(1) Department of Medical Imaging, University of Toronto, 101 College St, University Health Network, Toronto, ON, M5G 1L7, Canada
(2) Department of Otolaryngology-Head & Neck Surgery, University of Toronto, 190 Elizabeth Street, Room 3S-438, R. Fraser Elliott Building, University Health Network, Toronto, ON, M5G 2N2, Canada
(3) Department of Medical Imaging, University of Toronto, 610 University Ave, Princess Margaret Hospital, University Health Network, Toronto, ON, M5T 2M9, Canada
* Corresponding author Email: firstname.lastname@example.org
Modern advances in endonasal endoscopic surgery have expanded its utility in the management of malignant sinonasal tumours, and the degree to which cases are eligible. Knowledge of the indications, inclusion and exclusion criteria and imaging workup that empowers judicious patient selection is essential. This article discusses the endoscopic resection of malignant sinonasal tumours.
Studies describing the impact of patient quality of life will surely affect the fate of endoscopic resection, but the current trend suggests that this technique stands to replace an open approach when outcomes are similar. Combined craniofacial/endoscopic and craniofacial-only approaches to sinonasal malignancy remain an important option in the management of complex disease.
Over the past two decades, advances in endonasal endoscopic surgery in terms of both experience and technology have expanded its efficacy in the management of malignant sinonasal tumours. While an external approach remains the gold standard for resection of these tumours, undesirable morbidity can include facial incision and scarring, the need for craniotomy or facial bone osteotomy, surgical complications, longer hospitalisation period and slower recovery. When applicable, brain retraction with associated encephalomalacia and oedema can also be avoided[2,3].
Traditional external approaches seek to achieve an
Purported advantages of technical advances in endoscopic resection, particularly with the advent of angled endoscopes, include the ability to navigate difficult anatomic corners, with the subsequent provision of superior tumour visualisation. As opposed to
The authors have referenced some of their own studies in this review. These referenced studies have been conducted in accordance with the Declaration of Helsinki (1964), and the protocols of these studies have been approved by the relevant ethics committees related to the institution in which they were performed. All human subjects, in these referenced studies, gave informed consent to participate in these studies.
Endoscopic management of nasoethmoidal malignancies with or without involvement of the adjacent anterior skull base or orbital invasion has been shown to be an appropriate indication for pure endoscopic or cranioendoscopic surgery, with careful patient selection.
An endoscopic approach should be considered for tumours which occupy the central nasal cavity between the frontal and sphenoidal sinuses but do not extend to the lateral lamella of the pterygoid bone. Tumour invasion of the nasal bones, anterior/posterior table of the frontal sinus or frank orbital invasion are considered contraindications. Posteriorly, it is important to assess invasion of the carotid and cavernous sinus. If tumour is noted tracking along nerves (most importantly trigeminal), these are also relative contraindications to an endoscopic approach. Tumour can be resected from the periorbita, but the tumour-invading ocular muscle will typically require orbital exenteration and thus an open approach. Endoscopic orbital exenteration has been described but is not widely practiced[3,9]. Malignant tumour types that have been resected with favourable results include adenocarcinoma, adenoid cystic carcinoma, chordoma, malignant melanoma, olfactory neuroblastoma, osteosarcoma and squamous cell carcinoma[9,10,11,12].
In many situations, a lateral rhinotomy incision and craniotomy can be avoided by endoscopic resection of the extracranial portion of a sinonasal malignancy with transcranial resection at the anterior skull base[2,13]. In such cases, margins of the mucosa, periorbita and dura should be microscopically clear to ensure local control on par with open traditional approaches. A number of studies have shown that locally advanced malignancy across a variety of histopathological types can be treated endoscopically[1,14,15]. However, long-term data are lacking in terms of a direct comparison between open and endoscopic techniques. If the tumour extends lateral to the midpoint of the orbital roof, typically by way of dural spread, it becomes difficult to resect. Moreover, the dural reconstruction is challenging, leading to higher rates of cerebrospinal fluid (CSF) leak in the postoperative period. As a result, open craniotomy is combined with endoscopic resection to avoid lateral rhinotomy[12,13]. As robotic surgery and 3D high-definition endoscopy technologies advance, it is anticipated that these boundaries and the capabilities for skull base reconstruction will both improve.
Single olfactory bulb resection, such as in olfactory neuroblastoma, is challenging with a standard craniofacial approach, and thus an endoscopic approach is proposed as the preferred method of resection. However, contralateral olfactory bulb salvage is currently rare. In addition, these patients are only eligible in the absence of intraorbital or intracranial involvement, which underscores the importance of appropriate patient selection. In cases such as these where clear margins will be challenging, endoscopic resection should be avoided.
Endoscopic resection of sinonasal tumours that are centrally located in the nose and sinuses should be considered before the external approach is used. While adenocarcinoma, olfactory neuroblastoma, squamous cell carcinoma, adenoid cystic carcinoma and malignant melanoma are the most frequent indications for endoscopic surgery, it remains unclear which tumour types are most appropriate given anatomically favourable positions. An endoscopic approach is also recommended for surgical biopsies of tumours which penetrate into the sphenoid sinus from the sella, the petrous apex and/or adjacent areas.
Unfortunately, there are still no criteria which absolutely indicate whether a sinonasal malignancy is resectable by endoscopic means. Generally, however, tumours that demonstrate invasion of the cavernous sinus are associated with significant patient morbidity and should not be approached endoscopically. In addition, transdural extension is a relative contraindication for both open and endoscopic resection. In some cases, such as optic chiasm invasion, these tumours are not resectable by any means[4,19].
Other contraindications to a pure endoscopic approach include extension into facial and orbital soft tissues. Resection of a malignant tumour of the naso-ethmoidal complex through a purely endoscopic approach is also contraindicated as is extensive involvement of the lacrimal pathways, extension through orbital contents, the superior and lateral aspects of the frontal sinus and the floor of the nasal cavity. Patients with involvement of the medial wall of the maxilla and the medial portion of the posterior wall are often excellent candidates for endoscopic resection. However, involvement of the superior, anterior, inferior and lateral walls may require conversion to an open approach. Involvement of the brain or other extension which would require resection of the anterior cranial fossa dura lateral to approximately the mid-orbital roof is also contraindicated.
The main role of imaging in the setting of sinonasal malignancy is to identify malignant disease, its anatomical extent and any metastatic lymphadenopathy. Radiological findings suggestive of malignancy include bony involvement (erosion and destruction), a soft tissue component, a unilateral disease process and soft tissue necrosis with or without lymphadenopathy. The clinician should be aware of factors which are associated with particularly poor patient prognosis, such as tumour encasement of the carotid artery, extension to involve the periorbita or dura, cavernous sinus invasion and perineural tumour spread. Metastatic lymphadenopathy is suggested by the presence of nodal clustering, rounded shape, inhomogeneity on contrast-enhanced studies and peripheral spread, all in the context of a plausible drainage pathway. These factors are essential determinants in the planning of endoscopic surgery and potential postoperative radiation when applicable.
When tumours occur in the maxillary or ethmoid sinuses, the American Joint Commission on Cancer T-category is applied to imaging findings. Computed tomography (CT) represents the best modality with which to assess for the presence of bony remodelling or bony invasion, for example of the sinus walls, orbital margins and the floor of the anterior cranial fossa (Figure 1). Coronal and axial views are essential for complete assessment of the pterygoid plates, maxillary sinus, ethmoid bullae and sinus and sphenoethmoid recess. Magnetic resonance imaging (MRI) is an important tool for the assessment of bone marrow invasion, where the high T1 signal fatty marrow is replaced by that of tumour[20,23]. Displacement of the periorbita is typically diagnosed on CT, with a reported NPV of 86% and PPV of 75%. The essential orbital finding is invasion beyond the periorbita (i.e. fat, muscle) as periorbital invasion can be approached endoscopically, but involvement of the fat or muscle will usually necessitate orbital exenteration (Figure 2). MRI is also valuable in the assessment of dural involvement and brain invasion. The former is evidenced by focal thickening and enhancement of the dura. Brain invasion is suggested by the presence of both brain oedema and parenchymal enhancement and cortical disruption (Figure 3).
Sagittal CT bone algorithm image of a squamous cell carcinoma in the ethmoid sinus (asterisk) that has eroded the overlying floor of the anterior cranial fossa.
Coronal T1- (a) and T2 (b)-weighted images show an ethmoid sinus carcinoma that has disrupted the periorbita along the right medial orbital wall. Note the normal appearance of the periorbita on the left side which appears as a smooth curvilinear black line (white arrows in (b)). On the affected side, the black outline is attenuated and more irregular in contour. However, (a) still shows preservation of the orbital fat (arrow in (a)) and the rectus musculature.
Two coronal T2-weighted images of a patient with an ethmoid squamous cell carcinoma taken approximately 2 months apart. Note the breach of the normal black cortical outline of the ethmoid roof and cribriform plate. (a) Evidence of dural thickening and irregularity compatible with dural invasion (white arrow in (a)). (b) A new contiguous tumour involvement of the inferior grey matter/cortex of the right orbital gyrus (white arrow in (b)) and new oedema in the adjacent white matter (dashed arrow in (b)).
Purely endonasal resections within the frontal sinus are also limited by the lesion laterality. Using the the modified Lothrop procedure, the floor of the frontal sinus can be removed to approach tumours. However, this is typically limited to those tumours which have extended into the sinus with minimal or no bony involvement. Tumour occupying the frontal sinus is often heterogeneously enhancing on a gadolinium T1 sequence. Vascular encasement and nerve involvement, best identified with MRI, are also contraindications to pure endoscopic resection, but a combined external and endoscopic approach may facilitate circumferential dissection around these structures[23,25]. Last, MRI has been shown to have a 100% sensitivity and 94% specificity in the identification of jugular vein and venous sinus invasion.
Morbidity and mortality are significant concerns with both craniofacial-only and combined craniofacial/endoscopic resection, including but not limited to meningitis, encephalomalacia/oedema, pneumocephalus, trismus, blindness and bone flap necrosis[4,15,27]. However, the most common serious complication is CSF leak, and contemporary advances in endoscopy have yielded rates comparable with open resection. With the advent of the vascularised nasoseptal flap, CSF leak rates have dramatically fallen in most centres. A study of 800 endonasally treated patients revealed a CSF leak rate of 15.9%, in which all but one case was treated successfully with repeat endonasal endoscopy or lumbar drainage. These data are favourable for endoscopic over external resection, though the majority of the tumours in the study were benign.
A large series of malignant sinonasal tumours treated with pure endoscopic (72.8%) or cranioendoscopic (27.2%) techniques involving 184 patients showed a 5-year disease-specific survival of 91.4% and 58.8% respectively. The authors concluded that while operator experience is a significant factor, endoscopic management of malignancies eligible for either external or endoscopic resection is an equally efficacious alternative to an external approach. The largest US study to date addressing pure endoscopic (77.5%) and cranioendoscopic (22.5%) resection of malignant tumours corroborated the assertion that oncological outcomes are acceptable with these approaches. In their study of 120 patients, where the most common site of tumour origin was the nasal cavity (52%), positive margins were present in 15% of patients and the 5- and 10-year disease-specific survival rates were 87% and 80%, respectively. Recurrence rates did not differ between the two groups. Encouragingly, CSF leak occurred in 3% of patients and was also not significantly different between the groups. It is notable that in both studies, the proportion of olfactory neuroblastoma was high, as these are often associated with good prognosis for 5–10 years before possible recurrence. In addition, the proportions of patients in either treatment group were not necessarily matched for disease stage.
Last, there is great theoretical benefit for elderly patients undergoing resection of sinonasal malignancies. Outcome of craniofacial resection in patients 70 years of age and older are significantly worse in terms of mortality, complication and disease-specific long-term survival. Given that elderly patients often do not tolerate brain retraction well, pure endoscopic resection stands to improve outcomes in this population.
Endoscopic techniques are being employed with increasing frequency in the setting of sinonasal malignancy. In order to best utilise this approach, appropriate patient selection is essential. Concurrent advances in MRI are poised to assist in the discrimination of surgical candidates, particularly with improved detection of perineural and vascular involvement.
As international trials continue, long-term follow-up and further insight into oncologic outcomes will become apparent, ideally with disease-specific data. Additional information about the potential for improved quality of life will undoubtedly be a major determinant in the future use of endoscopic resection. Combined craniofacial/endoscopic and craniofacial-only approaches to sinonasal malignancy remain an important option in the management of complex disease.
CSF, cerebrospinal fluid; CT, computed tomography; MRI, magnetic resonance imaging.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.