Endoscopic Multiportal Approaches to Meckel's Cave: A Cadaveric Study and a Three-Dimensional Anatomical Video

Abstract

Objective  This cadaveric study aims to illustrate the lateral transorbital (LTO), transantral transpterygoid (TATP), and endoscopic endonasal approaches (EEA) to Meckel's cave (MC), defining their surgical freedom, area of exposure, and advantages and limitations; thus, refining their respective indications.

Design  Cadaveric study.

Setting  The Anatomical Laboratory for Visuospatial Innovations in Otolaryngology and Neurosurgery (ALT-VISION) at the Ohio State University, Wexner Medical Center.

Participants  Bilateral dissections of five injected cadavers (10 sides).

Main Outcome Measures  Distance to targets, surgical freedom, and area of exposure provided by the EEA, TATP, and LTO approaches.

Results  The TATP approach provides superior surgical freedom to foramen rotundum (167.70 ± 29.762 mm). However, surgical freedom to foramen ovale was best when using the LTO approach (75.01 ± 15.773 mm). The EEA provides a superior exposure of the medial MC (mean area of 587.69 ± 38.59 mm ² ). The LTO and TATP approaches provide equivalent access to the lateral MC (ranging from 468.90 ± 26.98 mm ² for TATP to 650.11 ± 35.76 mm ² for the LTO approach). Combining approaches increases the area of exposure significantly (1,237.79 ± 48.41 mm ² and 1,056.59 ± 48.12 mm ² for EEA and LTO vs. EEA and TATP).

Conclusion  This study thoroughly emphasizes the pros and cons of the aforementioned approaches. Each approach may be applied in selected cases as a single or as part of a combined technique. However, conventional approaches are still indicated according to extent and pathology.

Level of Evidence : V.

Introduction

Skull base lesions involving Meckel's cave (MC) and the middle cranial fossa (MCF) are surgically challenging. ^{1 2} In general, neoplasms in MC are rare with schwannomas and meningiomas being the commonest. ³ Anatomical relationships of the lesion with surrounding neurovascular structures and surgeon's experience frequently influence the choice for surgical access. ⁴ Minimally invasive approaches may be applied in selected cases with dual or multiport corridors.

This cadaveric study aims to illustrate the lateral transorbital (LTO), the transantral transpterygoid (TATP), and endoscopic endonasal (EEA) approaches to MC, elucidating surgical freedom, area of exposure, and limitations, as well as refining their associated indications.

Method

The study comprised five cadaveric specimens previously prepared with colored silicone vascular injections. Computed tomography scans (0.5-mm thick slices) of the specimens were uploaded to a navigation system (iNtellect, Stryker Inc., Kalamazoo, Michigan, United States) considering a registration error of <1 mm as acceptable. The cadaveric specimens were fixed with a 3-pin head holder to secure proper positioning and bilateral dissections (10 sides) were completed using standard neurosurgical instruments and 4 mm rod-lens endoscopes coupled to a high-definition camera and monitor (Karl Storz, Tuttlingen, Germany). The techniques for the different approaches, as previously described, ^{5 6 7} were completed in the following sequence: first the LTO approach, followed by the TATP, and finally the EEA. This yielded incremental surgical exposure and surgical freedom that was accurately measured for each approach. Illustrative examples of patients with tumors involving MC are also provided.

Statistical Analysis

Comparisons of the area of exposure and surgical freedom obtained for each approach were conducted using a one-way repeated-measure ANOVA with post-hoc analysis. For all comparisons, p -values <0.05 was considered statistically significant.

Results

Video 1

Download video file ^{(12.7MB, mp4)}

As can be seen in Figs. 1 and 2 , image guidance confirmed the feasibility of the approach, guided our drilling in critical regions, as well as helped calculating distances from surgical entry to different targets while approaching MC ( Table 1 ). The TATP approach provided a straightforward exposure to foramen rotundum (FR) and a superior surgical freedom (167.70 ± 29.762 mm). However, surgical freedom to foramen ovale (FO) was best when using the LTO approach (75.01 ± 15.773 mm).

Fig. 1.

Image guidance during greater wing of sphenoid (GWS) drilling in the lateral transorbital (LTO) approach (notice the demarcation between periorbita and middle cranial fossa [MCF] dura if peeling is done along the right dissection plane).

Fig. 2.

Image-guided craniotomy of pterygoid body in transantral transpterygoid (TATP) approach (notice V2 representing the medial limit of the drilling). GWS, greater wing of sphenoid.

Table 1. Distance to target.

		Mean and SD (mm)	Median	Range	IQR
EEA	FL	75.25 ± 4.62	75.90	69.50–81.60	70.05–79.52
	OS	72.07 ± 4.67	73.35	65.40–78.30	67.18–76.22
	VC	60.45 ± 6.88	59.25	50.10–71.40	55.23–66.53
	FR	63.20 ± 6.47	62.05	52.40–73.00	58.93–68.58
	FO	75.87 ± 5.20	76.95	69.60–82.10	70.30–80.97
	GG	78.77 ± 4.16	79.00	73.20–84.60	74.15–82.83
TATP	SPF	39.89 ± 2.76	40.35	35.00–44.10	37.83–42.05
	VC	45.79 ± 2.68	46.45	41.80–49.30	43.30–48.13
	FR	44.93 ± 2.23	45.10	41.80–48.10	42.60–46.75
	FO	57.88 ± 1.74	58.10	54.30–60.60	56.75–58.75
	GG	61.61 ± 1.65	61.50	59.20–64.30	60.38–63.28
LTO	ONC	43.56 ± 3.00	42.20	40.60–49.10	41.40–45.95
	SOF	40.42 ± 3.04	39.65	36.50–47.20	38.50–42.23
	FR	48.42 ± 2.24	48.45	45.30–52.40	46.58–50.13
	FO	60.07 ± 2.77	58.75	57.10–65.50	57.80–62.43
	GG	64.04 ± 2.66	63.65	61.40–70.00	61.80–65.08
	Drilled GWS-SA	309.72 ± 43.09	304.04	255.75–386.43	269.87–352.93

Abbreviations: EEA, expanded endonasal approach; FL, Foramen lacerum; FO, foramen ovale; FR, foramen rotundum; GG, Gasserian ganglion; GWS-SA, drilled greater wing of sphenoid surface area; IQR, interquartile range; LTO, lateral transorbital approach; ONC, optic nerve canal; OS, optic strut; SD, standard deviation; SOF, superior orbital fissure; SPF, sphenopalatine foramen; TATP, transantral transpterygoid approach; VC, Vidian canal.

Note: Data are expressed as mean and standard deviation, median, range, and interquartile range.

The shape of the quadrangular space was noticed as: trapezoid ( n  = 2), square ( n  = 4), or hourglass ( n  = 4). This finding had surgical implications as the working window in the trapezoid configuration is tighter making the approach more difficult.

A representative three-dimensional volumetric model connecting points of interest into lines and triangles demonstrating MC areas of exposure obtained with each route was created ( Figs. 3 4 5 ). Five macro-regions of areas of exposure were created through the sum of adjacent triangles: (1) medial exposure (equal to the sum of rhomboid 1 with triangles 2, 3, and 4); (2) posteromedial (triangles 2 and 4); (3) superolateral (triangles 5, 6, 7, and 8); (4) inferolateral (triangles 6, 7, and 8); (5) posterolateral equals triangles 7 and 8 ( Table 2 ).

Fig. 3.

Rhomboid 1 relates to the area extending from orbital apex (optic strut [OS] to foramen rotundum [FR]) to paraclival internal carotid artery (ICA) (foramen lacerum [FL] to posterior genu). Triangle 3 is outlined by FR, FL, and foramen ovale (FO).

Fig. 4.

Triangle 2 corresponds to the area extending backward from paraclival ICA to the cochlea (navigation-guided), while triangle 4 represents a backward extension of triangle 3 till the cochlea (two asterisks). Both triangles are separated by a line running along petrous ICA (one asterisk), so triangle 2 is comparable to premeatal Kawase triangle, while triangle 4 corresponds to Glasscock triangle from an anterior perspective.

Fig. 5.

Endoscopic view that represents the lateral surface of Gasserian ganglion (GG), triangle 5 correlates to the orbital apex from SOF side (OS to FR) (comparable to anteromedial triangle). Triangle 6 (a.k.a. anterolateral triangle) demarcated by FR, FO, and the intersection of both V2 and V3 at GG. Triangle 7 with boundaries between FO, foramen spinosum (FS), GG confluence. Finally, triangle 8 extends backward to the cochlea. A representative 3D volumetric Video 1 is provided (see attached link: https://drive.google.com/file/d/1evlKXA-d2vDRlWLegA0T8aR79iTTl-cp/view?fbclid=IwAR2aaw05XNsEvZm4d7v89MaawolJNuAaronaXb_ACfflD0jxRq07xK68vCc ). *ICA canal, **cochlea. FO, foramen ovale; FR, foramen rotundum; SOF, superior orbital fissure.

Table 2. Surface area.

		Mean ± SD (mm 2 )	Median	Range	IQR
EEA	Medial (1.2.3.4)	587.69 ± 38.59	597.41	508.73–636.21	563.30–613.92
	Postero-medial (2.4)	258.00 ± 35.87	258.22	213.86–319.74	219.48–290.42
TATP	Infero-lateral (6.7.8)	468.90 ± 26.98	472.30	422.12–518.81	449.49–483.53
	Posterolateral (7.8)	265.90 ± 20.19	265.96	219.08–292.24	223.84–281.51
LTO	Superior-lateral (5.6.7.8)	650.11 ± 35.76	657.22	583.16–697.73	621.25–675.74
EEA and TATP	Medial (1.2.3.4) and infero-lateral (6.7.8)	1,056.59 ± 48.12	1,051.75	988.97–1,155.02	1,031.97–1083.64
EEA and LTO	Medial (1.2.3.4) and superior-lateral (5.6.7.8)	1,237.79 ± 48.41	1,226.26	1,172.33–1331.58	1,202.56–1273.81
Combined-posterior	Postero-medial (2.4) and posterolateral (7.8)	523.90 ± 51.26	530.48	477.43–604.90	485.92–569.76

Abbreviations: IQR, interquartile range; SD, standard deviation.

Note: Data are expressed as mean and standard deviation, median, range, and interquartile range.

The EEA provided access to the medial portions of MC with a mean area of exposure of 587.69 ± 38.59 mm ² . Endoscopic LTO and TATP approaches provided equivalent access to the lateral MC; however, there was a gradual increase in the working area from TATP to LTO as the latter was superior visualizing triangle 5. The average area of exposure ranged from 468.90 ± 26.98 mm ² for TATP, to 650.11 ± 35.76 mm ² in the LTO approach. Table 3 summarizes the pros and cons of each single route.

Table 3. Summary of pros and cons of each approach.

	Pros	Cons
EEA	• No skin incision. • No nerve or vessel crossing. • No brain retraction. • Superior surgical freedom at FR. • Control of PPF, ITF • Control of anteroinferomedial MC. • Better exposure of PA while visualizing both paraclival and petrous ICA.	• Limited exposure in case of lateral/posterior involvement or major vessel encasement. • Need for reconstruction. • Nasal mucosal damage/crustation. • Injury of the VN and possible dry eye.
TATP	• Hidden sublabial incision • Control of inferior, anterior, and lateral MC. • Control of PPF, ITF.	• Inferior control of nasal area. • Inferior control of ICA compared with EEA. • Inferior visualization of V1 compared with LTO. • Need for reconstruction. • Need for angled instrumentations to facilitate maneuverability in MCF floor. • Exposure of PA necessitate marked drilling of MCF floor/crossing petrous ICA/angled instrumentation. • Reduction in mastication, trismus, or oroantral fistula are possible complications
LTO	• Small incision. • Spares the Vidian nerve. • Control of lateral MC from superior aspect. • No need for reconstruction. • Superior surgical freedom at FO	• No control of nasal area, PPF, ITF, anteroinferomedial MC. • Inferior control of ICA compared with EEA. • Exposure of PA necessitate marked retraction of both orbit and temporal lobe. • Inferior cosmosis.

Abbreviations: EEA, expanded endonasal approaches; FO, foramen ovale; FR, foramen rotundum; ICA, internal carotid artery; ITF, infratemporal fossa; LTO, lateral transorbital; MC, Meckel's cave; MCF, middle cranial fossa; PA, petrous apex; PPF, pterygopalatine fossa; TATP, transantral transpterygoid; VN, Vidian nerve.

Combining the approaches increased the area of exposure significantly (1,237.79 ± 48.41 mm ² , 1,056.59 ± 48.12 mm ² for EEA and LTO vs. EEA and TATP). Statistically speaking, adding both posteromedial and posterolateral regions of petrous apex (PA) increased the exposed surface area up to 523.90 ± 51.26 mm ² using either EEA and LTO or EEA and TATP.

Clinical Application

In the following section, we describe three patients with MC schwannoma. The surgical approach used differed according to the extension of the lesion:

Case 1 : a 44-year-old woman presented with a 4-year history of left teeth and lip numbness, a recent onset of mild proptosis, and blurred vision. Physical examination detected severe hypesthesia in the V2/V3 distribution. Magnetic resonance imaging (MRI) scans with contrast demonstrated a large T2-hypointense lesion with a central portion of necrosis and/or cystic change, measuring 39 × 46 × 37 mm occupying the central skull base. The lesion extended anteriorly to involve the left pterygopalatine fossae and the posterior aspect of the maxillary sinus, posteriorly to involve the anteromedial aspect of MC, superior orbital fissure, and the anterior aspect of the cavernous sinus, and medially to involve the sphenoid sinus and the posterior ethmoid air cells, laterally to involve the infratemporal fossa (ITF) and superiorly to involve the left orbit. The orbital component superiorly displaced the optic nerve (ON) significantly with possible involvement of inferior and medial rectus musculature. An intracranial component laterally displaced the temporal lobe parenchyma. This mass also abutted the clinoid and supraclinoid segments of the left internal carotid artery (ICA). Given the benign nature of the lesion and the lateral displacement of the neurovascular structures, a standard transpterygoid EEA was selected. MC was opened medially at the quadrangular space then V2, V3, ICA, and ON were identified and carefully dissected. Reconstruction was performed using an ipsilateral nasoseptal flap. Surgery and postoperative period proceeded uneventfully, and the patient's vision improved ( Fig. 6A ).

Fig. 6.

( A–C ) preoperative MRI (coronal, axial, sagittal cuts) of case 1, case 2, and case 3, respectively. MRI, magnetic resonance imaging.

Case 2 : a 45-year-old female patient with a known history of multiple sclerosis was found to have an incidental MC schwannoma on MRI, described as a well-circumscribed 14 × 14 mm T1-hypointense soft tissue mass involving the upper aspect of the left pterygopalatine fossa extending intracranially into MC along FR, widening the FO and extending into the ITF, and exerting some pressure deformity on the posterior aspect of the left maxillary sinus and the left pterygoid process. Through a sublabial TATP approach, the tumor was dissected from the infraorbital nerve and followed posteriorly to expose the Vidian nerve (VN), and FR which was widened by the tumor. The pterygoid base and greater wing of the sphenoid bones around FR were drilled off. After exposing the dura of the MCF, an interdural dissection was performed toward the V3, and Gasserian ganglion (GG). Reconstruction included multilayers of duragen. She was discharged 2 days after surgery with no postoperative complications or neurological deficits except for the expected facial numbness that improved over time ( Fig. 6B ).

Case 3 : a 33-year-old woman presented with trismus and ocular pain precipitated by eating, teeth brushing, or sneezing, occurring almost 3 to 4 times per week. MRI showed an ovoid, extra-axial, 1.7 × 2.1 cm, enhancing nodular mass, centered on the left MC, and associated with chronic remodeling of the adjacent osseous structures. The lesion demonstrated hypointense T1-weighted signal and heterogenous mildly increased T2 signal. A LTO approach through a superior eyelid incision and rim preservation provided adequate surgical freedom and exposure of that relatively small well-circumscribed lesion. The lesion was excised extradurally without any intraoperative cerebrospinal fluid (CSF) leak; thus, requiring no reconstruction. No ophthalmologic consequences were noted, and the patient was discharged 3 days later with very limited eyelid edema and a mild reduction in her mastication power ( Fig. 6C ).

Discussion

Despite advancements in optical technology and expanding familiarity with minimal access techniques, skull base lesions remain surgically demanding. It is understood that no single surgical corridor is optimal in every case. ^{8 9}

The current study demonstrates that the EEA provides a path toward the anteroinferomedial aspect of MC with an average area of exposure of 587.69 ± 38.59 mm ² , which is consistent with other similar studies found in the literature. ^{10 11} Therefore, it is indicated for medially located benign tumors displacing neurovascular structures laterally. It involves no skin incision or brain retraction; thus, resulting in decreased morbidity and faster recovery. ^{7 12} However, it necessitates a substantially longer dissection time and carries the risk of damaging the sinonasal mucosa with subsequent crusting and possible chronic sinusitis, albeit temporary. ³

Compared with the EEA, the TATP approach provides a shorter and a direct corridor to the anteroinferior and lateral portions with a wider surgical freedom (167.70 ± 29.762 mm). However, the TATP affords limited endoscope and instrument maneuverability to approach the upper MC as access is blocked by the maxillary strut. In addition, the visualization of the medial MC and paraclival ICA is inferior to that offered by EEA. Nevertheless, it could be expanded to expose the entire ITF and lateral skull base.

Although the EEA allows bilateral access as opposed to the unilateral access provided by the TATP approach, the TATP obviates the need to remove the entire anterior sphenoidal wall, medial maxillary wall, and posterior septectomy; thus, requiring less time to perform. One can follow V2 all the way back to GG making this approach ideal for tackling perineural spread of malignant tumors or schwannomas. Dural reconstruction using multi-layered grafting is advocated as vascularized flaps are not easily available. ⁶ Significant reduction in mastication force or trismus, oroantral fistula, and hypesthesia are rarely encountered complications. ^{13 14}

Conversely, the LTO gives accessibility to the upper lateral portion that is hard to reach through the TATP (area of exposure is 650.11 ± 35.76 mm ² in LTO compared with 468.90 ± 26.98 mm ² for TATP). Although approaching the PA is feasible, the narrow surgical corridor limited by the petrous ICA and tentorium and the complex anatomy would require significant retraction of both orbit and brain. ^{8 9 15 16 17 18} Additionally, the LTO approach spares the VN, as opposed to the EEA and the scar is hidden within the eyelid crease. The potential injury to the eye due to excessive retraction (>1 cm) seems more theoretical than practical. Furthermore, the risk of CSF leakage is a minor concern since the orbital contents bolster the reconstruction. ^{19 20 21}

Our preclinical experience combining EEA with TATP/LTO suggests that these hybrid approaches offer wider surgical window (1,237.79 ± 48.41 mm ² , 1,056.59 ± 48.12 mm ² for EEA and LTO vs. EEA and TATP) with nearly complete visualization of critical structures, as well as greater flexibility in terms of handling instruments and manipulating tissue. Of note, angled endoscopes were useful as they allow visualization of the most hidden aspects of MCF and enhance the visibility in such limited corridors.

Traditional open approaches (frontotemporal Dolenc's approach, ²² orbitozygomatic, ⁹ subtemporal (Kawase–Shiobara), ²³ modified Dolenc–Kawase approach, ²⁴ retrosigmoid intradural suprameatal, ^{25 26} or presigmoid approaches ²⁷ ) are still employed for extreme lateral and posterior lesions that may not be easily amenable to our multi-corridor approach. ^{28 29} Similarly, lesions that completely encase the GG should be individually tailored based on the pathology, extent of the lesion, availability of surgical navigation, nerve monitoring, and a skilled multidisciplinary team. ^{2 7 30}

As per any cadaveric study, the results obtained herein should be applied on larger clinical series to provide a more realistic evaluation of the new techniques. Additionally, cadavers may not reflect real-world scenarios in terms of brain shrinkage, consistency, absence of CSF or blood, and finally, lack of a lesion that may cause significant anatomical distortion is another limitation. Considering the rarity of pathologies affecting this location, multicenter studies of bigger sample sizes are strongly advocated.

Conclusion

Standard transcranial and endoscopic approaches overlap over the exposure of different areas of MC. This study helps surgeons to gain familiarity with the pros and cons of the studied approaches. These techniques can be applied selectively as a single or combined approaches. Routes provided by traditional open approaches may be required for extreme lateral or posterior lesions, or in those involving critical neurovascular structures.