Onodi cell (OC) is the most common anatomical variant of posterior ethmoidal air cells, which expand over the sphenoid sinus superolaterally, with near relation to the optic nerve (ON), sphenoid sinuses, and even the internal carotid artery (ICA) [1, 2]

Recognizing these cells is essential before performing endoscopic sinus surgery or skull base surgery, because of their close relationship to ON and ICA, which can result in catastrophic complications. The OC must be distinguished from the overriding posterior ethmoid cell since they are similar anatomically and both are very related to the optic nerve canal [3, 4].

The aforementioned reasons pushed us to investigate the prevalence of OC among the Iraqi Kurdistanian population, and classify it in relation to the sphenoid sinuses, optic nerve, and recognize their patterns of pneumatization.

Adolf Ondoi, in 1904, was the first otolaryngologist who describe the presence of the cell between the skull base and sphenoid sinus (SS), pneumatized posteriorly, laterally, and superiorly to SS [4–6].

General identification of the Onodi cell involve at least one of the following descriptions: having an identifiable optic canal bulge by endoscope (42–51% of studied cadavers were discovered to have Onodi cells according to this criterion) [7], or having lateral and superior pneumatization to the sphenoid sinus and being in close relation to the optic nerve, but no prominence of the optic nerve tubercle or ICA being definitely essential [8]. The most posterior point of the Onodi cell might be extended up to 1.5 cm beyond the anterior surface of the sphenoid sinus [9]. Therefore, in case an OC is identified, the sphenoid sinus must be entered through the inferomedial floor of the OC, and endoscopic anatomy should be correlated with the CT findings [10].

Moreover, anatomical variations of the paranasal sinus (PNS) can occur not only among individuals but also within the same individual between two sides [11]. It may be difficult to differentiate between the OC and the overriding posterior ethmoid cell (PEC) if both were coexistent. However, overriding posterior ethmoid cells did not have a completely lined bony septum, whereas OC have a completely lined bone septum, which is separated from the posterior ethmoid group of air cells [4]. An overriding posterior ethmoid cell can coexist in 36.8% of Onodi-positive cell specimens [8].

Onodi cell was found in (4.1-7%) of patients with PNS anatomical variants associated with CRS OCs displace the SS downward, reducing its volume, and therefore could be associated with sphenoiditis [12–14]. The important variations occur at the sphenoethmoidal junction, and most of these variations are related to the presence of the OC and intra-sinus protrusions of the ON. Onodi cell pneumatization might reach and surround the ON in various extensions [15–17].

Methods:

This study is a retrospective cross-sectional observational study regarding the prevalence of onodi cells and their types in the Kurdish ethnicity among Iraqi people, conducted between May 2018 and March 2019 at the Department of Radiology at the Sulaymaniyah Teaching Hospital in Sulaymaniyah city, Kurdistan region, Iraq. In terms of subjects, a total of subjects are 180 patients were included, who ranged between 18–80 years old, out of which 152 cases satisfied the inclusion criteria, patients who agreed to participate in the study of both genders, Kurdish ethnicity, with no history of previous sinonasal surgery, or sinonasal tumors were included in the current report.

Axial images of paranasal sinuses were taken on a multi-slice CT scanner SOMATOM Emotion (Siemens machine). The patient position was supine with head first, and axial slices were taken in helical format by fixing the tube current at 120 kV and 35 mA, section thickness ranged from 0.6 to 1 mm, rotation time 3.4 sec, and 12 sec. window width /window level (2000l400). Multiplanar reconstruction was done from axial images in the computer by the program of the RadiAnt DICOM viewer 4.2.1(62-bit). Axial sections were done in a plane lower part parallel to the hard palate to the roof of the frontal sinuses.

In the beginning, the scans are performed in the axial planes, and then by the sagittal and coronal reconstructions. All axes were used to identify OC presence, allowing the relationship of the most posterior ethmoid cell to the optic nerve canal (separately for left and right sides) to be identified.

The classification of the most posterior ethmoid cells (PEC) is three types based on the optic nerve canal (ONC) relationship to the cells as follows: Type A: No contact observed between the wall of the most posterior ethmoid cell (PEC) and the ONC, as illustrated by a yellow asterisk in Fig. 1(A, B, C). Type B: There is contact observed between the most posterior ethmoid cells (PEC) and ONC, as shown in Fig. 1(D, E, F). Type C: the most common PEC wall being bulged by the optic nerve, or there is a dehiscence of the ONC. This bulge is defined as a protrusion of the optic nerve into the PEC. The very thin imaging slices used ensure that any bulging of the optic nerve canal is not missed, as depicted in Fig. 1 (G, H, J) [6].

Since OC was identified as a posterior-most ethmoid cell that presents either superior/or lateral to the sphenoid sinus. Their pneumatization patterns and relationship with sphenoid sinus were studied by using coronal reconstructed images, analyzed and classified into three patterns in the present study for their easy recognition and to assess their relationship with sphenoid sinus by the following methodology; A horizontal line was drawn at the uppermost part of sphenoid sinus in coronal reformatted image at the side of OC presence where maximum size of OC was seen in contact with bony sphenoid sinus at that side, and another vertical line in the axial view from the outermost border of SS is drawn anteriorly toward the most PEC. If the whole cell was seen above and medial to the horizontal line drawn in the coronal view and medial to the line in the axial plane, or totally obscuring the SS view in the axial plane, then it was labeled as a superior pattern of pneumatization, as shown in Fig. 2 (A, B). If the air cell was seen both above and below the horizontal line in coronal view and lateral to the line in axial view, it will be considered a superolateral pneumatization pattern, as depicted in Fig. 2 (C, D). Furthermore, if it was seen below the horizontal line in coronal view and lateral to the line in horizontal view, it will be considered a lateral pneumatization pattern as in Fig. 2 (E, F) [7–9].

The optic nerve relation with each pattern of pneumatization is studied separately to view its contact, bulging, and dehiscence category within the OC [16].

Statistical Analysis: All the data were entered into a Microsoft Excel 2013 sheet for analysis. Categorical variables were reported as proportions. P-value of 0.05 is considered significant, as well as 95% confidence interval (CI) values were calculated using a confidence interval calculator for proportional values.

It is worth mentioning that among the total 152 patients, there were 64 males (42%) and 88 females (58%).

The ages of the studied CT scan samples are classified into four groups according to the individual’s age as follows: Age ranged 18–30 years are 74 individuals, 31–45 years are 44 individuals, 46–60 years are 28 individuals, and finally, more than 60 years are 6 individuals. According to the above four age groups, OCs' presence or not is demonstrated for each group, with frequency of presence of 5, 13, 8, and 2 OC in the aforementioned age groups, respectively, with a significant p-value of 0.004. However, it is distributed among 10 male and 18 female patients with no significant result P-value calculated by chi-square is < 0.05. Onodi cell prevalence in the whole studied sample is calculated as 35 (11.5%), and frequency with 95% CI for positive OCs (7.9–15%).

OC-positive patients are grouped into three categories according to the sides of OC in each patient. There were (7) 25% of patients who had bilateral OCs, (21)75% unilateral OCs, the right side constituted (17) (47.6%), and the left side constituted (18) (52.3%). The chi-square statistic of the P-value is < .00001. Therefore, the result is significant at p < 0.05. The 95% Confidence interval for OC location is calculated, and was 59% − 91% for unilateral, and 9%- 41% for bilateral location.

Relationship of the most PEC to the optic nerve classified into three groups as: Type (A); No contact Type (B); Just contact, and Type (C); Contact with bulging or Deh. of the ONC, and the commonest type was Type A with 88.4%.

Optic nerve relation to the most PEC in both sides and both genders is demonstrated in Table 1, with no significant result for any, as calculated by p-value with frequency of 0.84, 0.41, and 0.67, respectively.

Onodi cells, by definition, according to their relation with ON, will be included as type B and C. Type B (just contact with ON), constitutes 27 cells (77.2%), distributed between both sides as follows: 15 cells for the right side and 12 cells for the left side, respectively. Type C (contact with bulging or dehiscence of the ONC), which constitutes 22.8%, distributed 2 cells on the right side, and 6 cells on the left side, respectively.

Onodi cell frequency according to its location relevant to SS is grouped into three categories as a whole, and for each side, with the superolateral type popularity forming 80% (28) of OC type, the superior type forming 17% (6), and lastly the lateral type forming 2% (1) of OC types. The 95% CI for OC location-SS relevant type was as follows: for superolateral (66%-93%), superior (4.7%-29%), and lateral type (0%-8.4%), respectively, as shown in Fig. 3.

Onodi cell frequency according to its location with SS for each gender and for each side is demonstrated in Table 2, with no significant result as measured by p-values.

Furthermore, since the SS sinus is considered an important anatomical landmark that navigates for critical structures and for OC, and also the critical association of the ON with OC, for that reason OC was classified in this study according to both location (side) and its relation with SS and ON. Both sides were as follows: 21, 7, 5, 1, 1 for superolateral B, superolateral C, superior B, superior C, lateral B, respectively, as demonstrated in Fig. 4.

The relation of the most common types of OC (superolateral B, superolateral C, superior B) with gender reveals no significant p-value measurement, and 95%CI is calculated for each subtype as demonstrated in Table 3. The relation of the most common types of OC (superolateral B, superolateral C, superior B) with the side reveals no significant results, as shown in the p-value measurement, as demonstrated in Table 4.

In the 28/152 patients with OCs, the associated variation of sphenoid bone pneumatization is noted in 16 sides, with anterior clinoid process (ACP) being the most common associated variation, 46.8%, then greater wing of sphenoid (GWS), 31.25%, and lastly pterygoid plate (PP), 21.8%. Associated variations (ACP, GWS, PP) with the presence of OC in relation to side reveal no significant result with a p-value of 0.73.

Identification of anatomical variations of the sphenoid sinus and posterior ethmoid sinuses in preoperative assessment of patients planned for endoscopic sinus surgery is a valuable concern for rhinologists to prevent the possible catastrophic complication due to their close relation to the optic nerve canal, and variant pneumatization of PEC and SS into the adjacent structures. This is what encourages us to investigate these anatomical relations in ethnicity Kurdistan region.

The best technology to evaluate the anatomical relationship of ONC to PEC, SS, and the detection of OC is a thin-slice CT scan. Although the Onodi cell is best identified on coronal slices by locating an air cell above the sphenoid sinus first and recognizing its posterior ethmoid air cell continuity, three-dimensional reconstruction is warranted [17].

However, in other studies, the presence of an Onodi cell was not clearly identified in CT examinations, which is in contradiction to the current study. This is mainly related to not using a 3-dimensional study [3, 4], which may not be sufficient for Onodi cell detection.

In order to perfect the identification of OC, knowing the transition from the posterior ethmoids to the sphenoid is highly required. If a series of coronal CT scans of the posterior ethmoids were studied, reformatted to axial and sagittal views, the change from posterior ethmoids to sphenoid starts by the first CT scan when visualization of the posterior bony choanae can be detected. As soon as the posterior choanae solid bony rim is seen, the directly above this solid bone the cell lying should be the sphenoid sinus. Any horizontal bony lines above this cell in this or furthermore posterior coronal CT scans, this will be OC [18].

The horizontal septum (HS) is detected on a scan that is taken through the sphenoid, which is a very helpful pointer of a sphenoethmoidal (OC). So, the change from the posterior ethmoids to the sphenoid is essential to confirm prior to the horizontal septation that is sought. This septation is formed by the sphenoethmoidal (OC) over the sphenoid, pushing the anterior face of the sphenoid sinus to a more horizontal plane. Recognizing the posterior ethmoid cells from the sphenoid sinus is easier because the party wall is easily seen on a CT section, by following the lateral ethmoid and nasal wall, with the posterior end of the middle turbinate as a landmark. This point will be constant either by endoscopy or under the microscope if the surgeon follows the posterior end of the middle turbinate as a landmark strictly [19].

In the literature review, most articles report the prevalence of OC as 8–49% [20–23]. The current study reports a compatible investigation, which is 18%. The prevalence of onodi cells varies according to differences in how these cells are defined, differences in methods of identification, detection techniques, and racial factors [11]. According to Yeoh and Tan [6], in their CT scan study of OC, it was more frequent in the Asian population regarding other races as follows: a rate of 83.3% in Asians, 73.1% in Caucasians, 57.0% in African Americans, and 62.7% in Hispanics. However, among Asian populations, it has been found that Southeast Asians (e.g, Thailand 49%) are the most who have OC than other Asian populations (e.g., Saudi 14% and Kurdish 18%).

The relationship of ON with the PEC is classified into 3 groups in terms of contact, protrusion, and dehiscence (Figs. 1) [6]. When more than half (50%) of the nerve protruded into the sinus cavity as seen in coronal CT sections, protrusion was confirmed. Dehiscence of the bone was acknowledged when no bone density separating ON from the sinus was present in the CT images. This relationship was studied in detail by Onodi more than 100 years ago, and he reported that the bony wall between the most posterior part of the ethmoid cell progresses to the lesser wing of the sphenoid, and that the bony wall of the optic canal is thin in this region, which is identical to group 3 [24].

Delano classifies OC into 4 types as follows: Type 1 ONC course close to the sphenoid sinus without indentation of the wall (76%). Type 2 nerves course near the sphenoidal sinus, causing indentation of the sinus wall (15%). Type 3 nerves course through the sphenoid sinus (6%). Type 4 nerves course immediately adjacent to the sphenoidal sinus and the posterior ethmoidal air cell (3%) [8]. However, in this study, we classify onodi cells depending on their relation to the optic nerve and to the sphenoid sinus, and further merge both in one classification, following the concept that Excessive aeration of the sphenoid and surrounding structures can lead to the indentation of ONC into the sphenoid sinus and bony dehiscence. Increased pneumatization of posterior ethmoidal air cells may form part of the boundary of ONC and may predispose to injury during Functional Endoscopic Sinus Surgery (FESS) [3]. Furthermore, it is easier to interpret radiologically, and more practical during endoscopic sinus surgery, and it can be classified into three types: Superior to the sphenoid sinus, Lateral to the sphenoid sinus, and Superolateral to the sphenoid sinus [7–9].

In this study, it is noticed that the superolateral to sphenoid sinus type is the most common OC type, there was 28 (80%) superolateral onodi cells of total 35 onodi cells on both sides with CI 95% (66%-93%), which is in harmony with other investigations reported by Chmielik et al [12] was 39.8%, and Pornthep Kasemsiri, et al [25], where describe superolateral type as the most common type with 53%, then superior type with 46% and lateral type as the least type with 1% which is coincide with the current report.

So that, we found in our study that onodi cell superolateral to sphenoid sinus with only just contact to optic nerve (superolateral-B) is the most common subtype with 21 onodi cell (60%) among the 35 total onodi cells on both sides, CI 95% (43%-76%) followed by superolateral to sphenoid sinus with either bulging or dehiscence of the ONC (superolateral-C) of 7 onodi cells (20%), CI 95% (6.7–33%) followed by onodi cell superior to sphenoid sinus with just contact to optic nerve (superior-B) constitute 5 cells (14.2%), CI 95% (2.7%-25%) then subtypes superior to sphenoid with either bulging or dehiscence (superior C) and lateral to sphenoid with only contact to optic nerve (lateral B) come in the same frequency for 1 cell for each subtype (2.58%) with CI 95% (0–8%).

The aforementioned data can support the statement of excessive pneumatization of SS, and PEC can lead to indentation of ONC into SS and bony dehiscence, and may predispose to ON injury, FESS. The size of the sphenoid sinus is also taken into consideration in endoscopic surgeries, as the surgeons should be cautious, as high sphenoid sinus volume is a predisposing factor for optic nerve injury.

In the 28 patients with OC, there were 16 patients (57.1%) with other associated variations and 12 patients (42.8%) with no associated variation. Our study showed ACP is the most common associated variation of 46.8%, followed by GWS 31.25% and then PP 21.87%. Intra-sphenoid ON course is limited superiorly by ACP pneumatization, which may displace the ON more medially, making it vulnerable to injury during FESS. ACP forms a small space on the lateral wall of the sphenoid sinus between the optic canal superiorly and the carotid prominence inferiorly, which is known as the opticocarotid recess. This may result in ON protrusion into the sphenoid sinus [26, 27]. The current findings give us a robust impression of the coincidence of hyperpneumatized SS with the prevalence of OC with associated contact with the ON course.

Limitations: There are limitations regarding the restricted number of individuals’ numbers which might affect the conclusion of a robust statistical prevalence. Furthermore, the optional selection of non-diseased sphenoid sinus is warranted, as the inflammatory process usually affects the sinus anatomy through osteitis change, which makes the bony margin thicker, then might affect the relation of the sphenoid sinus and posterior ethmoid sinus to the optic nerve, whether dehiscent or not.

Sphenoid sinus anatomy gives high variability through its anatomical variation, where its recognition is paramount for sinus–skull endoscopic surgeons and radiologists to avoid unwanted complications. Therefore, explore their possible presence in each patient, especially for what concerns the close relationship with neurovascular elements is highly demanded. The current study observations point out the importance of careful preoperative CT analyses in preparation for surgery of the sphenoid sinus area, and of delineating a specific roadmap for each patient. Preoperative identification of the OC and its ONC relation is essential. In this study, we report the most significant superolateral PEC or OC with ONC contact as a more common variant in the included individuals, which reflects the significance of SS and PEC pneumatization.