Preoperative Squid embolization of carotid paragangliomas with direct puncture

Abstract

Background and purpose

Preoperative embolization of carotid paragangliomas is a common procedure in interventional neuroradiology. Direct puncture embolization has shown less morbidity and mortality than endovascular embolization and a higher percentage of devascularization. We describe our experience using Squid® as the only embolic agent in direct puncture glomus embolization.

Methods

We retrospectively reviewed pre-embolization imaging tests, emphasizing the volume of the lesion, clinical history data, technical aspects of the procedure, as well as the approximate amount of blood lost during the surgical procedure in all patients with preoperative embolization of carotid paragangliomas performed at our tertiary care hospital.

Results

Six patients met our criteria from May 2017 to August 2018. The volume of the mass ranged from 1.4–18.5 mL and the quantity of Squid® injected varied from 1.1–15 mL. Total devascularization was achieved in almost all cases (>90%), with one puncture needed in all but one patient, who was punctured two times. No hemorrhagic complications were described in surgery, no blood transfusions were needed, and the mean decrease of hemoglobin after surgery was 1.04 g/dL.

Conclusions

Direct puncture embolization of carotid paragangliomas only using Squid® is a safe and relatively simple procedure that facilitates the subsequent surgery with minimal blood loss.

Introduction

Head and neck paragangliomas are tumors derived from ectodermal cells of the neural crest. These are dysgenetic neuroendocrine lesions, whose characteristics of special neuroradiological interest are their multiplicity potential and cellular polymorphism. The former forces the discarding of the existence of other multifocal lesions, such as bilateral, ectopic, and in-syndromic association with other tumors, and multicompartment lesions, defined as multiple lesions in the same tumor mass. These lesions are highly vascularized and their irrigation shows a compartmental distribution as “puzzle pieces,” each irrigated by a specific artery. Superselective arteriography will opacify a segment of the lesion (tumor nidus) with an independent and specific venous drainage.¹

Although these tumors are benign, they may be locally destructive and require surgical resection,² but high irrigation may result in a notable amount of bleeding during surgery, in some cases hindering the complete removal of the lesion.³ Pre-surgical embolization is an intervention that has been shown to decrease the time of the surgical procedure, the amount of blood loss during surgery, and therefore morbidity and mortality.⁴

The most commonly used method for embolization of these lesions is transarterial, described for the first time by Borges et al.⁵ It involves supraselectively catheterizing each of the arterial feeders of the glomus and embolizing the lesion through pedicular closing with coils, with particles, or with liquid embolic agents. However, the transarterial pathway is conditioned to the vascular anatomy of the afferent arteries, that is, their tortuosity, atherosclerotic disease, or the possibility of producing a vasospasm during embolization. This results in a longer and more tedious intervention and often makes total devascularization of the paraganglioma impossible. In addition, special care must always be taken with dangerous anastomoses between the branches of the external carotid artery and the internal carotid, vertebral and ophthalmic arteries, especially if particles are used.

Paraganglioma embolization using a direct puncture technique was described by Casasco et al.⁶ This procedure, combined with surgery, is relatively simple for most interventional neuroradiology teams. Its effectiveness simplifies the surgical procedure and, therefore, facilitates the complete removal of the tumor, while reducing the surgical risk and the risk of recurrence.

The aim of our study is to review all patients who have had a preoperative embolization of carotid paragangliomas by direct puncture using only the liquid embolic agent Squid® (Balt Extrusion, Montmorency, France) in our tertiary care center. Although there are described in the literature cases treated with Onyx® (Medtronic, Minneapolis, USA), to our knowledge, this is the first time this technique has been described with this embolic agent.

Methods

We retrospectively reviewed all cases of carotid paragangliomas surgically removed in our tertiary care center, in which Squid®, an ethylene vinyl alcohol copolymer (EVOH), was used as the sole agent for direct percutaneous presurgical embolization. We reviewed the pre-embolization neck imaging studies (computed tomography (CT) or magnetic resonance imaging (MRI)), measuring the tumor size in the three dimensions, transverse, anteroposterior, and craniocaudal. Furthermore, we used the AW Server version 2.0 (GE Healthcare, Chicago, IL, USA) to perform three-dimensional (3D) reconstructions and thus be able to measure the tumor volume more accurately. With the Vascular Surgery department, we thoroughly examined the embolization angiographic imaging studies and the medical and operative notes.

All the interventions were performed under general anesthesia. We used the standard percutaneous transfemoral arterial access and placed a short 5F sheath. With a 5F diagnostic catheter we first performed a diagnostic cerebral angiography with the catheterization of common, external, and internal carotid arteries in our neurointerventional suite using biplane angiography. Appropriate pre-embolization angiographic analysis of the lesion helped us identify the arterial feeders, evaluate dangerous anastomosis with the pial circulation (mainly those between the ascending pharyngeal artery and the petrous and cavernous internal carotid artery), delineate the hypervascular lesion, and obtain the best projections to facilitate the percutaneous puncture and a safe embolization. The 5F catheter was left in the common carotid artery connected to a saline perfusion and was used to perform control series during embolization.

All paragangliomas were punctured with a 22-gauge spinal needle (BD, Oakville, Ontario, Canada) using a fluoroscopic biplanar arterial roadmap. The needle tip was considered properly positioned in the glomus when slow but continuous blood reflux was identified. At that time, iodized contrast was administered with a gentle injection through the puncture needle to verify the position. After this, the dead space of the needle was filled with dimethyl sulfoxide (DMSO) via fluoroscopy to see how the contrast continued to come out from the tip of the needle. By the time the contrast flow ceased, the DMSO had completely filled the dead space and Squid® was injected slowly but continuously using a 1 mL luer-lock syringe, without a connection tube always under a negative roadmap. This roadmap was refreshed multiple times to know exactly where the embolic agent was spreading. If the Squid® was directed to an unwanted area or vessel, the injection was stopped for approximately 1 minute to ensure adequate solidification of the embolic agent before we continued (Figure 1). This process was repeated as many times as necessary. Between injections, anteroposterior and lateral series were made from the arterial catheter to quantify the amount of lesion that is embolized. If it was seen that an area of the lesion was far from the puncture point and after several attempts the agent did not penetrate there, we proceeded to make a new puncture in the desired area to achieve an embolization of the lesion as completely as possible.

Figure 1.

Patient 1 in Table 1. (a) Sagittal maximum intensity projection contrast-enhanced computed tomography (CT) scan of the neck showing a left carotid body tumor. (b) Three-dimensional (3D) volume-rendering of the aforementioned lesion. (c) Lateral left common carotid artery angiogram demonstrating a hypervascular mass splaying the carotid bifurcation. (d) Lateral single-shot showing direct needle placement into the tumor with the first Squid® injection. (e) Lateral single-shot at the end of the procedure demonstrating the distribution of the embolizing agent. (f) Lateral left common carotid artery angiogram at the end of the procedure that depicts an almost complete devascularization of the paraganglioma.

We considered presurgical embolization to be a technical success when the degree of residual parenchymal staining of the lesion was ≤10% compared to the initial diagnostic angiography. All embolization procedures were completed in a single session.

The day after the procedure, our vascular surgeons removed the lesion. To quantify the amount of blood lost in the operation we looked for the number of blood transfusions needed and hemoglobin loss comparing blood tests immediately before and after surgery. In addition, in this study we compared the volume of the paraganglioma, measured in the pre-intervention imaging tests, with the total volume of embolic agent used.

Results

A total of six patients with carotid body paragangliomas underwent direct puncture preoperative embolization with only Squid® (Table 1).

Table 1.

Carotid body paraganglioma characteristics, treatment, and surgical blood loss.

Patient n°	Side	Tumor size (mm)	Tumor volume (mL)	N° of punctures	Squid (mL)	Devascularization	Hb loss (g/dL)
1	Left	42 × 29 × 50	16	1	15	>90%	1.4
2	Left	34 × 48 × 28	18.5	2	6	>90%	0.3
3	Right	22 × 17 × 27	5.3	1	4	>90%	2.4
4	Right	17 × 15 × 33	4.4	1	3	>90%	0.5
5	Left	12 × 10 × 19	1.3	1	1	>90%	0.6
6	Right	16 × 15 × 25	3.9	1	4	>90%	1.0

There were five women and one man aged between 17 and 72 years old. Tumor size varied from 1.3 mL to 18.5 mL. In all six cases an almost complete devascularization was achieved (at least 90%) using a direct percutaneous embolization technique with EVOH as the sole agent. A second puncture was performed in one tumor only. All punctures were done using biplane transarterial roadmap fluoroscopy without ultrasound guidance. The median duration of the procedure was 38 minutes, with a range between 32 and 42 minutes. The quantity of Squid® injected varied from 1.1–15 mL and in one patient was higher than 7 mL. The main objective was near-complete devascularization described as a decrease in the vascular enhance of over 90%. Near-complete devascularization was achieved in all cases evaluated using the last angiography run by two board-certified neuroradiologists (MM and SR).

Surgical removal was performed 24 h postembolization by vascular surgeons (JS and colleagues). Successful tumor resection was reported by the surgeons in all six cases through personal communication and clinical reports. No hemorrhagic complications were described in surgery and no blood transfusions were needed. The mean decrease of hemoglobin after surgery was 1.04 g/dL with an average postsurgical hemoglobin of 12.8 g/dL. All the loss of blood was reported to originate from the cervical incision with no blood from the tumor excision.

Non-target embolization was seen in all patients as the retrograde reflux through irrigating arteries or anterograde through veins; when either of these two migrations appeared, the injection was stopped for 1 minute or until no further embolization was seen. No patient suffered non-target embolization into the intracranial circulation and no complications related to direct puncture were described.

Discussion

Carotid body paragangliomas are a rare type of tumor derived from ectodermal cells of the neural crest. They need surgical removal because of the risk of growth and local invasion. Due to their high-grade vascularization, surgical removal is a demanding procedure with a high risk of blood loss. Therefore, presurgical embolization is a good option to decrease the risk of bleeding and the need for postsurgical transfusion.^1,3

Preoperative embolization for treating high vascular tumors has been done for the last 25 years.⁶

Carotid paragangliomas have been embolized with particles, Glubran® (GEM SRL, Lugano, Italy) or Onyx®. At its origins, this embolization procedure was achieved by intra-arterial treatment using microcatheters for the injection. This intravascular approach was limited by difficulty to navigate in certain anatomies or the small size of some feeders. Some stroke complications have been described after the use of intra-arterial particles for tumor embolization.⁷ Due to those complications in our institution, we decided to change from standard presurgical embolization with particles to Onyx® 4 years ago starting with intra-arterial embolization. We changed from intra-arterial to direct puncture of Onyx® because surgeons experienced some difficulties from the obliteration not only of the tumor but also of the pedicles used to introduce the agent. Some concerns have arisen in the surgical community due to the technical changes of surgery derived from the use of Onyx® instead of classic particles, making the removal of the tumor more difficult or leading to the appearance of sparks when using an electric scalpel.⁸ Moreover, if part of the embolic agent remains between the lesion and the skin puncture point, in addition to having more difficulties in accessing the lesion, surgeons could confuse it with the glomus.

Direct puncture was used as an adjuvant treatment at the beginning using acrylic glue, then as a sole technique with intra-arterial control. Due to the properties of the glue, several injections are needed because of the fast polymerization. Onyx® was then used, taking advantage of its slow thrombosis of the vessels and higher tumor penetration, diminishing the number of punctures needed.¹⁰

Squid® is a non-adhesive liquid embolic agent for embolization. Chemically, it is a copolymer of EVOH dissolved in a DMSO solution with suspended micronized tantalum powder for radio-opacity. It has four different presentations in two densities (12 and 18) and with low or standard doses of tantalum (low-density or normal). The use of Squid® in the treatment of AVM, fistulas, and tumors described it as a safe and effective embolic agent with satisfactory obliteration rate.^10,11

In our center, we have experience treating carotid paragangliomas with particles and Onyx® through an intra-arterial approach and direct puncture with Onyx®. We decided to change from Onyx® to Squid® 18 LD because in the latter, the grain size of tantalum powder is smaller, providing less sedimentation and higher homogeneity to the embolic mixture. This property leads to a more distal penetration of the agent into the vessels and lends a lesser radiopacity, which allows a better visualization of the distribution inside the lesion and therefore greater control over the injection of the liquid embolization material.

Other papers⁹ have described a complete obliteration of the lesion with several products and we did not want to get that level of occlusion, trying to diminish the possibilities of embolic agent migration to the internal carotid artery through the vasa vasorum. Even with a near-complete obliteration, no large blood loss was described during surgery and no blood transfusions were needed.

Squid® has another lower density presentation called Squid® 12 that could embolize further and through smaller pedicles. We have not used it because we thought it would be easier for this presentation to reach the internal carotid artery through the vasa vasorum, a problem that may appear in this kind of direct puncture embolization.

In this study, we also wanted to compare the paraganglioma volume to the quantity of Squid® needed in each intervention. We retrospectively made 3D reconstructions of the lesions using CT scans or MRI before the intervention; these were taken at least 2 weeks before the procedure. This showed that in all cases, a smaller quantity of embolic agent was used compared to the volume of the lesion, although this difference was minimal except in one patient (18.5 mL versus 6 mL). Reviewing the CT images of the latter, we realized the lesion presented substantial parenchymal necrosis, which could explain this striking difference between the volume of the injury and the amount of Squid® injected. In addition, in this patient we needed two different punctures, probably because the Squid® was not able to go through the necrosis (Figure 2). This calculation of the lesion volume would help predict how many bottles of Squid we should prepare and shake before the intervention.

Figure 2.

Patient 2 in Table 1. (a) Anteroposterior left common carotid artery angiogram demonstrating the hypervascular mass. (b) Anteroposterior parenchymogram demonstrating direct needle placement into the tumor with contrast injection showing the neovascularity of the tumor. (c) Anteroposterior left common carotid artery angiogram halfway through the procedure showing still enhancement of the most cranial and caudal portions of the lesion. (d) Anteroposterior parenchymogram demonstrating the new placement of the needle after a second puncture with contrast injection showing the cranial portion of the glomus.

This study has limitations. It is retrospective and we do not compare the efficacy of Squid® with other embolic agents or with transarterial embolization. The number of patients is low; therefore, we cannot generalize these findings in a larger cohort of patients.

Conclusions

After carrying out these six procedures with no complications and with good technical management, we believe Squid® is a safe and useful agent for this kind of presurgical embolization of tumors and may offer a higher degree of devascularization when compared with other embolic agents using an endovascular route.

Conflict of interest

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

Ethical approval

This is a retrospective study; formal consent is not required for this type of study.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

ORCID iD

Carlos Pérez-García https://orcid.org/0000-0002-5450-148X