OPERATIVE NUANCES

ANATOMIC AND SURGICAL BASIS OF THE SPHENOID

RIDGE KEYHOLE APPROACH FOR CEREBRAL ANEURYSMS

Edgar Nathal, M.D.,D.M.Sc.Division of Neurosurgery,Department of CerebrovascularSurgery, National Institute ofNeurology and Neurosurgery“Manuel Velasco Suarez” andInstituto Nacional de CienciasMedicas y de la Nutricion“Salvador Zubiran,”Mexico City, Mexico

Juan Luis Gomez-Amador,M.D.Division of Neurosurgery,Department of CerebrovascularSurgery, National Institute ofNeurology and Neurosurgery“Manuel Velasco Suarez,”Mexico City, Mexico

Reprint requests:Edgar Nathal, M.D.,Division de Neurocirugía, InstitutoNacional de Neurologiay Neurocirugia“Manuel Velasco Suarez,”Insurgentes Sur 3877,Tlalpan, 14269,Mexico D.F., Mexico.Email: nathal@edgar.to

Received, January 22, 2004.

Accepted, June 8, 2004.

IN VASCULAR NEUROSURGERY, the pterional approach has been used primarily for thetreatment of a wide variety of diseases (cavernous angiomas, arteriovenous malformations,etc.), and it is used to take advantage of naturally occurring planes and spaces to exposethe major structures of the circle of Willis. It provides access to the major part of theanterior circulation aneurysms and those occurring in the upper and most proximal part ofthe posterior circulation. Conversely, there has been an increasing interest in the so-calledminimally invasive procedures or keyhole approaches to treating cerebral aneurysms inspecific locations. In this work, we describe a novel keyhole approach that was conceivedto achieve the angle of vision and advantages of the classic pterional approach. Thissurgical approach is based on the anatomic location of the sphenoid ridge and itsrelationship with the sylvian fissure and basal cisterns. The initial incision is made over thehairline behind the external border of the eye on the side selected. A skin and muscularflap is reflected anteriorly, and a small 3 � 3-cm craniotomy is completed around theexternal landmarks of the sphenoid ridge. Further extradural drilling is completed down tothe anterior clinoid process. The dura is opened in a semilunar manner, and the sylvianfissure is opened completely to reach the sylvian and basal cisterns. Thereafter, theaneurysm is dissected and clipped according to the standard microtechnique of theneurosurgeon. A step-by-step description of the approach is offered in this work to facilitatea clear understanding of it. We recommend this approach for treatment of aneurysmsarising at the anterior part of the circle of Willis. It has the advantages of less operative time,fewer days of hospitalization, and similar morbidity and mortality compared with thestandard pterional craniotomy (5.7% on our service for nongiant ruptured aneurysms).

KEY WORDS: Cerebral aneurysm, Cerebrovascular surgery, Keyhole approach, Pterional approach,Sphenoid ridge, Surgical anatomy

Neurosurgery 56[ONS Suppl 1]:ONS-178–ONS-185, 2005 DOI: 10.1227/01.NEU.0000145967.66852.96

The pterional approach has been used inneurosurgery primarily for the treatmentof a wide variety of neurosurgical disor-

ders (2–6, 10, 11, 16). This approach provides theopportunity to treat lesions located in the ante-rior, middle, and upper parts of the posteriorfossae and can be used alone or in combinationwith other approaches to make possible themanagement of extensive and complex tumorsof the cranial base (1, 2, 9, 16–18). In vascularneurosurgery, the pterional craniotomy is usedto take advantage of naturally occurring planesand spaces that can be used to expose the base ofthe brain and the major structures of the circle ofWillis without significant brain retraction (20,23, 26). One of its primary anatomic landmarksis the sphenoid ridge, which represents the bonyfrontier between the frontal and temporal lobes.

During the treatment of vascular diseases, re-moval of the sphenoid ridge down to the ante-rior clinoid process allows a more comfortableopening of the sylvian fissure, first exposing themiddle cerebral artery (MCA), and with furtherdissection through the basal arachnoid cisterns,opening the space to manage the ipsilateral in-ternal carotid artery (ICA), the anterior cerebralartery, the anterior communicating artery com-plex, and even the contralateral ICA. Opening ofthe membrane of Liliequist provides access tothe posterior circulation vessels (31, 36, 37, 40).In other words, the standard pterional approachoffers the opportunity to treat more than 70% ofcerebral aneurysms.

Recently, there has been increasing interestin the keyhole surgery concept, and many ap-proaches have been reported in the literature

ONS-178 | VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 www.neurosurgery-online.com

designed to minimize the exposure and handling of cerebraltissue without compromising the surgical results and the viewof vascular and tumoral lesions (12–15, 19, 32–34, 44). Takinginto account that the pterional approach is probably the mostcommon approach in neurosurgery, we developed the sphe-noid ridge approach as the keyhole concept of the standardpterional craniotomy. This approach is based on a well-planned small craniotomy, extensive bone drilling of the sphe-noid ridge, and wide exposure and opening of the sylvianfissure and basal cisterns to reach aneurysms at the anteriorpart of the circle of Willis. In this work, we describe theanatomic and surgical basis of this approach.

PATIENTS AND METHODS

During the period from May 2000 to July 2003, 380 patientswith cerebral aneurysms were operated on at the Division ofNeurosurgery of the National Institute of Neurology and Neu-rosurgery “Manuel Velasco Suarez” and the National Instituteof Medical Sciences “Salvador Zubiran,” Mexico City. Of thetotal patients, 132 were operated on by the senior author (EN),and 85 patients harboring 98 aneurysms were selected for akeyhole sphenoid ridge approach (Table 1). The surgical tech-nique used in this group is reported here. The patients wereselected according to the position of the aneurysm in the circleof Willis, neurological grade on admission, and the possibleexistence of associated pathological conditions such as cere-bral hematoma, severe edema, angiographic vasospasm, orhydrocephalus. Only patients without associated pathologicalconditions were selected, as well as those with neurologicalGrade 0–III of the Hunt and Kosnik classification (Table 2).Theremaining 47 patients did not fulfill the above criteria andwere operated on via another approach according to the char-acteristics of the aneurysms.

SURGICAL PROCEDURE(see video at web site)

To clarify the anatomic concepts of this approach, we di-vided the surgical steps as follows.

Positioning of the Patient

The patient is positioned supine with the head elevated 10to 15 degrees to ensure that the final position of the head isabove the level of the heart. The neck is maximally extended toincrease the contribution of gravity and venous drainage. Thehead is rotated 45 degrees away from the side of the craniot-omy; however, the rotation of the head may vary according tothe aneurysm location from 15 to 20 degrees for anteriorcommunicating artery aneurysms to 60 degrees to some MCAaneurysms. The head is placed in a horseshoe headrest or athree-point fixation device, such as the Mayfield-Kees headholder.

Surgical Planning

The main sites to be identified on the scalp before applica-tion of the surgical drapes are 1) the frontozygomatic point, 2)the sylvian line, and 3) the pterion. The frontozygomatic pointis located on the orbital rim 2.5 cm above the level at which theupper edge of the zygomatic arch joins the orbital rim and justbelow the junction of the lateral and superior margins of theorbital rim. The sylvian fissure is located along a line thatextends backward from the frontozygomatic point across thelateral surface of the head to the three-quarter point (75% ofthe distance between the nasion and the inion on the midline)(29). The pterion, which represents the site on the externalsurface of the cranium approximating the lateral end of thesphenoid ridge, is located 3 cm behind the frontozygomaticpoint, on the path of the sylvian fissure line and in close

TABLE 1. Location of the aneurysmsa

Location No.

Paraclinoid-ICA 9Dorsal type 5Superior hypophyseal 3Ophthalmic 1

ICA-PComA 25

ICA-AChor 5

ICA-bifurcation 7

Middle cerebral artery 37

Anterior communicating artery 15

a ICA, internal carotid artery; PComA, posterior communicating arterysegment; AChor, anterior choroidal artery segment.

TABLE 2. Comparison between the neurological grade onadmission (Hunt and Kosnik) and the Glasgow OutcomeScale Scorea

H-K gradeGlasgow Outcome Scale score

Total1 2 3 4 5

0 10 1 1 12

1 17 1 18

2 30 2 1 33

3 20 1 1 22

4 0

5 0

Total 77 5 2 0 1 85

a H-K, Hunt and Kosnik.

SPHENOID RIDGE APPROACH

NEUROSURGERY VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 | ONS-179

relationship with the hairline (Figs. 1 and 2). Even when this areashows anatomic variations from one individual to another, thereis a bony depression following the main axis of the sylvian line.If we press firmly over the temporal area beneath the temporalmuscle, this depression is easily detected. If we follow this de-pression caudally until it crosses the hairline on the skin surface,we are actually following the course of the sphenoid ridge up tothe pterion. Thus, the sphenoid ridge is closely related to thevallecula of the sylvian fissure and to the MCA.

Incision

A 4- to 5-cm incision is planned at the level of the hairline,centered at the estimated location of the end of the sphenoidridge (Fig. 2). After the extension of the incision has beenplanned, the hair is strip-shaved 1 cm behind the hairline toavoid hair extending into the surgical field. After the surgicalarea has been wrapped, the skin is infiltrated with xylocainewith 2% epinephrine to avoid excessive bleeding from thesuperficial layers. The skin is incised over the planned line,and skin clips are not used to stop bleeding; rather, the bleed-ing points are cauterized with a bipolar coagulator of Malis toavoid a bulky skin flap. If it is planned to preserve the super-ficial temporal artery, we should locate the course of the arterybefore the incision is made, because, as a result of anatomicvariations, the artery may run rostral or caudal at the borderof the incision in the areolar tissue under the skin. In addition,the skin is dissected from the superficial temporal fascia andretracted with silk stitches or fishhooks.

Dissection of the Superficial Layers andTemporalis Muscle

After the skin and subcutaneous tissue are dissected, thesuperficial temporal fascia and muscle are opened in the same

FIGURE 1. Intraoperative photographs showing anatomic basis (left-sidedcraniotomy). A, external projection of the sphenoid ridge in relation to theexternal orbital rim; the red arrow shows the frontozygomatic point, andthe black arrow points to the sylvian line at the pterion. B, schematicposition of the initial burr hole and craniotomy. The triangle shows theadditional amount of bone removed with a high-speed drill down to theanterior clinoid process. C, intracranial projection of the sphenoid ridgerelated to the external position of the pterion. D, after dural opening, thesylvian veins overlying the sylvian fissure are seen. After dissection of theveins, the sylvian fissure is wide open up to the basal cisterns undermicroscopic magnification to release CSF and expand the surgical field.

FIGURE 2. Intraoperative photographs showing surgical procedure. A,incision planning. The lines show the relationship of the external orbitalrim and malar bone with the incision, the expected direction of the sphe-noid ridge (sylvian line), and the extension of the surgical incision. B, theskin flap is elevated without the use of skin hemostats. The fascia of thetemporal muscle is exposed. C, the muscular and skin flaps have beenretracted anteriorly with fishhooks. The bone is already exposed, and thebony depression of the sphenoid ridge, representing the anatomic andexternal landmarks of the frontal and temporal lobes, is visible. D, initialburr hole centered at the most caudal part of the surgical field centeredover the sphenoid ridge. The additional cuts are made with a craniotome.Sometimes, a small bridge of bone remains over the sphenoid ridge at themost rostral part of the craniotomy and is completed with a bone chisel. E,dural exposure over the frontal and temporal lobes after drilling the sphe-noid ridge. F, the dura mater has been opened in a semicircular mannerand retracted anteriorly. The brain is covered with blood at the subarach-noid space. No edema is present. G, after the aneurysm has been clipped,the dura mater is closed in a watertight manner and the bone fragmentrepositioned with silk sutures or a fixation device. H, the muscular flap isclosed with an absorbable 2-0 suture, and the skin incision is closedaccording to the preference of the neurosurgeon.

NATHAL AND GOMEZ-AMADOR

ONS-180 | VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 www.neurosurgery-online.com

direction as the skin incision, following the natural direction ofthe muscle fibers (Fig. 2). At this point, it is not necessary totake care of the frontal branch of the facial nerve that runsrostral to the point of incision. To date, we have not had anypatient with a postoperative lesion of the nerve. The bone isexposed after subperiosteal dissection, and the hooks are re-positioned to include the skin and muscular layers. The exter-nal representation of the end of the sphenoid ridge is identi-fied. Usually, at this point, the location of the sphenoid ridgeis felt by touch as a small depression over the bone surface,pointing up the separation between the frontal and temporallobes of the brain.

Craniotomy

A single burr hole is made with a craniotome placed at themost caudal aspect of the surgical exposure and centered overthe bony depression representing the sphenoid ridge on theexternal surface of the cranium (Figs. 1C and 2C). After com-pleting the burr hole, one can see the depression of the duracovering the sylvian fissure and the spike-shaped bone point-ing up the sphenoid ridge on the inner part of the cranium.Sometimes the burr hole is deviated slightly over the frontal ortemporal area, primarily if the sphenoid ridge is very promi-nent. In such a case, we can extend the opening with a high-speed drill or bone rongeurs to the opposite side to completelyexpose the dural folding between the frontal and temporalareas. Thereafter, a dural dissector is used to separate the duramater from the bone, especially in elderly patients, to avoidthe possibility of a dural tearing with the craniotome. Thebone cut starts superiorly up to 1 cm before the temporal line,then curving gradually anteriorly and inferiorly. At this point,a large sphenoid ridge is sometimes encountered, and furtherinferior cutting with the craniotome is blocked. In this case, asecond cut is made from the burr hole to the inferior edge ofthe sphenoid wing, symmetrical with the upper cut. In thisway, an oval-shape craniotomy approximately 2 � 2.5 cm indiameter is completed (Fig. 2). Sometimes, it is necessary touse a small bone chisel to complete the cut line over thesphenoid ridge when this structure is prominent to avoid bonefragmentation. After the bone flap is elevated, the dura isfreed in an epidural manner away from the orbital roof andthe sphenoid wing. By use of a high-speed drill with a dia-mond burr, the lateral part of the sphenoid wing is removed asdictated by the individual anatomy in each patient. It is best toremove the maximal amount of the sphenoid wing down tothe superior orbital fissure to enlarge the opening at the levelof the basal cisterns, especially for ICA or anterior communi-cating artery aneurysms. In patients with MCA aneurysms, itis not necessary to extend the drilling down to the anteriorclinoid process.

Dural Opening

The dura is opened in a semilunar manner, providing a2-mm margin from the craniotomy to facilitate the dural clo-

sure, creating a flap that can be retracted anteriorly and se-cured with sutures or small fishhooks. Even when we do notuse spinal drainage in our patients, if a spinal drainage cath-eter was placed for cerebrospinal fluid (CSF) release at thelumbar thecal sac, this is opened at this moment. Otherwise,furosemide or mannitol can be used at the convenience of theneurosurgeon. In a very few patients, we have found that theformer maneuvers were not enough to provide a slack brain;then, a ventricular tap was performed whenever ventricularenlargement was seen on a computed tomographic scan. Fromthis point, the effectiveness of the technique will rely on awide dissection of the sylvian fissure, avoiding the use ofbrain retractors to maximize the working space.

Sylvian Fissure Dissection

At this point of the operation, the operating microscope isbrought into the field, and the surgical view will include thesylvian fissure at the center and the frontal and temporal lobeson both sides (Fig. 1D). The sylvian veins are identified super-ficially. Usually, the veins run over the temporal side of thesylvian fissure, and one to three secondary veins cross fromthe frontal to the temporal side. These veins should be maxi-mally protected. The sylvian fissure is opened superficiallyover the frontal side, to avoid the veins, with the aid of anarachnoid knife or with the tip of a simple hypodermic needle(the preferred choice of the authors). This maneuver cuts thesuperficial layer of the arachnoid and provides access to thesylvian fissure itself. First, we go deeply into a small space,and thereafter, we open the fissure wide. As has been reportedpreviously, the sylvian fissure has many anatomic variations,and one should be familiar with them. The dissection is main-tained totally in the subarachnoid space to minimize trauma toany cortical surface. In cases of MCA bifurcation aneurysms,opening the distal part of the vallecula will expose the MCAbifurcation and the aneurysm itself. Before dissecting the an-eurysm, we should ensure a proximal control from M1, andthe sylvian fissure is opened as far as necessary to visualizethe main trunk of the MCA. Thereafter, the aneurysm is dis-sected and clipped. For deeper aneurysms, there is usually aneed to dissect the anterior portion of the vallecula, where thesylvian cistern communicates with the basal cisterns. Thesecisterns should be opened wide for visualization, dissection,and clipping of the ICA or anterior communicating arteryaneurysms. With proper brain relaxation, opening of the syl-vian fissure requires minimal brain retraction to visualize theparasellar area. In some patients, the brain is tight at thebeginning of the dissection, and if necessary, we can make asmall opening at the carotid cistern to release CSF and thenproceed to complete the sylvian fissure opening. After dissec-tion of the basal cisterns, we can reach the lamina terminalis toopen it and facilitate a complete relaxation of the brain. Withall these maneuvers, switching from a keyhole approach to alarger craniotomy has never been necessary.

SPHENOID RIDGE APPROACH

NEUROSURGERY VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 | ONS-181

Dural Closure and Bone Flap Placement

After the aneurysm is visualized, dissected, and clipped, aproper hemostasis is verified. The dura mater is closed tight ina water-seal manner with a 4-0 nylon suture. If necessary, asmall piece of muscle is used to seal small leakages of CSF. Thebone flap is fixed with silk, miniplates, or Craniofix (AesculapCo., Tuttlingen, Germany), and the temporal fascia and mus-cle are closed with a 2-0 absorbable suture (Fig. 2).

Skin Closure

The skin is closed with nylon (subdermic or continuousanchored suture) or skin staples, and usually, no drainages areleft in place. A small pad is positioned covering the line ofsuture. We remove the sutures 1 week after surgery unless anabsorbable intradermic suture was used for closure (Fig. 3).

DISCUSSION

Recently, the keyhole surgery concept applied to neurosur-gical procedures has gained acceptance among neurosur-geons, with the understanding that the keyhole surgery is notthe miniaturization of any standard technique but rather thenatural evolution into a more precise and refined act. Untilnow, many approaches have been designed to minimize thesurgical exposure. In 1971, Wilson (35) described a limitedexposure for aneurysm and tumor surgery with excellent re-sults. Later, Brock and Dietz (8) described a small frontolateralapproach for cerebral aneurysms using the subfrontal route.At present, one of the most used keyhole approaches fortreating cerebral aneurysms is the subfrontal transciliary ap-proach (synonyms: frontolaterobasal, eyebrow approach) (13–15, 34, 44). This approach provides access to the majority of the

anterior circulation aneurysms and basilar top aneurysms,except those located at the distal anterior cerebral artery, inwhich an interhemispheric craniotomy is commonly used.Some variations to the subfrontal transciliary approach havebeen reported, designed to gain additional angle of vision (27,28); however, in our experience, there are two main limitationsto this approach. At first, in some cases of MCA bifurcationaneurysms, the M1 segment is too long, and the direction ofthe dome is lateral and caudal to the surgical view from thisapproach. In such a case, the amount of dissection required isextensive, and the surgical view and work are in a very deepplane. Furthermore, the occurrence of a thick blood clotaround the aneurysm makes the dissection even more diffi-cult. Second, in cases of posterior communicating artery seg-ment aneurysms, the angle of vision with the subfrontal tran-sciliary approach is more rostral than that obtained with thepterional approach, and visualization of the neck area of theaneurysm is not good, primarily if the dome has a caudaldirection. In these patients, clipping may be difficult, and theneurosurgeon will never be sure whether a remnant dome isleft unless an angiogram is performed during or after thesurgical procedure. This problem could be solved in somepatients with the aid of a surgical neuroendoscope; however,this device is not available in all hospitals and requires previ-ous training to be used effectively in narrow spaces.

Taking these points into consideration, the standard pteri-onal approach provides a much better access to the MCAregion from the M1 segment to the insular branches, so dis-section and clipping of any kind of MCA aneurysm is usuallyin a more superficial plane. Also, visualization of the neck areain cases of ICA-posterior communicating artery aneurysms isfrom a more lateral angle; then, aneurysms with a dome witha caudal direction are better visualized through this route. Themain disadvantages of the standard pterional approach areprimarily functional and cosmetic. At first, most neurosur-geons continue to use extensive hair shaving to expose thesurgical area, even though there has been no clear demonstra-tion that extensive hair shaving is better in preventing skininfection, and this is sometimes psychologically distressing forsome patients. Second, even with the description of manytechniques to avoid damage to the frontal branch of the facialnerve and temporal muscle atrophy, these problems continueto occur frequently in neurosurgical practice; and third, if alarge incision is made over this area, there will be a markedsoft tissue edema, thus increasing the days of hospitalization.Previously, Chehrazi (12) described the temporal transsylvianapproach for anterior circulation aneurysms; however, evenwhen this approach is centered over the sylvian fissure andsolves some disadvantages of the pterional approach, the lineof incision is still located behind the hairline, proceeding fromthe zygomatic arch toward the anterior extent of the so-calledsuperior “keyhole,” which has disadvantages in patients witha receding hairline. Also, not enough emphasis was placed onan extensive drilling of the sphenoid ridge and the anteriorclinoid process, which provides a wide surgical passage for

FIGURE 3. Angiograms and photographs showing illustrative case. Thepatient is a 32-year-old woman with a wide-neck unruptured right MCAaneurysm. A and B, preoperative and postoperative angiograms. C, com-puted tomographic 3-D reconstruction and view of the bone flap in rela-tion to the cranium. The most rostral area corresponds to the drilled space.D and E, patient at the second postoperative day and 3 months after sur-gery. The incision is not cosmetically visible.

NATHAL AND GOMEZ-AMADOR

ONS-182 | VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 www.neurosurgery-online.com

deeply located aneurysms, relying in this way on the use ofmannitol and hyperventilation to attain enough space.

We focused our surgical technique on trying to avoid theseproblems associated with the classic pterional approach whilemaintaining the excellent angle of vision that it affords. Withthe sphenoid ridge approach we describe here, we do not needto shave the hair but only 1 cm behind the natural hairline inthe selected area (Fig. 2). Second, the incision over the fasciaand temporalis muscle is almost straight, following the naturalcourse of the muscular fibers, thus diminishing the possibilityof muscular atrophy. Also, we do not have to make any specialkind of dissection to avoid damage to the frontal branch of thefacial nerve, because it runs almost parallel to the skin inci-sion. Nevertheless, in our opinion, the most important point toconsider in the sphenoid ridge approach is the cisternal dis-section and location of the aneurysm. If we analyze any stan-dard craniotomy, our surgical corridor is through the spacecreated by drilling of the sphenoid ridge, which also creates atriangular prism over which lies the vallecula. Thus, it makesno sense to expose extensive areas of the brain behind thesphenoid ridge while we are actually trying to go anterior andmedial to it. This approach provides the opportunity to exposethe amount of brain necessary to ensure a comfortable open-ing of the sylvian fissure without special instrumentation andto direct the dissection to the basal cisterns to achieve enoughspace for gaining vascular control before facing the aneurysm.Again, it should be pointed out that an unavoidable prereq-uisite before trying to perform this approach is to have enoughskills to dissect and open the sylvian fissure and basal arach-noid cisterns without the need of using bulky brain retractors.The main indications for this approach are 1) all anteriorcirculation aneurysms except distal anterior cerebral arteryaneurysms (Table 1), 2) nonruptured aneurysms, and 3) rup-tured aneurysm in good neurological grade (Grades I–III ofthe Hunt and Kosnik classification). On the other side, themain contraindications to the use of this approach are 1)aneurysms associated with a large blood clot with mass effect,2) paraclinoid or giant aneurysms in which a high-flow bypassis precluded, 3) subarachnoid hemorrhage associated withbrain edema and small ventricular size, and 4) patients in poorneurological grade.

The goal of the surgical treatment for cerebral aneurysms isto occlude the lesion without causing harm to the patient.There was a great improvement in surgical results after intro-duction of the surgical microscope and the description of themicrotechniques that are still in use today (38–43). At present,a new alternative for treatment of cerebral aneurysms is avail-able through endovascular therapy. Also, a new generation ofneurosurgeons is emerging with new concepts based on min-imally invasive neurosurgery (14, 15, 19, 30, 32, 34, 44). In thisway, new technological advances, such as neuroendoscopesand navigators, have become a useful aid in surgical proce-dures; however, many people are afraid that all these ad-vances go against the precise anatomic knowledge that isrequired to treat a vascular lesion. However, because smallapproaches require much better anatomic planning, many

experienced neurosurgeons have moved away from classicapproaches to a keyhole surgery to treat vascular and tumorallesions (12, 24, 25, 34), so the reports about surgical results andnew techniques based on the keyhole concept are increasingworldwide (7, 13–15, 19, 21, 22, 27, 28, 30, 32–34, 44). Perhapsthe two main concerns about the keyhole surgery are, first,whether the surgical field we attain through one of theseapproaches is enough to deal with an intracranial aneurysm,and second, what is the possibility of solving a problem if itoccurs during operation (e.g., intraoperative aneurysm rup-ture, etc.). The response to the first question is related to thesurgical planning itself. First, the patient should be carefullyselected before surgery according to the recommended indi-cations. In addition, it has been our practice to achieve thesoftest brain possible before trying to make any dissectionover the aneurysm. If we work with a tight brain, the resultswill be as bad as with any large craniotomy under this con-dition. If we attain enough space by releasing CSF and usingsome kind of osmotic agent or any other described maneuver,the anatomic structures will be easily confirmed and the vas-cular control will be reached without any problems. At themoment of the aneurysm dissection and vascular control, yourdeep anatomic view should be the same as with any standardcraniotomy. The answer to the second question is related tothe surgical technique. If you have a comfortable surgical viewand vascular control before dealing with the aneurysm, anyunexpected complication, such as an intraoperative rupture,should be managed without space limitations. In this sense,the only condition that should be strictly observed is that thecraniotomy should be at least 3 � 2.5 cm in size (Fig. 2).

While using the sphenoid ridge approach, we are anticipat-ing that the neurosurgeon will take care of the brain tissue toavoid retraction and damage. The success of this and otherkeyhole approaches relies on a wide dissection of the cisternsand fissures to open the natural spaces surrounding the braintissue. It is necessary to acquire laboratory practice to knowthe use of the aspirator as dissector and retractor at the sametime, as well as the use of the bipolar forceps in the samesense. At present, on our service, 60% of cerebral aneurysmsare treated through a keyhole approach (sphenoid ridge, sub-frontal transciliary, micro-interhemispheric). The most impor-tant point is that our morbidity and mortality rates did notincrease with the use of this approach (5.7% for nongiantruptured anterior circulation aneurysms on our service) (Table2). We now have less operative time (3.9 h with this approachversus 5.7 h for a standard pterional craniotomy), fewer daysof hospitalization (7.3 versus 9.2 d), and fewer complicationsrelated to the approach itself, and overall, the patients aremuch more satisfied with the cosmetic results.

REFERENCES

1. Alaywan M, Sindou M: Fronto-temporal approach with orbito-zygomaticremoval: Surgical anatomy. Acta Neurochir (Wien) 104:79–83, 1990.

2. Alleyne CH Jr, Barrow DL, Oyesiku NM: Combined transsphenoidal andpterional craniotomy approach to giant pituitary tumors. Surg Neurol 57:380–390, 2002.

SPHENOID RIDGE APPROACH

NEUROSURGERY VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 | ONS-183

3. Al-Mefty O: Supraorbital-pterional approach to skull base lesions. Neuro-surgery 21:474–477, 1987.

4. Anson JA: Treatment strategies for intracranial fusiform aneurysms.Neurosurg Clin N Am 9:743–753, 1998.

5. Arnold H, Herrmann HD: Skull base chordoma with cavernous sinus involvement:Partial or radical tumor removal? Acta Neurochir (Wien) 83:31–37, 1986.

6. Backlund EO: Pterional approach for orbital decompression. ActaOphthalmol (Copenh) 46:535–540, 1968.

7. Bognar L, Czirjak S, Madarassy G: Frontolateral keyhole craniotomy through asuperciliary skin incision in children. Childs Nerv Syst 19:765–768, 2003.

8. Brock M, Dietz H: The small frontolateral approach for the microsurgical treatmentof intracranial aneurysms. Neurochirurgia (Stuttg) 21:185–191, 1978.

9. Cantore G, Santoro A, Da Pian R: Spontaneous occlusion of supraclinoidaneurysms after the creation of extra-intracranial bypasses using long grafts:Report of two cases. Neurosurgery 44:216–220, 1999.

10. Carmel PW, Antunes JL, Chang CH: Craniopharyngiomas in children.Neurosurgery 11:382–389, 1982.

11. Carrizo A, Basso A: Current surgical treatment for sphenoorbital meningi-omas. Surg Neurol 50:574–578, 1998.

12. Chehrazi BB: A temporal transsylvian approach to anterior circulation an-eurysms. Neurosurgery 30:957–961, 1992.

13. Czirjak S, Szeifert GT: Surgical experience with frontolateral keyhole crani-otomy through a superciliary skin incision. Neurosurgery 48:145–150, 2001.

14. Czirjak S, Nyary I, Futo J, Szeifert GT: Bilateral supraorbital keyhole ap-proach for multiple aneurysms via superciliary skin incisions. Surg Neurol57:314–324, 2002.

15. Dare AO, Landi MK, Lopes DK, Grand W: Eyebrow incision for combinedorbital osteotomy and supraorbital minicraniotomy: Application to aneurysmsof the anterior circulation—Technical note. J Neurosurg 95:714–718, 2001.

16. Day AL: Aneurysms of the ophthalmic segment: A clinical and anatomicalanalysis. J Neurosurg 72:667–691, 1990.

17. Day JD, Giannotta SL, Fukushima T: Extradural temporopolar approach tolesions of the upper basilar artery and infrachiasmatic region. J Neurosurg81:230–235, 1994.

18. Dolenc VV: A combined transorbital-transclinoid and transsylvian approachto carotid-ophthalmic aneurysms without retraction of the brain. ActaNeurochir Suppl (Wien) 72:89–97, 1999.

19. Fukushima T, Miyazaki S, Takusagawa Y, Reichman M: Unilateral inter-hemispheric keyhole approach for anterior cerebral artery aneurysms. ActaNeurochir Suppl (Wien) 53:42–47, 1991.

20. Hua SE, Gluckman TJ, Batjer HH: Middle cerebral artery occlusion afterpterional approach to basilar bifurcation aneurysm: Technical case report.Neurosurgery 39:1050–1054, 1996.

21. Kang SD: Pterional craniotomy without keyhole to supratentorial cerebralaneurysms: Technical note. Surg Neurol 60:457–462, 2003.

22. Kocaogullar Y, Avci E, Fossett D, Caputy A: The extradural subtemporalkeyhole approach to the sphenocavernous region: Anatomic considerations.Minim Invasive Neurosurg 46:100–105, 2003.

23. Nagasawa S, Ohta T, Tsuda E: Surgical results and the related topographicanatomy in paraclinoid internal carotid artery aneurysms. Neurol Res 18:401–408, 1996.

24. Perneczky A, Boecher-Schwarz HG: Endoscope-assisted microsurgery forcerebral aneurysms. Neurol Med Chir (Tokyo) 38[Suppl]:33–34, 1998.

25. Perneczky A, Fries G: Use of a new aneurysm clip with an inverted-springmechanism to facilitate visual control during clip application: Technicalnote. J Neurosurg 82:898–899, 1995.

26. Pritz MB, Chandler WF: The transsylvian approach to middle cerebral arterybifurcation/trifurcation aneurysms. Surg Neurol 41:217–220, 1994.

27. Ramos-Zuñiga R: The trans-supraorbital approach. Minim InvasiveNeurosurg 42:133–136, 1999.

28. Ramos-Zuñiga R, Velazquez H, Barajas MA, Lopez R, Sanchez E, Trejo S:Trans-supraorbital approach to supratentorial aneurysms. Neurosurgery51:125–131, 2002.

29. Rhoton AL Jr: Operative techniques and instrumentation for neurosurgery.Neurosurgery 53:907–934, 2003.

30. Schramm J, Kral T, Clusmann H: Transsylvian keyhole functional hemi-spherectomy. Neurosurgery 49:891–901, 2001.

31. Sugita K, Kobayashi S, Shintani A, Mutsuga N: Microneurosurgery foraneurysms of the basilar artery. J Neurosurg 51:615–620, 1979.

32. Taniguchi M, Perneczky A: Subtemporal keyhole approach to the suprasel-lar and petroclival region: Microanatomic considerations and clinical appli-cation. Neurosurgery 41:592–601, 1997.

33. Tobler WD, Stanley M: Stereotactic resection of brain metastases in eloquentbrain. Stereotact Funct Neurosurg 63:38–44, 1994.

34. van Lindert E, Perneczky A, Fries G, Pierangeli E: The supraorbital keyholeapproach to supratentorial aneurysms: Concept and technique. Surg Neurol49:481–490, 1998.

35. Wilson D: Limited exposure in cerebral surgery. J Neurosurg 34:102–106,1971.

36. Wright DC, Wilson CB: Surgical treatment of basilar aneurysms. Neuro-surgery 5:325–333, 1979.

37. Yamaura A, Ise H, Makino H: Treatment of aneurysms arising from theterminal portion of the basilar artery: With special reference to the radio-metric study and accessibility of trans-sylvian approach. Neurol Med Chir(Tokyo) 22:521–532, 1982.

38. Yasargil MG: The advent of microsurgery. Mt Sinai J Med 64:164–165, 1997.39. Yasargil MG, Fox JL: The microsurgical approach to intracranial aneurysms.

Surg Neurol 3:7–14, 1975.40. Yasargil MG, Antic J, Laciga R, Jain KK, Hodosh RM, Smith RD: Microsur-

gical pterional approach to aneurysms of the basilar bifurcation. SurgNeurol 6:83–91, 1976.

41. Yasargil MG, Boehm WB, Ho RE: Microsurgical treatment of cerebral aneu-rysms at the bifurcation of the internal carotid artery. Acta Neurochir(Wien) 41:61–72, 1978.

42. Yasargil MG, Reichman MV, Kubik S: Preservation of the frontotemporalbranch of the facial nerve using the interfascial temporalis flap for pterionalcraniotomy: Technical article. J Neurosurg 67:463–466, 1987.

43. Yasargil MG, Vise WM, Bader DC: Technical adjuncts in neurosurgery. SurgNeurol 8:331–336, 1977.

44. Zhao J, Wang S, Sui D: Treatment of intracranial anterior circulatory aneu-rysms via keyhole approach [in Chinese]. Zhonghua Yi Xue Za Zhi 81:323–325, 2001.

COMMENTS

Minimally invasive surgery constantly struggles to perfectthe balance between minimizing tissue trauma and

maintaining maximal effectiveness. In aneurysm surgery, ex-posure is everything. No approach should compromise theprincipal goals of surgery: safe dissection of the aneurysm,preservation of the perforating vessels, and complete exclu-sion of the aneurysm from the circulation in a manner thatneither endangers the parent vessels nor leaves a remnant.

The optimal angle of attack for dissection and clippingvaries for each aneurysm on the basis of its location on thevascular tree, the orientation of its neck, and the position ofadjacent neurovascular structures. The authors’ approachseems adequate to access aneurysms of the middle cerebralartery distribution and most aneurysms arising from branchesof the internal carotid artery. It does limit the range of avail-able angles in the axial plane from the frontal trajectory.

For most aneurysms involving the anterior circle of Willisand the upper posterior circulation, we favor variations of theorbitozygomatic approach. Although more invasive, the mod-ified supraorbital and full orbitozygomatic approaches offermuch wider access and a broader range of possible angles ofattack than the keyhole approach. The orbitozygomatic ap-proach is well tolerated, is associated with a low rate ofmorbidity, and is easily mastered. The larger opening and

NATHAL AND GOMEZ-AMADOR

ONS-184 | VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 www.neurosurgery-online.com

more extensive dissection are the costs paid to achieve the bestpossible clipping.

The above being said, the goals of the minimally invasiveapproach are laudable. In selected cases in which the anatomyis favorable, equally good results may well be achievedthrough a smaller opening. The authors’ technique is oneoption, and its efficacy is validated by their good results.

Peter NakajiRobert F. SpetzlerPhoenix, Arizona

In this well-written report, Nathal and Gomez-Amador de-scribe the sphenoid ridge “minimally invasive” approach for

cerebral aneurysms. The described approach has a number ofkey limitations. The limited bony opening (3 � 3 cm) limits theopening of the sylvian fissure and thus limits access and visual-ization of aneurysmal parent vessels and associated perforators.The limited access would in turn increase the tendency toward

more significant brain retraction. The strategy of clip reconstruc-tion of the aneurysm and circumferential study of and oblitera-tion of the aneurysm is limited. In situations in which untowardevents occur, i.e., intraoperative aneurysmal rupture, the surgeonwould be limited in effectively controlling the situation withtemporary clips in a reasonable manner and thus increasing themorbidity of the procedure.

The pterional transsylvian approach has been time-tested in theeffective management of cerebral aneurysms as well as in the sce-narios mentioned above. Interestingly, the incision described by theauthors (5 cm) is not substantially shorter than that of the pterionalapproach, which would add only a few additional centimeters, withno significant difference in the cosmetic outcome.

M. Gazi YasargilLittle Rock, ArkansasSaleem I. AbdulraufSt. Louis, Missouri

Drawing of an aneurysm surgery. Inset on theleft shows the operative entry with Dandy’s“concealed” incision. Inset on the right showsclip placed on the neck of the aneurysm and thecautery shriveling the sac. Drawing by DorcasHager Padget during her work with Dandy,1929 to 1946 (from, Dandy WE: The Brain,Hagerstown, W.F. Prior Co., Inc., 1966).

SPHENOID RIDGE APPROACH

NEUROSURGERY VOLUME 56 | OPERATIVE NEUROSURGERY 1 | JANUARY 2005 | ONS-185