Surgical Anatomy of the Retrosigmoid Approach with Endoscopic-Assisted Reverse Anterior Petrosectomy: Optimizing Meckel's Cave Access from the Posterior Fossa

Abstract

Objectives

We investigated the extent of access to Meckel's cave (MC) and the middle cranial fossa (MCF) protecting the internal carotid artery (ICA) using the retrosigmoid approach with endoscopic-assisted reverse anterior petrosectomy (EA-RAP).

Methods

Five specimens were dissected using the limited and extended EA-RAP. Based on the bone removal of the internal acoustic meatus (IAM) and subarcuate fossa, exposure of the MC and ICA were statistically compared.

Results

The limited and extended EA-RAP allowed access to the medial and anterior MC (4 mm posterior to the first genu of the cavernous ICA, and 20 mm posterior to foramen rotundum [FR]). The access to the lateral MC varied with distance of 12 and 8 mm medial to the foramen ovale for the limited and extended EA-RAP, respectively.

In the extended EA-RAP, the exposure of the ICA was gained by drilling with the 0-degree endoscope (3 mm) versus 45-degree endoscope (9 mm). The working distances from the midpoint of the IAM to the most medial point of the exposed ICA was 24 mm. The most lateral point of the exposed ICA varied between 0- and 45-degree endoscopes with a distance of 21 and 13 mm, respectively.

Conclusion

A coronal plane from the posterior genu of the cavernous ICA and a sagittal plane to the common crus of the semicircular canals can define the area of MCF accessed by the EA-RAP. Drilling of the temporal bone should be carefully customized according to the patient and can be aided by endoscopic assistance for direct visualization to minimize the risk of injuries to ICA.

Introduction

The retrosigmoid (RS) approach with endoscopic-assisted reverse anterior petrosectomy (EA-RAP) also referred to as suprameatal extension of the RS represents a key surgical approach to access the middle cranial fossa (MCF) from the cerebellopontine angle (CPA). ¹ From an infratentorial perspective, access to the MCF can be achieved through the drilling of the suprameatal tubercle (SMT), performing the so-called RAP, retrosigmoid intradural suprameatal approach (RISA), and internal petrosectomy. ² The SMT is a bony prominence located on the posterior surface of the petrous bone. It is bordered inferiorly by the superior aspect of the internal acoustic meatus (IAM) and acoustic-facial bundle, superiorly by the superior petrosal sinus (SPS) and tentorium cerebelli, superomedially by the trigeminal nerve (TN) and Meckel's cave (MC), and laterally by the subarcuate fossa (SF) and the superior and posterior semicircular canals. ^{3 4 5} Drilling the SMT enables exposure of the lateral margin of MC, Gasserian ganglion (GG), and inferomedial prepontine area. ⁶ RAP has traditionally been employed for surgical resection of the posterior fossa lesions with MC and MCF extension. ^{3 6 7} Several studies have sought to compare the exposure achieved by RAP with alternative transcranial approaches. ^{8 9 10} However, a few studies evaluated the maximal extension of bony removal in this region to obtain the furthest trajectory along the course of the trigeminal nerve, where many of these affecting tumors arise or may infiltrate. ^{5 11}

The primary aim of this cadaveric study was to investigate and compare the extent of access to the MC obtained by an EA-RAP with limited SMT removal and its maximally extended version with further endoscopic-assisted drilling of the petrous apex. Additionally, our secondary aim was to identify crucial landmarks to locate the internal carotid artery (ICA) during extended bony removal and, thus, to provide anatomical guidance on preventing iatrogenic vascular injuries in this specific region.

Materials and Methods

This study was reviewed and approved by the Mayo Clinic institutional review board and biospecimens committee (IRB Protocol-17–005898), as required standard protocols. Five formalin-fixed specimens (10 sides), injected with colored latex using a six-vessel technique previously described by our group, ¹² were dissected in the anatomy laboratory of our institution. A standard RS approach was performed. ¹³ Drilling of the petrous apex was tailored according to the inferior and lateral limits of the RAP, the CN VII/VIII in the IAM, and the semicircular canals in the SF, respectively. ^{4 5} We designated the RAP technique that preserved the IAM cortical bone and SF as “limited RAP.” Conversely, the RAP technique that involved extending the drilling to expose the pertinent contents of IAM and SF was described as “extended EA-RAP.” The limited and extended EA-RAPs were performed sequentially on each side for a total of 10 limited RAPs and 10 extended EA-RAPs. Exposure of the MCF through an extradural subtemporal approach was then performed to obtain a full view of MC and analyze the exposure of both posterior intradural approaches. Retraction of the cerebellar hemisphere was held constant at 1.5 cm to further ensure the validity of the measurements. Following each surgical approach, key measurements of the MC and ICA were gathered from both the RS corridor and the MCF exposure. Measurements were collected using a digital caliper. All dissections and measurements were conducted using an operating microscope (Zeiss S100/OPMI microscope; Carl Zeiss AG, Jena, Germany) and a 0-degree and 45-degree endoscope (Stryker Endoscopy, San Jose, California, United States). The dissection procedures were meticulously documented in a stepwise fashion through high-resolution photographic and endoscopic image-acquisition techniques. ¹²

An illustrative case involving the resection of a trigeminal schwannoma with MC extension was described to highlight the application of an EA-RAP in achieving maximal tumor resection and preventing vascular complications.

Limited RAP

The posterior fossa neurovascular structures were exposed after sharp arachnoid dissection of the CPA. The SMT was identified. A C-shaped incision of the dura mater was then performed along its lateral margin. The dural flap was mobilized, and the bony prominence of the SMT was exposed. A medial to lateral drilling of the SMT was performed using a straight, high-speed drill (Stryker, San Jose, California, United States). The bony removal was limited superiorly by the tentorium and the SPS, inferiorly by the superior wall of the IAM covering the CN VII/VIII complex, laterally by the anterior margin of the SF, covering the posterior and superior semicircular canals, and medially by MC with the porus trigeminus (PT) until the lateral aspect of the PT and MC dura mater was exposed. A horizontal incision of its dura mater was performed along its superior margin in a proximal to distal direction, exposing the entire length of the Meckel's segment of TN and the posterior part of the GG. Limiting the bony removal medial to the SF and preserving the IAM, the ICA was not exposed. Fig. 1 illustrates the limited RAP approach in a stepwise fashion.

Fig. 1.

Step-by-step dissection of the limited retrosigmoid approach with reverse anterior petrosectomy (RAP) in an anatomical specimen (right side). ( a ) An initial overview of the intradural exposure obtained after a standard retrosigmoid craniotomy. Here, cranial nerves (CN) IV–XI, superior and anterior inferior cerebellar arteries (SCA/AICA), and the superior petrosal vein (SPV) are identified within the cerebellopontine angle (CPA). ( b ) A magnified view, centered on CN V, reveals the relationship of the sensory and motor trigeminal divisions with the suprameatal tubercle (SMT), in which the latter obstructs the lateral view of their entrance into Meckel's cave (MC). The subarcuate fossa is identified superolateral to the internal acoustic meatus (IAM). ( c ) A rectangular region of dura overlying the bone of the SMT and IAM is removed to initiate exposure of MC. ( d ) A high-speed straight drill is used to perform the limited internal petrosectomy from the SMT to the superior border of the IAM, taking care to protect the IAM contents inferiorly and the internal carotid artery (ICA) anteromedially. The lateral and superior limits of bony removal are the subarcuate fossa and tentorium, respectively. ( e ) MC is opened via a single linear incision parallel and superior to the margin of CN V entering through the porus trigeminus (PT). ( f ) After gentle upward retraction of the dura, the entire length of the Meckel's segment of CN V and the posterior part of the Gasserian ganglion (GG) come into view. A, artery; AICA, anterior inferior cerebellar artery; CN, cranial nerve; IAM, internal acoustic meatus; JUNC, junction; SCA, superior cerebellar artery; SEG, segement; SMT, suprameatal tubercle; SPV, superior petrosal vein; SUP, superior.

Extended EA-RAP

The high-speed drill was utilized to perform the extended RAP to maximize MCF exposure. The boundaries of the limited RAP were maximized laterally to the edge of the superior and posterior semicircular canals, promptly discriminated from the surrounding bone due to the compact and yellow-colored otic capsule; inferiorly to the superior dura of the IAM; and anteromedially, the exposure reached the lacerum segment of the ICA. The bony removal inferior to the GG allowed access to the lacerum ICA. The petrous segment of the ICA was then exposed aided by a 45-degree endoscope. Enhanced exposure of the anteromedial portion of the petrous apex was achieved through further drilling toward the petroclival fissure and the inferior petrosal sinus, close to the abducens nerve. Fig. 2 depicts the extended EA-RAP.

Fig. 2.

Step-by-step dissection of the extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP) in an anatomical specimen (right side). ( a ) An overview of the exposure of the cerebellopontine angle (CPA) and its components through a right retrosigmoid craniotomy. A prominent suprameatal tubercle (SMT) obstructs the lateral view of the entrance of CN V into Meckel's cave (MC). The subarcuate fossa is identified superior and lateral to CN VIII entering the internal acoustic meatus (IAM). ( b ) A wide removal of dura overlying the temporal petrous bone is completed, and a high-speed straight drill is used to perform the extended internal petrosectomy from the SMT in multiple directions. Superiorly, the posterior petrous ridge is removed until the superior petrosal sinus (SPS) and tentorium are reached. Inferiorly, the superior and lateral borders of the IAM are removed, exposing the meatal segment of CN VIII. Laterally, the subarcuate fossa is drilled until a thin rim of dense, yellow otic capsule bone emerges, indicating the posterior curve of the superior semicircular canal and the superior curve of the posterior semicircular canal. CN V and its entrance into MC through the porus trigeminus (PT) are now visible. ( c ) 0-degree endoscopic images provide additional, magnified views of the anteromedial limit of bony removal, the point of contact between the short portion of the lacerum segment of the internal carotid artery (ICA) with the inferior and medial borders of MC; and ( d ) a detailed view of the superior limit of exposure, the tentorium, including the intradural entrance of CN IV into the posterior cavernous sinus. ( e ) 45-degree endoscopic images captured in a lateral orientation highlight the greater extent of exposure of the horizontal portion of the petrous ICA following optimal IAM unroofing and subarcuate fossa removal. MC is opened, and its dura is reflected laterally. CN VI entering Dorello's canal is identified inferiorly and laterally to this point. ( f ) Once MC dura is removed in its entirely and the course of the inferior petrosal sinus (IPS) in the inferior petrosal sulcus is clearly identified, a final view of the Gasserian ganglion (GG) overlying the entire length of the exposed petrous ICA is obtained. A, artery; AICA, anterior inferior cerebellar artery; CN, cranial nerve; ICA, internal carotid artery; IPS, inferior petrosal sinus; SCA, superior cerebellar artery; SEG, segement, SMT, suprameatal tubercle; SPS, superior petrosal sinus; SPV, superior petrosal vein; SUP, superior.

Meckel's Cave Measurements

For morphometric study purposes, the center of the medial border of the RS craniotomy was termed point X ( Fig. 3 ). Following the limited and extended EA-RAP, a straight Rhoton microdissector was placed on point X to estimate the most medial, anterior, and lateral accessible anatomical points within MC. The working distances between point X and these identified points were collected. Distances from these MC's anatomical points to reliable points of interest on the MCF were measured ( Fig. 4 , Table 1 ).

Fig. 3.

Craniotomy reference point and trajectories of dissection used in the limited and extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP). Right lateral view of an anatomical specimen in which both a right retrosigmoid and middle fossa craniotomies have been performed to visualize the operative route to reach the Meckel's cave (MC). A straight Rhoton microdissector is placed at the center of the medial border of the retrosigmoid craniotomy (X) as a reference point for the center of the working corridor for both the limited and extended RAP.ICA, internal carotid artery.

Fig. 4.

Comparative analysis of surgical limits of access to Meckel's cave (MC) between the limited and extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP) in an anatomical specimen (right side). The right middle cranial fossa was exposed via right temporal craniotomy with an extradural subtemporal approach to visualize, analyze, and compare MC's accessed zones from both a limited and extended RAP. Exposure of the middle cranial fossa (MCF) and drilled petrous apex area after ( a ) limited RAP and ( b ) extended RAP. ( c ) Removal of the dura comprising the lateral wall of the posterior cavernous sinus and MC to visualize the spatial relationship between the cavernous internal carotid artery (ICA) and each of the three postganglionic trigeminal divisions to the floor of the MCF. ( d ) Surgical limits of access to MC were identified after the limited RAP. Using a straight Rhoton microdissector positioned at the center of the medial border of the retrosigmoid craniotomy (see point X, Fig. 3 ), three black pins are placed to signify the most medial (a), anterior (b), and lateral (c) accessible anatomical points within the MC. ( e ) Surgical limits of access to the MC were identified after the extended RAP using the same method of analysis with red pins (A, B, C). ( f ) Merged representation of accessible MC's limits from both the approaches (green a/A and b/B for limited and extended RAP). The most medial (a/A) and anterior (b/B) reached limits are equivalent between the limited and extended RAP, whereas the most lateral reached limits (c, C) differ between both approaches. Measurements depicted in Table 1 are shown here: point a/A to the transition point between the posterior vertical segment and the posterior bending of the cavernous ICA; point b/B to the superior orbital fissure (SOF); point b/B to the foramen rotundum (FR); point c and C to the foramen ovale (FO). ( g–i ) Schematic representation of MC's accessed area with the limited ( g ) and extended RAP ( h ) in an anatomical position. Colors corresponding with the approach type (black, limited RAP; red, extended RAP) were used to shade the reached areas. ( i ) The final schematic merged view is shown. AE, arcuate eminence; CAV, cavernous; CN, cranial nerve; CS, cavernous sinus; FO, foramen ovale; FR, foramen rotundum; GSPN, greater superficial petrosal nerve; ICA, internal carotid artery; IPS, inferior petrosal sinus; MMA, middle meningeal artery; SCA, superior cerebellar artery; SOF, superior orbital fissure; SPS, superior petrous sinus.

Table 1. MCF measurements related to the limited and extended RAP.

A. Working distances from point X to the accessible limits of the Meckel's cave
Limited RAP		Extended EA-RAP
Distances	Mean ± SD (mm)	Distances	Mean ± SD (mm)	P Value*
X to a	82.2 ± 0.549	X to A	82.2 ± 0.573	1.000
X to b	84.8 ± 0.461	X to B	84.9 ± 0.443	0.832
X to c	85.7 ± 0.386	X to C	86.2 ± 0.391	0.015

B. Distance between the key landmarks of Meckel's cave and middle cranial fossa
Limited RAP		Extended EA-RAP
Distances	Mean ± SD (mm)	Distances	Mean ± SD (mm)	P Value*
a to post. C4	4.3 ± 0. 234	A to post. C4	4.1 ± 0.270	0.626
b to SOF	25.5 ± 0.157	B to SOF	25.1 ± 0.189	0.283
b to FR	20.3 ± 0.313	B to FR	19.8 ± 0.354	0.146
c to FO	12.2 ± 0.319	C to FO	7.9 ± 0.169	<0.001

Abbreviations: C4, cavernous segment of the internal carotid artery; EA-RAP, endoscopic-assisted reverse anterior petrosectomy; FO, foramen ovale; FR, foramen rotundum; MCF, middle cranial fossa; post, posterior; RAP, reverse anterior petrosectomy; SD, standard deviation; SOF, superior orbital fissure.

Note: * Statistical significance was considered when P value was <0.05.

Internal Carotid Artery Measurements

ICA exposure was obtained with the extended EA-RAP using a 0- and 45-degree endoscope through the RS corridor. The maximum lengths of the exposed ICA were measured and compared between the 0- and 45-degree endoscopic visualizations. The working distances from the middle point of the internal acoustic meatus (MIAM) to the medial (d) and lateral (e, f) points of exposed ICAs were also collected ( Fig. 5 , Table 2 ).

Fig. 5.

Internal carotid artery (ICA) analysis (right side). ( a ) Macroscopic transcranial view from the extradural subtemporal route of the middle cranial fossa (MCF) following the extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP). Meckel's cave (MC) dura was opened, and the rootlets of the Gasserian ganglion (GG) were dissected according to each of the three postganglionic trigeminal divisions. The medial aspect of MC overlies the ICA at its origin as the cavernous segment. ( b ) The GG and postganglionic trigeminal roots covering the ICA in the MCF, from the lacerum to the cavernous segment, have been removed to fully visualize the course of the ICA in the temporal petrous bone. The course of CN VI into the cavernous sinus and its relationship with the cavernous ICA can also be identified after the removal of Gruber's ligament and the unroofing of Dorello's canal. ( c ) 0-degree endoscopic images showed the exposed course of the ICA. A straight Rhoton microdissector is used to mark and sequentially measure the length of the artery from its medial (d) to lateral points (e). ( d ) A magnified transcranial view of the exposed ICA with the measured distances of interest following the limited and extended RAP. Distances between the key points of the ICA and the middle point of the internal acoustic meatus (MIAM) are illustrated. Please note that d represents the most medial exposure of ICA using both 0- and 45-degree endoscopes, while e illustrates the most lateral exposure acquired when using a 0-degree endoscope and f the most lateral exposure using the 45-degree endoscope. ( e ) 45-degree endoscopic images captured from a lateral orientation highlight further petrous bone removal inferior and lateral to the GG. A 45-degree angle Rhoton microdissector is used to denote the lateral limit of exposure after the extensive drilling (f). A., artery; AE, arcuate eminence; CAV, cavernous; CS, cavernous sinus; FO, foramen ovale; FR, foramen rotundum; GSPN, greater superficial petrosal nerve; IPS, inferior petrous sinus; MMA, middle meningeal artery; PLL; SCA, superior cerebellar artery; SOF, superior orbital fissure; SPS, superior petrous sinus; TN, trigeminal nerve.

Table 2. ICA measurements related to the extended EA-RAP.

C. Endoscopic-assisted ICA exposure in the extended EA-RAP
0-degree		45-degree
Length	Mean ± SD (mm)	Length	Mean ± SD (mm)	P Value*
		e to f	9.3 ± 0.218	<0.001
d to e	0.27 ± 0.072	d to f	12.3 ± 0.338	<0.001

D. Depth of the exposed ICA in the extended EA-RAP
0-degree		45-degree		P Value*
Distance	Mean ± SD (mm)	Distance	Mean ± SD (mm)
MIAM to d	24.0 ± 0.190	MIAM to e	21.0 ± 0.186	<0.001
		MIAM to f	13.2 ± 0.137	<0.001

Abbreviations: ICA, internal carotid artery; MIAM, middle point of the internal acoustic meatus; RAP, reverse anterior petrosectomy; SD, standard deviation.

Note: * Statistical significance was considered when P value was <0.05.

Statistical Analysis

A descriptive analysis was performed for all measurements of interest utilizing SPSS 25 (IBM, Armonk, New York). The mean values are expressed in ± SD. Measurements from the two variations were compared with a paired t -test. A p  < 0.05 was considered statistically significant.

Results

Meckel's Cave Exposure

Customizing the drilling of the petrous bone allowed access to different areas of the MC. By drilling the petrous apex from the SMT's base to the tentorium, direct access to GG in the medial and anterior parts of the MC was achieved. The superior petrosal vein was divided to increase maneuverability. Access to the inferior and lateral portions of the MC was achieved by drilling the petrous apex from the base of SMT to the IAM.

In the limited RAP, the superior osseous border of the IAM was preserved, while in the extended version, it was unroofed. This maneuver increased the inferolateral exposure of the GG. Furthermore, extended drilling of the SF, with the exposure of the limits of the posterior and superior semicircular canals, facilitated access to the inferolateral portions of the MC. Further bony removal of the petrous apex toward the abducens nerve and inferior petrosal sinus did not provide increased visualization of the MC but provided additional access to the inferior and medial parts of the trigeminal root.

Statistical analysis of the MCF anatomical measurements is reported in Table 1 . The mean working distances from the craniotomy to the most medial, anterior, and lateral MC points reached were similar in the limited and extended EA-RAP (significant differences between distances from the craniotomy to the most lateral MC points were found, p  = 0.015).For both the approaches, the most medially reached points within the MC (a, A) were located at a mean distance of approximately 4 mm posterior to the posterior bend of the cavernous ICA. The most anterior MC points reached (b, B) were posterior to the superior orbital fissure (SOF) with a mean distance of approximately 25 mm and posteromedial to the FR at a mean distance of 2 mm. The differences were not statistically significant between both the approach extensions.

Endoscopic Assistance in the Meckel's Cave Exposure

Once the unroofing of the IAM and the SF drilling were performed under high magnification in the extended EA-RAP, the posterior and superior semicircular canals could be seen as the most lateral limit of the exposure of the surgical corridor. Although the visualization of the intraoperative field can be improved by adjusting the microscope angle, the most inferior and lateral parts of the MC and the V3 still remain hidden. The introduction of a 45-degree endoscope played a key role. With regard to the mentioned lateral limits, it gained visualization and further increased the maneuverability of the drill and other straight surgical instruments in the exposure of the inferolateral MCF toward the FO.

The inferolateral exposure of the MC varied between the two approaches. The maximal lateral MC point was posterior to the foramen ovale (FO) at a mean of 12 mm for the limited RAP (c), and at a mean of 8 mm for the extended EA-RAP (C) (statistically significant differences between the two variations were found, p  < 0.001).

ICA Exposure

The horizontal segment of the petrous ICA (C2), after an anterior and medial curve in front of the cochlea, runs within the carotid canal toward the foramen lacerum. As it exits the carotid canal, the lacerum segment of the ICA (C3) is laterally covered by the petrolingual ligament. The ICA then moves toward the cavernous sinus and turns 90 degrees superiorly (the first genu or posterior bend) and 90 degrees anteriorly (the second genu or anterior bend) to the carotid groove of the sphenoid bone where it enters the medial aspect of the sinus (C4, cavernous segment). From the RS perspective, C2 (horizontal segment) lies inferior and posterolateral to the MC and V3; the C3 segment lies inferior and medial to the MC (▶ Fig. 6 ).

Fig. 6.

Detailed anatomical relationships between the internal carotid artery (ICA) and Gasserian ganglion (GG) as seen from a right extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP). Microsurgical anatomy of the middle cranial fossa (MCF) and its contents are exposed in detail from an axial ( a ) and sagittal perspective ( b ). ( c–e ) The ICA establishes a crucial relationship with the trigeminal nerve during its course in the MCF. In a retrosigmoid perspective, the cavernous segment of the ICA (C4) is located anterior and medial to the GG. The lacerum (C3) and horizontal segments of the petrous (C2) ICA lie medial and inferior to the GG and the mandibular division (V3). The entire course of the trigeminal nerve is situated superior and anterior to the ICA, according to a sagittal ( c ) and axial plane ( e ), respectively. The black dotted lines indicate the plane of the GG (1) and ICA (2). ( d ) Final image in which the anatomical specimen is positioned in a simulated supine position to demonstrate an operatively oriented view of the anatomical relationships of interest. Black dashed lines outline the representative area of the GG that can be reached after the extended EA-RAP. A, anterior; Ac., acoustic; Cav., cavernous; Clin., clinoidal; CN, cranial nerves; Endolymph, endolymphatic; Eust, Eustachian; FO, foramen ovale; FR, foramen rotundum; Gang., ganglion; Gen., geniculate; GSPN, greater superficial petrosal nerve; ICA, internal carotid artery; I, inferior; L, lateral; Lat, lateral; M, medial; MMA, middle meningeal artery; M, muscle; Petr., petrous; P, posterior; Post, posterior; S, superior; SOF, superior orbital fissure; Sup., superior; Supraclin, supraclinoid; Tens., tensor; Tymp., tympanic.

Endoscopic Assistance in the ICA Exposure

The ICA was not exposed during the limited RAP, in which the superior aspect of the bony IAM and the bone of the SF were intentionally preserved. The extended EA-RAP consistently exposed the C2 and C3 segments through endoscopic-assisted drilling of the petrous bone from the inferomedial portion of the GG toward the unroofed IAM.

Utilizing a 0-degree endoscope, the C3 segment of the ICA was exposed for a mean length of 3 mm from its most medial to most lateral points (d to e). Consequently, additional drilling of the petrous apex was performed aided by a 45-degree endoscope, and the C2 segment of the ICA was further exposed for a mean length of 9 mm (e to f) (significant, p  < 0.001).

The mean length of the ICA exposed by combining the use of the 0-degree and the 45-degree endoscope was a total of 12 mm (d to e + e to f). Therefore, the extended EA-RAP with tailored angled endoscopes significantly increased ICA exposure in this region.

The working distances from the MIAM to the most medial (d) and lateral (e, referring to drilling using the 0-degree endoscope; f, referring to drilling using a 45-degree endoscope) points of exposed ICA were 24 mm (MIAM to d), 21 mm (MIAM to e), and 13 mm (MIAM to f), respectively (significant, p  < 0.001).

Statistical analysis of the ICA anatomical measurements is reported in Table 2 .

Illustrative Case

A 27-year-old male presented to our institution with a progressively worsening imbalance and retroauricular headache. Imaging revealed an enlarging enhancing mass along the right trigeminal nerve extending from the CPA into the MC, suspicious for a trigeminal schwannoma.

The predominant bulk of the tumor was evident in the posterior fossa, yet its extension into the MCF beyond the PT prompted consideration for the extended EA-RAP. Specifically, the mass extended superiorly to the petrous segment of the ICA medial to its first genu and posteriorly to the posterior genu of the cavernous ICA. A subtemporal approach was not favored, given the asymmetric growth of the tumor within the CPA and its inferior extension below the IAM.

The patient underwent a successful RS craniotomy with an EA-RAP with gross total resection of the tumor. Histopathology confirmed schwannoma. Postoperatively, the patient experienced V2–V3 numbness and mild weakness of the right muscles of mastication, both of which significantly improved at 4 weeks ( Fig. 7 ). To date, no tumor recurrence has been noticed.

Fig. 7.

Case illustration demonstrating the limits of the extended endoscopic-assisted reverse anterior petrosectomy (EA-RAP) according to the internal carotid artery (ICA) segments. (a, b) Preoperative contrast-enhanced T1-weighted axial MRI demonstrating an enhancing tumor in the right cerebellopontine angle with extension into the Meckel's cave. The green dotted lines show a coronal plane passing through the transition point between the posterior vertical segment and the posterior bending of the cavernous ICA (1) and a tangential plane to the medial border of the semicircular canals (2). ( c ) Similar anatomical view within a cadaveric specimen (black dotted lines). ( d ) Postoperative magnetic resonance image showing gross total resection of the tumor. ICA, internal carotid artery; SSC, semicircular canals; SS, sigmoid sinus.

Discussion

First introduced by Samii et al in 1983 and further described by Seoane and Rhoton in 1999, the RAP offers an extended RS approach to access tumors with bicompartmental supratentorial/infratentorial extension, particularly into the PT, central MC, sellar and parasellar region, and anteromedial MCF. ^{1 2 4 6} Since then, and with the advantage of endoscopic assistance, numerous studies have focused on describing the microsurgical anatomy of the RAP and comparing its exposure with more traditional MCF approaches. ^{3 4 5 6 8 9 10 11 14 15 16} Previous studies aiming to extend this approach have primarily focused on maneuvers reaching the central skull base and cavernous sinus. ¹⁰ Only a few studies have focused on providing surgical landmarks based on measurements of intrapetrous structures, particularly the ICA, to increase intraoperative safety. ^{5 7 14 17}

To the best of our knowledge, this is the first anatomical study aimed to meticulously describe the range of MC's access through EA-RAP, and provide operatively oriented anatomical guidance to avoid ICA injuries during this extended approach.

Tumor types of particular relevance for accessing the MC with this approach include some petroclival meningiomas, MC meningiomas, trigeminal schwannomas, epidermoid cysts, and chondrosarcomas among others. ^{1 8 9 10 18} Trigeminal schwannomas may originate at the dorsal nerve root entry zone, the GG, or its branches and predominantly occupy the MC before following the path of less resistance to extend anteriorly along postganglionic segments in the MCF or posteriorly toward the brainstem. ¹⁸ Our study consistently established that the medial and anterior areas of the MC can be readily accessed via petrosectomy from the SMT base to the petrous ridge, with exposure to the proximal cavernous ICA. With the extended petrosectomy presented, the middle fossa's inferior and lateral areas were further effectively reached.

On the other hand, by increasing the anterior and inferolateral exposure toward the FO and FR, extending the RAP may optimize the resection of nerve sheath tumors by improving mobilization of the trigeminal nerve as well as uncovering a greater surface area of its infiltrative route along the postganglionic divisions. Maximal bony removal in an extended approach facilitates improved access to the anterolateral and central MC and postganglionic divisions of the trigeminal nerve toward their cranial foramina, thus serving as a potential strategy for accessing lesions with distribution in both the MCF and posterior cranial fossa (PCF). Despite advancements in intraoperative technology and microsurgical technique, surgical series report higher rates of surgical morbidities, lower rates of gross total resection, and higher remnants for tumors in this region. ³ Thus, a reappraisal of existing surgical corridors to safely optimize exposure and identify all surrounding neurovascular structures is warranted, and the extension of RAP in an anterior and inferolateral trajectory, as described in this study, offers one such strategy for select cases.

Based on our results, the aid of a 45-degree endoscope allowed us to obtain a significantly increased ICA exposure (3 mm vs. 9 mm). The extended RAP and the use of angled endoscopy are particularly useful for lesions that have significant involvement in this region, as the improved visualization assists in reaching further limits of the MCF and identifying residual tumors in this challenging anatomic corridor. ^{3 6} The consistency of epidermoid cysts, for example, causes the tumor to interdigitate between cranial nerves and potentially leave remnants hidden out of direct view of the microscope. Therefore, angled endoscopy and maximal bony removal become necessary to obtain gross total resection of epidermoids with significant MCF extension, ¹⁸ as well as for the removal of unrecognized tumor remnants of dumbbell-shaped trigeminal schwannomas unnoticed in the anterolateral and superolateral aspects of the MC, as previously reported by Samii et al. ³ Further, an extension of SMT drilling leads to minimized retraction of neurovascular structures, ⁶ which can also be appreciated in nononcologic pathology like trigeminal neuralgia, where the prominence of the SMT may obstruct the surgeon's ability to perform microvascular decompression of the trigeminal nerve effectively. ^{19 20}

Advantages of utilizing an extended RAP relative to approaches centered on the MCF, such as a subtemporal approach or extradural anterior petrosectomy, include better mobilization of the trigeminal nerve within the MC, early visualization of the cranial nerves and brainstem, lack of temporal lobe retraction, and increasing visibility of the trigeminal nerve. ^{3 11} However, customizing this internal petrosectomy ultimately depends on the tumor's spatial proportion, displacement, and invasion of local neurovascular structures, SMT size, and histology of the tumor, in addition to the patient's age, general fitness, preoperative cranial nerve deficits, and surgeon's experience. ^{5 6 9 10 16} Moreover, the anatomical course and proximity of the ICA impose specific boundaries that necessitate careful consideration and adherence.

As such, one of the main objectives of this study was to contextualize the course of the ICA in the middle fossa as a pertinent surgical landmark for this extended bicompartmental approach. In our opinion, the ICA lends itself as an optimal landmark, not only to aid in vascular injury avoidance, but also because its intrapetrous course reveals the exact tailoring of bony removal that can be achieved in each patient, which, in turn, predicts the ability to perform gross total resection of the tumor. Finally, the ICA is readily and easily identified in preoperative neuroimaging and intraoperative neuronavigation. For instance, we found that the maximal anteromedial exposure of the MCF was no less than 4 mm posterior to the posterior genu of the cavernous ICA for either the limited or extended RAP, thereby indicating that this approach would not be sufficient for tumors extending anteriorly beyond this margin. Therefore, a coronal plane passing through the posterior genu of the cavernous ICA could be used as a reliable method to estimate the most anterior reach of the RAP in the MCF. Similarly, based on the anatomical dissections herein, a tangential plane medial to the common crus of the semicircular canals could be used to estimate the most lateral reach of the extended RAP in the MCF. Therefore, we do believe that lesions located within these boundaries of the MCF can be accessed from the posterior route offered by the extended EA-RAP ( Fig. 7 ).

The second goal of this study was to improve understanding of the spatial relationship between the MC and the ICA through this approach and to provide anatomical guidance for avoiding intraoperative ICA injury during an extended EA-RAP. The transitions from the petrous segment (C2) to the lacerum segment (C3) of the ICA occur inferior to the MC, in a lateral-to-medial direction, from posterior to anterior. The MC is separated from C3 and the medial part of the horizontal section of C2 by the petrolingual ligament and the periosteum, which cover the top of the carotid canal. ^{21 22}

Despite the importance of understanding these anatomical relationships, the literature offers limited descriptions of how to safely extend the drilling inferior to the MC, under which the artery courses. Based on the available descriptions in the literature, most studies report average lengths of the ICA segments or their distances from the surrounding temporal bone structures, yet without correlating them with specific points of MC access, which we believe is of crucial functional importance when planning and performing tumor resection in this region. ^{5 7 10 17} Koerbel et al indicated that the bony removal of petrous apex in an anterior and superomedial direction inherently protects the MCF's neurovascular structures located in a lateral and inferior position within the temporal bone, like the petrous ICA, greater superficial petrosal nerve (GSPN), cochlea, and IAM contents. ¹⁰ As reported by Colasanti et al, the deep position of the petrous ICA within the petrous pyramid and the morphological variation of this region render anticipation of access to the petrous ICA challenging, even more so when distorted by tumors. ¹⁷ They provided indication on how to start the drilling immediately inferior to the IAM and to continue forward in the direction of the petrous apex, in which the area is devoid of important structures, yet guidance on the method to optimize inferolateral MC access while protecting the ICA was not discussed.

Our study addresses this gap with high-quality anatomical dissection with a further emphasis on the relationship of the ICA with the trigeminal nerve along its transpetrous route to its exiting foramina, points which more comprehensively represent the major areas of interest for tumor dissection in this region. Our findings demonstrate that the ICA is protected when performing a limited RAP by limiting drilling of the petrous apex inferior and medial to the MC and superior to the IAM without unroofing it. We found three particular moments where the ICA is most vulnerable to injury during the extended RAP: (1) unroofing of the IAM, after which the transition between the lacerum and the petrous segments of the ICA was found at average depths of 21 mm from the MIAM and (2) when angled drilling of the petrous apex is used to extend access of the MCF inferiorly and laterally where the horizontal segment of the ICA can be found at an average depth of 13 mm from the MIAM.

Therefore, to optimize the safety of the extended RAP, the authors recommend starting the drilling of the petrous apex from the superior border of the IAM and proceeding anteriorly carefully, leaving a protective shell of bone over the anticipated location of the lacerum and petrous segments of the ICA.

Finally, the risk of encountering the horizontal segment of the petrous ICA is minimized by gradually drilling the petrous bone utilizing angulated endoscopic assistance and intraoperative neuronavigation.

Limitations

The primary limitation of this study is related to surgical cadaveric analysis, which may introduce potential distortions in the neurovascular structures. Further, the embalmed brains' response to positional changes may not mirror that of a clinical setting. We implemented a standardized approach to address this limitation by applying a consistent amount of fixed retraction on the cerebellum. This simulated the gravitational retraction encountered in a clinical setting, thus aiding in mitigating the limitation above.

Conclusion

RS with EA-RAP provides access to the TN within the MC as well as part of its postganglionic divisions toward their respective foramina. The aid of the endoscope with different view angles could significantly improve the possibility of performing a tailored and careful bony removal, while ensuring the protection of the ICA. A coronal plane from the posterior genu of the cavernous ICA and a tangential plane to the common crus of the semicircular canals could define and help predict the area of the MCF accessed by an extended EA-RAP.

Funding Statement

Funding This study was funded by the Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota, the Joseph and Barbara Ashkins Endowed Professorship in Surgery and the Radiology Department, Mayo Clinic, Rochester, Minnesota, and the Charles B. and Ann L. Johnson Endowed Professorship in Neurosurgery, Mayo Clinic, Rochester, Minnesota.