Abstract
Introduction The petroclival region is an integral part of the skull base. It can harbor different pathologies and provides access to the petroclival junction and cerebellopontine angle. We present the results of the morphometric analysis of the posterior fossa and a prediction model to enable skull base surgeons to choose an optimal surgical corridor considering patient's bony anatomy.
Methods Ninety patients (14 to assess interobserver reliability) with temporal bone computed tomography were selected. Exclusion criteria included patients <18 years of age, radiographic evidence of trauma, infection, or previous surgery. The images were analyzed using OsiriX MD (Bernex, Switzerland). We recorded clival length, vertical angle, and surface area, and petroclival angle, petrous apex, and translabyrinthine corridors volume.
Results The average age was 49.5 years (55%) for males. The mean clival length and surface areas were 44.2 mm (standard deviation [SD] ± 4.1) and 8.1 cm 2 (SD ± 1.3). The mean petrous apex and translabyrinthine corridors volumes were 2.2 cm 3 (SD ± 0.6) and 10.1 cm 3 (SD ± 3.7). The mean petroclival angle at the internal auditory canal (IAC) was 154.9 degrees (SD ± 9). The clival length correlated positively with clival surface area (rho = 0.6, p <0.05), petrous apex volume (rho = 0.3, p < 0.05), and translabyrinthine volume (rho = 0.3, p < 0.05).
Conclusion The petroclival region is complex and with high variability of surgical significance. The use of preoperative measurements of the clival length and petroclival angle as part of surgical planning that could help the surgeon to choose an optimal surgical corridor by overcoming the anatomical variability elements.
Keywords: skull base, approaches, clivus, petroclival, petrous apex, prediction
Introduction
Skull base pathologies are challenging neurosurgical conditions due to their intimate relationships with critical neurovascular structures. 1 One of the tenets of skull base surgery is that of the strategic removal of bony elements as a means to create surgical corridors, thus minimizing the need for brain retraction. 1 Access to the petroclival region has been the focus of multiple reports on surgical approaches. 1 2 3 4 5 The existing algorithms for choosing skull base approaches put minimal emphasis on the variability of bony anatomy among patients and its effect on the adequacy of the different surgical corridors. In this study, we evaluated the existence of variability in the bony anatomy of the posterior fossa in adults. Besides, we attempted to identify the anatomic parameters that might affect the selection of surgical corridors to the petroclival region. The ultimate goal is to develop an objective and patient-specific method for the selection of surgical approaches.
Methods
Following approval from the local institutional board review (IRB), we screened a total of 200 consecutive thin-section head computed tomography (CT) scans of the temporal bone between January 2016 and January 2019. Exclusion criteria included: age younger than 18 years, radiographic evidence of trauma, neoplasm, inflammation/infection, previous surgery, or congenital malformation of the skull base. A total of 90 high resolution (512 × 512), 1-mm slice-thickness scans met inclusion criteria and were imported and analyzed using OsiriX MD (Bernex, Switzerland), reconstructing them into three-dimensional form after correcting for gantry angulation. Fourteen patients were used to assess the interobserver reliability (IOR) and the remaining 76 patients were used to perform the study.
Recorded parameters included patient age, gender, race, clival length, clival vertical angle, and clival maximum thickness. Additionally, we recorded the internal auditory canal (IAC) angle, clival surface area, IAC to jugular foramen (JF) distance, inter-IAC distance, inter-JF distance, petroclival angle (PCA), petrous apex volume, and translabyrinthine corridor volume. The methodology employed for the acquisition of all parameters in this study is detailed in ( Table 1 ). For the IOR, the two observers measured the variables of interest using a group of 14 patients, then we used the interclass correlation (ICC) analysis to assess the reliability of the observations. The two observers then shared the study population ( n = 76) equally.
Table 1. Measurement technique for posterior fossa anatomic parameters.
| Parameter | Measuring technique |
|---|---|
| Clival length | The distance between the basion and the posterior clinoid process in the midsagittal plane |
| Clival vertical angle | The angle between a line parallel to the posterior surface of the clivus and a line perpendicular to the plane between the hard palate and the opisthion (midsagittal) |
| Clival maximum thickness | The thickness of the clivus immediately caudal to the sphenoid rostrum in the midsagittal plane |
| IAC angle | The angle between line projects to the IAC and a line projecting along to the long axis of the skull base |
| IAC to jugular foramen distance (IAC–JF) | The distance between the midpoint of the IAC and the midpoint of the jugular foramen in the coronal plane |
| CSA (transclival corridor) | The surface area of the clivus from the anterior foramen magnum to the floor of the sella |
| Petrous apex corridor volume | The volume of the bone posterior to the carotid canal, lateral to petroclival fissure, and medial to IAC |
| Translabyrinthine corridor volume | The volume of the mastoid, lateral and posterior semicircular canal, excluding the IAC and fallopian tube |
| PCA at the level of the (PCA–CC) | The angle between two lines parallel to the dorsal surface of the petrous bone and the clivus at the level of the carotid canal |
| PCA at the level of IAC (PCA–IAC) | The angle between two lines parallel to the dorsal surface of the petrous bone and the clivus at the level of the IAC |
| Inter-IAC distance (IAC–IAC) | The distance between the midpoint of the porus acusticus on either side in the axial plane |
| Inter-JF distance (JF–JF) | The distance between the midpoint of the jugular foramen on each side in the coronal plane |
Abbreviations: CC, carotid canal; CSA, Clival surface area; IAC, internal auditory canal; JF, jugular foramen; PCA, petroclival angle.
Statistical Analysis
We reported the descriptive statistics as mean, range, and standard deviation (SD) values. We used the Wilcoxon's rank-sum test to compare continuous data and the Watson–Williams test for angular data. We used the Chi-square 2 test to compare categorical data. We identified the relationships between individual pairs of variables using the Spearman's correlation. We used univariate linear regression to create a prediction model (Y = B0 + B1 × X) that used the clival length as an input variable (X) to predict the size of the transclival, petrous apex, and translabyrinthine surgical corridors (Y). Additionally, we validated the regression model using the cross-folding technique. To assess the IOR, we used the ICC analysis. We used a significance level of p < 0.05 and statistical analysis was conducted using Stata version 13 (StataCorp, College Station, TX).
Results
Our study population included 76 patients (excluding the 14 patients that were used to assess the IOR) with an average age of 49.5 (range: 18–91) years, of which 42 were males (55%). All measurements are summarized in Table 2 , while the gender-based comparative analysis and the correlation analysis are summarized in Tables 3 and 4 .
Table 2. Normal large-scale distribution of skull base measurements.
| Variable | Mean | Range |
|---|---|---|
| Age (y) | 49.5 | 18–91 |
| Petrous apex volume (cm 3 ) | 2.2 | 1–4.1 |
| Translabyrinthine volume (cm 3 ) | 10.1 | 3.7–19.5 |
| Clival surface area (cm 2 ) (transclival corridor) | 8.1 | 5.4–11.6 |
| Clival length (mm) | 44.2 | 34–53.8 |
| Clival vertical angle (degrees) | 27.3 | 10–48 |
| Clival thickness (mm) | 15.7 | 6–26.7 |
| IAC angle (degrees) | 79.8 | 57–93.5 |
| IAC–JF (mm) | 15.2 | 9.8–21.2 |
| IAC–IAC distance (mm) | 49.7 | 38.5–58 |
| JF–JF distance (mm) | 58.5 | 49–76.3 |
| PCA at IAC (degrees) | 154.9 | 137–173.5 |
| PCA at CC (degrees) | 143.6 | 128–177 |
Abbreviations: CC, carotid canal; IAC, internal auditory canal; JF, jugular foramen; PCA, petroclival angle.
Table 3. Normal large-scale distribution of skull base measurements categorized by gender.
| Male | Female | ||||
|---|---|---|---|---|---|
| Variable | Mean | Range | Mean | Range | p -Value |
| Age (y) | 52.7 | 20–91 | 45.6 | 18–89 | 0.4 |
| Petrous apex volume (cm 3 ) | 2.3 | 1.1–4.1 | 2 | 1–3.2 | 0.02 |
| Translabyrinthine volume (cm 3 ) | 10.9 | 3.7–19.5 | 9.2 | 4–16.9 | 0.02 |
| Clival surface area (cm 2 ) (transclival corridor) | 8.7 | 6.4–11.6 | 7.3 | 5.4–9.7 | 1.1 |
| Clival length (mm) | 45.7 | 34–54 | 42.2 | 36–47 | 9.7 |
| Clival vertical angle (degrees) | 27 | 10–48 | 27.6 | 12–45 | 0.7 |
| Clival thickness (mm) | 16.2 | 9–27 | 15 | 6–25 | 0.1 |
| IAC angle (degrees) | 81.3 | 57–93.5 | 77.9 | 62–89 | 0.05 |
| IAC–JF (mm) | 15.5 | 12–21 | 14.8 | 10–20 | 0.1 |
| IAC–IAC distance (mm) | 51.3 | 44.5–58 | 47.8 | 38.5–54 | 3.7 |
| JF–JF distance (mm) | 59.3 | 49.5–76 | 57.4 | 49–71 | 0.1 |
| PCA at IAC (degrees) | 155.7 | 137–173 | 154 | 141–166 | 0.4 |
| PCA at CC (degrees) | 144.6 | 130–177 | 142.4 | 128–158.5 | 0.2 |
Abbreviations: CC, carotid canal; IAC, internal auditory canal; JF, Jugular foramen; PCA, Petroclival angle.
Table 4. Summary of the correlation analysis.
| PAC | TLC | Clival (SA) | Clival (L) | Clival (VA) | Clival (T) | IAC (A) | IAC–IAC | PCA @ IAC | PCA @ CC | JF–JF | IAC–JF | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PAC | 1.0 | 0.6 | 0.5 | 0.3 | −0.1 | 0.3 | −0.4 | 0.0 | −0.1 | −0.2 | 0.5 | |
| TLC | 0.6 | 1.0 | 0.4 | 0.3 | 0.0 | 0.2 | −0.3 | 0.2 | 0.0 | 0.5 | ||
| Clival (SA) | 0.5 | 0.4 | 1.0 | 0.6 | 0.0 | 0.3 | 0.1 | 0.3 | −0.1 | 0.0 | 0.2 | 0.4 |
| Clival (l) | 0.3 | 0.6 | 1.0 | −0.3 | 0.2 | 0.0 | 0.2 | 0.1 | −0.1 | 0.0 | 0.5 | |
| Clival (VA) | −0.1 | 0.0 | −0.3 | 1.0 | −0.1 | 0.2 | −0.1 | 0.0 | −0.2 | |||
| Clival (T) | 0.3 | 0.2 | 0.3 | 0.2 | −0.1 | 1.0 | −0.1 | 0.2 | −0.1 | 0.1 | 0.2 | |
| IAC (A) | −0.4 | −0.3 | 0.1 | 0.0 | 0.2 | −0.1 | 1.0 | 0.2 | 0.1 | 0.4 | −0.2 | |
| IAC–IAC | 0.0 | 0.2 | 0.3 | 0.2 | −0.1 | 0.2 | 1.0 | 0.2 | 0.3 | 0.1 | ||
| PCA @ IAC | 0.0 | −0.1 | 0.1 | −0.1 | 0.1 | 0.2 | 1.0 | 0.4 | 0.0 | |||
| PCA @ CC | −0.1 | 0.0 | −0.1 | 0.1 | 0.3 | 0.4 | 1.0 | 0.2 | −0.1 | |||
| JF–JF | −0.2 | 0.0 | 0.2 | 0.0 | 0.1 | 0.4 | 0.3 | 0.0 | 0.2 | 1.0 | −0.3 | |
| IAC–JF | 0.5 | 0.4 | 0.5 | −0.2 | 0.2 | −0.2 | 0.1 | 0.0 | −0.1 | −0.3 | 1.0 |
Abbreviations: A, angle; CC, carotid canal; CF, carotid foramen; IAC, internal auditory canal; JF, Jugular foramen; L, length; PAC, petrous apex corridor; PCA, petroclival angle; SA, surface area; T, thickness; TLC, translabyrinthine corridor; VA, vertical angle.
Mean values for the clival length, surface area, and vertical angle were 44.2 mm (SD ± 4.1), 8.1 cm 2 (SD ± 1.3), and 28 degrees (SD ± 8.6), respectively. The mean IAC angle, IAC–JF distance, IAC–IAC distance, and JF–JF distance were 79.8 degrees (SD ± 7.7), 15.2 mm (SD ± 2), 49.7 mm (SD ± 3.9), and 58.5 mm (SD ± 5.1) respectively. The mean petrous apex and translabyrinthine corridors volumes were 2.2 cm 3 (SD ± 0.6) and 10.1 cm 3 (SD ± 3.7), respectively. The mean PCAs at the level of the carotid canal and IAC were 143.6 (SD ± 8.9) and 154.9 (SD ± 9), respectively ( Table 2 ; Figs. 1 – 4 ).
Fig. 1.
CT scan images (sagittal view) show ( A ) clival length (short), ( B ) clival length (long).
Fig. 4.
CT scan images (coronal view) show ( A ) short IAC-JF distance, ( B ) long IAC-JF distance.
The gender-based comparative analysis showed that males had a significantly larger petrous apex (2.3 vs. 2 cm 3 , p = 0.02) and translabyrinthine (10.9 vs. 9.2 cm 3 , p = 0.02) corridors in comparison to females with no significant difference of the transclival corridor (8.7 vs. 7.3 cm 2 , p = 0.7; Table 3 ).
Correlation analysis showed that the clival length correlated positively with the clival surface area (rho = 0.6, p <0.05), IAC–JF distance (rho = 0.5, p < 0.05), petrous apex volume (rho = 0.3, p < 0.05), and translabyrinthine volume (rho = 0.3, p < 0.05. The clival vertical angle correlated positively with the IAC angle (rho = 0.2, p < 0.05) and negatively with the IAC–JF distance (rho = −0.2, p < 0.05), and clival length (rho = −0.3, p < 0.05; ( Figs. 5 and 6 ). The PCA at the levels of the IAC and carotid canal correlated positively with the IAC–IAC distance (rho = 0.2, 0.3, p < 0.05. The IAC–IAC distance correlated positively with the inter-JF distance (rho = 0.3, p <0.05; Fig. 7 ; Table 4 ).
Fig. 5.
CT scan images show ( A ) narrow clival angle (sagittal view), ( B ) narrow IAC angle (axial view).
Fig. 6.
CT scan images show ( A ) wide clival angle (sagittal view), ( B ) wide IAC angle (axial view).
Fig. 7.
A cross-correlation map shows the correlation between different anatomic variables. [CL, clival length; CSA, clival surface area; CT, clival thickness; CVA, clival vertical angle; CCD, central clival depression; IAC, internal auditory canal; PCF, petroclival fissure; JF, jugular foramen.
Fig. 2.
CT scan images show ( A ) clival surface area measurement (coronal view), ( B ) a 3-D reconstruction of the clivus.
Fig. 3.
CT scan images (axial view) show ( A ) wide petroclival angle, ( B ) narrow petroclival angle.
The linear regression model that predicts the size of the transclival (B0 = −1.5, B1 = 0.2, p < 0.05), petrous apex (B0 = −0.5, B1 = 0.06, p < 0.05), and translabyrinthine (B0 = −1.6, B1 = 0.2, p < 0.05) surgical corridors was statistically significant ( Fig. 8 ). A high degree of IOR was found and the average ICC was 0.00 with a 95% confidence interval (CI) from 0.991 to 0.999 (F [9.10.0] = 937.4, p < 0.0001).
Fig. 8.
A scatter plot shows the linear correlation between the clival length and translabyrinthine ( A ), Petrous apex ( B ), and transclival ( C ) corridors.
Discussion
The results of our study identified significant variability of the bony anatomy among patients that may affect the decisions for surgical approach selection. Notably, the clival length at the midsagittal plane was the most reliable predictor of corridor size as it showed a strong positive correlation with the size of each of the three surgical corridors (transclival, transpetrosal, and translabyrinthine).
Some authors attempted to quantify the anatomical variability of certain aspects of the petroclival region, but these studies not only had small sample sizes but also measured a limited number of petroclival region anatomic parameters, 2 3 4 5 6 thus failing to provide a more comprehensive set of data to evaluate different vectors of an approach. In our study, we found that PCAs correlated positively with each other at both levels (carotid canal and IAC). This suggests that the articulation of the petrous temporal bone with the clivus is the main determinant of the size of the PCAs. Abdel Aziz et al 2 found comparable results in their anatomic study of the PCA at the level of IAC. Also, a change in the PCA was associated with a change in the inter-IACs and inter-JFs distances. We also found that the length of the clivus in the midsagittal plane correlated positively with the size of the petrous apex, transclival, and translabyrinthine corridors. For example, a longer clivus was associated with a larger petrous apex, transclival, and translabyrinthine corridors, and vice versa. Conversely, a change in clival length was not associated with any morphometric changes in PCA. A narrow PCA can limit the size of the retrosigmoid and far lateral corridors due to the proximity of the petrous temporal bone to the cerebellum and brainstem. To quantify the exact effect of the change in PCA on the corridor size and cerebellar retraction a cadaveric study will be necessary, but the surgeon can use the PCA deviation from the mean to anticipate differences in the retrosigmoid or far lateral corridor size. If an alternative corridor is necessary, then the clival length can be used to predict the size of the three main alternatives and it can guide the surgeon toward the optimal corridor. Of note, there are other factors such as lesion location within the posterior fossa, size, patients' symptoms, and others that will factor in the final decision. Assessing the PCA and clival length can provide the surgeon with a patient-specific measurements which can be used to choose among the four major skull base approaches to the petroclival region (retrosigmoid, petrous apex, transclival, and translabyrinthine).
Multiple anatomic studies have evaluated the differences in the exposure provided by different surgical approaches. 7 8 9 10 These studies had a small sample size; therefore, their findings are not generalizable. Also, none of the studies examined the effects of the anatomic variability on the formation of different anatomic patterns of the petroclival region. For example, Safavi-Abbasi et al 8 in their anatomic study of the posterolateral approaches found that the far lateral extension of the standard retrosigmoid approach gives an additional exposure to the lower clivus. They didn't examine the effects of variability in the PCA on surgical exposure and the amount of cerebellar retraction needed. The target in their study was the petroclival area which they defined as the area limited by the Meckel's cave, IAC, JF, and clivus. By applying our findings, the surgeon can objectively select between an open retrosigmoid and endoscopic transclival approach to access the petroclival area by measuring the PCA and transclival window (using the clival length).
Our prediction model provides an effortless way to estimate the size of the petrous apex, transclival, and translabyrinthine corridors using any currently available clinical imaging workstation by using their respective image measuring tools. The significance of our findings, aside from application for the treatment of skull base pathology, can also extend to pain procedures like microvascular decompression for trigeminal neuralgia, where the occurrence of a narrow PCA can limit the view of the trigeminal nerve root entry zone. The ability to predict this eventuality in the preoperative setting would impact the course of the operation and selection of additional tools or maneuvers in anticipation of the procedure.
Historically, a surgeon's expertise on surgical approach selection and consideration of tumor size and extension has been the most commonly used element in decision making for the treatment of skull base pathology. Our findings indicate that variability in bony anatomy among patients results in variability in the size of surgical corridors. Therefore, in addition to routinely evaluated preoperative parameters, it appears important to objectively assess and consider each patients' bony anatomy in selecting a specific surgical approach. We anticipate that ongoing cadaveric studies can further expand on our findings, by creating a prediction model that can incorporate measurements of surgical freedom of movement for each corridor.
Conclusion
The petroclival region is an anatomically complex area of high variability with surgical significance. The incorporation of preoperative measurements of clival length and PCA as part of surgical planning could help overcome anatomical variability elements that impact the adequacy of surgical corridors and their selection.
Footnotes
Conflict of Interest B.J.W. reports other from Monteris medical, outside the submitted work. All the other authors report no conflict of interest.
References
- 1.Rangel-Castilla L, Russin J J, Spetzler R F. Surgical management of skull base tumors. Rep Pract Oncol Radiother. 2016;21(04):325–335. doi: 10.1016/j.rpor.2014.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Abdel Aziz K M, Sanan A, van Loveren H R, Tew J M, Jr, Keller J T, Pensak M L.Petroclival meningiomas: predictive parameters for transpetrosal approaches Neurosurgery 20004701139–150., discussion 150–152 [PubMed] [Google Scholar]
- 3.Desai S K, Baymon D, Sieloff E. Computed tomography morphometric analysis of the central clival depression and petroclival angle for application of the presigmoid approach in the pediatric population. J Pediatr Neurosci. 2016;11(02):109–111. doi: 10.4103/1817-1745.187625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mason E, Van Rompaey J, Carrau R, Panizza B, Solares C A. Anatomical and computed tomographic analysis of the transcochlear and endoscopic transclival approaches to the petroclival region. Laryngoscope. 2014;124(03):628–636. doi: 10.1002/lary.24378. [DOI] [PubMed] [Google Scholar]
- 5.Siwanuwatn R, Deshmukh P, Figueiredo E G, Crawford N R, Spetzler R F, Preul M C. Quantitative analysis of the working area and angle of attack for the retrosigmoid, combined petrosal, and transcochlear approaches to the petroclival region. J Neurosurg. 2006;104(01):137–142. doi: 10.3171/jns.2006.104.1.137. [DOI] [PubMed] [Google Scholar]
- 6.Jacquesson T, Berhouma M, Tringali S, Simon E, Jouanneau E. Which routes for petroclival tumors? A comparison between the anterior expanded endoscopic endonasal approach and lateral or posterior routes. World Neurosurg. 2015;83(06):929–936. doi: 10.1016/j.wneu.2015.02.003. [DOI] [PubMed] [Google Scholar]
- 7.Chang S W, Wu A, Gore P.Quantitative comparison of Kawase's approach versus the retrosigmoid approach: implications for tumors involving both middle and posterior fossaeNeurosurgery 2009;64(3, suppl)ons44–ons51, discussion ons51–ons52 [DOI] [PubMed]
- 8.Safavi-Abbasi S, de Oliveira J G, Deshmukh P. The craniocaudal extension of posterolateral approaches and their combination: a quantitative anatomic and clinical analysis. Oper Neurosurg (Hagerstown) 2010;66 01:54–64. doi: 10.1227/01.NEU.0000354366.48105.FE. [DOI] [PubMed] [Google Scholar]
- 9.Sharma M, Ambekar S, Guthikonda B, Nanda A. A Comparison between the Kawase and extended retrosigmoid approaches (retrosigmoid transtentorial and retrosigmoid intradural suprameatal approaches) for accessing the petroclival tumors. A cadaveric study. J Neurol Surg B Skull Base. 2014;75(03):171–176. doi: 10.1055/s-0033-1359305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tang C T, Kurozumi K, Pillai P, Filipce V, Chiocca E A, Ammirati M. Quantitative analysis of surgical exposure and maneuverability associated with the endoscope and the microscope in the retrosigmoid and various posterior petrosectomy approaches to the petroclival region using computer tomograpy-based frameless stereotaxy. A cadaveric study. Clin Neurol Neurosurg. 2013;115(07):1058–1062. doi: 10.1016/j.clineuro.2012.10.023. [DOI] [PubMed] [Google Scholar]