Infratemporal Fossa Vascular Anatomy Pertinent to Percutaneous Access to the Foramen Ovale for Treatment of Trigeminal Neuralgia: A Comparison of Cadaveric Dissection and Computed Tomography Analysis

Abstract

BACKGROUND:

Trigeminal neuralgia may be treated via percutaneous access to the foramen ovale (FO). Vascular complications associated with the needle trajectory can result in serious morbidity and mortality. This study aimed to correlate the vascular relationships of the FO at the skull base via cadaveric dissections and computed tomography (CT).

METHODS:

Two fresh cadaver heads were injected with red and blue latex to delineate arteries and veins. Neck and infratemporal fossa dissections were carried out to delineate the vascular relationships of the FO. High-resolution head CT images of adult patients undergoing neurosurgical evaluations or procedures were analyzed for distances and sizes of skull base foramina in the infratemporal fossa.

RESULTS:

Three infratemporal fossa dissections (2 cadaveric specimens) were performed. Mean distance of FO to internal carotid artery was 2.4 ± 0.12 cm, and mean distance of FO to middle meningeal artery was 0.8 ± 0.16 cm. Head CT images of 52 patients (104 sides) with 1-mm axial slice thickness were analyzed. Area of the FO was 31.1 ± 9.6 mm². Distance of FO to internal carotid artery was 1.70 ± 0.31 cm, and distance of FO to middle meningeal artery was 0.73 ± 0.61 cm.

CONCLUSIONS:

Cadaveric delineation of vascular structures in the infratemporal fossa correlates with head CT imaging and may be used to accurately plan percutaneous access to the FO. Inadvertent puncture of the extracranial internal carotid artery is nearly impossible with good technique. The most likely source of percutaneous vascular injury is the middle meningeal artery and distal branches of the maxillary artery

INTRODUCTION

Percutaneous access to the foramen ovale (FO) for the management of facial pain was first described by Hartel in 1912, who reported passing a needle into Meckel’s cave without injury to the third division of the trigeminal nerve.¹ Early improvements on the technique were developed by Penman,² who detailed an aiming grid, which was later modified by Ecker and Perl.³ The currently accepted technique was described by Lunsford and Bennett⁴ as a modification of a technique by Håkanson.⁵ The technique described by Lunsford and Bennett⁴ involved placing the patient in the supine position with anteroposterior and lateral fluoroscopy. Needle placement was in the cheek 2.5 cm from the corner of the mouth and directed toward the medial aspect of the pupil to a position 2.5 cm anterior to the external auditory canal. Final positioning was confirmed with fluoroscopy and egress of cerebrospinal fluid following stylet removal.

Since popularization of percutaneous treatment of trigeminal neuralgia, the anatomy around the FO has been well characterized.^6–11 The technique has been further improved with advancements in technology allowing for guidance of the needle with x-ray or computed tomography (CT). While many cadaveric studies have provided detailed descriptions of the anatomy of this region,^10,11 correlation of cadaveric anatomy with CT findings of the skull base foramina in the infratemporal region is lacking. We propose the importance of this relationship in that high-resolution CT imaging is readily available to the surgeon, and knowledge of correlation of cadaveric characterization with CT imaging would allow surgeons to better tailor their approach for each individual patient to ensure successful neurolysis and avoid complications. In this study, we analyzed cadaveric dissection of the infratemporal region and high-resolution CT imaging to establish a correlation of the 2 methods of deciphering anatomy of the region of the FO.

MATERIALS AND METHODS

This study was approved by the West Virginia University Institutional Review Board.

Procurement of Cadaver Heads

Two fresh whole human cadavers were procured from the West Virginia University Department of Anatomy. Cadaver heads were fixed, and vasculature was injected with latex as described in Supplementary Figure 1.

CT Imaging Analysis

High-resolution CT imaging of the brain was obtained from consecutive adult patients (>18 years of age) undergoing neurosurgical evaluations or procedures. No patients were excluded from analysis. All imaging was performed with 1-mm axial slices. Image analysis and measurements were done on the Synapse PACS system (FUJIFILM Medical Systems, Morrisville, North Carolina, USA).

Bilateral measurements were recorded from each imaging series. The FO, foramen spinosum, and proximal entrance to the carotid canal were identified in each infratemporal fossa. Width and length of the FO were recorded per measurements on axial imaging. To perform measurements, the position of the foramen spinosum and proximal entrance of the optic canal were extrapolated superiorly or posteriorly until in axial view with the FO. Linear anteroposterior and mediolateral measurements were recorded for the distance between each measurement. The distance of the FO to the internal carotid artery (ICA) was determined as the midpoint of the FO to the midpoint of the proximal entrance of the carotid canal. The distance of the FO to the middle meningeal artery (MMA) was determined as the midpoint of the FO to the midpoint of the foramen spinosum.

Statistical Analysis

Results of measurements are recorded as mean value with standard deviation. Comparison of right versus left was performed using a 2-tailed Student t test with P < 0.05 considered significant.

RESULTS

Cadaveric Results

The exploratory dissection of the neck and infratemporal fossa is shown in Figures 1 and 2. Latex injection showed good penetrance and delineation of arterial and venous architecture. The Hartel technique allowed for cannulation of the FO using an 18G spinal needle (Figure 2A). A small temporal craniotomy was performed to directly visualize the needle emerging from the FO in the middle fossa (Figure 1C).

Figure 1.

Incision to provide exposure for both the small temporal craniotomy and the infratemporal fossa dissection. (A) Musculocutaneous flap. (B) Cut surface of the mandible. (C) Temporal craniotomy depicting percutaneously placed needle traversing the foramen ovale (yellow asterisk).

Figure 2.

Extracranial cadaveric dissection of the infratemporal fossa showing the branches of the internal maxillary artery, the internal cerebral artery (red asterisk), the middle meningeal artery (blue asterisk), and their relationship to a needle passing through the foramen ovale (yellow asterisk) at the skull base. (A) Distance of needle in the foramen ovale to the middle meningeal artery. (B) Distance of needle in the foramen ovale to the internal cerebral artery.

Dissection of the infratemporal fossa was achieved via an extracranial, extranasal approach using a skin incision from the inferior edge and angle of the mandible extending superiorly into the temporal region and anteriorly along the hairline (Figure 1A). The internal maxillary artery and its branches, the ICA, and their relationship to the needle traversing the infratemporal fossa into the FO were visualized deep to the cut mandibular ramus (Figure 1B). The MMA was in close proximity to the FO (Figure 2A). Branches of the maxillary artery traversed the trajectory of the percutaneous needle (Figure 2C). The ICA was lateral and posterior to the plane of the FO (Figure 2A).

Figure 3 demonstrates the significant difference in trajectories of percutaneous cannulation of the FO and skull base cannulation of the extracranial ICA. The average cadaveric distance of the FO to the ICA was 2.4 ± 0.12 cm in 3 infratemporal fossa dissections. The average cadaveric distance of the FO to the MMA was 0.8 ± 0.16 cm. Distal branches of the maxillary artery remain in close proximity to the needle trajectory (Figure 2C).

Figure 3.

Comparison of percutaneous trajectories to the infratemporal fossa. (A and B) Anteroposterior (A) and lateral (B) radiographic depiction of percutaneous access of foramen ovale (yellow asterisk) and proximal entrance of the carotid canal (red asterisk). (C) Cadaveric dissection of the right infratemporal fossa with percutaneous needle located in the proximal entrance of the carotid canal (red asterisk).

CT Results

CT images of 52 patients (104 sides) obtained from October 24, 2018, to March 25, 2020, were analyzed. Scatter plots of data are shown in Figure 4. Anteroposterior and mediolateral measurements of the FO (Figure 4A and B) demonstrate limited variability compared with the cross-sectional area in Figure 4G. The high degree of cross-sectional area variance is the result of disparate shapes of the FO. The average area of the FO was 31.1 ± 9.6 mm²

Figure 4.

Scatter plots showing computed tomography measurements of the bilateral foramen ovale and distances to carotid canal and foramen spinosum. (A and B) Anteroposterior and mediolateral measurements of the foramen ovale. (C and D) Anteroposterior and mediolateral distance to carotid canal. (E and F) Anteroposterior and mediolateral distance to foramen spinosum. (G) Highly variable cross-sectional area of foramen ovale. AP, anteroposterior; ML, mediolateral.

The distances of the center of the FO to the center of the carotid canal and foramen spinosum were different. The average distance of the carotid canal to the FO was 1.70 ± 0.31 cm versus 0.73 ± 0.16 cm for the foramen spinosum (Figure 4C–F). The relatively large distance between the FO and carotid canal makes it nearly impossible to accidentally cannulate the ICA using standard percutaneous entry points and techniques. No significant difference existed in comparison of left versus right anatomy for distance between areas of interest or size of the FO.

DISCUSSION

The results of our study summarize the distances measured from the FO/trigeminal nerve to the carotid canal/ICA and foramen spinosum/MMA. While the distances measured by both modalities showed some variation, both showed that the relative distance from the FO/trigeminal nerve to the foramen spinosum/MMA was greater than the distance from the FO/trigeminal nerve to the carotid canal/ICA. This preservation of relative distances validates that CT imaging can guide preoperative planning.

The size of the FO has been previously reported. In an effort to characterize variations of the FO, Prakash et al.⁶ reported it to have a mean area of 31.310 ± 8.262 mm². Sepahdari and Mong¹² reported a mean short axis diameter of 4.57 mm on the right side and 4.66 mm on left side, which would give an approximate area of 16 mm². Kaplan et al.¹¹ reported a long axis diameter of 6.9–7.2 mm, which would give an approximate area of 38 mm². These prior data are in agreement with our findings and thus provide external validation of these measurements.

Neurovascular injury to the ICA during percutaneous treatment of trigeminal neuralgia via the FO has been previously reported. Nugent¹³ reported a 1% complication rate in 1456 cases reporting that ICA injury occurred when trajectories were too medial and inferior. Other neurovascular structures have also been reported at risk. Revuelta et al.¹⁴ reported a complication of arteriovenous fistula of the external carotid artery following microcompression of the gasserian ganglion, which they attributed to injury to the MMA. Rish¹⁵ reported an injury to the ICA in the foramen lacerum.

Extracranial carotid injury during percutaneous access of the FO has previously been reported to be nearly impossible using proper technique.¹⁶ This cadaveric dissection with visualization of the percutaneous needle traversing the FO is the first to visually confirm this clinical observation. Further, this anatomic analysis of the infratemporal fossa confirms that vascular injury during percutaneous approach to the FO is most likely secondary to injury to the MMA traversing the spinous foramen or MMA branches owing to their closer proximity

The first cadaveric anatomical characterization and quantification of the neurovascular structures of the infratemporal fossa was provided by Tubbs et al.¹⁰ They reported on extracranial dissection of the infratemporal fossa in 20 cadaveric heads (40 sides). Measuring the distance as the nearest aspect between 2 structures, they reported the distance from the FO to the MMA as 3 ± 1.13 mm and the distance from the FO to the ICA at the proximal entrance to the carotid canal as 12 ± 1.6 mm. Kaplan et al.¹¹ reported a further description of this region and included an intracranial, intradural analysis of associated structures in an attempt to better understand the risks of percutaneous access of the FO based on anatomy. They reported a mean distance of the FO to the foramen spinosum as 3 mm (range, 2.9–3.1 mm). They did not report a distance of the FO to the ICA at the proximal entrance of the carotid canal, but concluded that the most likely area of injury to the ICA was intradural at the medial side of the petrolingual ligament.

The discrepancy between the reported distances in this study and distances reported by Tubbs et al.¹⁰ and Kaplan et al.¹¹ can be largely accounted for by method of measurement. The prior publications reported distance between the nearest aspect of 2 structures, whereas the present study reports the distance between the central aspect of 2 structures. Furthermore, the distance as measured by CT took into account only the change in distance from medial to lateral and anterior to posterior. The superior-to-inferior distance was not included in the analysis. This was done purposely for ease of calculation to better mimic the readily available information to the surgeon evaluating a patient with thin-cut high-resolution axial images.

The analysis of cadaveric anatomy of the infratemporal fossa provides a three-dimensional understanding of this region, which is vital for the surgeon who is attempting to pass a needle into this region using only external landmarks. We propose that analysis of CT imaging preserves the relationships of vascular structures. Understanding these relationships on a patient-to-patient basis allows surgeons to better understand the specific anatomy in each patient and thus be better able to percutaneously access the FO without vascular complications.

CONCLUSIONS

Percutaneous access of the FO in the infratemporal fossa has been associated with the risk of injury to neurovascular structures. The closest neurovascular structure is the MMA, which is considerably closer than the ICA, posing a greater risk of injury. While use of CT to define neurovascular structures based on skull base foramina is not mandatory, this study shows that CT can reliably show the relationships and relative distances of these neurovascular structures to the FO and can help to guide the surgeon when performing percutaneous access to the infratemporal fossa.

Abbreviations and Acronyms

CT

Computed tomography

FO

Foramen ovale

ICA

Internal carotid artery

MMA

Middle meningeal artery