Sphenoethmoid Cell: The Battle for Places Inside of the Nose Between a Posterior Ethmoid Cell and Sphenoid Sinus: 3D-Volumetric Quantification.

BACKGROUND: Sphenoethmoid cells may be above the sphenoid sinus with/ or without con-tact to optical nerve. Although sphenoethmoid cells are theoretically considered to possibly influence the sphenoid sinus volume, we could not find any study in the literature on this issue. AIMS: The aim of our study was to detect sphenoethmoid cells and measure the sphenoid sinus vol-ume using multiplanar computerized tomography and also investigate the correlation between the presence of sphenoethmoid cells and the sphenoid sinus volume.
METHODS: Retrospectively 141 patients who had available paranasal computerized tomography images were included in this study. The sphenoid sinus volumes of each patient were calculated individually for each side, and the relationship between the presence of sphenoethmoid cell and sphenoid sinus volume was investigated.
RESULTS: Sphenoethmoid cells were detected at 106 (37.5%) of the total 282 sides in 141 patients. No gender difference was observed. The total sphenoid sinus volume was significantly lower in the group of patients who had bilateral sphenoethmoid cells than in the sphenoethmoid cell negative group. In patients with a unilateral sphenoethmoid cell, a significant decrease in the sphenoid sinus volume was observed only for the side where the sphenoethmoid cell was located.
CONCLUSION: It was observed that the sphenoethmoid cells caused a significant reduction in the sphe-noid sinus volume on the side where they were located. In the case of low sphenoid sinus aeration, the sphenoethmoid cell should be kept in mind. Further studies with an extended patient series are required to explore this issue.

INTRODUCTION

The sphenoethmoid air cells (SEC) (formerly known as “Onodi cells”) are located in a superior and lateral position to the sphenoid sinus (SS) [1]. The prevalence of sphenoethmoid cells is reported as 8-65.3% [2-4]. The SEC are also closely associated with the SS, the optic nerve (ON), and the internal carotid artery (ICA) [2, 5, 6].

The SS, which is deeply seated in the skull and surrounded by vital structures such as the ICA and the ON, is considered the most inaccessible paranasal sinus [7]. To prevent serious complications, such as injuries to these vital structures during transsphenoidal and endoscopic sinus surgery, a comprehensive knowledge of the variable anatomy of the SS is necessary [8, 9].

The paranasal sinuses begin their development from very early stage in utero and show very different features in shape and size [10]. While ethmoid sinuses originate from invagination of the lateral nasal wall, the SS originates from the posterior invagination of the nasal capsule. The posterior ethmoid cells and SS are pneumatized synchronously, and they exhibit major volumetric growth during the same period of life. It is natural to expect the SS and the SEC to have distinct pneumatization characteristics because these two structures have separate embryological origins. Possible inter-individual volumetric discrepancies and factors that determine the pneumatization patterns in the SS, posterior ethmoid cells, and particularly the SEC are yet to be discovered. Nomura et al. [11] stated that the SEC displaces the SS downward and reduces its volume. In general, the SEC is accepted to be an obstacle especially during the endoscopic transsphenoidal sellar surgery and must be opened for safety of the surgery and accessibility to the sellar region [6, 12].

Initial knowledge of human paranasal sinus pneumatization was obtained by anatomical measurements, injecting various materials into cadavers or performing plain radiography. Currently, computerized tomography (CT) and magnetic resonance imaging (MRI) provide more precise information and allow a more accurate assessment of the region. Computerized tomography is a gold standard tool for analyzing the sphenoid sinus and its surrounding structures [13]. It should be noted that during CT evaluation, all three dimensions (axial, coronal and sagittal) should be examined for proper identification of the SEC [5].

Several studies have evaluated the paranasal sinus volume using CT scans [14-16]. However, to the best of our knowledge, the correlation between the presence of SEC and the quantitative measurement of SS volume has not been reported. In this study, we aimed to detect the SEC and to measure SS volumes quantitatively using a high resolution multiplanar (axial, coronal and sagittal) CT scan and to analyze the correlations between the presence of the SEC and the SS volume.

MATERIALS AND METHODS

Following the Ethics Committee approval, retrospective data of 141 patients (80 males and 61 females) aged 18 years or older who underwent paranasal CT scans due to chronic sinusitis between March 2014 and March 2015, were screened and included in this study. Patients with traumatic signs or histories, malignancy, congenital malformation, or a history of endoscopic sinus surgery were excluded. The presence of SEC was evaluated in each CT scan, and patients were categorized into four groups: Group I: SEC negative (Control group); Group II: Bilateral SEC positive, Group III: Right SEC positive, Group IV: Left SEC positive. The relationship between the SEC and the SS volume was investigated according to the SS volume of each patient calculated individually for each side.

Image Acquisition

Routine paranasal CT scans were performed using a 128-slice multi-detector CT scanner (Ingenuity CT, Philips Healthcare, Andover, MA, USA). The exposure settings were 120 kV and 160 mA with a rotation time of 0.5 SECc and a collimation of 64 x 0.625. Axial images with 0.6 mm thickness were obtained, and coronal and sagittal CT scan images were 0.9 mm thick. All the images were sent to workstation (Syngo. via Work Station, Erlangen, Germany). The sphenoid sinus borders were drawn by the radiologist in each section of the sphenoid sinus area of interest. This process was done separately for each side. When all the sections of the sphenoid sinus area of interest were finished, the computer converted all of the obtained sections to volume measurement and obtained a sphenoid sinus volume using the volume programme of Syngo. via work station.

Statistical Analyses

The Shapiro-Wilk test was used to test normality, and the Levene test was used to test homogeneity of variance. An independent T test and one-way ANOVA (Robust Test: Brown- Forsythe) were used for the comparisons of two independent groups and multiple groups. Post-hoc analyses were performed with Fisher’s least significant difference (LSD) test. Quantitative variables are presented as the means ±SD (standard deviation), whereas numbers (n) and frequencies (%) are used to present categorical variables. A confidence level of 95% was adopted for the analyses, and a p value less than 0.05, indicated statistical significance. Statistical analyses were performed using SPSS 22.0 (IBM Corporation, Armonk, New York, USA).

RESULTS

The mean (±SD) age of the study population was 35.5 (±13.2) years, and the age of the patients ranged from 18 to 68 years. Eighty of 141 patients were female. The frequency of female patients was higher (56.7%); however, no significant difference was observed between the patient groups in terms of age (p=0.143) or gender (p=1.000).

Sphenoethmoid cells were detected at 106 (37.5%) of the total 282 sides in 141 patients (Table ). Among the SEC positive patients, the frequencies of bilateral, right side and left side SEC positive patients were 47.2%, 25.0% and 27.8%, respectively. The presence of bilateral SEC was more common (~47%) than unilateral existence in both genders. The existence of unilateral SEC was distributed similarly (25% vs. 28%) for each side in both genders (Table ).

In the absence of an SEC (Group I, Control), the mean total SS volume of the patients was 15.1 cm3, and the mean volume of each side was comparable (7.8 cm3 for the right side and 7.3 cm3 for the left side). The total SS volume was significantly lower in the group of patients who had bilateral SEC (Group II) than in the SEC-negative group (Group I, Control). However, in cases where SECs were present unilaterally (Group III and IV), a significant decrease in the SS volume was observed only for the side in which the SEC was present (Fig. , Table ). The volumes of each sides of the SEC-negative group in our study were compared with the previous studies (Table ). In addition, previous studies of authors, study modality, number of subjects and/or performed technique, sphenoethmoid cell prevalence were presented at Table .

DISCUSSION

Endonasal endoscopic sinus surgery may be required for indicated paranasal sinus diseases. In addition, transsphenoidal endonasal endoscopic surgery is preferred for the management of pituitary lesions [17]. Because the SS is near several vital structures; i.e. the ON, CA, and vidian nerve, complications in the interventions performed on this sinus may be more dangerous than those in other sinuses. Therefore, preoperative in-depth assessment of anatomical variations of the SS is extremely vital to prevent damage to these important and close structures.

Some anatomical cell variations may be present near the paranasal sinuses. The SEC is one of the cell variations around the sphenoid sinus and is known as a sphenoethmoidal cell. Săndulescu et al. [18] stated that important variations may be present at the sphenoethmoidal junction, and most of these are associated with the presence of an SEC and intrasinusal bulgings of the ON. Ozturan et al. [19] stated that the SEC aeration may achieve and enclose the ON in various extensions. The most posterior ethmoid cell is thought to be an obstacle for the endoscopic endonasal transsphenoidal sellar surgery and should be opened for exact view of the surgery and accessibility to the sellar region [6, 12].

Radiological techniques such as CT and MRI are widely used to improve diagnosis and preoperative assessments in the sphenoethmoidal region. The results suggested that three planes (axial, coronal and sagittal) for CT examinations should always be used in sinus CT examinations for proper detection and definition of the SEC [6].

Various prevalence rates have been reported for the SEC ranging from 8% to 65.3% [2-4]. With the evolution of radiological techniques, new CT studies have also been emerged. Technically, the inspection of all three dimensions (axial, coronal and sagittal) of CT scans has also been possible to detect the SEC. In this study, the prevalence of SEC was found to be 37.5%. The prevalence of SEC in this study is comparable with the previous reported prevalence rates of CT study by Hwang et al. [5], Nomura et al. [11], and Chmielik et al. [20]. The results of this study revealed higher prevalence of SEC than the CT studies of Al-Abri et al. [21], Leunig et al. [22] and Pérez-Pińas et al. [23], and our results were also lower than the CT reports of Kasemsiri et al. [24], Wada et al. [25] and Tomovic et al. [4]. In addition, in a CT study involving the patients with both rhinosinusitis and healthy volunteers, the SEC prevalence was found to be 9% in healthy volunteers and 7% in the patients with rhinosinusitis [26]. Data from healthy volunteers were not obtained in this study. In addition, Thanaviratananich et al. [2], Yeoh et al. [27] and Kainz and Stammberger [28] studied the prevalence of sphenoethmoid cells in cadavers and found 60%, 50.98%, and 42%, respectively. For comparing various studies on study modality, number of subjects, performed technique, including criteria and the prevalence of sphenoethmoid cell are presented in Table . In the light of the above data, regarding the inclusion criteria, studies on SEC prevalence have been performed either on chronic sinus disease as this study or on randomly paranasal sinus CTs or cadaveric specimens. The prevalence of SEC in our study was performed on chronic sinus disease and our results were compared with the CT studies of Nomura et al. [11] and Wada et al. [25]. This study's results were also higher than healthy individuals in the CT study (Table ). This may be caused by the inclusion criteria kept higher than the prevalence values obtained from healthy volunteers. Interestingly, studies performed on cadavers were also found to be higher than the prevalence of CT studies' values on healthy volunteers. Here again, these prevalence discrepancies may be due to in racial and regional differences as well as the different techniques performed.

Aeration of the sphenoid bone starts after birth, and expansion of the SS is age-related. In general, the volume of the SS increases until the third decade of life [29]. In our study, which included patients aged between 18 and 68 years, the mean age was 35 years, and no statistically significant difference was observed in terms of patient’s age (p=0.143).

In the literature, several studies have measured the SS volume using CT scanning [14-16]. In general, the results of the sphenoid sinus volumes were comparable with the results of the other contries' studies in the aspects of racial or regional differences (Table ). However, the correlation between the presence of an SEC and the sphenoidal sinus volume has not been studied. In this study, using a multiplanar CT-volume rendering technique, it was revealed that the unilateral presence of the SEC reduced the SS volume of the side on which it was located. For example, when a left-sided SEC was present, a statistically significant reduction was observed in the left-sided SS volume compared to the right side. However, bilateral SECs had a more statistically significant reductive effect on the total SS volume by reducing both the left and right sphenoid volumes.

Although it has been suggested that the SEC must be opened during the transsphenoidal endoscopic sellar surgery and the presence of the SEC caused a reduction in the SS volume, we could not find any volumetric study on this issue in the literature. While planning our study, it was decided to investigate the relationship between the SS and the SECs in the following manner. It was determined whether two cells competed to fill a certain volume or whether they showed increased pneumatization together with the action of similar stimulating factors. Based on the results of this study, it was concluded that the SECs compete with the SS to fill a certain volume, resulting in a reduction in the volume of the SS.

From the surgeon’s perspective, in patients with a small SS volume coexisting with SEC, surgery and instrumentation of the affected SS may be more difficult under pathological conditions, and thus the skull base, pituitary gland, and carotid artery may be entered incorrectly, resulting in a greater risk of injury without opening the SEC. On the other hand, with the surgical opening of the SEC, it may become possible to contribute to cleaning of the pathology of the affected SS to improve patients’ clinical condition.

CONCLUSION

According to our study, sphenoethmoid cells significantly reduce the sphenoid sinus volume of the side on which they are located. In the case of low sphenoid sinus aeration, the most posterior ethmoid cell should be kept in mind. With this study, the ideas that the most posterior ethmoid cell that is an obstacle to the sphenoid sinus surgery and the opening of this cell will increase the volume of the sphenoid sinus, were made clearly with the quantitative data. Further studies with an extended patient series are required to explore this issue.

Table 1

Location and prevalence of sphenoethmoid cells.

-	Right Side	Left Side	Total Number of Sides
Unilateral SEC	18	20	38
Bilateral SECs	34	34	68
Total Number of SECs	52/141(36.8%)	54/141(38.2%)	106/282 (37.5%)

Table 2

Location of sphenoethmoid cells by gender.

Location of Sphenoethmoid Cell	Total		Females		Males
	n	(%)	n	(%)	n	(%)
Bilateral	34	47.2	15	46.9	19	47.5
Right Side	18	25.0	8	25.0	10	25.0
Left Side	20	27.8	9	28.1	11	27.5
Total	72	100.0	32	100.0	40	100.0

Table 3

Distribution of sphenoethmoid cells and sphenoid sinus volumes.

Presence of Spheoethmoid Cell	n (%)	Mean (±SD) Volume of Right SS (mL)	Mean (±SD) Volume of Left SS (mL)	Mean (±SD) Total Volume of SS (mL)
Negative (Group I, Control)	69 (48.9)	7.82 (±3.55)	7.25 (±2.65)	15.07 (±5.37)
Bilateral (Group II)	34 (24.1)	4.92 (±2.94)	4.08 (±2.57)	9.00 (±4.37)
Right Side (Group III)	18 (12.8)	4.70 (±2.81)	7.85 (±3.47)	12.55 (±4.89)
Left Side (Group IV)	20 (14.2)	8.28 (±2.97)	4.41 (±2.49)	12.69 (±4.60)
Total	141 (100)	6.79 (±3.55)	6.16 (±3.12)	12.95 (±5.51)
p value	-	<0.001	<0.001	<0.001
Post Hoc Testsp value	I→II	<0.001	<0.001	<0.001
	II→III	0.817	<0.001	0.016
	II→IV	<0.001	0.676	0.010
	III→IV	0.001	<0.001	0.932
	II→IV	<0.001	0.676	0.010

SS: Sphenoid Sinus.

Table 4

Comparative data of our and previous studies about the volume of sphenoid sinus.

Authors [Reference Number]	Country	Number of Cases	Average Sphenoid Sinus Volume (mL)	Right Sphenoid Sinus Volume (mL)	Left Sphenoid Sinus Volume (mL)	Total Sphenoid Sinus Volume (mL)
Kawarai et al. [15]	Japan	20	N/A	N/A	N/A	15.4±6.9
Kim et al. [30]	Korea	60	N/A	N/A	N/A	13,766
Oliveira et al. [31]	Brazil	47	N/A	6.157±3.5	7.26±3.6	N/A
Selcuk et al. [32]	Turkey	115	7.81¶ and 6.35¶	N/A	N/A	N/A
Our Study	Turkey	141	N/A	7.82±3.55	7.25±2.65	15.07±5.37

N/A: Not avaliable. ¶: Detected volumes in two different regions in a country.

Table 5

Various studies on study modality, number of subjects, performed technique and prevalence of sphenoethmoid cell.

Author	Study Modality	Number of Subjects and/or Performed Technique, Including Criteria	Sphenoethmoid Cell Prevalence
Hwang et al. [5]	Computed Tomography	100 patients, retrospective image analysing	32%
Nomura et al. [11]	Computed Tomography	200 patients, septal or chronic sinonasal symptoms	34.3%
Chmielik et al. [20]	Computed Tomography	196 patients, retrospective image analysing	39.8%
Al-Abri et al. [21]	Computed Tomography	435 patients, chronic sinonasal symptoms	8%
Leunig et al. [22]	Computed Tomography	641 patients, chronic sinus conditions	8.4%
Pérez-Pińas et al. [23]	Computed Tomography	110 patients, inflammatory sinus pathology	10.9%
Kasemsiri et al. [24]	Computed Tomography	187 patients, retrospective image analysing	49.5%
Wada et al. [25]	Computed Tomography	261 patients, chronic rhinosinusitis	50.8%
Tomovic et al. [4]	Computed Tomography	170 patients, retrospective image analysing	65.3%
Thanaviratananich et al. [2]	Cadaveric Half-head	65 specimen, endoscopic ethmoidectomy	60%
Yeoh et al. [27]	Cadaveric Head	102 specimen, endoscopic sphenoethmoidectomy	50.98%
Kainz and Stammberger [28]	Cadaveric Half-head	52 specimen, endoscopic sphenoethmoidectomy	42%
Jones et al. [26]	Computed Tomography	100 patients with rhinosinusitis and 100 controls with intraorbital disease	7% in patients, 9% in controls
Our Study	Computed Tomography	141 patients, chronic sinusitis	37.5%