Three Dimensional And 2 Dimensional Endoscopic.21 - Visionsense

1116 | VOLUME 65 | NUMBER 6 | DECEMBER 2009 www.neurosurgery-online.com

SURGICAL ANATOMY AND TECHNIQUEJonathan Roth, M.D.Department of Neurological Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Ameet Singh, M.D.Department of Otolaryngology-Head and Neck Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Gurston Nyquist, M.D.Department of Otolaryngology-Head and Neck Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Justin F. Fraser, M.D.Department of Neurological Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Antonio Bernardo, M.D.Department of Neurological Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Vijay K. Anand, M.D.Department of Otolaryngology-Head and Neck Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Theodore H. Schwartz, M.D.Department of Neurological Surgery,and Department of Otolaryngology-Head and Neck Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,New York, New York

Reprint requests:Theodore H. Schwartz, M.D.,Department of Neurological Surgery,Weill Medical College ofCornell University,New York Presbyterian Hospital,525 E. 68th Street,New York, NY 10021.Email: [email protected]

Received, April 6, 2009.Accepted, August 14, 2009.

Copyright © 2009 by theCongress of Neurological Surgeons

Extended endoscopic endonasal approachesare becoming increasingly applied in themanagement of various intracranial pathol -

ogies along the midline cranial base. On oneend of the complexity spectrum lie lesions suchas pituitary adenomas and encephaloceles ofthe cribriform plate, which can be more easilyapproached through the endonasal corridor (5,

13, 16, 39, 43). On the other end, increasinglychallenging lesions such as chordomas, menin-giomas, esthesioneuroblastomas, and suprasel-lar craniopharyngiomas are more demandingto the endoscopic approaches, as they oftenrequire extensive drilling of the cranial base andcan surround or invade neurovascular struc-tures (12, 14, 15, 17, 22, 25–28, 30, 31, 33, 42, 47).However, these same challenges face the sur-geon using open transcranial approaches tosimilar lesions. Despite the availability of bothendonasal and transcranial corridors, littlework has been performed to directly comparetheir applicability and limitations in reachingcranial base targets (6–8).

ABBREVIATIONS: BA, basilar artery; CN, cranialnerve; ICA, internal carotid artery; OCR, optico-carotid recess; ON, optic nerve; OZ, orbitozygo-matic; PS, planum sphenoidale; 3D, 3-dimen-sional; TS, tuberculum sellae; 2D, 2-dimensional;VA, vertebral artery

THREE-DIMENSIONAL AND 2-DIMENSIONALENDOSCOPIC EXPOSURE OF MIDLINE CRANIALBASE TARGETS USING EXPANDED ENDONASALAND TRANSCRANIAL APPROACHES

OBJECTIVE: Endoscopic endonasal approaches provide an access method to the mid-line cranial base. To integrate these approaches into neurosurgical practice, the extentof their anatomic exposure must be compared with that provided by more traditionaltranscranial approaches.METHODS: Ten fresh cadaver heads were studied. Both endonasal and transcranialapproaches to the midline cranial base were performed. The midline cranial base wasdivided into several areas, and the relative exposure provided by each approach wasdescribed and presented in both 2-dimensional and 3-dimensional images. Limitationsand advantages of each approach are discussed.RESULTS: The endonasal approaches achieved a direct and wide exposure of the mid-line extracranial and intracranial cranial base anatomy. The main lateral limitations ofthe endonasal approaches were the optic nerves, lateral cavernous sinus, vidian nerve,internal carotid artery, abducens nerve in Dorello’s canal, jugular tubercle, and hypoglos-sal canals. Limitations of the transcranial approaches were the neurovascular structureswhich lie in the operative corridor and create narrow working spaces.CONCLUSION: The endonasal approaches achieve a direct and wide exposure of the mid-line cranial base bilaterally. Lateral exposure, beyond the cranial nerves and carotid artery,are challenging. Transcranial approaches are limited by the narrow corridors providedby the cranial nerves, and they do not visualize the contralateral paramedian cranial basevery well. Three-dimensional endoscopes augment the spatial orientation and may improvepatient safety and the learning curve for endoscopic approaches to the midline cranial base.KEY WORDS: Cranial base, Endonasal, Endoscopic, Minimal access, Minimally invasive, Skull base,Stereoscopic, 3-dimensional, Transcranial

Neurosurgery 65:1116–1130, 2009 DOI: 10.1227/01.NEU.0000360340.85186.7A www.neurosurgery- online.com

NEUROSURGERY VOLUME 65 | NUMBER 6 | DECEMBER 2009 | 1117

ENDOSCOPIC EXPOSURE OF MIDLINE CRANIAL BASE TARGETS

At present, standard state-of-the-art endoscopes providehigh-resolution 2-dimensional (2D) images. When the surgeonis working in extremely delicate intracranial spaces, where neu-ral and vascular structures are often separated by millimeters,depth perception is crucial to reduce the risk of injury to thesestructures (21). With 2D scopes, depth perception is based onthe surgeon’s knowledge of the spatial orientation and dis-tances between various anatomic structures, size discrepanciesbetween these structures, light and dark shadows, as well asvisual and tactile feedback during the various surgical maneu-vers. However, despite these visual and tactile senses, thesecues can be misleading (21, 49). Most neurosurgeons performsurgical approaches transcranially using a microscope, whichenables 3-dimensional (3D) visualization. Even experiencedsurgeons face a steep learning curve using the 2D endoscope(21). In such a setting, use of 3D endoscopy could help facilitatethe transition from a stereoscopic microscopic view to endo-scopic visualization (48). In this study, we performed a range ofwell-described transcranial cranial base approaches and com-pared the view with the corresponding extended endonasalendoscopic approaches. We then compared the transcranial andendonasal approaches using both a 2D endoscope and a novel3D endoscope.

MATERIALS AND METHODSTen fresh cadaver heads, injected with blue and red latex to the

venous and arterial systems, respectively, were used for anatomic dis-section. The endonasal approaches were performed using 2D 0-degreeand 30-degree endoscopes (4-mm diameter, 18-cm length; Telecam ccu20233103, Telecam camera head 20212134 (single-chip charge-coupleddevice); Karl Storz, Tuttlingen, Germany) and 3D 0-degree and 30-degree endoscopes (4.9-mm diameter, 30-cm length; Visionsense,Petach Tikva, Israel). The 3D endoscope uses a proprietary technol-ogy, incorporating a pair of pupils, a single lens, and a single chipmimicking a compound eye. Similar to an insect eye, the 3D image isfractionated by a matrix of microlenses to right and left images on apixel level. The electronic data composing the multiplexed right andleft channels are separated by the video processor and displayed usinga standard 3D display. The camera (including lens and chip) is about 3mm in diameter and about 15 mm long and is fixed at the tip of the

endoscope shaft (3, 21, 48). The image is projected as a fusion of left andright images on a screen placed directly in front of the surgical team,who are required to wear stereoscopic glasses to view the 3D image.Images were recorded and stored by the Karl Storz Aida system and bythe Visionsense software (version 1.6.03.01).

Transcranial approaches were performed using standard neurosur-gical techniques with microscopic visualization. A Zeiss NC4 micro-scope (Carl Zeiss, Inc., Oberkochen, Germany) was used for dissec-tions. Images were recorded and stored by the same endoscopes andrecording systems mentioned above.

RESULTSThe anatomic regions of the cranial base and the corresponding

endonasal and transcranial approaches are detailed in Table 1.

Endonasal DissectionsThe endonasal approaches all began similarly, as described in

previous publications (10, 41–43). In the cadaver, it is often use-ful to remove the turbinates to provide more exposure, sincethe fixed tissues are firm and not easily compressed or manip-ulated. However, we attempt to avoid resecting the turbinateswhenever possible during our surgical approaches in patients.In brief, using a bilateral approach, the natural ostia of thesphenoid sinus are enlarged in a medial and inferior directionwith curettes, sphenoid punches, and Kerrison rongeurs(Codman/Johnson & Johnson, Raynham, MA). Approximately2 cm of the posterior nasal septum is removed to create a largecavity and to facilitate binostril access to the sphenoid sinus.The entire anterior face of the sphenoid sinus is removed. Thisprovides exposure to the sella, tuberculum sellae (TS), planumsphenoidale (PS), and intrasphenoidal clivus in the midline.Often, the posterior ethmoids must be removed to fully exposethe anterior extent of the PS.

Areas A and B

Endoscopic Approach to Area AThe sellar floor is immediately identified in the center of the

field of view. Visualization of the lateral and medial optico-

TABLE 1. Endonasal and transcranial corridors used to approach each region of the cranial base

Area Anatomic area Endonasal approach Transcranial approach

A Sella, prechiasmatic cistern, Transsphenoidal, transplanum, 1. Pterionalsuprasellar cistern, lamina terminalis transtuberculum 2. Supraciliary

B Postchiasmatic cistern Transplanum, transtuberculum 1. Pterionalwith pituitary transposition 2. Supraciliary

C Basilar tip, interpeduncular cistern Transsphenoidal, transsellar, 1. Subtemporalremoval of posterior clinoids 2. Orbitozygomatic

D Prepontine cistern, basilar artery Transsphenoidal, transclival Middle fossa with anterior petrosectomyE Cavernous sinus, trigeminal nerve Transethmoidal, transsphenoidal OrbitozygomaticF Lower two-thirds of clivus and Transnasal, transclival 1. Presigmoid retrolabyrinthine, transpetrosal

cervicomedullary junction 2. Far lateral


ROTH ET AL.

carotid recesses (OCRs) bilaterally aids in the identification ofthe course of the optic nerves (ONs), as well as the parasellarand paraclival course of the internal carotid arteries (ICAs).The TS and PS are also identified (Fig. 1A). The sellar floor isdrilled off, and the dura is thereby exposed. The TS connecting

the medial OCRs is removed, and the PS is also removedbetween the lamina papyracea. Lateral exposure is defined bythe lamina papyracea forming the lateral borders of the PS, themedial OCR forming the lateral borders of the TS, and the cav-ernous sinuses forming the lateral borders of the sella. The supe-

FIGURE 1. A and B, 2-dimensional (2D) endoscopic images. A, sphenoidsinus anatomy. 1, lateral opticocarotid recess (OCR); 2, medial OCR; 3,internal carotid artery (ICA); 4, optic nerve (ON); 5, intersphenoid septuminsertion on sellar floor; 6, planum sphenoidale (PS). B, sphenoid sinusafter opening of the dura of the PS, tuberculum sellae (TS), and over thepituitary gland (PG). 1, PG; 2, superior intercavernous sinus (cut); 3, opticchiasm; 4, gyrus rectus; 5, lateral OCR; 6, lamina papyracea. C–E, 3-dimen-sional (3D) endoscopic images. C, left side of sphenoid sinus after drilling ofthe optic strut. 1, gyrus rectus; 2, olfactory nerve; 3, ON; 4, clinoid andproximal supraclinoid internal carotid artery; 5, ophthalmic artery; 6, resid-ual optic strut. D, suprasellar area. 1, pituitary stalk; 2, optic chiasm; 3,superior hypophyseal arteries; 4, right posterior communicating artery(PComA); 5, left P1 segment. E, suprasellar area. 1, optic chiasm; 2, laminaterminalis.

A B

C D

E



rior intercavernous sinus is identified, and the dura is openedboth above and below the intercavernous sinus (Fig. 1B).Anatomic structures exposed in this region include the anterioraspect of the optic chiasm and the medial aspect of the ONs, theproximal supraclinoid ICA as well as the several of its branches,such as the ophthalmic, superior hypophyseal, and proximalposterior communicating arteries (Fig. 1, C and D). Beneath thechiasm, the pituitary stalk is exposed. Above the chiasm, theanterior communicating artery complex, including A1, the ante-rior communicating artery, and A2 are visualized, as is the lam-ina terminalis, enabling a suprachiasmatic approach to the thirdventricle (Fig. 1E). The exposure is symmetrical; thus, bilateralstructures are easily approached. The caudal extent of the olfac-tory nerves is visualized, as are the gyri rectus. This approachexposes and enables treatment of pathologies involving the

gyrus rectus, prechiasmatic cistern, suprasellar cistern, medialaspect of both ONs, medial supraclinoid ICA, stalk, and sella.

Endoscopic Approach to Area BFull exposure of the postchiasmatic cistern requires addition

of the pituitary transposition. The diaphragma sellae is tran-sected, and the pituitary gland is dissected free from the lateraldura of the sella (medial wall of the cavernous sinus). The infe-rior hypophyseal arteries are transected (Fig. 2A). This enablestransposition of the pituitary gland superiorly, providing accessto the postchiasmatic cistern (29). The superior hypophysealarteries are preserved with this technique, and the gland is redi-rected anterosuperiorly. This provides exposure to the suprasel-lar area behind the pituitary stalk. The posterior boundary ofthis region is defined by the dorsum sellae (Fig. 2B).

A B

C D

FIGURE 2. A and B, 2D endoscopic images. A, 1, pituitary gland; 2, rightinferior hypophyseal artery; 3, diaphragma sellae orifice. B, anterosuperiortransposition of the pituitary gland. 1, pituitary gland; 2, pituitary stalk; 3,dorsum sellae; 4, supraclinoid ICA; 5, superior hypophyseal artery. C andD,

3D endoscopic images. C, view via a left pterional approach. 1, ONs; 2, opticchiasm; 3, supraclinoid ICA; 4, A1; 5, lamina terminalis; 6, CN3; 7, proxi-mal M1. D, left opticocarotid triangle. 1, ON; 2, ICA; 3, superior hypophy-seal artery; 4, pituitary stalk.


ROTH ET AL.

Transcranial Approach to Areas A and BTo reach both areas, a pterional and supraciliary (eyebrow

incision) approach are performed (37). Generally, bothapproaches yield similar exposures and working space, pro-ducing a wide exposure of the PS, TS, both ONs, prechiasmaticand suprasellar cisterns, both ICAs, including supraclinoid segments (the medial ipsilateral supraclinoid segment isobstructed by the ipsilateral ON, and the contralateral ICA isobstructed by the contralateral ON from above the supraclinoidsegment to just proximal to the bifurcation), both A1s, laminaterminalis, and anterior communicating arteries (Fig. 2C). Thesuprasellar area is typically approached anteriorly between

both ONs. Thus, a more frontal approach is needed, such as inthe supraciliary approach, or when the pterional approachincludes a more medial subfrontal component. A postfixed chi-asm enables a larger exposure of the suprasellar area, whereasa pre-fixed chiasm limits this approach, and a more lateral cor-ridor, such as through the opticocarotid or carotico-oculomotortriangle, is needed. Additionally, lesions immediately posteriorand inferior to the TS may not be visualized well if there is asteep angle between the TS and the sella, necessitating endo-scopic assistance or drilling of the TS.

The interpeduncular cistern and the stalk may be approachedthrough the opticocarotid triangle or the carotico-oculomotor

A B

C D

FIGURE 3. A–D, 3D endoscopic images. A, 1, basilar tip and interpedun-cular cistern area; 2, incidental right superior cerebellar artery (SCA)aneurysm; 3, SCA; 4, P1 segments of posterior cerebral arteries (PCAs); 5,PComAs with perforators; 6, right A1 segment of anterior cerebral artery; 7,mamillary bodies; 8, CN3; 9, pituitary stalk; 10, tuber cinereum. B, left-sidestructures. 1, CN3; 2, PComA and perforators; 3, supraclinoid ICA; 4, ante-rior choroidal artery; 5, superior hypophyseal artery; 6, temporal uncus; 7,

P1 segment of PCA; 8, mamillary bodies. C, interpeduncular cistern andbasilar tip as seen via the right carotico-oculomotor triangle. 1, CN3; 2, ICA;3, ON; 4, mamillary bodies; 5, dolichoectatic basilar artery (BA); 6, P2 seg-ment of PCA; 7, SCA; 8, anterior petroclinoid fold. D, interpeduncular cis-tern and basilar tip seen through the left carotico-oculomotor triangle. Notethe partial obstruction of the BA by the dorsum sellae. 1, dolichoectatic BA;2, SCA; 3, pituitary stalk; 4, dorsum sellae.



triangle, depending on the relative maneuverability of thesestructures (Fig. 2D). In comparison to the endonasal approach,the transcranial approach exposes the prechiasmatic and supra -sellar cisterns and the lamina terminalis with an extensive addi-tional lateral space. The medial side of the ipsilateral supra -clinoid ICA, however, is exposed better via the endonasalapproach, especially in pre-fixed chiasms. Similarly, approachingthe stalk is much easier using the endonasal approach, whichyields a panoramic view of the suprasellar area. Additionally,with the use of the endonasal approach, the lamina terminalis isexposed with no brain retraction, as opposed to the transcranialapproaches. The main limitation of the endonasal approach, asdemonstrated in our dissections, is exposure of the lateralaspects of the ONs and ICAs.

Area C

Endoscopic ApproachAfter exposure of area B, the inferior intercavernous sinus is

transected, and the posterior clinoids and the superior part of

the clivus are removed. The dura on the posterior aspect of theclivus is then opened. This exposes the basilar tip, mamillarybodies, interpeduncular cistern, origin of cranial nerve (CN) 3,proximal posterior cerebral arteries (P1), and superior cerebel-lar arteries bilaterally. The lateral boundaries of the exposureare CN3 and the posterior communicating arteries (Fig. 3, Aand B).

Transcranial ApproachTwo approaches to the basilar tip and interpeduncular cis-

tern are performed: subtemporal and orbitozygomatic (OZ). Inthe subtemporal approach, temporal lobe retraction is neededto achieve sufficient exposure over the tentorial edge, thusexposing P2, CN3, CN4, the ICA and its bifurcation, and theposterior communicating artery. Working through the carotico-oculomotor triangle exposes the interpeduncular cistern fromthe stalk anteriorly to the brainstem posteriorly. The upperexposure is the third ventricular floor (tuber cinereum), includ-ing the mamillary bodies, while the lower limit is bordered bythe anterior petroclinoid fold and CN3 (Fig. 3C).

C D

A B


ROTH ET AL.

The OZ approach exposes the interpeduncular cisternthrough the opticocarotid and carotico-oculomotor triangles.Removal of the anterior clinoid enlarges and improves theexposure. The posterior clinoid may also be partially removedto increase the exposure (Fig. 3D). The main differencebetween the subtemporal and OZ approaches is the angle ofview. In the subtemporal exposure, the view is horizontal andupward, toward the tuber cinereum, whereas, in the OZ expo-sure, the view is horizontal and downward, toward the pre-pontine area.

Comparing these approaches to the endonasal approach, theworking space is limited in the transcranial approaches by theneurovascular structures that form the opticocarotid andcarotico-oculomotor triangles. Temporal lobe retraction isneeded in the subtemporal approach, and frontal lobe retrac-tion is needed in the OZ approach.

The endonasal approach enables a panoramic view of theinterpeduncular cistern and upper basilar artery (BA). In addi-tion, proximal control of the BA is achieved more easily in theendonasal exposure. However, perforators coming off the back

of the BA are difficult to appreciate endonasally but are moreapparent with lateral transcranial approaches.

Area D

Endoscopic ApproachThe remainder of the intrasphenoidal clivus is removed to

expose the uppermost aspect of the intradural retroclivalvenous plexus (10, 31, 42). The dura is opened, exposing theBAs along with the proximal portions of CN6 before they enterDorello’s canal. Exposure is provided for the anterior aspect ofthe pons, the basilar perforators, and the lower part of themembrane of Liliequist. The lateral limits of this exposure are:superiorly, the cavernous sinus; and inferiorly, the ICA(lacerum and petrous segments). The superior limit is formedby the pituitary gland, which can be transposed, as describedabove (Fig. 4, A and B). Caudally, the floor of the sphenoidsinus limits the clival opening.

With the use of an angled endoscope, the fifth nerve originmay be visualized. However, actual tissue manipulation around

E F

FIGURE 4. A, 2D endoscopic image; panoramic view after removal of thedorsum sellae. 1, basilar tip; 2, SCA; 3, P1 segment of PCA; 4, PComAs andperforators; 5, superior hypophyseal arteries; 6, supraclinoid ICA; 7, A1 seg-ment of anterior cerebral artery; 8, mamillary bodies; 9, pituitary stalk; 10,optic nerves, chiasm, and tracts; 11, CN3. B, 3D endoscopic image of basilartip and interpeduncular cistern area. C, 2D endoscopic image using anangled endoscope, looking laterally to the left beyond the clival opening. 1,CN3; 2, P2 segment of PCA; 3, SCA; 4, tentorial edge; 5, CN5; 6, CN6; 7,petrosal vein; 8, CN4. D, 3D endoscopic image of same area as C. E and F,3D endoscopic images. E, middle fossa approach, right anterior petrosectomy.1, tentorial edge; 2, CN4; 3, CN 5; 4, CN6. F, right anterior petrosectomy. 1,tentorial edge; 2, PComA; 3, CN3; 4, CN4; 5, CN5; 6, SCA. G, middle fossaapproach, left anterior petrosectomy. 1, VA; 2, anteroinferior cerebellar artery(AICA); 3, CN6; 4, inferior petrosal sinus.

G



the fifth nerve origin is difficult to perform via this exposure(Fig. 4, C and D).

Transcranial ApproachA subtemporal anterior petrosectomy approach is per-

formed, including incision of the tentorium. This approachexposes the lateral pons and midbrain, with panoramic expo-sure of CN4 along the lateral side of the brainstem up to thecavernous sinus and the lateral aspect of CN5 from its originalong the lateral pons to the cavernous sinus and foramenovale (Fig. 4, E and F). CN6 is exposed from the pon-tomedullary junction, in the prepontine cistern, enteringDorello’s canal along the clivus, toward the cavernous sinus.With removal of the petrous apex, CN6 in Dorello’s canalserves as a medial limit of the exposure. Caudal exposure islimited by the inferior petrosal sinus (Fig. 4G). More caudal andmedial exposure is possible if the pathology creates a wide cor-ridor (such as occurs with a large petroclival meningioma).Midline structures are exposed to a limited degree. Clivusdrilling is limited by CN6, and the lateral aspect of the BA is

exposed. CN3 to CN5 limit the medial exposure, providing anarrow working space between the nerves and along the upperpart of the approach. The ICA is exposed along its petrous andlacerum segments and along the superior and lateral aspects ofthe intracavernous segment.

Area E

Endoscopic ApproachTo fully expose the cavernous sinus and the adjacent branches

of CN5, as well as CN3 and CN4, a more lateral exposure of thesphenoid sinus is facilitated by removing more of the ethmoidsinuses. The bone over the anterior and inferior aspects of thecavernous sinus is removed while preserving the dura. Thebony borders of the cavernous sinus are: anteriorly, the laminapapyracea and the posterior ethmoidal cells; inferiorly, thepterygoid bone; and superiorly, the optic canal. Care must betaken while removing this bone, which may be extremely thinor even absent. This is followed by a dural opening and removalof the venous latex surrounding the ICA. At this stage, CN6 is

C D

A B


ROTH ET AL.

of the ICA are exposed. CN4 is exposed along its course enter-ing the cavernous sinus and overriding CN3. CN6 is exposedlateral to the ICA.

The main conceptual difference between the transnasal andtranscranial approaches to the cavernous sinus is the medial/lateral exposure (8). As the cavernous sinus contains denseneurovascular structures, the surgical approach should use themost direct route to the relevant pathology. Lesions extendingfrom lateral medially (such as a medial sphenoid meningioma)should be approached transcranially. However, lesions invad-ing the medial aspect of the cavernous sinus (such as somepituitary adenomas) would probably be more securely resectedtransnasally (20).

Area F

Endoscopic ApproachExposure of the inferior clivus and cervicomedullary junc-

tion can be achieved completely below the sphenoid sinus,passing only through the nasal cavity parallel to the palate.However, opening the floor of the sphenoid superiorly is use-ful in providing landmarks for orientation. A wide, inferiorlybased U-flap consisting of nasopharyngeal mucosa, preverte-bral musculature, and clival periosteum is created to expose themiddle and lower clivus. The flap extends between the eusta -chian tubes, down to the base of C2. The muscles in this flapinclude the longus capitis, attached to the clivus, and longuscolli, inserting on C1. The pharyngeal tubercle is noted in themidline; it marks the junction between the middle and lower

E F

identified just lateral to the ICA and on the medial side of the V1branch of the trigeminal nerve. The interneural spaces of the lat-eral cavernous sinus, as described by Cavallo et al. (8) andAlfieri and Jho (1), are identified. The nerves CN3, CN6 (over-riding CN V1), and CN V2 form an “S” shape, forming a supe-rior triangular area and a quadrangular area. Inferolateral dis-section reveals V2 just under the cavernous sinus and extendingup to the foramen rotundum. On the upper side, CN3 is identi-fied. Only a short segment of the CN4 may be seen from thisangle, as it passes over CN3 within the cavernous sinus (Fig. 5,A–C). The carotid artery is exposed from the lacerum segment(in the lower lateral area of the exposure, just above the sphe-noid sinus floor) to the clinoid segment (which is obstructed bythe optic canal and medial OCR on the anteromedial aspect,and the optic strut, marked at its base as the lateral OCR, on thelateral aspect). Opening of the optic canal and removal of thebone over the medial OCR exposes the medial side of the cli-noid and supraclinoid segments, as well as the proximal part ofthe ophthalmic artery (Fig. 5, D and E).

Transcranial ApproachAn OZ approach to the cavernous sinus is performed. After

cutting the temporo-orbital ligament, the dura of the middlefossa is peeled off the lateral wall of the cavernous sinus, andan anterior clinoidectomy is performed. The lateral cavernoussinus is exposed, thus revealing the oculomotor, clinoid, supra-and infratrochlear, anteromedial, and lateral triangles (Fig. 5F)(38). The superolateral and medial aspects of the clinoidal seg-ment and the superior and lateral aspects of cavernous segment

FIGURE 5. A and B, 2D endoscopic images of the lateral cavernous sinusbefore (A) and after (B) skeletonization of the nerves. The typical “S” shape,as described by Cavallo et al. (8) is marked (dotted lines). 1, cavernous ICA;2, CN V2; 3, CN6 just medial to CN V1; 4, CN3. C, 3D endoscopic imageof a skeletonized lateral wall of right cavernous sinus. Similar view to B. D,2D endoscopic image after removal of medial wall of cavernous sinus andmedial and lateral OCR. The ICA is exposed, including the cavernous seg-ment (1), the clinoid segment (2), limited by the proximal (3) and distal (4)

dural rings, and the supraclinoid segment (5). The ophthalmic artery (6) isseen, too. E, 3D endoscopic image of the exposed ICA; similar view to D. Theproximal and distal dural rings (*) and ophthalmic artery (**) are indicated.F, 2D endoscopic image of lateral wall of cavernous sinus. 1, supraclinoidICA; 2, tentorial edge; 3, CN3; 4, CN4; 5, gasserian ganglion; 6, CN6; 7, cli-noid triangle; 8, oculomotor triangle; 9, infratrochlear triangle; 10–12,V1–V3 branches of CN5; 13, anteromedial triangle exposing sphenoid sinus;14, anterolateral triangle; 15, superior orbital fissure (circled).



dural segments of CN6 (Dorello’s canals); and at the lowerclivus, the hypoglossal canals (9). The dura is opened in an“H” fashion to prevent injuring of neurovascular structures(mainly the vertebral artery [VA], BA, and CN6).

Intradurally, a panoramic view of the VA, lower two-thirdsof the BA, and proximal anteroinferior cerebellar arteries isachieved. The intradural segment of CN6 is the lateral borderof the intradural working field at the level of the middle clivus(Fig. 6, A and B). In the lower clivus, the jugular tubercleobstructs the jugular foramen and exiting CNs; thus it needs tobe drilled to gain access to the jugular foramen. The anterome-dial parts of the occipital condyles are drilled up to thehypoglossal canals, serving as the lateral limits of the expo-sure in the lower clivus (Fig. 6, C–E). The corresponding brain-stem exposure from rostral to caudal is the anterior pons (lim-ited laterally by CN6), pontomedullary junction, and medulla.CN9–CN11 and CN12 are seen exiting the brainstem; however,the jugular tubercle and hypoglossal canal need to be drilled, asmentioned, to gain access to their contents.

C D

A B

thirds of the clivus. Elevation of soft tissue and exposure of theclivus continue laterally through the thick periosteal attach-ments and along the grooves of the lateral clivus. The floor ofthe sphenoid sinus is then drilled down to the floor of the nasalseptum to provide visualization of the clivus from the floor ofthe sella to the caudal basiocciput.

The clivus is drilled in a rostral to caudal direction, with thecarotid arteries (lacerum segments) and vidian nerves markingthe lateral extent of the dissection. The occipital tubercles arepreserved bilaterally to preserve CN12. All of the bone of theposterior clivus is removed to expose the tectorial membranebehind the middle and lower clivus. This is incised, along withthe external layer of dura from the sella superiorly to the infe-rior retroclival region, revealing the inferior intercavernoussinus superiorly and the basilar venous plexus. The borders ofthe clival opening are: superiorly, the sella; inferiorly, the tecto-rial membrane between the foramen magnum and anteriorarch of C1; laterally, in the upper clivus, the ascending carotidarteries; in the middle clivus, the vidian nerves and the inter-


ROTH ET AL.

Transcranial ApproachApproaching the lower two-thirds of the clivus and cervi-

comedullary junction is achieved, in its upper aspect, by a pre-sigmoid retrolabyrinthine transpetrosal approach. Posteriorretraction of the sigmoid sinus enlarges the exposure, enablingaccess to the lateral midbrain and pons and exposing CN4 inthe ambient and crural cisterns; the lateral and inferior parts ofthe cisternal segment of CN5; and the lateral, superior, andinferior aspects of CN7–CN8 and CN9–CN11 between theirorigins at the brainstem and the jugular foramen. CN6 and themidclival region may be seen through the corridor betweenCN5 and CN7–CN8. However, access is limited by the morelateral neurovascular structures (i.e., CN 5, CN7–CN11, andthe anteroinferior cerebellar artery) (Fig. 6F). The jugular tuber-cle obstructs access to the clivus and VA. The lower clivus isexposed via a far lateral approach (Fig. 6G). After opening ofthe dura, the lateral medulla is revealed; CN9–CN11, includingthe ascending accessory branch, are fully exposed up to thejugular foramen. The VA is exposed between the transverse

foramen of C1 and the premedullary cistern, including the pos-teroinferior cerebellar artery, from its origin to its ascendingsegment, beyond the tonsil. Retraction of the VA exposes thelower clivus below the level of the jugular tubercles.

The main limitations of the various transcranial approachesto the lower clivus are the narrow working spaces between thelower CNs and the jugular tubercle, which limits the workingangle when approaching posterolaterally. The lateral brainstemis exposed; however, working on the anterior brainstem is lim-ited. The transnasal approach enables direct access to the mid-line structures, including the anterior brainstem; however, thelateral working space is limited by the lower CNs.

DISCUSSIONThe field of cranial base surgery has evolved as a result of an

ongoing close collaboration between neurosurgery and otolaryn-gology. The philosophical motivation for the development of theextended transcranial approaches was that additional removal of

E F

FIGURE 6. A–D, 2D endoscopic images. A and B, views after removal ofclivus down to the foramen magnum (A). Magnification of the lower expo-sure (B), the lateral limit being CN6. 1, BA; 2, VAs; 3, AICA; 4, CN6; 5,lower CNs. C, lateral view to the left after removal of the lower clivus. Notethat the jugular tubercle (JT) obstructs the view of CN9–CN11. 1, JT; 2,CN9–CN11; 3, ascending branch of CN11; 4, CN12; 5, VA; 6, posteroinferiorcerebellar artery (PICA); 7, CN7–CN8 complex and arcuate artery. D, rightside of lower exposure after partial drilling of JT. 1, VA; 2, CN12 rootletsaccompanied by PICA; 3, CN 9–CN11 with adjacent choroid plexus; 4,CN7–CN8 complex with arcuate artery; 5, AICA; 6, CN6. E, 3D endoscopicimage of same area as D. F and G, 2D endoscopic images. F, view via a pre-sigmoid retrolabyrinthine posterior petrosectomy. 1, CN5; 2, CN6; 3,CN7–CN8 complex accompanied by the arcuate artery; 4, CN9–CN11 over-riding the JT. G, anterior view via a right far lateral approach. The lowerCN9–CN11 (1) override the JT; 2, CN12 rootlets; 3, PICA; 4, VA; 5, ascend-ing branch of CN11.

G

bone would minimize the degree of brain retraction and potentialbrain trauma. These principles remain valid for paramedian-based pathology. However, pathology arising from the midline isoften surrounded by CNs and by vital vascular structures, whichlimit the working corridors to this region provided by the tran-scranial approaches. In addition, midline pathology often extendsto the both sides of the cranial base, and lateral approaches maystill require significant brain retraction to achieve adequate expo-sure of the contralateral extent of the tumor. In certain cases, theexpansive nature of a lesion may provide a “natural” retractionby posteriorly displacing the brainstem, thus forming a corridorto the contralateral side. In response to these limitations of thetranscranial cranial base approaches, the endonasal approacheswere developed to provide a more direct midline approach tomidline pathology (10, 30, 31, 42). With these approaches, theCNs and ICA lie lateral and superior to the pathology. The epi-center of the pathology can be exposed and decompressed beforeextracapsular dissection of neurovascular structures is required.However, given the relative novelty of the endonasal approaches,the indications and limitations of these approaches, comparedwith the more traditional transcranial approaches, have not beenwell described. For this reason, we elected to perform this studyusing cadaveric dissections to directly compare the endoscopicwith the transcranial approaches to reach variety cranial base tar-gets. By implementing 3D technology, the spatial relationships, aswell as the relative advantages and limitations provided by eachapproach, can be more easily appreciated.

Paramedian cranial base regions were exposed well throughboth transcranial and endonasal approaches in a complemen-tary manner. The main limitations of the endonasal approachesare the lateral margins. Endoscopic endonasal corridors pro-vide excellent and direct access to midline structures (10, 30, 31,42). However, the neurovascular anatomy along the cranialbase acts to impede the extension of these corridors in the lat-eral direction. In particular, the ONs, the ICAs, the cavernoussinus, the vidian nerves, the abducens nerves, the jugulartubercles, and the hypoglossal canals create lateral boundaries,through which extension of endoscopic approaches will be dif-ficult (10, 32, 34, 46). However, it is important to note that spe-cific pathologies may distort these anatomic “boundaries”; dis-placement of these structures by the pathology may alleviatetheir role as lateral impediments. Alternatively, transcranialapproaches most often provide excellent lateral-to-medial tra-jectories to midline cranial base targets. Pathologies that proj-ect along these trajectories and extend laterally can be mostdirectly addressed with transcranial techniques. In addition,endoscopic assistance may enhance the deep view achievedtranscranially, enabling an around-the-corner view beyond thelinear view supplied by the microscope (11, 19, 36). However,other limitations of the transcranial approach remain: brainretraction and working through narrow neurovascular spaces,especially when the lesion is medial to these structures.

Ultimately, the choice of optimal approach to a given locationwill vary on the basis of specific size, extent, and consistency ofthe pathology. Endonasal approaches are more favorable forlesions that are midline with little lateral extent. Lesions whose



epicenters are lateral to the CNs and ICA will be more easilyexposed through transcranial approaches. If the pathology cre-ates a large working corridor, it is often possible to reach fartherthan expected, as the pathology may provide a dissection corri-dor around the normal anatomy. In addition, staging theremoval of large tumors may capitalize on combined approachesthat use both endonasal and transcranial corridors. In such a sit-uation, the medial extent of the tumor can be removed though anendonasal approach, while the lateral extent removed through atranscranial approach, with the choice of initial approach basedon the patient’s symptoms and the location of the highest degreeof neurological compression. Although our cadaveric dissectionsprovide views of the surgical anatomy that support such princi-ples in patient selection, clinical studies are needed to comparethese corridors in vivo for specific pathologies.

The 3D endoscope provided high-resolution images thatdetailed the depth relationships of vital neurovascular struc-tures. Qualitatively, 3D endoscopy also provided spatial orien-tation during the cadaveric dissections, making them easier toperform. Early preclinical studies with this technology supportits use in aiding in depth-oriented tasks (21). In addition, the useof 3D neuroendoscopy during pituitary tumor surgery providessubjectively improved depth perception for both the neurosur-geon and the otolaryngologist (3, 48). Previous studies in laparo-scopic surgery demonstrate that a 3D system improves dexter-ity, decreases errors, and improves the speed of novice as wellas experienced surgeons (4, 49). This study highlights the utilityof 3D endoscopy in appreciating anatomic relationships, whichis useful not only for educational purposes but also in assistingin depth perception and manual dexterity during surgery.

Limitations of the StudyThis study describes a qualitative comparison of various

approaches. However, quantitative comparisons were not per-formed. Future studies should focus on specific quantitativecomparisons, both of the exposure area and of the surgical lim-itations, i.e., the actual working space that is achieved by everyexposure. It is important to state that a surgical corridor is dic-tated not only by the actual visual exposure, but also by theactual instrumentation and optic system used. The closer theoptical tip to the field (endoscope versus microscope), thewider and less obstructed the visual exposure is, and by apply-ing suitable microinstruments, the working space may change.

In addition, we focused on the main transcranial approaches.We acknowledge that more complex cranial base approaches mayenable a wider exposure, with additional exposure of the clivusand anterior cranial base. These include, for example, the frontaltransbasal transsphenoidal approach to the anterior base includ-ing the interpeduncular cistern (44); the transcochlear and tran-sotic approaches to the upper and middle clival area (2, 45); anda combined subtemporal, presigmoid perilabyrinthine transtento-rial approach, which provides a wide exposure to the entire clivus,from the dorsum sellae down to the foramen magnum (23). Otherapproaches that provide larger exposures are a transotic approachor a transotic or retrolabyrinthine approach combined with atranstentorial incision for superior exposure (18, 24).

An additional surgical maneuver that may widen narrowworking corridors is opening of the oculomotor porus and inci-sion of the dural carotid rings (35, 40). This enables mobiliza-tion of these structures, widening the corridor to the interpe-duncular cistern.

CONCLUSIONSDirect comparison of endoscopic endonasal approaches and

traditional transcranial approaches to the median and parame-dian cranial base illustrates the respective advantages and dis-advantages of each approach in relation to target structures.Understanding the anatomic limits of each approach aids thesurgeon in providing a full surgical armamentarium to addresscranial base pathology. Three-dimensional visualizations aidin understanding the depth relationships of the denselypacked, highly sensitive neural and vascular elements in theparamedian cranial base. Clinical studies may be helpful incomparing surgical corridors for cranial base targets.Ultimately, however, the use of one or multiple approachesmust be tailored to the patient and the pathology.

DisclosureThe authors have no personal financial or institutional interest in any of the

drugs, materials, or devices described in this article.

REFERENCES1. Alfieri A, Jho HD: Endoscopic endonasal cavernous sinus surgery: An anatomic

study. Neurosurgery 48:827–837, 2001.2. Angeli SI, De la Cruz A, Hitselberger W: The transcochlear approach revis-

ited. Otol Neurotol 22:690–695, 2001.3. Brown SM, Tabaee A, Singh A, Schwartz TH, Anand VK: Three-dimensional

endoscopic sinus surgery: Feasibility and technical aspects. OtolaryngolHead Neck Surg 138:400–402, 2008.

4. Byrn JC, Schluender S, Divino CM, Conrad J, Gurland B, Shlasko E, Szold A:Three-dimensional imaging improves surgical performance for both noviceand experienced operators using the da Vinci Robot System. Am J Surg193:519–522, 2007.

5. Cappabianca P, Cavallo LM, Colao A, Del Basso De Caro M, Esposito F,Cirillo S, Lombardi G, de Divitiis E: Endoscopic endonasal transsphenoidalapproach: Outcome analysis of 100 consecutive procedures. Minim InvasiveNeurosurg 45:193–200, 2002.

6. Carrabba G, Dehdashti AR, Gentili F: Surgery for clival lesions: Open resectionversus the expanded endoscopic endonasal approach. Neurosurg Focus 25:E7,2008.

7. Catapano D, Sloffer CA, Frank G, Pasquini E, D’Angelo VA, Lanzino G: Com -pari son between the microscope and endoscope in the direct endonasal extendedtranssphenoidal approach: Anatomical study. J Neurosurg 104:419–25, 2006.

8. Cavallo LM, Cappabianca P, Galzio R, Iaconetta G, de Divitiis E,Tschabitscher M: Endoscopic transnasal approach to the cavernous sinus ver-sus transcranial route: Anatomic study. Neurosurgery 56:379–389, 2005.

9. Cavallo LM, Cappabianca P, Messina A, Esposito F, Stella L, de Divitiis E,Tschabitscher M: The extended endoscopic endonasal approach to the clivus andcranio-vertebral junction: Anatomical study. Childs Nerv Syst 23:665–671, 2007.

10. Cavallo LM, Messina A, Cappabianca P, Esposito F, de Divitiis E, Gardner P,Tschabitscher M: Endoscopic endonasal surgery of the midline skull base:Anatomical study and clinical considerations. Neurosurg Focus 19:E2, 2005.

11. Cheng WY, Chao SC, Shen CC: Endoscopic microvascular decompression ofthe hemifacial spasm. Surg Neurol 70 [Suppl 1]:40–46, 2008.

12. de Divitiis E, Cappabianca P, Cavallo LM, Esposito F, de Divitiis O, MessinaA: Extended endoscopic transsphenoidal approach for extrasellar cranio-pharyngiomas. Neurosurgery 61:219–228, 2007.


ROTH ET AL.

13. de Divitiis E, Cappabianca P, Gangemi M, Cavallo LM: The role of the endo-scopic transsphenoidal approach in pediatric neurosurgery. Childs Nerv Syst16:692–696, 2000.

14. de Divitiis E, Cavallo LM, Cappabianca P, Esposito F: Extended endoscopicendonasal transsphenoidal approach for the removal of suprasellar tumors:Part 2. Neurosurgery 60:46–59, 2007.

15. de Divitiis E, Cavallo LM, Esposito F, Stella L, Messina A: Extended endo-scopic transsphenoidal approach for tuberculum sellae meningiomas.Neurosurgery 61 [lsab]Suppl 2]:229–238, 2007.

16. Dehdashti AR, Ganna A, Karabatsou K, Gentili F: Pure endoscopic endonasalapproach for pituitary adenomas: Early surgical results in 200 patients and com-parison with previous microsurgical series. Neurosurgery 62:1006–1017, 2008.

17. Dehdashti AR, Karabatsou K, Ganna A, Witterick I, Gentili F: Expandedendoscopic endonasal approach for treatment of clival chordomas: Earlyresults in 12 patients. Neurosurgery 63:299–309, 2008.

18. Devèze A, Franco-Vidal V, Liguoro D, Guérin J, Darrouzet V: Transpetrosalapproaches for meningiomas of the posterior aspect of the petrous boneResults in 43 consecutive patients. Clin Neurol Neurosurg 109:578–588, 2007.

19. Ebner FH, Koerbel A, Kirschniak A, Roser F, Kaminsky J, Tatagiba M:Endoscope-assisted retrosigmoid intradural suprameatal approach to themiddle fossa: Anatomical and surgical considerations. Eur J Surg Oncol33:109–113, 2007.

20. Frank G, Pasquini E: Endoscopic endonasal cavernous sinus surgery, withspecial reference to pituitary adenomas. Front Horm Res 34:64–82, 2006.

21. Fraser JF, Allen B, Anand VK, Schwartz TH: Three-dimensional neurostere-oendoscopy? Subjective and objective comparison to 2D. Minim InvasiveNeurosurg 52:25–31, 2009.

22. Gardner PA, Kassam AB, Snyderman CH, Carrau RL, Mintz AH, Grahovac S,Stefko S: Outcomes following endoscopic, expanded endonasal resection ofsuprasellar craniopharyngiomas: A case series. J Neurosurg 109:6–16, 2008.

23. Hayashi N, Kurimoto M, Nagai S, Sato H, Hori S, Endo S: Tentorial incisionin a lateral-medial direction with minimal retraction of the temporal lobe inthe subtemporal transtentorial approach to the middle tentorial incisuralspace. Minim Invasive Neurosurg 51:340–344, 2008.

24. Horgan MA, Anderson GJ, Kellogg JX, Schwartz MS, Spektor S, McMenomeySO, Delashaw JB: Classification and quantification of the petrosal approach tothe petroclival region. J Neurosurg 93:108–112, 2000.

25. Jho HD, Carrau RL: Endoscopic endonasal transsphenoidal surgery:Experience with 50 patients. J Neurosurg 87:44–51, 1997.

26. Jho HD, Ha HG: Endoscopic endonasal skull base surgery: Part 1—The mid-line anterior fossa skull base. Minim Invasive Neurosurg 47:1–8, 2004.

27. Jho HD, Ha HG: Endoscopic endonasal skull base surgery: Part 3—The clivusand posterior fossa. Minim Invasive Neurosurg 47:16–23, 2004.

28. Kassam AB, Gardner PA, Snyderman CH, Carrau RL, Mintz AH, PrevedelloDM: Expanded endonasal approach, a fully endoscopic transnasal approachfor the resection of midline suprasellar craniopharyngiomas: A new classifi-cation based on the infundibulum. J Neurosurg 108:715–728, 2008.

29. Kassam AB, Prevedello DM, Thomas A, Gardner P, Mintz A, Snyderman C,Carrau R: Endoscopic endonasal pituitary transposition for a transdorsumsellae approach to the interpeduncular cistern. Neurosurgery 62:57–74, 2008.

30. Kassam A, Snyderman CH, Mintz A, Gardner P, Carrau RL: Expandedendonasal approach: The rostrocaudal axis. Part I. Crista galli to the sella tur-cica. Neurosurg Focus 19:E3, 2005.

31. Kassam A, Snyderman CH, Mintz A, Gardner P, Carrau RL: Expandedendonasal approach: The rostrocaudal axis. Part II. Posterior clinoids to theforamen magnum. Neurosurg Focus 19:E4, 2005.

32. Kassam AB, Vescan AD, Carrau RL, Prevedello DM, Gardner P, Mintz AH,Snyderman CH, Rhoton AL: Expanded endonasal approach: Vidian canal as alandmark to the petrous internal carotid artery. J Neurosurg 108:177–183, 2008.

33. Laufer I, Anand VK, Schwartz TH: Endoscopic, endonasal extendedtranssphenoidal, transplanum transtuberculum approach for resection ofsuprasellar lesions. J Neurosurg 106:400–406, 2007.

34. Magro F, Solari D, Cavallo LM, Samii A, Cappabianca P, Paternò V, LüdemannWO, de Divitiis E, Samii M: The endoscopic endonasal approach to the lateralrecess of the sphenoid sinus via the pterygopalatine fossa: Comparison ofendoscopic and radiological landmarks. Neurosurgery 59 [Suppl 2]:ONS237–ONS243, 2006.

35. Martins C, Yasuda A, Campero A, Rhoton AL Jr: Microsurgical anatomy ofthe oculomotor cistern. Neurosurgery 58 [Suppl 2]:ONS-220–ONS-228, 2006.



36. Rak R, Sekhar LN, Stimac D, Hechl P: Endoscope-assisted microsurgery formicrovascular compression syndromes. Neurosurgery 54:876–883, 2004.

37. Reisch R, Perneczky A, Filippi R: Surgical technique of the supraorbital key-hole craniotomy. Surg Neurol 59:223–227, 2003.

38. Rhoton AL Jr: The cavernous sinus, the cavernous venous plexus, and thecarotid collar. Neurosurgery 51:S375–S410, 2002.

39. Rudnik A, Zawadzki T, Wojtacha M, Bazowski P, Gamrot J, Galuszka-Ignasiak B, Duda I: Endoscopic transnasal transsphenoidal treatment ofpathology of the sellar region. Minim Invasive Neurosurg 48:101–107, 2005.

40. Sade B, Kweon CY, Evans JJ, Lee JH: Enhanced exposure of carotico-oculomo-tor triangle following extradural anterior clinoidectomy: A comparativeanatomical study. Skull Base 15:157–162, 2005.

41. Schwartz TH, Anand VK: The endoscopic endonasal transsphenoidalapproach to the suprasellar cistern. Clin Neurosurg 54:226–235, 2007.

42. Schwartz TH, Fraser JF, Brown S, Tabaee A, Kacker A, Anand VK: Endoscopiccranial base surgery: Classification of operative approaches. Neurosurgery62:991–1005, 2008.

43. Schwartz TH, Stieg PE, Anand VK: Endoscopic transsphenoidal pituitarysurgery with intraoperative magnetic resonance imaging. Neurosurgery58:ONS44–ONS51, 2006.

44. Sekhar LN, Nanda A, Sen CN, Snyderman CN, Janecka IP: The extendedfrontal approach to tumors of the anterior, middle, and posterior skull base.J Neurosurg 76:198–206, 1992.

45. Sincoff EH, McMenomey SO, Delashaw JB Jr: Posterior transpetrosalapproach: Less is more. Neurosurgery 60 [Suppl 1]:ONS53–ONS59, 2007.

46. Solari D, Magro F, Cappabianca P, Cavallo LM, Samii A, Esposito F, PaternoV, de Divitiis E, Samii M: Anatomical study of the pterygopalatine fossausing an endoscopic endonasal approach: Spatial relations and distancesbetween surgical landmarks. J Neurosurg 106:157–163, 2007.

47. Stippler M, Gardner PA, Snyderman CH, Carrau RL, Prevedello DM, KassamAB: Endoscopic endonasal approach for clival chordomas. Neurosurgery64:268–278, 2009.

48. Tabaee A, Anand VK, Fraser JF, Brown S, Singh A, Schwartz TH: Three-dimensional endoscopic pituitary surgery. Neurosurgery 65 [Suppl 2]:288–295, 2009.

49. Taffinder N, Smith SG, Huber J, Russell RC, Darzi A: The effect of a second-generation 3D endoscope on the laparoscopic precision of novices and expe-rienced surgeons. Surg Endosc 13:1087–1092, 1999.

COMMENTS

This article continues the work of previous anatomic studies per-formed on the same or similar topics by the groups in Gainesville,

Naples, Pittsburgh, and others. The contribution of the “New Yorkers”is interesting because it emphasizes the increasing role of endoscopicendonasal procedures in the management of cranial base lesions andthe need to perfectly understand the same relevant anatomy by com-paring the upper and the lower views of the same target area.Furthermore, it stresses the need for technological improvement andinstrument development.

Where I would be more cautious is in fixing the indications and theadvantages of the transcranial and/or the transnasal options. The sameprogress that is making it useful to have reports such as the present oneis slowly changing dogmas and paradigms of neurosurgical indica-tions. Therefore, the work produced by Kassam’s team in Pittsburghseems very important in extracting concepts from the anatomophysio-logical and surgical experience, as, for example: “Never cross thenerves.” According to such concepts and the actual possibilities pro-vided by such contemporary approaches, the surgeon can considerperforming a particular procedure rather than another one. Concepts,anatomic studies, and technical progress are, again, the keys toadvancement.

Paolo CappabiancaNaples, Italy

This is another excellent article from this distinguished cooperativegroup of neurosurgeons and otolaryngological surgeons. Of special

value is the comparison between transcranial (also, in part, endo-scopic!) and transnasal approaches to the midline cranium and theevaluation of advantages and limitations of both approaches. In recentyears, more and more reports are being published on the advantage ofthe extended endonasal approaches in this area. However, the existingadvantages of “traditional” transcranial approaches, which today arealso performed as “minimally invasive” procedures with neuronaviga-tion and endoscopic assistance, are often not considered in thesereports on the “modern” extended transnasal approach. But some ofthe advantages of transcranial approaches are obvious: the small risk ofcerebrospinal fluid leakage, which continues to be a significant chal-lenge in intracranial transnasal approaches; no damage to nasal airflow,with often minimal side effects of precise transcranial approaches usinglimited retraction; and, especially, the advantage of binocular stereo-scopic microsurgery in a more familiar anatomic environment for theneurosurgeon.

Here, this article presents another highlight: the use of 3-dimen-sional (3D) endoscopy for the preparation, which, however, can beused not only transnasally, but also in transcranial approaches. Theauthors have previously reported on the use of 3D endoscopes; in thisstudy, they used 3D endoscopy in the transnasal as well as transcranialapproaches, so that the comparison of both ways is quite evident (withthe additional combination of 3D endoscopy and microscopy in mostof the transcranial approaches).

This article presents very illustrative figures of excellent anatomicpreparations in vascular injections of fresh cadaver heads; the 3D endo-scopic images are given with the “insect eye” Visionsense technology(Visionsense, Petach Tikva, Israel), whose features are briefly men-tioned and are given in greater detail in the authors’ 3D pituitary study(1). The results (advantages and limitations) of each approach aredivided into the areas A, B (pre- and postchiasmatic), C (posterioraspect of the clivus), D (lower clivus, to below the abducens nerve), E(lateral to the internal carotid artery), and F (to the cervicomedullaryjunction); also, the different possibilities with the transcranialapproaches in pre- and postfixed chiasms are addressed.

The conclusions are convincing and well illustrated; the main “con-ceptual difference” is defined as medial/lateral exposure; thus, lesionsexpanding from lateral to medial should be approached transcranially,so this approach retains its value; however, lesions invading the medialbasal structures (such as pituitary lesions, which are “classic”transnasal targets, but today extending also to more complex lesions)are favored for the transnasal route, which has its limitations at the lat-eral margins, as presented by the internal carotid artery, cavernoussinus, abducens nerves, and hypoglossus canal. Also, a combination ofboth methods (medial parts of the lesion transnasally, lateral parts tran-scranially) is convincingly described; the limitations of such anatomicstudies and the focus on only selected transcranial approaches arefrankly admitted. As the authors stress, clinical studies ultimately mustdefine the surgical corridors, and approaches must be individually tai-lored for each patient. However, this study is very useful for the indi-vidual decision; the high-quality figures can even be used by the neu-rosurgeon for the preparation of the clinical operation.

Michael R. GaabHannover, Germany

1. Tabaee A, Anand VK, Fraser JF, Brown S, Singh A, Schwartz TH: Three-dimen-sional endoscopic pituitary surgery. Neurosurgery 65 [Suppl 2]:288–295, 2009.


ROTH ET AL.

oth et al. present an interesting anatomic study comparing theexposure provided by expanded endonasal and transcranial cranial

base approaches. The investigation was performed in 10 fresh cadaverheads. Access to various areas of the midline cranial base was evalu-ated. The authors conclude that the endonasal approaches achieved awide and direct exposure of the midline, whereas transcranialapproaches were limited by narrow operative corridors formed by neu-rovascular structures. Additionally, a comparison of 2D and 3D endo-scopic visualization was performed.

Although I agree with most of the statements in the article, I do notgenerally agree that “transcranial approaches…do not visualize thecontralateral paramedian cranial base very well,” as stated in theAbstract. In transcranial approaches to the frontal cranial base, forexample, much more overview of the cranial base can be achieved thanwith the endonasal approach. With proper positioning of the head andearly release of cerebrospinal fluid, significant brain retraction is notrequired to visualize the contralateral paramedian cranial base.Furthermore, in most transcranial tumor operations, visualization ofthe contralateral cranial base is easily possible because the tumors oftendisplace the brain posteriorly or superiorly and provide wide spacewhen being resected.

I agree that endonasal approaches have several advantages, com-pared with transcranial approaches to the cranial base. First of all,there is no brain retraction. Especially in subchiasmatic lesions withpre-fixed chiasm, the endonasal approach is much better than thetranscranial route, because retraction of the neurovascular structures,which is clearly necessary in transcranial resection, can be avoidedcompletely. However, a major limitation of the endonasal approachis the poor overview of the entire cranial base. Only the midline canbe visualized from below. Lesions with significant extensions to oneside or the other cannot completely and safely be resected by theendonasal approach.

I have had a chance to test the 3D endoscope used by the authors,and I was very impressed by the image quality and the stereoscopicview. Much progress has been made, in comparison to the 3D endo-scopes I have used in the past. However, my impression is that theimage resolution and, in particular, the color fidelity of the recentlyintroduced high-definition video cameras is significantly better.Today, high-definition imaging is the state-of-the-art in endoscopicvisualization. It provides excellent resolution (1920 � 1080 = approx-imately 2 Mio. pixels) and superb color information. I would neveraccept a decrease in resolution and color fidelity as the price forobtaining a 3D view.

Nevertheless, I am sure that, in the future, high-resolution 3D endo-scopes will replace our current 2D endoscopic equipment. Neuro -navigation will be easily integrated while superimposing segmented

magnetic resonance imaging data on the endoscopic images, and thisinformation will be displayed on a single screen.

That said, the authors have presented a nice study. This article is wellwritten and adds important information to the growing body of liter-ature on endonasal cranial base surgery.

Henry W.S. SchroederGreifswald, Germany

The authors have presented a fine anatomic study. They comparedendonasal and transcranial approaches to the cranial base on

cadaver specimens. These approaches have been well documentedusing 3D video and still images. This article takes us one step furtherto defining the optimal surgical approaches to cranial base lesions. Myonly criticism is the glaring omission of the use of endoscopic assis-tance with the transcranial approaches. Many of the limitations of thetranscranial approaches may well have been overcome with the use ofangled endoscopes. I have removed a brainstem cavernoma thatextended down to the midmedulla using a subtemporal approach andendoscopic assistance.

My other issue is more a warning than a criticism. Having becomefamiliar and comfortable with both transcranial and endonasalapproaches to the cranial base, I still favor the transcranial approach inthe majority of cases. I have formulated this surgical algorithm for thefollowing reasons: 1) standard microsurgical techniques are more famil-iar; 2) if one uses endoscopic assistance, the size of the craniotomy canbe limited, and many previously hidden areas are adequately visual-ized; 3) endonasal instrumentation is still suboptimal; 4) stereoscopicvision is better with the microscope than with current 3D endoscopes;and 5) the cerebrospinal fluid leakage rate is lower.

There are, however, several specific tumors of the cranial base thatI believe are clearly better removed endonasally. The first is the suprasellar/infrachiasmatic craniopharyngioma. The second is anolfactory groove meningioma that does not extend laterally past theoptic nerves nor posteriorly to the tuberculum sellae. The third is thepurely extradural chordoma of the clivus. A questionable fourth lesionis the schwannoma or meningioma of the medial cavernous sinus. Itruly believe that this list will rapidly expand with the development ofendonasal instruments, especially bipolar coagulators.

In conclusion, although this elegant study shows us that theendonasal approach exposes the midline and paramedian anatomybetter than some transcranial approaches, this should not be extrapo-lated to the clinical setting, because of other practical considerationssuch as instrumentation and surgeon familiarity.

Charles TeoSydney, Australia

R

Documents

Three Dimensional And 2 Dimensional Endoscopic.21 - Visionsense