Functional endoscopic sinus surgery through transnasal approach is a common modality of treatment for disorders of the nasal cavity, paranasal sinuses as well as cranial cavity [1]. The variations in morphology of the paranasal sinuses may alter the results during minimally invasive procedures and therefore anatomical considerations of ethmoid sinus (ES) were reported as early as in 1929 by great Mosher that ‘If the ES were placed in any other part of the body, it would be an insignificant and harmless collection of bony cells. In the place where nature has put it, it has major relationships so diseases and surgery of the labyrinth often lead to tragedy. Any surgery in this region should be simple but has proven one of the easiest ways to harm a patient’ [2].

The ethmoid bone has a delicate cribriform plate, which separates the nasal cavity from the anterior cranial fossa and is the thinnest most variable part of the cranial base. The olfactory fossa (OF) is located along the superior aspect of cribriform plate which varies in shape and depth. This variable measurement of the depth of OF or the height of ethmoid roof is mostly responsible for a greater risk of intracranial infiltration during endoscopic procedures in and around the nasal cavity [3-5]. Previous authors and Kero’s described high risk cases where the depth of OF was between 7–16 mm and categorized them as type III category on paranasal sinus CT scans. Further, currently there is limited research on measurements of OF [5-7].

The morphology of frontal and ES and frontoethmoid recess vary from simple to complex. The frontal sinus (FS) is located in the bone of the same name, posterior to the superciliary arches, and drains through frontonasal duct to ethmoid infundibulum which opens into the semilunar hiatus of middle nasal meatus. The size of FS is directly related to prevalence of sinusitis. The medium and larger FS are more prone to infection in comparison to smaller FS. These are the only paranasal sinuses that are absent at birth, and usually, these sinuses do not extend up into the frontal bone until the age of six years. As the left and right FS develop independently, a significant asymmetry between FS can be observed in the same individual [8]. The ES are the only paranasal air sinuses that are formed of multiple thin walled cavities in the ethmoid labyrinth. Clinically the ES are categorized into anterior and posterior groups on each side by a tortuous barrier [9]. Although, there is continuous progress in three-dimensional imaging techniques during past two decades but with lack of necessary depth of perception [3-6]. This cadaveric study is planned to improve the ability of the otolaryngologist, radiologist, and anatomist to understand the possible morphological variations and plan steps of less invasive “precision surgery” to have a safe and complication free procedures and communicate these complexities when teaching or reporting outcomes.

The study was carried out on 37 cadavers available in the department of Anatomy, AIIMS Bhopal, India. Permission through letter number IHEC-LOP/2018/IM0175 from the Institutional Ethical Committee of AIIMS, Bhopal was taken before starting the project. The cadavers were fixed in 10% neutral formalin and were partially dissected by medical students. The anatomic specimens belonged to subjects with an age range from 31 to 80 years (mean age, 63 years). All cadavers included were free of any gross signs of nasal and paranasal sinus deformity. The specimens with pathological appearances such as deformity, fracture, cysts, or tumors were excluded from the study.

Functional endoscopic sinus surgical procedures on one hand enhanced the outcomes of surgery in and around paranasal air sinuses for almost three decades and on the other side has also made it essential to have precise knowledge about the anatomy of paranasal air sinuses and floor of anterior cranial fossa [1, 3, 5]. The depth of OF was studied extensively by Kero’s P. The author categorized the OF into three categories where the depth of the OF was estimated as type 1 (1–3 mm), type 2 (4–7 mm), and type 3 (8–16 mm) [6]. We found that in the original Kero’s classification the measurements between 0–0.9, 3.1–3.9, and 7.1–7.9 mm were difficult to classify in given types and there was a breach of almost 0.99 mm between the mentioned first, second and third categories. To maintain the continuity in measurements of the depth of OF, this Modified Kero’s classification is proposed and applied in the present study. The Modified Kero’s classification was applied and measurements were grouped as modified Keros type I (0–3.99 mm), type II (4–7.99 mm), type III (8–16 mm) respectively. In Kero’s study, Type I was described in 26.3% of population, versus 29.7% in our study. Keros Type II OF described in 73.3% of the population, versus 60.8% in this study. The OF of Keros III was described in 0.5% of the population, versus 9.5% in our study. Almushayti et al. [4] also found the type II category as the most common followed by type I then type III. Few authors reported that type one was the most common followed by type two then type three [7, 10, 11]. This significant difference in the result between the present study and the previous ones may be attributed to the difference in the populations, and variation in techniques of measurement (radiographic CT study versus cadaveric specimen) and variation in grouping the measurements into different types. In Table 5 we have compared the results of Kero’s classification by various authors from different countries [10-16].

An absence of pneumatization in the frontal bone results in FS aplasia. Bilateral absence of the FS has been reported in 2% to 33%, whereas the incidence of a unilateral absence had been reported to be between 0.8% and 7.4% [17-19]. In the present study absence of FS is observed in 9.46% cases. Krogman [20] had observed its absence in 5% adults, while Gulisano et al. [21] observed its absence in 24.7% of the cases. Ozdemir et al. [22] found frontal sinusitis more common in medium and large FS which we encountered in 19% and 9.5% of specimens respectively.

Agger nasi cell is the anterior most ethmoid air cell, located below the FS forming a significant part of the anterior wall of the frontal recess. Agger nasi cells may impinge upon the FS drainage tract, extending infero-laterally to lacrimal fossa and located antero-superior to the hiatus semilunaris. AN incidence was reported at different rates in different studies. Kayalioglu et al. [23] reported AN incidence as 7.8%. In contrast, in another study, Bradley and Kountakis [24] reported this incidence very high, 93%. Özdemir and Arslan [22] found the incidence of AN cells was 51.9%, and bilateral AN cells 25.2%. We found 14.8% agger nasi cells during the study.

The anatomical variation of posterior most ethmoid cells is called Onodi cells (spheno-ethmoid cells), seen extending posterior, lateral and superior to the sphenoid sinus, medial to the optic nerve. Extensive pneumatization can expose the circumference of optic nerve, surrounded by air spaces. In presence of Onodi cell, both the internal carotid artery and optic nerve may be exposed within the posterior ethmoid cells and result in most serious surgical complications. Driben et al. [25] noted the prevalence of Onodi cell in 39% on endoscopic examination in cadavers vs. 7% on an axial view of CT scans. Weinberger et al. [26] informed that the prevalence of the Onodi cell was 14% on endoscopic examination in cadavers versus 8% on a coronal view of CT scans. Arslan et al. [27], studied CT scans using two views (axial and coronal) and reported prevalence as 12%. Yeoh et al. [28] and Thanaviratananich et al. [29] found the prevalence as 51% (95% CI: 41.4–60.5) and 60% (95% CI: 47.9–71.0) respectively on endoscopic examinations of the posterior ES. In our study, we found 8% Onodi cells.

In conclusion, an attempt has been made to identify and reclassify the depth of OF as modified Kero’s classification as few dimensions were ignored in original Kero’s classification. Study of different shapes, dimensions, septa in FS and number of cells along with Onodi cell and agger nasi cell of ES presented in study provide more precise information.