COMPLICATION
Brain Herniation After Endoscopic Transnasal Resection of Anterior Skull Base Malignancies BACKGROUND: Endoscopic endonasal approaches, when appropriate, allow a less invasive method to remove anterior skull base cancer than traditional external transfacial/craniofacial approaches. The resultant skull base defect can be significantly large, potentially extending from the posterior table of the frontal sinus to the tuberculum sellae in the sagittal plane, and from one lamina papyracea to the other in the coronal plane. However, frontal lobe herniation after such expanded endoscopic resection has been considered more of a theoretical than a practical occurrence. OBJECTIVE: To report the occurrence of frontal lobe herniation into the sinonasal cavity after expanded endonasal approaches, and to analyze causes and pathogenetic mechanisms of this unusual complication, proposing how it could have been prevented. METHODS: Two cases have been observed in 2 different skull base referral centers in the United States and Italy. Surgical and perioperative complications, postoperative course, and need for revisions were analyzed. RESULTS: Available data support the hypothesis that this complication is not attributable to the size of the anterior skull base defect, to the surgical technique, or to the materials used for the reconstruction. We found that 1 possible contributing factor may be the presence of increased intracranial pressure associated with obesity and obstructive sleep apnea, observed in both patients. CONCLUSION: Frontal lobe herniation must be considered as a possible, albeit rare, complication of expanded endoscopic anterior skull base resection. Preoperative investigations concerning the presence of obstructive sleep symptoms as well as proper identification of neuroimaging signs of intracranial hypertension are recommended for such cases.
Paolo Battaglia, MD* Mario Turri-Zanoni, MD* Paolo Castelnuovo, MD* Daniel M. Prevedello, MD‡ Ricardo L. Carrau, MD§ *Division of Otorhinolaryngology, Department of Biotechnology and Life Sciences, University of Insubria, Varese, Italy; ‡Department of Neurological Surgery, The Ohio State University, Columbus, Ohio; §Department of Otolaryngology, Head and Neck Surgery, The Ohio State University, Columbus, Ohio Correspondence: Mario Turri-Zanoni, MD, Division of Otorhinolaryngology, Department of Biotechnology and Life Sciences, University of Insubria, Varese, Italy, Via Guicciardini 9, Varese, Italy. E-mail: [email protected] Received, March 1, 2015. Accepted, May 21, 2015. Published Online, June 24, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
KEY WORDS: Brain herniation, Endoscopic endonasal, Intracranial hypertension, Sinonasal cancer, Skull base, Sleep apnea syndrome Operative Neurosurgery 11:457–462, 2015
D
uring the past two decades, significant advances in endoscopic skull base surgery have allowed the extirpation of tumors involving the anterior skull base by the use of a pure endonasal approach. This technique has shown promising oncologic outcomes that are comparable to those obtained with external approaches and that have been reproduced among various skull base centers worldwide. Therefore, these data suggest that endoscopic endonasal resection, when properly planned, and
ABBREVIATION: BMI, body mass index
OPERATIVE NEUROSURGERY
DOI: 10.1227/NEU.0000000000000859
performed by experienced surgeons, is an acceptable treatment for select skull base malignancies.1 Major advantages of endoscopic over external approaches include the absence of facial incisions or osteotomies, less brain manipulation, decreased hospitalization time, improved visualization of tumor borders, and reduced morbidity and mortality rates.2 Furthermore, retraction on the frontal lobes, with the ensuing associated complications, is avoided. Endoscopic endonasal techniques afford a less invasive approach to remove lesions encroaching the skull base than traditional external transfacial/ craniofacial approaches.3 However, the resultant skull base defect can be significant, potentially
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 457
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
extending from the posterior table of the frontal sinus to the tuberculum sellae in the sagittal plane, and from one lamina papyracea to the other in the coronal plane. Therefore, failure of transnasal reconstruction with a resultant postoperative cerebrospinal fluid (CSF) leak is considered a significant caveat. Moreover, the extent of the disease often requires resection of the dura mater, which further increases the likelihood of a postoperative CSF leak. In recent years, technical refinements in skull base reconstruction, including the use of a multilayer technique and the adoption of pedicled flaps, have significantly reduced the rate of postoperative CSF leaks.4 Frontal lobe sagging after reconstruction of these large anterior skull base defects has not been considered to be a significant complication. In fact, frontal lobe herniation, as a complication of an expanded endoscopic resection of the anterior skull base, has not been reported in the literature regardless of the technique and material used for the reconstruction.5 This report analyzes 2 unusual cases of frontal lobe herniation after endoscopic endonasal resection of a sinonasal malignancy. Its pathophysiology and management are discussed within the confines of our experience and a pertinent literature review.
METHODS We retrospectively reviewed the combined experience of more than 500 cases of endoscopic endonasal anterior skull base reconstruction after malignant tumor resection performed in 2 skull base referral centers in the United States and Italy from 1998 to 2014, reporting 2 cases of brain parenchyma herniation into the sinonasal cavity. Parameters including clinical data, preoperative imaging studies, surgical and perioperative complications, length of postoperative stay, and the need for surgical revisions were analyzed. The present study was conducted according to policies approved by the local institutional review boards. Informed consent was obtained from all patients involved in this survey.
RESULTS Patient 1 A 72-year-old man with an occupational history of woodworking was referred for an intestinal-type adenocarcinoma of the left ethmoid sinuses (Figures 1A and 1B). Clinical comorbidities included a history of a previous myocardial infarction treated with percutaneous transluminal coronary angioplasty (10 years before), arterial hypertension, obesity (body mass index [BMI], 36.2 kg/m2), and obstructive sleep apnea syndrome. The patient underwent an endoscopic resection with transnasal craniectomy, following a previously described surgical technique.6 Given the biology of this lesion, the entire ethmoid sinuses (bilateral) were included in the resection (ie, carcinogenic exposure to wood dust renders the mucosa of the nasoethmoidal complex vulnerable to developing multiple foci of adenocarcinoma).7 A nasal septectomy was required as part of the oncologic resection, therefore, precluding the use of a pedicled nasoseptal flap for the reconstruction. The resulting dural defect extended
458 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
from the posterior wall of the frontal sinus back to the planum sphenoidale and from orbit to orbit. It was repaired by using a multilayer closure technique comprising 3 layers of autologous iliotibial tract (first layer, subdural; second layer, intracranial epidural; third layer, extracranial). Finally, oxidized methylcellulose was placed around the edges of the extracranial layer to secure it to the adjacent bone, and the surgical cavity was packed with expandable sponge nasal packing (Merocel, Medtronic Xomed; Jacksonville, Florida) covered with bacitracin ointment. Postoperatively, the patient was kept on bed rest with the head of the bed elevated to 30°, under the coverage of a thirdgeneration cephalosporin and stool softeners. Two days later, the nasal packing was removed under endoscopic control and a scheduled brain computed tomography (CT) scan ruled out any intracranial complication (Figures 1C and 1D). At the fifth postoperative day, an endoscopic examination showed the presence of a CSF leak with brain exposure due to the displacement of the duraplasty. A head CT scan surprisingly revealed a herniation of the frontal lobe through the surgical anterior skull base defect (Figures 1E and 1F). Remarkably, the patient was asymptomatic, maintaining an alert and oriented mental state, without localizing neurological findings. This situation required an urgent endoscopic surgical revision of the duraplasty to prevent tension pneumocephalus and intracranial infections, such as ascending bacterial meningitis or abscess. Intraoperatively, the herniated brain parenchyma could not be reduced, because it extended well below the bony margins of the defect; therefore, the exposed brain was covered with fascia lata, and autologous free fat was grafted to eliminate the dead space and to flatten the residual denuded anterior skull base. A final overlay layer was made of a large iliotibial tract free graft, fixed with fibrin glue and buttressed with gelatin foam pledgets (Gelfoam; Pharmacia, Kalamazoo, Michigan). Moreover, a lumbar spinal drain was placed for 5 days to reduce intracranial pressure, therefore helping the healing and consolidation of the anterior skull base reconstruction. The patient was discharged home 10 days after the surgical revision with clear instructions to avoid nose blowing and any activity that could raise intracranial pressure (ie, straining). Considering the low grade (G1 adenocarcinoma) and limited stage (pT2, American Joint Committee on Cancer staging system, 7th edition, 2010) of the tumor on the final histological analysis and the radical resection with tumor-free margins, no adjuvant radiotherapy was recommended. One year after surgery, the patient continues to be completely asymptomatic, and a contrasted MR scan demonstrated a stable frontal lobe herniation (Figures 1G and 1H). Patient 2 A 64-year-old man was referred with a biopsy-proven esthesioneuroblastoma that had initially presented with left nasal bleeding and obstruction. The tumor extended bilaterally into the nasal cavities, ethmoid sinuses, and nasal septum, and invaded the left anterior cranial fossa (Figures 2A and 2B). Clinical comorbidities
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BRAIN HERNIATION AFTER EXPANDED ENDONASAL APPROACHES
FIGURE 1. Preoperative contrast-enhanced MRI in coronal (A) and sagittal (B) view, showing a left ethmoidal intestinal-type adenocarcinoma, involving the nasal septum and encroaching the left anterior skull base. Although an early postoperative CT scan looks normal (C, D), the fifth postoperative day CT scan reveals a herniation of the frontal lobe through the surgical anterior skull base defect (E, F). The postoperative MRI performed 1 year after the surgical revision demonstrated a stable frontal lobe herniation (G, H).
OPERATIVE NEUROSURGERY
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 459
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
FIGURE 2. Pre-operative contrast-enhanced coronal MRI (A) and CT scan (B) showing a left esthesioneuroblastoma (T), involving the nasal septum, right ethmoid sinuses and anterior cranial fossa (white arrow in A). Postoperative coronal MRI (C) and CT scan (D) within 24 hours of the resection reveal an adequate resection and no brain sagging (white arrows). Subsequent coronal MRI (E) and CT scan (F) show significant sagging of the frontal lobe, marked by the white arrows (B, brain). Radionecrosis of the reconstructive flap led to significant brain herniation (BH) demonstrated By sagittal (G) and coronal (H) MRI views (white arrows indicate the brain herniation). A Lactasorb plate and fascia lata were required to maintain the reduction of the brain herniation as shown by sagittal (I) and coronal (J) MRI views (white arrows point out the successful skull base reconstruction). K and L demonstrate progressive parenchymal radionecrosis (white arrows). T, tumor; FB, Foley Balloon catheter; B, brain; BH, brain herniation.
included a history of arterial hypertension, obesity (BMI, 38.8 kg/m2), and obstructive sleep apnea syndrome. The patient underwent an endoscopic resection with transnasal craniectomy, following a previously described technique.8 Given the biology and extent of this lesion, the entire ethmoid labyrinth (bilaterally), lamina papyraceae, septum, dura, and olfactory bulbs were included in the resection. Sacrifice of the nasal septum precluded the use of a pedicled nasoseptal flap for the skull base reconstruction. Therefore, the resulting skull base and dural defect, which extended from the posterior wall of the frontal sinus to the planum sphenoidale and from the orbit to orbit, was repaired by using a multilayer closure technique comprising a subdural layer of collagen matrix and an extracranial (onlay) pericranial flap. The latter was harvested via a coronal incision and inserted into the nasal cavity via a bone window through the nasion (transfrontal technique). The flap was bolstered against
460 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
the defect by using Nasopore (Stryker Corporation; Kalamazoo, Michigan) and expandable sponge nasal packing (Merocel, Medtronic, Jacksonville, Florida). An immediate postoperative head CT scan ruled out any intracranial complication. Magnetic resonance imaging (MRI) confirmed the adequacy of the resection, good enhancement of the flap following contrast, and minimal sagging of the brain (Figures 2C and 2D). Postoperatively, the patient continued receiving prophylactic third-generation cephalosporin for 48 hours and stool softeners. The patient was discharged home from the hospital 3 days after surgery with instructions to avoid nose blowing and any activity that could raise intracranial pressure, such as straining. At the sixth postoperative day, the patient was evaluated as an outpatient, the sponge packing was removed and the lower nasal cavity was gently debrided. A slight sagging of the flap was noted, but it looked viable and in good contact.
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BRAIN HERNIATION AFTER EXPANDED ENDONASAL APPROACHES
Considering the extent of the tumor (pT3 University of California, Los Angeles staging system9 or Kadish10 stage C) and Hyams grade II histopathology, adjuvant proton radiotherapy was recommended. A contrast-enhanced MRI performed during the proton therapy planning revealed further herniation of the brain (Figures 2E and 2F). Three months after completion of the proton therapy, the patient developed changes in personality and shortterm memory loss. MRI revealed severe radiation edema of the frontal lobes. Despite corticosteroids and hyperbaric oxygen, the edema and mental changes persisted, although somewhat improved. Treatment with bevacizumab (Avastin; Genentech/ Roche, San Francisco, California) resulted in a complete resolution of the symptoms and cerebral edema (MRI proven). Seven months after completion of the proton therapy, the patient developed clear rhinorrhea followed by changes in mentation requiring emergent hospitalization. He was diagnosed with bacterial meningitis associated with exposed brain herniating into the nasal cavity (Figures 2G and 2H). This required a surgical repair that included debridement of the nasal cavity and endonasal repair using 2-layered epidural fascia lata graft. Nasal examination during that surgery revealed that the pericranial flap, mucosa, and dura were completely necrotic, and the brain had herniated though the original skull base defect. Extensive radionecrosis was also noted over the periorbital and sphenoid and maxillary sinuses. This attempt failed to repair the CSF fistula (the brain herniated into the nasal cavity); thus, it was followed by a cranioendoscopic repair using a 2-layer epidural fascia lata graft resulting in a similar outcome: the brain herniated into the nasal cavity again. Finally, using a cranioendoscopic approach, we implanted an absorbable Lorenz Lactosorb mesh (Biomet Corporation; Jacksonville, Florida) enveloped in fascia lata to cover the defect (subdural), and, thus, stop the brain herniation (Figures 2I and 2J). An extracranial lateral thigh myofascial microvascular flap was subsequently transferred to reinforce the reconstruction after the mesh was noted to be sagging. Furthermore, an adjunctive external ventricular device was placed to decrease the intracranial pressure. Unfortunately, the patient developed progressive radionecrosis of the frontal lobe and deterioration of neurological function (Figures 2K and 2L). Following an intensive rehabilitation program, repeated courses of corticosteroids and bevacizumab (Avastin; Genentech/Roche, San Francisco, California), he finally started improving. Examination of the proton planning revealed that the patient had received unintended radiation to the frontal lobes due to the herniation of brain through the defect.
DISCUSSION Frontal lobe herniation through the anterior skull base defect has been considered more of a theoretical than practical consideration, as confirmed by the fact that we found no other reports in the English literature. In our combined experience of more than 500 cases of endoscopic endonasal anterior skull base reconstruction after malignant tumor resection, these are the first 2 cases of brain parenchyma herniation. To the best of our knowledge, no similar cases have been
OPERATIVE NEUROSURGERY
described as yet in the English literature. We routinely perform skull base reconstruction according to the site, extension, and the biology of the lesion, using a multilayer technique with reliable success rates. A retrospective report by Eloy and colleagues5 showed no significant frontal lobe descent after endoscopic repair of large cribriform defects utilizing a triple-layer reconstruction technique with fascia lata, acellular dermal allograft, and pedicled nasoseptal flap. This argues against the need of a rigid structural graft such as cartilage, bone, or alloplastic materials (eg, titanium plate) to reinforce the repair of the anterior skull base. Furthermore, available data support the hypothesis that frontal lobe herniation into the sinonasal cavities is not attributable to the size of the anterior skull base defect, to surgical technique, or to the materials used for the reconstruction. We believe that a factor that could be involved in the pathogenesis of this complication may be the presence of increased intracranial pressure associated with obesity and obstructive sleep apnea. Preoperative and postoperative neuroimaging of these patients revealed no evidence of abnormalities of the ventricular system, and neurodiagnostic studies were otherwise normal, although the preoperative MRI of patient 1 showed an empty sella, lateral sinus collapse (smoothwalled venous stenoses), and buckling of the optic nerves with increased perineural fluid. All these are specific signs included in the modified Dandy criteria for idiopathic intracranial hypertension.11 Moreover, no other secondary cause for intracranial hypertension could be found in both patients herein reported. Obesity and sleep apnea syndrome may have played a role in the pathogenesis of the brain herniation. Several studies have documented increased intracranial pressure during apnea periods.12 Because this is a plausible mechanism to explain increased intracranial pressure, BMI assessment and preoperative sleep studies in patients with symptoms suggestive of obstructive sleep apnea may be prudent. When intracranial hypertension is suspected to be involved in the pathogenesis, lumbar spinal puncture and drainage could be useful to both document and reduce intracranial pressure. One should note, however, that CSF pressure fluctuates throughout the day and, even in pathological states, at times could be normal; thus, a single normal CSF measurement does not exclude intracranial hypertension definitively. From a clinical viewpoint, in high-risk patients, weight loss, sodium restriction often with the addition of acetazolamide, must be considered to minimize the long-term failure rates of an anterior skull base reconstruction or progressive herniation. We acknowledge that this strategy is empirical and may elicit controversy. From a technical viewpoint, other reasonable considerations for the management of this clinical situation include an external transcranial approach to reduce the herniated frontal lobe, the surgical resection of the herniated brain parenchyma, or permanent CSF diversion. However, the most important consideration is to achieve separation between the herniated frontal lobes and the sinonasal tract to avoid ascending infections and their complications. Progressive brain herniation and symptomatic manifestations would require more assertive surgical maneuvers such as the ones previously mentioned. In such challenging cases, the use of rigid structural grafts to reinforce
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 461
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
the skull base reconstruction is controversial because of the increased risk of chronic inflammation, infection, radionecrosis, and/or extrusion of both autologous (cartilage or bone) and heterologous materials.13 In light of these risks, the use of rigid grafts has been avoided even for the reconstruction of high-flow compartments such as the posterior cranial fossa after transclival transdural endoscopic approaches, where the use of fat graft combined with the vascularized nasoseptal flap is generally adopted to minimize the risk of pontine herniation.14 Another critical issue is the planning of adjuvant radiotherapy in case of progressive brain herniation that complicates obtaining an accurate definition of radiation doses and volumes on the frontal lobes. For this reason, early recognition of this complication by repeating appropriate imaging (CT and/or MR scan) is paramount, in order to balance the oncologic purpose of cancer treatment with the increased risks of radiation-induced brain edema and necrosis. To this effect, a multidisciplinary team approach with strict cooperation between surgeons and radiation oncologists is mandatory for a comprehensive management of such challenging cases.
CONCLUSION Frontal lobe herniation must be considered as a possible, albeit rare, complication of expanded endoscopic anterior skull base resection. Preoperative investigations concerning the presence of obstructive sleep symptoms as well as proper identification of neuroimaging signs of intracranial hypertension are recommended to ascertain the risk of this rare but important complication. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Castelnuovo P, Turri-Zanoni M, Battaglia P, Bignami M, Bolzoni Villaret A, Nicolai P. Endoscopic endonasal approaches for malignant tumours involving the skull base. Curr Otorhinolaryngol Rep. 2013;1(4):197-205. 2. Batra PS. Minimally invasive endoscopic resection of sinonasal and anterior skull base malignant neoplasms. Expert Rev Med Devices. 2010;7(6):781-791. 3. Castelnuovo P, Lepera D, Turri-Zanoni M, et al. Quality of life following endoscopic endonasal resection of anterior skull base cancers. J Neurosurg. 2013; 119(6):1401-1409.
462 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
4. Kassam AB, Thomas A, Carrau RL, et al. Endoscopic reconstruction of the cranial base using a pedicled nasoseptal flap. Neurosurgery. 2008;63(1 suppl 1):ONS44ONS52. 5. Eloy JA, Shukla PA, Choudhry OJ, Singh R, Liu JK. Assessment of frontal lobe sagging after endoscopic endonasal transcribriform resection of anterior skull base tumors: is rigid structural reconstruction of the cranial base defect necessary? Laryngoscope. 2012;122(12):2652-2657. 6. Castelnuovo P, Battaglia P, Turri-Zanoni M, et al. Endoscopic endonasal surgery for malignancies of the anterior cranial base. World Neurosurg. 2014;82(6 suppl): S22-S31. 7. Antognoni P, Turri-Zanoni M, Gottardo S, et al. Endoscopic resection followed by adjuvant radiotherapy for sinonasal intestinal-type adenocarcinoma: retrospective analysis of 30 consecutive patients. Head Neck. 2015;37(5): 677-684. 8. Kasemsiri P, Prevedello DM, Otto BA, et al. Endoscopic endonasal technique: treatment of paranasal and anterior skull base malignancies [in English, Portuguese]. Braz J Otorhinolaryngol. 2013;79(6):760-779. 9. Dulguerov P, Calcaterra T. Esthesioneuroblastoma: the UCLA experience 19701990. Laryngoscope. 1992;102(8):843-849. 10. Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma. A clinical analysis of 17 cases. Cancer. 1976;37(3):1571-1576. 11. Wall M. Idiopathic intracranial hypertension. Neurol Clin. 2010;28(3): 593-617. 12. Jennum P, Børgesen SE. Intracranial pressure and obstructive sleep apnea. Chest. 1989;95(2):279-283. 13. Gil Z, Abergel A, Leider-Trejo L, et al. A comprehensive algorithm for anterior skull base reconstruction after oncological resections. Skull Base. 2007;17(1): 25-37. 14. Koutourousiou M, Filho FV, Costacou T, et al. Pontine encephalocele and abnormalities of the posterior fossa following transclival endoscopic endonasal surgery. J Neurosurg. 2014;121(2):359-366.
COMMENT
T
he authors present 2 cases of brain herniation through the frontal fossa defect after extended endonasal transsphenoidal surgery for skull base tumors. This rare, delayed complication of the extended endonasal approach should be recognized, particularly when rigid buttresses cannot be used to augment the closure and in patients with signs or risk factors of elevated intracranial pressure (sleep apnea, obesity, Chiari I malformation). When unrecognized, as seen in the second patient, the radiation therapy planning can inadvertently include the encephalocoele in the center of the radiation field. This may explain the severe radiation necrosis that this patient experienced. This supports the need for an open dialogue between the skull base surgeons and the radiation oncologists. We commend the authors for describing this rare complication of extended endonasal surgery. Garni Barkhoudarian Santa Monica, California
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
Brain Herniation After Endoscopic Transnasal Resection of Anterior Skull Base Malignancies BACKGROUND: Endoscopic endonasal approaches, when appropriate, allow a less invasive method to remove anterior skull base cancer than traditional external transfacial/craniofacial approaches. The resultant skull base defect can be significantly large, potentially extending from the posterior table of the frontal sinus to the tuberculum sellae in the sagittal plane, and from one lamina papyracea to the other in the coronal plane. However, frontal lobe herniation after such expanded endoscopic resection has been considered more of a theoretical than a practical occurrence. OBJECTIVE: To report the occurrence of frontal lobe herniation into the sinonasal cavity after expanded endonasal approaches, and to analyze causes and pathogenetic mechanisms of this unusual complication, proposing how it could have been prevented. METHODS: Two cases have been observed in 2 different skull base referral centers in the United States and Italy. Surgical and perioperative complications, postoperative course, and need for revisions were analyzed. RESULTS: Available data support the hypothesis that this complication is not attributable to the size of the anterior skull base defect, to the surgical technique, or to the materials used for the reconstruction. We found that 1 possible contributing factor may be the presence of increased intracranial pressure associated with obesity and obstructive sleep apnea, observed in both patients. CONCLUSION: Frontal lobe herniation must be considered as a possible, albeit rare, complication of expanded endoscopic anterior skull base resection. Preoperative investigations concerning the presence of obstructive sleep symptoms as well as proper identification of neuroimaging signs of intracranial hypertension are recommended for such cases.
Paolo Battaglia, MD* Mario Turri-Zanoni, MD* Paolo Castelnuovo, MD* Daniel M. Prevedello, MD‡ Ricardo L. Carrau, MD§ *Division of Otorhinolaryngology, Department of Biotechnology and Life Sciences, University of Insubria, Varese, Italy; ‡Department of Neurological Surgery, The Ohio State University, Columbus, Ohio; §Department of Otolaryngology, Head and Neck Surgery, The Ohio State University, Columbus, Ohio Correspondence: Mario Turri-Zanoni, MD, Division of Otorhinolaryngology, Department of Biotechnology and Life Sciences, University of Insubria, Varese, Italy, Via Guicciardini 9, Varese, Italy. E-mail: [email protected] Received, March 1, 2015. Accepted, May 21, 2015. Published Online, June 24, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
KEY WORDS: Brain herniation, Endoscopic endonasal, Intracranial hypertension, Sinonasal cancer, Skull base, Sleep apnea syndrome Operative Neurosurgery 11:457–462, 2015
D
uring the past two decades, significant advances in endoscopic skull base surgery have allowed the extirpation of tumors involving the anterior skull base by the use of a pure endonasal approach. This technique has shown promising oncologic outcomes that are comparable to those obtained with external approaches and that have been reproduced among various skull base centers worldwide. Therefore, these data suggest that endoscopic endonasal resection, when properly planned, and
ABBREVIATION: BMI, body mass index
OPERATIVE NEUROSURGERY
DOI: 10.1227/NEU.0000000000000859
performed by experienced surgeons, is an acceptable treatment for select skull base malignancies.1 Major advantages of endoscopic over external approaches include the absence of facial incisions or osteotomies, less brain manipulation, decreased hospitalization time, improved visualization of tumor borders, and reduced morbidity and mortality rates.2 Furthermore, retraction on the frontal lobes, with the ensuing associated complications, is avoided. Endoscopic endonasal techniques afford a less invasive approach to remove lesions encroaching the skull base than traditional external transfacial/ craniofacial approaches.3 However, the resultant skull base defect can be significant, potentially
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 457
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
extending from the posterior table of the frontal sinus to the tuberculum sellae in the sagittal plane, and from one lamina papyracea to the other in the coronal plane. Therefore, failure of transnasal reconstruction with a resultant postoperative cerebrospinal fluid (CSF) leak is considered a significant caveat. Moreover, the extent of the disease often requires resection of the dura mater, which further increases the likelihood of a postoperative CSF leak. In recent years, technical refinements in skull base reconstruction, including the use of a multilayer technique and the adoption of pedicled flaps, have significantly reduced the rate of postoperative CSF leaks.4 Frontal lobe sagging after reconstruction of these large anterior skull base defects has not been considered to be a significant complication. In fact, frontal lobe herniation, as a complication of an expanded endoscopic resection of the anterior skull base, has not been reported in the literature regardless of the technique and material used for the reconstruction.5 This report analyzes 2 unusual cases of frontal lobe herniation after endoscopic endonasal resection of a sinonasal malignancy. Its pathophysiology and management are discussed within the confines of our experience and a pertinent literature review.
METHODS We retrospectively reviewed the combined experience of more than 500 cases of endoscopic endonasal anterior skull base reconstruction after malignant tumor resection performed in 2 skull base referral centers in the United States and Italy from 1998 to 2014, reporting 2 cases of brain parenchyma herniation into the sinonasal cavity. Parameters including clinical data, preoperative imaging studies, surgical and perioperative complications, length of postoperative stay, and the need for surgical revisions were analyzed. The present study was conducted according to policies approved by the local institutional review boards. Informed consent was obtained from all patients involved in this survey.
RESULTS Patient 1 A 72-year-old man with an occupational history of woodworking was referred for an intestinal-type adenocarcinoma of the left ethmoid sinuses (Figures 1A and 1B). Clinical comorbidities included a history of a previous myocardial infarction treated with percutaneous transluminal coronary angioplasty (10 years before), arterial hypertension, obesity (body mass index [BMI], 36.2 kg/m2), and obstructive sleep apnea syndrome. The patient underwent an endoscopic resection with transnasal craniectomy, following a previously described surgical technique.6 Given the biology of this lesion, the entire ethmoid sinuses (bilateral) were included in the resection (ie, carcinogenic exposure to wood dust renders the mucosa of the nasoethmoidal complex vulnerable to developing multiple foci of adenocarcinoma).7 A nasal septectomy was required as part of the oncologic resection, therefore, precluding the use of a pedicled nasoseptal flap for the reconstruction. The resulting dural defect extended
458 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
from the posterior wall of the frontal sinus back to the planum sphenoidale and from orbit to orbit. It was repaired by using a multilayer closure technique comprising 3 layers of autologous iliotibial tract (first layer, subdural; second layer, intracranial epidural; third layer, extracranial). Finally, oxidized methylcellulose was placed around the edges of the extracranial layer to secure it to the adjacent bone, and the surgical cavity was packed with expandable sponge nasal packing (Merocel, Medtronic Xomed; Jacksonville, Florida) covered with bacitracin ointment. Postoperatively, the patient was kept on bed rest with the head of the bed elevated to 30°, under the coverage of a thirdgeneration cephalosporin and stool softeners. Two days later, the nasal packing was removed under endoscopic control and a scheduled brain computed tomography (CT) scan ruled out any intracranial complication (Figures 1C and 1D). At the fifth postoperative day, an endoscopic examination showed the presence of a CSF leak with brain exposure due to the displacement of the duraplasty. A head CT scan surprisingly revealed a herniation of the frontal lobe through the surgical anterior skull base defect (Figures 1E and 1F). Remarkably, the patient was asymptomatic, maintaining an alert and oriented mental state, without localizing neurological findings. This situation required an urgent endoscopic surgical revision of the duraplasty to prevent tension pneumocephalus and intracranial infections, such as ascending bacterial meningitis or abscess. Intraoperatively, the herniated brain parenchyma could not be reduced, because it extended well below the bony margins of the defect; therefore, the exposed brain was covered with fascia lata, and autologous free fat was grafted to eliminate the dead space and to flatten the residual denuded anterior skull base. A final overlay layer was made of a large iliotibial tract free graft, fixed with fibrin glue and buttressed with gelatin foam pledgets (Gelfoam; Pharmacia, Kalamazoo, Michigan). Moreover, a lumbar spinal drain was placed for 5 days to reduce intracranial pressure, therefore helping the healing and consolidation of the anterior skull base reconstruction. The patient was discharged home 10 days after the surgical revision with clear instructions to avoid nose blowing and any activity that could raise intracranial pressure (ie, straining). Considering the low grade (G1 adenocarcinoma) and limited stage (pT2, American Joint Committee on Cancer staging system, 7th edition, 2010) of the tumor on the final histological analysis and the radical resection with tumor-free margins, no adjuvant radiotherapy was recommended. One year after surgery, the patient continues to be completely asymptomatic, and a contrasted MR scan demonstrated a stable frontal lobe herniation (Figures 1G and 1H). Patient 2 A 64-year-old man was referred with a biopsy-proven esthesioneuroblastoma that had initially presented with left nasal bleeding and obstruction. The tumor extended bilaterally into the nasal cavities, ethmoid sinuses, and nasal septum, and invaded the left anterior cranial fossa (Figures 2A and 2B). Clinical comorbidities
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BRAIN HERNIATION AFTER EXPANDED ENDONASAL APPROACHES
FIGURE 1. Preoperative contrast-enhanced MRI in coronal (A) and sagittal (B) view, showing a left ethmoidal intestinal-type adenocarcinoma, involving the nasal septum and encroaching the left anterior skull base. Although an early postoperative CT scan looks normal (C, D), the fifth postoperative day CT scan reveals a herniation of the frontal lobe through the surgical anterior skull base defect (E, F). The postoperative MRI performed 1 year after the surgical revision demonstrated a stable frontal lobe herniation (G, H).
OPERATIVE NEUROSURGERY
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 459
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
FIGURE 2. Pre-operative contrast-enhanced coronal MRI (A) and CT scan (B) showing a left esthesioneuroblastoma (T), involving the nasal septum, right ethmoid sinuses and anterior cranial fossa (white arrow in A). Postoperative coronal MRI (C) and CT scan (D) within 24 hours of the resection reveal an adequate resection and no brain sagging (white arrows). Subsequent coronal MRI (E) and CT scan (F) show significant sagging of the frontal lobe, marked by the white arrows (B, brain). Radionecrosis of the reconstructive flap led to significant brain herniation (BH) demonstrated By sagittal (G) and coronal (H) MRI views (white arrows indicate the brain herniation). A Lactasorb plate and fascia lata were required to maintain the reduction of the brain herniation as shown by sagittal (I) and coronal (J) MRI views (white arrows point out the successful skull base reconstruction). K and L demonstrate progressive parenchymal radionecrosis (white arrows). T, tumor; FB, Foley Balloon catheter; B, brain; BH, brain herniation.
included a history of arterial hypertension, obesity (BMI, 38.8 kg/m2), and obstructive sleep apnea syndrome. The patient underwent an endoscopic resection with transnasal craniectomy, following a previously described technique.8 Given the biology and extent of this lesion, the entire ethmoid labyrinth (bilaterally), lamina papyraceae, septum, dura, and olfactory bulbs were included in the resection. Sacrifice of the nasal septum precluded the use of a pedicled nasoseptal flap for the skull base reconstruction. Therefore, the resulting skull base and dural defect, which extended from the posterior wall of the frontal sinus to the planum sphenoidale and from the orbit to orbit, was repaired by using a multilayer closure technique comprising a subdural layer of collagen matrix and an extracranial (onlay) pericranial flap. The latter was harvested via a coronal incision and inserted into the nasal cavity via a bone window through the nasion (transfrontal technique). The flap was bolstered against
460 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
the defect by using Nasopore (Stryker Corporation; Kalamazoo, Michigan) and expandable sponge nasal packing (Merocel, Medtronic, Jacksonville, Florida). An immediate postoperative head CT scan ruled out any intracranial complication. Magnetic resonance imaging (MRI) confirmed the adequacy of the resection, good enhancement of the flap following contrast, and minimal sagging of the brain (Figures 2C and 2D). Postoperatively, the patient continued receiving prophylactic third-generation cephalosporin for 48 hours and stool softeners. The patient was discharged home from the hospital 3 days after surgery with instructions to avoid nose blowing and any activity that could raise intracranial pressure, such as straining. At the sixth postoperative day, the patient was evaluated as an outpatient, the sponge packing was removed and the lower nasal cavity was gently debrided. A slight sagging of the flap was noted, but it looked viable and in good contact.
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BRAIN HERNIATION AFTER EXPANDED ENDONASAL APPROACHES
Considering the extent of the tumor (pT3 University of California, Los Angeles staging system9 or Kadish10 stage C) and Hyams grade II histopathology, adjuvant proton radiotherapy was recommended. A contrast-enhanced MRI performed during the proton therapy planning revealed further herniation of the brain (Figures 2E and 2F). Three months after completion of the proton therapy, the patient developed changes in personality and shortterm memory loss. MRI revealed severe radiation edema of the frontal lobes. Despite corticosteroids and hyperbaric oxygen, the edema and mental changes persisted, although somewhat improved. Treatment with bevacizumab (Avastin; Genentech/ Roche, San Francisco, California) resulted in a complete resolution of the symptoms and cerebral edema (MRI proven). Seven months after completion of the proton therapy, the patient developed clear rhinorrhea followed by changes in mentation requiring emergent hospitalization. He was diagnosed with bacterial meningitis associated with exposed brain herniating into the nasal cavity (Figures 2G and 2H). This required a surgical repair that included debridement of the nasal cavity and endonasal repair using 2-layered epidural fascia lata graft. Nasal examination during that surgery revealed that the pericranial flap, mucosa, and dura were completely necrotic, and the brain had herniated though the original skull base defect. Extensive radionecrosis was also noted over the periorbital and sphenoid and maxillary sinuses. This attempt failed to repair the CSF fistula (the brain herniated into the nasal cavity); thus, it was followed by a cranioendoscopic repair using a 2-layer epidural fascia lata graft resulting in a similar outcome: the brain herniated into the nasal cavity again. Finally, using a cranioendoscopic approach, we implanted an absorbable Lorenz Lactosorb mesh (Biomet Corporation; Jacksonville, Florida) enveloped in fascia lata to cover the defect (subdural), and, thus, stop the brain herniation (Figures 2I and 2J). An extracranial lateral thigh myofascial microvascular flap was subsequently transferred to reinforce the reconstruction after the mesh was noted to be sagging. Furthermore, an adjunctive external ventricular device was placed to decrease the intracranial pressure. Unfortunately, the patient developed progressive radionecrosis of the frontal lobe and deterioration of neurological function (Figures 2K and 2L). Following an intensive rehabilitation program, repeated courses of corticosteroids and bevacizumab (Avastin; Genentech/Roche, San Francisco, California), he finally started improving. Examination of the proton planning revealed that the patient had received unintended radiation to the frontal lobes due to the herniation of brain through the defect.
DISCUSSION Frontal lobe herniation through the anterior skull base defect has been considered more of a theoretical than practical consideration, as confirmed by the fact that we found no other reports in the English literature. In our combined experience of more than 500 cases of endoscopic endonasal anterior skull base reconstruction after malignant tumor resection, these are the first 2 cases of brain parenchyma herniation. To the best of our knowledge, no similar cases have been
OPERATIVE NEUROSURGERY
described as yet in the English literature. We routinely perform skull base reconstruction according to the site, extension, and the biology of the lesion, using a multilayer technique with reliable success rates. A retrospective report by Eloy and colleagues5 showed no significant frontal lobe descent after endoscopic repair of large cribriform defects utilizing a triple-layer reconstruction technique with fascia lata, acellular dermal allograft, and pedicled nasoseptal flap. This argues against the need of a rigid structural graft such as cartilage, bone, or alloplastic materials (eg, titanium plate) to reinforce the repair of the anterior skull base. Furthermore, available data support the hypothesis that frontal lobe herniation into the sinonasal cavities is not attributable to the size of the anterior skull base defect, to surgical technique, or to the materials used for the reconstruction. We believe that a factor that could be involved in the pathogenesis of this complication may be the presence of increased intracranial pressure associated with obesity and obstructive sleep apnea. Preoperative and postoperative neuroimaging of these patients revealed no evidence of abnormalities of the ventricular system, and neurodiagnostic studies were otherwise normal, although the preoperative MRI of patient 1 showed an empty sella, lateral sinus collapse (smoothwalled venous stenoses), and buckling of the optic nerves with increased perineural fluid. All these are specific signs included in the modified Dandy criteria for idiopathic intracranial hypertension.11 Moreover, no other secondary cause for intracranial hypertension could be found in both patients herein reported. Obesity and sleep apnea syndrome may have played a role in the pathogenesis of the brain herniation. Several studies have documented increased intracranial pressure during apnea periods.12 Because this is a plausible mechanism to explain increased intracranial pressure, BMI assessment and preoperative sleep studies in patients with symptoms suggestive of obstructive sleep apnea may be prudent. When intracranial hypertension is suspected to be involved in the pathogenesis, lumbar spinal puncture and drainage could be useful to both document and reduce intracranial pressure. One should note, however, that CSF pressure fluctuates throughout the day and, even in pathological states, at times could be normal; thus, a single normal CSF measurement does not exclude intracranial hypertension definitively. From a clinical viewpoint, in high-risk patients, weight loss, sodium restriction often with the addition of acetazolamide, must be considered to minimize the long-term failure rates of an anterior skull base reconstruction or progressive herniation. We acknowledge that this strategy is empirical and may elicit controversy. From a technical viewpoint, other reasonable considerations for the management of this clinical situation include an external transcranial approach to reduce the herniated frontal lobe, the surgical resection of the herniated brain parenchyma, or permanent CSF diversion. However, the most important consideration is to achieve separation between the herniated frontal lobes and the sinonasal tract to avoid ascending infections and their complications. Progressive brain herniation and symptomatic manifestations would require more assertive surgical maneuvers such as the ones previously mentioned. In such challenging cases, the use of rigid structural grafts to reinforce
VOLUME 11 | NUMBER 3 | SEPTEMBER 2015 | 461
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
BATTAGLIA ET AL
the skull base reconstruction is controversial because of the increased risk of chronic inflammation, infection, radionecrosis, and/or extrusion of both autologous (cartilage or bone) and heterologous materials.13 In light of these risks, the use of rigid grafts has been avoided even for the reconstruction of high-flow compartments such as the posterior cranial fossa after transclival transdural endoscopic approaches, where the use of fat graft combined with the vascularized nasoseptal flap is generally adopted to minimize the risk of pontine herniation.14 Another critical issue is the planning of adjuvant radiotherapy in case of progressive brain herniation that complicates obtaining an accurate definition of radiation doses and volumes on the frontal lobes. For this reason, early recognition of this complication by repeating appropriate imaging (CT and/or MR scan) is paramount, in order to balance the oncologic purpose of cancer treatment with the increased risks of radiation-induced brain edema and necrosis. To this effect, a multidisciplinary team approach with strict cooperation between surgeons and radiation oncologists is mandatory for a comprehensive management of such challenging cases.
CONCLUSION Frontal lobe herniation must be considered as a possible, albeit rare, complication of expanded endoscopic anterior skull base resection. Preoperative investigations concerning the presence of obstructive sleep symptoms as well as proper identification of neuroimaging signs of intracranial hypertension are recommended to ascertain the risk of this rare but important complication. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Castelnuovo P, Turri-Zanoni M, Battaglia P, Bignami M, Bolzoni Villaret A, Nicolai P. Endoscopic endonasal approaches for malignant tumours involving the skull base. Curr Otorhinolaryngol Rep. 2013;1(4):197-205. 2. Batra PS. Minimally invasive endoscopic resection of sinonasal and anterior skull base malignant neoplasms. Expert Rev Med Devices. 2010;7(6):781-791. 3. Castelnuovo P, Lepera D, Turri-Zanoni M, et al. Quality of life following endoscopic endonasal resection of anterior skull base cancers. J Neurosurg. 2013; 119(6):1401-1409.
462 | VOLUME 11 | NUMBER 3 | SEPTEMBER 2015
4. Kassam AB, Thomas A, Carrau RL, et al. Endoscopic reconstruction of the cranial base using a pedicled nasoseptal flap. Neurosurgery. 2008;63(1 suppl 1):ONS44ONS52. 5. Eloy JA, Shukla PA, Choudhry OJ, Singh R, Liu JK. Assessment of frontal lobe sagging after endoscopic endonasal transcribriform resection of anterior skull base tumors: is rigid structural reconstruction of the cranial base defect necessary? Laryngoscope. 2012;122(12):2652-2657. 6. Castelnuovo P, Battaglia P, Turri-Zanoni M, et al. Endoscopic endonasal surgery for malignancies of the anterior cranial base. World Neurosurg. 2014;82(6 suppl): S22-S31. 7. Antognoni P, Turri-Zanoni M, Gottardo S, et al. Endoscopic resection followed by adjuvant radiotherapy for sinonasal intestinal-type adenocarcinoma: retrospective analysis of 30 consecutive patients. Head Neck. 2015;37(5): 677-684. 8. Kasemsiri P, Prevedello DM, Otto BA, et al. Endoscopic endonasal technique: treatment of paranasal and anterior skull base malignancies [in English, Portuguese]. Braz J Otorhinolaryngol. 2013;79(6):760-779. 9. Dulguerov P, Calcaterra T. Esthesioneuroblastoma: the UCLA experience 19701990. Laryngoscope. 1992;102(8):843-849. 10. Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma. A clinical analysis of 17 cases. Cancer. 1976;37(3):1571-1576. 11. Wall M. Idiopathic intracranial hypertension. Neurol Clin. 2010;28(3): 593-617. 12. Jennum P, Børgesen SE. Intracranial pressure and obstructive sleep apnea. Chest. 1989;95(2):279-283. 13. Gil Z, Abergel A, Leider-Trejo L, et al. A comprehensive algorithm for anterior skull base reconstruction after oncological resections. Skull Base. 2007;17(1): 25-37. 14. Koutourousiou M, Filho FV, Costacou T, et al. Pontine encephalocele and abnormalities of the posterior fossa following transclival endoscopic endonasal surgery. J Neurosurg. 2014;121(2):359-366.
COMMENT
T
he authors present 2 cases of brain herniation through the frontal fossa defect after extended endonasal transsphenoidal surgery for skull base tumors. This rare, delayed complication of the extended endonasal approach should be recognized, particularly when rigid buttresses cannot be used to augment the closure and in patients with signs or risk factors of elevated intracranial pressure (sleep apnea, obesity, Chiari I malformation). When unrecognized, as seen in the second patient, the radiation therapy planning can inadvertently include the encephalocoele in the center of the radiation field. This may explain the severe radiation necrosis that this patient experienced. This supports the need for an open dialogue between the skull base surgeons and the radiation oncologists. We commend the authors for describing this rare complication of extended endonasal surgery. Garni Barkhoudarian Santa Monica, California
www.operativeneurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited
