. Numer-
mors measured less than 2.0 cm in maximal ous publications have established the merits and
diameter (Figure 2). A complete resection was success of using the minimally invasive, endo-
accomplished in all cases with no complications or scopic approach for what has always been con-
need for postoperative shunting. There have been sidered a challenging microsurgical procedure with
no recurrence or need for adjuvant therapy in inherent potential morbidity [8,17–22]. Expectedly,
these three patients, although the follow up period the success of endoscopic colloid cyst removal is
at the time of this review is only 14 months. heavily dependent upon the surgeon’s experience
Endoscopic resection of solid tumors has also and the technique. The operating sheath coupled
been described in small numbers by others [12,14]. to a stereotactic navigational system is often used
Gaab and Schroeder made similar observations in in selecting a precoronal entry site with a trajec-
their experience with endoscopic resection of three tory as tangential to the roof of the third ventricle
solid tumors: a choroid plexus papilloma, a sub- as possible. The 30° endoscope is then inserted into
ependymoma, and a metastatic medulloblastoma the operating sheath with the angle of view direc-
[14]. Those authors, although advocating the use ted toward the roof of the third ventricle. The cyst
of the Nd:YAG laser, similarly recommended wall and any overlapping choroid plexus are gen-
using an endoscopic resection only for tumors less erously coagulated with bipolar diathermy. Sharp
than 2 cm in diameter. In an effort to enhance dissection is used to incise the cyst wall and a
tumor removal some have advocated the use of suction tube is placed within the cyst and aspira-
multiple endoscopes or entry sites to accommodate tion is then gently applied. Caution needs to be
additional instrumentation [15,16]. However, the utilized in applying intraventricular suction. Spe-

Figure 2. Preoperative (left) and postoperative (right) axial MRI scans (FLAIR sequences) from a patient treated with endoscopic
resection of a low-grade glioneuronal tumor. The tumor was completely resected using endoscopic aspiration and the patient did not
require further CSF diversionary procedures.
214

cifically, only when the lumen of the catheter removal with no reported increase in recurrence
approximates the target tissue should suction be rates [18,21]. With regard to incomplete resection,
applied so as to avoid rapidly evacuating CSF. where tumor remnants remain, but are controlled
Cyst contents can be highly varied, and blunt with coagulation, there is no noted higher rate of
dissection and repeated manipulations with the recurrence [8,17,18,21,22]. In fact, in their review
suction catheter are frequently required in most of 20 patients undergoing endoscopic colloid cyst
situations. Once the contents have been emptied, ablation, Hellwig et al. [21] stated in their recent
the cyst wall is thoroughly coagulated using report ‘that parts of the cyst membrane were left
bipolar diathermy. The cyst wall can then be dis- behind in all patients’. During their follow up
sected away from the roof of the third ventricle period ranging from 12 to 117 months, only one
primarily using a rotational grasping forceps. patient had a clinically recognized recurrence
Remnants of the cyst if adherent to large venous 6 years following surgery, this being treated suc-
tributaries are not forcibly removed, but rather cessfully with repeat endoscopic resection. It is
managed with bipolar coagulation. Ventricular logical that the recurrence rate following endo-
drainage is typically used for monitoring intra- scopic treatment remains substantially lower than
cranial pressure on an individual basis depending that seen in simple stereotactic aspiration due to
upon the degree of intraventricular hemorrhage. the combination of thorough cyst evacuation and
Of 15 colloid cysts operated by one of the au- aggressive cyst wall extirpation [23–25]. It would
thors (MMS) using endoscopic approach, only one thus seem that total tumor ablation rather than
cyst was not completely resected. The incomplete removal is as an acceptable goal of endoscopic
resection occurred during the author’s first at- colloid cyst surgery.
tempt at colloid cyst resection. In that particular
case, a fiberoptic endoscope was used with a
coronal burr hole and no navigational guidance. Pituitary tumors
Since, all endoscopic colloid cysts resections have
been performed using a solid lens system with a Since its first description a century ago, trans-
precoronal entry site and navigational, stereotactic sphenoidal pituitary surgery has undergone sig-
guidance. Using this technique, the author has nificant modification. In 1907 Schloffer reported
conducted 14/14 total resections of a colloid cyst the first successful removal of a pituitary tumor
without radiographic recurrence (mean follow through a transnasal approach using an extended
up ¼ 31 months). rhinotomy incision [26]. The transsphenoidal ap-
Although no randomized study between open proach was further modified by surgical pioneers
resection and endoscopic removal of colloid cysts such as, Hirsch, Halstead, and Cushing. After the
has been conducted, Lewis et al. [20] did report on procedure was abandoned for several years, it was
their single-institutional series of 15 patients once again revived by Dott et al. [27]. With
undergoing either procedure. That review con- introduction of operative microscope by Jules
cluded that the endoscopic group (n ¼ 7) had Hardy in 1967, surgical outcomes of transsphe-
shorter surgical times (127 min vs. 206 min; noidal pituitary surgery significantly improved and
P ¼ 0.01), a shorter hospitalization (4.0 days vs. this technique became the primary surgical ap-
9.5 days; P ¼ 0.05), less complications (P ¼ 0.09), proach for sellar tumors [28]. Because of its great
less risk of ventriculoperitoneal shunting, and a success and low morbidity, the procedure under-
shorter interval before returning to work (26 days went only minor modifications until more
vs. 59 days; P ¼ 0.05) compared with the group recently when the endonasal modification was
undergoing an open microsurgical resection introduced.
(n ¼ 8).
Valid concerns that persist with respect to Transnasal approaches to the sella
endoscopic removal of colloid cyst include the
relatively short duration of follow up and the po- A detailed description of the conventional trans-
tential for incomplete resection. With respect to labial and transspetal approaches to the sella can
the former concern, recent series include patients be found elsewhere [29]. Both approaches involve
followed for over 10 years following endoscopic mucosal incisions and submucosal dissection of
 215

the anterior nasal septum, which then allow for series of 100 patients, Zada et al. [39] also reported
complete immobilization of the nasal septum and similar outcomes in patients undergoing the direct
exposure of the sella. While unproven, the sub- endonasal approach using a pure microscopic
mucosal dissection of the septum was incorporated technique. In this authors own comparison of
into the procedure to possibly minimize the risk of transseptal vs. endonasal endoscopic-assisted ap-
infection and meningitis. This extensive mucosal proaches to the sella there was a reduction in the
dissection and soft tissue trauma, however, has operative time, recovery time and postoperative
been a significant drawback to the procedure and pain in patients treated with the endonasal endo-
mostly responsible for postoperative pain and scopic approach [40].
discomfort, and sinonasal complications such as Although transseptal approaches provide wide
septal perforation. In 1987 the modern version of view of the sella and are suitable for tumors with
the direct endonasal approach, which involves an bilateral parasellar extension, the endonasal route
anterior sphenoidotomy without mucosal dissec- offers adequate exposure for resection of most
tion, was described by Griffith and Veerapen [30]. pituitary tumors. Our endoscopic-guided pituitary
Although this approach was used thereafter by operations are performed through only one nostril,
others, it was not until the introduction of usually the one opposite to the side of the tumor.
endoscopy that the endonasal approach became For midline tumors, the wider nasal cavity is se-
even more popular. The use of the endoscope to lected. The endoscope and a frameless navigation
enhance tumor visualization during transsphenoi- system are routinely used for every surgery. Al-
dal pituitary surgery was first described by Guiot though some authors fracture the middle turbinate
in 1963 [31]. It was perhaps the popularization of to reach the sphenoid ostium, we use the larger
endoscopic sinonasal surgery by otolaryngologists space between the inferior turbinate and the nasal
and their collaboration with neurosurgeons that septum to reach the nasopharynx, and as the scope
ultimately led to the development of pure endo- is angulated in a cephalad direction, the posterior
scopic endonasal pituitary surgery. attachment of the middle turbinate is used as a
landmark to localize the sphenoid sinus. The mu-
cosa over the sphenoid sinus is cut using a suction-
Endoscopic endonasal approach monopolar-coagulator and the posterior aspect of
the nasal septum is fractured from the vomer. To
Jho and Carrau were first to introduce the pure minimize injury to the nasal mucosa during tumor
endoscopic technique for resection of pituitary removal, a modified narrow nasal speculum is then
tumors in 1996 [32]. Since then, this technique has introduced. While the endoscope is still used to
been described by a number of surgeons optimize the extent of anterior sphenoidotomy and
throughout the world [33–35]. In Jho’s series of exposure of the sella, the actual tumor removal is
150 endoscopic pituitary operations, surgical performed using a microscope. By freeing the sur-
safety and efficacy were shown to be comparable geon’s hands and providing stereoscopic visuali-
to the microscopic approach [36]. Endoscopic ap- zation, the microscope can reduce the operative
proach reduced patient discomfort as a result of time and allow for easier control of bleeding. After
elimination of postoperative nasal packing and tumor removal, the nasal speculum is removed and
shortened hospital stay [37]. Cappabianca also the septum and middle turbinate are pushed into
reported a decrease in hospital stay from an their normal anatomical orientation. Postoperative
average of 6 days for his transnasal transsphenoi- nasal packing is not used. Since our first reported
dal operations to 3 days for the endoscopic ap- experience with the endoscopic-guided direct en-
proach [34]. This author also reported a decrease donasal route to the sella 4 years ago, we have used
in surgical morbidity as compared to prior his- this approach in nearly 150 patients with pituitary
torical studies of traditional transsphenoidal ap- tumors. We have not experienced any cosmetic
proaches [38]. complications or nasal perforations using this
The reported reduction in postoperative dis- technique.
comfort in patients undergoing the direct endo- Recent surveys of surgeons who had performed
nasal route is not due to the use of the endoscope traditional microscopic approaches to the sella
per se, but due to the surgical approach [36]. In a have shown an improvement in morbidity and
216

mortality after performing 200 operations [41]. least traumatic. Thus, a number of keyhole ap-
Similarly, authors have reported that the learning proaches have been described that provide the
curve for pure endoscopic procedures is quite steep least invasive, yet adequate exposure to reach
and that after a seemingly small number of slow deep-seated skull base tumors [6,44–52]. Technical
operations the speed and efficacy of there tech- details and indications of these keyhole ap-
nique improves [36]. The criticisms of the pure proaches can be found in the references provided.
endoscopic technique are that it increases opera- Some of these approaches, such as the retrosigm-
tive time because the surgeon can only manipulate oid approach to cerebellopontine tumors, are
tools in one hand, or if an endoscope holder is currently used by most oncologic neurosurgeons.
employed, the holder crowds the operative field. Others, like suproorbital approach are not as
Furthermore, the endoscope often requires fre- popular. In this paper we will describe one such
quent cleanings because of debris collecting on the approach which we have found very useful in
lens. Finally, the view provided by endoscopes is resection of anterior cranial fossa tumors.
two dimensional thus anatomical relationships The ‘eyebrow orbitotomy’, also named orbital
may be difficult to ascertain. This later limitation roof craniotomy is a modification of subfrontal
can lead to inadvertent exposure of the cavernous craniotomy through a limited eyebrow incision
sinus during early experience with endoscopes [38]. [44,53–55]. This keyhole approach can provide a
Despite these limitations, it is clear that the use of small corridor into frontal skull base and supra-
endoscopes in pituitary surgery, either as the sole sellar region while limiting temporalis muscle dis-
visualizing instrument or as an adjunct to the section (Figure 4). The operation is performed
microscope, has greatly impacted this field. The through an eyebrow incision. After cutting
wide-view afforded by the endoscope can guide the through the frontalis muscle, the supraorbital
surgeons during such surgeries (Figure 3). In the nerve is dissected. Although this nerve can be
future, further refinement of instrumentation and preserved for small anterior tumors, it is cut in
endoscopes will undoubtedly promote the use of most cases to obtain a larger corridor. After a
endoscopes in pituitary surgery. small bur hole is placed laterally, a limited orbi-
totomy which may extend into the frontal sinus
and orbital roof is performed using a side-cutting
Keyhole Approaches to Brain Tumors drill, an oscillating saw and osteotomes. The mu-
cosa of the frontal sinus is removed completely
The term ‘keyhole surgery’ was first introduced by and the sinus packed with pieces of gelfoam, bone
Yasargil in 1967 in the context of microsurgical wax and sometime abdominal fat. The dura is
techniques used through a ‘Spalte’ (slot) for tissue opened in a standard fashion. For the closure, the
manipulation [42]. Donald H Wilson used this bone is fixed by small titanium plates, the frontalis
term in 1971 to represent ‘‘a swift, simple method muscle is approximated and skin closed with 6–0
of opening and closing’’ [43]. More recently, this nylon sutures. The sutures are removed within
concept has been further developed to include 5 days to minimize scar formation. Although pa-
surgical corridors which provide the shortest dis- tients experience frontalis palsy and hypoesthesia,
tance to the pathological tissue while minimizing these symptoms often resolve in 2–3 months. We
brain retraction and damage to normal tissue [6]. have used the eyebrow orbitotomy as a less inva-
These keyhole approaches do not necessarily sive alternative to fronto-temporal approach to
translate into the small size of the incision or the resect anterior skull base meningiomas or para-
craniotomy, but instead reflect the initial surgical sellar tumors such as pituitary adenomas or cra-
planning that makes every attempt to minimize niopharyngiomas (Figure 5). An even smaller
iatrogenic damage to normal brain structures as a version of this approach can also be used to repair
result of their manipulation. In neuro-oncology frontal fossa defects such as meningoceles.
such approaches are most appropriate for deep-
seated skull base tumors where significant brain Conclusion
retraction may be necessary. Although many cra-
niofacial techniques offer a short access to deep We are in the midst of an incredible ubiquitous era
lesions, these approaches are not necessarily the of technologic advancement driven by the ability
 217

Figure 3. Microscopic (left) and endoscopic (right) view of the sella during direct endonasal approach for resection of a microade-
noma. The so-called fish-eye effect provided by the endoscope can be very helpful in the visualization of sella and parasellar structures.

tions such as radiotherapy are becoming available

to patients with brain tumors. These challenges
have motivated neurosurgeon to develop less
invasive techniques that minimize iatrogenic injury
and patient discomfort. The field of minimally
invasive neurosurgery, however, is at its infancy.
New technologies such as robotics will undoubt-
edly foster this field. It is our responsibility to be at
Figure 4. Schematic drawing of the keyhole ‘eyebrow orbitot-
omy’ used for resection of frontal fossa and parasellar tumors.
the forefront of these innovations. It is also our
responsibility not to lose the site of the purpose
behind these developments. The drive to develop
new and minimally invasive approaches must be
of science and engineering to produce safer and matched by the desire to provide scientific evidence
more effective medical tools. Such innovations that these approaches are as effective (if not more
have not only helped us improve patient care, but effective) as their predecessors in the management
also they have introduced new challenges into our of brain tumor patients.
field. Improvements in diagnostic imaging, for
example, have allowed for earlier diagnosis of References
smaller and often asymptomatic brain tumors.
Neurosurgeons are now confronted with per- 1. Nylen CO: The microscope in aural surgery its first use
forming invasive procedures on ‘normal’ patients. and later developments. Acta Otolaryng. (Stockholm)
Concurrently, new noninvasive therapeutic op- suppl. 116: 226–240, 1954

Figure 5. Preoperative (left) and postoperative (right) axial MRI scans of a recurrent suprasellar pituitary macroadenoma resected
through a left ‘eyebrow orbitotomy’ approach.
218

2. Rand R, Jannetta P: Microneurosurgery: application of 20. Lewis AI, Crone KR, Taha J, van Loveren HR, Yeh HS,
the binocular surgical microscope in brain tumors, intra- Tew Jr JM: Surgical resection of third ventricle colloid
cranial aneurysms, spinal cord disease and nerve recon- cysts. Preliminary results comparing transcallosal micro-
struction. Clin Neurosurg 15: 319–342, 1968 surgery with endoscopy. J Neurosurg 81: 174–178, 1994
3. Mixter W: Ventriculoscopy and puncture of the floor of 21. Hellwig D, Benes L, Bertalanffy H, Bauer BL: Endoscopic
the third ventricle. Bost Med Surg J 188: 277–278, stereotaxy – an eight year’s experience. Stereotact Funct
1923 Neurosurg 68: 90–97, 1997
4. Dandy W: Treatment of non-capsulated brain tumors by 22. Decq P, Le Guerinel C, Brugieres P, Djindjian M, Silva D,
extensive resection of contiguous braintissue. Bull Johns Keravel Y, Melon E, Nguyen JP: Endoscopic management
Hopkins Hosp 33: 189–190, 1922 of colloid cysts. Neurosurgery 42: 1288-1294, discussion
5. Perneczky A, Fries G: Endoscope-assisted brain surgery: 1294–1286, 1998
Part 1 – Evolution, basic concept, and current technique. 23. Mathiesen T, Grane P, Lindquist C, von Holst H: High
Neurosurgery 42: 219–225, 1998 recurrence rate following aspiration of colloid cysts in the
6. Perneczky A, Muller-Forell W, van Lindert E, Fries G: third ventricle. J Neurosurg 78: 748–752, 1993
Keyhole Concept in Neurosurgery: With Endscope- 24. Kondziolka D, Lunsford LD: Stereotactic management of
Assisted Microsurgery and Case Studies. Thieme, Stutt- colloid cysts: factors predicting success. J Neurosurg 75:
gart, New York, 1999 45–51, 1991
7. Souweidane MM, Sandberg DI, Bilsky MH, Gutin PH: 25. Kondziolka D, Lunsford LD: Stereotactic techniques for
Endoscopic biopsy for tumors of the third ventricle. colloid cysts: roles of aspiration, endoscopy, and micro-
Pediatr Neurosurg 33: 132–137, 2000 surgery. Acta Neurochir Suppl (Wien) 61: 76–78,
8. Abdou MS, Cohen AR: Endoscopic treatment of colloid 1994
cysts of the third ventricle. Technical note and review of 26. Schloffer H: Ergolgreiche Operation eines Hypophysentu-
the literature. J Neurosurg 89: 1062–1068, 1998 mors auf nasalem Wege. Wien Klin Wochenschr 21: 621,
9. Robinson S, Cohen AR: The role of neuroendoscopy in 1907
the treatment of pineal region tumors. Surg Neurol 48: 27. Liu JK, Das K, Weiss MH, Laws ER Jr., Couldwell WT:
360–365, discussion 365–367, 1997 The history and evolution of transsphenoidal surgery. J
10. Fukushima T: Endoscopic biopsy of intraventricular Neurosurg 95: 1083–1096, 2001
tumors with the use of a ventriculofiberscope. Neurosur- 28. Hardy J: Transsphenoidal microsurgery of the normal and
gery 2: 110–113, 1978 pathological pituitary. Clin Neurosurg 16: 185–217, 1969
11. Macarthur DC, Buxton N, Punt J, Vloeberghs M, 29. Liu JK, Weiss MH, Couldwell WT: Surgical approaches to
Robertson IJ: The role of neuroendoscopy in the manage- pituitary tumors. Neurosurg Clin N Am 14: 93–107, 2003
ment of brain tumours. Br J Neurosurg 16: 465–470, 30. Griffith HB, Veerapen R: A direct transnasal approach to
2002 the sphenoid sinus. Technical note. J Neurosurg 66: 140–
12. Macarthur DC, Buxton N, Vloeberghs M, Punt J: The 142, 1987
effectiveness of neuroendoscopic interventions in children 31. Guiot G, Rougerie J, Fourestler A, et al.: Une nouvelle
with brain tumours. Childs Nerv Syst 17: 589–594, technique endoscopique: Exploration endoscopiques in-
2001 tracraniennes. Presse Med 71: 1225–1228, 1963
13. Pople IK, Athanasiou TC, Sandeman DR, Coakham HB: 32. Jho H-D, Carrau RL: Endoscopy assisted transsphenoidal
The role of endoscopic biopsy and third ventriculostomy surgery for pituitary adenoma: Technical Note. Acta
in the management of pineal region tumours. Br J Neurochir (Wien) 138: 1416–1425, 1996
Neurosurg 15: 305–311, 2001 33. Kawamata T, Iseki H, Ishizaki R, Hori T: Minimally
14. Gaab MR, Schroeder HW: Neuroendoscopic approach to invasive endoscope-assisted endonasal trans-sphenoidal
intraventricular lesions. J Neurosurg 88: 496–505, 1998 microsurgery for pituitary tumors: experience with 215
15. Jallo GI, Morota N, Abbott R: Introduction of a second cases comparing with sublabial trans-sphenoidal ap-
working portal for neuroendoscopy. A technical note. proach. Neurol Res 24: 259–265, 2002
Pediatr Neurosurg 24: 56–60, 1996 34. Cappabianca P, Cavallo LM, Colao A, Del Basso De Caro
16. Cohen AR: Endoscopic ventricular surgery. Pediatr Neu- M, Esposito F, Cirillo S, Lombardi G, de Divitiis E:
rosurg 19: 127–134, 1993 Endoscopic endonasal transsphenoidal approach: out-
17. Schroeder HW, Gaab MR: Endoscopic resection of come analysis of 100 consecutive procedures. Minim Invas
colloid cysts. Neurosurgery 51: 1441–1444, discussion Neurosurg 45: 193–200, 2002
1444–1445, 2002 35. Nasseri SS, Kasperbauer JL, Strome SE, McCaffrey TV,
18. Rodziewicz GS, Smith MV, Hodge CJ Jr.: Endoscopic Atkinson JL, Meyer FB: Endoscopic transnasal pituitary
colloid cyst surgery. Neurosurgery 46: 655-660, discussion surgery: report on 180 cases. Am J Rhinol 15: 281–287,
660–652, 2000 2001
19. King WA, Ullman JS, Frazee JG, Post KD, Bergsneider 36. Jho HD, Alfieri A: Endoscopic endonasal pituitary
M: Endoscopic resection of colloid cysts: surgical consid- surgery: evolution of surgical technique and equipment
erations using the rigid endoscope. Neurosurgery 44: 1103- in 150 operations. Minim Invasive Neurosurg 44: 1–12,
1109, discussion 1109–1111, 1999 2001
 219

37. Jho H-D: Endoscopic transsphenoidal surgery. J Neuro- 48. Grand W, Landi MK, Dare AO: Transorbital keyhole
Oncol 54: 187–195, 2001 approach to anterior communicating artery aneurysms.
38. Cappabianca P, Cavallo LM, Colao A, de Divitiis E: Neurosurgery 49: 483–484, 2001
Surgical complications associated with the endoscopic 49. Kocaogullar Y, Avci E, Fossett D, Caputy A: The
endonasal transsphenoidal approach for pituitary adeno- extradural subtemporal keyhole approach to the spheno-
mas. J Neurosurg 97: 293–298, 2002 cavernous region: anatomic considerations. Minim Inva-
39. Zada G, Kelly DF, Cohan P, Wang C, Swerdloff R: sive Neurosurg 46: 100–105, 2003
Endonasal transsphenoidal approach for pituitary adenomas 50. Lin Y, Qiu Y: Microanatomy of endoscope-assisted
and other sellar lesions: an assessment of efficacy, safety, and glabellar nasal keyhole approach. Minim Invasive Neuro-
patient impressions. J Neurosurg 98: 350–358, 2003 surg 46: 155–160, 2003
40. Badie B, Nguyen P, Preston JK: Endoscopic-guided direct 51. Paladino J, Pirker N, Stimac D, Stern-Padovan R:
endonasal approach for pituitary surgery. Surg Neurol 53: Eyebrow keyhole approach in vascular neurosurgery.
168-172, discussion 172–163, 2000 Minim Invasive Neurosurg 41: 200–203, 1998
41. Ciric I, Ragin A, Baumgartner C, Pierce D: Complications 52. van Lindert E, Perneczky A, Fries G, Pierangeli E: The
of transsphenoidal surgery: results of a national survey, supraorbital keyhole approach to supratentorial aneu-
review of the literature and personal experience. Neuro- rysms: concept and technique. Surg Neurol 49: 481-489,
surgery 40: 225–237, 1997 discussion 489–490, 1998
42. Yasargil MG: Microneurosurgery, vol. IVB. Thieme, 53. Jho HD: Orbital roof craniotomy via an eyebrow incision:
Stuttgart, New York, 1996, a simplified anterior skull base approach. Minim Invasive
43. Wilson D: Limited exposure in cerebral surgery: Technical Neurosurg 40: 91–97, 1997
Note. J Neurosurg 34: 102–106, 1971 54. Czirjak S, Szeifert GT: Surgical experience with fronto-
44. Steiger HJ, Schmid-Elsaesser R, Stummer W, Uhl E: lateral keyhole craniotomy through a superciliary skin
Transorbital keyhole approach to anterior communicating incision. Neurosurgery 48: 145–149, discussion 149–150,
artery aneurysms. Neurosurgery 48: 347–351, discussion 2001
351–342, 2001 55. Sanchez-Vazquez MA, Barrera-Calatayud P, Mejia-Villela
45. Cristante L, Puchner MA: A keyhole middle fossa M, Palma-Silva JF, Juan-Carachure I, Gomez-Aguilar
approach to large cholesterol granulomas of the petrous JM, Sanchez-Herrera F: Transciliary subfrontal craniot-
apex. Surg Neurol 53: 64-70; discussion 70–61, 2000 omy for anterior skull base lesions. Technical note. J
46. Czirjak S, Nyary I, Futo J, Szeifert GT: Bilateral supra- Neurosurg 91: 892–896, 1999
orbital keyhole approach for multiple aneurysms via
superciliary skin incisions. Surg Neurol 57: 314–323, Address for offprints: Behnam Badie, Department of Neuro-
discussion 323–314, 2002 logical Surgery, University of Wisconsin Hospital and Clinics,
47. Fukushima T, Miyazaki S, Takusagawa Y, Reichman M: Room K3/805, Clinical Science Center, 600 Highland Ave.,
Unilateral interhemispheric keyhole approach for anterior Madison, WI 53792, USA; Tel.: +1-608-263-1411; Fax: +1-
cerebral artery aneurysms. Acta Neurochir Suppl (Wien) 608-263-1728; E-mail: [email protected]
53: 42–47, 1991