The vidian canal: radiological features in Japanese population and clinical implications.

The vidian canal (VC), a bony tunnel in which the vidian artery and nerve pass, has been widely known as an important landmark to identify the anterior genu of the petrous carotid artery (AGPCA) especially during lateral extended endoscopic endonasal approachs (LEEEAs). The objectives of this study in the Japanese population are to describe the radiological anatomic features and relationships between VC and its surrounding structures, and discuss the clinical implications. We studied 231 high-resolution computed tomography (CT) scans with a slice thickness of 0.5 mm. All the patients had known sellar or parasellar pathologies but without any involvement of VC. The following VC-related parameters were examined: its length, relationship to AGPCA, course from the pterygopalatine fossa to the carotid canal, its position relative to the medial pterygoid plate and pneumatization pattern of the sphenoid sinus. Mean length of VC is 14.6 mm. There is more tendency of straight-running VC compared to other populations. VC locates infero-lateral to AGPCA in all the cases. The protrusion of VC and the paraclival carotid artery to the sphenoid sinus, as well as well-pneumatization of the sinus is also observed more frequently in almost a half of the population. Surgeons who perform LEEEAs in Japanese patients must know these anatomical features. The characteristics particular to Japanese populations may facilitate better identification of VC and exposure to AGPCA intraoperatively.

Introduction

Expanded endonasal approaches (EEAs) have been widely accepted to treat selected skull base lesions.[1–7)] Kassam et al. classified EEAs according to the accessibility to specific areas of the skull base defined by the sagittal and the coronal planes.[3,5,6,8)] In coronal planes, EEAs allow access to the middle third of the clivus, the petrous bone, the middle cranial fossa, the inferior and superior portions of the cavernous sinus, and the medial portion of the infratemporal fossa through a series of five modular approaches.[9)] These modules can be divided into suprapetrous and infrapetrous. The difference between these two is the position relative to the horizontal segment of the internal carotid artery (ICA).[9)] Identification of the anterior genu of the ICA is essential when EEAs are performed to access lesions located laterally in the skull base with intimate relationship with the segments of the ICA.[10)] The vidian canal (VC) and its contents are known to be a consistent and reliable surgical landmark to the petrous ICA, allowing its identification at the point of transition from the horizontal petrous to the vertical paraclival segment of ICA at the level of the foramen lacerum (FL).[9,11,12)]

VC is located in the skull base along the line of fusion of the pterygoid process and the body of the sphenoid bone. It connects the pterygoid fossa with FL, and transmits the vidian nerve and vessels.[13)] The anterior opening of VC is located infero-medial to the foramen rotundum (FR). Anatomic and radiological features of VC have been described in detail to facilitate the perception of its relationships and distances with other important bony, neural, and vascular structures.[12–16)]

Recently, Lu et al. have described differences in the pneumatization patterns of the sphenoid sinus (SS) between Chinese and Caucasian individuals.[17)] These different features, related to poor lateral and clival pneumatization in Chinese, might make the EEAs more dangerous in Chinese population. However, no similar study has been performed in Japanese population. Thus, the aim of this study in Japanese population is to describe the anatomical features of VC radiologically and its relationships between VC and surrounding key landmarks including the anterior genu of the petrous ICA. We also discuss the differences between Japanese and others subjects.[12,18)]

Materials and Methods

We reviewed a total of 231 high-resolution computed tomography (CT) scans. Only adult patients were included. Mean age was 51 years old. Patients with sellar or parasellar pathology, but without disruption of VC were selected among patients operated neuroendoscopically in Chiba University Hospital (CUH) from January 2008 to September 2012. Patients without skull base pathology were randomly selected from CUHs radiographic database. Finally, 115 CT scans presented skull base pathology, being the most frequent diagnosis pituitary adenoma, while 116 did not present skull base pathology. High-resolution CT scans were obtained at 0.5 mm thickness using the Aquilion one multidetector CT scanner (Toshiba Medical Systems, Otawara, Tochigi). Matrix is 512 × 512 and field of view is about 200 mm, adjusting to a patient’s head size. The images were analyzed on PACS workstation.

VC-related parameters were studied in axial and reconstructed-coronal planes. The length and course of VC from anterior to posterior and depth of the pterygopalatine fossa (Fig. 1A) were studied in the axial plane. The coronal plane was used to measure the distance from VC to FR, the distance from VC to the base of the medial pterygoid plate (MPTG) (Fig. 1B), the plane of MTGP relative to VC (Fig. 2), the position of the posterior end of VC relative to anterior genu of the petrous carotid ICA (Fig. 3), the pneumatization patterns of SS, especially its lateral portion relative to VC, and the frequency of VC protrusion into SS (Fig. 4). SPSS software version 19.0 (SPSS, London, UK) was used for the statistical analysis. All the abbreviations of the anatomical landmarks discussed in this study are listed in Table 1.

Fig. 1.

A: Axial computed tomography (CT) scan slide showing measurements obtained: a: vidian canal (VC) length, b: VC course, c: pterygopalatine fossa (PTG) depth, arrow: VC. B: Coronal CT scan showing measurements obtained: d: distance from VC to foramen rotundum (FR), e: distance between VC and medial pterygoid plate (MPTG) base, m: midline. Arrow: MPTG, arrowhead: VC, asterisk: FR.

Fig. 2.

Coronal computed tomography (CT) scans showing three different VC (arrowhead) positions relative to MPTG (dotted line). A: medial position, B: straight position, C: lateral position. Arrow: palatovaginal canal, VC: vidian canal.

Fig. 3.

Coronal computed tomography (CT) scan showing the relationship between the posterior end of VC (arrow) and anterior genu of the petrous portion of the internal carotid artery (encircled). VC: vidian canal.

Fig. 4.

Coronal computed tomography (CT) scans showing the four different patterns of pneumatization degree of VC (arrowhead) according to Vescan et al. A: Grade I, B: Grade II, C: Grade III, D: Grade IV. Total protruded canal is showed in D image.

Table 1

Abbreviation list of surgical landmarks in endoscopic endonasal approaches

Abbreviation	Description
AGPCA	anterior genu of the petrous carotid artery
FL	foramen lacerum
FR	foramen rotundum
ICA	internal carotid artery
MPTG	medial pterygoid plate

Results

The length of VC was measured in the axial plane along its medial wall, from the posterior wall of the pterygopalatine fossa to its end on the carotid canal (Fig. 1A-a). The length averaged 14.4 mm (range 7.9–20.3) on the right side, and 14.7 mm (8–20.8) on the left.

The direction of VC from the pterygopalatine fossa to FL was classified into three groups: medial to lateral, lateral to medial, and straight course (Fig. 1A-b). On the right side, VC had medial to lateral course in 212 (91.8%) and straight course in 19 (8.2%). Similarly, it was medial to lateral in 209 (90.5%), straight in 21 (9.1%), and lateral to medial in 1 (0.4%) on the left.

The mean depth of the pterygopalatine fossa measured in the axial plane at the level of VC (Fig. 1A-c) was 6.7 mm (2.4–10.5) on the right and 6.7 mm (2.4–11.4) on the left. The anterior opening of VC is generally located infero-medial to FR. We investigated the distance from the lateral aspect of VC to the medial aspect of FR at their anterior opening to the pterygopalatine fossa (Fig. 1B-d). The mean distance on the right side was 6.5 mm (1.0–13) and on the left side was 7.2 mm (1.0–15.7).

Another important bony structure to localize anterior opening of VC is MPTG. We analyzed the distance between the origin of MPTG and VC, and the position of VC relative to MPTG. Both measurements were done in the most rostral coronal image, which exposed both structures (Figs. 1B-e, 2). The mean distance on the right side was 8.5 mm (1.8–15.2) and 8.9 mm (3.1–15.1) on the left. The position of VC relative to MPTG was classified according to a vertical line from MPTG and determined whether VC was medial, straight, or lateral respectively (Fig. 2A–C). On the right side, VC was located medially in 210 (90.9%), straight in 20 (8.7%), and laterally in 1 (0.4%). Similarly the VC position on the left was medial in 199 (86.1%), straight in 28 (12.1%), and lateral in 4 (1.7%).

Furthermore we examined the position of the posterior end of VC relative to the anterior genu of the petrous ICA in the coronal plane (Fig. 3), and revealed that VC was without exception located infero-lateral to the anterior genu of the petrous ICA.

SS pneumatization pattern, specifically its lateral portion relative to VC, was studied on the coronal plane by using the classification system proposed by Vescan et al.,[12)] which evaluates the degree of pneumatization relative to the circumference of VC (Fig. 4). The pneumatization pattern graded as 0% = grade 1 (Fig. 4A), 1% to 33% = grade 2 (Fig. 4B), 34% to 66% = grade 3 (Fig. 4C), and 67% to 100% = grade 4 (Fig. 4D). Table 2 describes the results of pneumatization patterns, determining the incidence of VC protrusion into SS. Although VC protrusion was of course absent in Vescan’s grade 1 group, it was always obvious in subjects with grade 3 or 4. Interestingly, we identified VC protrusion on the right side in 3 of 47 cases classified into grade 2, and in 2 of 49 cases on the left. Thus, the overall incidence of VC protrusion was 52.8% (122 cases) on the right side, and 51.9% (120 cases) on the left. We found a complete protrusion of VC with a thin bony stalk anchoring to the sphenoid bone (Fig. 4D) in 26 (11.2%) cases on the right side and 27 (11.7%) on the left. No statistical differences were found between sides in any measurements.

Table 2

Lateral SS pneumatization relative to the vidian canal classified by the method proposed by Vescan et al.

Pneumatization grade	Right sphenoid	Left sphenoid
Grade 1	65 (28.1%)	62 (26.8%)
Grade 2	47 (20.3%)	51 (22.1%)
Grade 3	95 (41.1%)	90 (38.9%)
Grade 4	24 (10.4%)	28 (12.1%)

Discussion

The nasal cavity and SS are considered as a corridor to the anterior, middle, and posterior fossa. Precise detection of the important surgical landmarks allows the surgeon to recognize important neurovascular structures and to avoid their damages. Within SS, the external bony protuberances and recesses on the posterior and lateral wall are the key landmarks that will guide the surgeon. However, a well-pneumatized SS is essential for such landmarks. VC is a consistent and reliable landmark not depending on the pneumatization degree. It helps the surgeons to determine the position of the anterior genu and lacerum segment of the ICA as it emerges from the petrous bone.[9,12,13)] Therefore, the identification of VC is an important step when EEAs are performed to access lesions located lateral in the skull base with intimate relationship with the segments of the ICA.[10)]

VC is an osseous tunnel that contains the vidian nerve and artery, and connects the pterygopalatine fossa with FL. It has been shown to have variability in its course and length. The most frequent course from anterior to posterior is medial to lateral. We found this direction in 91.8% of VC on the right side, and in 90.5% on the left. Straight course has been found to be 8.2% of the time on the right side and 9.1% on the left. One lateral to medial direction course was found in the left side. These results are different from other reports previously published, as we found a significant higher incidence of straight course,[12)] and next to nothing frequency of lateral to medial VC course compared to a study by Kim et al., who reported an incidence of lateral to medial course of 19.5% in Korean population.[14)] These data demonstrate the considerable and interesting differences of VC course not only between the Japanese and Caucasian populations but also between the Japanese and Korean who are thought to be very common genetically.

The length of VC reported in previous papers varies from 10 mm to 18 mm.[2,13,19)] We found that the mean length of VC was 14.4 mm on the right side and 14.7 mm on the left. This is an important distance, because the amount of bone we need to drill during the approach will be related to the length and the pneumatization degree of VC.[13)] Kassam described the drilling technique on the inferior and medial surface of VC to identify the petrous ICA, safely.[9,12)] This is why we measured the length of VC along its medial wall. Moreover, the posterior wall of the pterygopalatine fossa has an oblique orientation in the axial plane, making the medial wall of VC shorter than the lateral wall. This circumstance could explain the difference detected with the length reported by Vescan et al., as they measured the length along the lateral wall.[12)]

One important step during EEAs is to localize the anterior opening of VC on the posterior wall of the pterygopalatine fossa. The distance between the base of MPTG and VC gives us useful information to localize the anterior opening of VC, and also about the amount of bone we usually need to drill to identify VC during the transpterygoid approach.[8,10)] We found an average distance of 8.5 mm on the right side and 8.9 mm on the left, similar to the distance reported in the literature.[12)] The position of VC relative to MPTG is also a valuable data to easily find the anterior vidian opening. The most frequent position in our study was medial (right 90.9%–left 86.1%), straight in line the second most frequent (right 8.7%–left 12.1%), and lateral to it the least frequent (right 0.4%–left 1.7%). This relationship helps us radiologically to avoid confusion with palatovaginal canal, located medially and usually smaller, or with FR located supero-laterally to the anterior opening of VC. The distance between the lateral aspect of VC and medial aspect of FR is relevant for endonasal approaches to the quadrangular space, which is limited by the horizontal portion of the petrous ICA inferiorly, the second division of the trigeminal nerve laterally, the paraclival carotid medially, and the abducens nerve superiorly.[3)] An average distance in our study was 6.5 mm on the right side, and 7.2 mm on the left side, compared to an average of 5 mm to 9 mm in the previous reports.[12–14,19)]

The pterygopalatine fossa needs to be opened to visualize the anterior opening of VC.[11,12)] The average depth of the pterygopalatine fossa in the axial plane at the level of VC previously reported was 6.9 mm on the right side and 7.1 mm on the left.[12)] These results are consistent with our findings, mean depth 6.7 mm on both sides.

VC courses in the floor of SS and may protrude into it.[13,14,20)] The degree to which VC protrudes into the sinus depends on the degree of pneumatization. We specifically studied the pneumatization patterns of SS relative to VC by using the grading system adopted by Vescan et al.[12)] Our results showed that the most frequent pattern of pneumatization was grade 3, a canal surrounded by air for 1/3 to 2/3 of its circumference (Fig. 4C). We divided the pneumatization pattern into two subgroups, poor- and well-pneumatized. The poor-pneumatized group included grade 1 (right 28.1%–left 26.8%) and grade 2 (right 20.3%–left 22.1%), and the well-pneumatized group was composed of grade 3 (right 41.1%–left 38.9%) and grade 4 (right 10.4%–left 12.1%). These data demonstrated that almost a half of VCs might be hard to deal with during EEAs. Our results, summarized in Table 2, showed that well-pneumatized subjects in Japanese populations are slightly less frequent than those in Caucasian subjects reported by Vescan et al. with the well-pneumatized rate of 60%.

VC protrusion has been recognized unilaterally or bilaterally in approximately one-third of radiological studies already published.[12,18,20)] Our rate was 52.5% on the right side and 51.9% on the left side, confirming the more frequent rate of VC protrusion in Japanese. We further found a complete protrusion of the canal in 11.2% on the right side and in 11.7% on the left side. The pneumatization degree of SS and the protrusion of the canal are two important factors. The more pneumatization, the less bone drilling will be required[12)]; and an evident protrusion could help us to locate the canal during endoscopic approaches. Thus, VC with complete protrusion into SS would be the most suitable to perform an EEA. Although poor-pneumatized and non-protruded VCs are thought to be more difficult to deal with, even in these cases VC is useful and reliable to safely localize the ICA during EEAs, because no other landmarks will help the surgeon.

The posterior end of VC was located infero-lateral to the anterior end of the carotid canal and anterior genu of the petrous ICA in all the cases (Fig. 3). No other relationship has been published.[12,13)] Thus, drilling on the infero-medial surface of VC allows the safe identification of the petrous ICA, as it is located in a superior and medial plane.[3,9)] This is an important concept to avoid an inadvertent injury to the ICA.

Although we found some differences between our data and other already published articles performed in other populations, no changes are needed to be done when an EEA is performed in Japanese population. Compiling our data we can define a standard Japanese VC by an average length of 14.5 mm, with a medial to lateral course from anterior to posterior, surrounded by air from 34% to 66% and protruding into SS. The mean distance between VC and FR is 6.8 mm, and between VC and medial pterygoid plate 8.9 mm, being medially relative to MPTG. Its posterior end is always infero-medial to ICA. These data will help the surgeons to understand the surgical anatomy and relationships of VC, allowing a safe procedure to the anterior genu of the petrous ICA.

Conclusion

Although VCs length and depth from MPTG or FR in Japanese populations are equivalent to those in Caucasians, the pneumatization of SS is slightly less frequent. Interestingly, subjects with VC protrusion into SS, the potentially suitable candidates for EEAs, are more frequently observed. Surgeons who performed EEAs especially in Japanese patients must know these anatomical characteristics for safe and efficient procedures.