Neurosurg Rev DOI 10.1007/s10143-015-0634-2
ORIGINAL ARTICLE
Image-guided, microsurgical topographic anatomy of the endolymphatic sac and vestibular aqueduct via a suboccipital retrosigmoid approach Roberto Colasanti 1,2 & Al-Rahim Abbasali Tailor 1 & Jun Zhang 3 & Mario Ammirati 1
Received: 9 August 2014 / Revised: 3 February 2015 / Accepted: 14 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract The endolymphatic sac (ES) and the vestibular aqueduct (VA) are often in the surgical field when posterior fossa lesions are targeted using retrosigmoid approaches. The purpose of this work is to validate neuronavigator accuracy in predicting VA location as well as to give guidelines to preserve the ES and VA. A retrosigmoid approach was performed bilaterally in six specimens in the semisitting position. Preoperatively, we registered in the CT scans the position of the VA genu (virtual genu). After the approach execution, ES and VA genu topographic relationships with evident posterolateral cranial base structures were measured using neuronavigation. Next, we exposed the VA genu: its position coincided with the virtual VA genu in all the specimens. On the average, the ES was 17.93 mm posterosuperolateral to the XI nerve in the jugular foramen, 12.26 mm posterolateral to the internal acoustic meatus, 20.13 mm anteromedial to the petro-sigmoid intersection at a point 13.30 mm inferior to the petrous ridge. The VA genu was located 7.23 mm posterolateral to the internal acoustic meatus, 18.11 mm superolateral to the XI nerve in the jugular foramen, 10.27 mm inferior to
* Mario Ammirati [email protected] 1
Dardinger Microneurosurgical Skull Base Laboratory, Department of Neurological Surgery, Wexner Medical Center, The Ohio State University, N1025 Doan Hall, 410 West 10th Avenue, Columbus, OH 43210, USA
2
Department of Neurosurgery, Umberto I General Hospital, Università Politecnica delle Marche, Ancona, Italy
3
Department of Radiology and Wright Center of Innovation in Biomedical Imaging, Wexner Medical Center, The Ohio State University, Columbus, OH, USA
the petrous ridge, and 6.28 mm anterolateral to the endolymphatic ledge at a depth of 3.46 mm from the posterior pyramidal wall. Our study demonstrates that is possible to use neuronavigation to reliably predict the location of the VA genu. In addition, neuronavigation may be effectively used to create a topographical framework that may help maintaining the integrity of the ES/VA during retrosigmoid approaches. Keywords Endolymphatic sac . Hearing preservation . Intraoperative navigation . Microsurgical anatomy . Retrosigmoid approach . Vestibular aqueduct
Introduction Drilling of the temporal bone is often necessary during skull base procedures. The area between the internal acoustic meatus and the sigmoid sinus is frequently drilled, and it houses two important structures, the endolymphatic sac (ES) and endolymphatic duct (ED). While the ES is identifiable, being contained in a dural duplication, the ED is not, as it is embedded in the temporal bone. Though there is no consensus on whether or not manipulation or damage to the ES can cause inner ear malfunction, damage to the ED is associated with hearing compromise [2, 5, 14]. Several papers have given anatomical topographical landmarks that are useful to preserve these two structures [1, 2, 4, 6–8, 14]. As neuronavigator use has become standard in neurosurgical procedure, we hypothesized that the use of neuronavigators could accurately predict the location of the vestibular aqueduct (VA), which is the bony tunnel surrounding the ED. We tested this hypothesis on cadavers by identifying the VA genu using the neuronavigator CT dataset (virtual genu) and then exposing the real genu and
Neurosurg Rev
determining its correlation with the virtual one. In the same cadavers, we also measured, always using the neuronavigator, distances between surgically exposed readily identifiable landmarks evident during a suboccipital retrosigmoid approach and the VA genu/ES (Fig. 1). The purpose of these measurements was to give neuronavigationverified guidelines to preserve the ED and ES during retrosigmoid approaches.
Materials and methods Six embalmed adult cadaveric heads were used in this study providing 12 sides for dissection and measurements. Highresolution computed tomographic scans (slice thickness, 0.6 mm) with bone fiducials were obtained prior to dissection and utilized for intraoperative navigation (Stryker Instruments, Kalamazoo, MI, USA) according to a previously published protocol [11]. The procedures were performed using standard microsurgical instruments, a high-speed drill (Midas Rex, Medtronic, Inc., Minneapolis, MN, USA), a surgical microscope (Moller,
Fig. 1 Drawing depicting the topographic relationships between the endolymphatic sac, the vestibular aqueduct genu, and readily evident posterolateral skull base landmarks. Blue arrows point to the distances between these landmarks and the endolymphatic sac. Yellow arrows illustrate their relationships with the vestibular aqueduct genu, 1=endolymphatic sac; 2=uppermost portion of the vestibular aqueduct genu; 3=petrous ridge; 4=internal acoustic meatus; 5=petro-sigmoid intersection; 6=XI nerve in the jugular foramen; 7=posterior pyramidal wall; 8 = sigmoid sinus; 9 = VII–VIII complex; 10 = cochlea; 11=intrapetrous internal carotid artery; 12=V nerve; 13=common crus; 14 = posterior semicircular canal; 15 = superior semicircular canal; 16=lateral semicircular canal (adapted with permission from AANS: Journal of Neurosurgery, Ammirati M, Ma J, Cheatham ML, Maxwell D, Bloch J, Becker DP., Drilling the posterior wall of the petrous pyramid: a microneurosurgical anatomical study., 1993 Mar;78(3):452–5)
Wedel, Germany), and Budde Halo retractor system (Omi Surgical Products, Cincinnati, OH, USA).
Surgical procedure The head was placed in a Mayfield fixation device in the semisitting position with slight anterior flexion. The cadaver was registered to the navigation system, and the putative position of the genu of the VA (virtual VA) was identified as a landmark. A standard suboccipital retrosigmoid craniotomy was performed exposing the inferior and medial edges of the transverse and sigmoid sinuses, respectively. The dura was opened via a c-shaped incision based on the sigmoid sinus. Next, the arachnoid membrane forming the posterior wall of the cerebellomedullary and the cerebellopontine cisterns was opened identifying the lower cranial nerves as well as the VII–VIII complex entering the internal auditory canal. A cerebellar retractor was used to support the cerebellum to simulate operative conditions. Next, the ES was identified visually and by palpation with a dissector before and after the excision of the dura mater covering the posterior surface of the petrous pyramid. In addition, intraoperative navigation was used to confirm its position as well as to visualize the putative course of the VA (Fig. 2). Then, the navigation system was used to measure the shortest distance from the preoperatively defined landmark (i.e., the uppermost portion of the VA genu) to the posterior pyramidal wall, the XI nerve in the jugular foramen, the petrous ridge, the posterior lip of the internal acoustic meatus and the upper limit of the endolymphatic ledge. Moreover, we recorded the shortest distance from the center of the ES to the XI nerve in the jugular foramen and to the posterior lip of the internal acoustic meatus. Next, the shortest distance from the center of the ES and the petrous ridge was measured as well as the distance from the same point on the petrous ridge to the petro-sigmoid intersection (i.e., the point where the sigmoid sinus intersects the petrous ridge). Finally, the height and the width of the ES were also registered. Every measurement was taken five times, and the mean of the measurements was used for final analysis. All distances were calculated using the three-dimensional distance formula in which the distance between points (X1, Y1, Z1) and (X2, Y2, Z2) can be defined as: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi d ¼ ðX 1 −X 2 Þ2 þ ðY 1 −Y 2 Þ þ ðZ 1 −Z 2 Þ2 Next, we exposed the genu of the VA and placed the tip of the neuronavigator on it and checked the position of the VA genu surgically exposed (real VA genu) versus the virtual VA genu (Fig. 3).
Neurosurg Rev Fig. 2 Microscopic photograph (a) and corresponding intraoperative navigation image (b) obtained while placing the neuronavigator probe at the level of the endolymphatic depression
Results The position of the surgically exposed VA genu (real VA genu) coincided with one of the virtual VA genu in all the specimens. Figures 2 and 3 show the ES and VA genua identified in the operative field and their correspondent neuronavigation position. Our results are summarized in Tables 1 and 2. On the average, the center of the ES was 17.93 mm posterosuperolateral
to the XI nerve in the jugular foramen and 12.26 mm posterolateral to the internal acoustic meatus. Moreover, the center of the ES was located 20.13 mm anteromedial to the petro-sigmoid intersection at a point 13.30 mm inferior to the petrous ridge (mean distances for both measurements). The median width and height were, respectively, 4.19 mm (mean, 4.28 mm) and 4.27 mm (mean, 4.38 mm) thus conferring to the ES a circular shape in most cases.
Neurosurg Rev Fig. 3 Microscopic photograph (a) and corresponding intraoperative navigation image (b) obtained while placing the neuronavigator probe at the level of the vestibular aqueduct genu. The vestibular aqueduct has been partially opened
On the average, the uppermost portion of the VA genu was located 7.23 mm posterolateral to the internal acoustic meatus, 18.11 mm superolateral to the XI nerve in the jugular foramen, 10.27 mm inferior to the petrous ridge, and 6.28 mm anterolateral to the endolymphatic ledge at a depth of 3.46 mm from the posterior pyramidal wall.
Discussion The ES and VA are often in the surgical field when posterior fossa lesions are targeted using the retrosigmoid approach. The ES may be divided in two parts. The proximal portion, pars rugosa, is essentially a continuation of the ED and is
Neurosurg Rev Table 1
Topographic anatomy of the endolymphatic sac
Distance
Mean (mm)
Median (mm)
Range (mm)
IAM-ES XI-ES Ridge-ES PSI-ESa ES (width/height)
12.26 17.93 13.30 20.13 4.28/4.38
12.20 17.59 14.43 19.63 4.19/4.27
10.06–15.48 16.48–19.60 7.14–15.72 18.08–23.10 2.80–7.06/3.48–5.28
ES center of the endolymphatic sac, IAM posterior margin of the internal acoustic meatus, PSI petro-sigmoid intersection, Ridge petrous ridge, XI XI nerve in the jugular foramen a
This distance was measured from the petrous-sigmoid intersection to the same point on the petrous ridge analyzed in the previous row
partially covered by the bony operculum of the ES. The distal smooth portion is extraosseus and resides in a fold between the two layers of dura terminating near the sigmoid sinus [2]. Recent histologic papers have described the ES as a highly complex structure of interconnecting tubules, cisterns, and crypts [10]. The VA travels vertically inside the posterior petrous bone and contains the ED, which connects the ES to the utricular and saccular ducts [8, 10]. The ES/ED system plays an important role in inner ear dynamics since it is involved in endolymph resorption and production. Its malfunction, may determine the symptoms of Meniere disease (endolymphatic hydrops). It is generally believed that while injury to the ES may or may not be associated with hearing or vestibular problems injury to the ED are linked to significant hearing compromise [2, 3, 5, 14]. During removal of cerebellopontine angle tumors, the ES and the ED may be injured when the dural flap is created as well as during the drilling of the posterior wall of the petrous pyramid in order to accomplish a wide exposure of the internal acoustic canal and of the infralabyrinthine region [14].
Table 2 Topographic anatomy of the uppermost portion of the vestibular aqueduct genu Distance
Mean (mm)
Median (mm)
Range (mm)
IAM-VA XI-VA Ridge-VA Endolymphatic ledge-VA PPW-VA
7.23 18.11 10.27 6.28 3.46
7.10 18.62 9.97 6.25 3.33
6.30–9.30 15.00–20.62 7.20–13.30 4.5–8.5 2.20–5.00
IAM posterior margin of the internal acoustic meatus, Endolymphatic Ledge upper limit of the endolymphatic ledge, PPW posterior pyramidal wall, Ridge petrous ridge, VA uppermost portion of the vestibular aqueduct genu, XI XI nerve in the jugular foramen
While the ES is usually recognizable in a dural duplication on the posterior pyramidal wall, the ED is embedded inside the petrous pyramid; hence, its preservation during surgical procedures is more challenging. Since neuronavigators have gained more and more popularity in the last 20 years in neurosurgery, we reasoned that it would be worthwhile to validate their accuracy in predicting the location of the VA, which is usually identifiable on computed tomography. Our prosections confirmed that neuronavigators may safely guide the drilling around the VA. Indeed, the position of the surgically exposed VA genu (real VA genu) coincided with the one that was identified using the neuronavigator CT dataset (virtual VA genu) in all the specimens. In other words, we demonstrated in a laboratory setting that a neuronavigator may represent a reliable tool for identifying the location of a sub-millimetric structure like the VA: this finding may result in a direct benefit in the operative theater. Moreover, the knowledge of the topographic relationships between the ES, the VA, and anatomic landmarks readily identifiable during a suboccipital retrosigmoid intradural approach may help to determine where the ES and the VA are located even in the absence of neuronavigation. Clearly in a clinical setting, the location and the extension of the tumor, the resulting distortion of the petrous bone anatomy together with the anatomic individual variability will determine the area that needs to be exposed as well as the extent of safe drilling. Nonetheless, we believe the topographic data presented may strengthen the knowledge of temporal bone anatomic features thereby providing general guidelines for a retrosigmoid labyrinth sparing petrosectomy. Our study shows that in most cases, the ES occupied a region of approximately 4.00 mm in diameter on the posterior pyramidal wall. Moreover, in 83.3 % of our specimens, the center of the ES was located between 10.88 and 14.66 mm posterolateral to the internal acoustic meatus, between 16.54 and 19.26 mm posterosuperolateral to the XI nerve in the jugular foramen, at a point between 11.26 and 15.58 mm inferior to the petrous ridge. A careful examination of the region defined by the above landmarks may allow the identification of the dural duplication in which the ES is located as well as of its bony ledge. The senior author, in a previous paper under different conditions (caliper measurements), observed slightly shorter values on the distance from the ES to the XI nerve in the jugular foramen (mean, 15.70 mm; range, 11.00–18.50 mm) and to the petrous ridge (mean, 11.50 mm; range, 8.00– 17.00 mm) and longer values on the distance between the petro-sigmoid intersection and the point on the petrous ridge that was the closest to the center of the ES (mean, 24.10 mm; range, 20.00–28.00 mm) [2].
Neurosurg Rev Fig. 4 Preoperative gadoliniumenhanced axial MRI of a 59-year-old female with disabling vertigo and a 2-mm intracanalicular vestibular schwannoma unresponsive to gamma knife therapy (a). Postoperative gadoliniumenhanced axial MRI showing complete tumor removal with preservation of preoperative hearing (b). Axial CT images demonstrating extensive drilling of the posterior meatal wall with preservation of vestibular aqueduct, arrows (c and d)
Our results on the distance between the ES and the posterior meatal lip are in accordance with previous investigations [2, 7, 14]. Geurkink [7] and Shimizu [14] reported shorter distances because they measured from the meatal lip to the bony ledge of the ES and not to its center. Moreover, our values on the size of the ES reflect the great variability observed in previous studies [2, 6]. The VA genu is usually located more medially than the common crus and, for this reason, during drilling, it is usually exposed before the common crus [14]. In 83.3 % of our specimens, the uppermost portion of the VA genu was encountered between 15.50 and 20.30 mm superolateral to the XI nerve in the jugular foramen, between 8.33 and 12.88 mm inferior to the petrous ridge, and between 4.60 and 7.60 mm anterolateral to the endolymphatic ledge. Furthermore, on the average, 7.23 mm of petrous bone removal may be carried out posterior to the internal acoustic meatus without damaging the VA. At this level, it was found to be at a depth between 2.60 and 4.50 mm from the posterior pyramidal wall in 83.3 % of cases. Shimizu noted similar distances between the posterior meatal lip and the VA (mean, 7.51 mm), but the low value reported (4.68 mm) was smaller than ours (6.30 mm) [14].
Regarding the depth of the petrous bone needed to expose the uppermost portion of the VA, Ammirati reported values similar (mean, 3.20 mm; range, 2.60–4.00 mm) to ours (mean, 3.46 mm; 2.20–5.00 mm) while Shimizu reported shorter ones (mean, 1.48 mm; range, 0.60–3.03 mm) than ours [1, 14]. Previous clinical and laboratory studies have proposed to guide the posterior meatal wall drilling using an endoscope [9], a frameless stereotaxy system [13], or both [12] in order to maximize its unroofing without violating the integrity of the labyrinth or injuring the cochlear nerve. Other authors have advised to gain an optimized working angle on the internal acoustic meatus through a medial extension of the classic retrosigmoid craniectomy in order to expose the intrameatal portion of the tumor while at the same time preserving the VA [5]. However, the knowledge of the ES and VA topographic relationships we provide, combined with a careful assessment of the preoperative imaging, may facilitate this petrous bone removal. Moreover, considering that the ES is the inferiormost labyrinthine structure in the region behind the internal acoustic meatus, its identification may also make safer and more extensive the drilling of the inframeatal/infralabyrinthine area. In conclusion, our study demonstrates that is possible to use neuronavigation to reliably predict the location of the VA genu. In addition, neuronavigation may be effectively
Neurosurg Rev
used to create a topographical framework that may help maintain the integrity of the ES/VA during retrosigmoid approaches (Fig. 4).
13.
14. Conflict of interest The authors report no conflict of interest.
Samii A, Brinker T, Kaminsky J, Lanksch WR, Samii M (2000) Navigation-guided opening of the internal auditory canal via the retrosigmoid route for acoustic neuroma surgery: cadaveric, radiological, and preliminary clinical study. Neurosurgery 47(2):382– 387, discussion 388 Shimizu S, Tanaka R, Oka H, Fujii K (2006) Risk of damage to the endolymphatic sac and duct during removal of the posterior meatal wall: an anatomic study. Neurosurgery 59(4 Suppl 2):ONS435– ONS439, discussion ONS439–440
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Ammirati M, Ma J, Cheatham ML, Maxwell D, Bloch J, Becker DP (1993) Drilling the posterior wall of the petrous pyramid: a microneurosurgical anatomical study. J Neurosurg 78(3):452–455 Ammirati M, Spallone A, Feghali J, Ma J, Cheatham M, Becker D (1995) The endolymphatic sac: microsurgical topographic anatomy. Neurosurgery 36(2):416–419 Arenberg IK, Marovitz WF, Shambaugh GE Jr (1970) The role of the endolymphatic sac in the pathogenesis of endolymphatic hydrops in man. Acta Otolaryngol Suppl 275:1–49 Domb GH, Chole RA (1980) Anatomical studies of the posterior petrous apex with regard to hearing preservation in acoustic neuroma removal. Laryngoscope 90(11 Pt 1):1769–1776 Ebner FH, Kleiter M, Danz S, Ernemann U, Hirt B, Löwenheim H, Roser F, Tatagiba M (2014) Topographic changes in petrous bone anatomy in the presence of a vestibular schwannoma and implications for the retrosigmoid transmeatal approach. Neurosurgery. doi:10.1227/NEU.0000000000000454 Friberg U, Jansson B, Rask-Andersen H, Bagger-Sjöbäck D (1988) Variations in surgical anatomy of the endolymphatic sac. Arch Otolaryngol Head Neck Surg 114(4):389–394 Geurkink NA (1977) Surgical anatomy of the temporal bone posterior to the internal auditory canal: an operative approach. Laryngoscope 87(6):975–986 Kartush JM, Telian SA, Graham MD, Kemink JL (1986) Anatomic basis for labyrinthine preservation during posterior fossa acoustic tumor surgery. Laryngoscope 96(9 Pt 1):1024–1028 King WA, Wackym PA (1999) Endoscope-assisted surgery for acoustic neuromas (vestibular schwannomas): early experience using the rigid Hopkins telescope. Neurosurgery 44(5):1095–1100, discussion 1100–1102 Lo WW, Daniels DL, Chakeres DW, Linthicum FH, Ulmer JL, Mark LP, Swartz JD (1997) The endolymphatic duct and sac. AJNR Am J Neuroradiol 18(5):881–887 Pillai P, Sammet S, Ammirati M (2008) Application accuracy of computed tomography-based, image-guided navigation of temporal bone. Neurosurgery 63(4 Suppl 2):326–332, discussion 332–333 Pillai P, Sammet S, Ammirati M (2009) Image-guided, endoscopicassisted drilling and exposure of the whole length of the internal auditory canal and its fundus with preservation of the integrity of the labyrinth using a retrosigmoid approach: a laboratory investigation. Neurosurgery 65(6 Suppl):53–59, discussion 59
Comments Nicholas C. Bambakidis, Cleveland, USA This is an anatomic paper investigating the usefulness of neuronavigation in aiding identification of and preservation of the endolymphatic sac (ES) and vestibular aquaduct (VA). The authors have produced a nice set of data from their cadaveric dissections; as with all such papers, the actual utility of the data is dubious. Nevertheless, we have found that careful study of the anatomic relationships between these landmarks can be of use in many surgical procedures around the internal auditory canal and cerebellopontine angle. Atul Goel, Mumbai, India The endolymphatic sac and duct are important structures that innocuously occupy a position in the petrous bone limited anteriorly by the 7th to 11th nerve complexes, superiorly by the petrous ridge and posteriorly and inferiorly by the sigmoid sinus. Though the retrosigmoid suboccipital approach is a very commonly used approach, the anatomy and preservation of the endolymphatic sac and duct is not a common subject of discussion. The authors have elaborately discussed the anatomy of the endolymphatic system and have elegantly showed the use of neuronavigation to identify and preserve these structures. The significance of endolymphatic sac and vestibular aqueduct is unclear even to an otolaryngologist. The exact role this anatomical structure plays in hearing and in ‘vertigo’ is more of speculation and individual conjectures. Translabyrinthine approaches will certainly sacrifice the ES and VA. Retrosigmoid operations have the advantage that endolymphatic sac and VA can be preserved. Currently, my surgical strategy in vestibular schwannomas is to avoid drilling of the internal acoustic meatus in most cases. In general, drilling of the internal acoustic meatus should be minimized to 5–6 mm posterolaterally. This avoids injury to the vestibular apparatus, the endolymphatic structures, and a high-riding jugular bulb and avoids opening of the air cells. Ammirati et al. present a cadaveric study and show the advantages of using navigation system to identify the structure and preserve it. Although technological assistance is always welcome, the well-defined anatomic landmarks and parameters that are the basis of all temporal bone surgery can be foolproof. Essentially, while technology is helpful, the importance of anatomical understanding cannot be ignored. Despite this, the endolymphatic sac is a well-defined anatomical entity and should be preserved wherever and whenever possible.
ORIGINAL ARTICLE
Image-guided, microsurgical topographic anatomy of the endolymphatic sac and vestibular aqueduct via a suboccipital retrosigmoid approach Roberto Colasanti 1,2 & Al-Rahim Abbasali Tailor 1 & Jun Zhang 3 & Mario Ammirati 1
Received: 9 August 2014 / Revised: 3 February 2015 / Accepted: 14 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract The endolymphatic sac (ES) and the vestibular aqueduct (VA) are often in the surgical field when posterior fossa lesions are targeted using retrosigmoid approaches. The purpose of this work is to validate neuronavigator accuracy in predicting VA location as well as to give guidelines to preserve the ES and VA. A retrosigmoid approach was performed bilaterally in six specimens in the semisitting position. Preoperatively, we registered in the CT scans the position of the VA genu (virtual genu). After the approach execution, ES and VA genu topographic relationships with evident posterolateral cranial base structures were measured using neuronavigation. Next, we exposed the VA genu: its position coincided with the virtual VA genu in all the specimens. On the average, the ES was 17.93 mm posterosuperolateral to the XI nerve in the jugular foramen, 12.26 mm posterolateral to the internal acoustic meatus, 20.13 mm anteromedial to the petro-sigmoid intersection at a point 13.30 mm inferior to the petrous ridge. The VA genu was located 7.23 mm posterolateral to the internal acoustic meatus, 18.11 mm superolateral to the XI nerve in the jugular foramen, 10.27 mm inferior to
* Mario Ammirati [email protected] 1
Dardinger Microneurosurgical Skull Base Laboratory, Department of Neurological Surgery, Wexner Medical Center, The Ohio State University, N1025 Doan Hall, 410 West 10th Avenue, Columbus, OH 43210, USA
2
Department of Neurosurgery, Umberto I General Hospital, Università Politecnica delle Marche, Ancona, Italy
3
Department of Radiology and Wright Center of Innovation in Biomedical Imaging, Wexner Medical Center, The Ohio State University, Columbus, OH, USA
the petrous ridge, and 6.28 mm anterolateral to the endolymphatic ledge at a depth of 3.46 mm from the posterior pyramidal wall. Our study demonstrates that is possible to use neuronavigation to reliably predict the location of the VA genu. In addition, neuronavigation may be effectively used to create a topographical framework that may help maintaining the integrity of the ES/VA during retrosigmoid approaches. Keywords Endolymphatic sac . Hearing preservation . Intraoperative navigation . Microsurgical anatomy . Retrosigmoid approach . Vestibular aqueduct
Introduction Drilling of the temporal bone is often necessary during skull base procedures. The area between the internal acoustic meatus and the sigmoid sinus is frequently drilled, and it houses two important structures, the endolymphatic sac (ES) and endolymphatic duct (ED). While the ES is identifiable, being contained in a dural duplication, the ED is not, as it is embedded in the temporal bone. Though there is no consensus on whether or not manipulation or damage to the ES can cause inner ear malfunction, damage to the ED is associated with hearing compromise [2, 5, 14]. Several papers have given anatomical topographical landmarks that are useful to preserve these two structures [1, 2, 4, 6–8, 14]. As neuronavigator use has become standard in neurosurgical procedure, we hypothesized that the use of neuronavigators could accurately predict the location of the vestibular aqueduct (VA), which is the bony tunnel surrounding the ED. We tested this hypothesis on cadavers by identifying the VA genu using the neuronavigator CT dataset (virtual genu) and then exposing the real genu and
Neurosurg Rev
determining its correlation with the virtual one. In the same cadavers, we also measured, always using the neuronavigator, distances between surgically exposed readily identifiable landmarks evident during a suboccipital retrosigmoid approach and the VA genu/ES (Fig. 1). The purpose of these measurements was to give neuronavigationverified guidelines to preserve the ED and ES during retrosigmoid approaches.
Materials and methods Six embalmed adult cadaveric heads were used in this study providing 12 sides for dissection and measurements. Highresolution computed tomographic scans (slice thickness, 0.6 mm) with bone fiducials were obtained prior to dissection and utilized for intraoperative navigation (Stryker Instruments, Kalamazoo, MI, USA) according to a previously published protocol [11]. The procedures were performed using standard microsurgical instruments, a high-speed drill (Midas Rex, Medtronic, Inc., Minneapolis, MN, USA), a surgical microscope (Moller,
Fig. 1 Drawing depicting the topographic relationships between the endolymphatic sac, the vestibular aqueduct genu, and readily evident posterolateral skull base landmarks. Blue arrows point to the distances between these landmarks and the endolymphatic sac. Yellow arrows illustrate their relationships with the vestibular aqueduct genu, 1=endolymphatic sac; 2=uppermost portion of the vestibular aqueduct genu; 3=petrous ridge; 4=internal acoustic meatus; 5=petro-sigmoid intersection; 6=XI nerve in the jugular foramen; 7=posterior pyramidal wall; 8 = sigmoid sinus; 9 = VII–VIII complex; 10 = cochlea; 11=intrapetrous internal carotid artery; 12=V nerve; 13=common crus; 14 = posterior semicircular canal; 15 = superior semicircular canal; 16=lateral semicircular canal (adapted with permission from AANS: Journal of Neurosurgery, Ammirati M, Ma J, Cheatham ML, Maxwell D, Bloch J, Becker DP., Drilling the posterior wall of the petrous pyramid: a microneurosurgical anatomical study., 1993 Mar;78(3):452–5)
Wedel, Germany), and Budde Halo retractor system (Omi Surgical Products, Cincinnati, OH, USA).
Surgical procedure The head was placed in a Mayfield fixation device in the semisitting position with slight anterior flexion. The cadaver was registered to the navigation system, and the putative position of the genu of the VA (virtual VA) was identified as a landmark. A standard suboccipital retrosigmoid craniotomy was performed exposing the inferior and medial edges of the transverse and sigmoid sinuses, respectively. The dura was opened via a c-shaped incision based on the sigmoid sinus. Next, the arachnoid membrane forming the posterior wall of the cerebellomedullary and the cerebellopontine cisterns was opened identifying the lower cranial nerves as well as the VII–VIII complex entering the internal auditory canal. A cerebellar retractor was used to support the cerebellum to simulate operative conditions. Next, the ES was identified visually and by palpation with a dissector before and after the excision of the dura mater covering the posterior surface of the petrous pyramid. In addition, intraoperative navigation was used to confirm its position as well as to visualize the putative course of the VA (Fig. 2). Then, the navigation system was used to measure the shortest distance from the preoperatively defined landmark (i.e., the uppermost portion of the VA genu) to the posterior pyramidal wall, the XI nerve in the jugular foramen, the petrous ridge, the posterior lip of the internal acoustic meatus and the upper limit of the endolymphatic ledge. Moreover, we recorded the shortest distance from the center of the ES to the XI nerve in the jugular foramen and to the posterior lip of the internal acoustic meatus. Next, the shortest distance from the center of the ES and the petrous ridge was measured as well as the distance from the same point on the petrous ridge to the petro-sigmoid intersection (i.e., the point where the sigmoid sinus intersects the petrous ridge). Finally, the height and the width of the ES were also registered. Every measurement was taken five times, and the mean of the measurements was used for final analysis. All distances were calculated using the three-dimensional distance formula in which the distance between points (X1, Y1, Z1) and (X2, Y2, Z2) can be defined as: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi d ¼ ðX 1 −X 2 Þ2 þ ðY 1 −Y 2 Þ þ ðZ 1 −Z 2 Þ2 Next, we exposed the genu of the VA and placed the tip of the neuronavigator on it and checked the position of the VA genu surgically exposed (real VA genu) versus the virtual VA genu (Fig. 3).
Neurosurg Rev Fig. 2 Microscopic photograph (a) and corresponding intraoperative navigation image (b) obtained while placing the neuronavigator probe at the level of the endolymphatic depression
Results The position of the surgically exposed VA genu (real VA genu) coincided with one of the virtual VA genu in all the specimens. Figures 2 and 3 show the ES and VA genua identified in the operative field and their correspondent neuronavigation position. Our results are summarized in Tables 1 and 2. On the average, the center of the ES was 17.93 mm posterosuperolateral
to the XI nerve in the jugular foramen and 12.26 mm posterolateral to the internal acoustic meatus. Moreover, the center of the ES was located 20.13 mm anteromedial to the petro-sigmoid intersection at a point 13.30 mm inferior to the petrous ridge (mean distances for both measurements). The median width and height were, respectively, 4.19 mm (mean, 4.28 mm) and 4.27 mm (mean, 4.38 mm) thus conferring to the ES a circular shape in most cases.
Neurosurg Rev Fig. 3 Microscopic photograph (a) and corresponding intraoperative navigation image (b) obtained while placing the neuronavigator probe at the level of the vestibular aqueduct genu. The vestibular aqueduct has been partially opened
On the average, the uppermost portion of the VA genu was located 7.23 mm posterolateral to the internal acoustic meatus, 18.11 mm superolateral to the XI nerve in the jugular foramen, 10.27 mm inferior to the petrous ridge, and 6.28 mm anterolateral to the endolymphatic ledge at a depth of 3.46 mm from the posterior pyramidal wall.
Discussion The ES and VA are often in the surgical field when posterior fossa lesions are targeted using the retrosigmoid approach. The ES may be divided in two parts. The proximal portion, pars rugosa, is essentially a continuation of the ED and is
Neurosurg Rev Table 1
Topographic anatomy of the endolymphatic sac
Distance
Mean (mm)
Median (mm)
Range (mm)
IAM-ES XI-ES Ridge-ES PSI-ESa ES (width/height)
12.26 17.93 13.30 20.13 4.28/4.38
12.20 17.59 14.43 19.63 4.19/4.27
10.06–15.48 16.48–19.60 7.14–15.72 18.08–23.10 2.80–7.06/3.48–5.28
ES center of the endolymphatic sac, IAM posterior margin of the internal acoustic meatus, PSI petro-sigmoid intersection, Ridge petrous ridge, XI XI nerve in the jugular foramen a
This distance was measured from the petrous-sigmoid intersection to the same point on the petrous ridge analyzed in the previous row
partially covered by the bony operculum of the ES. The distal smooth portion is extraosseus and resides in a fold between the two layers of dura terminating near the sigmoid sinus [2]. Recent histologic papers have described the ES as a highly complex structure of interconnecting tubules, cisterns, and crypts [10]. The VA travels vertically inside the posterior petrous bone and contains the ED, which connects the ES to the utricular and saccular ducts [8, 10]. The ES/ED system plays an important role in inner ear dynamics since it is involved in endolymph resorption and production. Its malfunction, may determine the symptoms of Meniere disease (endolymphatic hydrops). It is generally believed that while injury to the ES may or may not be associated with hearing or vestibular problems injury to the ED are linked to significant hearing compromise [2, 3, 5, 14]. During removal of cerebellopontine angle tumors, the ES and the ED may be injured when the dural flap is created as well as during the drilling of the posterior wall of the petrous pyramid in order to accomplish a wide exposure of the internal acoustic canal and of the infralabyrinthine region [14].
Table 2 Topographic anatomy of the uppermost portion of the vestibular aqueduct genu Distance
Mean (mm)
Median (mm)
Range (mm)
IAM-VA XI-VA Ridge-VA Endolymphatic ledge-VA PPW-VA
7.23 18.11 10.27 6.28 3.46
7.10 18.62 9.97 6.25 3.33
6.30–9.30 15.00–20.62 7.20–13.30 4.5–8.5 2.20–5.00
IAM posterior margin of the internal acoustic meatus, Endolymphatic Ledge upper limit of the endolymphatic ledge, PPW posterior pyramidal wall, Ridge petrous ridge, VA uppermost portion of the vestibular aqueduct genu, XI XI nerve in the jugular foramen
While the ES is usually recognizable in a dural duplication on the posterior pyramidal wall, the ED is embedded inside the petrous pyramid; hence, its preservation during surgical procedures is more challenging. Since neuronavigators have gained more and more popularity in the last 20 years in neurosurgery, we reasoned that it would be worthwhile to validate their accuracy in predicting the location of the VA, which is usually identifiable on computed tomography. Our prosections confirmed that neuronavigators may safely guide the drilling around the VA. Indeed, the position of the surgically exposed VA genu (real VA genu) coincided with the one that was identified using the neuronavigator CT dataset (virtual VA genu) in all the specimens. In other words, we demonstrated in a laboratory setting that a neuronavigator may represent a reliable tool for identifying the location of a sub-millimetric structure like the VA: this finding may result in a direct benefit in the operative theater. Moreover, the knowledge of the topographic relationships between the ES, the VA, and anatomic landmarks readily identifiable during a suboccipital retrosigmoid intradural approach may help to determine where the ES and the VA are located even in the absence of neuronavigation. Clearly in a clinical setting, the location and the extension of the tumor, the resulting distortion of the petrous bone anatomy together with the anatomic individual variability will determine the area that needs to be exposed as well as the extent of safe drilling. Nonetheless, we believe the topographic data presented may strengthen the knowledge of temporal bone anatomic features thereby providing general guidelines for a retrosigmoid labyrinth sparing petrosectomy. Our study shows that in most cases, the ES occupied a region of approximately 4.00 mm in diameter on the posterior pyramidal wall. Moreover, in 83.3 % of our specimens, the center of the ES was located between 10.88 and 14.66 mm posterolateral to the internal acoustic meatus, between 16.54 and 19.26 mm posterosuperolateral to the XI nerve in the jugular foramen, at a point between 11.26 and 15.58 mm inferior to the petrous ridge. A careful examination of the region defined by the above landmarks may allow the identification of the dural duplication in which the ES is located as well as of its bony ledge. The senior author, in a previous paper under different conditions (caliper measurements), observed slightly shorter values on the distance from the ES to the XI nerve in the jugular foramen (mean, 15.70 mm; range, 11.00–18.50 mm) and to the petrous ridge (mean, 11.50 mm; range, 8.00– 17.00 mm) and longer values on the distance between the petro-sigmoid intersection and the point on the petrous ridge that was the closest to the center of the ES (mean, 24.10 mm; range, 20.00–28.00 mm) [2].
Neurosurg Rev Fig. 4 Preoperative gadoliniumenhanced axial MRI of a 59-year-old female with disabling vertigo and a 2-mm intracanalicular vestibular schwannoma unresponsive to gamma knife therapy (a). Postoperative gadoliniumenhanced axial MRI showing complete tumor removal with preservation of preoperative hearing (b). Axial CT images demonstrating extensive drilling of the posterior meatal wall with preservation of vestibular aqueduct, arrows (c and d)
Our results on the distance between the ES and the posterior meatal lip are in accordance with previous investigations [2, 7, 14]. Geurkink [7] and Shimizu [14] reported shorter distances because they measured from the meatal lip to the bony ledge of the ES and not to its center. Moreover, our values on the size of the ES reflect the great variability observed in previous studies [2, 6]. The VA genu is usually located more medially than the common crus and, for this reason, during drilling, it is usually exposed before the common crus [14]. In 83.3 % of our specimens, the uppermost portion of the VA genu was encountered between 15.50 and 20.30 mm superolateral to the XI nerve in the jugular foramen, between 8.33 and 12.88 mm inferior to the petrous ridge, and between 4.60 and 7.60 mm anterolateral to the endolymphatic ledge. Furthermore, on the average, 7.23 mm of petrous bone removal may be carried out posterior to the internal acoustic meatus without damaging the VA. At this level, it was found to be at a depth between 2.60 and 4.50 mm from the posterior pyramidal wall in 83.3 % of cases. Shimizu noted similar distances between the posterior meatal lip and the VA (mean, 7.51 mm), but the low value reported (4.68 mm) was smaller than ours (6.30 mm) [14].
Regarding the depth of the petrous bone needed to expose the uppermost portion of the VA, Ammirati reported values similar (mean, 3.20 mm; range, 2.60–4.00 mm) to ours (mean, 3.46 mm; 2.20–5.00 mm) while Shimizu reported shorter ones (mean, 1.48 mm; range, 0.60–3.03 mm) than ours [1, 14]. Previous clinical and laboratory studies have proposed to guide the posterior meatal wall drilling using an endoscope [9], a frameless stereotaxy system [13], or both [12] in order to maximize its unroofing without violating the integrity of the labyrinth or injuring the cochlear nerve. Other authors have advised to gain an optimized working angle on the internal acoustic meatus through a medial extension of the classic retrosigmoid craniectomy in order to expose the intrameatal portion of the tumor while at the same time preserving the VA [5]. However, the knowledge of the ES and VA topographic relationships we provide, combined with a careful assessment of the preoperative imaging, may facilitate this petrous bone removal. Moreover, considering that the ES is the inferiormost labyrinthine structure in the region behind the internal acoustic meatus, its identification may also make safer and more extensive the drilling of the inframeatal/infralabyrinthine area. In conclusion, our study demonstrates that is possible to use neuronavigation to reliably predict the location of the VA genu. In addition, neuronavigation may be effectively
Neurosurg Rev
used to create a topographical framework that may help maintain the integrity of the ES/VA during retrosigmoid approaches (Fig. 4).
13.
14. Conflict of interest The authors report no conflict of interest.
Samii A, Brinker T, Kaminsky J, Lanksch WR, Samii M (2000) Navigation-guided opening of the internal auditory canal via the retrosigmoid route for acoustic neuroma surgery: cadaveric, radiological, and preliminary clinical study. Neurosurgery 47(2):382– 387, discussion 388 Shimizu S, Tanaka R, Oka H, Fujii K (2006) Risk of damage to the endolymphatic sac and duct during removal of the posterior meatal wall: an anatomic study. Neurosurgery 59(4 Suppl 2):ONS435– ONS439, discussion ONS439–440
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Ammirati M, Ma J, Cheatham ML, Maxwell D, Bloch J, Becker DP (1993) Drilling the posterior wall of the petrous pyramid: a microneurosurgical anatomical study. J Neurosurg 78(3):452–455 Ammirati M, Spallone A, Feghali J, Ma J, Cheatham M, Becker D (1995) The endolymphatic sac: microsurgical topographic anatomy. Neurosurgery 36(2):416–419 Arenberg IK, Marovitz WF, Shambaugh GE Jr (1970) The role of the endolymphatic sac in the pathogenesis of endolymphatic hydrops in man. Acta Otolaryngol Suppl 275:1–49 Domb GH, Chole RA (1980) Anatomical studies of the posterior petrous apex with regard to hearing preservation in acoustic neuroma removal. Laryngoscope 90(11 Pt 1):1769–1776 Ebner FH, Kleiter M, Danz S, Ernemann U, Hirt B, Löwenheim H, Roser F, Tatagiba M (2014) Topographic changes in petrous bone anatomy in the presence of a vestibular schwannoma and implications for the retrosigmoid transmeatal approach. Neurosurgery. doi:10.1227/NEU.0000000000000454 Friberg U, Jansson B, Rask-Andersen H, Bagger-Sjöbäck D (1988) Variations in surgical anatomy of the endolymphatic sac. Arch Otolaryngol Head Neck Surg 114(4):389–394 Geurkink NA (1977) Surgical anatomy of the temporal bone posterior to the internal auditory canal: an operative approach. Laryngoscope 87(6):975–986 Kartush JM, Telian SA, Graham MD, Kemink JL (1986) Anatomic basis for labyrinthine preservation during posterior fossa acoustic tumor surgery. Laryngoscope 96(9 Pt 1):1024–1028 King WA, Wackym PA (1999) Endoscope-assisted surgery for acoustic neuromas (vestibular schwannomas): early experience using the rigid Hopkins telescope. Neurosurgery 44(5):1095–1100, discussion 1100–1102 Lo WW, Daniels DL, Chakeres DW, Linthicum FH, Ulmer JL, Mark LP, Swartz JD (1997) The endolymphatic duct and sac. AJNR Am J Neuroradiol 18(5):881–887 Pillai P, Sammet S, Ammirati M (2008) Application accuracy of computed tomography-based, image-guided navigation of temporal bone. Neurosurgery 63(4 Suppl 2):326–332, discussion 332–333 Pillai P, Sammet S, Ammirati M (2009) Image-guided, endoscopicassisted drilling and exposure of the whole length of the internal auditory canal and its fundus with preservation of the integrity of the labyrinth using a retrosigmoid approach: a laboratory investigation. Neurosurgery 65(6 Suppl):53–59, discussion 59
Comments Nicholas C. Bambakidis, Cleveland, USA This is an anatomic paper investigating the usefulness of neuronavigation in aiding identification of and preservation of the endolymphatic sac (ES) and vestibular aquaduct (VA). The authors have produced a nice set of data from their cadaveric dissections; as with all such papers, the actual utility of the data is dubious. Nevertheless, we have found that careful study of the anatomic relationships between these landmarks can be of use in many surgical procedures around the internal auditory canal and cerebellopontine angle. Atul Goel, Mumbai, India The endolymphatic sac and duct are important structures that innocuously occupy a position in the petrous bone limited anteriorly by the 7th to 11th nerve complexes, superiorly by the petrous ridge and posteriorly and inferiorly by the sigmoid sinus. Though the retrosigmoid suboccipital approach is a very commonly used approach, the anatomy and preservation of the endolymphatic sac and duct is not a common subject of discussion. The authors have elaborately discussed the anatomy of the endolymphatic system and have elegantly showed the use of neuronavigation to identify and preserve these structures. The significance of endolymphatic sac and vestibular aqueduct is unclear even to an otolaryngologist. The exact role this anatomical structure plays in hearing and in ‘vertigo’ is more of speculation and individual conjectures. Translabyrinthine approaches will certainly sacrifice the ES and VA. Retrosigmoid operations have the advantage that endolymphatic sac and VA can be preserved. Currently, my surgical strategy in vestibular schwannomas is to avoid drilling of the internal acoustic meatus in most cases. In general, drilling of the internal acoustic meatus should be minimized to 5–6 mm posterolaterally. This avoids injury to the vestibular apparatus, the endolymphatic structures, and a high-riding jugular bulb and avoids opening of the air cells. Ammirati et al. present a cadaveric study and show the advantages of using navigation system to identify the structure and preserve it. Although technological assistance is always welcome, the well-defined anatomic landmarks and parameters that are the basis of all temporal bone surgery can be foolproof. Essentially, while technology is helpful, the importance of anatomical understanding cannot be ignored. Despite this, the endolymphatic sac is a well-defined anatomical entity and should be preserved wherever and whenever possible.
