Transnasal devascularisation of a sinonasal hypervascular tumour (glomangiopericytoma) with direct injection of liquid polymer agent (Squid®)

Abstract

The technique of direct intratumoral injection of permanent liquid agents has evolved significantly over the past few decades. The percutaneous direct puncture technique is being utilised more and more for embolisation of head and neck hypervascular lesions. We describe a case treated via the direct percutaneous transnasal (TN) puncture technique for embolisation of a sinonasal hypervascular tumour, performed in a 21-year-old male patient. Due to extensive vascularisation of the tumour, preoperative embolisation was requested by our otolaryngologist team in order to minimise intraoperative bleeding and reduce surgical morbidity. A primary attempt at trans-arterial embolisation through selective ophthalmic artery catheterisation was unsuccessful due to unfavourable anatomy. An extensive devascularisation of the nasal fossa lesion was obtained using a single injection of low-density polymeric embolic agent (SQUID®). Post-operative eradication was uneventful. Final histopathology exam revealed vascular tumour consistent with a rare sinonasal glomangiopericytoma. To our knowledge, this is the first reported case of a rare sinonasal glomangiopericytoma treated by direct TN injection, utilising the liquid polymer agent SQUID®.

Introduction

Sinonasal glomangiopericytomas (GPC) are low-malignancy soft-tissue tumours composed of perivascular modified smooth cells (pericytic cells). GPC account for 0.5% of sinonasal tumours, with 15–25% of GPC occurring in the head and neck, with only 5% found in the nose or paranasal sinuses. Sinonasal GPC affects middle-aged patients but is possible at any age.¹ Clinical presentation includes nasal obstruction, epistaxis, facial pain, headache and visual disturbances. GPC are characterised by low tendency of metastases and exhibit a recurrence rate of 25%. The therapy of choice is surgical resection via an endonasal approach.² A preoperative embolisation helps to avoid intraoperative massive bleeding and affects tumour size and the area of resection positively.^3–6 Primary radiotherapy of GPC is associated with high rate of recurrence, while adjuvant radiotherapy may reduce risk of recurrence in patients with incomplete surgical resection.

When the conventional transarterial route is inaccessible, a direct percutaneous transnasal (TN) approach can be a valid alternative, after carefully ruling out all the dangerous anastomotic feeders during pre-embolisation diagnostic angiography,^7–9 such as in our case in which the technique was feasible, safe and effective.

Case presentation

This study was carried out according with Declaration of Helsinki principles. The patient gave informed consent to publish this work and to report individual data.

A 21-year-old male patient was admitted to our Ear Nose Throat (ENT) department after a 3-month history of headache, left ocular pain and mild proptosis. A maxillary MRI done elsewhere showed a large inhomogeneous lesion in the left nasal fossa, with ipsilateral orbital and anterior cranial fossa extension. A CT scan, performed at our department, confirmed the presence of a heterogeneous enhancing lesion in the left nasal fossa, with erosion of ipsilateral nasal lacrimal duct, lamina papyracea and ethmoidal planum, resulting in orbital and intracranial extension. No enlargement or erosion of the sphenopalatine foramen and Holmann-Miller sign were demonstrated at CT scan.

Nasal endoscopic examination demonstrated a solid, easily bleeding mass occupying left middle meatus reaching the top of the nasal fossa. Due to the presence of a strong post-contrast enhancement on CT and MR images, suggesting a hypervascular tumour, a diagnostic digital subtraction angiography (DSA) was performed and surgical removal was planned 48 h after embolisation (Figure 1).

Figure 1.

Contrast-enhancement high resolution CT multiplanar reconstructions (a-b: axial plane; c-d: coronal plane; e-f: sagittal plane) demonstrate left middle meatus lesion, characterised by strong, inhomogeneous enhancement. The lesion extends to the left orbit through the lamina papiracea (red arrows) and toward the anterior cranial fossa through the ethmoidal planum (white arrows). T2-MR (g) and postcontrast fat saturated-T1 (h) coronal images, better delineate the anatomical relationship between the lesion and the surrounding structures.

Endovascular procedure

DSA demonstrated an intense tumour blush supplied mostly from ethmoidal arteries (EA) branches of the left ophthalmic artery (OA), with appearance of a direct ‘arterio-tumour’ (A–T) shunt, and from ipsilateral internal maxillary artery (IMA) branches. Preoperative devascularisation was considered crucial in order to limit bleeding loss during surgery and to avoid lesioning the ophthalmic artery and the orbit. Because of its deep position close to orbit structures, the A–T shunt, supplied by EA–OA branches, was considered difficult to manage during TN surgery. Due to a very unfavourable anatomy, several attempts of superselective left OA micro-catheterisation failed to reach an appropriate position for glue injection (Figure 2).

Figure 2.

Left external carotid artery DSA, LL subtracted view (a), demonstrates supply to the anteroinferior part of the lesion by IMA branches; the left internal carotid artery DSA, LL (b) and AP (c) views, show a prominent blush in the postero-superior part of the lesion supplied by EA originating from OA. Direct A-T shunt from EA is best recognisable in the magnified picture (red arrow in d).

Transnasal embolisation

A few days after endovascular procedure, as an alternative approach, we considered to embolise the lesion via a direct TN puncture. Ipsilateral external and internal carotid arteries angiograms were obtained to assess the feasibility of percutaneous embolisation, to delineate the target, to evaluate dangerous vascular collaterals to the intracranial circulation, and to determine the appropriate working projection. Under general anaesthesia, using biplane angiography road-map guidance and bone landmarks, the lesion was punctured transnasally utilising a 20-gauge needle. The needle was placed into the lesion close to the A–T shunt under road-map guidance, with the intent to obliterate the deeper compartment of the tumour. After ‘trans-needle’ angiography to check the correct tip position, a total of 15 ml of non-adhesive embolic agent (SQUID-12®, Emboflu, Switzerland) was injected for a total fluoroscopy time of 25 min, directly through the needle, utilising the short DMSO-compatible plastic tube available in the commercial kit on request. The injection of SQUID® was performed in a continuous manner, due to the absence of dangerous reflux. Unexpectedly, the glue filled the entire tumour in a total time of 25 min, obtaining almost complete devascularisation of the lesion, demonstrated also by Xper-CT DSA (Allura Philips, The Netherlands) (Figure 3).

Figure 3.

The direct TN puncture (a) is performed under transarterial road-map guidance (b, LL view), the SQUID injections are performed under live roadmap (LL view in c); the final cast is well visible in the native DSA (LL view in d), showing the deep penetration of the agent into the vascular tumour bed.

Endoscopic surgery and histopathology

Two days after embolisation, endoscopic TN middle and inferior turbinectomy and left maxillectomy were performed, obtaining tumour debulking. Despite the embolisation, it was necessary to isolate and clip the left IMA, and both EAs had to be cauterised. Periorbital and anterior basicranial lesion components were removed and a Hadad-flap was made to reconstruct the skull base defect. Final histopathology exam showed a hypervascular, partially embolised necrotic lesion, characterised by uniform spindle-cells proliferation with lack of atypia. Despite the initial suspicion of a juvenile angiofibroma, the immunohistochemistry was positive for beta-catenin and cyclin-D1, typical of sinonasal glomangiopericytoma (Figure 4).

Figure 4.

Histological features of GPC, highly vascularised neoplasia with sign of embolisation x 100 (H&E stain in a); in a well preserved area of the tumour the neoplastic cells express nuclear beta-catenin (b), with variable expression of Cyclin D1 x100 (c). In the T2w-coronal MR image post-operative changes are visible (d).

Discussion

To our knowledge, this is the first reported case of TN embolisation of a sinonasal GPC with non-adhesive embolic agent. Only one study has reported percutaneous embolisation of an orbital GPC with Onyx™ (Medtronic, Irvine CA).⁸ The technique of direct intratumoral embolisation was introduced in the early 1990s by Casasco et al. to overcome the limitations of transarterial devascularisation.⁹ Generally, preoperative selective embolisation of internal maxillary artery is performed utilising polyvinyl alcohol (PVA), particulate microspheres or n-butylcyanoacrylate glue (n-BCA).⁵ We decided to use SQUID-12® instead of the SQUID-18® or Onyx™ based on our personal experience and prior reports, considering the lower density and the micronised tantalum particles (smaller than in Onyx), supposedly offering a more homogeneous solution with a deeper penetration into the tumour vascular network.¹⁰ However, we are not aware of any study comparing the behaviour of these different agents’ in tumours.

In our case, there were no adverse effects, and preoperative GPC direct TN embolisation with SQUID-12® was feasible, safe and effective.

Despite this, some limitations and warnings must be taken into account. Primarily, this procedure requires the availability of a multidisciplinary team (neuroradiologist, otolaryngologist and anaesthetist), and must be performed under general anaesthesia in a dedicated neuroangiography suite. Secondly, proper patient selection and a very experienced interventionalist team are mandatory in order to avoid unnecessary procedural risk, ensuring high-quality care. Finally, the high price of non-adhesive polymeric agents is a critical aspect that should be taken into consideration.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.