Modeling alterations in sinonasal physiology after skull base surgery Dennis O. Frank-Ito, Ph.D.,1 Mirabelle Sajisevi, M.D.,1 C. Arturo Solares, M.D.,2 and David W. Jang, M.D.1
Y P
ABSTRACT
Background: Endonasal endoscopic skull base surgery (EESBS) often requires significant alterations in intranasal anatomy. For example, posterior septectomy (PS) with middle turbinate resection (MTR) is frequently performed to provide access to large sellar and clival tumors. However, little is known about the alterations that occur in sinonasal physiology. This study was designed to assess changes in sinonasal physiology after virtually performed endoscopic skull base surgery. Methods: Three-dimensional models of the sinonasal passage were created from computed tomography scans in three subjects with varying anatomy: no SD (SD), right anterior SD, and left anterior SD, respectively. Four additional surgery types were performed virtually on each model: endoscopic transsphenoidal approach (ETSA) with small (1 cm) PS (smPS), ETSA with complete (2 cm) PS, ETSA with smPS and right MTR, and ETSA with complete PS and right MTR. Computational fluid dynamics (CFD) simulations were performed on the 3 presurgery and 12 virtual surgery models to assess changes from surgery types. Results: Increased nasal airflow corresponded to amount of tissue removed. Effects of MTR on unilateral airflow allocation were unchanged in subject with no SD, worsened in leftward SD, and reversed in rightward SD. Severity of airflow and mucosal wall interactions trended with amount of tissue removed. MTR hindered flow interactions with the olfactory mucosa in subjects with SD. Conclusion: CFD simulations on virtual surgery models are able to reasonably detect changes in airflow patterns in the computer-generated nasal models. In addition, each patient’s unique anatomy influences the magnitude and direction of these changes after virtual EESBS. Once future studies can reliably correlate CFD parameters with patient symptoms, CFD will be a useful clinical tool in surgical planning and maximizing patient outcomes. (Am J Rhinol Allergy 29, 145–150, 2015; doi: 10.2500/ajra.2015.29.4150)
T
T
he endonasal endoscopic skull base surgery (EESBS) technique involves the removal of skull base tumors and lesions through the nose using high-definition endoscopy. As the application of EESBS continues to grow, large tumors can now be safely and effectively removed from all areas of the skull base. Extirpation of such tumors and reconstruction of defects requires significant disruptions in normal sinonasal anatomy. For example, endoscopic pituitary tumor resection frequently involves a posterior septectomy (PS) to allow for a binostril approach. Large sinusotomies with partial or complete resection of turbinates, septum, and pterygoid plates are commonplace with other more extensive approaches.1,2 However, very little is known about the changes in sinonasal physiology that occur with these surgeries. What remains clear is that patients are at risk for crusting, empty nose syndrome, and olfactory deficits as a result.3–5 Although the complication rate associated with EESBS is low, reported complications include nasal septal perforation, saddle nose, altered sense of smell, cerebrospinal fluid leak, meningitis, sinusitis, nasal obstruction, and chronic crusting.6–9 These complications could invariably lead to disruption in sinonasal airflow pattern, poor airconditioning capability due to inadequate humidification of inspired air, drying of the nasal, and sinus mucosa, which may result in crust formation, poor nasal filtration, and olfactory impairment. One of the tasks of the otolaryngologist is to develop innovative ways to predict and minimize long-term sequelae and maximize sinonasal quality-of-life (QOL) outcomes after EESBS. To this end, postoperative assessment of sinonasal function was investigated using computational fluid dynamics (CFD) modeling to quantify air-
O D
O N
From the 1Division of Otolaryngology–Head and Neck Surgery, Department of Surgery, Duke University Medical Center, Durham, North Carolina, and 2Georgia Skull Base Center, Georgia Regents University, Augusta, Georgia The authors have no conflicts of interest to declare pertaining to this article Address correspondence to Dennis Onyeka Frank-Ito, Ph.D., Division of Otolaryngology– Head and Neck Surgery, Duke University Medical Center, Box 3805, Durham, NC 27710 E-mail address: [email protected] Copyright © 2015, OceanSide Publications, Inc., U.S.A.
O C
flow variables that will provide objective information regarding the effects of surgery on the nose. Computational methods for quantifying sinonasal function show significant promise in predicting physiological changes, as evidenced from the growing number of studies that have used CFD techniques to assess physiological changes in the nose.10–17 The focus of this pilot study is to investigate whether CFD is able to evaluate changes in sinonasal ventilation rate in virtual surgery models for endoscopic skull base surgery, as well as study the impact of EESBS in the setting of a deviated nasal septum, and whether a unilateral obstruction from septal deviation (SD) influences nasal airflow after surgery.
METHODS Subjects This is a retrospective study approved by the Duke University Health System Institutional Review Board for Clinical Investigations. Computed tomography (CT) images of three subjects with no evidence of sinonasal disease were obtained for this study. The slice increments for all CT scans were within 0.7–1.25 mm, with pixel sizes of 0.313–0.391 mm and slice thicknesses of 0.75–1.25 mm. Subject 1 is a 61-year-old woman with no clinical evidence of deviated nasal septum; subject 2 is a 64-year-old man with a moderate leftward anterior SD; and subject 3 is a 52-year-old woman with a moderate rightward anterior SD (Fig. 1 A). We defined a moderate SD as a deviation 50–75% to the distance of the lateral nasal wall.
Sinonasal Airway Reconstruction CT scans were imported into a medical imaging software package (Mimics16.0; Materialise, Inc., Leuven, Belgium) and three-dimensional (3D) reconstructions of each subject’s natural sinonasal cavities were created. These served as previrtual surgery models. Each subject’s sinonasal airway was then digitally modified by the surgeon (DWJ) to mimic four different endoscopic transsphenoidal approach (ETSA) techniques: (1) ETSA with a 1-cm (small) PS (smPS), ETSA with a 2-cm (large) complete PS (lgPS), ETSA with a smPS and right
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
145
Y P
Figure 1. (A) Coronal view of CT showing the anatomic shape of each subject’s septum. (B) Side view of 3D reconstructed model for each subject (notice in subject 3, narrow nasopharynx). 3D, three-dimensional; CT, computed tomography.
O C
middle turbinate resection (MTR) (smPST), and ETSA with a lgPS and right MTR (lgPST). Virtual bilateral sphenoidotomy was performed on all of the digitally modified 3D models (Fig. 2). A total of 15 3D models of the sinonasal airspace were reconstructed, with each subject having 5 different 3D reconstructions (1 presurgery and 4 virtual surgery).
ments the finite volume method to numerically solve the conservation of mass and momentum equations for laminar, incompressible flow, described by the equations,
Mesh Generation
where u is the velocity vector field, ⫽ 1.204 kg/m3 is fluid density, ⫽ 1.825 ⫻ 10⫺5 kg/m ⫺ s is dynamic viscosity, and p is pressure. The boundary conditions to determine airflow in each 3D model were specified as follows: a “wall” condition assuming that the walls were stationary with zero air velocity at the air–wall interface, a “pressure inlet” condition at the nostrils with gauge pressure set to zero; and a “pressure outlet” condition at the outlet with gauge pressure set to ⫺20 Pa. For the constant inspiratory breathing pressure specified, the present study assumes simulations to be low-tomoderate breathing rate and the flow regime in the sinonasal cavity to be laminar.
All 15 sinonasal 3D models were exported from Mimics in stereolithography file format into the CAD and mesh-generating software package ICEM-CFD14.5 (ANSYS, Inc., Canonsburg, PA). Planar inlet (nostrils) and outlet (at tail end of the nasopharynx) surfaces were constructed. The 3D models were discretized by constructing 6 million unstructured tetrahedral mesh elements in the interior of each sinonasal cavity. Mesh refinement analysis conducted with the current discretization method (by varying mesh density from 0.25 to 8 million cells) found asymptotic behavior for outlet flow rate occurring around 4 million cells, indicating that a finer grid of 6 million elements will produce a consistent result. Mesh check and quality analysis were performed to prevent distorted elements from affecting the accuracy of the numerical simulation.
CFD Simulation
O D
ⵜ 䡠 uជ ⫽ 0
共uជ 䡠 ⵜ兲uជ ⫽ ⫺ ⵜp ⫹ ⵜ2uជ ,
O N
T
Steady-state laminar incompressible inhalation was simulated in each 3D model using the CFD software package Fluent14.5 (ANSYS, Inc.) under physiological pressure-driven conditions. Fluent imple-
RESULTS Total Flow Rate For subject 1 (no SD), total airflow rate was much higher in the lgPS (27.7 L/min) and lgPST (27.6 L/min) models (Fig. 3 A). The same was true for the lgPS (33.9L/min) and lgPST (35.0L/min) models in subject 2 (leftward anterior SD; Fig. 3 B). The effects of right MTR on total
Figure 2. (A–E) Subject 1 coronal view of CT and (F–J) 3D side view. (A) Presurgery airspace, (B) ETSA with smPS (1 cm) (C) ETSA with lgPS (2 cm), (D) ETSA with smPS (1 cm) and right MTR, (E) ETSA with lgPS (2 cm) PS and right MTR, (F) presurgery septum, (G) septum showing smPS, (H) septum showing lgPS, (I) presurgery lateral nasal wall, and (J) lateral nasal wall showing right MTR. ESTA, endoscopic transsphenoidal approach; IT, inferior turbinate; lgPS, large posterior septectomy; MT, middle turbinate; MTR, middle turbinate resection; smPS, small posterior septectomy.
146
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
Y P
Figure 3. Total airflow rate in the sinonasal cavity in presurgery and virtual surgery models for (A) subject 1, (B) subject 2, and (C) subject 3. (D) Flow streamlines in subject 3. lgPS, large posterior septectomy; lgPST, large posterior septectomy and right middle turbinate resection; smPS, small posterior septectomy; smPST, small posterior septectomy and right middle turbinate resection.
O N
airflow was more evident in subject 3 (rightward anterior SD), in which lgPST had the highest airflow rate at 6.7 L/min, followed by smPST at 6.6 L/min. The lgPS and smPS models in subject 3 had lower airflow rates at 6.5 and 6.4 L/min, respectively (Fig. 3 C). The previrtual surgery model in all three subjects had the least airflow. In general, total flow rate in the cavity was much smaller in subject 3 (between 6 and 7 L/min) than in subjects 1 (between 24 and 28 L/min) and 2 (between 30 and 35 L/min). This is because subject 3 had a much narrower nasopharynx and outlet than subjects 1 and 2 (Fig. 1 B). In addition, Fig. 3 D shows streamlines generated from 14 randomly distributed seeds points on both nostril surfaces in subject 3. There is more shunting of airflow through the larger PS models than in the smaller PS models. The effect of right MTR on airflow distribution in the sinonasal cavity was not evident from the streamlines plotted.
O D
T
Unilateral Airflow
In subject 1, PS of any size and unilateral right MTR had no measurable effect on flow partition (Fig. 4 A). Airflow partition was ⬃56% on the left side and 44% on the right side across the previrtual surgery model and all virtual surgery types. For subject 2, unilateral airflow partition in the previrtual surgery model was 34% on the left and 66% on the right, whereas the PS only models (smPS and lgPS) produced a further imbalance in flow allocation (left, 33%; right, 67%; Fig. 4 B). Additional imbalance in flow partition (left, 31%; right, 69%) occurred in the models that included the right MTR (smPST and lgPST). In this subject (subject 2), both small-sized and large-sized septectomies had equal effect on unilateral airflow partition, whereas right-sided MTR resulted in an additional increase in flow partition on the right side (Fig. 4 B). Results of airflow partition in subject 3
O C
were more dramatic (Fig. 4 C). Flow partition in the previrtual surgery model was 53% on the left and 47% on the right. Performing virtual PS produced a further imbalance in flow allocation (smPS: left, 55%, and right, 45%; lgPS: left, 56%, and right, 44%). Airflow partition was reversed with the addition of a right MTR, resulting in more airflow now entering the right side (smPST: left, 47%, and right, 53%; lgPST: left, 48%, and right, 52%).
Wall Shear Stresses Average wall shear stress (WSS) values were computed to investigate the magnitude of the interaction between airflow and the sinonasal mucosa wall (Figs. 5, A–C). Average WSS in subject 1 was lower in the lgPS models (lgPS, 0.06 Pa; lgPST, 0.058 Pa), and elevated in the smPS models (smPS, 0.063 Pa; smPST, 0.062 Pa; Fig 5 A) when compared with all other models. In all subjects regardless of the size of PS, performing right MTR tended to reduce average WSS. WSS values in subject 2 models with PS alone were 0.069 (smPS) and 0.068 Pa (lgPS), and 0.065 Pa for both smPST and lgPST, respectively (Fig. 5 B). In subject 3, there was a stepwise reduction in average WSS in the following order smPS, lgPS, smPST, and lgPST (Fig. 5 C).
Olfactory Shear Stresses To investigate the effects of these virtual surgery techniques on the olfactory cleft region, spatial distribution contours of WSS in the olfactory cleft for subject 3 are shown in Fig. 6 A. Elevated levels of shear stress were located mostly in the posterior region of the olfactory mucosa, with smPS and lgPS having the most olfactory cleft surface area with elevated levels of WSS. In addition,
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
147
Y P
average WSS in the olfactory cleft for all 15 models are given in Fig. 6 B. The virtual surgery models for subject 1 had the highest average WSS values in the olfactory cleft. In subjects 2 and 3, the virtual surgery models with right MTR had lower average WSS in the olfactory cleft.
O D
T
O N
Figure 4. Unilateral airflow partition in the left and right sides of the sinonasal cavity in presurgery and virtual surgery models: (A) subject 1, (B) subject 2, and (C) subject 3.
O C
Figure 5. Average WSS in presurgery and virtual surgery models: (A) subject 1, (B) subject 2, and (C) subject 3. WSS, wall shear stress.
DISCUSSION
EESBS is most commonly used for transsphenoidal pituitary tumor resections but has expanded to include lesions of the anterior skull base, clivus, petrous apex, and infratemporal fossa.1 In many of these cases, normal intranasal anatomy needs to be altered to gain access to lesions and to perform reconstructions of skull base defects.2 However, little is known about the physiological alterations and the resulting QOL changes that occur with EESBS. Studies assessing sinonasal QOL after endoscopic pituitary surgery have shown that olfaction and QOL scores are negatively affected at least in the initial postoperative period.3,4 Minimal surgical manipulation of the septum, turbinates, and paranasal sinuses has been associated with better QOL outcomes.5 However, such manipulation can not be avoided with more extensive EESBS. The septum is often partially resected in EESBS, leaving a permanent septal perforation. Although posterior septal perforations are not as symptomatic as anterior perforations,18 the extent and location of septal resection can vary greatly. With the ETSA, a smPS with removal of the sphenoid rostrum is performed to access bilateral sphenoid sinuses through a binostril approach. However, tumors of the clivus or the nasopharynx often require large or complete posterior septectomies, and large tumors of the anterior skull base require
148
complete or partial resection of the anterior septum as well.19,20 In some instances, when a pedicled nasoseptal flap is used for reconstruction, a PS may also be performed as part of a reverse septal flap.21 In addition to partial septectomy, MTR is sometimes used to improve access and visualization.2 For the ETSA, the middle turbinates are often outfractured bilaterally,22 while others prescribe MTR.23 In one cadaver study, unilateral MTR provided enhanced exposure to the middle third of the clivus and ipsilateral sphenopalatine artery, but bilateral MTR did not confer any advantage over unilateral MTR.24 However, controversy persists because the concern for scarring, atrophic rhinitis, and empty nose syndrome exists with turbinate resection. Inferior turbinate resection is also a factor when performing an endoscopic medial maxillectomy or when harvesting an inferior turbinate mucosal graft. Finally, large surgical openings of the paranasal sinuses are sometimes performed in EESBS. However, this is likely to negatively affect sinonasal physiology and diminish QOL in patients who do not have chronic sinusitis. Crusting, scarring, diminished olfaction, and impaired mucociliary clearance are all concerns. As shown in this pilot CFD study, increases in total airflow were directly related to the amount of virtually removed tissue. For uni-
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
can not be made. Rather, our findings simply point to the potential of CFD to provide individualized predictions for each patient’s unique anatomy and surgery. In this sense, this study is only a proof-ofconcept study showing that CFD in virtual surgery is able to detect changes that we may expect to see postoperatively in EESBS. Extensive studies need to be performed before CFD in virtual skull base surgery can have clinical application. Once CFD in virtual surgery is established as a reliable clinical tool for predicting patient outcomes, it can be used to apply any surgical approach to each individual patient’s anatomy. The ability to anticipate and quantify patientspecific alterations in nasal physiology after EESBS would be helpful in preoperative planning and patient counseling. At the same time, we acknowledge that outcomes after EESBS are not solely dependent on changes in nasal airflow.
Y P
CONCLUSION
O C
This pilot study showed that CFD simulations on virtual surgery models are able to reasonably detect changes in simulated nasal physiology after virtual EESBS. Moreover, these changes were influenced not only by virtual surgical technique, but also by the presence of anatomic variations.
REFERENCES 1. 2.
T 3.
Figure 6. (A) Wall shear stress contours in the olfactory cleft region for Subject 3. (B) Average wall shear stress in the olfactory cleft for all subjects. smPS, small posterior septectomy; lgPS, large posterior septectomy; smPST, small posterior septectomy and right middle turbinate resection; lgPST, large posterior septectomy and right middle turbinate resection.
O N
lateral flow partition, the changes were minimal in subject 1 (subject with no SD). Flow allocation did not change with MTR or PS, regardless of size. However, this was not the case in the setting of SD. In the subjects 2 and 3, PS produced a further imbalance in flow partition. Moreover, MTR on the opposite side of an SD produced more of an imbalance. However, MTR on the same side as a deviation redistributed flow partition to that side. Our findings in these computergenerated sinonasal models suggest that a carefully planned MTR in certain patients may improve nasal flow partition and, in the least, not worsen the sensation of unilateral obstruction. Simulated WSS attest to the fact that septectomy size and MTR do not produce consistent changes in all patients (Fig. 5, A–C). For example, the smPST model in subject 1 had higher shear stress than the presurgical model, whereas the opposite was true for subjects 2 and 3. Similarly, the lgPS model increased WSS over presurgery model in subject 2, but decreased WSS in subjects 1 and 3. The effects of virtual surgery type on air interacting with olfactory cleft mucosa indicated that MTR tended to inhibit flow interactions with the olfactory mucosa in subjects with a moderate anterior septal deflection (subjects 2 and 3); this was not true in subject 1. In addition, the magnitude of air interacting with olfactory mucosa as computed by WSS in the virtual surgery models was higher in subject 1 (no SD) than in subjects 2 and 3. Consequently, one could postulate that a subject with clinical evidence of nasal anatomic anomaly undergoing EESBS may experience diminished olfactory function postoperatively, with a likelihood of more severe symptoms if right MTR is performed. We emphasize that our findings are specific to the virtually created surgeries modeled in these three patients; therefore, generalizations
O D
Lee SC, and Senior BA. Endoscopic skull base surgery. Clin Exp Otorhinolaryngol 1:53–62, 2008. Schwartz TH, Fraser JF, Brown S, et al. Endoscopic cranial base surgery: Classification of operative approaches. Neurosurgery 62: 991–1002, 2008. Bedrosian JC, McCoul ED, Raithatha R, et al. A prospective study of postoperative symptoms in sinonasal quality-of-life following endoscopic skull-base surgery: Dissociations based on specific symptoms. Int Forum Allergy Rhinol 3:664–669, 2013. Wang YY, Srirathan V, Tirr E, et al. Nasal symptoms following endoscopic transsphenoidal pituitary surgery: Assessment using the General Nasal Patient Inventory. Neurosurg Focus 30:E12, 2011. Thompson CF, Suh JD, Liu Y, et al. Modifications to the endoscopic approach for anterior skull base lesions improve postoperative sinonasal symptoms. J Neurol Surg B Skull Base 75:65–72, 2014. Charalampaki P, Ayyad A, Kockro RA, and Perneczky A. Surgical complications after endoscopic transsphenoidal pituitary surgery. J Clin Neurosci 16:786–789, 2009. Black PM, Zervas NT, and Candia GL. Incidence and management of complications of transsphenoidal operation for pituitary adenomas. Neurosurgery 20:920–924, 1987. Laws ER Jr, and Kern EB. Complications of trans-sphenoidal surgery. Clin Neurosurg 23:401–416, 1976. Barrow DL, and Tindall GT. Loss of vision after transsphenoidal surgery. Neurosurgery 27:60–68, 1990. Cannon DE, Frank DO, Kimbell JS, et al. Modeling nasal physiology changes due to septal perforations. Otolaryngol Head Neck Surg 148:513–518, 2013. Martonen TB, Zhang Z, Yue G, and Musante CJ. Fine particle deposition within human nasal airways. Inhal Toxicol 15:283–303, 2003. Frank DO, Kimbell JS, Cannon D, and Rhee JS. Computed intranasal spray penetration: Comparisons before and after nasal surgery. Int Forum Allergy Rhinol 3:48–55, 2013. Kleven M, Melaaen MC, Reimers M, et al. Using computational fluid dynamics (CFD) to improve the bi-directional nasal drug delivery concept. Food Bioprod Process 83:107–117, 2005. Frank DO, Zanation AM, Dhandha VH, et al. Quantification of airflow into the maxillary sinuses before and after functional endoscopic sinus surgery. Int Forum Allergy Rhinol 3:834–840, 2013. Chen XB, Lee HP, Chong VF, and Wang de Y. Assessments of nasal bone fracture effects on nasal airflow: A computational fluid dynamics study. Am J Rhinol Allergy 25:e39–e43, 2011. Faramarzi M, Baradaranfar MH, Abouali O, et al. Numerical investigation of the flow field in realistic nasal septal perforation geometry. Allergy Rhinol (Providence) 2014. Kimbell JS, Garcia GJ, Frank DO, et al. Computed nasal resistance compared with patient-reported symptoms in surgically treated na-
4.
5.
6.
7.
8. 9. 10.
11. 12.
13.
14.
15.
16.
17.
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
149
18.
19.
20.
sal airway passages: A preliminary report. Am J Rhinol Allergy 26:e94–e98, 2012. Re M, Paolucci L, Romeo R, et al. Surgical treatment of nasal septal perforations. Our experience. Acta Otorhinolaryngola Ital 26:102–109, 2006. Hong Jiang W, Ping Zhao S, Hai Xie Z, et al. Endoscopic resection of chordomas in different clival regions. Acta Otolaryngol 129:71–83, 2009. Schaberg MR, Farrell CJ, Rabinowitz M, et al. Septal transposition: A novel technique for preservation of the nasal septum during endoscopic endonasal resection of olfactory groove meningiomas. Neurol Surg B 73:A165, 2012. doi: 10.1055/s-0032–1312213.
21.
Kasemsiri P, Carrau RL, Otto BA, et al. Reconstruction of the pedicled nasoseptal flap donor site with a contralateral reverse rotation flap: Technical modifications and outcomes. Laryngoscope 123:2601–2604, 2013. 22. Nyquist GG, Anand VK, Brown S, et al. Middle turbinate preservation in endoscopic transsphenoidal surgery of the anterior skull base. Skull Base 20:343–347, 2010. 23. Paluzzi A, Fernandez-Miranda JC, Pinheiro-Neto C, et al. Endoscopic endonasal infrasellar approach to the sellar and suprasellar regions: Technical note. Skull Base 21:335–342, 2011. 24. Guthikonda B, Nourbakhsh A, Notarianni C, et al. Middle turbinectomy for exposure in endoscopic endonasal transsphenoidal surgery: When is it necessary? Laryngoscope 120:2360–2366, 2010. e
Y P
T
O D 150
O C
O N
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
Y P
ABSTRACT
Background: Endonasal endoscopic skull base surgery (EESBS) often requires significant alterations in intranasal anatomy. For example, posterior septectomy (PS) with middle turbinate resection (MTR) is frequently performed to provide access to large sellar and clival tumors. However, little is known about the alterations that occur in sinonasal physiology. This study was designed to assess changes in sinonasal physiology after virtually performed endoscopic skull base surgery. Methods: Three-dimensional models of the sinonasal passage were created from computed tomography scans in three subjects with varying anatomy: no SD (SD), right anterior SD, and left anterior SD, respectively. Four additional surgery types were performed virtually on each model: endoscopic transsphenoidal approach (ETSA) with small (1 cm) PS (smPS), ETSA with complete (2 cm) PS, ETSA with smPS and right MTR, and ETSA with complete PS and right MTR. Computational fluid dynamics (CFD) simulations were performed on the 3 presurgery and 12 virtual surgery models to assess changes from surgery types. Results: Increased nasal airflow corresponded to amount of tissue removed. Effects of MTR on unilateral airflow allocation were unchanged in subject with no SD, worsened in leftward SD, and reversed in rightward SD. Severity of airflow and mucosal wall interactions trended with amount of tissue removed. MTR hindered flow interactions with the olfactory mucosa in subjects with SD. Conclusion: CFD simulations on virtual surgery models are able to reasonably detect changes in airflow patterns in the computer-generated nasal models. In addition, each patient’s unique anatomy influences the magnitude and direction of these changes after virtual EESBS. Once future studies can reliably correlate CFD parameters with patient symptoms, CFD will be a useful clinical tool in surgical planning and maximizing patient outcomes. (Am J Rhinol Allergy 29, 145–150, 2015; doi: 10.2500/ajra.2015.29.4150)
T
T
he endonasal endoscopic skull base surgery (EESBS) technique involves the removal of skull base tumors and lesions through the nose using high-definition endoscopy. As the application of EESBS continues to grow, large tumors can now be safely and effectively removed from all areas of the skull base. Extirpation of such tumors and reconstruction of defects requires significant disruptions in normal sinonasal anatomy. For example, endoscopic pituitary tumor resection frequently involves a posterior septectomy (PS) to allow for a binostril approach. Large sinusotomies with partial or complete resection of turbinates, septum, and pterygoid plates are commonplace with other more extensive approaches.1,2 However, very little is known about the changes in sinonasal physiology that occur with these surgeries. What remains clear is that patients are at risk for crusting, empty nose syndrome, and olfactory deficits as a result.3–5 Although the complication rate associated with EESBS is low, reported complications include nasal septal perforation, saddle nose, altered sense of smell, cerebrospinal fluid leak, meningitis, sinusitis, nasal obstruction, and chronic crusting.6–9 These complications could invariably lead to disruption in sinonasal airflow pattern, poor airconditioning capability due to inadequate humidification of inspired air, drying of the nasal, and sinus mucosa, which may result in crust formation, poor nasal filtration, and olfactory impairment. One of the tasks of the otolaryngologist is to develop innovative ways to predict and minimize long-term sequelae and maximize sinonasal quality-of-life (QOL) outcomes after EESBS. To this end, postoperative assessment of sinonasal function was investigated using computational fluid dynamics (CFD) modeling to quantify air-
O D
O N
From the 1Division of Otolaryngology–Head and Neck Surgery, Department of Surgery, Duke University Medical Center, Durham, North Carolina, and 2Georgia Skull Base Center, Georgia Regents University, Augusta, Georgia The authors have no conflicts of interest to declare pertaining to this article Address correspondence to Dennis Onyeka Frank-Ito, Ph.D., Division of Otolaryngology– Head and Neck Surgery, Duke University Medical Center, Box 3805, Durham, NC 27710 E-mail address: [email protected] Copyright © 2015, OceanSide Publications, Inc., U.S.A.
O C
flow variables that will provide objective information regarding the effects of surgery on the nose. Computational methods for quantifying sinonasal function show significant promise in predicting physiological changes, as evidenced from the growing number of studies that have used CFD techniques to assess physiological changes in the nose.10–17 The focus of this pilot study is to investigate whether CFD is able to evaluate changes in sinonasal ventilation rate in virtual surgery models for endoscopic skull base surgery, as well as study the impact of EESBS in the setting of a deviated nasal septum, and whether a unilateral obstruction from septal deviation (SD) influences nasal airflow after surgery.
METHODS Subjects This is a retrospective study approved by the Duke University Health System Institutional Review Board for Clinical Investigations. Computed tomography (CT) images of three subjects with no evidence of sinonasal disease were obtained for this study. The slice increments for all CT scans were within 0.7–1.25 mm, with pixel sizes of 0.313–0.391 mm and slice thicknesses of 0.75–1.25 mm. Subject 1 is a 61-year-old woman with no clinical evidence of deviated nasal septum; subject 2 is a 64-year-old man with a moderate leftward anterior SD; and subject 3 is a 52-year-old woman with a moderate rightward anterior SD (Fig. 1 A). We defined a moderate SD as a deviation 50–75% to the distance of the lateral nasal wall.
Sinonasal Airway Reconstruction CT scans were imported into a medical imaging software package (Mimics16.0; Materialise, Inc., Leuven, Belgium) and three-dimensional (3D) reconstructions of each subject’s natural sinonasal cavities were created. These served as previrtual surgery models. Each subject’s sinonasal airway was then digitally modified by the surgeon (DWJ) to mimic four different endoscopic transsphenoidal approach (ETSA) techniques: (1) ETSA with a 1-cm (small) PS (smPS), ETSA with a 2-cm (large) complete PS (lgPS), ETSA with a smPS and right
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
145
Y P
Figure 1. (A) Coronal view of CT showing the anatomic shape of each subject’s septum. (B) Side view of 3D reconstructed model for each subject (notice in subject 3, narrow nasopharynx). 3D, three-dimensional; CT, computed tomography.
O C
middle turbinate resection (MTR) (smPST), and ETSA with a lgPS and right MTR (lgPST). Virtual bilateral sphenoidotomy was performed on all of the digitally modified 3D models (Fig. 2). A total of 15 3D models of the sinonasal airspace were reconstructed, with each subject having 5 different 3D reconstructions (1 presurgery and 4 virtual surgery).
ments the finite volume method to numerically solve the conservation of mass and momentum equations for laminar, incompressible flow, described by the equations,
Mesh Generation
where u is the velocity vector field, ⫽ 1.204 kg/m3 is fluid density, ⫽ 1.825 ⫻ 10⫺5 kg/m ⫺ s is dynamic viscosity, and p is pressure. The boundary conditions to determine airflow in each 3D model were specified as follows: a “wall” condition assuming that the walls were stationary with zero air velocity at the air–wall interface, a “pressure inlet” condition at the nostrils with gauge pressure set to zero; and a “pressure outlet” condition at the outlet with gauge pressure set to ⫺20 Pa. For the constant inspiratory breathing pressure specified, the present study assumes simulations to be low-tomoderate breathing rate and the flow regime in the sinonasal cavity to be laminar.
All 15 sinonasal 3D models were exported from Mimics in stereolithography file format into the CAD and mesh-generating software package ICEM-CFD14.5 (ANSYS, Inc., Canonsburg, PA). Planar inlet (nostrils) and outlet (at tail end of the nasopharynx) surfaces were constructed. The 3D models were discretized by constructing 6 million unstructured tetrahedral mesh elements in the interior of each sinonasal cavity. Mesh refinement analysis conducted with the current discretization method (by varying mesh density from 0.25 to 8 million cells) found asymptotic behavior for outlet flow rate occurring around 4 million cells, indicating that a finer grid of 6 million elements will produce a consistent result. Mesh check and quality analysis were performed to prevent distorted elements from affecting the accuracy of the numerical simulation.
CFD Simulation
O D
ⵜ 䡠 uជ ⫽ 0
共uជ 䡠 ⵜ兲uជ ⫽ ⫺ ⵜp ⫹ ⵜ2uជ ,
O N
T
Steady-state laminar incompressible inhalation was simulated in each 3D model using the CFD software package Fluent14.5 (ANSYS, Inc.) under physiological pressure-driven conditions. Fluent imple-
RESULTS Total Flow Rate For subject 1 (no SD), total airflow rate was much higher in the lgPS (27.7 L/min) and lgPST (27.6 L/min) models (Fig. 3 A). The same was true for the lgPS (33.9L/min) and lgPST (35.0L/min) models in subject 2 (leftward anterior SD; Fig. 3 B). The effects of right MTR on total
Figure 2. (A–E) Subject 1 coronal view of CT and (F–J) 3D side view. (A) Presurgery airspace, (B) ETSA with smPS (1 cm) (C) ETSA with lgPS (2 cm), (D) ETSA with smPS (1 cm) and right MTR, (E) ETSA with lgPS (2 cm) PS and right MTR, (F) presurgery septum, (G) septum showing smPS, (H) septum showing lgPS, (I) presurgery lateral nasal wall, and (J) lateral nasal wall showing right MTR. ESTA, endoscopic transsphenoidal approach; IT, inferior turbinate; lgPS, large posterior septectomy; MT, middle turbinate; MTR, middle turbinate resection; smPS, small posterior septectomy.
146
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
Y P
Figure 3. Total airflow rate in the sinonasal cavity in presurgery and virtual surgery models for (A) subject 1, (B) subject 2, and (C) subject 3. (D) Flow streamlines in subject 3. lgPS, large posterior septectomy; lgPST, large posterior septectomy and right middle turbinate resection; smPS, small posterior septectomy; smPST, small posterior septectomy and right middle turbinate resection.
O N
airflow was more evident in subject 3 (rightward anterior SD), in which lgPST had the highest airflow rate at 6.7 L/min, followed by smPST at 6.6 L/min. The lgPS and smPS models in subject 3 had lower airflow rates at 6.5 and 6.4 L/min, respectively (Fig. 3 C). The previrtual surgery model in all three subjects had the least airflow. In general, total flow rate in the cavity was much smaller in subject 3 (between 6 and 7 L/min) than in subjects 1 (between 24 and 28 L/min) and 2 (between 30 and 35 L/min). This is because subject 3 had a much narrower nasopharynx and outlet than subjects 1 and 2 (Fig. 1 B). In addition, Fig. 3 D shows streamlines generated from 14 randomly distributed seeds points on both nostril surfaces in subject 3. There is more shunting of airflow through the larger PS models than in the smaller PS models. The effect of right MTR on airflow distribution in the sinonasal cavity was not evident from the streamlines plotted.
O D
T
Unilateral Airflow
In subject 1, PS of any size and unilateral right MTR had no measurable effect on flow partition (Fig. 4 A). Airflow partition was ⬃56% on the left side and 44% on the right side across the previrtual surgery model and all virtual surgery types. For subject 2, unilateral airflow partition in the previrtual surgery model was 34% on the left and 66% on the right, whereas the PS only models (smPS and lgPS) produced a further imbalance in flow allocation (left, 33%; right, 67%; Fig. 4 B). Additional imbalance in flow partition (left, 31%; right, 69%) occurred in the models that included the right MTR (smPST and lgPST). In this subject (subject 2), both small-sized and large-sized septectomies had equal effect on unilateral airflow partition, whereas right-sided MTR resulted in an additional increase in flow partition on the right side (Fig. 4 B). Results of airflow partition in subject 3
O C
were more dramatic (Fig. 4 C). Flow partition in the previrtual surgery model was 53% on the left and 47% on the right. Performing virtual PS produced a further imbalance in flow allocation (smPS: left, 55%, and right, 45%; lgPS: left, 56%, and right, 44%). Airflow partition was reversed with the addition of a right MTR, resulting in more airflow now entering the right side (smPST: left, 47%, and right, 53%; lgPST: left, 48%, and right, 52%).
Wall Shear Stresses Average wall shear stress (WSS) values were computed to investigate the magnitude of the interaction between airflow and the sinonasal mucosa wall (Figs. 5, A–C). Average WSS in subject 1 was lower in the lgPS models (lgPS, 0.06 Pa; lgPST, 0.058 Pa), and elevated in the smPS models (smPS, 0.063 Pa; smPST, 0.062 Pa; Fig 5 A) when compared with all other models. In all subjects regardless of the size of PS, performing right MTR tended to reduce average WSS. WSS values in subject 2 models with PS alone were 0.069 (smPS) and 0.068 Pa (lgPS), and 0.065 Pa for both smPST and lgPST, respectively (Fig. 5 B). In subject 3, there was a stepwise reduction in average WSS in the following order smPS, lgPS, smPST, and lgPST (Fig. 5 C).
Olfactory Shear Stresses To investigate the effects of these virtual surgery techniques on the olfactory cleft region, spatial distribution contours of WSS in the olfactory cleft for subject 3 are shown in Fig. 6 A. Elevated levels of shear stress were located mostly in the posterior region of the olfactory mucosa, with smPS and lgPS having the most olfactory cleft surface area with elevated levels of WSS. In addition,
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
147
Y P
average WSS in the olfactory cleft for all 15 models are given in Fig. 6 B. The virtual surgery models for subject 1 had the highest average WSS values in the olfactory cleft. In subjects 2 and 3, the virtual surgery models with right MTR had lower average WSS in the olfactory cleft.
O D
T
O N
Figure 4. Unilateral airflow partition in the left and right sides of the sinonasal cavity in presurgery and virtual surgery models: (A) subject 1, (B) subject 2, and (C) subject 3.
O C
Figure 5. Average WSS in presurgery and virtual surgery models: (A) subject 1, (B) subject 2, and (C) subject 3. WSS, wall shear stress.
DISCUSSION
EESBS is most commonly used for transsphenoidal pituitary tumor resections but has expanded to include lesions of the anterior skull base, clivus, petrous apex, and infratemporal fossa.1 In many of these cases, normal intranasal anatomy needs to be altered to gain access to lesions and to perform reconstructions of skull base defects.2 However, little is known about the physiological alterations and the resulting QOL changes that occur with EESBS. Studies assessing sinonasal QOL after endoscopic pituitary surgery have shown that olfaction and QOL scores are negatively affected at least in the initial postoperative period.3,4 Minimal surgical manipulation of the septum, turbinates, and paranasal sinuses has been associated with better QOL outcomes.5 However, such manipulation can not be avoided with more extensive EESBS. The septum is often partially resected in EESBS, leaving a permanent septal perforation. Although posterior septal perforations are not as symptomatic as anterior perforations,18 the extent and location of septal resection can vary greatly. With the ETSA, a smPS with removal of the sphenoid rostrum is performed to access bilateral sphenoid sinuses through a binostril approach. However, tumors of the clivus or the nasopharynx often require large or complete posterior septectomies, and large tumors of the anterior skull base require
148
complete or partial resection of the anterior septum as well.19,20 In some instances, when a pedicled nasoseptal flap is used for reconstruction, a PS may also be performed as part of a reverse septal flap.21 In addition to partial septectomy, MTR is sometimes used to improve access and visualization.2 For the ETSA, the middle turbinates are often outfractured bilaterally,22 while others prescribe MTR.23 In one cadaver study, unilateral MTR provided enhanced exposure to the middle third of the clivus and ipsilateral sphenopalatine artery, but bilateral MTR did not confer any advantage over unilateral MTR.24 However, controversy persists because the concern for scarring, atrophic rhinitis, and empty nose syndrome exists with turbinate resection. Inferior turbinate resection is also a factor when performing an endoscopic medial maxillectomy or when harvesting an inferior turbinate mucosal graft. Finally, large surgical openings of the paranasal sinuses are sometimes performed in EESBS. However, this is likely to negatively affect sinonasal physiology and diminish QOL in patients who do not have chronic sinusitis. Crusting, scarring, diminished olfaction, and impaired mucociliary clearance are all concerns. As shown in this pilot CFD study, increases in total airflow were directly related to the amount of virtually removed tissue. For uni-
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
can not be made. Rather, our findings simply point to the potential of CFD to provide individualized predictions for each patient’s unique anatomy and surgery. In this sense, this study is only a proof-ofconcept study showing that CFD in virtual surgery is able to detect changes that we may expect to see postoperatively in EESBS. Extensive studies need to be performed before CFD in virtual skull base surgery can have clinical application. Once CFD in virtual surgery is established as a reliable clinical tool for predicting patient outcomes, it can be used to apply any surgical approach to each individual patient’s anatomy. The ability to anticipate and quantify patientspecific alterations in nasal physiology after EESBS would be helpful in preoperative planning and patient counseling. At the same time, we acknowledge that outcomes after EESBS are not solely dependent on changes in nasal airflow.
Y P
CONCLUSION
O C
This pilot study showed that CFD simulations on virtual surgery models are able to reasonably detect changes in simulated nasal physiology after virtual EESBS. Moreover, these changes were influenced not only by virtual surgical technique, but also by the presence of anatomic variations.
REFERENCES 1. 2.
T 3.
Figure 6. (A) Wall shear stress contours in the olfactory cleft region for Subject 3. (B) Average wall shear stress in the olfactory cleft for all subjects. smPS, small posterior septectomy; lgPS, large posterior septectomy; smPST, small posterior septectomy and right middle turbinate resection; lgPST, large posterior septectomy and right middle turbinate resection.
O N
lateral flow partition, the changes were minimal in subject 1 (subject with no SD). Flow allocation did not change with MTR or PS, regardless of size. However, this was not the case in the setting of SD. In the subjects 2 and 3, PS produced a further imbalance in flow partition. Moreover, MTR on the opposite side of an SD produced more of an imbalance. However, MTR on the same side as a deviation redistributed flow partition to that side. Our findings in these computergenerated sinonasal models suggest that a carefully planned MTR in certain patients may improve nasal flow partition and, in the least, not worsen the sensation of unilateral obstruction. Simulated WSS attest to the fact that septectomy size and MTR do not produce consistent changes in all patients (Fig. 5, A–C). For example, the smPST model in subject 1 had higher shear stress than the presurgical model, whereas the opposite was true for subjects 2 and 3. Similarly, the lgPS model increased WSS over presurgery model in subject 2, but decreased WSS in subjects 1 and 3. The effects of virtual surgery type on air interacting with olfactory cleft mucosa indicated that MTR tended to inhibit flow interactions with the olfactory mucosa in subjects with a moderate anterior septal deflection (subjects 2 and 3); this was not true in subject 1. In addition, the magnitude of air interacting with olfactory mucosa as computed by WSS in the virtual surgery models was higher in subject 1 (no SD) than in subjects 2 and 3. Consequently, one could postulate that a subject with clinical evidence of nasal anatomic anomaly undergoing EESBS may experience diminished olfactory function postoperatively, with a likelihood of more severe symptoms if right MTR is performed. We emphasize that our findings are specific to the virtually created surgeries modeled in these three patients; therefore, generalizations
O D
Lee SC, and Senior BA. Endoscopic skull base surgery. Clin Exp Otorhinolaryngol 1:53–62, 2008. Schwartz TH, Fraser JF, Brown S, et al. Endoscopic cranial base surgery: Classification of operative approaches. Neurosurgery 62: 991–1002, 2008. Bedrosian JC, McCoul ED, Raithatha R, et al. A prospective study of postoperative symptoms in sinonasal quality-of-life following endoscopic skull-base surgery: Dissociations based on specific symptoms. Int Forum Allergy Rhinol 3:664–669, 2013. Wang YY, Srirathan V, Tirr E, et al. Nasal symptoms following endoscopic transsphenoidal pituitary surgery: Assessment using the General Nasal Patient Inventory. Neurosurg Focus 30:E12, 2011. Thompson CF, Suh JD, Liu Y, et al. Modifications to the endoscopic approach for anterior skull base lesions improve postoperative sinonasal symptoms. J Neurol Surg B Skull Base 75:65–72, 2014. Charalampaki P, Ayyad A, Kockro RA, and Perneczky A. Surgical complications after endoscopic transsphenoidal pituitary surgery. J Clin Neurosci 16:786–789, 2009. Black PM, Zervas NT, and Candia GL. Incidence and management of complications of transsphenoidal operation for pituitary adenomas. Neurosurgery 20:920–924, 1987. Laws ER Jr, and Kern EB. Complications of trans-sphenoidal surgery. Clin Neurosurg 23:401–416, 1976. Barrow DL, and Tindall GT. Loss of vision after transsphenoidal surgery. Neurosurgery 27:60–68, 1990. Cannon DE, Frank DO, Kimbell JS, et al. Modeling nasal physiology changes due to septal perforations. Otolaryngol Head Neck Surg 148:513–518, 2013. Martonen TB, Zhang Z, Yue G, and Musante CJ. Fine particle deposition within human nasal airways. Inhal Toxicol 15:283–303, 2003. Frank DO, Kimbell JS, Cannon D, and Rhee JS. Computed intranasal spray penetration: Comparisons before and after nasal surgery. Int Forum Allergy Rhinol 3:48–55, 2013. Kleven M, Melaaen MC, Reimers M, et al. Using computational fluid dynamics (CFD) to improve the bi-directional nasal drug delivery concept. Food Bioprod Process 83:107–117, 2005. Frank DO, Zanation AM, Dhandha VH, et al. Quantification of airflow into the maxillary sinuses before and after functional endoscopic sinus surgery. Int Forum Allergy Rhinol 3:834–840, 2013. Chen XB, Lee HP, Chong VF, and Wang de Y. Assessments of nasal bone fracture effects on nasal airflow: A computational fluid dynamics study. Am J Rhinol Allergy 25:e39–e43, 2011. Faramarzi M, Baradaranfar MH, Abouali O, et al. Numerical investigation of the flow field in realistic nasal septal perforation geometry. Allergy Rhinol (Providence) 2014. Kimbell JS, Garcia GJ, Frank DO, et al. Computed nasal resistance compared with patient-reported symptoms in surgically treated na-
4.
5.
6.
7.
8. 9. 10.
11. 12.
13.
14.
15.
16.
17.
American Journal of Rhinology & Allergy
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
149
18.
19.
20.
sal airway passages: A preliminary report. Am J Rhinol Allergy 26:e94–e98, 2012. Re M, Paolucci L, Romeo R, et al. Surgical treatment of nasal septal perforations. Our experience. Acta Otorhinolaryngola Ital 26:102–109, 2006. Hong Jiang W, Ping Zhao S, Hai Xie Z, et al. Endoscopic resection of chordomas in different clival regions. Acta Otolaryngol 129:71–83, 2009. Schaberg MR, Farrell CJ, Rabinowitz M, et al. Septal transposition: A novel technique for preservation of the nasal septum during endoscopic endonasal resection of olfactory groove meningiomas. Neurol Surg B 73:A165, 2012. doi: 10.1055/s-0032–1312213.
21.
Kasemsiri P, Carrau RL, Otto BA, et al. Reconstruction of the pedicled nasoseptal flap donor site with a contralateral reverse rotation flap: Technical modifications and outcomes. Laryngoscope 123:2601–2604, 2013. 22. Nyquist GG, Anand VK, Brown S, et al. Middle turbinate preservation in endoscopic transsphenoidal surgery of the anterior skull base. Skull Base 20:343–347, 2010. 23. Paluzzi A, Fernandez-Miranda JC, Pinheiro-Neto C, et al. Endoscopic endonasal infrasellar approach to the sellar and suprasellar regions: Technical note. Skull Base 21:335–342, 2011. 24. Guthikonda B, Nourbakhsh A, Notarianni C, et al. Middle turbinectomy for exposure in endoscopic endonasal transsphenoidal surgery: When is it necessary? Laryngoscope 120:2360–2366, 2010. e
Y P
T
O D 150
O C
O N
March–April 2015, Vol. 29, No. 2
Delivered by Ingenta to: Guest User IP: 191.101.55.202 On: Thu, 30 Jun 2016 09:30:01 Copyright (c) Oceanside Publications, Inc. All rights reserved. For permission to copy go to https://www.oceansidepubl.com/permission.htm
