Introduction

The optic canal’s proximal or cranial aperture (CAOC), transmits the ophthalmic artery, sympathetic fibers (around the ophthalmic artery), and optic nerve between the middle cranial base and orbital cavity [26, 41]. Due to traumatic (e.g., fracture) or pathologic lesions (e.g., craniopharyngioma, optic nerve compression, optic nerve meningioma, optic glioma, mucocele, and ophthalmic aneurysm) located around the CAOC, surgeons perform different surgical procedures such as the endoscopic endonasal transsphenoidal approach, pterional approach, subfrontal approach, and supraorbital approach [7, 8, 19, 35, 38, 44, 47]. The sellar region including the CAOC has an extremely complex structure leading to technical challenge in application of those approaches, but morbidity and mortality rates decrease owing to the developing technology (e.g., intraoperative image-guidance/neuronavigation), increasing anatomical knowledge, and novel surgical methods [12, 14, 18, 27, 29]. Despite these positive advances, one of the main criticisms is that the vast majority of information in the literature (e.g., surgical approach analysis, morphometric evaluations, and exhaustive anatomical definitions) is obtained from limited age groups (mostly adults) [3,4,5, 9]. In this context, further investigation conducted on children may be useful for surgeons to avoid complications (e.g., the optic nerve or ophthalmic artery injury).

Some pathologic lesions such as craniopharyngioma, optic glioma, and optic nerve compression are diagnosed in childhood [8, 9, 38]. Unintentional damage of the optic nerve is one of the common complications during the treatment of those lesions [7, 9, 19]. Considering the differences of pediatric patients compared to adult cases (e.g., management, types or metastasis pattern of tumoral lesions, and also anatomic features of the skull base) [13, 39, 45, 46], safe dissection route in children should be known in detail by surgeons [9, 28, 31, 40]. However, the majority of existing articles provided information based on adult subjects [3, 9]. This situation may create difficulties for surgeons familiar with adult data (depth, angle and diameter measurements) to ensure surgical orientation in children (e.g., implant positioning) [9]. In this context, similar guidelines with adult reference measurements should be adapted to children to position the patient, to choose the appropriate approach, and to treat lesions [9, 13, 39, 45, 46]. An extended morphometric dataset focused on characteristic features of the CAOC in pediatric patients may be helpful for otolaryngologist, neurosurgeons, and ophthalmologist to better enlighten safe dissection zone. The present work aimed to peruse the growth pattern of the CAOC in pediatric subjects between the ages of one and 20.

Materials and methods

The ethical approval was obtained from the review committee of Mersin University for the present retrospective computed tomography (CT) study (2020/412). The inclusion criteria of the work were planned as follows: (a) 200 (100 boys and 100 girls) pediatric patients aged from one to 20 years, (b) ten patients, five boys and girls, for all ages, (c) patients without fractures, infectious diseases, tumoral lesions, vascular disorders, syndromic or genetic malformations, (d) patients with good quality CT images, and (e) files of child patients who admitted to the hospital between January 1, 2019 and December 30, 2019. In this context, the files (which contained information about hospital admission date, anamnesis data, complaints, sex, age, radiological images, diagnostic info, management procedures, and hospital discharge date) of 4592 pediatric patients (2859 boys and 1733 girls) with different grievances (blow to head, dizziness, falling from high, headache, etc.) in 2019 were reviewed retrospectively. The exclusion criteria consisted of the following: (a) patients after 200 (100 boys and girls) randomly chosen pediatric subjects aged from one to 20 years, (b) patients after 10 (five boys and girls) randomly chosen child subjects for all ages, (c) patients over 20 years, (d) patients with fractures in craniofacial bones (e.g., orbit or skull base), (e) patients with syndromic, vascular or genetic malformations (e.g., aneurysms or Crouzon syndrome), (f) patients with infectious diseases, (g) patients with tumoral lesions (e.g., craniopharyngioma), and (h) patients with low quality CT images. The raw data from CT scans (64-slice scanner, Aquillion 64, matrix: 512 × 512, FOV: 240 mm, pixel size: 0.46 mm, 0.5 mm thick slices, 230 mA, 120 kV, 0.3 mm interval, Toshiba Medical Systems Tokyo, Japan) were reformatted in different planes to form three-dimensional multiplanar reconstruction images. All measurements were performed by the same neuroradiologist (BT) once in different images (sagittal, coronal and axial planes). In 200 randomly chosen pediatric subjects (sex: 100 boys and girls, age: 10.50 ± 5.77 years) according to the inclusion and exclusion criteria, the following parameters were measured (Figs. 1, 2):

  • The height (vertical diameter) of the CAOC in coronal plane on CT.

  • The width (horizontal diameter) of the CAOC in coronal plane on CT.

  • The surface area of the CAOC in coronal plane on CT.

  • The distance (Dis-to-AB) between the CAOC (the frontmost point) and the anterior boundary of the anterior skull base in axial plane on CT.

  • The distance (Dis-to-LB) between the CAOC (the most lateral point) and the lateral boundary of the anterior skull base in axial plane on CT.

  • The distance (Dis-to-MSL) between the CAOC (the most medial point) and the midsagittal line in axial plane on CT.

  • The angle (Ang-in-AP) between the optic canal and sagittal horizontal line (i.e., the midline axis) in axial plane on CT.

  • The angle (Ang-in-SP) between the optic canal and sagittal horizontal line (i.e., the line parallel to the ground) in sagittal plane on CT.

  • The CAOC shape in coronal plane on CT.

Fig. 1
figure 1

Photographs show the parameters. a the CAOC width, (b) the CAOC height, (c) the CAOC area, (d) Dis-to-MSL, (e) Dis-to-LB, and (f) Dis-to-AB. *The location of the CAOC

Full size image
Fig. 2
figure 2

Photographs show the parameters. a Ang-in-AP, and (b) Ang-in-SP. *the location of the CAOC

Full size image

The normality control of the morphometric values belonging to the CAOC parameters was checked with Shapiro–Wilk test. One-way ANOVA (post-hoc Bonferroni test) was used to examine the changes in the CAOC parameters during the transition from 1 to 20 years (Table 1), and from infancy period to postpubescent period (patients aged between 0 and 2 years: infancy group, patients aged between 3 and 5 years: early childhood group, patients aged between 6 and 9 years: later childhood group, patients aged between 10 and 13 years: prepubescent group, and patients aged between 14 and 20 years: postpubescent group) (Table 2) [5, 43]. Moreover, those tests were utilized to display the alterations of the parameters relative to the CAOC shape (Table 3). Right–left (the paired test) or boys–girls comparisons (the independent test) were carried out with Student’s t test (Table 4). Correlations between the depth, angle and diameter measurements of the CAOC were analyzed with Pearson correlation coefficient test (Table 5). The relation of the CAOC shape with child growth periods (between infancy period to postpubescent period) were evaluated with Chi square test (Table 6). The regression equations and scatter plots related to the CAOC parameters were obtained via the simple linear regression test. The “p < 0.05” was set as threshold during statistical evaluations.

Table 1 Measurements related to the CAOC in children aged from 1 to 20 years
Full size table
Table 2 Comparison of the parameters according to child growth periods
Full size table
Table 3 Comparison of the parameters according to the CAOC shape
Full size table
Table 4 Comparison of the parameters in terms of sex and side
Full size table
Table 5 Correlations between the CAOC parameters
Full size table
Table 6 Statistical evaluation of the CAOC shape according to child growth periods
Full size table

Results

The numerical values related to the CAOC parameters were presented as mean data ± standard deviations in tables due to the normal distribution. Our findings were as follows:

  • All parameters (depth, angle and diameter measurements) varied depending on growth in childhood (Table 1). The CAOC height, Ang-in-AP, and Ang-in-SP were decreasing during the transition from 1 to 20 years, while the CAOC width, Dis-to-AB, Dis-to-LB, and Dis-to-MSL were increasing (p < 0.001) (Table 1).

  • The CAOC height, CAOC width, Ang-in-SP, and Dis-to-MSL did not alter from prepubescent period, while the Ang-in-AP did not vary from early childhood period. The surface area of the CAOC displayed irregular alteration between child age groups. The Dis-to-LB did not vary from the age of the six (i.e., later childhood period), while the Dis-to-AB were increasing proportionally with age between infancy and postpubescence (Table 2).

  • The CAOC shape were described as the tear-drop (186 foramina, 46.5%), triangular (156 foramina, 39%), oval (47 foramina, 11.8%), and round (11 foramina, 2.8%) (Figs. 3, 4).

  • The CAOC shape in boys was observed as the triangular (86 foramina, 43%) > tear-drop (83 foramina, 41.5%) > oval (27 foramina, 13.5%) > round (four foramina, 2%), while in girls as the tear-drop (103 foramina, 51.5%) > triangular (70 foramina, 35%) > oval (20 foramina, 10%) > round (seven foramina, 3.5%).

  • The CAOC shape in right sides was found as the triangular (89 foramina, 44.5%) > tear-drop (81 foramina, 40.5%) > oval (24 foramina, 12%) > round (six foramina, 3%), while in left sides as the tear-drop (105 foramina, 52.5%) > triangular (67 foramina, 33.5%) > oval (23 foramina, 11.5%) > round (five foramina, 2.5%).

  • The measured parameters except that the CAOC area (p = 0.223) and Dis-to-LB (p = 0.225) varied according to the shapes of the CAOC (Table 3).

  • The parameters other than the Ang-in-SP (p = 0.348), Ang-in-AP (p = 0.730), and Dis-to-MSL (p = 0.118) were greater in boys than girls. In terms of sides, the parameters did not display significant differences (Table 4).

  • Correlation findings between the measurements are given in Table 5. Correlations did not found between the vertical and horizontal diameters (p = 0.052, r = 0.097), and between the Ang-in-SP and Ang-in-AP (p = 0.125, r = 0.077) (Table 5).

  • The CAOC shape distribution by child growth periods displayed that the age affected the shape (p < 0.001) (Table 6).

  • The CAOC height was smaller than the width (p < 0.001).

  • The Ang-in-SP was smaller than the Ang-in-AP (p < 0.001).

  • The Dis-to-AB was greater than the Dis-to-LB (p < 0.001).

  • The regression equations for the CAOC width (y = 5.760 + 0.035× years), the CAOC height (y = 4.626—0.026×years), the CAOC area (y = 17.678—0.014× years), the Ang-in-SP (y = 22.218—0.592×years), the Ang-in-AP (y = 41.085—0.172×years), the Dis-to-AB (y = 46.457 + 0.723×years), the Dis-to-LB (y = 40.247 + 0.220×years), and the Dis-to-MSL (y = 5.701 + 0.141 × years) were calculated (Figs. 5, 6).

Fig. 3
figure 3

Photographs show the CAOC shapes (dashed white lines)

Full size image
Fig. 4
figure 4

Charts show the rates of the CAOC shapes

Full size image
Fig. 5
figure 5

Charts show the linear functions. a the CAOC width, (b) the CAOC height, (c) the CAOC area, and (d) Dis-to-MSL

Full size image
Fig. 6
figure 6

Charts show the linear functions. a Ang-in-SP, (b) Ang-in-AP, (c) Dis-to-AB, and (d) Dis-to-LB

Full size image

Discussion

The anatomic features of the CAOC in association with neighboring structures such as the anterior clinoid process, prechiasmatic sulcus, optic strut may be important for surgeons during the management of tumoral entities such as meningioma and craniopharyngioma using transcranial or transnasal approaches [8, 18, 21, 29, 38]. For example, Beer-Furlan et al. used the pterional port approach for removal of the bones surrounding the optic nerve (anterior clinoid process and optic strut) to obtain 270° nerve decompression [2]. However, Shibata et al. used the endonasal transsphenoidal technique (endoscopically) to make urgent optic nerve decompression in an infant with craniopharyngioma [38]. Guthikonda et al. identified the prechiasmatic sulcus as four different types (a: wide-flat, b: wide-steep, c: narrow-flat, and d: narrow-steep), and suggested that the narrow-steep type was more suitable for the pterional approach, while the wide-flat type for the subfrontal approach [18]. Moreover, they stated that the narrow prechiasmatic sulcus was more suitable for transcranial procedures, whereas the wide sulcus for transnasal procedures [18]. Beger et al. studied on the prechiasmatic sulcus in pediatric cases aged from one to 20 years, and observed that the sulcal length (antero-posterior length of the sulcus) did not alter after birth, but the sulcal angle (the angle between the planum sphenoidale and prechiasmatic sulcus) was decreasing from the postpubescent period [5]. In the light of all these clinical comments, we thought that this morphometrically expanded study focused on the CAOC in children might be beneficial for multidisciplinary operating teams to understand alterations in its morphology in children aged between infancy to postpubescence periods, to avoid unintentional damage of neurovascular structures (e.g., the optic nerve), to guess its location and diameter in different age groups, and to fathom the difference of the sellar region in children from adults.

The CAOC area, width and height in this radiologic study were found as 17.53 ± 2.80 mm2, 6.12 ± 0.84 mm, and 4.35 ± 0.64 mm, respectively. The existing papers related to these parameters are given in Table 7, which displayed that limited studies focused on the CAOC anatomy in children were available in the literature [33]. Prado et al. stated that the diameter of the CAOC (3.76/3.79 mm for fetuses, 4.67/4.72 mm for children, and 5.24/5.43 mm for adults) were increasing from prenatal period to adulthood [33]. Our average values obtained from children aged with 10.50 ± 5.78 years were within the mean data range given for adults in previous articles (4.45–19.36 mm2 for the area, 4.59–7.38 mm for the width, and 3.60–5.17 mm for the height) [1, 6, 16, 20, 23, 26, 28, 33, 34, 36, 37, 40, 42]. The CAOC height and width did not change from the age of 10 (prepubescent period), while its area displayed irregular alteration between child growth periods. The existing papers (Table 7) displayed that the numerical data related to the CAOC in previous studies were quite different. Possible reasons for this situation were listed by Beger as follows: (a) measurement technique (e.g., CT and digital caliper), (b) demographic differences (e.g., sex and sides), and (c) specimen selection (e.g., pediatric or adult patients) [3]. Kalthur et al. observed that the diameters measured by caliper were greater than measured by CT [23]; however, Berlis et al. stated that the measurements obtained from CT and caliper were similar [6]. Kalthur et al. reported that the diameters in the left (4.51 ± 0.79 mm for the width, and 3.54 ± 0.71 mm for the height) were smaller than in the right (4.68 ± 0.86 mm for the width, and 3.67 ± 0.82 mm for the height) [23]. Kumar et al. found that the height in the left (4.82 ± 0.77 mm) was smaller than in the right (5.14 ± 0.73 mm) [26]. Similar to the other studies [3, 20, 33, 34, 36], we did not find significant differences between right and left sides. In addition, Kalthur et al. observed that the diameters in females (4.24 ± 0.67 mm for the width, and 3.51 ± 0.82 mm for the height) were smaller than in males (4.75 ± 0.83 mm for the width, and 3.63 ± 0.74 mm for the height) [23]. In this study, the parameters including the area, height and width of the CAOC were greater in boys than girls. Area calculation method might be one of the reasons that trigger the difference between studies [36, 42]. The area in the study of Tao et al. (4.45 ± 0.46 mm2, using computer morphometric analysis system) was smaller than in the study Radunovic et al. (19.36 ± 1.87 mm2, using “width×height×π” formula) [36, 42].

Table 7 Morphometric data related to the CAOC dimension
Full size table

The Ang-in-AP and Ang-in-SP in this study were found as 39.28 ± 5.13° and 16.01 ± 6.76°, respectively. The existing articles related to the angles are given in Table 8, which showed that limited studies focused on the CAOC anatomy in children were available [21]. The Ang-in-AP in this work was compatible with the value (39.5°) of Harwood-Nash’s work focused on children [21]. Our average values obtained from children were within the mean data range given for adults in previous articles (29.56–45.32° for the Ang-in-AP, and 10.27–18.2° for the Ang-in-SP) [6, 11, 15, 16, 21, 22, 31, 32, 40, 48]. Guthikonda et al. suggested that the position of the nerve might change due to the optic canal disruption caused by tumoral lesions; in such cases, the safest way of intervention in avoiding iatrogenic injury might be the normal course of the nerve [18]. The Ang-in-SP did not alter from the age of 10 (prepubescent period), while the Ang-in-AP did not vary from the age of three (early childhood period). The Ang-in-SP in infants (24.87 ± 5.74°) was almost twice that of adolescents (12.81 ± 5.53° for postpubescent period), and in addition, the Ang-in-AP in infants was statistically greater than the other age groups. On the other hand, the Dis-to-MSL, Dis-to-AB, and Dis-to-LB in this study were found as 7.17 ± 1.48 mm, 54.04 ± 5.23 mm, and 42.55 ± 3.28 mm, respectively. Our measurement related to the Dis-to-MSL was compatible with the previous measurements (at mean range: 6.48–7.48 mm) (Table 8) [11, 16, 31, 48]. The orbital depth measurements (i.e., the Dis-to-AB) were presented between 39–50 mm in the previous articles [1, 10, 24, 25, 30, 31]. Selection of different reference points during the depth measurements may be the reason that trigger the difference between studies (Table 9) [9]. Chang et al. measured orbital floor depth (the distance between the inferior orbital rim and the optic canal’s inferomedial rim) in children aged between 3 months and 18 years using magnetic resonance imaging views, and found that the depth was increasing rapidly until 6 years, and it seemed to reach almost adult size at 13 years [9]. The Dis-to-MSL did not alter from the age of 10 (prepubescent period), while the Dis-to-LB did not change from the age of six (later childhood period). In addition, the Dis-to-AB were increasing in the transition from infancy period to postpubescent period. In a nutshell, these angular and locational changes in childhood should be kept in mind by surgeons in terms of interventions such as implant positioning.

Table 8 Morphometric data related to the location of the CAOC
Full size table
Table 9 Morphometric data related to the distance of the CAOC to certain anatomical landmarks
Full size table

The CAOC shapes were described as the tear-drop in 186 cases (46.5%), triangular in 156 cases (39%), oval in 47 cases (11.8%), and round in 11 cases (2.8%). The existing papers containing shape classifications are given in Table 10, which showed that shape definitions were quite different between the studies. The CAOC was described as elliptical in all skulls by Akdemir et al. and Maniscalco and Habal, as oval by Govsa et al. and Purohit and Singh, and as round by Slavin et al. and Kumar et al. [1, 16, 26, 28, 34, 40]. In the study of Beger, the CAOC was identified as round (28%) and oval (72%) [3]. The shape of the CAOC was defined by Guseva and Denisov as triangular, polygonal, rhomboidal, horizontal–oval, and round [17]. Demographic information such as population, small-scale research, or imperfect techniques may be the reasons that trigger the difference in shape description between studies [16, 17, 28, 34]. Beger observed that the width of the CAOC in fetuses (round shape: 1.48 ± 0.34 mm, oval shape: 1.97 ± 0.61 mm) were varying according to the CAOC shape [3]. Except that the CAOC area and Dis-to-LB, the other parameters in this study varied according to the shapes of the CAOC. Moreover, the distribution of the CAOC shapes by child growth periods displayed that its shape was affected by age. Knowledge related to microsurgical anatomy of the CAOC may be useful for surgeons to prevent unintentional damage of neurovascular structures (the optic nerve, ophthalmic artery, etc.).

Table 10 Literature data raleted to the CAOC shape
Full size table

Conclusion

The CAOC height, CAOC width, Ang-in-SP, and Dis-to-MSL did not change from the age of 10 (prepubescent period). The Dis-to-LB did not change from the age of six (later childhood period), while the Dis-to-AB were increasing in the transition from infancy period to postpubescent period. In addition, the Ang-in-AP did not vary from the age of three (early childhood period). The depth, angle and diameter measurements belonging to the CAOC were changing according to the CAOC shape or demographic data (age and sex). From these perspectives, preoperative radiologic evaluation containing the shape, location and size of the CAOC may be beneficial for multidisciplinary operating teams during surgical interventions such as implant positioning.