Acta Neurochir (2014) 156:1393–1401 DOI 10.1007/s00701-014-2107-8
TECHNICAL NOTE - NEUROSURGICAL TECHNIQUES
Avoidance of postoperative epistaxis and anosmia in endonasal endoscopic skull base surgery: a technical note Chester F. Griffiths & Aaron R. Cutler & Huy T. Duong & Gal Bardo & Kian Karimi & Garni Barkhoudarian & Ricardo Carrau & Daniel F. Kelly
Received: 30 January 2014 / Accepted: 19 April 2014 / Published online: 9 May 2014 # Springer-Verlag Wien 2014
Abstract Background Most endoscopic transsphenoidal approaches jeopardize the sphenopalatine artery and septal olfactory strip (SOS), increasing the risk of postoperative anosmia and epistaxis while precluding the ability to raise pedicled nasoseptal flaps (NSF). We describe a bilateral “rescue flap” technique that preserves the mucosa containing the nasal-septal vascular pedicles and the SOS. This approach can reduce the risk of postoperative complications, including epistaxis and anosmia. Methods A retrospective analysis was conducted of all patients who underwent endoscopic transsphenoidal surgery with preservation of both sphenopalatine vascular pedicles and SOS. In a recent subset of patients, olfactory assessment was performed. Results Of 174 consecutive operations performed in 161 patients, bilateral preservation of the sphenopalatine vascular pedicle and SOS was achieved in 139 (80 %) operations, including 31 (22 %) with prior transsphenoidal surgery. Of the remaining 35 operations, 18 had a planned formal NSF and 17 had prior surgery or extensive lesions precluding use of this technique. Of pituitary adenomas, RCCs or sellar arachnoid cysts, 118 (94 %) underwent this approach, including C. F. Griffiths : G. Bardo : K. Karimi : G. Barkhoudarian (*) : D. F. Kelly Saint Johns Medical Center, John Wayne Cancer Institute, 2200 Santa Monica Blvd, Santa Monica, CA 90404, USA e-mail: [email protected] A. R. Cutler University of Utah Medical Center, Salt Lake City, UT, USA H. T. Duong Kaiser Permanente Health Center, Sacramento, CA, USA R. Carrau The Ohio State University Medical Center, Columbus, OH, USA
91 % of patients who had prior surgery. Preoperative olfaction function was maintained in 97 % of patients that were tested. None of the patients had postoperative arterial epistaxis. Conclusion Preservation of bilateral sphenopalatine vascular pedicles and the SOS is feasible in over 90 % of patients undergoing endonasal endoscopic surgery for pituitary adenomas and RCCs. This approach, while not hindering exposure or limiting instrument maneuverability, preserves the nasoseptal vasculature for future NSF use if needed and appears to minimize the risks of postoperative arterial epistaxis and anosmia. Keywords Anosmia . Cerebrospinal fluid leak . Endonasal endoscopic surgery . Epistaxis . Nasoseptal flap . Olfaction . Pituitary adenoma
Introduction The endonasal endoscopic approach has become increasingly utilized for most pathologies of the median anterior skull base. Although effective for safe tumor removal, recognized complications of this technique include postoperative cerebrospinal fluid (CSF) leaks, life-threatening epistaxis secondary to transection of the sphenopalatine arterial branches, prolonged nasal crusting with chronic bacterial colonization, endonasal synechiae causing nasal obstruction, and loss of olfaction [5, 6, 20, 30, 33, 41]. The use of the Hadad-Bassagaisteguy nasoseptal flap (NSF) has considerably reduced postoperative CSF leak rates in extended approaches with large dural openings (Grade 3 CSF leak). However, we find it is rarely necessary for smaller dural defects (Grade 1 and 2 CSF leaks) encountered in the removal of more common sellar lesions such as pituitary adenomas and Rathke’s cleft cysts (RCCs), and when raised, is associated with an increase in postoperative nasal crusting [13, 15, 20].
1394
In traditional endonasal transsphenoidal approaches, if an NSF is not elevated at the start of the procedure, one or both of the vascular pedicles may often be sacrificed, either intentionally or inadvertently, while performing the standard posterior septectomy and sphenoidotomy. This eliminates the NSF as a potential reconstruction option if a large CSF leak is encountered or if an extensive reconstruction is needed during future procedures [15, 20]. The unilateral “rescue” flap technique as first described by Rivera-Serrano et al. [35] in cadaveric dissections circumvents this problem by preserving the posterior nasoseptal arterial pedicle on one side without obstructing access to the rostrum of the sphenoid, allowing for the elevation of a full NSF should it become necessary. The availability of a unilateral NSF, identified at the start of the operation, may have limitations when, ultimately, the contralateral side would have been more optimal. In addition, the risk of postoperative epistaxis from the posterior nasal cavity also increases if terminal branches of the sphenopalatine vascular complex are ligated as may occur when the vascular pedicle is not preserved [5, 6, 30, 33]. An additional complication associated with excessive mucosal sacrifice during an endonasal endoscopic approach is the destruction of the olfactory epithelium in the area of the superior septal olfactory mucosa, a region we have termed the “septal olfactory strip” (SOS). SOS trauma or sacrifice may often result in varying degrees of postoperative hyposmia or anosmia. Although generally underreported, this loss of olfaction and associated taste dysfunction can significantly impact patients’ quality of life [18, 36]. In the initial description of the NSF, the authors recognize this potential postoperative complication and advise to incise the mucosa 1–2 cm below the superior most aspect of the septum in order to preserve best the olfactory epithelium in this region when creating the posterior superior limb of the flap [15, 20]. Although recent studies have confirmed that diminished olfaction can occur regardless of whether or not an NSF is created, if the olfactory mucosa is not adequately recognized and preserved, higher rates of postoperative olfactory dysfunction are still associated with NSF-based reconstructions [41]. Mucosal destruction and NSF elevation may also result in significant postoperative nasal crusting and discharge. In this technical note, we describe an endonasal endoscopic approach technique that aims to preserve the olfactory epithelium containing septal olfactory strip (SOS) mucosa and maintains the posterior nasal vasculature including the posterior nasal artery and its terminal septal branches bilaterally. An important aspect of this approach is the effort made to preserve as much nasal septal mucosa as possible. A minimal amount of mucosa on the face of the sphenoid is excised, and the middle turbinates are preserved in the majority of cases except when more lateral exposure is necessary. Overall the technique strives to promote rapid mucosal healing, reduce the risk of postoperative epistaxis, while not compromising sellar
Acta Neurochir (2014) 156:1393–1401
and parasellar exposure or the ability to perform an skull base repair with a NSF should a large CSF leak be encountered. It also ensures the availability of a NSF on either side for future procedures, if needed.
Clinical material and methods Patients We performed a retrospective analysis of operative reports and postoperative follow up records in all patients who underwent an endonasal endoscopic approach for tumor or cyst removal at Saint John’s Healthcare Center from August 2011 to July 2013. We specifically identified patients who received a bilateral nasal mucosal incisional technique during the initial approach with preservation of both vascular pedicles and the SOS. The study was reviewed by our institutional review board (IRB) and was granted an exemption waiver. Patient selection A bilateral nasal mucosal incisional technique was used in all patients who underwent first-time surgery for a pituitary adenoma or RCC, as well as for other intrasellar and/or primarily extradural tumors. In contrast, a vascularized NSF was routinely used in extended skull base approaches for intradural tumors such as tuberculum sellae meningiomas and craniopharyngiomas, in which a large dural and arachnoid defect was created (Grade 3 CSF leak), as well as for clival chordomas without dural invasion given the likely need for postoperative radiotherapy. Patients who have undergone sinus or transsphenoidal surgery in the past were also potential candidates for a unilateral or bilateral nasal mucosal incisional technique, but this determination is made intraoperatively. We were unable to perform a full bilateral mucosal and vascular pedicle preserving approach only in patients where one of the vascular pedicles had been obliterated, either from previous surgery or because of extensive tumor invasion. Surgical technique Following the induction of general anesthesia and endotracheal intubation, the patient’s head is angled 20°–30° towards the left shoulder, either in a horseshoe headholder—for standard endonasal cases—or in three-point Mayfield fixation for procedures expected to exceed 4 h. Frameless image-guided navigation is used in all cases (Stryker Surgical Navigation System II, Kalamazoo, MI). The nasal cavity is prepared by placing oxymetazolinesoaked ½×3 in. cottonoids in both nares for several minutes. The nasal region and right lower abdominal area (for potential fat graft) are prepped and draped. The initial approach to the sphenoid sinus is performed by the otolaryngologist with a 0°, 4 mm rigid endoscope (Karl Storz, Tuttlingen, Germany). A solution of 1 % lidocaine with epinephrine 1:100,000 is first injected into the middle turbinates bilaterally. A thorough endonasal endoscopic evaluation for septal or sinus abnormalities (i.e. septal spurs, conchae bullosa, sinonasal polyps, etc.)
Acta Neurochir (2014) 156:1393–1401
is performed, and a decision is made whether any of these abnormalities need to be addressed to enhance surgical exposure and ingress of instruments. The inferior and middle turbinates are then out-fractured bilaterally and the sphenoid ostia are identified at the root of the superior turbinate on both sides (Fig. 1). In some cases, the sphenoid ostium is not immediately visualized. In these situations, the posterior insertion of the superior turbinate is utilized as an anatomic landmark to safely estimate the location of the sphenoid ostium, which is typically superior to this insertion. Incisions are never made below the point of the superior turbinate insertion to preserve the posterior nasal artery vascular pedicle. Routine resection of the middle turbinates is not required in this approach. After the sphenoid ostia have been identified, bilateral nasal mucosal incisions are made with preservation of both the right and left vascular pedicles and the SOS mucosa (Fig. 2). As previously described by Pinheiro-Nieto et al. [31], the distance between the sphenoid ostium and nasoseptal arterial pedicle averages 9 mm, allowing for a horizontal incision to be safely made 2–3 mm below the ostium without fear of injuring the vascular pedicle (Fig. 3). The mucoperiosteal incisions are made with an extended shaft, manually bent, curved micro-tip electrocautery. (Megadyne E-Z Clean 6.0”/152 mm ref: 0016 M, Megadyne, Draper, Utah, USA), beginning just inferior to the sphenoid ostium and extending laterally for a few millimeters. The incision is carried anteriorly and horizontally for approximately 2 cm along the perpendicular plate of the ethmoid and posterior nasal septum, and ending at a point opposite to the anterior most portion of the middle turbinate. It is then extended further anteriorly and superiorly in a hockey stick-shaped fashion to facilitate flap mobilization; this is the same superior
Fig. 1 Intraoperative photograph of the right nasal cavity with the 0°, 4 mm endoscope of initial endonasal approach to the sphenoid sinus. The inferior (not seen) and middle turbinates (MT) are out-fractured and displaced laterally and the sphenoid ostium (SO) is identified at the root of the superior turbinate (ST). The course of the planned mucoperiosteum incision is demarcated by the blue line; the mucosa of the septal olfactory strip (SOS) above and the “rescue” flap (RF) below the incision is noted. (NS—nasal septum)
1395
incision used to raise the formal NSF and preserves the posterior septal mucosa that contains much of the olfactory epithelium [35]. However, if a pedicled nasoseptal flap is not required, the incision does not extend further anteriorly, leaving the anterior mucosa intact on the nasal septum. Once completed on one side, the mucosal incisional technique is made contralaterally. This positioning of the mucoperiosteal incisions preserves the bilateral vascular pedicles running inferior and parallel to the incision while also maintaining the SOS located superiorly and posteriorly. Both “rescue” flaps with their intact vascular pedicles are then dissected downwards towards the nasopharynx to minimize visual obstruction into the sphenoid sinus and are maintained in this inferior position by placing bilateral 1/2×3 in. cottonoids in the sub-periosteal space at the base of the sphenoid “keel”. The olfactory mucosa containing SOS is also preserved with gentle elevation along the remaining superior portion of the nasal septum. These SOS mucosal flaps are reflected laterally and usually adhere to the superior turbinate tissue outside of the operative field of vision. Once both “rescue” and SOS flaps have been safely placed at the periphery of the operative corridor, a generous posterior bony septectomy and wide sphenoidotomy are performed. An attempt is made to remove the vomer in one piece so as to preserve the bone for sellar floor reconstruction if necessary at the time of closure. A 30° rod-lens endoscope (Karl Storz Endoscopy, Tuttlingen, Germany) is used to visualize the superior nasal cavity, lateral sphenoid wall, and planum sphenoidale, while opening the superior aspect of the sphenoidotomy. Posterior ethmoid air cells are removed as necessary. Sellar exposure, tumor removal and sellar floor reconstruction is then performed in standard fashion. At the conclusion of the sellar phase of the operation, blood is aspirated from the sphenoid sinus, nasal cavity, and nasopharynx, and nasal mucosal hemostasis is obtained to minimize the amount of blood swallowed in the immediate postoperative period. In the vast majority of cases in which a nasoseptal flap is not needed, cottonoids are removed and the preserved mucosa is elevated back along the remaining vomer and inferior nasal septum to their original location. The middle turbinates are repositioned anatomically and the SOSs on both sides are lowered back into a normal anatomic position (Fig. 4). In rare cases that a nasoseptal flap is necessary for skull base repair, the flap incisions may be extended in standard fashion. Typically, nasal packing is not used unless it is needed as a temporary sellar buttress. Patient follow-up Outpatient nasal debridements are typically performed on or around postoperative day 10 and then two more times at 2-week intervals until the nasal cavity is completely healed and free of crusting (Fig. 5). Objective olfactory analysis is measured preoperatively and postoperatively at a minimum of 3 months post-surgery using the Brief
1396
Fig. 2 Intraoperative photographs with 0° endoscope of bilateral nasal mucosal incisional technique being performed. a Within the left nasal cavity, the mucoperiosteum of the nasal septum is incised using an extended shaft, curved micro-tip Bovie (Megadyne E-Z Clean 6.0″/ 152 mm ref: 0016 M, Megadyne, Draper, Utah, USA). The incision begins a few mm below the sphenoid ostium and extends approximately 2 cm anteriorly at the mid-level of the middle turbinate. b The superior olfactory strip (SOS) has been released upward and the “rescue” flap (RF) is being dissected inferiorly towards the nasopharynx with a Cottle
Acta Neurochir (2014) 156:1393–1401
elevator. c Protecting the “rescue” flap (RF) with placement of ½×3 in. cottonoids in the sub-periosteal space. d Posterior septectomy completed and sphenoid ostia opened with vomer left intact and “rescue” flaps (RF) displaced inferiorly with bilateral ½×3 in. cottonoids. e Sphenoidotomy completed and vomer removed revealing unobstructed view of sellae and upper clivus. f Working instruments in sphenoid sinus and sellae during removal of pituitary adenoma without obstruction from “rescue” flaps (RF)
Smell Identification Test (BSIT, Sensonics Corporation; Haddon Heights, NJ, USA). Adequate follow-up was obtained in a subset of patients, as routine post-operative BSIT analysis was initiated in May, 2012.
Results One hundred and seventy-four consecutive operations were performed in 161 patients who underwent endoscopic endonasal transsphenoidal, resection of a sellar or parasellar lesion from August 2011 to July 2013, (mean age 51.7+/− 14.5 years). Pathologies included 112 pituitary adenomas, 11 RCC, eight meningiomas, seven craniopharyngiomas, 10 metastatic lesions, three sellar arachnoid cysts, three clival chordomas, and 16 other parasellar lesions. In total, 134 (77 %) operations allowed for preservation of bilateral nasoseptal vascular pedicles and the SOS (Table 1). Of the remaining 40 operations, five had unilateral preservation of the nasoseptal pedicle and SOS due to their pathology, 18 had a full NSF elevated at the start of the procedure and 17 had undergone prior surgery or had extensive tumor invasion with compromised vascular pedicles on both sides. Only one patient required intraoperative conversion from a rescue flap to a complete nasoseptal flap. The mean age of patients undergoing the 139 operations in which preservation of unilateral or bilateral nasoseptal
Fig. 3 Illustration of the bilateral nasal mucosal incisional technique concept with preservation of sphenopalatine and posterior nasoseptal arteries along with the olfactory epithelium containing septal olfactory strip (SOS). The mucosal cuts are started several millimeters below the inferior aspect of the sphenoid ostia and carried anteriorly as shown by the green line. The dotted line shows mucosal cuts if a complete nasoseptal flap is needed. The double arrow denotes the typical 9 mm distance between the inferior edge of the ostium and posterior nasal septal artery. The blue shading indicates the extent of the posterior septectomy (PS)
Acta Neurochir (2014) 156:1393–1401
Fig. 4 Following the completion of tumor removal if a nasoseptal flap (NSF) is not required the a nasoseptal “rescue” flap (RF) and b superior olfactory strip (SOS) are carefully dissected back into their original anatomic locations
vascular pedicles and SOS was possible, was 51.6 years. Their pathology included 105 pituitary adenomas (34 functional, 71 nonfunctional), 11 RCCs, and 23 other lesions (see Table 1). Thirty-eight tumors (27.3 %) had suprasellar and/or parasellar extension, 47 (33.8 %) had cavernous sinus invasion as determined during surgery. Thirty-one operations (22.3 %) were performed in patients who had undergone previous endonasal or sublabial transsphenoidal surgery. Indications for the 174 operations included 126 (72.4 %) pituitary adenomas, Rathke’s Cleft Cysts, or arachnoid cysts, which were primarily sellar or with suprasellar extension. We
Fig. 5 Postoperative outpatient follow up with endoscopic visualization of the nasal cavity demonstrating mucosal preservation and good healing of the previously elevated a nasoseptal “rescue” flap (RF) and b septal
1397
were able to perform bilateral mucosal preservation technique in 118 (94 %) of these cases (Table 2). This was slightly less successful in patients undergoing re-operative surgery (90.6 %) compared to primary operations (94.7 %). All seven craniopharyngioma patients underwent pedicled nasoseptal flap elevation for grade 3 CSF leaks. Of the operations that had unilateral or bilateral rescue flaps, an intraoperative CSF leak was encountered in 77 operations (55.4 %) including 28 Grade 1, 41 Grade 2, and eight Grade 3 leaks. Two operations (1.5 %), both with Grade 2 intraoperative CSF leaks, developed postoperative leaks requiring reoperation. In both, there was evidence of dislodgement of the rigid sellar buttress, (bone or titanium mesh) noted at reoperation. Both leaks were repaired uneventfully with new placement of fat, collagen sponge and a rigid buttress. Two operations with Grade 3 CSF leaks, repaired with a nasoseptal mucosal flap, developed postoperative leaks requiring reoperation. There were no instances of postoperative arterial epistaxis in the total cohort of 161 patients. Preoperative and postoperative objective smell analysis (BSIT) was collected in 35 of 128 (27 %) of these patients. There was no significant difference in the lesion types or extent in this subset of patients. Preoperative olfactory dysfunction was identified in four patients (11.4 %). Of these 35 patients, one patient (3 %), a 74 year old man, had new, mild postoperative deterioration (BSITscore: 9 to 7). Median follow-up duration was 6 months. None of the patients with preoperative hyposmia had further deterioration of their olfactory dysfunction.
Discussion The use of vascularized NSFs has transformed endoscopic cranial base surgery by improving the quality of CSF leak reconstructions associated with extended endonasal approaches to pathologies within the subarachnoid space [15, 38, 43]. Overall NSFs have contributed to reducing the
olfactory strip (SOS) at 3 weeks and again at c 3 months with progressive opposition of the “rescue” flap (RF) and SOS and with minimal nasal crusting seen
1398
Acta Neurochir (2014) 156:1393–1401
Table 1 Detailed demographic and operative characteristics of patients who underwent a mucosal preserving technique for transsphenoidal surgery Variable
N=139/174 (79.9 %)
Mean age in years Men Women Tumor type Pituitary adenoma Endocrine-inactive adenoma Functional pituitary adenoma Cushing’s disease Prolactinoma
51.6 70 (50.4 %) 69 (49.6 %) 105 71 34 17 9
Acromegaly Rathke’s Cleft Cysts (RCCs) Sellar metastases Meningioma Arachnoid Cysts B-Cell lymphoma Fibrous dysplasia Negative exploration for Cushing’s Clival chordoma Hemangioma Cholesterol granuloma Mucocele Postoperative CSF Leak Repair Postoperative Hematoma Re-do operations Suprasellar/parasellar extension Cavernous sinus invasion Intraoperative cerebrospinal fluid (CSF) leak
8 11 5 3 3 2 2 2 1 1 1 1 1 1 31 (22.3 38 (27.3 47 (33.8 77 (55.4
Grade 1 Grade 2 Grade 3 Postoperative CSF leak New olfactory deficit (B-UPSIT) Postoperative epistaxis
28 41 8 2 (1.4 %) 1/35 (2.9 %) 0
%) %) %) %)
incidence of postoperative CSF leak rates in extended endonasal endoscopic surgery from 20 % or higher to
TECHNICAL NOTE - NEUROSURGICAL TECHNIQUES
Avoidance of postoperative epistaxis and anosmia in endonasal endoscopic skull base surgery: a technical note Chester F. Griffiths & Aaron R. Cutler & Huy T. Duong & Gal Bardo & Kian Karimi & Garni Barkhoudarian & Ricardo Carrau & Daniel F. Kelly
Received: 30 January 2014 / Accepted: 19 April 2014 / Published online: 9 May 2014 # Springer-Verlag Wien 2014
Abstract Background Most endoscopic transsphenoidal approaches jeopardize the sphenopalatine artery and septal olfactory strip (SOS), increasing the risk of postoperative anosmia and epistaxis while precluding the ability to raise pedicled nasoseptal flaps (NSF). We describe a bilateral “rescue flap” technique that preserves the mucosa containing the nasal-septal vascular pedicles and the SOS. This approach can reduce the risk of postoperative complications, including epistaxis and anosmia. Methods A retrospective analysis was conducted of all patients who underwent endoscopic transsphenoidal surgery with preservation of both sphenopalatine vascular pedicles and SOS. In a recent subset of patients, olfactory assessment was performed. Results Of 174 consecutive operations performed in 161 patients, bilateral preservation of the sphenopalatine vascular pedicle and SOS was achieved in 139 (80 %) operations, including 31 (22 %) with prior transsphenoidal surgery. Of the remaining 35 operations, 18 had a planned formal NSF and 17 had prior surgery or extensive lesions precluding use of this technique. Of pituitary adenomas, RCCs or sellar arachnoid cysts, 118 (94 %) underwent this approach, including C. F. Griffiths : G. Bardo : K. Karimi : G. Barkhoudarian (*) : D. F. Kelly Saint Johns Medical Center, John Wayne Cancer Institute, 2200 Santa Monica Blvd, Santa Monica, CA 90404, USA e-mail: [email protected] A. R. Cutler University of Utah Medical Center, Salt Lake City, UT, USA H. T. Duong Kaiser Permanente Health Center, Sacramento, CA, USA R. Carrau The Ohio State University Medical Center, Columbus, OH, USA
91 % of patients who had prior surgery. Preoperative olfaction function was maintained in 97 % of patients that were tested. None of the patients had postoperative arterial epistaxis. Conclusion Preservation of bilateral sphenopalatine vascular pedicles and the SOS is feasible in over 90 % of patients undergoing endonasal endoscopic surgery for pituitary adenomas and RCCs. This approach, while not hindering exposure or limiting instrument maneuverability, preserves the nasoseptal vasculature for future NSF use if needed and appears to minimize the risks of postoperative arterial epistaxis and anosmia. Keywords Anosmia . Cerebrospinal fluid leak . Endonasal endoscopic surgery . Epistaxis . Nasoseptal flap . Olfaction . Pituitary adenoma
Introduction The endonasal endoscopic approach has become increasingly utilized for most pathologies of the median anterior skull base. Although effective for safe tumor removal, recognized complications of this technique include postoperative cerebrospinal fluid (CSF) leaks, life-threatening epistaxis secondary to transection of the sphenopalatine arterial branches, prolonged nasal crusting with chronic bacterial colonization, endonasal synechiae causing nasal obstruction, and loss of olfaction [5, 6, 20, 30, 33, 41]. The use of the Hadad-Bassagaisteguy nasoseptal flap (NSF) has considerably reduced postoperative CSF leak rates in extended approaches with large dural openings (Grade 3 CSF leak). However, we find it is rarely necessary for smaller dural defects (Grade 1 and 2 CSF leaks) encountered in the removal of more common sellar lesions such as pituitary adenomas and Rathke’s cleft cysts (RCCs), and when raised, is associated with an increase in postoperative nasal crusting [13, 15, 20].
1394
In traditional endonasal transsphenoidal approaches, if an NSF is not elevated at the start of the procedure, one or both of the vascular pedicles may often be sacrificed, either intentionally or inadvertently, while performing the standard posterior septectomy and sphenoidotomy. This eliminates the NSF as a potential reconstruction option if a large CSF leak is encountered or if an extensive reconstruction is needed during future procedures [15, 20]. The unilateral “rescue” flap technique as first described by Rivera-Serrano et al. [35] in cadaveric dissections circumvents this problem by preserving the posterior nasoseptal arterial pedicle on one side without obstructing access to the rostrum of the sphenoid, allowing for the elevation of a full NSF should it become necessary. The availability of a unilateral NSF, identified at the start of the operation, may have limitations when, ultimately, the contralateral side would have been more optimal. In addition, the risk of postoperative epistaxis from the posterior nasal cavity also increases if terminal branches of the sphenopalatine vascular complex are ligated as may occur when the vascular pedicle is not preserved [5, 6, 30, 33]. An additional complication associated with excessive mucosal sacrifice during an endonasal endoscopic approach is the destruction of the olfactory epithelium in the area of the superior septal olfactory mucosa, a region we have termed the “septal olfactory strip” (SOS). SOS trauma or sacrifice may often result in varying degrees of postoperative hyposmia or anosmia. Although generally underreported, this loss of olfaction and associated taste dysfunction can significantly impact patients’ quality of life [18, 36]. In the initial description of the NSF, the authors recognize this potential postoperative complication and advise to incise the mucosa 1–2 cm below the superior most aspect of the septum in order to preserve best the olfactory epithelium in this region when creating the posterior superior limb of the flap [15, 20]. Although recent studies have confirmed that diminished olfaction can occur regardless of whether or not an NSF is created, if the olfactory mucosa is not adequately recognized and preserved, higher rates of postoperative olfactory dysfunction are still associated with NSF-based reconstructions [41]. Mucosal destruction and NSF elevation may also result in significant postoperative nasal crusting and discharge. In this technical note, we describe an endonasal endoscopic approach technique that aims to preserve the olfactory epithelium containing septal olfactory strip (SOS) mucosa and maintains the posterior nasal vasculature including the posterior nasal artery and its terminal septal branches bilaterally. An important aspect of this approach is the effort made to preserve as much nasal septal mucosa as possible. A minimal amount of mucosa on the face of the sphenoid is excised, and the middle turbinates are preserved in the majority of cases except when more lateral exposure is necessary. Overall the technique strives to promote rapid mucosal healing, reduce the risk of postoperative epistaxis, while not compromising sellar
Acta Neurochir (2014) 156:1393–1401
and parasellar exposure or the ability to perform an skull base repair with a NSF should a large CSF leak be encountered. It also ensures the availability of a NSF on either side for future procedures, if needed.
Clinical material and methods Patients We performed a retrospective analysis of operative reports and postoperative follow up records in all patients who underwent an endonasal endoscopic approach for tumor or cyst removal at Saint John’s Healthcare Center from August 2011 to July 2013. We specifically identified patients who received a bilateral nasal mucosal incisional technique during the initial approach with preservation of both vascular pedicles and the SOS. The study was reviewed by our institutional review board (IRB) and was granted an exemption waiver. Patient selection A bilateral nasal mucosal incisional technique was used in all patients who underwent first-time surgery for a pituitary adenoma or RCC, as well as for other intrasellar and/or primarily extradural tumors. In contrast, a vascularized NSF was routinely used in extended skull base approaches for intradural tumors such as tuberculum sellae meningiomas and craniopharyngiomas, in which a large dural and arachnoid defect was created (Grade 3 CSF leak), as well as for clival chordomas without dural invasion given the likely need for postoperative radiotherapy. Patients who have undergone sinus or transsphenoidal surgery in the past were also potential candidates for a unilateral or bilateral nasal mucosal incisional technique, but this determination is made intraoperatively. We were unable to perform a full bilateral mucosal and vascular pedicle preserving approach only in patients where one of the vascular pedicles had been obliterated, either from previous surgery or because of extensive tumor invasion. Surgical technique Following the induction of general anesthesia and endotracheal intubation, the patient’s head is angled 20°–30° towards the left shoulder, either in a horseshoe headholder—for standard endonasal cases—or in three-point Mayfield fixation for procedures expected to exceed 4 h. Frameless image-guided navigation is used in all cases (Stryker Surgical Navigation System II, Kalamazoo, MI). The nasal cavity is prepared by placing oxymetazolinesoaked ½×3 in. cottonoids in both nares for several minutes. The nasal region and right lower abdominal area (for potential fat graft) are prepped and draped. The initial approach to the sphenoid sinus is performed by the otolaryngologist with a 0°, 4 mm rigid endoscope (Karl Storz, Tuttlingen, Germany). A solution of 1 % lidocaine with epinephrine 1:100,000 is first injected into the middle turbinates bilaterally. A thorough endonasal endoscopic evaluation for septal or sinus abnormalities (i.e. septal spurs, conchae bullosa, sinonasal polyps, etc.)
Acta Neurochir (2014) 156:1393–1401
is performed, and a decision is made whether any of these abnormalities need to be addressed to enhance surgical exposure and ingress of instruments. The inferior and middle turbinates are then out-fractured bilaterally and the sphenoid ostia are identified at the root of the superior turbinate on both sides (Fig. 1). In some cases, the sphenoid ostium is not immediately visualized. In these situations, the posterior insertion of the superior turbinate is utilized as an anatomic landmark to safely estimate the location of the sphenoid ostium, which is typically superior to this insertion. Incisions are never made below the point of the superior turbinate insertion to preserve the posterior nasal artery vascular pedicle. Routine resection of the middle turbinates is not required in this approach. After the sphenoid ostia have been identified, bilateral nasal mucosal incisions are made with preservation of both the right and left vascular pedicles and the SOS mucosa (Fig. 2). As previously described by Pinheiro-Nieto et al. [31], the distance between the sphenoid ostium and nasoseptal arterial pedicle averages 9 mm, allowing for a horizontal incision to be safely made 2–3 mm below the ostium without fear of injuring the vascular pedicle (Fig. 3). The mucoperiosteal incisions are made with an extended shaft, manually bent, curved micro-tip electrocautery. (Megadyne E-Z Clean 6.0”/152 mm ref: 0016 M, Megadyne, Draper, Utah, USA), beginning just inferior to the sphenoid ostium and extending laterally for a few millimeters. The incision is carried anteriorly and horizontally for approximately 2 cm along the perpendicular plate of the ethmoid and posterior nasal septum, and ending at a point opposite to the anterior most portion of the middle turbinate. It is then extended further anteriorly and superiorly in a hockey stick-shaped fashion to facilitate flap mobilization; this is the same superior
Fig. 1 Intraoperative photograph of the right nasal cavity with the 0°, 4 mm endoscope of initial endonasal approach to the sphenoid sinus. The inferior (not seen) and middle turbinates (MT) are out-fractured and displaced laterally and the sphenoid ostium (SO) is identified at the root of the superior turbinate (ST). The course of the planned mucoperiosteum incision is demarcated by the blue line; the mucosa of the septal olfactory strip (SOS) above and the “rescue” flap (RF) below the incision is noted. (NS—nasal septum)
1395
incision used to raise the formal NSF and preserves the posterior septal mucosa that contains much of the olfactory epithelium [35]. However, if a pedicled nasoseptal flap is not required, the incision does not extend further anteriorly, leaving the anterior mucosa intact on the nasal septum. Once completed on one side, the mucosal incisional technique is made contralaterally. This positioning of the mucoperiosteal incisions preserves the bilateral vascular pedicles running inferior and parallel to the incision while also maintaining the SOS located superiorly and posteriorly. Both “rescue” flaps with their intact vascular pedicles are then dissected downwards towards the nasopharynx to minimize visual obstruction into the sphenoid sinus and are maintained in this inferior position by placing bilateral 1/2×3 in. cottonoids in the sub-periosteal space at the base of the sphenoid “keel”. The olfactory mucosa containing SOS is also preserved with gentle elevation along the remaining superior portion of the nasal septum. These SOS mucosal flaps are reflected laterally and usually adhere to the superior turbinate tissue outside of the operative field of vision. Once both “rescue” and SOS flaps have been safely placed at the periphery of the operative corridor, a generous posterior bony septectomy and wide sphenoidotomy are performed. An attempt is made to remove the vomer in one piece so as to preserve the bone for sellar floor reconstruction if necessary at the time of closure. A 30° rod-lens endoscope (Karl Storz Endoscopy, Tuttlingen, Germany) is used to visualize the superior nasal cavity, lateral sphenoid wall, and planum sphenoidale, while opening the superior aspect of the sphenoidotomy. Posterior ethmoid air cells are removed as necessary. Sellar exposure, tumor removal and sellar floor reconstruction is then performed in standard fashion. At the conclusion of the sellar phase of the operation, blood is aspirated from the sphenoid sinus, nasal cavity, and nasopharynx, and nasal mucosal hemostasis is obtained to minimize the amount of blood swallowed in the immediate postoperative period. In the vast majority of cases in which a nasoseptal flap is not needed, cottonoids are removed and the preserved mucosa is elevated back along the remaining vomer and inferior nasal septum to their original location. The middle turbinates are repositioned anatomically and the SOSs on both sides are lowered back into a normal anatomic position (Fig. 4). In rare cases that a nasoseptal flap is necessary for skull base repair, the flap incisions may be extended in standard fashion. Typically, nasal packing is not used unless it is needed as a temporary sellar buttress. Patient follow-up Outpatient nasal debridements are typically performed on or around postoperative day 10 and then two more times at 2-week intervals until the nasal cavity is completely healed and free of crusting (Fig. 5). Objective olfactory analysis is measured preoperatively and postoperatively at a minimum of 3 months post-surgery using the Brief
1396
Fig. 2 Intraoperative photographs with 0° endoscope of bilateral nasal mucosal incisional technique being performed. a Within the left nasal cavity, the mucoperiosteum of the nasal septum is incised using an extended shaft, curved micro-tip Bovie (Megadyne E-Z Clean 6.0″/ 152 mm ref: 0016 M, Megadyne, Draper, Utah, USA). The incision begins a few mm below the sphenoid ostium and extends approximately 2 cm anteriorly at the mid-level of the middle turbinate. b The superior olfactory strip (SOS) has been released upward and the “rescue” flap (RF) is being dissected inferiorly towards the nasopharynx with a Cottle
Acta Neurochir (2014) 156:1393–1401
elevator. c Protecting the “rescue” flap (RF) with placement of ½×3 in. cottonoids in the sub-periosteal space. d Posterior septectomy completed and sphenoid ostia opened with vomer left intact and “rescue” flaps (RF) displaced inferiorly with bilateral ½×3 in. cottonoids. e Sphenoidotomy completed and vomer removed revealing unobstructed view of sellae and upper clivus. f Working instruments in sphenoid sinus and sellae during removal of pituitary adenoma without obstruction from “rescue” flaps (RF)
Smell Identification Test (BSIT, Sensonics Corporation; Haddon Heights, NJ, USA). Adequate follow-up was obtained in a subset of patients, as routine post-operative BSIT analysis was initiated in May, 2012.
Results One hundred and seventy-four consecutive operations were performed in 161 patients who underwent endoscopic endonasal transsphenoidal, resection of a sellar or parasellar lesion from August 2011 to July 2013, (mean age 51.7+/− 14.5 years). Pathologies included 112 pituitary adenomas, 11 RCC, eight meningiomas, seven craniopharyngiomas, 10 metastatic lesions, three sellar arachnoid cysts, three clival chordomas, and 16 other parasellar lesions. In total, 134 (77 %) operations allowed for preservation of bilateral nasoseptal vascular pedicles and the SOS (Table 1). Of the remaining 40 operations, five had unilateral preservation of the nasoseptal pedicle and SOS due to their pathology, 18 had a full NSF elevated at the start of the procedure and 17 had undergone prior surgery or had extensive tumor invasion with compromised vascular pedicles on both sides. Only one patient required intraoperative conversion from a rescue flap to a complete nasoseptal flap. The mean age of patients undergoing the 139 operations in which preservation of unilateral or bilateral nasoseptal
Fig. 3 Illustration of the bilateral nasal mucosal incisional technique concept with preservation of sphenopalatine and posterior nasoseptal arteries along with the olfactory epithelium containing septal olfactory strip (SOS). The mucosal cuts are started several millimeters below the inferior aspect of the sphenoid ostia and carried anteriorly as shown by the green line. The dotted line shows mucosal cuts if a complete nasoseptal flap is needed. The double arrow denotes the typical 9 mm distance between the inferior edge of the ostium and posterior nasal septal artery. The blue shading indicates the extent of the posterior septectomy (PS)
Acta Neurochir (2014) 156:1393–1401
Fig. 4 Following the completion of tumor removal if a nasoseptal flap (NSF) is not required the a nasoseptal “rescue” flap (RF) and b superior olfactory strip (SOS) are carefully dissected back into their original anatomic locations
vascular pedicles and SOS was possible, was 51.6 years. Their pathology included 105 pituitary adenomas (34 functional, 71 nonfunctional), 11 RCCs, and 23 other lesions (see Table 1). Thirty-eight tumors (27.3 %) had suprasellar and/or parasellar extension, 47 (33.8 %) had cavernous sinus invasion as determined during surgery. Thirty-one operations (22.3 %) were performed in patients who had undergone previous endonasal or sublabial transsphenoidal surgery. Indications for the 174 operations included 126 (72.4 %) pituitary adenomas, Rathke’s Cleft Cysts, or arachnoid cysts, which were primarily sellar or with suprasellar extension. We
Fig. 5 Postoperative outpatient follow up with endoscopic visualization of the nasal cavity demonstrating mucosal preservation and good healing of the previously elevated a nasoseptal “rescue” flap (RF) and b septal
1397
were able to perform bilateral mucosal preservation technique in 118 (94 %) of these cases (Table 2). This was slightly less successful in patients undergoing re-operative surgery (90.6 %) compared to primary operations (94.7 %). All seven craniopharyngioma patients underwent pedicled nasoseptal flap elevation for grade 3 CSF leaks. Of the operations that had unilateral or bilateral rescue flaps, an intraoperative CSF leak was encountered in 77 operations (55.4 %) including 28 Grade 1, 41 Grade 2, and eight Grade 3 leaks. Two operations (1.5 %), both with Grade 2 intraoperative CSF leaks, developed postoperative leaks requiring reoperation. In both, there was evidence of dislodgement of the rigid sellar buttress, (bone or titanium mesh) noted at reoperation. Both leaks were repaired uneventfully with new placement of fat, collagen sponge and a rigid buttress. Two operations with Grade 3 CSF leaks, repaired with a nasoseptal mucosal flap, developed postoperative leaks requiring reoperation. There were no instances of postoperative arterial epistaxis in the total cohort of 161 patients. Preoperative and postoperative objective smell analysis (BSIT) was collected in 35 of 128 (27 %) of these patients. There was no significant difference in the lesion types or extent in this subset of patients. Preoperative olfactory dysfunction was identified in four patients (11.4 %). Of these 35 patients, one patient (3 %), a 74 year old man, had new, mild postoperative deterioration (BSITscore: 9 to 7). Median follow-up duration was 6 months. None of the patients with preoperative hyposmia had further deterioration of their olfactory dysfunction.
Discussion The use of vascularized NSFs has transformed endoscopic cranial base surgery by improving the quality of CSF leak reconstructions associated with extended endonasal approaches to pathologies within the subarachnoid space [15, 38, 43]. Overall NSFs have contributed to reducing the
olfactory strip (SOS) at 3 weeks and again at c 3 months with progressive opposition of the “rescue” flap (RF) and SOS and with minimal nasal crusting seen
1398
Acta Neurochir (2014) 156:1393–1401
Table 1 Detailed demographic and operative characteristics of patients who underwent a mucosal preserving technique for transsphenoidal surgery Variable
N=139/174 (79.9 %)
Mean age in years Men Women Tumor type Pituitary adenoma Endocrine-inactive adenoma Functional pituitary adenoma Cushing’s disease Prolactinoma
51.6 70 (50.4 %) 69 (49.6 %) 105 71 34 17 9
Acromegaly Rathke’s Cleft Cysts (RCCs) Sellar metastases Meningioma Arachnoid Cysts B-Cell lymphoma Fibrous dysplasia Negative exploration for Cushing’s Clival chordoma Hemangioma Cholesterol granuloma Mucocele Postoperative CSF Leak Repair Postoperative Hematoma Re-do operations Suprasellar/parasellar extension Cavernous sinus invasion Intraoperative cerebrospinal fluid (CSF) leak
8 11 5 3 3 2 2 2 1 1 1 1 1 1 31 (22.3 38 (27.3 47 (33.8 77 (55.4
Grade 1 Grade 2 Grade 3 Postoperative CSF leak New olfactory deficit (B-UPSIT) Postoperative epistaxis
28 41 8 2 (1.4 %) 1/35 (2.9 %) 0
%) %) %) %)
incidence of postoperative CSF leak rates in extended endonasal endoscopic surgery from 20 % or higher to
