Surg Radiol Anat DOI 10.1007/s00276-015-1497-5
ORIGINAL ARTICLE
Anatomic comparison of anterior petrosectomy versus the expanded endoscopic endonasal approach: interest in petroclival tumors surgery Timothe´e Jacquesson1,2 • Emile Simon2 • Moncef Berhouma1 • Emmanuel Jouanneau1
Received: 3 February 2015 / Accepted: 23 May 2015 Ó Springer-Verlag France 2015
Abstract Purpose Since the petroclival region is deep-seated with close neurovascular relationships, the removal of petroclival tumors still represents a fascinating surgical challenge. Although the classical anterior petrosectomy (AP) offers a meaningful access to this petroclival region, the expanded endoscopic endonasal approach (EEEA) recently leads to overcome difficulties from trans-cranial approaches. Herein, we present an anatomic comparison of AP versus EEEA. We aim to describe the limits of these both approaches helping the choice of the optimal surgical route for petroclival tumors. Methods Six fresh cadaveric heads were harvested and injected with colored latex. Each approach was step-bystep detailed until its final surgical exposure. Results The AP provided a narrow direct supero-lateral access to the petroclival area that can also reach the cavernous sinus, the retrochiasmatic region and perimesencephalic cisterns. However, this corridor anterior to the internal acoustic meatus passed on each side of the trigeminal nerve. Moreover, tumor extensions toward the foramen jugularis, inside the clivus or behind the internal acoustic meatus were difficult to control. The EEEA brought a straightforward access to the clivus but the petrous apex was hidden behind the internal carotid artery.
Several variants were described: a medial transclival, a lateral through the Meckel’s cave and an inferior transpterygoid route. Elsewhere, tumor extension behind the internal acoustic meatus or above the tentorium could not be satisfactorily assessed. Discussion and conclusion PA and EEEA have their own limits in reaching the petroclival region in accordance with the tumor characteristics. The AP should be preferred for radical removal of middle-sized petrous apex intradural tumors like meningiomas. The EEEA would be of interest for extradural midline tumors like chordomas or for petrous apex cysts drainage.
& Timothe´e Jacquesson [email protected]
Introduction
1
Skull Base Multi-disciplinary Unit, Department of Neurosurgery B, Neurological Hospital Pierre Wertheimer, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Lyon Cedex, France
2
Department of Anatomy, University of Lyon 1, 8 Avenue Rockefeller, 69003 Lyon, France
Keywords Skull base Petroclival Petrosectomy Endoscopy Endonasal Abbreviations AP Anterior petrosectomy EEEA Expanded endoscopic endonasal approach IAM Internal acoustic meatus ICA Internal carotid artery SPN Superficial petrosal nerve UPR Upper petrous ridge
The petroclival region appears rather as a ‘‘surgical’’ space limited anteriorly by the clivus, laterally by the petrous apex, medially by the brainstem and posteriorly by the internal acoustic meatus (IAM) [14] (Fig. 1). Since it is deep-seated and closely related with numerous neurovascular structures, reaching the petroclival region remains a
123
Surg Radiol Anat b Fig. 1 The petroclival region, photographs. Superior views of the
middle cranial fossa (a), posterior side of the petrous bone (b) and superior side of the petrous bone (c). The petroclival region includes the petrous apex and the clivus, from the dorsum sellae (DS) to the foramen jugularis (FJ), medially to the internal acoustic meatus (IAM). Around the petroclival region are described: anteriorly at the midline the sella turcica (ST) and the tuberculum sellae (TS), posteriorly the foramen magnum and laterally from front to back the optic canal (OC), the anterior clinoid process (AC), the superior orbital fissure (SOF), the foramen rotondum (FR), the foramen ovale (FO), the foramen spinosum (FS) and the sigmoid sinus (SS). A stairstep can be found at the upper petrous ridge (UPR) and falsely delimits this last. The horizontal intrapetrous carotid canal (CC) is parallel and medial to the superficial petrosal nerves (SPN) which are also parallel and medial to the line formed by the foramen ovale and spinosum. The axis of the IAM is indicated by the bisectrix of the SPN and the eminentia arcuata (EA), whose summit is the geniculate ganglion. Laterally, the tegmen (T) appears as a thin bone layer separating the middle cranial fossa and the middle ear
described first by Kawase then by Velut [30, 24]. This transpetrous route called anterior petrosectomy (AP) required a dangerous drilling around the otologic structures, the trigeminal nerve and the internal carotid artery (ICA). Third, the development of endoscopy in neurosurgery allowed for an anterior endonasal transsphenoidal access to the cranial base thus preventing cerebrum retraction and intrapetrous structures damages [4, 17]. Likewise, the expanded endoscopic endonasal approach (EEEA) provided a greater exposure, vertically from the cribriform plate to the craniovertebral junction and laterally to the parasellar or infratemporal fossas [6, 17, 18]. Thereby, we performed an anatomical comparison of PA and EEEA describing the limits of each and helping petroclival tumors surgical strategy.
Materials and methods Specimen preparation
fascinating challenge in skull base surgery. Thereafter, petroclival tumors have long been considered inoperable due to incomplete resection and high morbidity [5, 10]. More recently, improvements in surgical technique, operative microscope and electrophysiological monitoring raised hope to achieve a more safe efficient surgery in this petroclival area. Then, several renowned neurosurgeons reported their experiences and various surgical routes to these tumors [1, 2, 11, 24, 25]. First, a posterior, retrosigmoid approach was proposed, but its corridor crossed cranial nerves V to IX with significant risks [22, 29]. Second, a lateral approach was
123
Six fresh cadaver heads were harvested at the anatomy laboratory of University Lyon 1 (Lyon, France) and prepared with 10 % formaldehyde. Within 48 h of death, the common carotid arteries, the vertebral arteries (VA) and the internal jugular veins were catheterized, then the entire vascular network was washed and injected with red (arteries) and blue (veins) latex (Phoce´enne de chimie—Marseille/France; Ae´rographe Colorex Technics, Magenta et Cyan). Anatomic study Six APs were performed with microsurgical instruments (ClimdalÒ Instrumentation Me´dico-Chirurgicale, Lyon, France) and operative microscope (Carl ZeissÒ Microscopy Gmbh, Go¨ttingen, Germany) according to the technique
Surg Radiol Anat
described by Kawase and Velut [20, 31]. Similarly, six EEEAs were performed controlaterally with dedicated endoscopic devices (Karl StorzÒ Endoskope GmbH, Tuttlingen, Germany) according to the technique described by Kassam [17]. The surgical anatomy was precisely detailed at each step for both approaches. Between each dissection phase, specimens were conserved in baths of 10 % ethanol and 1 % formaldehyde. Craniotomies and drilling necessitated a continuous irrigated high-speed motor (Integrated Power Console and Handpieces, MedtronicÒ, Louisville, USA). Post-approach skull base study After completion of AP and contralateral EEEA on each specimen, the calvaria and brain were cautiously removed, leaving in place the cranial nerves, the lower part of the brainstem and the arterial circle of Willis. No further modifications were made to the surgical routes. The osteodural opening, the working space, the neurovascular relationships and the possibilities of extension were specified for both approaches.
Results Microsurgical anterior petrosectomy Surgical technique Specimens were placed to obtain a nearly vertical alignment of the external acoustic meatus. After a curvilinear incision, a classic temporal craniotomy was made as low as possible then the dura mater was elevated posteriorly to anteriorly discovering the superior side of the petrous part of the temporal bone. Having sectioned the middle meningeal artery, the ‘‘rhomboid construct’’ [12] was discovered with: the foramen spinosum, the foramen ovale, the eminencia arcuata, the superficial petrosal nerves (SPN), the carotid canal and the upper petrous ridge (UPR) (Figs. 1, 2). The axis of the foramen spinosum and the foramen ovale gave medially the SPN and then the horizontal canal of the internal carotid artery (ICA). A 100° on average angle was observed between the arcuate eminence and the SPN and the bisectrix of this angle drew the anterior edge of the internal acoustic meatus (IAM) [27]. At the summit of this angle sat the geniculate ganglion and just forward the cochlea. A ‘‘staircase’’ groove for the superior petrosal sinus might be seen at the UPR. The meningeal sheet of foramen ovale dura mater had to be cut easing an anterior displacement of the mandibular nerve and releasing the petrous apex. The Kawase’s triangle was carefully drilled from the trigeminal print to the IAM and
from the UPR to the inferior petrosal sinus with approximatively 20 mm (mm) width and 10 mm depth. Care had to be taken to ICA or SPN laterally and cochlea or acoustic facial bundle posteriorly. Then, it allowed for tentorium splitting and posterior and middle cranial fossa dura mater opening. It thus offered a broad supra/infra-tentorial exposure of the petroclival region through the drilled petrous apex. Limits This approach provided a direct, narrow, superolateral access to the petroclival region bounded: superiorly by the IVth nerve and the temporal lobe; inferiorly by the inferior petrosal sinus and the ascending cisternal segment of the VIth nerve with its course into the Dorello’s canal; anteriorly by the clivus; medially by the brainstem and the vertebrobasilar system and posteriorly by the IAM which contained the VIIth and VIIIth nerves. In addition to give access to Meckel’s cave and the cavernous sinus, this surgical corridor could be widened to the entire middle cranial fossa, the retrochiasmatic space and perimesencephalic cisterns. Nonetheless, this route required a retraction of the temporal lobe working dangerously on either side of the trigeminal nerve. Drilling was also delicately close to intrapetrous otologic structures like geniculate ganglion, cochlea, semicircular canals and tegmen tympani. It came with risk of facial palsy, deafness, balance disorders and CSF leak, respectively. Also, tumor extensions could not be controlled inferiorly toward the foramen jugularis, anteriorly toward the clivus and posteriorly behind the IAM (Fig. 4). Finally, enlarging toward the midline appeared possible only if the basilar artery and the brainstem were displaced by a tumor. Expanded endoscopic endonasal approach Surgical technique of the transsphenoidal approach Specimens were placed facing the operator in a supine position. A 180-mm length and 4-mm diameter endoscope was used with 0° or 30° lenses (Karl StorzÒ Endoskope GmbH, Tuttlingen, Germany). A bi-nostril path was necessary and a self-retaining endoscope holder allowed for a two-hands free technique by a single operator. Moving along the nasal floor and the inferior turbinate led to the choanal arch, which was a reliable sagittal landmark matching the floor of the sphenoid sinus and containing the septal branch of the sphenopalatine artery. The pharyngeal wall was visualized with the ostium of the Eustachian tube laterally and the nasal septum medially. The ostium sphenoidale was most frequently located 10 mm on average above the choanal arch. A septal mucosa incision
123
Surg Radiol Anat
Fig. 2 Anterior petrosectomy. Operative views. Posterior and inferior to the coronal and squamosal bone sutures, a classic low temporal craniotomy is performed. Then, a careful dural dissection discovers the ‘‘rhomboid construct’’ at the top of the petrous bone with, from back to front (a, b): the upper petrous ridge (UPR), the eminentia arcuata (AE), the internal acoustic meatus (IAM), the superficial petrosal nerve (SPN), internal carotid artery (ICA), the middle meningeal artery (MMA) in its foramen spinosum (FS) and the three branches of the trigeminal nerve. Drilling starts ahead of the IAM until the trigeminal print and inferiorly toward the posterior cranial fossa dura. Sometimes uncovered by bone, the horizontal petrous segment of the ICA is located just medial and parallel to the SPN.
After opening the temporal and posterior cranial fossas dura and the tentorium, the final exposition is obtained (c, d). The trigeminal nerve crosses the corridor and gives its branches V1, V2 and V3 with their respective destination: the superior orbital fissure (SOF), the foramen rotondum (FR) and the foramen ovale (FO). Deeply, the antero-lateral side of the brainstem is reached. Superiorly, the IVth nerve passes under the tentorium before entering the cavernous sinus (CS) whom dura mater (DM) can be peeled. Anteriorly, the VIth nerve courses before its Dorello’s canal. Inferiorly, the inferior petrosal sinus marks out the lower limit of the access while the IAM can be opened posteriorly
started from this ostium discovering the rostral spur of the vomer then the contralateral ostium sphenoidale. A wide sphenoidotomy was performed, identifying the classical intrasphenoid landmarks: the paracavernous and paraclival segments of ICAs, the optic canals, the lateral and medial opticocarotid recesses, the sella turcica, the paraclival recess, the planum sphenoidale and posterior ethmoidal cells. Thereby, the posterior third of the nasal septum and the middle turbinate were removed. The bone opening was adapted on demand to provide sufficient exposure and maneuverability. Finally, access to the entire clivus was straightforward, but access to the petrous apex needed circumventing the ICA. Thus, three variants could be decribed: medial, lateral and below the ICA (Fig. 3).
Variants and limits
123
The first medial variant was trans-sphenoidal trans-clival. Indeed, this route medial to the ICA went through the clivus with its venous plexus and reached the tip of the petrous pyramid surrounded: superiorly by the pituitary fossa, inferiorly by the sphenoid floor and posteriorly by the basilar artery with the brainstem and the VIth nerve. The abducens nerve emerged from the pons at the midline level of the sphenoid floor. It then coursed obliquely to join Dorello’s canal at the superior third of the vertical part of the ICA paraclival segment. Finally, the aperture could be easily extended toward the contralateral side contrary to a lateral extension which would be hindered by the ICA.
Surg Radiol Anat
The second lateral variant was trans-sphenoidal transMeckel’s cave involving a lateral enlargement of the corridor as middle turbinectomy and posterior ethmoidectomy. Likewise, the vertical paraclival ICA segment was cautiously skeletonized; the orbital apex, the cavernous sinus and the Meckel’s cave were exposed. The petrous apex was thus reached through the Meckel’s cave. This corridor passing between the trigeminal branches and the ICA was
limited: superiorly by the cavernous sinus with ascending V1, VIth nerve and orbital apex, inferiorly and medially by the petrous and paracavernous ICA and laterally by the sphenoid bone with its foramen rotondum and ovale. Opening these last allowed for pursuing V2 and V3 anteriorly. Having medially displaced the anterior loop of the ICA, the IIIrd and IVth nerves were also identified into the cavernous sinus.
Fig. 3 Expanded endoscopic endonasal approach. Endoscopic views. A classic endoscopic endonasal trans-sphenoidal approach is performed with a 0° lens endoscope via the two nostrils. First, a large sphenoidotomy exposes the classical intra-sphenoidal landmarks (a, b) from up to down: ethmoidal cells (EC) with the posterior ethmoidal artery (PEA) and planum, optic canals (OC), medial and lateral optico-carotid recesses (mOCR, lOCR), sella turcica (ST), internal carotid artery (ICA) with its paraclival and paracavernous segments, paraclival recess (PCR) and the always medial vomer bone. The lOCR known as ‘‘optic strut’’ separates the paracavernous ICA and the ON and leads to the anterior clinoid process. Then, access to the clivus is large and straightforward, but the ICA needs to be circumvented to reach the petrous apex. Three variants could be described: a medial trans-clival route, a lateral route through the Meckel’s cave and an inferior trans-pterygoid route (C, D). The first variant (red square) passes through the clivus to reach the tip of the petrous pyramid medial to the ICA under the VIth nerve. The second
variant (red circle) requires a lateral bone extension including ICA skeletonization. This way the orbital apex, the cavernous sinus and Meckel’s cave are exposed with their own contents: cranial nerves III, IV, V1, V2, V3 and VI. The third variant (red star) needs a long drilling of the sphenoid floor (Sph) and the medial pterygoid plate. The vidian nerve (VN) leads to the foramen lacerum (FL) and the path continues below the horizontal ICA and above the Eustachian tube (ET) (c). Through the middle clivus, the pons is discovered with the basilar artery (BA) and its branches: the superior cerebellar artery (SCA) near the IIIrd nerve (III) and the anterior inferior cerebellar artery (AICA). The VIth nerve crosses obliquely from the pons to Dorello’s canal behind the vertical paraclival ICA. Deeply, the drilling attains the internal acoustic meatus with its acoustic-facial bundle (VII–VIII). Finally, this approach can be also expanded up to the sellar, parasellar and suprasellar areas (d) including the pituitary gland (PG), the optic and olfactory nerves (ON, Olf) (color figure online)
123
Surg Radiol Anat Fig. 4 Skull base photographs after right anterior petrosectomy and c left expanded endoscopic endonasal approach. These views accurately show the anatomical limits of each approach. From laterally (a), the anterior petrosectomy (AP) follows the top of the petrous bone to find out successively from back to front: the upper petrous ridge (UPR), the middle meningeal artery (MMA) in its foramen spinosum (FS), the three trigeminal branches (V3, V2 and V1). Drilling the petrous apex gives a narrow access to the petroclival region working on each side of the trigeminal nerve (V). This corridor can be enlarged to expose the whole middle cranial fossa until the lesser sphenoid wing (lSw) as the cavernous sinus and the Meckel’s cave with theirs contents: cranial nerves III, IV, V1, V2, V3 and VI. Likewise, any supra-tentorial tumor extensions can be controlled as the retrochiasmatic space (behind the chiasma, Ch) or perimesencephalic cisterns. From medially (b), the AP’s corridor appears parallel to the acoustic facial bundle (VII–VIII) and just behind the VIth nerve. The basilar artery (BA), the anterior inferior cerebellar artery (AICA) and the brainstem (BT) are crossing through. That is why a contralateral or midline exposure is possible only if a tumor opens the way. From posteriorly (c), the expanded endoscopic endonasal approach (EEEA) offers a large osteodural aperture that can be widened to the pituitary and anterior cranial fossas. The medial transclival variant crosses the clivus and jeopardizes the VIth nerve and the basilar artery. The lateral trans-Meckel’s cave variant needs a tumor that pushes aside V2 and V3 branches from the ICA. The inferior trans-pterygoid route passes between the Eustachian tube and the ICA; it then gets close to the internal acoustic meatus and the foramen jugularis with cranial nerves VII to XI. Despite the dangerous path around the ICA to the hidden petrous apex, the EEEA provides a large access to the clivus with multiple specific expansions as into the pterygopalatine fossa, the infratemporal fossa, the cavernous sinus, the Meckel’s cave, the contralateral side and the suprasellar area
The inferior third variant was trans-sphenoidal transpterygoid. The maxillary sinus and pterygopalatine fossa opening and the sphenopalatine artery cutting were required. The pterygopalatine ganglion met the maxillary, vidian and great palatine nerves and was then lateralized. The pterygoid canal with its vidian nerve and artery was found out at the intersection of the sphenoid floor and the medial pterygoid plate [19]. A long drilling of the upper portion of the medial pterygoid plate along the pterygoid canal thus led to the foramen lacerum with its fibrous proximal ring. The drilling was continued under the horizontal portion of the ICA releasing the cartilaginous part of the Eustachian tube and reaching the ‘‘inferior’’ petrous apex [30] surrounded: superiorly by the horizontal ICA, inferiorly by the Eustachian tube, laterally by the greater wing of the sphenoid and medially by the median sphenoid floor. Posteriorly, the foramen jugularis as far as the IAM was attained. Similarly, the acoustic facial bundle was identified with the labyrinthine artery and the choroid plexus in the lateral aperture of the fourth ventricle. Finally, for all variants, tumor extensions behind the ICA or the IAM are not accessible contrary to sellar, parasellar and supra-sellar areas (Fig. 4).
123
Discussion Principal findings AP and EEEA offer different routes to the petroclival region with their own limits (Table 1). AP provides a narrow access to the petrous apex, the upper clivus, the trigeminal porus and the antero-lateral side of the brainsteam [8, 15, 20, 23, 31]. Having tailored the craniotomy, the corridor could be enlarged toward the retrochiasmatic space, perimesencephalic cisterns and the whole middle cranial fossa. It eases the control of tumor vascular supply and bone invasion. However, it runs anterior to the acoustic facial
Surg Radiol Anat Table 1 Pros and cons of anterior petrosectomy versus expanded endoscopic endonasal approach Anterior petrosectomy
Expanded endoscopic endonasal approach
Technique
Microscope
Endoscope
Working space
Wide superficially, narrow deeply
Narrow superficially, wide deeply
Distance to the target
Short
Long
Surgical corridor
oblique
Straightforward
Handiness
Limited deeply
Limited superficially
Visibility
3D, microscope
Panoramic, 2D, ‘‘barrel effect’’
Skills
Skull base surgery
Endoscopy
Anatomy
Petrous bone
Nasal pits, sphenoid bone
Advantages
Classic sub temporal approach
Minimally invasive
Warnings
Side effects
Vascular supply access
Corner and close up views
extension to supratentorial area extension to MC, CS
Extension to suprasellar area Extension to CVJ, ITF, MC, CS
Crossing IVth, Vth and VIth nerves
Crossing VIth and vidian nerves
Risks of cochlea, SCC, SPN damages
Risks of ICA, ET damages
Temporal lobe retraction, craniotomy
CSF leak, skull base reconstruction
Limited inferior access
Limited superior access
Scar, headaches
Empty nose sensation, rhinitis
This table summarizes the technical subtleties, advantages and disadvantages of anterior petrosectomy and the expanded endoscopic endonasal approach to help the choice of the optimal surgical route for petroclival tumors surgery 2D/3D bi/tridimensional, CS cavernous sinus, CSF cerebrospinal fluid, CVJ craniovertebral junction, ET eustachian tube, ICA internal carotid artery, MC Meckel’s cave, ITF infratemporal fossa, SCC semicircular canals, SPN superficial petrosal nerves
bundle, AP still have to cross the pathetic and trigeminal nerves. Also, it jeopardizes the horizontal ICA, the intrapetrous otologic structures, the venous drainage system and the temporal lobe. Consequently, the AP should be preferred for radical removal of medium-sized, intradural, vascularized petroclival tumors like meningiomas. EEEA provides a broad access to the clivus and a unique infero-lateral exposure of the Meckel’s cave and the pterygopalatine or infratemporal fossas. It brings a close-up panoramic view of the operative field avoiding craniotomy, cerebral retraction and otologic intrapetrous structures damages [17]. Nonetheless, the three variants dangerously circumventing the ICA have their own pitfalls: the transclival variant turns around the ICA and deals with the VIth nerve and venous basilar plexus; the trans-Meckel’s cave variant crossing a virtual space between the ICA, the maxillary and mandibulary nerves requires a tumor that opens the way; the trans-pterygoid variant involves a timeconsuming drilling of the spheno-pterygoid complex with potential damages to the vidian and great palatine nerves. However, a potential CSF leak has always to be kept in mind for such large endoscopic skull base approach as well as the learning curve the endoscopy needs. To the end, it should be reserved for extradural midline tumors like chondrosarcomes or chordomas and cysts like cholesterol granulomas or epidermoid cysts which can be treated by fenestration.
Study contributions and weaknesses Many surgical and anatomic studies have already described the technical subtleties of AP or EEEA. However, cadaveric studies do not reproduce the anatomical variations resulting from a tumoral process as the displacement of cranial nerves or vessels and the structural modification of the bone. Even so, our study precisely describes both of these approaches with surgical nuances as for operative conditions. The skull base view brings a new insight of osteodural aperture, neurovascular relationships and expected surgical difficulties. This way, it could be helpful for surgeons unfamiliar with the petroclival anatomy. The illustrative quality of our study is built upon the formaldehyde preparation, the colored latex vascular injection and the use of dedicated microsurgical and endoscopic instruments. Obviously, this cadaveric study will benefit from a larger in vivo study to support our results. Literature review Recently, in 2014, a study compared AP and EEEA with CT-scan measures of bone resection. They established with reason that the two approaches reach the petroclival apex complementary and thus address to tumors with different characteristics: more medial and cystic for EEEA, more
123
Surg Radiol Anat
superolateral and fibrous for AP. Van Gompel et al. interestingly suggested that AP [30] be renamed ‘‘superior anterior petrosectomy’’ and EEEA ‘‘inferior anterior petrosectomy’’. Although quantitative analysis of bone aperture on post-dissection CT-scan would be of interest, it does not accurately fit with the surgical exposure. Indeed, the EEEA provides a wide bone opening but it is still constrained by the ICA or the cavernous sinus. Likewise, despite a small-sized bone resection, the AP includes tentorium opening, cerebrospinal fluid draining and V3 displacement that substantially broadens the surgical corridor. Therefore, it appears more relevant to consider the neurovascular structures endangered and extension possibilities of each surgical route in order to choose the best one to the petroclival region. In 2009, De Notaris et al. explored the anatomy of the petroclival region in an endoscopic study comparing the trans-sphenoidal and retro-sigmoid approaches [13]. They brought a precise description of the limits of each routes from a strictly endoscopic point of view. However, in their study, the transsphenoidal approach was purely transclival; they did not consider the possibility of lateral or inferior extension around the ICA. Also, De Notaris et al. conceded that the manipulation of the endoscope and instruments in the retrosigmoid space induced major risks for the crossed cranial nerves. Additionally, they emphasized that a specimen study cannot adequately mimic true operative conditions with their unexpected complications: arterial and venous bleeding or CSF leak. Finally, they concluded that it may be possible to use both approaches in a combined or consecutive way to allow optimal removal of petroclival tumors with lateral or posterior extensions unreachable with a single route. In 2005, Cavallo et al. reviewed cavernous sinus anatomy by correlating findings from a transcranial microsurgical route to those of EEEA [6]. They afforded a precise and applicable description of the surgical anatomy of this region and presented the advantages and limits of the EEEA compared to the microsurgical frontotemporal orbitozygomatic approach. This renowned skull base surgery team also described corridors specific to the ‘‘endoscopic transnasal route’’ reaching the cavernous sinus: a C-shaped corridor medial to the ICA and a large triangular space lateral to the ICA. They emphasized the difficulties in getting around the cranial nerves in the transcranial approach, and around the ICA in the transnasal approach. As we did, they concluded that the two approaches had their own limits: the transcranial route offered a restrained and difficult access to the ‘‘multilevel’’ architecture of the cavernous sinus, whereas the transnasal route could interestingly reach parasellar or infratemporal regions. Finally, they discussed anatomic variations caused by tumors, and how it should ease and orient the surgical approach.
123
Other approaches to the petroclival region Studies on areas of exposure in the petroclival region have shown that the surgical approach must be carefully chosen to achieve a reasonable balance between the extent of opening and the morbidity [16]. The retrosigmoid route offers a posterior access to the endocranial posterior surface of the petrous bone, the clivus and the ventrolateral side of the brainstem. This well-known simple route remains popular among neurosurgeons for access to the cerebellopontine angle and even the petroclival region without hearing damage. However, it obliges cranial nerves V to XI working through, cerebellum retraction and late access to vascular supply [9]. Siwanuwatn et al. reported that the retrosigmoid route may offer, without invasive bone resection, a working area in the petroclival region almost equivalent to that of combined petrosectomy (i.e., 292.4 ± 59.9 mm2 vs 354.1 ± 60. 3 mm2, respectively) [28]. In the retrosigmoid intradural suprameatal variant (RISA), the suprameatal tuberculum is drilled to extend access to the trigeminal porus, the cavernous sinus and Meckel’s cave. However, it requires a risky drilling close to the acoustic facial bundle and around the trigeminal nerve [7, 26]. The posterior transpetrosal approaches may be useful for attaining the petroclival region, particularly in case of tumor extension posterior to the IAM. Indeed, the retrolabyrinthine, translabyrinthine and transcochlear approaches provide an increasing petroclival exposure which comes with a longer time of drilling and greater risk of facial palsy [3, 21, 28, 32]. Finally, their corridor endangers the acoustic facial bundle and does not expose the whole petroclival area. This way they should be ‘‘combined’’ with others supra-tentorial approaches turning around the petrous bone and accessing big-sized petroclival tumors.
Conclusion From this anatomical study, it appears that each of AP and EEEA should be selectively chosen depending on their own limits and tumors characteristics. In fact, extradural midline tumors or cysts would be candidates for this anterior endoscopic approach. In contrast, intradural tumors like meningiomas should be reached by AP, particularly in case of supratentorial tumor extensions. Finally, the skull base surgeon has to master both approaches to propose the optimal one for the petroclival tumors surgical strategy. Acknowledgments We thank the technical staff of the Department of Anatomy for their preparation of specimens. Conflict of interest I hereby confirm that the authors have no conflict of interest in this manuscript.
Surg Radiol Anat
References 1. Al-Mefty O, Fox JL, Smith RR (1988) Petrosal approach for petroclival meningiomas. Neurosurgery 22:510–517 2. Al-Mefty O, Sekhar LN, Sen C, van Loveren HR (2001) Petroclival meningioma: case history and responses. Skull Base 11:143–148 3. Bambakidis NC, Kakarla UK, Kim LJ, Nakaji P, Porter RW, Daspit CP et al (2007) Evolution of surgical approaches in the treatment of petroclival meningiomas: a retrospective review. Neurosurgery 61:202–211 4. Berhouma M, Messerer M, Jouanneau E (2012) Shifting paradigm in skull base surgery: roots, current state of the art and future trends of endonasal endoscopic approaches. Rev Neurol (Paris) 168:121–134 5. Campbell E, Whitfield RD (1948) Posterior fossa meningiomas. J Neurosurg 5:131–153 6. Cavallo LM, Cappabianca P, Galzio R, Iaconetta G, de Divitiis E, Tschabitscher M (2005) Endoscopic transnasal approach to the cavernous sinus versus transcranial route: anatomic study. Neurosurgery 56:379–389 7. Chanda A, Nanda A (2006) Retrosigmoid intradural suprameatal approach: advantages and disadvantages from an anatomical perspective. Neurosurgery 59:ONS–1–ONS–6 10.1227 8. Chang SW, Wu A, Gore P, Beres E, Porter RW, Preul MC et al (2009) Quantitative comparison of Kawase’s approach versus the retrosigmoid approach: implications for tumors involving both middle and posterior fossae. Neurosurgery 64:ons44–ons52 9. Chen L, Yu X, Bu B, Xu B, Zhou D (2011) The retrosigmoid approach to petroclival meningioma surgery. J Clin Neurosci Off J Neurosurg Soc Australas 18:1656–1661 10. Cherington M, Schneck SA (1966) Clivus meningiomas. Neurology 16:86–92 11. Couldwell WT, Fukushima T, Giannotta SL, Weiss MH (1996) Petroclival meningiomas: surgical experience in 109 cases. J Neurosurg 84:20–28 12. Day JD, Fukushima T, Giannotta SL (1994) Microanatomical study of the extradural middle fossa approach to the petroclival and posterior cavernous sinus region: description of the rhomboid construct. Neurosurgery 34:1009–1016 (discussion 1016) 13. De Notaris M, Cavallo LM, Prats-Galino A, Esposito I, Benet A, Poblete J et al (2009) Endoscopic endonasal transclival approach and retrosigmoid approach to the clival and petroclival regions. Neurosurgery 65:42–50 (discussion 50–520) 14. Fournier H-D, Mercier P, Roche P-H (2007) Surgical anatomy of the petrous apex and petroclival region. Adv Tech Stand Neurosurg 32:91–146 15. Franc¸ois P, Ben Ismail M, Hamel O, Bataille B, Jan M, Velut S (2010) Anterior transpetrosal and subtemporal transtentorial approaches for pontine cavernomas. Acta Neurochir (Wien) 152:1321–1329 (discussion 1329) 16. Horgan MA, Anderson GJ, Kellogg JX, Schwartz MS, Spektor S, McMenomey SO et al (2000) Classification and quantification of
17.
18.
19.
20.
21. 22. 23. 24.
25. 26.
27. 28.
29. 30.
31.
32.
the petrosal approach to the petroclival region. J Neurosurg 93:108–112 Kassam AB, Gardner P, Snyderman C, Mintz A, Carrau R (2005) Expanded endonasal approach: fully endoscopic, completely transnasal approach to the middle third of the clivus, petrous bone, middle cranial fossa, and infratemporal fossa. Neurosurg Focus 19:E6 Kassam AB, Prevedello DM, Carrau RL, Snyderman CH, Gardner P, Osawa S et al (2009) The front door to Meckel’s cave: an anteromedial corridor via expanded endoscopic endonasal approach—technical considerations and clinical series. Neurosurgery 64:ons71–ons82 (discussion ons82–83) Kassam AB, Vescan AD, Carrau RL, Prevedello DM, Gardner P, Mintz AH et al (2008) Expanded endonasal approach: vidian canal as a landmark to the petrous internal carotid artery: technical Note. J Neurosurg 108:177–183 Kawase T, Toya S, Shiobara R, Mine T (1985) Transpetrosal approach for aneurysms of the lower basilar artery. J Neurosurg 63:857–861 McElveen JT, Dorfman BE, Fukushima T (2001) Petroclival tumors: a synthesis. Otolaryngol Clin N Am 34:1219–1230 Rhoton AL (1979) The suboccipital approach to removal of acoustic neuromas. Head Neck Surg 1:313–333 Roche P-H, Lubrano VF, Noudel R (2011) How I do it: epidural anterior petrosectomy. Acta Neurochir (Wien) 153:1161–1167 Samii M, Ammirati M, Mahran A, Bini W, Sepehrnia A (1989) Surgery of petroclival meningiomas: report of 24 cases. Neurosurgery 24:12–17 Samii M, Tatagiba M (1992) Experience with 36 surgical cases of petroclival meningiomas. Acta Neurochir (Wien) 118:27–32 Samii M, Tatagiba M, Carvalho GA (2000) Retrosigmoid intradural suprameatal approach to Meckel’s cave and the middle fossa: surgical technique and outcome. J Neurosurg 92:235–241 Sennaroglu L, Slattery WH (2003) Petrous anatomy for middle fossa approach. Laryngoscope 113:332–342 Siwanuwatn R, Deshmukh P, Figueiredo EG, Crawford NR, Spetzler RF, Preul MC (2006) Quantitative analysis of the working area and angle of attack for the retrosigmoid, combined petrosal, and transcochlear approaches to the petroclival region. J Neurosurg 104:137–142 Tedeschi H, Rhoton AL (1994) Lateral approaches to the petroclival region. Surg Neurol 41:180–216 Van Gompel JJ, Alikhani P, Tabor MH, van Loveren HR, Agazzi S, Froelich S et al (2014) Anterior inferior petrosectomy: defining the role of endonasal endoscopic techniques for petrous apex approaches. J Neurosurg 120:1321–1325 Velut S, Jan M (1988) Petrectomy of the point during approach to the clivus: technic, values and limitations. Apropos of a case of meningioma. Neurochirurgie 34(1):17–25 Xu F, Karampelas I, Megerian CA, Selman WR, Bambakidis NC (2013) Petroclival meningiomas: an update on surgical approaches, decision making, and treatment results. Neurosurg Focus 35:E11
123
ORIGINAL ARTICLE
Anatomic comparison of anterior petrosectomy versus the expanded endoscopic endonasal approach: interest in petroclival tumors surgery Timothe´e Jacquesson1,2 • Emile Simon2 • Moncef Berhouma1 • Emmanuel Jouanneau1
Received: 3 February 2015 / Accepted: 23 May 2015 Ó Springer-Verlag France 2015
Abstract Purpose Since the petroclival region is deep-seated with close neurovascular relationships, the removal of petroclival tumors still represents a fascinating surgical challenge. Although the classical anterior petrosectomy (AP) offers a meaningful access to this petroclival region, the expanded endoscopic endonasal approach (EEEA) recently leads to overcome difficulties from trans-cranial approaches. Herein, we present an anatomic comparison of AP versus EEEA. We aim to describe the limits of these both approaches helping the choice of the optimal surgical route for petroclival tumors. Methods Six fresh cadaveric heads were harvested and injected with colored latex. Each approach was step-bystep detailed until its final surgical exposure. Results The AP provided a narrow direct supero-lateral access to the petroclival area that can also reach the cavernous sinus, the retrochiasmatic region and perimesencephalic cisterns. However, this corridor anterior to the internal acoustic meatus passed on each side of the trigeminal nerve. Moreover, tumor extensions toward the foramen jugularis, inside the clivus or behind the internal acoustic meatus were difficult to control. The EEEA brought a straightforward access to the clivus but the petrous apex was hidden behind the internal carotid artery.
Several variants were described: a medial transclival, a lateral through the Meckel’s cave and an inferior transpterygoid route. Elsewhere, tumor extension behind the internal acoustic meatus or above the tentorium could not be satisfactorily assessed. Discussion and conclusion PA and EEEA have their own limits in reaching the petroclival region in accordance with the tumor characteristics. The AP should be preferred for radical removal of middle-sized petrous apex intradural tumors like meningiomas. The EEEA would be of interest for extradural midline tumors like chordomas or for petrous apex cysts drainage.
& Timothe´e Jacquesson [email protected]
Introduction
1
Skull Base Multi-disciplinary Unit, Department of Neurosurgery B, Neurological Hospital Pierre Wertheimer, Hospices Civils de Lyon, 59 Bd Pinel, 69677 Lyon Cedex, France
2
Department of Anatomy, University of Lyon 1, 8 Avenue Rockefeller, 69003 Lyon, France
Keywords Skull base Petroclival Petrosectomy Endoscopy Endonasal Abbreviations AP Anterior petrosectomy EEEA Expanded endoscopic endonasal approach IAM Internal acoustic meatus ICA Internal carotid artery SPN Superficial petrosal nerve UPR Upper petrous ridge
The petroclival region appears rather as a ‘‘surgical’’ space limited anteriorly by the clivus, laterally by the petrous apex, medially by the brainstem and posteriorly by the internal acoustic meatus (IAM) [14] (Fig. 1). Since it is deep-seated and closely related with numerous neurovascular structures, reaching the petroclival region remains a
123
Surg Radiol Anat b Fig. 1 The petroclival region, photographs. Superior views of the
middle cranial fossa (a), posterior side of the petrous bone (b) and superior side of the petrous bone (c). The petroclival region includes the petrous apex and the clivus, from the dorsum sellae (DS) to the foramen jugularis (FJ), medially to the internal acoustic meatus (IAM). Around the petroclival region are described: anteriorly at the midline the sella turcica (ST) and the tuberculum sellae (TS), posteriorly the foramen magnum and laterally from front to back the optic canal (OC), the anterior clinoid process (AC), the superior orbital fissure (SOF), the foramen rotondum (FR), the foramen ovale (FO), the foramen spinosum (FS) and the sigmoid sinus (SS). A stairstep can be found at the upper petrous ridge (UPR) and falsely delimits this last. The horizontal intrapetrous carotid canal (CC) is parallel and medial to the superficial petrosal nerves (SPN) which are also parallel and medial to the line formed by the foramen ovale and spinosum. The axis of the IAM is indicated by the bisectrix of the SPN and the eminentia arcuata (EA), whose summit is the geniculate ganglion. Laterally, the tegmen (T) appears as a thin bone layer separating the middle cranial fossa and the middle ear
described first by Kawase then by Velut [30, 24]. This transpetrous route called anterior petrosectomy (AP) required a dangerous drilling around the otologic structures, the trigeminal nerve and the internal carotid artery (ICA). Third, the development of endoscopy in neurosurgery allowed for an anterior endonasal transsphenoidal access to the cranial base thus preventing cerebrum retraction and intrapetrous structures damages [4, 17]. Likewise, the expanded endoscopic endonasal approach (EEEA) provided a greater exposure, vertically from the cribriform plate to the craniovertebral junction and laterally to the parasellar or infratemporal fossas [6, 17, 18]. Thereby, we performed an anatomical comparison of PA and EEEA describing the limits of each and helping petroclival tumors surgical strategy.
Materials and methods Specimen preparation
fascinating challenge in skull base surgery. Thereafter, petroclival tumors have long been considered inoperable due to incomplete resection and high morbidity [5, 10]. More recently, improvements in surgical technique, operative microscope and electrophysiological monitoring raised hope to achieve a more safe efficient surgery in this petroclival area. Then, several renowned neurosurgeons reported their experiences and various surgical routes to these tumors [1, 2, 11, 24, 25]. First, a posterior, retrosigmoid approach was proposed, but its corridor crossed cranial nerves V to IX with significant risks [22, 29]. Second, a lateral approach was
123
Six fresh cadaver heads were harvested at the anatomy laboratory of University Lyon 1 (Lyon, France) and prepared with 10 % formaldehyde. Within 48 h of death, the common carotid arteries, the vertebral arteries (VA) and the internal jugular veins were catheterized, then the entire vascular network was washed and injected with red (arteries) and blue (veins) latex (Phoce´enne de chimie—Marseille/France; Ae´rographe Colorex Technics, Magenta et Cyan). Anatomic study Six APs were performed with microsurgical instruments (ClimdalÒ Instrumentation Me´dico-Chirurgicale, Lyon, France) and operative microscope (Carl ZeissÒ Microscopy Gmbh, Go¨ttingen, Germany) according to the technique
Surg Radiol Anat
described by Kawase and Velut [20, 31]. Similarly, six EEEAs were performed controlaterally with dedicated endoscopic devices (Karl StorzÒ Endoskope GmbH, Tuttlingen, Germany) according to the technique described by Kassam [17]. The surgical anatomy was precisely detailed at each step for both approaches. Between each dissection phase, specimens were conserved in baths of 10 % ethanol and 1 % formaldehyde. Craniotomies and drilling necessitated a continuous irrigated high-speed motor (Integrated Power Console and Handpieces, MedtronicÒ, Louisville, USA). Post-approach skull base study After completion of AP and contralateral EEEA on each specimen, the calvaria and brain were cautiously removed, leaving in place the cranial nerves, the lower part of the brainstem and the arterial circle of Willis. No further modifications were made to the surgical routes. The osteodural opening, the working space, the neurovascular relationships and the possibilities of extension were specified for both approaches.
Results Microsurgical anterior petrosectomy Surgical technique Specimens were placed to obtain a nearly vertical alignment of the external acoustic meatus. After a curvilinear incision, a classic temporal craniotomy was made as low as possible then the dura mater was elevated posteriorly to anteriorly discovering the superior side of the petrous part of the temporal bone. Having sectioned the middle meningeal artery, the ‘‘rhomboid construct’’ [12] was discovered with: the foramen spinosum, the foramen ovale, the eminencia arcuata, the superficial petrosal nerves (SPN), the carotid canal and the upper petrous ridge (UPR) (Figs. 1, 2). The axis of the foramen spinosum and the foramen ovale gave medially the SPN and then the horizontal canal of the internal carotid artery (ICA). A 100° on average angle was observed between the arcuate eminence and the SPN and the bisectrix of this angle drew the anterior edge of the internal acoustic meatus (IAM) [27]. At the summit of this angle sat the geniculate ganglion and just forward the cochlea. A ‘‘staircase’’ groove for the superior petrosal sinus might be seen at the UPR. The meningeal sheet of foramen ovale dura mater had to be cut easing an anterior displacement of the mandibular nerve and releasing the petrous apex. The Kawase’s triangle was carefully drilled from the trigeminal print to the IAM and
from the UPR to the inferior petrosal sinus with approximatively 20 mm (mm) width and 10 mm depth. Care had to be taken to ICA or SPN laterally and cochlea or acoustic facial bundle posteriorly. Then, it allowed for tentorium splitting and posterior and middle cranial fossa dura mater opening. It thus offered a broad supra/infra-tentorial exposure of the petroclival region through the drilled petrous apex. Limits This approach provided a direct, narrow, superolateral access to the petroclival region bounded: superiorly by the IVth nerve and the temporal lobe; inferiorly by the inferior petrosal sinus and the ascending cisternal segment of the VIth nerve with its course into the Dorello’s canal; anteriorly by the clivus; medially by the brainstem and the vertebrobasilar system and posteriorly by the IAM which contained the VIIth and VIIIth nerves. In addition to give access to Meckel’s cave and the cavernous sinus, this surgical corridor could be widened to the entire middle cranial fossa, the retrochiasmatic space and perimesencephalic cisterns. Nonetheless, this route required a retraction of the temporal lobe working dangerously on either side of the trigeminal nerve. Drilling was also delicately close to intrapetrous otologic structures like geniculate ganglion, cochlea, semicircular canals and tegmen tympani. It came with risk of facial palsy, deafness, balance disorders and CSF leak, respectively. Also, tumor extensions could not be controlled inferiorly toward the foramen jugularis, anteriorly toward the clivus and posteriorly behind the IAM (Fig. 4). Finally, enlarging toward the midline appeared possible only if the basilar artery and the brainstem were displaced by a tumor. Expanded endoscopic endonasal approach Surgical technique of the transsphenoidal approach Specimens were placed facing the operator in a supine position. A 180-mm length and 4-mm diameter endoscope was used with 0° or 30° lenses (Karl StorzÒ Endoskope GmbH, Tuttlingen, Germany). A bi-nostril path was necessary and a self-retaining endoscope holder allowed for a two-hands free technique by a single operator. Moving along the nasal floor and the inferior turbinate led to the choanal arch, which was a reliable sagittal landmark matching the floor of the sphenoid sinus and containing the septal branch of the sphenopalatine artery. The pharyngeal wall was visualized with the ostium of the Eustachian tube laterally and the nasal septum medially. The ostium sphenoidale was most frequently located 10 mm on average above the choanal arch. A septal mucosa incision
123
Surg Radiol Anat
Fig. 2 Anterior petrosectomy. Operative views. Posterior and inferior to the coronal and squamosal bone sutures, a classic low temporal craniotomy is performed. Then, a careful dural dissection discovers the ‘‘rhomboid construct’’ at the top of the petrous bone with, from back to front (a, b): the upper petrous ridge (UPR), the eminentia arcuata (AE), the internal acoustic meatus (IAM), the superficial petrosal nerve (SPN), internal carotid artery (ICA), the middle meningeal artery (MMA) in its foramen spinosum (FS) and the three branches of the trigeminal nerve. Drilling starts ahead of the IAM until the trigeminal print and inferiorly toward the posterior cranial fossa dura. Sometimes uncovered by bone, the horizontal petrous segment of the ICA is located just medial and parallel to the SPN.
After opening the temporal and posterior cranial fossas dura and the tentorium, the final exposition is obtained (c, d). The trigeminal nerve crosses the corridor and gives its branches V1, V2 and V3 with their respective destination: the superior orbital fissure (SOF), the foramen rotondum (FR) and the foramen ovale (FO). Deeply, the antero-lateral side of the brainstem is reached. Superiorly, the IVth nerve passes under the tentorium before entering the cavernous sinus (CS) whom dura mater (DM) can be peeled. Anteriorly, the VIth nerve courses before its Dorello’s canal. Inferiorly, the inferior petrosal sinus marks out the lower limit of the access while the IAM can be opened posteriorly
started from this ostium discovering the rostral spur of the vomer then the contralateral ostium sphenoidale. A wide sphenoidotomy was performed, identifying the classical intrasphenoid landmarks: the paracavernous and paraclival segments of ICAs, the optic canals, the lateral and medial opticocarotid recesses, the sella turcica, the paraclival recess, the planum sphenoidale and posterior ethmoidal cells. Thereby, the posterior third of the nasal septum and the middle turbinate were removed. The bone opening was adapted on demand to provide sufficient exposure and maneuverability. Finally, access to the entire clivus was straightforward, but access to the petrous apex needed circumventing the ICA. Thus, three variants could be decribed: medial, lateral and below the ICA (Fig. 3).
Variants and limits
123
The first medial variant was trans-sphenoidal trans-clival. Indeed, this route medial to the ICA went through the clivus with its venous plexus and reached the tip of the petrous pyramid surrounded: superiorly by the pituitary fossa, inferiorly by the sphenoid floor and posteriorly by the basilar artery with the brainstem and the VIth nerve. The abducens nerve emerged from the pons at the midline level of the sphenoid floor. It then coursed obliquely to join Dorello’s canal at the superior third of the vertical part of the ICA paraclival segment. Finally, the aperture could be easily extended toward the contralateral side contrary to a lateral extension which would be hindered by the ICA.
Surg Radiol Anat
The second lateral variant was trans-sphenoidal transMeckel’s cave involving a lateral enlargement of the corridor as middle turbinectomy and posterior ethmoidectomy. Likewise, the vertical paraclival ICA segment was cautiously skeletonized; the orbital apex, the cavernous sinus and the Meckel’s cave were exposed. The petrous apex was thus reached through the Meckel’s cave. This corridor passing between the trigeminal branches and the ICA was
limited: superiorly by the cavernous sinus with ascending V1, VIth nerve and orbital apex, inferiorly and medially by the petrous and paracavernous ICA and laterally by the sphenoid bone with its foramen rotondum and ovale. Opening these last allowed for pursuing V2 and V3 anteriorly. Having medially displaced the anterior loop of the ICA, the IIIrd and IVth nerves were also identified into the cavernous sinus.
Fig. 3 Expanded endoscopic endonasal approach. Endoscopic views. A classic endoscopic endonasal trans-sphenoidal approach is performed with a 0° lens endoscope via the two nostrils. First, a large sphenoidotomy exposes the classical intra-sphenoidal landmarks (a, b) from up to down: ethmoidal cells (EC) with the posterior ethmoidal artery (PEA) and planum, optic canals (OC), medial and lateral optico-carotid recesses (mOCR, lOCR), sella turcica (ST), internal carotid artery (ICA) with its paraclival and paracavernous segments, paraclival recess (PCR) and the always medial vomer bone. The lOCR known as ‘‘optic strut’’ separates the paracavernous ICA and the ON and leads to the anterior clinoid process. Then, access to the clivus is large and straightforward, but the ICA needs to be circumvented to reach the petrous apex. Three variants could be described: a medial trans-clival route, a lateral route through the Meckel’s cave and an inferior trans-pterygoid route (C, D). The first variant (red square) passes through the clivus to reach the tip of the petrous pyramid medial to the ICA under the VIth nerve. The second
variant (red circle) requires a lateral bone extension including ICA skeletonization. This way the orbital apex, the cavernous sinus and Meckel’s cave are exposed with their own contents: cranial nerves III, IV, V1, V2, V3 and VI. The third variant (red star) needs a long drilling of the sphenoid floor (Sph) and the medial pterygoid plate. The vidian nerve (VN) leads to the foramen lacerum (FL) and the path continues below the horizontal ICA and above the Eustachian tube (ET) (c). Through the middle clivus, the pons is discovered with the basilar artery (BA) and its branches: the superior cerebellar artery (SCA) near the IIIrd nerve (III) and the anterior inferior cerebellar artery (AICA). The VIth nerve crosses obliquely from the pons to Dorello’s canal behind the vertical paraclival ICA. Deeply, the drilling attains the internal acoustic meatus with its acoustic-facial bundle (VII–VIII). Finally, this approach can be also expanded up to the sellar, parasellar and suprasellar areas (d) including the pituitary gland (PG), the optic and olfactory nerves (ON, Olf) (color figure online)
123
Surg Radiol Anat Fig. 4 Skull base photographs after right anterior petrosectomy and c left expanded endoscopic endonasal approach. These views accurately show the anatomical limits of each approach. From laterally (a), the anterior petrosectomy (AP) follows the top of the petrous bone to find out successively from back to front: the upper petrous ridge (UPR), the middle meningeal artery (MMA) in its foramen spinosum (FS), the three trigeminal branches (V3, V2 and V1). Drilling the petrous apex gives a narrow access to the petroclival region working on each side of the trigeminal nerve (V). This corridor can be enlarged to expose the whole middle cranial fossa until the lesser sphenoid wing (lSw) as the cavernous sinus and the Meckel’s cave with theirs contents: cranial nerves III, IV, V1, V2, V3 and VI. Likewise, any supra-tentorial tumor extensions can be controlled as the retrochiasmatic space (behind the chiasma, Ch) or perimesencephalic cisterns. From medially (b), the AP’s corridor appears parallel to the acoustic facial bundle (VII–VIII) and just behind the VIth nerve. The basilar artery (BA), the anterior inferior cerebellar artery (AICA) and the brainstem (BT) are crossing through. That is why a contralateral or midline exposure is possible only if a tumor opens the way. From posteriorly (c), the expanded endoscopic endonasal approach (EEEA) offers a large osteodural aperture that can be widened to the pituitary and anterior cranial fossas. The medial transclival variant crosses the clivus and jeopardizes the VIth nerve and the basilar artery. The lateral trans-Meckel’s cave variant needs a tumor that pushes aside V2 and V3 branches from the ICA. The inferior trans-pterygoid route passes between the Eustachian tube and the ICA; it then gets close to the internal acoustic meatus and the foramen jugularis with cranial nerves VII to XI. Despite the dangerous path around the ICA to the hidden petrous apex, the EEEA provides a large access to the clivus with multiple specific expansions as into the pterygopalatine fossa, the infratemporal fossa, the cavernous sinus, the Meckel’s cave, the contralateral side and the suprasellar area
The inferior third variant was trans-sphenoidal transpterygoid. The maxillary sinus and pterygopalatine fossa opening and the sphenopalatine artery cutting were required. The pterygopalatine ganglion met the maxillary, vidian and great palatine nerves and was then lateralized. The pterygoid canal with its vidian nerve and artery was found out at the intersection of the sphenoid floor and the medial pterygoid plate [19]. A long drilling of the upper portion of the medial pterygoid plate along the pterygoid canal thus led to the foramen lacerum with its fibrous proximal ring. The drilling was continued under the horizontal portion of the ICA releasing the cartilaginous part of the Eustachian tube and reaching the ‘‘inferior’’ petrous apex [30] surrounded: superiorly by the horizontal ICA, inferiorly by the Eustachian tube, laterally by the greater wing of the sphenoid and medially by the median sphenoid floor. Posteriorly, the foramen jugularis as far as the IAM was attained. Similarly, the acoustic facial bundle was identified with the labyrinthine artery and the choroid plexus in the lateral aperture of the fourth ventricle. Finally, for all variants, tumor extensions behind the ICA or the IAM are not accessible contrary to sellar, parasellar and supra-sellar areas (Fig. 4).
123
Discussion Principal findings AP and EEEA offer different routes to the petroclival region with their own limits (Table 1). AP provides a narrow access to the petrous apex, the upper clivus, the trigeminal porus and the antero-lateral side of the brainsteam [8, 15, 20, 23, 31]. Having tailored the craniotomy, the corridor could be enlarged toward the retrochiasmatic space, perimesencephalic cisterns and the whole middle cranial fossa. It eases the control of tumor vascular supply and bone invasion. However, it runs anterior to the acoustic facial
Surg Radiol Anat Table 1 Pros and cons of anterior petrosectomy versus expanded endoscopic endonasal approach Anterior petrosectomy
Expanded endoscopic endonasal approach
Technique
Microscope
Endoscope
Working space
Wide superficially, narrow deeply
Narrow superficially, wide deeply
Distance to the target
Short
Long
Surgical corridor
oblique
Straightforward
Handiness
Limited deeply
Limited superficially
Visibility
3D, microscope
Panoramic, 2D, ‘‘barrel effect’’
Skills
Skull base surgery
Endoscopy
Anatomy
Petrous bone
Nasal pits, sphenoid bone
Advantages
Classic sub temporal approach
Minimally invasive
Warnings
Side effects
Vascular supply access
Corner and close up views
extension to supratentorial area extension to MC, CS
Extension to suprasellar area Extension to CVJ, ITF, MC, CS
Crossing IVth, Vth and VIth nerves
Crossing VIth and vidian nerves
Risks of cochlea, SCC, SPN damages
Risks of ICA, ET damages
Temporal lobe retraction, craniotomy
CSF leak, skull base reconstruction
Limited inferior access
Limited superior access
Scar, headaches
Empty nose sensation, rhinitis
This table summarizes the technical subtleties, advantages and disadvantages of anterior petrosectomy and the expanded endoscopic endonasal approach to help the choice of the optimal surgical route for petroclival tumors surgery 2D/3D bi/tridimensional, CS cavernous sinus, CSF cerebrospinal fluid, CVJ craniovertebral junction, ET eustachian tube, ICA internal carotid artery, MC Meckel’s cave, ITF infratemporal fossa, SCC semicircular canals, SPN superficial petrosal nerves
bundle, AP still have to cross the pathetic and trigeminal nerves. Also, it jeopardizes the horizontal ICA, the intrapetrous otologic structures, the venous drainage system and the temporal lobe. Consequently, the AP should be preferred for radical removal of medium-sized, intradural, vascularized petroclival tumors like meningiomas. EEEA provides a broad access to the clivus and a unique infero-lateral exposure of the Meckel’s cave and the pterygopalatine or infratemporal fossas. It brings a close-up panoramic view of the operative field avoiding craniotomy, cerebral retraction and otologic intrapetrous structures damages [17]. Nonetheless, the three variants dangerously circumventing the ICA have their own pitfalls: the transclival variant turns around the ICA and deals with the VIth nerve and venous basilar plexus; the trans-Meckel’s cave variant crossing a virtual space between the ICA, the maxillary and mandibulary nerves requires a tumor that opens the way; the trans-pterygoid variant involves a timeconsuming drilling of the spheno-pterygoid complex with potential damages to the vidian and great palatine nerves. However, a potential CSF leak has always to be kept in mind for such large endoscopic skull base approach as well as the learning curve the endoscopy needs. To the end, it should be reserved for extradural midline tumors like chondrosarcomes or chordomas and cysts like cholesterol granulomas or epidermoid cysts which can be treated by fenestration.
Study contributions and weaknesses Many surgical and anatomic studies have already described the technical subtleties of AP or EEEA. However, cadaveric studies do not reproduce the anatomical variations resulting from a tumoral process as the displacement of cranial nerves or vessels and the structural modification of the bone. Even so, our study precisely describes both of these approaches with surgical nuances as for operative conditions. The skull base view brings a new insight of osteodural aperture, neurovascular relationships and expected surgical difficulties. This way, it could be helpful for surgeons unfamiliar with the petroclival anatomy. The illustrative quality of our study is built upon the formaldehyde preparation, the colored latex vascular injection and the use of dedicated microsurgical and endoscopic instruments. Obviously, this cadaveric study will benefit from a larger in vivo study to support our results. Literature review Recently, in 2014, a study compared AP and EEEA with CT-scan measures of bone resection. They established with reason that the two approaches reach the petroclival apex complementary and thus address to tumors with different characteristics: more medial and cystic for EEEA, more
123
Surg Radiol Anat
superolateral and fibrous for AP. Van Gompel et al. interestingly suggested that AP [30] be renamed ‘‘superior anterior petrosectomy’’ and EEEA ‘‘inferior anterior petrosectomy’’. Although quantitative analysis of bone aperture on post-dissection CT-scan would be of interest, it does not accurately fit with the surgical exposure. Indeed, the EEEA provides a wide bone opening but it is still constrained by the ICA or the cavernous sinus. Likewise, despite a small-sized bone resection, the AP includes tentorium opening, cerebrospinal fluid draining and V3 displacement that substantially broadens the surgical corridor. Therefore, it appears more relevant to consider the neurovascular structures endangered and extension possibilities of each surgical route in order to choose the best one to the petroclival region. In 2009, De Notaris et al. explored the anatomy of the petroclival region in an endoscopic study comparing the trans-sphenoidal and retro-sigmoid approaches [13]. They brought a precise description of the limits of each routes from a strictly endoscopic point of view. However, in their study, the transsphenoidal approach was purely transclival; they did not consider the possibility of lateral or inferior extension around the ICA. Also, De Notaris et al. conceded that the manipulation of the endoscope and instruments in the retrosigmoid space induced major risks for the crossed cranial nerves. Additionally, they emphasized that a specimen study cannot adequately mimic true operative conditions with their unexpected complications: arterial and venous bleeding or CSF leak. Finally, they concluded that it may be possible to use both approaches in a combined or consecutive way to allow optimal removal of petroclival tumors with lateral or posterior extensions unreachable with a single route. In 2005, Cavallo et al. reviewed cavernous sinus anatomy by correlating findings from a transcranial microsurgical route to those of EEEA [6]. They afforded a precise and applicable description of the surgical anatomy of this region and presented the advantages and limits of the EEEA compared to the microsurgical frontotemporal orbitozygomatic approach. This renowned skull base surgery team also described corridors specific to the ‘‘endoscopic transnasal route’’ reaching the cavernous sinus: a C-shaped corridor medial to the ICA and a large triangular space lateral to the ICA. They emphasized the difficulties in getting around the cranial nerves in the transcranial approach, and around the ICA in the transnasal approach. As we did, they concluded that the two approaches had their own limits: the transcranial route offered a restrained and difficult access to the ‘‘multilevel’’ architecture of the cavernous sinus, whereas the transnasal route could interestingly reach parasellar or infratemporal regions. Finally, they discussed anatomic variations caused by tumors, and how it should ease and orient the surgical approach.
123
Other approaches to the petroclival region Studies on areas of exposure in the petroclival region have shown that the surgical approach must be carefully chosen to achieve a reasonable balance between the extent of opening and the morbidity [16]. The retrosigmoid route offers a posterior access to the endocranial posterior surface of the petrous bone, the clivus and the ventrolateral side of the brainstem. This well-known simple route remains popular among neurosurgeons for access to the cerebellopontine angle and even the petroclival region without hearing damage. However, it obliges cranial nerves V to XI working through, cerebellum retraction and late access to vascular supply [9]. Siwanuwatn et al. reported that the retrosigmoid route may offer, without invasive bone resection, a working area in the petroclival region almost equivalent to that of combined petrosectomy (i.e., 292.4 ± 59.9 mm2 vs 354.1 ± 60. 3 mm2, respectively) [28]. In the retrosigmoid intradural suprameatal variant (RISA), the suprameatal tuberculum is drilled to extend access to the trigeminal porus, the cavernous sinus and Meckel’s cave. However, it requires a risky drilling close to the acoustic facial bundle and around the trigeminal nerve [7, 26]. The posterior transpetrosal approaches may be useful for attaining the petroclival region, particularly in case of tumor extension posterior to the IAM. Indeed, the retrolabyrinthine, translabyrinthine and transcochlear approaches provide an increasing petroclival exposure which comes with a longer time of drilling and greater risk of facial palsy [3, 21, 28, 32]. Finally, their corridor endangers the acoustic facial bundle and does not expose the whole petroclival area. This way they should be ‘‘combined’’ with others supra-tentorial approaches turning around the petrous bone and accessing big-sized petroclival tumors.
Conclusion From this anatomical study, it appears that each of AP and EEEA should be selectively chosen depending on their own limits and tumors characteristics. In fact, extradural midline tumors or cysts would be candidates for this anterior endoscopic approach. In contrast, intradural tumors like meningiomas should be reached by AP, particularly in case of supratentorial tumor extensions. Finally, the skull base surgeon has to master both approaches to propose the optimal one for the petroclival tumors surgical strategy. Acknowledgments We thank the technical staff of the Department of Anatomy for their preparation of specimens. Conflict of interest I hereby confirm that the authors have no conflict of interest in this manuscript.
Surg Radiol Anat
References 1. Al-Mefty O, Fox JL, Smith RR (1988) Petrosal approach for petroclival meningiomas. Neurosurgery 22:510–517 2. Al-Mefty O, Sekhar LN, Sen C, van Loveren HR (2001) Petroclival meningioma: case history and responses. Skull Base 11:143–148 3. Bambakidis NC, Kakarla UK, Kim LJ, Nakaji P, Porter RW, Daspit CP et al (2007) Evolution of surgical approaches in the treatment of petroclival meningiomas: a retrospective review. Neurosurgery 61:202–211 4. Berhouma M, Messerer M, Jouanneau E (2012) Shifting paradigm in skull base surgery: roots, current state of the art and future trends of endonasal endoscopic approaches. Rev Neurol (Paris) 168:121–134 5. Campbell E, Whitfield RD (1948) Posterior fossa meningiomas. J Neurosurg 5:131–153 6. Cavallo LM, Cappabianca P, Galzio R, Iaconetta G, de Divitiis E, Tschabitscher M (2005) Endoscopic transnasal approach to the cavernous sinus versus transcranial route: anatomic study. Neurosurgery 56:379–389 7. Chanda A, Nanda A (2006) Retrosigmoid intradural suprameatal approach: advantages and disadvantages from an anatomical perspective. Neurosurgery 59:ONS–1–ONS–6 10.1227 8. Chang SW, Wu A, Gore P, Beres E, Porter RW, Preul MC et al (2009) Quantitative comparison of Kawase’s approach versus the retrosigmoid approach: implications for tumors involving both middle and posterior fossae. Neurosurgery 64:ons44–ons52 9. Chen L, Yu X, Bu B, Xu B, Zhou D (2011) The retrosigmoid approach to petroclival meningioma surgery. J Clin Neurosci Off J Neurosurg Soc Australas 18:1656–1661 10. Cherington M, Schneck SA (1966) Clivus meningiomas. Neurology 16:86–92 11. Couldwell WT, Fukushima T, Giannotta SL, Weiss MH (1996) Petroclival meningiomas: surgical experience in 109 cases. J Neurosurg 84:20–28 12. Day JD, Fukushima T, Giannotta SL (1994) Microanatomical study of the extradural middle fossa approach to the petroclival and posterior cavernous sinus region: description of the rhomboid construct. Neurosurgery 34:1009–1016 (discussion 1016) 13. De Notaris M, Cavallo LM, Prats-Galino A, Esposito I, Benet A, Poblete J et al (2009) Endoscopic endonasal transclival approach and retrosigmoid approach to the clival and petroclival regions. Neurosurgery 65:42–50 (discussion 50–520) 14. Fournier H-D, Mercier P, Roche P-H (2007) Surgical anatomy of the petrous apex and petroclival region. Adv Tech Stand Neurosurg 32:91–146 15. Franc¸ois P, Ben Ismail M, Hamel O, Bataille B, Jan M, Velut S (2010) Anterior transpetrosal and subtemporal transtentorial approaches for pontine cavernomas. Acta Neurochir (Wien) 152:1321–1329 (discussion 1329) 16. Horgan MA, Anderson GJ, Kellogg JX, Schwartz MS, Spektor S, McMenomey SO et al (2000) Classification and quantification of
17.
18.
19.
20.
21. 22. 23. 24.
25. 26.
27. 28.
29. 30.
31.
32.
the petrosal approach to the petroclival region. J Neurosurg 93:108–112 Kassam AB, Gardner P, Snyderman C, Mintz A, Carrau R (2005) Expanded endonasal approach: fully endoscopic, completely transnasal approach to the middle third of the clivus, petrous bone, middle cranial fossa, and infratemporal fossa. Neurosurg Focus 19:E6 Kassam AB, Prevedello DM, Carrau RL, Snyderman CH, Gardner P, Osawa S et al (2009) The front door to Meckel’s cave: an anteromedial corridor via expanded endoscopic endonasal approach—technical considerations and clinical series. Neurosurgery 64:ons71–ons82 (discussion ons82–83) Kassam AB, Vescan AD, Carrau RL, Prevedello DM, Gardner P, Mintz AH et al (2008) Expanded endonasal approach: vidian canal as a landmark to the petrous internal carotid artery: technical Note. J Neurosurg 108:177–183 Kawase T, Toya S, Shiobara R, Mine T (1985) Transpetrosal approach for aneurysms of the lower basilar artery. J Neurosurg 63:857–861 McElveen JT, Dorfman BE, Fukushima T (2001) Petroclival tumors: a synthesis. Otolaryngol Clin N Am 34:1219–1230 Rhoton AL (1979) The suboccipital approach to removal of acoustic neuromas. Head Neck Surg 1:313–333 Roche P-H, Lubrano VF, Noudel R (2011) How I do it: epidural anterior petrosectomy. Acta Neurochir (Wien) 153:1161–1167 Samii M, Ammirati M, Mahran A, Bini W, Sepehrnia A (1989) Surgery of petroclival meningiomas: report of 24 cases. Neurosurgery 24:12–17 Samii M, Tatagiba M (1992) Experience with 36 surgical cases of petroclival meningiomas. Acta Neurochir (Wien) 118:27–32 Samii M, Tatagiba M, Carvalho GA (2000) Retrosigmoid intradural suprameatal approach to Meckel’s cave and the middle fossa: surgical technique and outcome. J Neurosurg 92:235–241 Sennaroglu L, Slattery WH (2003) Petrous anatomy for middle fossa approach. Laryngoscope 113:332–342 Siwanuwatn R, Deshmukh P, Figueiredo EG, Crawford NR, Spetzler RF, Preul MC (2006) Quantitative analysis of the working area and angle of attack for the retrosigmoid, combined petrosal, and transcochlear approaches to the petroclival region. J Neurosurg 104:137–142 Tedeschi H, Rhoton AL (1994) Lateral approaches to the petroclival region. Surg Neurol 41:180–216 Van Gompel JJ, Alikhani P, Tabor MH, van Loveren HR, Agazzi S, Froelich S et al (2014) Anterior inferior petrosectomy: defining the role of endonasal endoscopic techniques for petrous apex approaches. J Neurosurg 120:1321–1325 Velut S, Jan M (1988) Petrectomy of the point during approach to the clivus: technic, values and limitations. Apropos of a case of meningioma. Neurochirurgie 34(1):17–25 Xu F, Karampelas I, Megerian CA, Selman WR, Bambakidis NC (2013) Petroclival meningiomas: an update on surgical approaches, decision making, and treatment results. Neurosurg Focus 35:E11
123
