Analysis of Various Tracts of Mastoid Air Cells Related to CSF Leak After the Anterior Transpetrosal Approach

CLINICAL ARTICLE J Neurosurg 130:360–367, 2019

Analysis of various tracts of mastoid air cells related to CSF leak after the anterior transpetrosal approach

Ryota Tamura, MD,1 Ryosuke Tomio, MD, PhD,2 Farrag Mohammad, MD,1,3 Masahiro Toda, MD, PhD,1 and Kazunari Yoshida, MD, PhD1

1Department of Neurosurgery, Keio University School of Medicine, Tokyo; 2Department of Neurosurgery, Mihara Memorial Hospital, Gunma, Japan; and 3Department of Neurosurgery, Assiut University, Assiut, Egypt

OBJECTIVE The anterior transpetrosal approach (ATPA) was established in 1984 and has been particularly effective for petroclival tumors. Although some complications associated with this approach, such as venous hemorrhage in the temporal lobe and nervous disturbances, have been resolved over the years, the incidence rate of CSF leaks has not greatly improved. In this study, some varieties of air cell tracts that are strongly related to CSF leaks are demonstrated. In addition, other pre- and postoperative risk factors for CSF leakage after ATPA are discussed. METHODS Preoperative and postoperative target imaging of the temporal bone was performed in a total of 117 patients who underwent ATPA, and various surgery-related parameters were analyzed. RESULTS The existence of air cells at the petrous apex, as well as fluid collection in the mastoid antrum detected by a postoperative CT scan, were possible risk factors for CSF leakage. Tracts that directly connected to the antrum from the squamous part of the temporal bone and petrous apex, rather than through numerous air cells, were significantly related to CSF leak and were defined as “direct tract.” All patients with a refractory CSF leak possessed “unusual tracts” that connected to the attic, tympanic cavity, or eustachian tube, rather than through the mastoid antrum. CONCLUSIONS Preoperative assessment of petrous pneumatization types is necessary to prevent CSF leaks. Direct and unusual tracts are particularly strong risk factors for CSF leaks. https://thejns.org/doi/abs/10.3171/2017.9.JNS171622 KEY WORDS anterior transpetrosal approach; skull base; CSF leak; petrous apex; mastoid air cell

he anterior transpetrosal approach (ATPA) is effec- Development of the surgical technique, monitoring tive and often used for petroclival meningiomas, devices, and preoperative radiographic technologies was trigeminal schwannomas, and basilar trunk artery helpful to resolve complications, such as venous hemor- aneurysms.T 10,36 The ATPA is conducted by temporal cra- rhage in the temporal lobe and nervous disturbances.26,32 niotomy with anterior petrosectomy. After temporal crani- However, the incidence rate of postoperative CSF leak- otomy, the dura mater of the middle fossa floor is elevated age at our hospital was still not low—12.9% in the first to avoid tearing the lesser and greater superficial petrosal 15 years and 13.2% in the last 15 years. In most cases, nerves until the trigeminal impression is visible. Kawase’s the ATPA cannot be performed without opening the air triangle (the area enclosed by the greater superficial petro- cells, because it requires drilling of the squamous part of sal nerve, geniculate ganglion, arcuate eminence, and pe- the temporal bone and petrous apex, as described above. trosal ridge) is resected to expose the dura of the posterior In general, the large dural defect is covered by a flap of cranial fossa. The temporal lobe dura is opened toward vascularized temporalis fascia, and the opened air cells the posterior cranial fossa dura.10 Since the establishment are replaced by autologous tissues (e.g., fat or muscle tis- of the ATPA by Kawase et al. in 1984,10 as of March 2017, sue) and fibrin glue to prevent a CSF leak.10,36 However, in 325 patients had undergone surgery by ATPA at Keio Uni- large series of patients who underwent skull base surgery, versity School of Medicine. including the ATPA, up to 15% suffered from a postoper-

ABBREVIATIONS ATPA = anterior transpetrosal approach. SUBMITTED July 6, 2017. ACCEPTED September 11, 2017. INCLUDE WHEN CITING Published online March 16, 2018; DOI: 10.3171/2017.9.JNS171622.

360 J Neurosurg Volume 130 • February 2019 ©AANS 2019, except where prohibited by US copyright law

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al. ative CSF leak.5,14,15,19,20,24, 30, 31, 33,35 Although other modified methods (e.g., multilayer sealing techniques using fat, fascia, and inlay techniques) have been developed, CSF leaks have not been completely resolved.7,11,16,18,21,25,34 In recent years, skull base surgery via the endonasal endoscopic approach has demonstrated outstanding prog- ress.3,6,8 The risk of CSF leak after the endonasal endoscopic approach has been decreased by using fat, fascia, or muscle to plug the surgical defects.28,29 Therefore, more efficient ways to prevent and predict CSF leaks following an ATPA should be developed. Some anatomical analyses of the temporal bone and pneumatization have been published.1,13,27 Air cells in the squamous part of the temporal bone and petrous apex typically connect to the attic through the mastoid antrum.1,27 In this study, we revealed different tracts that are related to CSF leaks after an ATPA. Determination of the pre- and postoperative risk factors in this study may lead to meticulous packing of the opened air cells and decrease the incidence ratio of CSF leaks. Methods FIG. 1. Preoperative and postoperative radiographic characteristics of The current retrospective study, performed at the Keio CSF leaks after the ATPA. A: Preoperative axial CT scan showing the University Hospital (Tokyo, Japan), was approved by the development of air cells in the petrous apex. B: Preoperative axial CT institutional review board. For this retrospective analy- scan showing the development of air cells in the squamous part of the sis, we collected all CT images obtained in 168 patients temporal bone. C: Preoperative coronal CT scan showing an extremely who underwent surgery via the petrosal approach at our thin tegmen tympani. D: Postoperative axial CT scan showing fluid col- hospital between January 2005 and March 2017. The CT lection in the mastoid antrum that is strongly related to CSF leaks after examinations were performed on a 64–detector row CT the ATPA. Figure is available in color online only. scanner (LightSpeed VCT; GE Healthcare), a 320–detector row CT scanner (Aquilion ONE; Toshiba Medical Sys- tems), or a 256–multidetector row CT scanner (Revolution the perioperative period; and 3) postoperative factors— CT; GE Healthcare). fluid collection in the mastoid air cells or antrum detected Exclusion criteria of this study were as follows: 1) pa- by CT scan 1 day postoperatively (Fig. 1D). tients who underwent a combined petrosal approach (n = The chi-square test was used to compare the incidence 31), and 2) patients for whom there were no retained rec rates of CSF leak in the presence and absence of the pa- ords (n = 20). Thus, data in 117 patients (45 males and rameters listed above. Analyses were performed with IBM 72 females, 6–76 years of age) who underwent the ATPA SPSS Statistics (IBM Corp.). were analyzed in this study. The anatomy of the petrous bone air cells was evaluated by pre- and postoperative bone window CT using direct axial imaging with 1.0- or Results 5.0-mm collimation parallel to the orbitomeatal line or The results are summarized in Table 1. Of 117 patients, target imaging of the petrous bone. To avoid the possibil- 17 (14.5%) suffered a CSF leak. One hundred fifteen pa- ity of biasing the results, this retrospective review of CT tients had a tumor, and the remaining 2 patients had neu- scans obtained in 117 patients was performed blindly by 2 ralgia. authors (R. Tamura and R. Tomio), who studied the scans independently. Surgical data were retrieved from opera- Preoperative Factors tive reports, information on tumor histology was obtained In general, pneumatization in the temporal bone con- from pathology reports, and all other perioperative infor- nects to the antrum through numerous mastoid air cells mation was collected from hospital medical records. (Fig. 2A).1,27 In this analysis, tracts that directly connected Parameters analyzed to assess their relationship to CSF to the antrum from the squamous part of the temporal leaks were as follows: 1) preoperative factors—pneumati- bone and petrous apex, rather than through numerous air zation in the petrous apex (Fig. 1A) or squamous part of cells, were identified, thereby defining “squamous direct the temporal bone (Fig. 1B), bone erosion by tumor, the tract” and “petrous direct tract” (Fig. 2B). All of the air tegmen tympani (Fig. 1C), and the anatomical features of cells in the temporal bone (including mastoid, squamous, air cell tracts around the mastoid antrum; 2) intraoperative and petrous parts) typically connect to the attic through factors—the size of craniotomy and drilled petrous apex, the mastoid antrum.1,27 However, our analyses revealed the surgeon’s intraoperative awareness of the opened air different tracts that connected to the attic, tympanic cav- cells, replacement air cells of the petrous apex by fat or ity, or eustachian tube, rather than through the mastoid an- muscle tissue, and the insertion of a spinal drain during trum, thereby defining “unusual tracts” (Fig. 2C).

J Neurosurg Volume 130 • February 2019 361

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

TABLE 1. Incidence ratio of CSF leaks in 117 patients Incidence Ratio of CSF Leaks Related to Each Parameter Parameter Yes (%) No (%) p Value Preop parameters Air cell of petrous apex 14/56 (25) 3/61 (5) 0.002 Air cell in squamous part of temporal bone 16/95 (17) 1/22 (5) 0.14 Thin tegmen tympani <0.9 mm 3/10 (30) 14/48 (29) 0.957 Erosion of petrous bone 0/6 (0) 17/111 (15) 0.299 Squamous direct tract 10/38 (26) 7/79 (9) 0.012 Petrous direct tract 13/47 (28) 4/70 (6) 0.005 Unusual tract 14/19 (74) 3/96 (3) <0.001 Intraop parameters Craniotomy >55 mm 7/51 (14) 10/66 (15) 0.835 Petrous apex defect >15 mm 9/66 (14) 8/51 (16) 0.688 Intraop awareness of air cell opening 3/17 (18) 4/41 (10) 0.401 Replacement of air cell by fat or muscle 7/37 (19) 6/39 (15) 0.682 Spinal drain 2/10 (20) 5/48 (10) 0.397 Postop parameters Fluid collection in mastoid air cell 17/82 (21) 0/35 (0) 0.004 Fluid collection in mastoid antrum 17/53 (32) 0/64 (0) <0.001

The incidence ratio of CSF leakage did not differ sig- less of replacement of the petrous apex air cells by fat or nificantly between patients with and without air cells in muscle tissue and insertion of a spinal drain during the the squamous part of the temporal bone (incidence ratio perioperative period, there were no significant differences 16 of 95 [17%] vs 1 of 22 [5%], p = 0.14). In contrast, the in the incidence ratio of CSF leakage (replacement of air incidence ratio of CSF leakage differed significantly be- cells by fat or muscle incidence ratio 7 of 37 [19%] vs 6 tween patients with and without air cells in the petrous of 39 [15%], p = 0.682; spinal drain incidence ratio 2 of apex (incidence ratio 14 of 56 [25%] vs 3 of 61 [5%], p = 10 [20%] vs 5 of 48 [10%], p = 0.397). These procedures 0.002). The incidence ratio of CSF leakage differed sig- were more likely to be performed for patients with a high nificantly between patients with and without squamous or risk of CSF leaks. The surgeon’s awareness of opened air petrous direct tract (the squamous direct tract incidence cells also had no significant association with CSF leakage ratio 10 of 38 [26%] vs 7 of 79 [9%], p = 0.012; the petrous (incidence ratio 3 of 17 [18%] vs 4 of 41 [10%], p = 0.401). direct tract incidence ratio 13 of 47 [28%] vs 4 of 70 [6%], p = 0.005). Squamous and petrous direct tracts were dem- Postoperative Factors onstrated in 80% of patients with a CSF leak. Notably, the The incidence ratio of fluid collection in the antrum or incidence ratios of CSF leakage differed significantly be- mastoid air cells detected by a postoperative CT scan var- tween patients with and without unusual tracts (incidence ied considerably (antrum incidence ratio 17 of 53 [32%] vs ratio 14 of 19 [74%] vs 3 of 98 [3%], p < 0.001). In addi- 0 of 64 [0%], p < 0.001; mastoid air cells incidence ratio 17 tion, all patients who exhibited a refractory CSF leak that of 82 [21%] vs 0 of 35 [0%], p = 0.004). All patients with required closure had unusual tracts. CSF leaks had fluid collection in the mastoid antrum and The tegmen tympani was also evaluated in this analy- mastoid air cells. Patients with fluid collection in mastoid sis. A thin tegmen tympani was defined as a thickness < air cells but not in the mastoid antrum did not suffer from 0.9 mm.30 For 58 patients in whom the coronal CT image was obtained, 10 patients (17%) had a thin tegmen tym- CSF leaks. pani, and 3 of these 10 patients had a CSF leak. In contrast, 14 of 48 patients with a normal tegmen tympani had Discussion a CSF leak. Thin tegmen tympani may not be a risk factor Risk factors for a CSF leak after skull base surgery, for a CSF leak (p = 0.957). including the ATPA, are summarized in Table 2.5,14,15, 19, 20, 24,30, 31, 33,35 Regardless of whether exposed air cells were re- Intraoperative Factors placed by fat or muscle tissues in these reports, the CSF The size of craniotomy (> 55 mm) and drilled petrous leak did not completely resolve. Although the incidence apex (> 15 mm) also had no significant association with ratio of postoperative CSF leakage in our hospital was not CSF leaks (craniotomy incidence ratio 7 of 51 [14%] vs high compared with reported results by others, some pa- 10 of 66 [15%], p = 0.835; drilled petrous apex incidence tients demonstrated a refractory course and required reop- ratio 9 of 66 [14%] vs 8 of 51 [16%], p = 0.688). Regard- eration.17,23 Among the studies listed in Table 2, pneuma-

362 J Neurosurg Volume 130 • February 2019

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

FIG. 2. Tracts of pneumatization in the temporal bone. A: Typical type of tracts from the squamous part of the temporal bone and petrous apex connecting to the mastoid antrum through numerous air cells. B: The squamous and petrous direct tracts connecting to the mastoid antrum, rather than through numerous air cells. C: Unusual tracts from the squamous part of the temporal bone and petrous apex to the attic, tympanic cavity, or eustachian tube. 1) The posterosuperior cell tract connecting to the attic directly from the petrous apex, rather than through the mastoid antrum. 2) The hypotympanic tract connecting to the tympanic cavity through the carotid canal and cochlea from the apical area of the petrous apex. 3) The peritubal tract connecting around the eustachian tube from the apical area of the petrous apex. 4) The squamous tympanic tract connecting to the tympanic cavity from the squamous part of the temporal bone, rather than through the mastoid antrum.

tization of the petrous apex was the most important risk ticulous packing of air cells and decrease the incidence factor, as in our study. Furthermore, we analyzed various ratio of CSF leak. types of pneumatization and found air cell tracts that were closely related to CSF leakage after the ATPA. Our analy- Air Cells in the Squamous Part of the Temporal Bone and sis indicates that patients at high risk for CSF leaks should Petrous Apex be identified before the operation, which may lead to me- In general, air cells in the mastoid, squamous part of

J Neurosurg Volume 130 • February 2019 363

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

TABLE 2. Literature review on risk factors of CSF leaks after skull base surgery No. of Patients Authors & Year (ratio of CSF leaks) Risk Factors Surgical Approach Nutik & Korol, 1995 156 (15%) Opened petrous apex air cell (not opened mastoid air cell at the lat side of the RA craniotomy). Communication between the IAC & the middle ear Leonetti et al., 2001 589 (6%) CSF in the antrum in postop CT TLA, RA, CA Yamakami et al., 2003 NE Pneumatized petrous bone RA Fishman et al., 2004 196 (6.6%) Communication w/ the middle ear space & patent eustachian tube. Extensively RA, MCA, TLA pneumatized air cell Little et al., 2005 137 (7.3%) The wider exposure of bone after drilling CA, RA, MCA, TMA Stieglitz et al., 2010 519 (5.2%) Petrous bone air cell vol RA Nanda et al., 2011 50 (4%) Reop TPA, OZA Walcott et al., 2012 32 (6%) Reop w/o vascularized temporalis muscle flap. Congenital abnormality in the RA, TPA middle ear. Prior radiation Scheich et al., 2016 148 (13%) Pneumatized petrous bone MCA Stevens et al., 2016 48 (14.5%) Abnormally thin tegmen tympani TMA, MCA, TPA Present study 117 (14.5%) Pneumatized petrous bone. Petrous & squamous direct tracts. Unusual tracts.* ATPA Fluid collection in the mastoid antrum on postop CT CA = combined petrosal approach; IAC = internal auditory canal; MCA = middle cranial fossa approach; NE = not examined; OZA = orbitozygomatic approach; RA = retrosigmoid approach; TLA = translabyrinthine approach; TMA = transmastoid approach; TPA = transpetrosal approach (anterior or posterior transpetrosal approach). * Unusual tracts include the posterosuperior cell tract, hypotympanic tract, peritubal tract, and squamous tympanic tract. the temporal bone, and petrous apex typically connect to petrous apex were a significant risk factor for CSF leaks. the attic through the mastoid antrum.12 Therefore, air cells We tried to visualize the petrous bone of a patient with a in the squamous part of the temporal bone was a possible refractory CSF leak after the ATPA by using a 3D printer risk factor for CSF leaks. However, air cells in the squa- (Fig. 3A). Although autologous tissues were used to fill in mous part of the temporal bone can be replaced by fat and the drilled part of the petrous apex, microscopic confirma- muscle tissues macroscopically and can be compressed by tion of complete coverage is not easy because of the blind a bone flap; only 3 of 17 cases had CSF leaks in the squa- area (Fig. 3B). Therefore, some patients in whom opened mous part of the temporal bone. In contrast, air cells in the air cells were noticed intraoperatively may suffer from a

FIG. 3. Blind area through microscopic angle demonstrated by a 3D printer. A: A petrous bone model made by a 3D printer of a patient with a CSF leak shows the resected area of the petrous bone after operation. The red area contains opened air cells. The figure on the left shows anatomical orientation, and the one on the right shows the microscopic angle. B: Autologous tissue (fat tissue) is placed into opened air cells of the petrous apex. The opened air cells seem to be closed by microscopic angle. However, incomplete coverage becomes apparent when viewed from above.

364 J Neurosurg Volume 130 • February 2019

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

FIG. 4. Operative strategies for CSF leak after the ATPA in different tracts. A: Typical and direct tracts. 1) Opened air cells in the squamous part of the temporal bone or petrous apex should be packed with autologous tissue (fat tissue). 2) If this method is not sufficient to prevent a CSF leak, packing the antrum with fat tissue may be required. B: Unusual tracts. 1) Opened air cells in the squamous part of the temporal bone or petrous apex should be carefully packed with autologous tissue. 2) If this method is not sufficient to prevent a CSF leak, packing the eustachian tube with autologous tissue (fat tissue) may be required.

CSF leak postoperatively. Pneumatization of the petrous nized by a preoperative CT scan can also be strong risk fac- apex was an important risk factor, and careful packing was tors for CSF leakage (Fig. 2B). The air cells in the squamous needed. An endoscope may help to evaluate incomplete part of the temporal bone and petrous apex typically con- coverage. nect to the antrum through numerous air cells (Fig. 2A). In contrast, squamous and petrous direct tracts do not connect Direct Tracts to the antrum through numerous mastoid air cells. These The squamous and petrous direct tracts clearly recog- direct tracts were short, resulting in a high risk of CSF leak.

J Neurosurg Volume 130 • February 2019 365

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

Unusual Tracts Other Parameters Yamakami et al.35 have reported that mastoid and pe- With regard to postoperative risk factors, fluid collec- trous apex air cells were risk factors for CSF leakage in tion in mastoid air cells and the mastoid antrum (Fig. 1D) skull base surgery. In their report, 31% of patients had should be checked. All patients with CSF leaks had fluid petrous apex air cells, and they suggested that preopera- collection in the mastoid antrum. Most patients also had tive assessment of petrous pneumatization using a bone fluid collection in mastoid air cells. However, patients with window CT scan and meticulous intraoperative sealing of fluid collection in mastoid air cells and not the mastoid the exposed air cells may decrease the incidence of post- antrum did not suffer from CSF leaks. Fluid collection in operative CSF leaks. The petrous apex air cells sometimes the antrum should be evaluated by postoperative CT scans. expand to the carotid canal and articular tubercle.2 Ana- Stevens et al.30 showed a relationship between abnor- tomical analysis of the temporal bone has been reported; mally thin tegmen tympani and CSF leaks after skull base the air cells of the temporal bone were classified into the surgery. We have also observed a patient with a CSF leak mastoid region (antrum, periantral cell area, tegmental cell after the ATPA, which was caused by thin tegmen tym- area, sinodural cell area, perisinus cell area, perifacial cell pani. A thin tegmen tympani was not explicitly a risk fac- area, and mastoid tip cell area), the perilabyrinthine re- tor for CSF leaks in this analysis. A thin tegmen tympani gion (supralabyrinthine area and infralabyrinthine area), rarely causes a CSF leak, as shown in other reports and the petrous apex region (peritubal area and apical area), our experience. and the accessory region (zygomatic cell area, squamous According to our results, if air cells of the squamous cell area, occipital cell area, and styloid cell area).5,13,27 It part of the temporal bone or the petrous apex, or unusual or has been reported that some tracts directly lead to the attic, direct tracts, are identified by a preoperative CT scan, au- tympanic cavity, or eustachian tube rather than through tologous tissue should be carefully packed into the opened the antrum.4,5,9,14 air cells to prevent a refractory CSF leak. If fluid collection In our analysis, in addition to the presence of squamous in the mastoid antrum is increased several weeks postop- and petrous direct tracts, the presence of unusual tracts eratively, a refractory CSF leak should be considered. was associated with a high risk of CSF leaks after the ATPA (Fig. 2C). The posterosuperior cell tract directly Conclusions connects to the attic from the petrous apex. The hypotympanic tract courses between the carotid canal, cochlea, Unusual tracts of air cells in the temporal bone to the and tympanic cavity to reach the apical area of the petrous attic, tympanic cavity, or eustachian tube, rather than bone. The peritubal tract is an extension around the audi- through the antrum, are strong risk factors for refractory tory canal air cells into the apical area. Although only 18 CSF leaks after the ATPA. Squamous and petrous direct patients presented with these unusual tracts in our study, tracts also carry a strong risk of CSF leaks. all patients who required reoperation to close a CSF leak had unusual tracts. Four of 7 patients with the posterosu- References perior cell tract, 3 of 6 patients with the hypotympanic 1. Allam AF: Pneumatization of the temporal bone. Ann Otol tract, and 5 of 5 patients with the peritubal tract had a re- Rhinol Laryngol 78:49–64, 1969 fractory CSF leak. In addition, 1 patient had the squamous 2. Bronoosh P, Shakibafard A, Mokhtare MR, Munesi Rad T: tympanic tract from the squamous part of the temporal Temporal bone pneumatisation: a computed tomography bone to the tympanic cavity, leading to a refractory CSF study of pneumatized articular tubercle. Clin Radiol 69:151– leak (Fig. 2C). The peritubal tract may be the tract of most 156, 2014 concern, because the peritubal tract sometimes opens into 3. Castelnuovo P, Mauri S, Locatelli D, Emanuelli E, Delù G, Giulio GD: Endoscopic repair of cerebrospinal fluid rhinor- the eustachian tube. The existence of these unusual tracts rhea: learning from our failures. Am J Rhinol 15:333–342, was not related to the level of development of air cells. 2001 These direct and unusual tracts strongly influence the 4. Curtin HD, Som PM: The petrous apex. Otolaryngol Clin operative method to repair when a refractory CSF leak oc- North Am 28:473–496, 1995 curs. If a patient who has the squamous or petrous direct 5. Fishman AJ, Marrinan MS, Golfinos JG, Cohen NL, Roland tracts has a refractory CSF leak, regardless of replacing JT Jr: Prevention and management of cerebrospinal fluid leak the opened air cells by autologous tissue, packing of the following vestibular schwannoma surgery. Laryngoscope 114:501–505, 2004 antrum is required (Fig. 4A). In contrast, if a patient with 6. Hegazy HM, Carrau RL, Snyderman CH, Kassam A, Zweig unusual tracts has a refractory CSF leak, it cannot be im- J: Transnasal endoscopic repair of cerebrospinal fluid rhinor- proved only by packing the antrum. If all else fails, pack- rhea: a meta-analysis. Laryngoscope 110:1166–1172, 2000 ing the eustachian tube may also be required (Fig. 4B), 7. Hoffmann TK, El Hindy N, Müller OM, Schuler PJ, Berg- but it carries the risk of conduction deafness. In our study, mann C, Hierner R, et al: Vascularised local and free flaps in with regard to autologous tissues, there was no significant anterior skull base reconstruction. Eur Arch Otorhinolar- difference in the incidence of CSF leaks between patients yngol 270:899–907, 2013 in whom air cells in the petrous apex were replaced by 8. Ismail AS, Costantino PD, Sen C: Transnasal transsphenoidal endoscopic repair of CSF leakage using multilayer acellular muscle and those in whom air cells were replaced by fat. dermis. Skull Base 17:125–132, 2007 However, fat tissue is considered the better tissue for fill- 9. Jen A, Sanelli PC, Banthia V, Victor JD, Selesnick SH: Re- ing large air cells because some reports have pointed out lationship of petrous temporal bone pneumatization to the the advantage that it contains rich mesenchymal stem cells eustachian tube lumen. Laryngoscope 114:656–660, 2004 differentiated into the connective tissue.18,22 10. Kawase T, Toya S, Shiobara R, Mine T: Transpetrosal ap-

366 J Neurosurg Volume 130 • February 2019

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC R. Tamura et al.

proach for aneurysms of the lower basilar artery. J Neuro- 27. Smith PG, Leonetti JP, Grubb RL: Management of cerebro- surg 63:857–861, 1985 spinal fluid otorhinorrhea complicating the retrosigmoid ap- 11. Laufer I, Anand VK, Schwartz TH: Endoscopic, endonasal proach to the cerebellopontine angle. Am J Otol 11:178–180, extended transsphenoidal, transplanum transtuberculum 1990 approach for resection of suprasellar lesions. J Neurosurg 28. Stamm AC, Pignatari SS: Nasal septal cross-over flap tech- 106:400–406, 2007 nique: a choanal atresia micro-endoscopic surgical repair. 12. Lavezzi P, Bondioni MP, Chiesa A, Ettorre GC: [The anat- Am J Rhinol 15:143–148, 2001 omy of the temporal region viewed by magnetic resonance.] 29. Stamm AC, Pignatari SS, Vellutini E: Transnasal endoscopic Radiol Med (Torino) 91:33–38, 1996 (Ital) surgical approaches to the clivus. Otolaryngol Clin North 13. Lee DH, Kim MJ, Lee S, Choi H: Anatomical factors influ- Am 39:639–656, xi, 2006 encing pneumatization of the petrous apex. Clin Exp Oto- 30. Stevens SM, Rizk HG, McIlwain WR, Lambert PR, Meyer rhinolaryngol 8:339–344, 2015 TA: Association between lateral skull base thickness and sur- 14. Leonetti JP, Anderson D, Marzo S, Moynihan G: Prevention gical outcomes in spontaneous CSF otorrhea. Otolaryngol and management of cerebrospinal fluid fistula after transtem- Head Neck Surg 154:707–714, 2016 poral skull base surgery. Skull Base 11:87–92, 2001 31. Stieglitz LH, Giordano M, Gerganov V, Raabe A, Samii A, 15. Little KM, Friedman AH, Sampson JH, Wanibuchi M, Fuku- Samii M, et al: Petrous bone pneumatization is a risk factor shima T: Surgical management of petroclival meningiomas: for cerebrospinal fluid fistula following vestibular schwanno- defining resection goals based on risk of neurological mor- ma surgery. Neurosurgery 67 (2 Suppl Operative):509–515, bidity and tumor recurrence rates in 137 patients. Neurosur- 2010 gery 56:546–559, 2005 32. Tomio R, Akiyama T, Ohira T, Horikoshi T, Yoshida K: Use- 16. Locatelli D, Rampa F, Acchiardi I, Bignami M, De Bernardi fulness of facial nerve monitoring for confirmation of greater F, Castelnuovo P: Endoscopic endonasal approaches for re- superficial petrosal nerve in anterior transpetrosal approach. pair of cerebrospinal fluid leaks: nine-year experience. Neu- Acta Neurochir (Wien) 156:1847–1852, 2014 rosurgery 58 (4 Suppl 2):ONS-246–ONS-257, 2006 33. Walcott BP, Nahed BV, Sarpong Y, Kahle KT, Sekhar LN, 17. Lüdemann WO, Stieglitz LH, Gerganov V, Samii A, Samii Ferreira MJ: Incidence of cerebrospinal fluid leak following M: Fat implant is superior to muscle implant in vestibular petrosectomy and analysis of avoidance techniques. J Clin schwannoma surgery for the prevention of cerebrospinal fluid Neurosci 19:92–94, 2012 fistulae. Neurosurgery 63 (1 Suppl 1):ONS38–ONS43, 2008 34. Wormald PJ, McDonogh M: The bath-plug closure of anterior 18. Nakajima S, Fukuda T, Hasue M, Sengoku Y, Haraoka J, skull base cerebrospinal fluid leaks. Am J Rhinol 17:299– Uchida T: New technique for application of fibrin sealant: 305, 2003 rubbing method devised to prevent cerebrospinal fluid leak- 35. Yamakami I, Uchino Y, Kobayashi E, Yamaura A: Computed age from dura mater sites repaired with expanded polytetra- tomography evaluation of air cells in the petrous bone—re- fluoroethylene surgical membranes. Neurosurgery 49:117– lationship with postoperative cerebrospinal fluid rhinorrhea. 123, 2001 Neurol Med Chir (Tokyo) 43:334–339, 2003 19. Nanda A, Javalkar V, Banerjee AD: Petroclival meningio- 36. Yoshida K, Kawase T: Trigeminal neurinomas extending into mas: study on outcomes, complications and recurrence rates. multiple fossae: surgical methods and review of the literature. J Neurosurg 114:1268–1277, 2011 J Neurosurg 91:202–211, 1999 20. Nutik SL, Korol HW: Cerebrospinal fluid leak after acoustic neuroma surgery. Surg Neurol 43:553–557, 1995 21. Park JS, Kong DS, Lee JA, Park K: Intraoperative management to prevent cerebrospinal fluid leakage after microvas- cular decompression: dural closure with a “plugging muscle” method. Neurosurg Rev 30:139–142, 2007 Disclosures 22. Rimondini L, Mele S: Stem cell technologies for tissue regen- The authors report no conflict of interest concerning the materi- eration in dentistry. Minerva Stomatol 58:483–500, 2009 als or methods used in this study or the findings specified in this 23. Roberti F, Sekhar LN, Kalavakonda C, Wright DC: Posterior paper. fossa meningiomas: surgical experience in 161 cases. Surg Neurol 56:8–21, 2001 Author Contributions 24. Scheich M, Ginzkey C, Ehrmann-Müller D, Shehata-Dieler W, Hagen R: Management of CSF leakage after microsurgery Conception and design: Tamura. Acquisition of data: Tamura, for vestibular schwannoma via the middle cranial fossa ap- Tomio. Analysis and interpretation of data: Tamura, Tomio. Draft- proach. Eur Arch Otorhinolaryngol 273:2975–2981, 2016 ing the article: Tamura. Critically revising the article: Tamura, 25. Seifert V, Raabe A, Zimmermann M: Conservative (laby- Mohammad, Toda. Reviewed submitted version of manuscript: rinth-preserving) transpetrosal approach to the clivus and Tomio, Mohammad, Toda. Statistical analysis: Tamura. Admin- petroclival region—indications, complications, results and istrative/technical/material support: Tomio, Mohammad, Toda, lessons learned. Acta Neurochir (Wien) 145:631–642, 2003 Yoshida. Study supervision: Yoshida. 26. Shibao S, Toda M, Orii M, Fujiwara H, Yoshida K: Various patterns of the middle cerebral vein and preservation of ve- Correspondence nous drainage during the anterior transpetrosal approach. J Ryota Tamura: Keio University School of Medicine, Tokyo, Japan. Neurosurg 124:432–439, 2016 [email protected].

J Neurosurg Volume 130 • February 2019 367

Unauthenticated | Downloaded 10/04/21 02:03 AM UTC