Surgery of the Anterior Infratemporal Fossa (AITF) and the Pterygopalatine Fossa (PPF) lesions: critical assessment of operative exposure, loss of functioning tissue and morbidity between Extradural Subtemporal Infratemporal Approach (ESITA) and Endonasal Endoscopic Transmaxillary Transpterygoid Approach (EEMP)

Abstract

Background

Surgery of the anterior infratemporal fossa (AITF) and pterygopalatine fossa (PPF) is technically demanding. The Extradural Subtemporal Infratemporal Approach (ESITA) and the Endonasal Endoscopic Transmaxillary Transpterygoid (EEMP) Approach are established routes for accessing these regions. This study quantitatively compares the surgical exposure, surgical freedom, angles of attack, and clinical outcomes achieved with these approaches.

Methods

ESITA and EEMP were performed on ten sides of five cadaveric heads. Resection volumes were calculated using 3D reconstructions from thin-slice CT scans obtained before and after dissection. Additional measurements—including exposure length, surgical freedom, and angles of attack—were obtained with a neuronavigation system. A retrospective review of twenty-two clinical cases was also conducted to evaluate outcomes.

Results

ESITA provided a greater resection volume for the AITF than EEMP (92% vs 78.7%, p = 0.043), while EEMP achieved a larger resection volume for the PPF (100% vs 73.1%, p = 0.043). EEMP offered superior vertical exposure to the PPF, whereas both approaches provided comparable AITF exposure. ESITA demonstrated wider surgical freedom and more favorable angles of attack. Clinically, GTR and NTR was achieved in 56% of transcranial, 88% of endoscopic, and 60% of combined cases. Minor complications such as nasal crusting and transient facial numbness occurred mainly in the endoscopic group, while one patient experienced meningitis after transcranial surgery and another developed visual impairment following embolization.

Conclusions

ESITA is advantageous for AITF lesions, while EEMP is better suited for PPF pathology. ESITA also provides greater surgical maneuverability compared to EEMP.

Introduction

Surgery for lesions in the AITF (Anterior Infratemporal Fossa) and PPF (Pterygopalatine Fossa) remains difficult and challenging for neurosurgeons and ENT specialists. The conventional preauricular or postauricular Fisch infratemporal (ITF) approaches were standard operative techniques; however, these ITF surgeries (Fisch type A, B, C) were associated with extensive and aggressive dissection and significant risks of postoperative morbidity [3, 6, 7, 14]. Over the past two decades, AITF lesions have been operated on using the less invasive transcranial Extradural Subtemporal Infratemporal Approach (ESITA) [13, 15, 21]. The ESITA technique is simple, safe, and provides a wide operative corridor with an extensive surgical field in the depth between the V2 and V3 branches (vidian triangle) with negligible postoperative morbidity. In recent years, the Endonasal Endoscopic Transmaxillary Transpterygoid (EEMP) Approach has been described and increasingly adopted as the preferred minimally invasive operative technique for the AITF and PPF [1–5, 8–10, 12, 17–20, 22, 23].

This study aims to provide quantitative measurements of surgical exposure, surgical freedom, angles of attack, and clinical outcomes offered by the ESITA and EEMP approaches to lesions involving the AITF and PPF.

Material and methods

Definition

The AITF is defined by the posterolateral surface of the maxillary sinus anteriorly, the lateral pterygoid plate anteromedially, the medial pterygoid muscle and tensor veli palatine muscle medially, the ramus of the mandible laterally and the greater wing of the sphenoid bone superiorly. The posterior limit of the AITF is the foramen rotundum anterior to the tympanic part of the temporal bone. The inferior boundary of the AITF connects with the neck but there are no clear definitive landmarks. In this study, the horizontal line parallel to the hard palate is set as an inferior limit of the AITF.

The PPF is bordered by the posterior wall of the maxillary sinus anteriorly, the pterygoid process posteriorly, the pterygomaxillary fissure laterally and the perpendicular plate of palatine bone medially. The inferior limit is the superior opening of the greater palatine canal.

Description of the dissection techniques

ESITA

The head was positioned supine with the head moderately rotated to lateral and fixated in a 3-pin skull clamp. A preauricular sickle or question mark incision was made from the level of the inferior zygomatic point and coursing superiorly to the squamosal suture (Fig. 1A). A 4 × 5 cm rectangular temporal bone flap was turned with minimal bone loss. Remaining subtemporal bone along the inferior margin of the craniotomy is drilled away toward the foramen rotundum and ovale exposing the infratemporal fascia.

Fig. 1.

Photographs of stepwise cadaveric dissection illustrating the ESITA approach. A A preauricular question mark incision from the inferior zygomatic point and the squamosal suture. B The subtemporal dura is held gently. Foramen rotundum and ovale are skeletonized. C Along the V2-V3 vidian loop toward the wider AITF, AITF structures were identified. D After pterygoid plate and muscle had been removed, PPF was exposed. The orientation in picture (D) is shown as A = anterior, I = inferior, P = posterior, S = superior. ATN = auriculotemporal nerve, BN = buccal nerve, DPN = descending palatine nerve, DTN = deep temporal nerve, IAN = inferior alveolar nerve, ION = infraorbital nerve, LN = lingual nerve, MMA = middle meningeal artery, MN = masseteric nerve, PF = pharyngobasilar fascia, PPG = pterygopalatine ganglion, VN = vidian nerve

The subtemporal dura is held gently using a pair of 2 mm tip-10 mm steel rigid spatula. The foramen rotundum and the ovale are gently skeletonized to expose the peripheral branches (Fig. 1B). Between the V2 and V3, the vidian nerve bony canal will be exposed. Along the V2-V3 vidian loop toward the wider AITF (lateral vidian triangle), we can easily expose and identify AITF structures and boundaries (Fig. 1C). PPF was exposed after pterygoid plate and muscle had been removed from posterior aspect (Fig. 1D).

EEMP

We investigated the standard Endoscopic Endonasal Transmaxillary Transpterygoid Approach using the binostril, 4-handed surgical technique.

Step 1: Extended Transmaxillary Transsphenoidal Approach

A wide medial maxillectomy should be performed extending from the nasolacrimal duct to the posterior wall of the antrum and from the inferior orbital wall to the floor of the nasal cavity. A posterior septectomy was performed to improve access from the contralateral nostril and to allow 2 surgeons, 4 hands working simultaneously. An approach began with trimming the inferior portion of the middle and superior turbinates while preserving the olfactory mucosa. Antero-posterior ethmoidectomy, large maxillary antrostomy with removal of the inferior turbinate and sphenoidotomy were performed with complete exposure of the lamina papyracea and orbital apex (Fig. 2A).

Fig. 2.

Photographs of stepwise cadaveric dissection illustrating the EEMP approach. A A wide medial maxillectomy followed by removal of the inferior turbinate and sphenoidotomy. B The posteromedial wall of the maxillary sinus was removed, exposing the PPF. C By removing the vertical process of the palatine bone, the DPN and DPA were uncovered. D After opening the periosteal plane behind the maxillary antrum and removing fat pad, the AITF was demonstrated. BN divided two heads of LPM. ATN = auriculotemporal nerve, BN = buccal nerve, DPA = descending palatine artery, DPN = descending palatine nerve, DTN = deep temporal nerve, ET = eustachian tube, IAN = inferior alveolar nerve, ION = infraorbital nerve, LN = lingual nerve, LPM = lateral pterygoid muscle, Max = maxillary sinus, MMA = middle meningeal artery, MN = masseteric nerve, NC = nasal cavity, PP = pterygoid plate, PPG = pterygopalatine ganglion, Sp = sphenoid sinus, SPA = sphenopalatine artery, VN = vidian nerve, TM = temporalis muscle

Step 2: The Transpterygoid Approach: Exposure of the PPF and LRS

The sphenopalatine artery behind the crista ethmoidalis of the palatine bone was identified, then the orbital and sphenoid processes of the palatine bone were removed. The posteromedial wall of the maxillary sinus was removed in a medial to lateral direction, exposing the PPF. During the removal of the bone, the periosteum behind the posterior wall of the maxillary sinus was carefully kept intact. By removing the vertical process of the palatine bone, the descending palatine artery and palatine nerve were uncovered and divided. The vidian canal (medially) and the foramen rotundum (superolaterally) could be identified (Fig. 2B, C). The vidian canal was drilled posteriorly to locate the lacerum segment of the ICA. Completely removing the bone medial to the V2, above the medial pterygoid plate, and anterior to the foramen lacerum resulted in full exposure of the LRS.

Step 3: Removal of the pterygoid plate and muscle: Exposure of the AITF and parapharyngeal space

The sagittal plane passing through the infraorbital nerve is a landmark dividing the ITF (lateral) and the PPF (medial). Adequate exposure of the lateral part of ITF required removal of the posterolateral wall of the maxillary sinus from the floor to the roof. After opening the periosteal plane behind the maxillary antrum, the fat pad was carefully removed. The origin of the lateral pterygoid muscle at the infratemporal surface and lateral surface of the lateral pterygoid plate can be identified and detached in a subperiosteal fashion.

The important contents of the AITF space were then identified at this stage. The deep portion of the temporalis muscle outlined at the lateral border of the ITF. The buccal nerve was identified between the 2 heads of the lateral pterygoid muscle. The lingual and inferior alveolar nerves passed behind the inferior margin of the lateral pterygoid muscle. The deep temporal and masseteric nerves coursed below the inferior surface of the greater sphenoid wing (Fig. 2D). The second part of maxillary artery can be identified medial or lateral to the inferior head of the lateral pterygoid muscle. Then, the root of the lateral pterygoid plate was drilled out posteriorly leading to the V3. After separating the medial pterygoid and tensor veli palatine muscles from the bony attachments, the pterygoid processes could be removed resulting in exposure of the parapharyngeal space. By complete removal of the pterygoid processes, the cartilaginous eustachian tube medial to the tensor veli palatine muscle and the levator veli palatine muscles underneath the eustachian tube could be identified.

Quantitative measurements

The microanatomical dissection of both the ESITA and EEMP methods was conducted using five injected head specimens at Chula Soft Cadaver Center. Before and after the dissection, fine-cut 0.625 mm CT examinations were obtained with all head specimens covered and protected. The EEMP was performed on the right side (5 sides) using a standard endoscope (4 mm in diameter and 18 cm in length, with 0° and 30° lenses) attached to a high-definition camera. An ESITA was performed on the left side (5 sides) with operative microscopes. Measurements were obtained using the BRAINLAB Cranial Navigation system. Each measurement was performed 3 times by one examiner, to reduce intra-observer variability, and then the average was calculated. We also evaluated the loss of normal structures in each step during dissection.

After complete exposure of the PPF, the length of the descending palatine nerves was directly measured, which represented the height of the PPF. The length of the masseteric (MN) and deep temporal nerves (DTN) coursed almost horizontally underneath the temporal floor, representing the width of the AITF. Similarly, the lingual (LN) and inferior alveolar nerves (IAN) passed through the AITF in a vertical direction, representing the height of the AITF.

Resection volume (%)

The percentage of resection volume of AITF and PPF was calculated by comparing the pre- and post-dissection volumes using CT scans and BRAINLAB navigation software (Fig. 3).

Fig. 3.

Three-dimensional reconstructions of (A) the AITF and (B) the PPF in all five head specimens from antero-posterior view. The EEMP and the ESITA were performed from the right side (R) and the left side (L), respectively. Yellow and orange represented the resectable and the unresectable volume of each location, respectively

Length of exposure (%)

The percentage of the nerve exposure representing the height of PPF, the height of AITF and the width of AITF was computed by dividing the length of each nerve measured by navigation and CT scan.

Surgical freedom (cm2) and angles of attack (degrees)

Our study measured surgical freedom and angles of attack in four targets including the foramen rotundum, foramen ovale, the most inferior point of IAN and LN, and the most lateral point of DTN and MN (Fig. 4).

Fig. 4.

The three-dimensional reconstruction of (A) the ESITA and (B) the EEMP approaches created illustrating the method of measurement for the surgical freedom (light green trapezoid), vertical angle of attack (red angle) and horizontal angle of attack (green angle) for the target of interest; foramen ovale. The reconstructions were obtained with BRAINLAB Cranial Navigation system

The maximal permissible working area at the proximal end of a 12-cm BRAINLAB navigation probe was placed along the four extreme-most positions in space, forming an imaginary trapezium for each target of interest.

The angles of attack were calculated by moving navigation probe as far as possible in the vertical and horizontal planes, while fixating its distal end on each target of interest.

Statistical analysis

Continuous outcomes from two surgical approaches group were compared using the median with Related-Samples Wilcoxon Signed Rank Test. Statistical significance was set at p < 0.05. Analysis was performed using commercially available statistical software (SPSS, version 29.0, IBM corporation, Armonk, New York).

Loss of functioning tissue

Functioning tissue was defined as any nerves, vascular structures, bony structures and muscles. The tissue loss was recorded during each step of cadaveric dissection.

Clinical outcome evaluation

A retrospective review of clinical charts and operative records at King Chulalongkorn Memorial Hospital from January 2018 to October 2022 included 22 cases. The inclusion criteria included patients with various pathological processes involving the AITF and PPF areas. Patients underwent surgical treatment via transcranial, endoscopic, or combined methods. The extent of resection was categorized as gross total resection (no residual tumor in post-operative imaging), near total resection (residual tumor less than 10%) and subtotal resection (residual tumor more than 10%).

Results

Quantitative measurements (Fig. 5)

Fig. 5.

Quantitative comparisons of the ESITA and EEMP approaches. A Length of exposure (B) Resection volume (C) Surgical freedom (D) Angles of attack

Length of exposure (2-dimensional exposure)

The height and width of AITF measured from two approaches were not statistically different. In contrast, the EEMP provided better exposure for the height of PPF than the ESITA (96.7% vs 86.3%, p = 0.043).

Resection volume (3-dimensional exposure)

In the AITF, the ESITA provided a larger resection volume than the EEMP (92.0% vs 78.7%, p = 0.043). Conversely, for the PPF, the EEMP contributed 100% of the resection volume (73.1% vs 100%, p = 0.043).

Surgical freedom and angles of attack

ESITA demonstrated greater surgical freedom than EEMP for all targets of interest, including the foramen rotundum, foramen ovale, most inferior point of AITF and most lateral point of AITF. ESITA also provided wider vertical and horizontal angles of attack for each target, enhancing surgical maneuverability.

Loss of functioning tissue

Functioning tissue loss was recorded from lateral to medial structures. In the EEMP approach to lateral targets in the AITF, removal of functioning tissue on the medial side included the pterygopalatine ganglion, descending palatine nerve, posterior superior alveolar nerve, vidian nerve, pterygoid plates, and muscles. Conversely, no functioning tissue was sacrificed when approaching median structures in the PPF. ESITA required more tissue destruction on the lateral side to access medial targets in the PPF, sacrificing the zygoma, deep temporal nerve, masseteric nerve, and pterygoid plates and muscles. In contrast, approaching the AITF from the ESITA led to sacrificing only the zygoma, deep temporal nerve, and masseteric nerve.

Clinical outcome

We included a total of 22 cases undergoing surgery (Table 1), comprising both transcranial and endoscopic approaches in the AITF and PPF area. Of these, nine patients (41%) received transcranial surgery, eight patients (36%) received endoscopic surgery, and the remaining five patients (23%) were in the combined group. The mean follow-up time was 26.2 months. The pathology of lesions included angiofibroma (36%), schwannoma (18%), adenoid cystic carcinoma (14%), meningioma (9%) and other pathologies (23%). Gross total resection (GTR) and near total resection (NTR) were achieved in 56%, 88% and 60% of patients in the transcranial, endoscopic, and combined groups, respectively.

Table 1.

Summary of subjects

Number	Age/sex	Histology	Location	Approach	Resection	Complications	F/U (months)
1	48/F	Schwannoma	AITF, PPF	ESITA	STR		46
2	55/F	Schwannoma	PPF	ESITA	GTR		45
3	57/F	Adenoid cystic carcinoma	AITF	ESITA	STR		7
4	74/M	Metastatic carcinoma	AITF, PPF, NC	ESITA	GTR	Pneumocephalus, meningitis	3
5	67/M	Meningioma	AITF, PPF	ESITA	NTR		15
6	39/F	Meningioma	AITF, PPF	ESITA	STR		4
7	32/M	Giant cell tumor	AITF	ESITA	NTR	Wound dehiscence	46
8	62/F	Plasmacytoma	AITF, PPF	ESITA	NTR		13
9	73/F	Lymphoma	AITF, PPF	ESITA	STR		19
10	64/F	Schwannoma	PPF	EEMP	GTR	Transient facial numbness	35
11	55/F	Schwannoma	AITF, PPF	EEMP	NTR	Temporary nasal crusting	24
12	20/M	Angiofibroma	PPF, NC	EEMP	GTR		52
13	17/M	Angiofibroma	PPF, NC	EEMP	GTR	Temporary nasal crusting, Transient facial numbness	46
14	13/M	Angiofibroma	PPF, NC	EEMP	GTR	Visual impairment from pre-operative embolization, nasal crusting	43
15	14/M	Angiofibroma	PPF, NC	EEMP	GTR		7
16	27/F	Adenoid cystic carcinoma	PPF, NC	EEMP	NTR	Temporary nasal crusting, Transient facial numbness	41
17	20/F	Osteosarcoma	AITF, PPF, NC	EEMP	STR	Temporary nasal crusting	21
18	30/M	Angiofibroma	PPF, NC	Combined	GTR	Transient facial numbness	4
19	21/M	Angiofibroma	AITF, PPF	Combined	GTR		11
20	21/M	Angiofibroma	AITF, PPF, NC	Combined	STR	Temporary nasal crusting, Transient facial numbness	48
21	14/M	Angiofibroma	AITF, PPF, NC	Combined	GTR	Temporary nasal crusting, Transient facial numbness	3
22	36/F	Adenoid cystic carcinoma	AITF, PPF, NC	Combined	STR		43

AITF anterior infratemporal fossa, PPF pterygopalatine fossa, NC nasal cavity, GTR gross total resection, NTR near total resection, STR subtotal resection, F/U follow-up

Minor complications such as nasal crusting and transient facial numbness occurred in nine patients with endoscopic approaches. Severe complications included post-operative meningitis in a patient with large metastatic carcinoma after transcranial surgery and visual impairment from pre-operative embolization in another case.

Discussion

Surgery in deep locations such as the AITF and the PPF presents significant challenges. Adequate exposure and ease of instrumentation are considered paramount. Few studies have explored the differences between open and endoscopic approaches for these lesions. Youssef et al. conducted a comparative anatomical study and concluded that the Endoscopic Endonasal Transpterygoid Approach provided better visualization and more direct exposure of median structures compared to the lateral open approach [23]. Similarly, Fahmy et al. performed the first volumetric analysis of these approaches to the infratemporal fossa. The excision volumes for each approach were similar, suggesting that the field of instrumentation and visualization were comparable [4]. However, there is a lack of definitive studies comparing both surgical exposure and ease of instrumentation using numerical results. Therefore, our study aims to illustrate the statistical comparison between open (ESITA) and endoscopic (EEMP) surgery in the AITF and the PPF.

Based on the results from our study, the ESITA method provides appropriate visualization and surgical freedom to address lesions in the AITF. Approaching more lateral lesion from a more lateral surgical technique offers direct exposure and creates a larger corridor, facilitating the dissection of lesions with ease and accuracy. Conversely, lesions in the PPF are better treated through the EEMP, which provides a more direct approach to median structures in the PPF. These findings suggest that ESITA and EEMP each offer more direct exposure and better visualization of the AITF and PPF, respectively.

Despite the different outcomes for surgical exposure in these two areas, the ease of instrumentation appears to be identical. The larger external opening from a craniotomy provides a benefit to the ESITA approach. Surgeon maneuverability is more flexible when performing ESITA compared to EEMP in all targeted areas of interest. However, the advantage of EEMP is that it avoids brain retraction and provides panoramic visualization in the surgical field.

The consideration of loss of functioning tissue is crucial in selecting proper surgical approaches. When approaching the AITF via ESITA, there is less tissue loss, such as the temporal base only. More medial structures, such as the pterygopalatine ganglion and its branches, vidian nerve, and pterygoid bone and muscles, could be preserved. In contrast, EEMP causes less tissue loss for lesions in the PPF, with only partial pterygoid bone removal, while more lateral structures, including the temporal base and branches of CN V3, are not involved. Our results suggest that more medial and lateral locations like the PPF and AITF should be accessed from more medial and lateral approaches like EEMP and ESITA, respectively, to preserve functioning tissue as much as possible.

The EEMP approach for accessing lateral lesions in AITF involves extensive anterior septectomy, leading to possible septal perforation and loss of nasal tip support [16]. To mitigate these complications, advanced endoscopic techniques like Denker's approach, prelacrimal recess approach, and Caldwell-Luc approach have been developed. Anteromedial maxillectomy or Denker's approach could provide better surgical freedom for endoscope and instrument movement but may disrupt the medial maxillary buttress and result in loss of lateral alar support [1]. The prelacrimal recess approach can avoid nasolacrimal duct resection but offers less surgical freedom compared to Denker's approach [10]. Anterior maxillary antrostomy or Caldwell-Luc approach provide a direct route to the lesion in AITF; however, complications such as facial swelling, pain, and numbness in the face and gums may occur [11].

From our retrospective review of case series, GTR and NTR rates are similar among those surgical techniques. Moreover, we have found that the extent of resection does not depend only on surgical approaches but also on the type of pathology and tissue infiltration. Well-circumscribed benign lesions such as angiofibroma and schwannoma appear to be more resectable. Furthermore, most complications associated with the endoscopic approach are mild and transient. Conversely, more severe complications, such as temporal lobe contusion, CSF leakage, and meningitis, could occur in transcranial surgery for a complicated lesion.

Strengths and limitations

Our study has several limitations. First, the sample size of five head specimens is small. Conducting further studies using a larger number of cadaveric specimens would help validate our results. Second, as this is a cadaveric study, several factors may preclude the direct application of our results into clinical practice. Tissue consistency and the lack of bleeding should be considered. Finally, a large clinical trial conducted in multicenter institutes would emphasize the results.

However, our strength lies in the quantitative measurements using the navigation system and a volumetric analysis using CT scan to enhance accuracy.

Conclusions

To select an appropriate approach for patients with lesions in the AITF and PPF, we should consider surgical exposure, ease of instrumentation, and tissue loss. The ESITA is recommended for treating AITF lesions while the EEMP is more suitable for PPF lesions, as they provide enhanced visibility and minimize the loss of functioning tissue. The ESITA also allows for superior surgical maneuverability when compared to the EEMP for each specific target point.

Author contributions

N.J.: Manuscript writing, figure preparation, data analysis, and assistance with cadaveric dissection. U.B.: Supervision, manuscript review, and primary responsibility for cadaveric dissection.

Funding

This study was funded by Ratchadapiseksompotch Fund, Graduate Affairs, Faculty of Medicine, Chulalongkorn University, Grant number GA66/075.

Data availability

All data is provided within the manuscript.

Declarations

Ethics approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (COA No. 0098/2023, IRB No. 0846/65), and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent to participate

Not applicable. For this type of retrospective study, formal consent from individual patients was not required. A waiver of consent was granted by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University Authority to allow the inclusion of anonymized data without introducing selection bias.

Disclosures

The authors have no personal, financial, or institutional interest to disclose.

Competing interests

The authors declare no competing interests.