A Novel International Endoscopic Sphenoid Surgery Classification (IESSC): A Delphi Consensus

ABSTRACT

Background

Advancements in endoscopic sinus and skull base surgery created a need for standardized terminology to describe sphenoid sinus surgery. Although classification systems exist for other sinuses, one for endoscopic sphenoid sinus surgery is lacking. Developing such a system would standardize procedure descriptions and promote a common language among surgeons. This study aimed to develop a new classification system for endoscopic sphenoid surgery.

Methods

Consensus on a novel endoscopic sphenoid surgery classification system by running the Delphi procedure with 16 rhinology experts from around the world.

Results

Four Delphi rounds were required to reach a consensus on all stages of the classification. The average percentage of agreement on the stages of classification progressively increased from 70.83% in the first round to 87.68% in the last round. The rejection rates continuously decreased from 8.81% in the first round to 4.44% in the last round. The classification system was developed as follows: stage 1, presphenoid surgery; stage 2A, partial sphenoidotomy; stage 2B, complete sphenoidotomy; stage 2C, transpterygoid sphenoidotomy; stage 3A, Rostral sphenoidectomy; and stage 3B, extended sphenoid drill‐out.

Conclusions

This novel endoscopic sphenoid surgery classification system facilitates the description of different sphenoid sinus procedures, providing surgeons with better opportunities for discussion and communication.

1. Introduction

Endoscopic sinus surgery (ESS) and skull base surgery have advanced considerably, owing to improvements in imaging technology, endoscopic equipment, and surgical techniques. These procedures allow for minimally invasive procedures with enhanced precision. Technologies such as navigation systems and high‐definition imaging have increased the accuracy and safety of accessing the sinonasal and skull base regions, thereby reducing complications. These advancements have established endoscopic endonasal surgery as the standard of care for managing chronic rhinosinusitis, nasal polyps, and appropriately selected complex skull base tumors [1, 2].

Being the most challenging sinus, the frontal sinus has received the most attention. Kuhn (1994) was the first to classify the different cells around the frontal sinus [3]. He described an integrated approach to the frontal sinuses, applying the concept “to choose the least invasive procedure with which the surgeon is comfortable and which will accomplish the task. Additional procedures then may be employed as the situation dictates” [4]. Draf developed three degrees of frontal sinus opening: (1) simple drainage (type I), (2) extended drainage (type II), and (3) endonasal median drainage (type III) [5]. Recently, the International Frontal Sinus Anatomy Classification and the Classification of the Extent of Endoscopic Frontal Sinus Surgery were published based on a global consensus of numerous rhinologists [6].

Similarly, the sphenoid sinus and its anatomical variations have gained increasing attention over the last decade. Numerous anatomical studies have categorized its variable pneumatization and relationships to surrounding vital structures [7, 8, 9, 10]. In addition to its role in sinus diseases, the sphenoid sinus is often part of the main corridor for endoscopic skull base surgery, raising its significance for both otolaryngologists and neurosurgeons.

Sphenoid sinus surgery has evolved over recent decades from open approaches to transnasal microscopic and presently endonasal endoscopic approaches [11]. Adopting the microscope for transsphenoidal approaches to remove pituitary microadenomas was a major advancement brought by Hardy in 1965 [12]. After the rod‐lens endoscope was introduced by Hopkins in 1950, it gradually became a preferred tool of visualization for sinus and transsphenoidal surgeries [13]. Currently, endoscopic endonasal transsphenoidal surgery is considered by many as the standard of care for pathologies such as pituitary adenoma, planum meningioma, craniopharyngioma, and clival chordoma [13].

Having common definitions facilitates communication among physicians and improves teaching and illustrations. Despite various anatomical classifications of the sphenoid sinus, a classification indicating the surgical extent is still lacking. This study aimed to establish a consensus regarding a novel surgical grading system for the sphenoid sinuses.

2. Materials and Methods

2.1. Study Design

This study was designed to develop a new classification grading system for endoscopic sphenoid sinus surgery. The study group consisted of a steering committee and an international consensus expert group. All members of both groups contributed to the authorship of this study. A list of 25 internationally recognized experts in rhinology and skull base surgery was identified by the steering committee, ensuring broad geographical representation across different continents. Selection was based on academic contributions, clinical expertise, and prior involvement in classification systems or consensus‐building efforts. To ensure objectivity and balance, an independent vote was conducted among the four steering committee members, resulting in the final selection of 16 experts. They were invited to reach at least 12 respondents, the minimum required to perform the Delphi procedure and achieve consensus [14].

2.2. Proposed Classification

The steering committee (H. B., A. R., A. Q., and S. S.) conducted a thorough literature review to identify all described endoscopic approaches to the sphenoid sinus. A preliminary classification system was designed after discussing these techniques in five separate meetings with international advisors (A. J., P. C., and R. C.). The classification consisted of six stages: Stage 1, stage 2A, 2B, 2C, and stage 3A, 3B.

2.3. Process of Achieving Consensus

All members of the international consensus group were asked to complete a questionnaire sent via email, using the Delphi method to reach a consensus on the classification. The Google Forms questionnaire, designed by the steering committee, was sent via email anonymously to each member of the consensus group to avoid bias. Each participant was asked to evaluate each stage of the classification from two perspectives: the title and description, in addition to the question on resecting the superior turbinate to the stages, and another question on the appropriate sequence of the classification system. Each item of the classification was scored on a Likert scale from 1 (strongly disagree) to 7 (strongly agree). In addition, participants could also provide free‐text comments or advice. At the start of each round, the consensus group had three weeks to answer the questionnaire and send it back. The steering committee was blinded to responses to avoid bias. Based on this feedback, the classification was modified and revised for participants’ re‐evaluation. Items reaching consensus were excluded from the next round. Participants were informed of the results from each round before proceeding to the next, ensuring transparency in decision‐making. Once a stage reached the predefined consensus threshold, it was considered stable, and subsequent rounds focused on refining areas still under discussion. This approach maintains the integrity of the consensus process by preventing unnecessary repetition and maintaining forward progress. Iterations were performed until a consensus was achieved (Figure 1).

FIGURE 1.

Flow diagram showing the process of reaching consensus.

2.4. Criteria for Achieving Consensus

Consensus definitions in Delphi studies vary, but a recent systematic review of the Delphi methodology identified a ≥80% agreement threshold as the most commonly used criterion for establishing consensus. This threshold was chosen in our study to ensure a high level of agreement among international experts while maintaining methodological rigor. Additionally, our definition aligns with established Delphi best practices, where thresholds between 70–80% are widely accepted for consensus in healthcare‐related classifications [15].

In this study, consensus was achieved when the mean agreement score for each stage was ≥6. Additionally, 80% needed to fall within the two upper categories (6: agree or 7: strongly agree). No more than 10% of responses could be outliers in the opposite direction (3: somewhat disagree, 2: disagree, or 1: strongly disagree). Only one panel member was allowed to skip the questionnaire [15].

2.5. Data Analyses

Data were analyzed using Microsoft Excel 365 with standard descriptive statistics (mean, standard deviation [SD], median, and range). Agreement and disagreement percentages were calculated for each stage. Each stage was labeled as passing or not passing a round. The stages that reached consensus were excluded from the next Delphi round. The steering committee continuously reviewed the classification sequence to avoid any gaps.

3. Results

Sixteen international rhinology experts from 12 countries were invited to participate, and all agreed to be members of the consensus group. Four Delphi rounds were required to reach a consensus on all stages of classification. All the participants completed the four rounds, except for one member who dropped out in the last round. The scoring and progress processes for reaching consensus at each stage are listed in Table 1. Of the 12 items, two reached consensuses in the first round, four in the second, three in the third, and three in the fourth. The average percentage of agreement on the stages of classification progressively increased from 70.83% in the first round to 87.68% in the last round. On the contrary, rejection rates continuously decreased from 8.81% in the first round to 4.44% in the last round (Figure 2). All the participants agreed on the manner of the 6 stages however, the question of adding resection of the superior turbinate to the stages did not reach a consensus. The classification numbering is based on the sequential opening of the sphenoid sinus, where stage 1 is to approach the sinus without widening the ostium, stage 2 is widening the ostium in different degrees A–C but without connecting both sides together and stage 3 is when we connect both sides of the sinuses together in different degrees. In this classification, the term “widening” refers to an increase in the size of the ostium but not reaching total removal of the anterior sphenoidal wall.’’ Complete removal’’ is referred to as completely taking out the structure, for example, complete removal of the anterior sphenoid sinus wall. The final classification system is presented in Table 2, with all stages of the classification illustrated in Figure 3.

TABLE 1.

Agreement scores for items in all four rounds of Delphi.

Stage	Mean ± SD	Median	Range	Agree (%)	Outliers (%)
First round
1: Title	5.6 ± 1.3	6	3–7	68.75	6.25
1: Description	5.1 ± 1.8	6	2–7	62.5	25
2A: Title a	6.0 ± 0.9	6	3–7	81.25	6.25
2A: Description	5 ± 1.8	6	1–7	56.25	25
2B: Title	5.7 ± 1.6	6	2–7	68.75	12.5
2B: Description b	6.1 ± 0.8	6	5–7	75	0
2C: Title	5.5 ± 1.3	6	2–7	56.25	6.25
2C: Description	5.8 ± 1.1	6	4–7	68.75	0
3A: Title	5.5 ± 1.7	6	1–7	68.75	12.5
3A: Description	5.9 ± 0.8	6	4–7	75	0
3B: Title	5.8 ± 1.6	6	2–7	81.25	12.5
3B: Description a	6.2 ± 0.6	6	5–7	87.5	0
Second round
1: Title a	6.1 ± 1.5	7	1–7	81.25	6.25
1: Description a	6.3 ± 1.5	7	1–7	87.5	6.25
2A: Description a	6.2 ± 1.5	7	1–7	87.5	6.25
2B: Title b	6.0 ± 1.7	7	1–7	75	12.5
2B: Description a	6.4 ± 1.0	7	3–7	87.5	6.25
2C: Title	5.1 ± 2.0	6	1–7	62.5	18.75
2C: Description	5.9 ± 1.1	6	4–7	75	0
3A: Title	5.9 ± 1.8	7	1–7	81.25	12.5
3A: Description b	6.0 ± 1.4	7	3–7	81.25	12.5
3B: Title	6.0 ± 1.5	7	2–7	81.25	12.5
Third round
2B: Title a	6.5 ± 0.8	7	4–7	87.5	0
2C: Title a	6.2 ± 0.6	6	5–7	87.5	0
2C: Description	5.9 ± 1.2	6	3–7	81.25	12.5
3A: Title	5.9 ± 1.4	7	3–7	68.75	12.5
3A: Description b	6.1 ± 1.2	7	3–7	75	6.25
3B: Title a	6.6 ± 0.6	7	5–7	93.75	0
Fourth round
2C: Description a	6.4 ± 0.9	7	4–7	86.67	0
3A: Title a	6.2 ± 1.0	6	3–7	86.67	6.67
3A: Description a	6.1 ± 1.1	6	3–7	86.67	6.67

^a Reached full criteria for consensus.

^b Reached mean score criteria for consensus.

FIGURE 2.

Average percentage of agreement and disagreement plotted against the Delphi rounds.

TABLE 2.

The sphenoid sinus surgery classification system.

#	Type	Description
1	Presphenoid surgery	Clearance of the sphenoethmoidal recess and/or posterior ethmoid cells without widening the sphenoid ostium
2A	Partial sphenoidotomy	Partial widening of the sphenoid ostium
2B	Complete sphenoidotomy	Complete removal of the anterior sphenoid sinus wall
2C	Transpterygoid sphenoidotomy	Partial or complete removal of the anterior sphenoid sinus wall with partial resection of the pterygoid process
3A	Rostral sphenoidectomy (RS)	Complete removal of bilateral anterior sphenoid sinus walls, rostrum, and intersinus septum
3B	Extended sphenoid drill‐out	Complete removal of bilateral anterior sphenoid sinus walls, rostrum, intersinus septum, and sphenoid sinus floor

FIGURE 3.

Illustration showing the stages of the final endoscopic sinus surgery classification system. ASW, anterior sphenoid wall; CR, clival recess; ET, eustachian tube; IT, inferior turbinate; LR, lateral recess of sphenoid sinus; MT, middle turbinate; MS, maxillary sinus; NP, nasopharynx; NS, nasal septum; SO, sphenoid ostium; SS, sphenoid sinus; SB, skull base.

3.1. Type 1: Presphenoid Surgery

This procedure can be performed by clearing any pathology in the sphenoethmoidal recess and/or posterior ethmoid cells without widening the sphenoid ostium. Clearance of the posterior ethmoid was added to the description, as most authors agree that presphenoid clearance can be performed through a transnasal or transethmoid approach. The same description was found in a previous classification by Simmen and Jones for type 1 sphenoidotomy [16].

3.2. Type 2A: Partial Sphenoidotomy

This stage included any degree of sphenoid ostium widening that did not completely remove the anterior sphenoidal wall. Different preferences exist among authors regarding technique and direction. Therefore, a consensus was achieved on the partial widening of the sphenoid ostium without determining size and direction.

3.3. Type 2B: Complete Sphenoidotomy

The term “complete sphenoidotomy” was preferred by the authors to total sphenoidotomy. Moreover, considering that this classification describes an endoscopic staging system, the term “removal” was favored over “resection.” This stage describes the complete removal of the anterior sphenoid sinus wall, similar to type 3 sphenoidotomy, in which the opening extends from the skull base superiorly to the sphenoid floor inferiorly and from the nasal septum medially to the orbit laterally.

3.4. Type 2C: Transpterygoid Sphenoidotomy

The pterygoid process projects inferiorly from the junction of the body and the greater wing of the sphenoid bone and bifurcates into medial pterygoid and lateral pterygoid plates. Bolger first described this technique in 1999 to repair an encephalocele caused by a skull base defect in the lateral recess of the sphenoid sinus [17]. The procedure primarily involved wide maxillary antrostomy, ethmoidectomy, and complete sphenoidotomy, followed by removing the medial superior posterior maxillary wall to expose the pterygoid process for removal, allowing wider access to the lateral sphenoid recess. By removing the posterior maxillary wall, the pterygopalatine fossa (PPF) will be exposed. The sphenopalatine artery is encountered first when approaching the PPF from the nasal cavity, while the vidian nerve is more posterior and lateral. After cutting the SPA, the pterygoid process can be exposed by dissecting the PPF laterally. Then, the pterygoid process is drilled to access the lateral sphenoid recess, keeping the vidian nerve inferiomedial and maxillary nerve V2 superolateral. The terms “transethmoidal–pterygoidal–sphenoidal approach” and retrograde transsphenoidal approach have been used in the literature to describe this. [18, 19]. The term “pterygoid process” was used instead of “pterygoid wedge,” as the latter is not a recognized anatomical term.

3.5. Type 3A: Rostral Sphenoidectomy

The sphenoid rostrum is a triangular bony spine that protrudes from the lower surface of the body of the sphenoid bone. The narrow edge of the spine fits into the fissure between the two alae of the vomer bone. It subsequently extends forward to join the sphenoidal crest on the front surface of the body of the sphenoid bone [20]. This stage was named after the sphenoid rostrum because it was the main structure that was removed to communicate with both sphenoid sinus cavities. The name is preferred over “sphenoid drillout” which is what it is described as in the literature [21]; the procedure can be performed without a drill. It is commonly used to access the anterior skull base during endonasal endoscopic transsphenoidal approaches and in other conditions, such as refractory chronic sphenoid sinusitis [21, 22, 23, 24]. This stage is achieved via the complete removal of the bilateral anterior sphenoid sinus walls, rostrum, and intersinus septum.

3.6. Type 3B: Extended Sphenoid Drill‐Out

This stage is achieved by completing a 3A and then drilling out the sphenoid floor to achieve an exposure extending from the planum sphenoidale superiorly to the nasopharynx inferiorly. This is referred to as “radical sphenoidectomy” or “sphenoid marsupialization” in some publications [25, 26]. With the advancement of endoscopic skull base surgery, this procedure is applicable for persistent sphenoid sinusitis, resecting benign or malignant neoplasms involving the sphenoid sinus or nasopharynx, and extended approaches to the middle and posterior cranial fossae [27, 28].

4. Discussion

Since the development of endoscopic sinus surgery, several techniques for accessing the sphenoid sinus have been described. These techniques/methods are scattered throughout the literature without a clear, common, agreed‐upon classification that describes and stages the different techniques. Simmen and Jones proposed a sphenoidotomy classification with three types of sphenoidotomies: [16] Type 1, only identified the sphenoid ostium without further instrumentation; type 2 was opening the sphenoid inferiorly to half its height and upward to the skull base; and type 3, extending the opening to the sinus floor and laterally to vital structures. However, this classification is limited and not applicable to extended endonasal skull base approaches. Table 3 compares this classification system with the existing system. Other authors have described individual sphenoid sinus procedures (e.g., drillout and marsupialization). However, a stepwise approach or classification is lacking [21, 25].

TABLE 3.

Comparison between the old classification system and the new system.

New (IESSC) classification	Simmen and Jones classification
The title of each stage describes the procedure	The title is only by number, that is, sphenoidal sinusotomy i, ii, iii
It uses the terms (partial and complete sphenoidotomy) without mentioning the direction of widening. This keeps it more flexible and fit to all maneuvers of doing sphenoidotomy.	There are many ways to perform sphenoid sinusotomy that do not fit into this classification. e.g., opening the sphenoid inferiorly only, etc.
Describes transpterygoid stage to deal with pathologies in the lateral sphenoid recess	Cannot deal with pathologies in the lateral sphenoid recess
Describe expanded sphenoidotomy techniques to deal with extensive inflammatory disease or to approach skull base pathologies	Do not have stages describing extended sphenoidotomy

Establishing a unified classification system for endoscopic sphenoid sinus surgery is crucial for standardizing surgical procedures, ensuring consistent reporting of research outcomes, and improving patient care by facilitating better surgical planning and postoperative evaluation.

4.1. Strengths and Limitations

This novel endoscopic sphenoid classification overcomes previous limitations by including stepwise increasingly complex techniques for treating different inflammatory or neoplastic sphenoid and/or skull base pathologies. This classification was designed using a standardized Delphi technique, incorporating the opinions of international experts from different schools, enhancing the classification's strength.

Our proposed classification underwent internal validation through a rigorous Delphi consensus method, ensuring expert agreement and practical applicability. Regarding external validation, a direct comparison with existing systems, such as those by Simmen and Jones, is not feasible, as these classifications have not themselves undergone validation and remain proposed frameworks by individual authors rather than widely adopted standards. Given this limitation, we believe that future clinical studies assessing real‐world applicability, reproducibility, and interobserver reliability will provide the most meaningful external validation for our system.

Consensus was reached on all stages of the classification, making it more reliable for clinical practice. However, some aspects of the classification, such as partial or complete resection of the superior turbinate at stage 2, encountered some resistance in reaching full agreement. Numerous items were omitted to simplify the classification during the conceptualization period. For example, approaches to the sphenoid sinus, whether transseptal, transnasal, or transethmoid, were omitted as they will add another layer of complexity and may lower the chances of achieving consensus and worldwide adoption (Table 4). The present classification looks over the extent (end‐result) of sphenoid sinus surgery opening for both neoplastic and inflammatory, regardless of the surgical approach. However, both transethmoid and transnasal approaches can be used to reach the targeted stage. It also does not look for the type of pneumatization within or around the sphenoid sinus. While factors such as sphenoethmoidal air cells and pneumatization types may influence the surgical approach, they do not alter the staging of the surgical extent, which remains the primary focus of this classification. The main objective of ESS in chronic rhinosinusitis is to preserve mucociliary function and ventilation of the paranasal sinuses, regardless of the performance of the bone. It also aims to facilitate the installation of topical treatments. Recent classification has been published including the “mucosal concept” using the Lamella Ostium Extent Mucosa system [29]. Finally, considering that the participants were from different countries, conducting online meetings to discuss controversies or comments was technically challenging, owing to time differences (Table 4).

TABLE 4.

Detailed descriptions that did not reach consensus.

Number	Item description
1	Adding partial or complete resection of superior turbinate to stage 2A, B, C
2	Adding the direction of widening to the description of partial Sphenoidotomy.
3	Adding the approaches to the sphenoid sinus, whether transseptal, transnasal, or transethmoid, to the classification
4	Adding the sphenopalatine artery and vidian nerve in the description of transpterygoid sphenoidotomy

This novel classification system for endoscopic sphenoid surgery was developed through consensus among global rhinology experts. The idea of this project is to format reliable classification through a consensus from the Delphi method that can be the base for future work. It aimed to establish a standardized language for endoscopic sphenoid sinus surgery to facilitate communication among surgeons.

Conflicts of Interest

The authors declare no conflicts of interest.