1 Introduction
The pterygopalatine fossa (PPF) is an inverted cone-shaped hollow on the skull base. The PPF is defined by the pterygoid process posteriorly, the maxillary tuberosity anteriorly, and the orbital apex superiorly. While this space communicates with the pharynx and nasal cavity, infratemporal fossa, oral cavity, orbit, and middle cranial fossa through various foramina and pathways, it is small and challenging to access. Neoplastic diseases in this space are rare, with the most frequent entities being juvenile nasopharyngeal angiofibroma, schwannomas, and extension of sinonasal malignancies
[1,
2]. A thorough understanding of PPF anatomy is therefore indispensable for safe and effective surgical management.
Our previous studies have shown that the endoscopic transpterygoid approach (ETPA) is an effective and safe approach with low morbidity to treat various pathologies
[3,
4,
5]. The ETPA was first described for access to the lateral recess of the sphenoid sinus
[6]. Nicolai et al. described the ETPA to the PPF and demonstrated the endoscopic anatomy
[7]. The key step in exposure of the PPF is to remove the palatine bone and the posterolateral wall of the maxillary sinus
[4,
7,
8]. However, a reusable teaching model that preserves the entire bony anatomy without destruction remains lacking in clinical education.
Herein, we utilize a new method to demonstrate the complex three-dimensional osseous anatomy of the PPF via endoscopes through natural bony corridors, including the inferior orbital fissure (IOF) and pterygomaxillary fissures (PMF). We also provide a detailed endoscopic description of the neurovascular structures within the PPF as correlated with computed tomography (CT) multiplanar imaging and cadaveric dissection, addressing the longstanding challenge of teaching PPF anatomy.
2 Materials and Methods
2.1 Specimen
This study was approved by the Institutional Review Board of Beijing Tongren Hospital (No. TRECKY2017-033). Three dry skulls (five sides) were used to examine bony landmarks within the PPF and skull base. One PPF of the dry skulls was damaged, and five sides were available for analysis. Three fresh-frozen heads (five sides) were prepared for dissection by injecting the heads with colored silicone; the arterial system was injected with red silicone and the venous system with blue. All dissections were performed from September 2020 to July 2023 at the Department of Rhinology, Beijing Tongren Hospital, Capital Medical University.
2.2 Equipment
A 4-mm-diameter, 18-cm-long, 0° rigid endoscope coupled to a high-definition (HD) camera (Karl Storz Endoscopy Inc., Tuttlingen, Germany) and a cold light source provided visualization during dissection. An IMAGE S system (Karl Storz Endoscopy, Inc., Tuttlingen, Germany) was used to provide high-quality documentation of the dissections. The pictures were taken with a digital camera (Canon 5D Mark III, Tokyo, Japan) using a standard endoscopic adapter (Beijing Fanxing Medical Equipment Co., Ltd., Beijing, China).
A 2.7-mm-diameter, 110-mm-long, 0° endoscope (Beijing Fanxing Medical Equipment) was used for the IOF approach, while a 4-mm-diameter, 180-mm-long 0° endoscope was used for transnasal and PMF approaches.
2.3 Endoscopic approaches and surgical techniques
2.3.1 Endoscopic intranasal approach
In Fig. 1a, the 0°, 2.7-mm or 4-mm rigid endoscope was placed into the nasal cavity. The ostium of the maxillary sinus and posterior wall of the maxillary sinus were anterior to the SF (Fig. 1b). The ethmoidal crest could be traced anteriorly from the SF; It is an important landmark for ligation and electrocoagulation of the sphenopalatine artery (Fig. 1b). The orbital and sphenoid processes of the palatine bone form the SF with the anterolateral wall of the sphenoid bone (Fig. 1c). The sphenoid sinus ostium was ovoid, lying superomedial to the SF (Fig. 1c). The vertical process of the palatine bone forms the medial wall of the PPF.
When a 0° 4 mm diameter endoscope was used, the foramen rotundum and vidian canal could not be seen through the SF. However, when a 0° 2.7 mm endoscope was applied, the foramen rotundum was identifiable (Fig. 1d–e). The foramen rotundum and vidian canal can be seen only on the three sides of the three dry skulls with large sphenopalatine foramina (Fig. 1f).
2.3.2 Endoscopic IOF approach
In all dry skulls (five sides), the posterior wall of the PPF was fully exposed through the IOF, representing the communication between the orbit and PPF in the anterior-posterior direction (Fig. 2a, f). During the approach, the foramen rotundum was first visualized, followed by the vidian canal, palatovaginal canal, and sphenopalatine canal located medially (Fig. 2b–e, g–i).
The foramen rotundum was separated from the superior orbital fissure by the maxillary strut. It opens into the PPF and the vidian canal going through the pterygoid process below and medial to the foramen rotundum (Fig. 2b, g). The palatovaginal canal was identified medial to the vidian canal and ran posteromedially. We noted that the vidian and palatovaginal canals have a conjunct ‘funnel-shaped’ channel (Fig. 2c, g).
However, the anatomy of the PPF shows individual diversity. When the anterior-posterior distance of the PPF is short, the vidian canal is often poorly exposed. The right vidian canal was more poorly exposed in three cadavers than the left ones because the right PPF had smaller dimensions. (Fig. 2g–h).
2.3.3 Endoscopic trans-PMF approach
The PMF is located between the posterior maxillary wall and the pterygoid process. The space of the PMF was adequate for multi-angle observation with a 4-mm endoscope (Fig. 3a, g). The bottom and medial walls of the PPF were completely visible through the endoscopic trans-PMF approach (Fig. 3b–e, h–j). In addition, parts of the anterior wall and posterior wall could be seen (Fig. 3b, h).
The roof of the PPF is formed by the great wing of the sphenoid bone and the basis of the pterygoid process. The maxillary strut is an important anatomical landmark in skull base surgery, separating the foramen rotundum from the superior orbital fissure (Fig. 3f, k). The medial wall of the PPF is formed by the perpendicular plate of the palatine bone, the orbital process, and the sphenoid process. The medial wall articulates with the maxilla anteriorly, the sphenoid posteriorly, the pterygoid process laterally, and the alveolar process inferiorly. The bottom of the PPF is the pterygopalatine canal, which is divided into two canals: the greater palatine canal, which descends within the medial wall of the maxillary bone, and the lesser palatine canal, posterior to the greater palatine canal, which traverses the pyramidal process of the palatine bone.
2.3.4 Cadaveric endoscopic transmaxillary-pterygopalatine approach
All endoscopic dissections were completed with 0° 4-mm rigid endoscopes (Karl Storz Endoscopy Inc., Tuttlingen, Germany). A two-surgeon, four-handed technique was used. Three fresh-frozen cadaveric heads (five sides) were dissected and observed with a 0° endoscope. A vertical mucosal incision was made between the anterior aspect of the inferior turbinate and the inner nostril on the lateral wall of the nasal cavity. For better access to the posterior wall of the maxillary sinus, the mucosal flap with the inferior turbinate and membranous nasolacrimal duct was transected. The infraorbital nerve and vessel were identified, forming the landmark of the PMF, which separates the PPF from the infratemporal fossa (Fig. 4a).
The anterior wall was removed to optimize visibility. The mucosa and bone on the posterior wall of maxillary sinus (MS) were removed to the angle between the roof and posterior wall of the maxillary sinus superiorly, to the vertical process of the palatine bone medially. The anterior fascia layer and fat tissue in the PPF were exposed and meticulously removed (Fig. 4b). The inferior orbital artery nerves, sphenopalatine artery, SF, and sphenopalatine ganglion became accessible (Fig. 4c). By tracking along the infraorbital nerve, the maxillary nerve was seen. Next, the removal of the medial wall of the PPF was continued. Part of the perpendicular plate of the palatine bone was removed posteriorly to the sphenoid process, superiorly to the orbital process, and inferiorly to the pterygopalatine canal. The pharyngeal, vidian, and greater palatine artery nerves were visualized (Fig. 4d–e). The posterior wall of the PPF became visible. The perpendicular plate of the palatine bone was drilled into the medial wall of the pterygoid plate. Simultaneously, the pharyngeal and vidian artery nerves were transected. The palatovaginal canal, the vidian canal, the foramen rotundum, “those backward corridors” were completely exposed (Fig. 4f).
2.4 Imaging study
CT scans of the entire three dry skull bones were performed on 256-multidetector-row CT scanner (Brilliance 256, Philips Medical Systems, Cleveland, OH; scan parameters: 120 kV; 300 mA; matrix size 512 × 512; rotation time 0.5 s; slice thickness 0.5 mm), at Beijing Tongren Hospital, Capital Medical University, Beijing, China. Images were reconstructed and windowed for bone (window width: 2000 Hounsfield units [HU]; window level: 200 HU), axial plane: parallel to the infraorbital line, from the top of the frontal sinus to the hard palate; coronal plane: perpendicular to the hard palate, anterior to the frontal sinus to the rear of the sphenoid sinus; sagittal plane: parallel to the midsagittal plane, including the lateral edges of the maxillary sinus on both sides.
2.5 Educational assessments
All the trainees (Otolaryngologist) watched the anatomy videos (Suppl. Video) and took the survey form. The data on respondent characteristics and demographics were collected.
2.5.1 Comparison of exposure quality scores of key anatomic structures in the PPF in dry skulls versus frozen cadaveric heads
A 0–5 point scale was used for assessment of exposure: 0 (no exposure), 1 (poor exposure), 2 (fair exposure), 3 (suboptimal exposure), 4 (good exposure), and 5 (excellent exposure).
2.5.2 Comparison of scores for overall educational effectiveness and practicality evaluation
A 0–5 point scale was employed to assess teaching effectiveness, with scores ranging from poor to excellent.
2.6 Qualitative and statistical analysis
Predefined metrics, including the identification of foramina, boundaries of the dry skull, surgical boundaries, and neurovascular and musculoskeletal structures encountered, were calculated. GraphPad Prism 8.0 (GraphPad Software; USA) and SPSS 27 (IBM Corporation, USA) were used to perform. χ2 test (fisher’s exact test) was used to analyze the differences in the percentage of exposure of anatomy structure among the different approaches. Mann-Whitney U test was used to analyzed the differences in exposure quality scores and the scores of overall educational effectiveness and practicality evaluation. The proportion of 5-point scores and average scores were used for the analysis. All P values were two-sided, and statistical significance was set at P < 0.05.
3 Results
3.1 Exposure of osseous anatomical structures
The rates of exposure of bony anatomical structures of dry skulls are summarized in Table 1. The endoscopic PMF approach offered a complete exposure of the bony anatomy the PPF in dry skulls, whereas the endoscopic IOF and intranasal approaches offered an adequate exposure of the PPF.
3.2 Exposure of bony and neurovascular structures and pathways
The exposure of bony and neurovascular structures of the PPF is shown in Fig. 4. All the boundaries, foramina, and neurovascular structures were completely exposed in the endoscopic transmaxillary-pterygopalatine approach (ETMPA). In this approach, the posterior and medial walls of the PPF were resected to demonstrate the internal structure of the PPF.
3.3 Details of foramina of the PPF on CT scans and endoscopy
On the axial CT images of the five sides, two distinct canals behind the PPF could be seen (Fig. 5a). The lateral canal was the vidian canal that connect the PPF to the lateral margin of the foramen lacerum. The palatovaginal canal ran posteromedial to the nasopharyngeal cavity. The SF lay in the anteromedial region of the PPF, which opened into the nasal cavity. We could obtain a full range of the posterior portion of the PPF from the IOF.
On coronal CT images, the palatovaginal canal was medial to the vidian canal, both of which were separated by a thin bony septum that thickened gradually from anterior to posterior. The foramen rotundum was lateral to the sphenoid sinus and separated from the vidian canal and palatovaginal canal, based on the pterygoid process (Fig. 5b).
The full trajectory of the pterygopalatine canal could be tracked using the endoscopic trans-PMF approach. The pterygopalatine canal was at the end of the inverted pyramid, housing the descending palatine artery and nerves. This canal was simultaneously identified in three dimensions of the CT scan (Fig. 6).
3.4 Educational assessment
3.4.1 Characteristics of trainees
A total of 51 trainees/otolaryngologists completed the training and returned the questionnaire. The majority of participants were attending physicians (31, 60.78%), followed by resident physicians (8, 15.69%), associate chief physicians (8, 15.69%), and chief physicians (4, 7.84%). Most trainees were from level 3 hospitals (42, 82.35%), while 9 were from level 2 hospitals and none were from level 1 hospitals (Suppl. Table 1). None of the trainees reported being highly familiar with the anatomy of the PPF. Instead, 28 (54.90%) were slightly familiar, 13 (25.49%) were unfamiliar, and 10 (19.61%) were familiar with the PPF anatomy (Suppl. Table 2).
3.4.2 Comparison of exposure quality scores of key anatomic structures in the PPF via IOF and PMF
The exposure quality scores (proportion of 5-point scores and average scores) of key bony anatomic structures of dry skull specimens were significantly higher than those of fresh cadaveric specimens (p < 0.01). It included all the boundaries and foramina of the PPF. The comparison of exposure quality scores of key bony anatomic structures in the PPF was summarized in Suppl. Table 3.
3.4.3 Comparison of scores for overall educational effectiveness and practicality evaluation via IOF and PMF
The scores (proportion of 5-point scores and average scores) of 3D comprehension, structure identification and improvement of learning confidence of dry skull specimens were significantly higher than those of fresh cadaveric specimens (p < 0.05) (Suppl. Table 4).
The scores (proportion of 5-point scores and average scores) of reusability and applicability of specimens, and convenience of specimens for clinical teaching of dry skull specimens were significantly higher than that of fresh cadaveric specimens (p < 0.01) (Suppl. Table 3). In terms of exposing and identifying the structures of the PPF, 45% of the trainees chose the dry skull and only about 21% chose fresh specimens (Suppl. Table 5).
4 Discussion
In our study, we used novel and innovative endoscopic approaches to provide a quantitative description of the endoscopic osseous anatomy of the PPF in dry skulls. An objective comparison was performed by analyzing the rate of exposure of the PPF. The main limitation of this study is the limited application of these novel endoscopic approaches to the PPF in cadaveric heads, as well as the small sample size of this preclinical anatomical study. Currently, there is no practical surgical approach that is associated with the IOF and the PMF. The utility of this article is that it provides in depth detail of intact osseous anatomic structure of the PPF and may assist surgeons in mastering the complex anatomy of this area.
The anatomy of the PPF could be appreciated by removing the posterior wall of the maxillary sinus through an ETMPA
[8]. In comparison, one of the drawbacks of the endoscopic transmaxillary-pterygopalatine approach is destroying the integrity of the PPF. In our study, we demonstrated the bone landmark of the PPF via the endoscopic PMF and IOF approaches. These approaches to the PPF may help the surgeon in mastering the knowledge of complex anatomic structures.
The IOF is an important anatomical landmark during transorbital endoscopic approaches to the skull base. Gerges et al.
[9] described a novel approach to access the ITF via a transorbital endoscopic approach. This approach can serve as an excellent alternative to conventional transnasal approaches. However, they did not expose the anatomy of the PPF either. The IOF was divided into three segments: anterolateral, intermediate, and posteromedial (Fig. 7a–b)
[9]. In our present study, we found that the sizes of the three segments of the IOF were different, and we examined the relationship between the orbital wall and the PPF (Fig. 7b). The posteromedial edge of the IOF was the narrowest segment. Hence, we chose a 2.7 mm rigid endoscope for the intermediate segment of the IOF and demonstrated the anatomy of the PPF. The endoscopic IOF approach offered adequate exposure of the osseous anatomy of the PPF. However, applying this approach to the PPF in fresh cadaver heads is warranted.
The PMF is the lateral border of the PPF, and is best accessed from the infratemporal fossa (Fig. 7c). The maxillary artery enters the PMF, acting as an important landmark. Previous studies have investigated the size of the PMF and showed that the mean size of the PMF is 4.94 ± 1.20 mm and 4.57 ± 1.18 mm
[10,
11]. Through the endoscopic PMF approach, the corridor space is larger, the 4-mm endoscope makes it easier to approach the PPF, and the exposure of each wall is more sufficient. However, the PMF is unsuitable for the surgical approach in cadaveric heads because the mandible and soft tissue of the infratemporal fossa obstruct the endoscopic PMF approach. Our results indicate that the endoscopic PMF approach could be a better way to learn the bony landmarks of the PPF.
The greater and lesser palatine neurovascular structures enter the oral cavity through the greater palatine foramen, so these canals should be approached carefully during surgical procedures to avoid damage to these neurovascular structures
[12,
13]. One of the unique findings from our study was the demonstration of the greater and lesser palatine canal endoscopically via the PMF. We found that it was possible to master the greater and lesser palatine canals and correlate their bony landmarks with sinus CT findings.
The maxillary strut is the bone that separates the foramen rotundum and superior orbital fissure. The maxillary strut and foramen rotundum are landmarks of anteromedial and anterolateral middle fossa triangles. The maxillary strut and foramen rotundum were well demonstrated by multiple studies. Most of these studies used cadaver heads, and the posterior wall of the maxillary sinus was required to be removed to expose the maxillary strut and foramen rotundum
[14,
15]. In our study, the osseous anatomy of the PPF was demonstrated via the endoscopic transmaxillary-pterygopalatine approach in cadaver heads and the IOF and PMF approaches in dry skull heads. The relationship between the maxillary strut, foramen rotundum, and middle cranial fossa was demonstrated within the PPF via the IOF and PMF approaches.
Accurate identification of important landmarks with multiplanar CT is a vital learning method because it optimizes understanding of local anatomical structures and cultivates awareness of adjacent key structures in three dimensions
[16]. Several studies have created a three-dimensional (3D) printed model of the PPF as a visual and detachable learning tool for completely visualizing the fossa and its boundaries
[17]. However, it was simplified, certain structures were inaccurate, or their course was slightly modified for better visibility and disassembly. In the future, rebuilding a precise 3D model based on multiplanar CT and MRI may better demonstrate the complex relationships between the PPF and the adjacent spaces. Our study explored the endoscopic view of the PPF from different perspectives while keeping it intact, providing a new and precise teaching method for beginners.
Regarding educational assessment, there were 51 Chinese otolaryngologists participated in the study and completed the questionnaire. Our results showed dry skull specimens received significantly higher exposure quality scores for the key PPF anatomic structures compared to fresh cadaveric specimens. Dry skull specimens were also rated significantly higher than fresh cadaveric specimens in terms of 3D comprehension, structure identification, learning confidence, reusability, and convenience for clinical teaching. Our results indicated that when using the novel approach (IOF and PMF), dry skull specimens proved superior to fresh cadaveric specimens for the PPF bony anatomy training based on trainee-perceived effectiveness and practicality from questionnaire feedback.
It should be noted that fresh cadaveric specimens received lower trainee ratings for the IOF- and PMF-endoscopic teaching views in this study. This finding can be largely attributed to soft-tissue obstruction: in fresh cadavers, muscles, fat, and the mandibular apparatus within the infratemporal fossa occupy the natural bony corridors of the IOF and PMF, limiting direct endoscopic visualization of the PPF bony landmarks without extensive tissue removal. By contrast, dry skull specimens have no overlying soft tissues, leaving these bony pathways fully patent for endoscopic observation of osseous anatomy. We emphasize that this perceived advantage of dry skulls is specific to our bone-focused teaching objective for the intact PPF osseous anatomy. Fresh cadavers remain indispensable for learning soft-tissue neurovascular relationships and hands-on surgical dissection and are not inferior for general anatomical training. Additional practical factors including reusability and long-term teaching convenience also contributed to higher questionnaire scores for dry-skull models.
5 Conclusion
The demonstration of osseous anatomy of the PPF was provided via the endoscopic IOF, the PMF and intranasal approaches. When compared with the innovative endoscopic IOF approach, the PMF approach seems to achieve a similar extent of exposure to the PPF in dry skulls. The anatomical characteristics of the PMF, combined with CT scans and cadaveric dissection, are well understood. For training in the PPF bony anatomy, dry skull specimens were rated by participants as more effective and practical than fresh cadaveric specimens using the novel approach, according to questionnaire-based subjective assessments. This dry-skull anatomy model may be used for endoscopic anatomy teaching. A larger sample size of dry skull heads is needed to validate the reproducibility.
6 Supplementary files
Supplementary material is available in the online version of this article at
https://doi.org/10.15302/ENTD.2026.090009 and is accessible for authorized users.
The Author(s). This article is published by Higher Education Press at journal.hep.com.cn.
This is an open access article distributed under the Creative Commons Attribution License 4.0 (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.