Financial support provided by the Department of Otolaryngology, Wake Forest University School of Medicine.
Preservation of Olfaction in Anterior Skull Base Surgery †
Article first published online: 2 JAN 2009
Copyright © 2000 The Triological Society
Volume 110, Issue 8, pages 1317–1322, August 2000
How to Cite
Dale Browne, J. and Mims, J. W. (2000), Preservation of Olfaction in Anterior Skull Base Surgery . The Laryngoscope, 110: 1317–1322. doi: 10.1097/00005537-200008000-00017
- Issue published online: 2 JAN 2009
- Article first published online: 2 JAN 2009
- Manuscript Accepted: 9 FEB 2000
- Anterior skull base surgery;
- craniofacial resection;
Objectives/Hypothesis In selected unilateral tumors and defects of the anterior skull base, the preservation of contralateral olfaction is achievable through a localized subcranial approach without compromising surgical objectives of resection or repair.
Study Design Description of a functional adaptation of anterior skull base surgical techniques through a retrospective patient series.
Methods Nine patients underwent anterior skull base surgery for unilateral cribriform plate disease including four malignant and two benign tumors, two encephaloceles, and one iatrogenic cribriform injury with cerebrospinal rhinorrhea. All nine patients consented to a localized subcranial approach to the anterior skull base to preserve the contralateral olfactory nerves. In four patients with benign disease a portion of the ipsilateral nerves was additionally conserved. Postoperative olfaction was assessed objectively with a commercially available smell test. Indications, technique, results, and complications are reported and discussed.
Results All patients had eradication of disease with preservation of functional olfaction
Conclusions Conservation of olfaction is possible in selected cases of anterior skull base surgery when the lesion is unilaterally confined.
Anterior skull base surgery was pioneered in the 1940s for orbital tumors 1,2 and later expanded to paranasal sinus tumors. 3 Modifications in the approach to the anterior skull base have evolved to accommodate broad applications for congenital, traumatic, and neoplastic disease of this region. Initial access to the anterior skull base was established through a frontal craniotomy with retraction of the frontal lobe and exposure of the orbital roofs and cribriform plate. 3 Later, others advocated a subcranial approach from an anterior perspective to minimize morbidity associated with frontal lobe retraction. 4–6 Additionally, advances in reconstruction, imaging, antibiotics, and multiplanar radiation therapy have furthered the success of surgery in this region. 7
Independent of approach, full superior exposure of the anterior skull base requires lifting the olfactory bulbs from the cribriform plate with resultant loss of smell. Anosmia deprives the patient of an important sense. It reduces their appreciation of food, prevents them from sensing harmful smoke or fumes, and impairs their discernment of spoilage. Although loss of olfaction is frequently necessary to eradicate disease of the anterior skull base, contralateral olfaction can be preserved during surgery for selective unilateral disorders.
This article discusses nine patients treated at our institution who underwent anterior skull base surgery for unilateral disorders with preservation of olfaction. A localized frontal subcranial approach minimized frontal lobe retraction and utilized the crista galli and perpendicular plate of the ethmoid as natural barriers to the spread of disease.
MATERIALS AND METHODS
Over the past 6 years nine patients have undergone anterior craniofacial surgery through a frontal subcranial approach to treat unilateral disease involving the anterior cranial fossa skull base. All surgeries were performed in the Department of Otolaryngology at the Wake Forest University Baptist Medical Center by the senior author (j.d.b.). The cases are summarized in Table I. The ages ranged from 16 to 79 years old, with a mean of 46 years. Eight were male; one was female. Six patients were diagnosed with tumors, two with encephaloceles, and one with an iatrogenic cerebral spinal fluid leak. Four of the tumors were malignancies: an adenocarcinoma of the ethmoid, a neuroendocrine malignancy of the cribriform area, an olfactory neuroblastoma, and a local recurrence of a previously resected ethmoid melanoma. Two of the tumors were benign: an inverting papilloma invading the left cribriform plate and a cosmetically deforming fibro-osseous tumor involving the cribriform plate. Histologically, the fibro-osseous tumor was within the spectrum of densely ossified fibrous dysplasia versus osteoma.
NED = no evidence of disease; CSF = cerebrospinal fluid; L = left; R = right; B = bilateral.
In all cases the disease was unilateral and each involved the cribriform plate. The medial bony orbit was involved in four cases and nasolacrimal reconstruction was required in four cases. Various other surgical procedures were performed in conjunction with the anterior craniofacial resection as necessitated by specific disease extent. Orbital reconstruction was required in one case. One patient underwent a medial maxillectomy through a mid-facial degloving approach and two via the coronal approach alone. Reconstruction of the cribriform defect included a pericranial flap in each case. Additionally, fascia from the temporalis or rectus abdominus muscle, titanium mesh, primary closure, thrombin-soaked collagen sponge, and fibrin glue were other adjuvants used to close the cribriform or any dural defect.
Postoperative testing for olfaction occurred no sooner than 4 weeks—after packing was removed and before any radiation therapy. Olfactory testing was both subjective and objective. In subjective testing patients were asked if their olfactory sense was preserved and whether they could tell a difference between sides. For objective testing the Pocket Smell Test (Sensonics, Haddon Heights, NJ) was used. Each test required correct identification of three odors (lemon, lilac, and smoke) in a multiple-choice format. Patients performed the test for each side of the nose while the contralateral nostril was occluded.
A bicoronal flap was raised from behind the hairline in the subgaleal plane forward to the superior orbital rims. Separate left and right pericranial flaps were elevated from the coronal incision to their respective supratrochlear vessels. The bicoronal flap was inverted and advanced to the medial canthi while the pericranial flaps (neurovascular bundles) were protected. A sterile template of the frontal sinus, cut from a 6-foot Caldwell plain X-ray, demonstrated the outline of the frontal sinus on the cranium. Titanium mini-plates were contoured for later repositioning of the osteoplastic flap and the holes were drilled in advance. The anterior wall of the frontal sinus was cut with a narrow side-cutting high-speed bur and elevated with an osteotome. The frontal sinus mucosa was completely removed with a blunt elevator and an otologic burr to prevent mucocele formation. The portion of the posterior wall of the frontal sinus removed with a high-speed drill depends on the exposure required. In small sinuses removal of all the posterior frontal sinus bony wall enables safe retraction of anterior cranial fossa dura. However, in larger sinuses, removal of half of the posterior wall usually provides adequate exposure.
Blunt elevation of the dura from the anterior-most edge of the ipsilateral cribriform plate exposes the first olfactory nerve filament. From anterior to posterior, the olfactory nerve filaments were serially divided after bipolar cautery and cut superior to their entrance into the cribriform plate. Depending on the size of the lesion, the dissection could extend posteriorly onto the planum sphenoidum. For malignancies, the crista galli and superior perpendicular plate of the ethmoid were always removed as the medial margin of resection, leaving the contralateral septal mucosa and olfactory nerve roots intact. Removal of the crista galli permits osteotomies just medial to the contralateral olfactory bulb and adjacent dura. Once the ipsilateral frontal dura was safely separated from the diseased area, the remainder of the resection was completed. In anterior nonmalignant disorders, only the anterior one half of the ipsilateral cribriform plate and respective olfactory nerves were removed.
For reconstruction the ipsilateral pericranial flap was spread over the resultant cribriform plate defect and positioned against the exposed dura. The closure was reinforced as needed with combinations of collagen sponge soaked with thrombin, as well as fibrin glue, fascia, or primary suture closure of any dural defect. The contralateral pericranial flap filled the frontal sinus space if needed. The frontal ducts were blocked with the pericranial flap or free muscle. Abdominal fat was added to fill any residual space. The anterior wall of the frontal sinus was replaced and plated into its original anatomic position. The bicoronal flap was returned and sutured with closed suction drains underneath.
A 54-year-old man presented with right-sided epistaxis and underwent endoscopic evaluation (case 3, Table I). Biopsy of a suspicious polyp revealed olfactory neuroblastoma. He was referred for further evaluation at the Department of Otolaryngology at the Wake Forest University Baptist Medical Center. Magnetic resonance imaging (Figs. 1 and 2) and computed tomography detailed a superior mass extending along the undersurface of the right nasal roof with bone expansion. The perpendicular plate of the ethmoid retained tumor to the right side. A frontal subcranial approach was used (Fig. 3) to remove the ipsilateral ethmoids, sphenoid, cribriform plate, crista galli, and superior bony septum with ipsilateral mucosa (Fig. 4), leaving the contralateral mucosa intact. The resultant cribriform defect was reconstructed with a pericranial flap and the frontal sinus osteoplastic flap was plated back into position (Fig. 5). Hospitalization lasted for 5 days for recovery and included administration of intravenous ampicillin and sulbactam. Nasal packs were removed on the fourth day with no cerebrospinal fluid leak. At 18 months the patient is asymptomatic and there has been no radiological evidence of disease (Fig. 6). When interviewed concerning his ability to smell, the patient stated, “normal—can't notice any difference from before surgery. ” Interestingly, despite complete resection of his ipsilateral olfactory nerves and cribriform plate, this patient could not readily distinguish a loss of smell from either side. On endoscopic examination he was found to have developed a small superior septal perforation just below the cribriform area, presumably allowing contralateral airflow.
Preservation of contralateral olfactory nerve roots was achieved in all nine cases and a functional sense of smell existed after surgery in all nine patients as determined subjectively by interview and objectively by testing. All patients identified the three correct smells on olfactory testing on the side of the preserved cribriform plate. In those patients with only a partial anterior resection of the ipsilateral cribriform plate, there was no loss of smell on the operated side and correct smells were identified in cases 5, 6, 7, and 8.
All patients were pleased with their sense of olfaction and considered it functionally equal to its preoperative state. Only one of the five patients who underwent complete unilateral cribriform plate resection could distinguish that the sense of smell was absent on one side before objective testing. As mentioned, one of these five patients could smell on the operated side despite complete resection of all ipsilateral olfactory nerves; postoperative endoscopic examination revealed a superior septal perforation that presumably allowed the sensation of bilateral stimuli by the nonresected side. Of the six neoplasms presented, no tumors have recurred, although two patients (cases 1 and 4) are at only their fifteenth postoperative month and one (case 9) is at his twelfth at the time of this report. Both encephaloceles and the cerebrospinal fluid leak were successfully sealed with no episodes of recurrence or meningitis. Follow-up ranges from 12 months to 5 years.
Surgery of the anterior skull base originated in the 1940s 1,2,8 and gained popularity in the 1950s and 1960s because of the work of surgeons such as Smith 9 and Ketcham. 3 Wide superior exposure of the anterior skull base was afforded through a frontal craniotomy and retraction of the frontal lobe. Raveh later described a subcranial approach through a frontal-naso-orbital osteotomy for repair of traumatic disruptions 4 and neoplasms. 5 This approach benefited those patients by reducing the frontal lobe retraction while avoiding facial incisions with a bicoronal flap. 4–6 In 1995 Raveh discussed a subgroup of his study in which contralateral olfaction was preserved. 6 In a series of a 104 neoplasms, he noted seven (5 malignant, 2 benign) unilateral tumors in which the contralateral olfactory filaments were preserved, the obvious advantage being conservation of the patient's sense of smell.
In our series nine patients with diverse lesions of the anterior skull base were treated through a localized subcranial approach with preservation of the contralateral olfactory filaments. In selected cases of benign disease it was possible to perform only a partial resection of the ipsilateral olfactory nerves without compromise of the surgical result. The common factor among all nine patients was that of preoperative imaging and symptoms demonstrating unilateral disease at the anterior skull base. In malignant disease, this functional modification cannot be employed if the disease process crosses the midline at the level of the anterior skull base. Fortunately, the perpendicular plate of the ethmoid and crista galli can provide a barrier to help confine disease to one side. In benign disease that may affect only the anterior portions of the cribriform plate on each side, preservation of smell should be possible with a bilateral anterior partial resection, judging by the success of such unilateral techniques in this series.
A large frontal sinus is not necessary to perform the dissection at the cribriform plate. In case 1, the frontal sinus was of small size and still provided adequate exposure of the cribriform plate. Although absence of a frontal sinus was not encountered in this series, osteotomies could create a similar frontal-glabellar osteoplastic flap. 5
In cases of malignancy the goals of surgery entail complete resection of the tumor with negative margins. 10–12 As described by others, involvement of the optic chiasm, clivus, sella turcica, and cavernous sinus remain as posterior limits in the resection of anterior skull base malignancies through this frontal avenue. 7 The guidelines for orbital preservation stated by Matthog can also be applied to the contralateral olfactory bundle: for elimination of disease, the surgeon should resect one anatomic border beyond the tumor. 13 As applied to the cribriform and ethmoid malignancy, the crista galli may be resected to place the medial osteotomy immediately adjacent to the contralateral olfactory groove. This allows for resection of the perpendicular plate of the ethmoid where adjacent to tumor, leaving the contralateral mucosa intact. The value of negative margins for anterior skull base malignancies was recently questioned by Shah, who showed identical recurrence rates (35%) for positive and negative margins. 7 However, we do not advocate preserving the contralateral olfactory bulb at the expense of positive margins. Additionally, the subcranial approach used in this series may be inadequate exposure for malignancies with intradural extension. Further, this anterior craniofacial resection can be combined with ablative procedures of the orbit and maxilla as necessary.
Reconstruction of the cribriform and dural defect consistently involved pericranial flaps. Primary dural closure, abdominal wall fat, titanium mesh, fascia, collagen sponge, and fibrin glue were also used. Pericranial flaps have received mixed support in the literature. Kraus wrote, “a pedicled pericranial flap was unsatisfactory and associated with an increased incidence of local complications.”14 However, others have strongly advocated pericranial flaps for small to moderate defects. 15,16 Our only complication of the series was one postoperative cerebrospinal fluid leak that resolved by successful endoscopic repair of a dural defect related to the pericranial flap failing to seat initially. We view this complication not as a deficiency of the pericranial flap but one of individual technique and continue to view the pericranial flap as a reliable tool in anterior skull base reconstruction.
Castalano reported a 63% morbidity rate for tumors of the paranasal sinus in 1994. 10 However, he noted that the complication rate decreased with time, from 52% during 1980 to 1987 to 28% after 1988. Five-year survival has been documented at 59% for nonesthesioblastoma malignancy and 90% for esthesioblastoma. 17 The low complication rate in our study (11%) may reflect a bias toward limited disease and inclusion of non-tumor cases.
Loss of olfaction as a result of cribriform plate resection is frequently necessary to eradicate disease in the anterior skull base. However, using a localized frontal subcranial approach for unilateral disease can allow for the preservation of the contralateral olfactory nerve bundle. Similarly, additional preservation of ipsilateral olfaction can be accomplished in benign disease when only a partial anterior resection of cribriform plate on the pathological side is adequate to eradicate disease. This functional modification of more extensive anterior craniofacial resections allows selected patients to maintain their sense of smell. Unilateral olfactory preservation is contraindicated in bilateral disease involving the whole of both cribriform plates, and in instances of malignancy in which resection of the contralateral olfactory apparatus is required to obtain clear surgical margins. A relative contraindication is intradural disease, with the feasibility of olfaction preservation being dependent on pathology and intraoperative assessment of the extent of disease. Nevertheless, the possibility of olfactory preservation should be considered when possible in patients with unilateral disease of the anterior skull base.
- 1.Orbital Tumors: Results Following the Transcranial Operative Attack. New York: Oskar Priest, 1941:1154..
- 2.Combined intracranial and orbital operation for retinoblastoma. Arch Ophthalmol 1943; 30:437–445., .
- 8.Ein Beitrag zur versorgung frontobasaler Hinverletzungen. Pract Otorhinolaryngol 1944; 6:326–336..