Pre-operative direct puncture embolization of head and neck hypervascular tumors using SQUID 12

Abstract

Objective

The authors have evaluated their experience in pre-operative direct puncture embolization of hypervascular tumors of the head and neck using SQUID 12, an embolic liquid agent.

Methods

Between July 2016 and March 2019, the authors retrospectively reviewed clinical, embolization and surgical data of 11 consecutive patients with 12 hypervascular head and neck tumors who had undergone pre-operative embolization using SQUID 12. Percutaneous embolizations were performed by inserting a 19-22 Gauge needle directly into the tumor under ultrasound, fluoroscopic and/or endoscopic guidance. The hub of the needle was connected to a 15-cm DMSO-compatible extension tube, and the SQUID 12 was injected.

Results

Total or near-total devascularization was achieved in 11 over 12 cases. Complete en-bloc tumor removal by surgery was achieved in all cases. Only one patient required blood transfusion. No major periprocedural adverse events were recorded.

Conclusions

Direct puncture embolization of hypervascular tumors of the head and neck using SQUID 12 seems to be safe and effective. It may offer almost complete devascularization due to homogenous, deep penetration in the tumor, with optimal visibility of the agent throughout the percutaneous procedure. It may reduce intraoperative blood loss and the need for transfusion, thus facilitating complete surgical resection.

Introduction

Hypervascular head and neck tumors such as paragangliomas and juvenile angiofibromas are surgically challenging due to the risk of significant intra-operative blood loss.¹ Endovascular pre-operative embolization has become an important tool in the management of such tumors, by reducing blood loss and improving visualization of surgical planes.² There is a reasonable correlation between the degree of devascularization and the intra-operative blood loss.¹ Similarly, the deep compartments of the neoplasm, especially those located near the skull bases, are the most difficult for a surgeon to reach and manipulate, and therefore they are also the goal of endovascular treatment.³

In recent years, reports of trans-arterial or direct puncture embolization of hypervascular tumors of the head and neck using different agents have gradually emerged. Traditionally, this has been performed using resorbable materials like gelfoam,⁴ fibrin glue, polyvinyl alcohol (PVA),⁵ or permanent embolic agents such as ethanol or N-butyl-cyanoacrylate.^6,7 Recently, an increasing number of studies^2,3,7–15 reporting the use of non-adhesive liquid agents (ethylene vinyl alcohol copolymers – EVOH) such as Onyx seem to show a greater effectiveness of the procedures in question due to a slower rate of precipitation and improved control of the agent during injection, resulting in a better rate of devascularization.^7,16

SQUID is an EVOH polymer that has become recently available. We review our preliminary experience of preoperative direct puncture embolization of hypervascular tumors of the head and neck using SQUID 12.

Materials and methods

Data collection

We retrospectively evaluated all hypervascular tumors that had been consecutively embolized by direct puncture using SQUID 12, between July 2016 and March 2019. The following information was collected and analyzed: the age of patients, the stage and characteristics of the tumor (size and extension evaluated on the pre-operative contrast-enhanced MRI), angiographic complications (e.g. non-target diffusion of embolization material), the devascularization rate, the number of needles used and punctures, the amount of embolic agent, estimated blood loss (EBL) during surgery, pre- and post-hemoglobin and hematocrit levels and clinical outcome (assessed by clinical examination three months after the surgical resection). Before the embolization procedure, a catecholamine secretion test was performed in order to exclude this condition that contraindicates a percutaneous procedure.

Written informed consent about the interventional procedure was obtained in each case.

The study was performed in accordance with the principles set out in the World Medical Association Declaration of Helsinki (1975). Approval for data collection was based on local regulations. Ethical approval not required in our country for literature review and retrospective case series.

Embolization technique

Endovascular treatment was performed using a monoplane flat panel angiographic system (Allura Xper FD 20, Philips Medical Systems, The Netherlands). All procedures were performed under general anesthesia without neuromonitoring.

Preliminary diagnostic angiographic studies were obtained in order to define the extent of tumor blush, feeding arteries, draining veins, and extracranial-to-intracranial anastomoses. In each case, a needle (19-22 Gauge, 10 mm) was inserted in the tumor core under both ultrasound and fluoroscopic guidance, in order to guide the needle tip into the target components of the lesion, and to avoid any puncturing of non-target vessels (e.g. internal carotid artery – ICA). In one case of juvenile angiofibroma, the needles were inserted into the lesion under endoscopic and fluoroscopic guidance. Once the target area was reached, an intra tumoral angiogram was performed to confirm the needle’s position, to evaluate the extension of the tumor parenchyma and to detect the presence of potentially dangerous anastomoses with the internal carotid or vertebral arteries. Subsequently, the hub of the spinal needle was connected to a 15-cm DMSO-compatible extension tube (total dead space estimated at 1 mL). The needle was then flushed with saline solution and the empty space was filled with DMSO. Subsequently, SQUID 12 was injected through the needle using a 1 mL luer-lock syringe.

During injection of the embolic agent, a balloon microcatheter (SCEPTER 4 mm × 11 mm XC; Microvention, CA, USA) was inflated at the origin of the external carotid artery (ECA) to prevent retrograde reflux. Where present, feeding arteries directly emerging from common carotid artery bifurcation were controlled with opportune projections, and the balloon has always been inflated as close as possible to the ostium of the ECA so as to best protect the origin of these small branches.

Where required, multiple punctures were performed in the event of occlusion of the needle or when the embolic agent failed to progress into all compartments of the lesion.

The radiologic success of all embolization procedures was determined by the degree of residual neovascularity and parenchymal staining observed on control angiography in orthogonal planes at the end of therapy. A semiquantitative assessment of the extent of tumor devascularization was performed using the following grading system: poor (0%–30%); moderate (30%–70%); subtotal (70%–95%); near-total (95%–99%); and total (100%).⁵

At the end of the procedure, after deflation of the balloon, a trans-arterial angiogram was performed to assess the degree of devascularization of the lesion and to ensure that there was no inadvertent embolization into the brain circulation. After the embolization, all patients received a twice-a-day administration of dexamethasone 4 mg in order to reduce the inflammatory reaction due to the embolization; patients who were directly transferred into the Operating room did not receive this medication.

Results

A total of 8 carotid body paragangliomas, 3 vagal paragangliomas and 1 juvenile angiofibroma in 11 patients (6 male, 5 female, age range 17–79) underwent percutaneous embolization with SQUID 12. The mean volume of the lesion, assessed by cross-sectional imaging, was 335 mm³. In 3 out of the 11 paragangliomas, there was an upper extension to the skull base (assessed on MRI). Mean injection time was 57 min (range 40–77 min) evaluated considering the time between the first and the last injection of SQUID. Constant radiopacity of the embolic agent was observed throughout the entire procedure. The mean volume of SQUID 12 used per case was 19.3 mL. The mean number of needles inserted was 2.8. Successful near-total and total devascularization was obtained in 11 cases over 12, demonstrated by both the lack of tumor opacification in the final control angiograms and by the absence of a color-Doppler signal at the final control sonography in those cases with adequate acoustic windows. We could not achieve a complete devascularization of one huge paraganglioma (patient n.7 – subtotal devascularization) and a blood transfusion (800 mL) was required during the surgical time. No major endovascular-related complications were recorded; no inadvertent intracranial embolization occurred. One patient experienced ipsilateral hypoglossal nerve palsy after carotid body paraganglioma embolization, which was resolved by the oral administration of steroids.

The mean time between embolization and surgery was 1.8 days (range 1–3 days). Complete excision of the tumor was achieved in all cases. In all vagal paragangliomas, a transcervical-transmandibular approach was needed. The average duration of surgery was 184 min (range 86–262 min).

Mean estimated intraoperative blood loss was 367 cc (range 150–800 cc). The mean postoperative hematocrit level was 37.6% (range 34.1%–40.9%), and the mean postoperative level change was 1.91%. Mean hemoglobin level change was 1.1 g/dL. The main results are summarized in Table 1.

Table 1.

Patients’ characteristics and intraoperative variables.

No.	Age	Tumor	Size (mm3)	Amount of SQUID (mL)	No. of needles	EBL (mL)	Surgical time (m)	Hct level change (%)	Blood transf. (mL)	Complications
1	69	R Vagal Paraganglioma	300	18	2	780	86	4.2	–	–
2	35	R Vagal Paraganglioma	323	27	2	230	156	1.7	–	–
2	35	L Carotid Paraganglioma	315	24	5	410	220	2.8	–	XII nerve palsy
3	59	L Vagal Paraganglioma	351	16.5	4	370	170	1.9	–	–
4	17	Juvenile Angiofibroma	331	15	2	400	262	2.2	–	–
5	78	L Carotid Paraganglioma	260	10.5	4	180	195	0.6	–	–
6	65	R Carotid Paraganglioma	295	14.5	3	460	235	2.9	–	–
7	73	R Carotid Paraganglioma	650	32	2	800	180	0.9	800	–
8	46	R Carotid Paraganglioma	280	16	1	150	135	0.9	–	–
9	64	R Carotid Paraganglioma	290	17	1	180	165	1.1	–	–
10	37	R Carotid Paraganglioma	325	25	2	250	210	1.7	–	–
11	79	R Carotid Paraganglioma	300	24	2	195	190	2	No	–

EBL: estimated blood loss; Hct: hematocrit.

Illustrative cases

Case 1 (Figure 1)

Figure 1.

A left carotid paraganglioma in a 78-year-old patient. (a) and (b): left CCA angiogram showing the hypervascular nature of the tumor widening the carotid bifurcation and fed by hypertrophic vessels arising from branches of the ECA. (c) After inflation of the balloon in the ECA distal to the origin of the superior thyroid artery, multiple other tiny feeders are shown arising at the origin of the ECA itself. (d) For this reason, the balloon was replaced and inflated at the very origin of the ECA in order to avoid any unwanted reflux of the embolic agent. (e) Intra-tumoral angiogram obtained by direct puncture of the mass. (f) Control angiogram during embolization with SQUID 12 (balloon still inflated in the ECA). (g) Final angiogram after deflation of the balloon, showing the complete devascularization of the lesion. (h) and (i) Surgical procedure after pre-operative embolization with complete resection of the tumor fully filled with SQUID 12. CCA: common carotid artery; ECA: external carotid artery.

A 78-year-old woman presented progressive swelling of a left latero-cervical mass. Ultrasound and MR examinations demonstrated a carotid paraganglioma (maximum size: 4 cm). Pre-operative DSA showed a hypervascular mass widening the carotid bifurcation.

Under ultrasound and fluoroscopic guidance, a 20-Gauge needle was inserted directly into the mass. A balloon catheter was then inflated at the origin of the ECA to prevent retrograde reflux of the embolic agent. A volume of 10.5 cc of SQUID 12 was then injected into the mass, resulting in the complete devascularization of the lesion. No neurological deficit or adverse events were recorded after the embolization.

The following day, the patient underwent complete resection performed using the trans-mandibular approach. No intra-operative complications occurred. There was no need for any blood transfusion. The EBL was 180 cc. Post-procedural Hb was 11.7 g/dL (vs. 12.0 g/dL pre-treatment) with a hematocrit level of 36.7% (compared to 37.3% pre-treatment).

Case 2 (Figure 2)

Figure 2.

A juvenile angiofibroma in a 17-year-old boy. (a) Sagittal contrast-enhanced T1-weighted MR showing a contrast-enhancing tumor within the left nasal cavity extending posteriorly into the nasopharynx. (b) Left internal and (c) external subtracted angiograms demonstrating a hypervascular mass fed by multiple arterial branches of external carotid artery and a small artery arising from a petrous portion of the internal carotid. Lateral radiogram (d) showing a 19-Gauge needle inserted within the tumor by trans-nasal means. (e) Post-embolization lateral radiogram and (f) subtracted angiograms showing the SQUID 12 cast and the complete devascularization of the tumor.

A 17-year-old young adult presented with epistaxis and nasal stuffiness. Diagnostic work-up (MR) revealed a FISH 2¹⁷ juvenile angiofibroma of the left nasal cavity measuring 3 × 2 × 5 cm. DSA showed a hypervascular mass with arterial supply from multiple feeders arising from both the internal carotid artery (ICA) and the ECA; no feeders arising from vertebral artery (VA) was noted. Under endoscopic guidance, a 19-Gauge needle was inserted into two different portions of the mass. A volume of 15 cc of SQUID 12 was injected resulting in total devascularization of the lesion. No complications occurred during and following the embolization. Complete surgical resection was performed the next day through both trans-oral and trans-nasal approaches. There were no surgical complications, EBL was 400 cc and no intraoperative blood transfusion was needed. No significant changes in hemoglobin levels were recorded; the postoperative hematocrit level change was 2.2%.

Case 3 (Figure 3)

Figure 3.

A right vagal paraganglioma in a 34-year-old patient. (a) Sagittal reformatted CTA image showing a large vagal paraganglioma extending upward to the skull base; (b) Lateral view of right CCA angiogram showing the hypervascular nature of the tumor displacing both the ICA and ECA anteriorly, fed by multiple branches of the ECA; (c) anterior view of intra tumoral angiogram obtained by means of the direct puncture of the mass; (d) Anterior view of right CCA control angiogram showing the hypervascular tumor (widening the carotid bifurcation) and the correct position of the needle within the mass; (e) Anterior and lateral (f) views of right CCA angiogram showing the complete devascularization of the tumor at the end of the embolization procedure by direct puncture injection of SQUID 12. CCA: common carotid artery; ICA: internal carotid artery; ECA: external carotid artery.

A 35-year-old male patient was referred to our institution with a right cervical mass. MRA and CTA indicated an extensive vagal paraganglioma with superior extension to the skull base. An angiography revealed that the tumor was supplied by multiple branches of the ECA; no feeders were found arising from ICA or vertebral artery. The focus of the embolization was to achieve a complete devascularization of the paraganglioma. Under ultrasound and fluoroscopic guidance, one 19-Gauge needle was inserted into the tumor and a balloon was inflated in the proximal part of the ECA in order to avoid intra-arterial reflux during injection. A volume of 27 cc of SQUID 12 was injected into the tumor resulting in the total devascularization of the tumor. No angiographic or clinical complications were recorded. The resection was performed two days later through a transmandibular approach. There were no surgical complications, EBL was 230 cc, there was no need of intraoperative blood transfusion and the hematocrit level changed from 38.6% to 36.9%.

Discussion

In this report, we present our series of head and neck hypervascular tumors treated by direct puncture embolization with SQUID 12.

The highly vascular nature of head and neck paragangliomas, as well as that of juvenile angiofibromas, often represents a challenge for the surgeon.

In recent decades, endovascular therapy has become an important instrument for the management of head and neck hypervascular tumors. The aims of preoperative endovascular treatment are: bleeding control and the reduction of surgery-related blood loss, the intra-arterial administration of chemotherapy and a reduction in the progression of inoperable lesions. Pre-operative trans-arterial embolization was introduced in the 1970s.¹⁸ However, the trans-arterial route can be limited, since the feeding arteries may be inaccessible owing to one, or a combination, of the following factors: extremely small size, unfavorable angle or site of origin, excessive tortuosity or previous surgical ligation of the proximal external carotid trunk. In addition, the presence of dangerous extracranial-to-intracranial anastomoses and the risk of inadvertent reflux into the intracranial circulation increase the risk of inadvertent neurological sequelae.^5,11 These limitations can lead to non-target embolization and incomplete devascularization of the tumor with the consequent futility of the procedure for surgical purposes.

In recent years, EVOH copolymers, a new class of liquid embolic agents, have been widely used for the percutaneous treatment of vascular malformations.⁶ The well-known characteristics of this liquid embolic agent are a slower precipitation time, allowing deep intratumoral penetration, and a more controlled injection of the agent which can reduce the risk of non-target embolization via dangerous anastomoses.

Thanks to such properties, EVOH agents are increasingly used for the embolization of hypervascular head and neck tumors.

Alaraj et al.⁹ reported the first case of the trans-arterial embolization of a vagal paraganglioma using an EVOH copolymer. They achieved an adequate tumor devascularization but encountered a non-target migration of Onyx in the vertebral artery. Many other authors report the experience of trans-arterial embolization, with Onyx, of head and neck tumors.^7,12,14 Trans-arterial Onyx injection with a balloon-assisted technique has been proposed in order to reduce non-target embolization and dangerous anastomoses. Ladner et al., in their series of five patients, used a dual-lumen balloon inflated into the feeding pedicle for the injection of Onyx.¹⁴ With this technique, minimal proximal reflux and better intra tumoral penetration were achieved, although the resulting rate of devascularization was only 64%.

To overcome the limits of trans-arterial embolization, in 1994, Casasco et al. first described the embolization of hypervascular tumors using the direct puncture technique, with the injection of cyanoacrylate in the majority of cases and reported a good rate of devascularization with a low rate of complications.¹⁹ The main advantage offered by this technique, as also demonstrated by preliminary data from experiments conducted on animal models, consists in its ability to produce a complete filling of the tumor microvascular bed, creating a distinctive intra tumoral casting of the neovascular bed that is likely to result in a more complete and irreversible occlusion of the embolized vessels compared to traditional trans-arterial endovascular techniques.⁵ Furthermore, it allows easier access to the vascular tumor bed, which is not limited by arterial tortuosities, atherosclerotic disease or induced vasospasm. This technique could also reduce inadvertent embolization through extracranial–intracranial anastomoses, or to vasa nervorum, thanks to the fact that the embolic material needs to first traverse the tumor parenchyma before reaching the feeding pedicles.¹⁰

However, this does not fully overcome the complex vascular anatomy and multiplicity of the feeders often seen in these tumors. Feeding vessels arising from the ICAs or VAs still represent a technical challenge and can cause the reflux of the embolic agent into the intracranial circulation.^6,20 Wanke et al.⁸ first reported a small series of direct puncture embolizations of carotid body tumors with an EVOH copolymer (Onyx). Currently, the largest case series of paragangliomas treated by direct puncture injection with Onyx is that reported by Elhammady et al.¹⁰; nine carotid body tumors and three jugular paragangliomas were treated, resulting in a mean devascularization rate of 88%.

Recently, a case report described the use of PHIL (an injectable precipitating hydrophobic liquid) in pre-operative embolization of carotid-body paraganglioma with good results.²¹

Similar results were achieved using the direct puncture injection of EVOH copolymers in the pre-operative embolization of juvenile angiofibromas.³ Nine patients were treated and underwent resection using a standard open surgical or endoscopic approach; almost complete devascularization was achieved in all cases.

SQUID is an EVOH copolymer that has become recently available. It is composed of EVOH (ethylene vinyl alcohol copolymer) with suspended micronized Tantalum powder for radiopacity and DMSO (dimethyl sulfoxide) solvent. SQUID is available in four different formulations depending on viscosity (18 and 12) and on the percentage of tantalum powder dissolved (18 LD and 12 LD). While SQUID 18 has similar characteristics to Onyx 18, SQUID 12 is an innovative, unique formulation in terms of viscosity. With its lower viscosity, SQUID 12 is more fluid than other agents. In our series, we believe that this feature aided deep penetration into tumoral parenchyma and the diffusion of the embolic agent in areas of the tumor distant from the puncture site, including regions near the skull base which are difficult to reach for a surgeon. Other authors have found the same advantageous feature during the treatment of arteriovenous malformations (AVMs) with SQUID 12, compared to that using SQUID 18 or Onyx.²² We also noticed a constant radiopacity with optimal embolic agent displays right up to the end of each single procedure, that may be better than that achieved using Onyx18 in the case of prolonged treatment exceeding 30 min, as reported by Mason et al.²³

The devascularization rate we achieved was higher than 95% in 11/12 cases, which is comparable to the results achieved in previous series.^3,10 In addition, our results in terms of intraoperative EBL (mean value 367 mL) were slightly better than those of other reports: Gemmete et al.³; Elhammady et al.¹⁰, respectively, reported mean EBL of 506 mL and 567.7 mL. One paraganglioma was not completely filled by SQUID (80%) due to its huge dimensions that limited the ability to reach deeper zones, and the extended duration of the interventional procedure.

To our knowledge, this is the first consecutive series of head and neck hypervascular tumors entirely treated only with SQUID 12 by means of direct puncture. Akmangit et al. described their experience in the treatment of AVM, arteriovenous fistula and tumors with SQUID 12 and 18 using the endovascular approach.²²

In our small series, there were no treatment-related major complications and just one minor complication (hypoglossal nerve palsy probably due to the occlusion of tiny vasa nervorum), which was partially managed by the administration of steroids.

Other authors reporting their experiences of direct puncture embolization pointed to no, or a low rate of, treatment-related complications.^10,16

After embolization treatment with SQUID 12, in 11 over 12 tumors, our surgeons experienced minimal bleeding in the operating field which facilitated resections and the en bloc removal of the tumor. They reported a lower-than-expected intraoperative blood loss, better visualization of surrounding tissues and the well-defined demarcation of pathological margins since the tumor had been transformed into a rigid avascular mass which clearly stood out against the surrounding tissues, as also reported by other authors.^5,6 Moreover, they noticed reduced inflammation of the periadventitial soft tissues, when compared with cases previously embolized using more traditional techniques such as trans-arterial particle injection. In our series, the mean interval between the embolization procedure and the surgery was short (1.8 days); however, one of the advantages of embolization with EVOH copolymers compared to the embolization with particles is a wider time window available before surgery. This option must be kept in mind when planning a pre-operative embolization treatment with EVOH copolymers.

Conclusions

Our preliminary results with the direct puncture embolization of hypervascular tumors of the head and neck using SQUID 12 seem promising, as they reveal the effective and safe nature of devascularization therapy, with the prolonged, constant visibility of agent radiopacity throughout the procedures performed. We did not experience any relevant inadvertent migration of the embolic agent into non-target vessels, and we achieved a high degree of devascularization due to homogenous, deep penetration into the tumor, probably due to the low viscosity of the agent. In our limited series, this technique reduced intraoperative blood loss and the need for transfusion, thus facilitating en bloc tumor removal in all cases. The authors believe that a valid in vivo comparison between SQUID and other EVOH copolymers is not possible; in any case, it was not the purpose of this study.

Contributorship

All the authors listed gave substantial contributions to the conception of the work, to the acquisition, analysis and interpretation of data, also revising it critically for important intellectual content. All the authors gave their final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Declaration of conflicting interests

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

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent of all patients has been collected.

Funding

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

ORCID iDs

Iacopo Valente https://orcid.org/0000-0002-0451-2105 Andrea Alexandre https://orcid.org/0000-0002-8080-3916