Multimodality management of large carotid-space paragangliomas: A single-center case series

Abstract

Background:

Large carotid-space paragangliomas (LCSPs) are uncommon, highly vascular tumors that present significant operative challenges because of their intimate relationship to the carotid circulation, lower cranial nerves, and skull base. The central hazard in surgical management is not tumor size alone but uncertainty regarding vascular anatomy, collateral circulation, and operative conditions before incision. We describe a structured multimodality strategy to reduce this uncertainty before critical dissection.

Case Description:

A retrospective review was performed of eight consecutive patients treated for LCSPs using a standardized multidisciplinary workflow involving neurovascular surgery, head and neck surgery, and neurointerventional surgery. Five tumors were carotid body tumors, and three were glomus vagale tumors (GVTs). Mean age was 55 years and mean maximum tumor diameter was 6.1 cm. Multifocal disease occurred in two patients (25%); one tumor was biochemically active. All patients underwent diagnostic angiography and preoperative embolization. Balloon test occlusion was performed in seven patients; six passed. Gross total resection was achieved in six patients (75%). Intentional subtotal resection was performed in two GVTs due to skull base extension. One patient subsequently underwent stereotactic radiosurgery for residual disease. Median estimated blood loss was 100 mL (mean 122 mL; range 25–350 mL). No patient required transfusion, had perioperative stroke, carotid sacrifice, or major vascular complication. Two patients developed new postoperative vagal/recurrent laryngeal nerve dysfunction at 6-week follow-up. Median length of stay was 2 nights.

Conclusion:

A structured multimodality strategy emphasizing preoperative uncertainty reduction was associated with favorable outcomes in our case series. Surgical endpoints were determined by anatomy rather than dogmatic pursuit of complete resection, providing a reproducible framework for LCSP management.

INTRODUCTION

Paragangliomas are rare neuroendocrine tumors arising from neural crest-derived chromaffin cells within the autonomic nervous system, with an estimated incidence of 1–8 cases/million annually.[9] Paragangliomas arising within the carotid space represent some of the most technically demanding lesions encountered in head and neck surgery. Although typically benign, these tumors are frequently large, highly vascular, and intimately associated with the carotid circulation, internal jugular vein (IJV), and lower cranial nerves (LCNs). As tumor size increases, operative risk rises accordingly. Hemorrhage can rapidly obscure critical anatomy, vascular injury may require immediate repair or sacrifice decisions, and cranial nerve (CN) morbidity is often determined by the quality of dissection achieved around densely adherent neurovascular structures. Unlike many oncologic operations, large carotid-space paraganglioma (LCSP) resection is often less forgiving of uncertainty.[2,8]

Carotid body tumors (CBTs) and glomus vagale tumors (GVTs) comprise the majority of carotid-space paragangliomas encountered in clinical practice. CBTs arise at the carotid bifurcation and are commonly classified according to the Shamblin grading system, which reflects the degree of carotid vessel involvement. GVTs arise from paraganglionic tissue associated with the vagus nerve and frequently extend superiorly toward the skull base. While both lesions share a hypervascular nature, the principal challenge in large tumors is not simply tumor removal but safe management of a lesion that may involve long segments of the carotid circulation, distort normal anatomy, and extend beyond areas of reliable visualization.[3,4]

Successful treatment therefore depends on more than operative technique alone. Before incision, the surgeon must understand the relationship of the tumor to the carotid arteries and IJV, identify the dominant arterial supply, determine whether collateral circulation is sufficient to tolerate carotid injury or sacrifice, and anticipate the anatomical limitations that may be encountered at the skull base. Without this information, an operation that appears controlled at the outset may become high-risk rapidly when bleeding obscures anatomy or unexpected vascular involvement is encountered.[3,4]

A multimodality treatment strategy offers an opportunity to reduce this uncertainty before surgery begins. Advanced imaging can define anatomical relationships and tumor extent. Diagnostic angiography characterizes vascular supply, vessel integrity, and collateral circulation. Balloon test occlusion (BTO) provides contingency planning should carotid injury occur during resection.[6] Preoperative embolization may substantially reduce tumor vascularity and transform the operative field from a pulsatile, hemorrhagic mass into a more controlled, devascularized lesion. When combined with coordinated multidisciplinary surgery, these techniques have the potential to improve operative safety and reproducibility.

At our institution, repeated management of LCSP led to the development of a structured neurovascular workflow incorporating advanced imaging, diagnostic angiography, BTO, preoperative embolization, multidisciplinary surgical resection, and selective use of stereotactic radiosurgery. In this study, we review our experience with this treatment paradigm, evaluate operative outcomes, and describe a structured decision-making framework for multimodality management and maximal safe resection of LCSP.

METHODS

Study design and patient selection

After the Institutional Review Board exemption was granted by Wellstar Health System, a retrospective review was performed of consecutive patients who underwent treatment for LCSP at a single institution. All patients were managed using a standardized multidisciplinary treatment protocol involving neurovascular surgery, head and neck surgery, and neurointerventional surgery, with additional input from radiology, pathology, and radiation oncology as clinically indicated [Figure 1].

Figure 1:

Institutional management algorithm for large carotid-space paragangliomas. The treatment strategy integrates advanced imaging, biochemical testing, diagnostic angiography, balloon test occlusion, preoperative embolization, and multidisciplinary surgical resection, magnetic resonance imaging surveillance, and radiosurgery if progression. Surgical endpoints are determined by anatomy and visualization rather than tumor classification alone.

Patient demographics, imaging characteristics, procedural details, operative reports, pathology results, and postoperative outcomes were reviewed. Variables collected included age, sex, presenting symptoms, tumor location, tumor dimensions, multifocal disease status, angiographic findings, BTO results, embolization technique, extent of resection, estimated blood loss (EBL), perioperative complications, CN outcomes, stroke, transfusion requirements, and length of hospital stay.

For the purposes of this review, a LCSP was defined as a lesion measuring ≥4 cm in maximum diameter on preoperative imaging or pathological assessment. Although there is currently no universally accepted size threshold defining a large paraganglioma, a cutoff of 4 cm was selected because tumors of this size are more frequently associated with increased operative complexity, greater neurovascular involvement, and higher rates of CN morbidity reported in the literature.

Preoperative imaging evaluation

All patients underwent contrast-enhanced computed tomography (CT) imaging of the neck. Magnetic resonance imaging (MRI) of the neck with and without contrast became part of the routine institutional evaluation during the study period and was obtained when available. Imaging was used to evaluate tumor size, vascular relationships, carotid displacement or encasement, skull base extension, airway deviation or compression, IJV involvement, and the presence of multifocal disease.

Patients were considered for surgical intervention based on a combination of symptom burden, tumor size, anatomical considerations, growth characteristics, and functional status. Particular concern was given to lesions demonstrating substantial carotid involvement, progressive enlargement, airway deviation, or significant mass effect.

Routine biochemical evaluation with plasma free or urinary fractionated metanephrines was performed before any intervention to identify catecholamine-secreting tumors to avoid a hypertensive crisis.

Diagnostic angiography and Balloon Test Occlusion

All patients underwent diagnostic cerebral angiography to characterize tumor vascular supply and evaluate the surrounding cerebrovascular anatomy. Particular attention was directed toward identifying arterial feeders, assessing carotid artery integrity, evaluating venous patency, and defining collateral circulation through the Circle of Willis.

BTO became a near-routine component of the institutional workflow for LCSP and was performed whenever feasible. The purpose of BTO was not to plan carotid sacrifice, but rather to provide contingency planning should vascular injury occur during surgical resection.

BTO was performed under general anesthesia with continuous electroencephalographic and somatosensory evoked potential monitoring. General anesthesia was favored because angiography, BTO, and embolization were frequently performed during a single procedural session, allowing prolonged evaluation while minimizing patient discomfort and movement. Patients were systemically heparinized to maintain an activated clotting time >250 s. A compliant balloon was positioned within the internal carotid artery (ICA) and inflated until complete arrest of antegrade flow was confirmed angiographically.

Neuromonitoring was assessed at 1-min intervals throughout the procedure. Control angiography was performed every 2 min to confirm persistent occlusion. After ten minutes of stable occlusion, mean arterial pressure was reduced by approximately 20% to create a physiologic stress challenge. Occlusion was then maintained for an additional 10 min. Any neuromonitoring change prompted immediate balloon deflation and restoration of baseline blood pressure. Failure of the test was defined by any reproducible neuromonitoring abnormality during the occlusion period.

BTO results were incorporated into operative planning and provided information regarding management options should carotid injury occur.

Endovascular embolization strategy

Preoperative embolization was performed in all patients, typically 1–2 days before surgical resection. This timing allowed confirmation of procedural stability before surgery while preserving the benefits of embolization and minimizing the potential for collateral reconstitution. Separating embolization from resection also permitted scheduling of the open procedure as a first-start operation with a fully rested operative team.

The primary goal of embolization was not complete tumor infarction, but rather reduction of tumor perfusion sufficient to transform a pulsatile, hemorrhagic lesion into a more controlled operative target. Dominant arterial feeders were identified angiographically and selectively catheterized. Onyx was used as the primary embolic agent in most cases because of its visibility and ability to achieve durable feeder occlusion. In selected large tumors, microparticles were used before Onyx administration to penetrate the distal tumor bed. When vertebral artery contributions were present, coil embolization was utilized selectively to eliminate feeder inflow while minimizing risk to the posterior circulation.

The ascending pharyngeal artery represented the dominant arterial supply in most lesions. Embolization focused on tumor-feeding branches while preserving skull base branches whenever possible to minimize the risk of ischemic cranial neuropathy.

Multidisciplinary surgical technique

All resections were performed jointly by neurovascular and head and neck surgeons. Head and neck surgery performed cervical exposure and facilitated identification of critical neurovascular structures. Tumor mobilization, vascular dissection, and skull base exposure were performed collaboratively.

The operative strategy emphasized early vascular control. Vessel loops were routinely placed around the common carotid artery, ICA, external carotid artery, and hypoglossal nerve when anatomically feasible. Carotid repair instruments, vascular clamps, and intraoperative ultrasound were immediately available throughout the procedure.

Two attending surgeons were simultaneously engaged during the critical portions of tumor mobilization and vascular dissection. Continuous intraoperative communication regarding anatomy, plane development, and vascular proximity was considered an intentional component of risk mitigation rather than simply a staffing arrangement.

When tumors demonstrated dense adherence to surrounding structures, subcapsular dissection was used selectively to develop a safe plane between tumor and vessel wall while minimizing manipulation of the ICA, IJV, and LCNs.

Resection philosophy

The operative objective was maximal safe resection, and the intended surgical endpoint differed according to anatomy and visualization rather than tumor classification alone.

For CBTs, gross total resection (GTR) was pursued when safely achievable because vascular anatomy and tumor margins could generally be visualized directly throughout the dissection.

For GVTs, superior extension toward the skull base frequently limited direct visualization. When complete visualization could be safely maintained, GTR was performed. However, when dissection progressed beyond a point of reliable visualization and became dependent primarily on palpation near the skull base, residual tumor was intentionally left behind in favor of a maximal safe resection strategy. Management of residual disease was individualized and included surveillance or stereotactic radiosurgery based on patient age, disease burden, anatomical considerations, and multidisciplinary consensus.

Postoperative management

Patients were routinely admitted to a neurocritical care unit following surgery for close neurological and airway monitoring. Closed-suction drains were left in place because of concern for postoperative neck hematoma and were removed when output became minimal.

Physical therapy, speech-language pathology evaluation, and swallowing assessment were obtained when clinically indicated. Patients demonstrating postoperative cranial neuropathy were followed collaboratively with otolaryngology. Discharge was typically to home once neurological status, swallowing function, and pain control were stable. Final clinical outcomes, including the presence of cranial neuropathies and other postoperative sequelae, were determined during the routine 6-week postoperative head and neck surgery outpatient follow-up. Extent of resection was determined from the operative report, supplemented by postoperative imaging when available; routine postoperative imaging was not available for all patients.

Outcomes and statistical analysis

Primary outcomes included extent of resection, EBL, transfusion requirement, stroke, new cranial neuropathy, and length of hospital stay. Secondary outcomes included angiographic findings, BTO results, embolization technique, and postoperative management requirements.

Given the small cohort size, statistical analysis was limited to descriptive measures including means, medians, ranges, and percentages.

RESULTS

Patient and tumor characteristics

Eight patients underwent multimodality treatment for LCSP during the study period. The cohort included six women and two men with a mean age of 55 years (range 21–76 years). Five tumors were CBTs and three were GVTs. Two patients (25%) demonstrated multifocal paraganglioma disease. Patient demographics, tumor characteristics, and presenting symptoms are summarized in Table 1.

Table 1:

Patient and tumor characteristics.

Tumor size was substantial throughout the cohort. Maximum tumor diameter ranged from 4.1 cm to 9.2 cm, with a mean diameter of 6.1 cm and median diameter of 5.8 cm. CBTs arose at the carotid bifurcation, whereas GVTs occupied the parapharyngeal space with varying degrees of superior skull base extension.

Presenting symptoms commonly reflected mass effect, vascular pulsatility, and LCN involvement, including palpable neck mass, dysphagia, cough, dysphonia, aspiration, pulsatile tinnitus, headache, and vocal cord dysfunction.

Preoperative evaluation

All patients underwent contrast-enhanced CT imaging of the neck and diagnostic cerebral angiography. MRI of the neck with and without contrast was obtained in six patients (75%). A representative LCSP demonstrating substantial mass effect and complex neurovascular relationships is shown in Figure 2.

Figure 2:

Contrast-enhanced coronal computed tomography demonstrating a large carotid-space paraganglioma with substantial mass effect and intimate relationship to the carotid circulation and skull base. The lesion displaces adjacent neurovascular structures and illustrates the anatomical complexity that motivates the multimodality treatment strategy described in this series.

In addition, biochemical testing for plasma metanephrines was performed in all patients with one patient (12.5%) demonstrating a biochemically active secretory lesion associated with episodic flushing. This patient underwent appropriate preoperative α-adrenergic blockade before surgical intervention without perioperative hemodynamic complications. Diagnostic angiography demonstrated a hypervascular tumor blush in all cases and identified dominant arterial supply patterns that informed embolization planning. BTO was performed in seven patients (87.5%). One patient underwent angiography and embolization before routine incorporation of BTO into the institutional workflow.

Six patients successfully passed BTO. One patient failed BTO without clinical sequelae, reinforcing the importance of carotid preservation during subsequent surgical resection and influencing operative risk tolerance during carotid dissection.

Endovascular embolization

Preoperative embolization was performed in all eight patients. Onyx was utilized as the primary embolic agent in seven cases. One patient underwent combined microparticle and Onyx embolization because of extensive tumor vascularity.

Embolization was technically successful in all cases and was completed without procedural stroke, vascular injury, or other embolization-related complications. Representative angiographic findings and embolization techniques are shown in Figure 3.

Figure 3:

Diagnostic angiography and embolization of a large carotid-space paraganglioma. (A) Preembolization angiography demonstrates dense tumor vascularity and prominent tumor blush. (B) Selective microcatheter angiography defines tumor angioarchitecture and dominant arterial feeders. (C) Fluoroscopic image demonstrating Onyx embolization of the feeding vasculature. (D) Postembolization angiography demonstrates marked reduction in tumor perfusion and vascularity, facilitating subsequent surgical resection.

Surgical resection

All patients subsequently underwent surgical resection through a combined neurovascular and head and neck surgical approach.

GTR was achieved in six patients (75%). All five CBTs underwent GTR. Among GVTs, GTR was achieved in one patient in whom complete visualization could be safely maintained to the skull base. Subtotal resection was intentionally performed in the remaining two GVTs because of superior extension to the skull base, where further dissection was limited by visualization, carotid artery proximity, and LCN risk.

Pathological examination confirmed paraganglioma in all eight cases.

One patient (12.5%) underwent CyberKnife stereotactic radiosurgery (25 Gray in 5 fractions) for residual GVT following intentional subtotal resection and for a contralateral CBT as part of management of multifocal disease. No other patients underwent adjuvant radiation therapy during the study period. In the remaining patient with multifocal disease, the unresected lesion was managed with surveillance, with radiation therapy reserved for future progression or clinical indication.

Operative and clinical outcomes

Despite substantial tumor size and complex neurovascular anatomy, operative outcomes were favorable throughout the case series.

Median EBL was 100 mL (mean 122 mL; range 25– 350 mL). No patient required blood transfusion. No patient experienced perioperative stroke, carotid artery sacrifice, or major vascular complication.

Two patients (25%) developed new postoperative vagal/recurrent laryngeal nerve (RLN) dysfunction, both of which remained present at the approximately 6-week postoperative assessment. No additional cranial neuropathies were present at the 6-week follow-up, including first-bite syndrome, Horner syndrome, or hypoglossal and marginal mandibular nerve deficits.

Median length of hospital stay was 2 nights (range 1–8 nights). The longest hospitalization occurred in a patient with severe preexisting dysphagia. Placement of a percutaneous endoscopic gastrostomy tube was performed to address this baseline swallowing impairment and was not related to postoperative neurological deterioration. No patient experienced worsening neurological function attributable to stroke or major vascular injury.

Overall, the multimodality treatment strategy was associated with GTR in six of eight patients, low EBL, absence of transfusion requirement, and no major cerebrovascular complications despite the large size and anatomical complexity of the tumors treated. Operative details and clinical outcomes are summarized in Table 2.

Table 2:

Treatment workflow and clinical outcomes.

DISCUSSION

The central hazard in LCSP surgery is not the tumor itself but the uncertainty surrounding it. A surgeon who does not know whether the carotid artery is displaced or invaded, which vessels supply the tumor, whether the patient can tolerate carotid sacrifice, and how the operative field will behave when dissection begins is operating with a significant informational deficit. That deficit becomes dangerous when bleeding obscures anatomy, vascular injury demands an immediate decision, or skull base extension limits visualization.[3,4] In this setting, the relevant question is not simply how to remove the tumor, but how to reduce uncertainty before the critical dissection begins.

The workflow described in this series was designed around that principle. Advanced imaging defined tumor extent and anatomical relationships. Routine preoperative biochemical screening is vital, as unrecognized catecholamine secretion may result in significant perioperative hemodynamic instability. In patients with confirmed secretory tumors, preoperative α-adrenergic blockade, with β-blockade added only after adequate α-blockade when indicated, remains the standard approach to optimize anesthetic and surgical safety.[3] Diagnostic angiography characterized vascular supply, carotid integrity, venous patency, and collateral circulation. BTO provided contingency planning – not because carotid sacrifice was intended, but because knowing the answer before surgery altered operative decision-making if vascular injury occurred. Preoperative embolization reduced tumor vascularity and transformed a pulsatile, hemorrhagic lesion into a more controlled operative target. Multidisciplinary surgical execution ensured that two experienced surgeons shared a common understanding of the anatomy before and throughout the dissection. Each component addressed a distinct category of uncertainty. By the time resection began, the questions most likely to become dangerous intraoperatively had already been answered.

BTO deserves particular emphasis as a contingency-planning tool rather than simply a preoperative screening study.[1] The risk of carotid injury during resection of LCSP is low but real, particularly in lesions demonstrating extensive vascular adherence or skull base extension. Although carotid sacrifice was not anticipated in any patient in this series, knowledge of a patient’s ability – or inability – to tolerate carotid occlusion informed operative planning and preparedness for a potential vascular complication. In the present cohort, BTO results did not directly determine the extent of resection; both subtotal resections occurred in patients who passed BTO and were driven by skull base extension and limited visualization. Rather, BTO provided information regarding potential management should carotid injury or sacrifice become necessary during resection. In our patient who failed BTO, the planned extent of resection was unchanged, as carotid sacrifice was not ultimately required.

Preoperative embolization represented another important component of the treatment strategy.[7] The objective was not complete tumor infarction but reduction of vascularity sufficient to improve operative conditions. Intraoperatively, embolized tumors frequently appeared more rubbery, devascularized, and less hemorrhagic than would otherwise be expected. This observation supports the practical value of embolization in reducing the likelihood that bleeding will obscure anatomy and force critical decisions under suboptimal conditions. Rather than simply reducing blood loss, embolization changed the character of the operation itself in our experience. This aligns with the recent institutional experience reported by Keles et al. in the surgical management of CBTs.[5] Importantly, no embolization-related stroke, vascular injury, or procedural complication occurred in this series.

The multidisciplinary nature of treatment also warrants emphasis. Resection of LCSPs frequently requires simultaneous consideration of vascular anatomy, skull base exposure, CN preservation, and contingency planning for vascular injury. Continuous collaboration between neurovascular and head and neck surgeons provided complementary expertise and allowed real-time confirmation of anatomy throughout the operation. Two experienced surgeons actively engaged in the critical portions of the dissection functioned as a deliberate risk-mitigation strategy rather than a staffing convenience.

Perhaps the most important benefit of a structured workflow is not technical but cognitive. When vascular anatomy, collateral circulation, embolization strategy, and contingency planning have been addressed before incision, the operative team is free to focus its attention on anatomy and execution rather than crisis management. Standardization reduces distraction, simplifies decision-making, and improves the ability to respond effectively when unexpected challenges arise.

The experience with GVTs in this series illustrates the importance of visualization-driven decision-making. All CBTs underwent GTR. Among GVTs, one lesion underwent GTR because complete visualization could be safely maintained to the skull base, whereas two underwent intentional subtotal resection because superior extension limited reliable visualization near the skull base. This distinction reflects anatomy and visualization rather than surgical ambition. The operative objective was maximal safe resection, not GTR at any cost. When visualization remained adequate, complete removal was achievable. When visualization deteriorated and risk increased, subtotal resection represented the safer endpoint. These findings emphasize that the extent of resection was dictated by anatomical constraints rather than surgical intent, with maximal safe resection prioritized over aggressive pursuit of GTR. One patient in this series subsequently underwent CyberKnife stereotactic radiosurgery approximately 3 months after surgery for residual disease following intentional subtotal resection, illustrating how surgery and radiosurgery can function as complementary modalities when complete resection would require disproportionate operative risk.

Despite substantial tumor size, outcomes were favorable throughout the cohort. GTR was achieved in six of eight patients. No patient required transfusion, no patient experienced stroke, and no major vascular complications occurred. CN morbidity was limited to vagal RLN dysfunction in two patients. Although the sample size is small, these findings suggest that systematic preoperative planning can produce controlled operative conditions even in anatomically complex tumors.

This study has several limitations. The cohort is small and reflects the experience of a single institution. The retrospective design introduces the potential for selection bias and limits statistical analysis. Clinical outcomes were assessed at approximately 6 weeks after surgery, and longer-term clinical follow-up was not systematically collected for this retrospective analysis. Accordingly, cranial neuropathies persisting at that assessment cannot be considered definitively permanent. Postoperative surveillance imaging was also heterogeneous and was not available at standardized intervals for all patients; therefore, operative assessments of gross versus subtotal resection were not uniformly confirmed radiographically, and this study was not designed to evaluate long-term radiographic progression or tumor control. In addition, the structured workflow evolved during the study period, resulting in minor variations in preoperative evaluation among earlier cases. Genetic counseling and germline testing were not routinely available in this retrospective cohort and were therefore not included in the present analysis. Nevertheless, the consistency of the treatment philosophy, operative strategy, and outcomes provides a useful framework for management of these uncommon lesions.

CONCLUSION

LCSPs present substantial technical challenges because of their vascularity, neurovascular proximity, and frequent extension toward the skull base. A structured multimodality treatment strategy incorporating advanced imaging, diagnostic angiography, BTO, preoperative embolization, and multidisciplinary surgical resection provided a structured framework for operative planning and facilitated safe tumor resection.

In this series, the approach was associated with GTR in six of eight patients, low EBL, absence of transfusion requirement, and no major cerebrovascular complications. Most importantly, surgical endpoints were determined by anatomy and visualization rather than dogmatic pursuit of complete resection. When complete resection could not be safely achieved, residual disease could be effectively incorporated into a multimodality treatment strategy including stereotactic radiosurgery. This workflow provides a practical framework for management of LCSPs and may facilitate the safety and reproducibility of treatment in appropriately selected patients.

Ethical approval:

The Institutional Review Board has waived the ethical approval for this study.

Declaration of patient consent:

After the Institutional Review Board exemption was granted by Wellstar Health System, a retrospective review was performed of consecutive patients who underwent treatment for LCSP at a single institution.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.