Endoscopic Transnasal Approach for Deep Lateral Orbital Decompression: A Cadaver Study

Abstract

Objectives

This study evaluates the anatomical feasibility of an endoscopic transnasal approach for deep lateral orbital decompression using cadaver models.

Design

Cadaver study.

Participants

Four fresh frozen cadaver heads (eight sides) were used.

Main Outcome Measures

Measurements of the sphenoid trigone before and after bone removal were assessed using CT scans. Key outcome measures included the width, height, depth, and volume of the trigone.

Results

The transnasal approach achieved a 53.9% reduction in trigone volume, with significant decreases in height (65.0%), width (84.3%), and depth (76.8%). Preoperative measurements revealed an average orbital surface width of 20.0 mm, later reduced to 16.7 mm postoperatively. The average trigone depth was reduced from 16.1 to 12.0 mm, and height decreased from 21.3 to 13.4 mm in the postoperative assessment. The procedure showed a mild limitation in reduction along the cephalocaudal axis, with residual regions measuring 4.5 and 3.0 mm in the superior and inferior directions, respectively.

Conclusion

The endoscopic transnasal approach effectively reduces the size of the sphenoid trigone, providing a promising alternative for orbital decompression with potential clinical applications. Future research should explore long-term outcomes and integration into surgical practice.

Introduction

Orbital decompression is applicable to patients with exophthalmos caused by conditions such as Graves' orbitopathy (GO), addressing both cosmesis and functional issues. ¹ It is also indicated in cases of orbital abscesses, periorbital hematoma, orbital hematoma, and neoplasm. ¹

Numerous surgical techniques for orbital decompression have been reported, ^{2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26} which may result in the reduction of exophthalmos. This is supposed to depend on the number of the removed orbital walls, the extent of their removal, and the anatomy of the bony orbit and periorbital spaces. ²⁰ Among these, deep lateral orbital decompression is noteworthy for its approach to expanding the orbit by removing bone from the orbital portion of the greater sphenoid wing. ²⁷ On axial computed tomography (CT) slices, the deep lateral wall appears as a triangle, hence the denomination sphenoid trigone. ^{26 27 28 29} Removing this portion can significantly reposition the globe posteriorly. ²⁶

Various open approaches are employed in deep lateral orbital decompression, ^{4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25} with the eyelid crease approach being the most commonly used today due to its cosmetic advantages and minimal incision length. ^{11 12 13 14 15 16 17 18 19} However, this approach may carry risks of scarring and facial nerve injury. ¹ Recently, a novel anatomical landmark between the periorbita and the periosteum of the pterygopalatine fossa (which is located at the inferolateral periorbital periosteal line [ILPPL]) has enabled a transnasal approach to the anteromedial temporal fossa, ³⁰ allowing for lateral orbital access without external incisions.

The primary objective of this study is to evaluate the anatomical feasibility of the endoscopic transnasal approach for deep lateral orbital decompression using preserved cadavers. The study endpoints include measuring the extent of the sphenoid trigone removed through this approach using paranasal CT and assessing the reachability of the approach.

Method

Population

This study was conducted using four preserved cadaver heads (eight sides) purchased from Science Care Inc. (Phoenix, Arizona). The cadaveric dissections were performed at the University of North Carolina Surgical Skills Laboratory, adhering to all relevant institutional guidelines and approvals for anatomical research. High-resolution cone-beam CT scans (J. Morita Mfg. Corp., Kyoto, Japan) were performed on each specimen before and after dissection. Contiguous 1-mm axial and coronal CT images were obtained, with axial images reconstructed based on the planes parallel to the skull base.

Surgical Procedure ( Video 1 )

Download video file ^{(11.8MB, mp4)}

Video 1 Endoscopic transnasal approach for deep lateral orbital decompression.

The procedure utilized a 0-degree endoscope (4 mm diameter and 18 cm length), which was paired with a high-definition camera and monitor (Karl Storz Endoscopy, Tuttlingen, Germany) to capture the endoscopic view. Video images were recorded and stored using an AIDA ^® system (Karl Storz Endoscopy). Osseous dissections were performed using a high-speed drill (Stryker Co., Kalamazoo, MI) with a straight handpiece and 4-mm coarse diamond burrs. A comprehensive maxillary antrostomy and total sphenoethmoidectomy were conducted to ensure optimal access to the orbital floor, medial wall, and pterygopalatine fossa (PPF). Additionally, a transseptal approach was employed to expand the operating range, facilitating the use of three or four hands during the orbital surgery. ^{31 32} Using a diamond drill and Kerrison rongeurs, the lamina papyracea and the posterior walls of the maxillary sinus were removed to expose the periorbita and the ILPPL ( Fig. 1A ). After incising the ILLPL, the periosteum was elevated up to the lateral wall of the orbit ( Fig. 1B ). The transseptal corridor was used to elevate the orbital periosteum while drilling out the lateral wall of the orbit (sphenoid trigone) ( Fig. 1C ). On the central side, bone was removed extending to the lateral part of the superior orbital fissure, and the depth of dissection reached the dura of the temporal fossa ( Fig. 1D ).

Fig. 1.

Endoscopic transnasal approach for deep lateral orbital decompression. ( A ) After performing a comprehensive maxillary antrostomy and total sphenoethmoidectomy, the medial and inferior walls are removed, exposing the orbital periosteum and the inferolateral periorbital periosteal line (ILPPL). MS, maxillary sinus; SS, spenoid sinus. White dot line shows ILPPL. ( B ) Incise the ILPPL, elevate the periosteum, and expose the lateral orbital wall. Asterisk (*) shows lateral orbital wall. ( C ) Using assistance from 3 to 4 hands, drill out the lateral orbital wall. ( D ) The temporal dura is visible following the removal of bone from the lateral orbital wall. White arrow shows temporal dura.

Outcome Measures

The primary outcome was the extent of resection of sphenoid trigone, assessed by comparing preoperative and postoperative CT scans. Specific measurements included.

Preoperative CT Assessments

As previously described in the report, ²⁶ the axial section selected showed the globe, optic nerve, and medial rectus muscle ( Fig. 2A ), and the coronal section showed the most posterior portion of the sphenoid trigone not passing through the mid-cranial fossa ( Fig. 2B ). In the axial section, measurements included the width (orbital surface) of the sphenoid trigone, defined as the distance from the anterior vertex to the posteromedial vertex, and the depth, defined as the distance between the posteromedial vertex and the posterolateral vertex. Additionally, the distance from the orbital rim to the anterior vertex was measured. Using OsiriX imaging software (Pixmeo, Switzerland), the sphenoid trigone was identified and outlined in the axial scans, enabling the calculation of the region of interest (ROI) in square millimeters. ²⁹ In the coronal section, the height of the trigone was defined as the vertical distance between the lowest part of the superior edge and the lowest part of the inferior edge of the bone marrow inside the trigone. The model assumed that the trigone was shaped as a triangular prism, as proposed by Kitaguchi et al, ²⁶ and the volume was calculated as the product of the ROI and the height.

Fig. 2.

CT images of the trigone before and after bone removal with measurements between anatomical landmarks. Both the arrows indicate the measured distances between each anatomical landmark. ( A ) Axial section CT before bone removal. A, anterior vertex of the trigone; PL, posterolateral vertex of the trigone; PM, posteromedial vertex of the trigone; R, edge of the lateral orbital rim. ( B ) Coronal section CT before bone removal. Height: The vertical distance between the lowest point of the superior edge and the lowest point of the inferior edge of the bone marrow within the trigone. ( C ) Axial section CT after bone removal. AL, the anterior limit of the approach; LL, the lateral limit of the approach; PL, the posterior limit of the approach; R, edge of the lateral orbital rim. ( D ) Coronal section CT after bone removal. Α, the height of the resected trigone; Β, the distance from the lowest point of the superior edge of the bone marrow within the trigone to the upper limit of resection, representing the residual superior part; γ, the distance from the lower limit of resection to the lowest point of the inferior edge of the bone marrow within the trigone, defined as the residual inferior part.

Postoperative CT Assessments

Postoperative CT assessments involved extracting the same areas as the preoperative ones in both axial and coronal sections. In the axial section, the width of the orbital surface removed, the depth from the orbital surface, and the distance from the orbital rim to the anterior limit of the approach were measured ( Fig. 2C ). Similarly, using OsiriX imaging software, the removed area was calculated as the ROI. In the coronal section, the height of the resected trigone was measured. The distance from the lowest part of the superior edge of the bone marrow inside the trigone to the upper limit of resection was defined and measured as the residual superior part. Similarly, the distance from the lower limit of resection to the lowest part of the inferior edge of the bone marrow inside the trigone was measured as the residual inferior part ( Fig. 2D ).

Statistical Analyses

Patient's age, measurement results, and calculated volumes were expressed as means ± standard deviation and range. All data analyses were conducted using GraphPad Prism 8 software (GraphPad Software, Inc., La Jolla, CA, USA).

Results

The study evaluated a cohort of preserved cadaver heads, totaling eight sides. The average age of the specimens was 70.8 years (± 7.3), with an age range from 65 to 81 years, and all were female. Preoperative measurements showed that the average distance from the orbital rim to the anterior vertex was 16.0 mm (± 4.4), with a range of 10.7 to 23.0 mm. The width of the orbital surface of the sphenoid trigone averaged 20.0 mm (± 2.8), ranging from 16.1 to 24.9 mm, while the depth averaged 16.1 mm (± 3.7), with a variation between 11.6 and 20.7 mm. The height of the trigone was 21.3 mm (± 3.4), ranging from 16.9 to 26.6 mm, and the trigone's volume was calculated to be 1.8 cm ³ (± 0.7), spanning 0.9 to 2.9 cm ³ ( Table 1 ).

Table 1. Patient's background and corresponding measurements of the trigone before bone removal.

Total cohort ( n  = 8 sides)	Overall	Range
Age (years)	70.8 ± 7.3	65–81
Sex ( n , %)
Female	4	100
Male	0	0
Distance from the rim to the anterior vertex (mm)	16.0 ± 4.4	10.7–23.0
Width of the orbital surface of the trigone (mm)	20.0 ± 2.8	16.1–24.9
Depth of the trigone (mm)	16.1 ± 3.7	11.6–20.7
Height of the trigone (mm)	21.3 ± 3.4	16.9–26.6
Trigone volume (cm 3 )	1.8 ± 0.7	0.9–2.9

Note: Width of orbital surface: The distance from the anterior vertex of the trigone to the posterior vertex of the trigone. Data are expressed as means ± standard deviation, along with ranges.

Postoperative measurements indicated a mean distance from the margin to the anterior limit of the approach of 17.9 mm (± 3.0), with a range of 13.7 to 22.2 mm. The width of the orbital plane of the reachable sphenoidal trigone was 16.7 mm (± 2.5), with a range of values from 13.8 to 20.9 mm. The depth of the reachable trigone was 12.0 mm (± 1.4), with a range of 10.2 to 14.1 mm, and the height averaged 13.4 mm (± 1.3), with a range of 11.3 to 15.4 mm. The volume of the excised trigone was 0.9 cm ³ (± 0.2) and ranged from 0.6 to 1.2 cm ³ ( Table 2 ).

Table 2. Corresponding measurements of the trigone after bone removal.

Postoperative measurements	Overall	Range
Distance from the rim to the anterior limit (mm)	17.9 ± 3.0	13.7–22.2
Width of the orbital surface of the removed trigone (mm)	16.7 ± 2.5	13.8–20.9
Depth of the removed trigone (mm)	12.0 ± 1.4	10.2–14.1
Height of the removed trigone (mm)	13.4 ± 1.3	11.3–15.4
Volume of removed trigon (cm 3 )	0.9 ± 0.2	0.6–1.2

Note: Data are expressed as means ± standard deviation, along with ranges.

The approach resulted in significant reductions in the size of the trigone. The orbital surface width was reduced by a mean of 84.3% (± 12.4), with variations from 56.6 to 100.0%. The depth saw a mean reduction of 76.8% (± 12.1), ranging from 60.9 to 100.0%. The height was reduced by a mean of 65.0% (± 14.7), with a range from 48.5 to 86.4%. Overall, the volume removal was 53.9% (± 13.7), varying from 35.6 to 72.4% ( Table 3 ).

Table 3. Trigone removal rate associated with this approach.

Removal rate of trigone by approach	Overall	Range
Width of the orbital surface (%)	84.3 ± 12.4	56.6–100.0
Depth (%)	76.8 ± 12.1	60.9–100.0
Height (%)	65.0 ± 14.7	48.5–86.4
Volume (%)	53.9 ± 13.7	35.6–72.4

Note: Data are expressed as means ± standard deviation, along with ranges.

Additionally, the residual areas identified with the transnasal approach were evaluated. The superior part of the residual area measured 4.5 mm (± 1.3), with a range of 3.1 to 6.5 mm, while the inferior part was 3.0 mm (± 1.8), ranging from 0.1 to 5.2 mm ( Table 4 ).

Table 4. Residual regions in the cephalocaudal direction associated with this approach.

Residual trigone regions with transnasal approach	Overall	Range
Superior part (mm)	4.5 ± 1.3	3.1–6.5
Inferior part (mm)	3.0 ± 1.8	0.1–5.2

Note: Data are expressed as means ± standard deviation, along with ranges.

Discussion

This study evaluated anatomical changes of the sphenoid trigone following the endoscopic transnasal approach for deep lateral decompression in eight sides of four preserved cadaver heads. The results demonstrated substantial reductions in the width, depth, height, and volume of the orbital surface of the sphenoid trigone post-surgery. Specifically, the mean reduction in orbital surface width was 84.3%, depth was 76.8%, height was 65.0%, and volume was 53.9%. These findings underscore the effectiveness of this approach in achieving significant anatomical alterations conducive to treating conditions requiring orbital decompression.

Kitaguchi et al reported that the depth of the trigone, defined as the distance between the posteromedial vertex and the posterolateral vertex, is the only significant positive predictive factor for the reduction of exophthalmos. ²⁶ A greater depth means a larger volume of the trigone. By surgically removing the trigone, a larger new space is created. This allows more of the tissue behind the eye to move into this space, leading to a more effective reduction in exophthalmos. The endoscopic transnasal approach adopted in this study offers a different surgical access route compared with traditional orbital approaches. In particular, access from the opposite nasal cavity via the transeptal corridor facilitates bone resection of the trigone, allowing for the procedure to be performed with a good field of view ( Fig. 3A ). In the study by Noiphithak et al, the surgical freedom of the lateral orbital region was compared between the preservation and osteotomy of the lateral orbital rim and their impact on operative freedom. The results indicated that although lateral orbital rim preservation was associated with surgical freedom at the entry site, there was no significant change in operative freedom in the deep lateral orbital region, particularly in the trigone area. ³³ They pointed out that due to the crowding of surgical instruments, surgeons must become accustomed to uncomfortable technical maneuvers in a very narrow working space. ^{34 35} Kong et al examined the effectiveness of the endoscopic transorbital approach for spheno-orbital meningiomas. They pointed out that the rate of gross-total resection in the medial one-third of the greater sphenoidal wing is poor, highlighting the issue of a very narrow working space associated with the transorbital approach. ³⁶ In contrast, this method allows for sufficient utilization of the nasal cavity space. As a result, on the central side, bone was removed extending to the lateral part of the superior orbital fissure, leading to an 84.3% reduction in the amount of bone resection. On other hand, this approach has limitations in the craniocaudal direction. Specifically, the dimensions of the residual superior and inferior areas post-decompression were 4.5 and 3.0 mm, respectively. Since this approach accesses from below the globe and above the PPF, the orbital contents pose restrictions for superior bone resection, while the PPF contents limit inferior bone resection.

Fig. 3.

Comparison of transnasal and transorbital approaches to the deep lateral orbital wall. Green: Operable range by the transnasal approach. Red: Operable range by the transorbital approach. ( A ) CT axial section—solid line shows transnasal approach; dotted line shows transorbital approach. ( B ) CT coronal section.

Since the lamina papyracea has been removed in this approach, the risk of globe compression and optic nerve traction due to medial displacement is reduced compared with the transorbital approach. However, it is important to consider the risk associated of superior displacement for this approach, as it remains a concern. In addition, this approach is not only suitable for deep lateral orbital osteotomy but also applicable to skull base lesions such as spheno-orbital meningiomas and middle cranial fossa lesions. However, in cases with extension above the superior part of the middle cranial fossa, additional access through an endoscopic transorbital approach directed toward the superolateral part of the middle cranial fossa is preferred (see Fig. 3B ). ^{35 36 37 38}

In this study, we did not measure the average dissection time; therefore, we cannot provide specific data on surgical duration. However, it is anticipated that the endonasal approach, which includes a comprehensive maxillary antrostomy and a transseptal approach, may take longer compared with the transorbital approach. This is due to the complexity of the procedures involved. Nevertheless, as the technique is like endoscopic sinus surgery, it is expected that with accumulated experience, surgeons will be able to reduce the overall surgical time. Although this study demonstrates a promising alternative for orbital decompression, it is essential to evaluate long-term outcomes and complications associated with the endoscopic transnasal approach in a clinical setting. This would entail a comprehensive review of patient cases, focusing on both aesthetic results and functional recovery.

Conclusion

In conclusion, the endoscopic transnasal approach for deep lateral orbital decompression is a viable option that significantly reduces the size of the sphenoid trigone while minimizing the risks associated with traditional surgical techniques. The anatomical feasibility observed in this cadaver study paves the way for future clinical applications and has the potential to improve patient outcomes in the management of exophthalmos and related orbital disorders.