Vagal paraganglioma: The Jefferson experience

Vagal paraganglioma: The Jefferson experience RANDY B. MILLER, MD, MAURITS S. BOON, MD, JOSEPH P. ATKINS, MD, and LOUIS D. LOWRY, MD, Philadelphia, Pennsylvania

Vagal paraganglioma is a rare tumor of neural crest origin. Although the literature is in agreement with regard to epidemiology, diagnosis, and tumor biology, there is some controversy over treatment modalities for these patients. We performed a nonrandomized retrospective study in a large singleinstitution series of patients (n = 19) in whom vagal paraganglioma was diagnosed. General statistics included age, male/female ratio, tumor size, and duration of follow-up. Other variables such as signs and symptoms at presentation, family history, multicentricity, metastatic disease, and secretion of catecholamines were included. CT scan, MRI, and angiography were used in combination for diagnostic purposes as well as for treatment planning. Preoperative embolization was performed in 5 of the more recently treated patients. Current issues regarding the use of preoperative embolization and choice of surgical approach were analyzed. In this article the possibility and sequela of vagus nerve–sparing procedures will be presented. Operative complications and postoperative morbidity related to cranial neuropathies will be discussed. The rationale for performing adjunct procedures, including cricopharyngeal myotomy and vocal fold medialization, to facilitate the rehabilitation of patients with postoperative cranial nerve deficits will be given. Our findings and recommendations will be compared with currently accepted treatment protocols in conjunction with a review of the literature. (Otolaryngol Head Neck Surg 2000;122: 482-7.)

Vagal paraganglioma is a rare tumor of neural crest origin. Neoplasms in this subset, previously grouped into tumors of the amine precursor uptake and decarboxylaFrom the Department of Otolaryngology–Head and Neck Surgery, Thomas Jefferson University. Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, San Francisco, CA, September 7-10, 1997. Reprint requests: Randy B. Miller, MD, 7200 Almeda, #502, Houston, TX 77054. Copyright © 2000 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/2000/$12.00 + 0 23/1/102807 doi:10.1067/mhn.2000.102807 482

tion (APUD) system, have been defined more recently as neoplasms of the diffuse neuroendocrine system (DNES). Paraganglia or cells of the DNES are found in the aortic and carotid bodies, lung, middle ear, jugular vein in the area of the jugular foramen, vagus nerve, adrenal medulla, and ciliary ganglion of the eye.1 The function of the vagal paraganglionic tissue and the cause of its associated neoplasms are unknown. The term nonchromaffin paraganglioma was formerly used to describe extra-adrenal tumors of neuroectodermal origin. This description was based on a lack of affinity for chromium salts used in the histologic staining process. This reaction has been determined to be nonspecific, and the term has since been abandoned.1 Chemodectoma has also been used to describe tumors of vagal paraganglionic origin. This term incorrectly points to the chemoreceptor function of these cells. Only DNES cells of the aortic and carotid bodies have been shown to have chemoreceptive activity.2 The term chemodectoma is therefore mentioned for historical interest only. Glomus vagale is another term that has been used in the past to describe neoplasms of vagal paraganglion origin. Strictly speaking, a glomus body is a subcutaneous collection of modified smooth muscle elements that are unrelated to cells of the DNES.3 The abovementioned misnomers are included to clarify the currently accepted term, vagal paraganglioma.2 Batsakis1 stated that vagal paragangliomas most commonly arise from the inferior or nodose ganglion of the vagus nerve. It has, however, been noted that these tumors may arise at any point along the course of the vagus nerve. Tumors arising from the middle vagal ganglion often erode into the jugular foramen and may present with otologic manifestations. Paragangliomas of the superior vagal ganglion are often dumbbell shaped (because of a combination of intracranial and extracranial components) and are commonly associated with multiple cranial nerve deficits. It is widely accepted that the primary modality for treatment of a vagal paraganglioma is surgical excision. It is important to note that the major morbidity associated with surgical excision is related to postoperative cranial nerve dysfunction. The vagus nerve is commonly sacrificed during extirpation of these tumors. A postoperative (acute) vagal paralysis is frequently complicated by an associated hypoglossal and/or glossopharyngeal

Otolaryngology– Head and Neck Surgery Volume 122 Number 4

MILLER et al

Table 1. General statistics

Table 2. Tumor biology Data

No. of patients M/F ratio Average age (y) Average tumor size (cm) Follow-up range (y)

483

19 8:11 49 5×4 2.2-15.1

nerve palsy. The issue of postsurgical cranial nerve deficits should be the single most important consideration when formulating a treatment plan for cure or control of a vagal paraganglioma. This article retrospectively analyzes 19 patients treated for vagal paragangliomas. Current recommendations regarding treatment are outlined on the basis of a review of the literature combined with our own institutional experience. The purpose of this article is to further define the safest and most effective way to evaluate and treat patients with vagal paragangliomas. METHODS Between 1977 and 1996, vagal paragangliomas were diagnosed and treated in 19 patients at Thomas Jefferson University Hospital. The charts were identified on the basis of pathologic diagnosis coding. These cases were then reviewed in a retrospective, nonrandomized, and unblinded fashion. Data pertaining to the female/male ratio, age at presentation, tumor size, and duration of follow-up were included. Presenting signs and symptoms, family history, multicentricity, catecholamine secretion, metastatic disease, and tumor-related death were tabulated. Radiographic evaluation, tumor size, and choice of surgical approach were recorded. Morbidity was summarized according to the incidence of carotid artery injury, deficits of cranial nerves X and XII, and associated aspiration. The development of postoperative aspiration was correlated with pre-existing and postoperative (acute) cranial nerve deficits. The efficacy of performing adjunctive procedures, including cricopharyngeal myotomy and vocal fold medialization, for the treatment of postoperative aspiration was evaluated. Comparisons were made between preoperatively embolized and nonembolized patients with respect to blood loss, operative time, and hospital stay. Specifically, the Student t test was used to analyze the association between preoperative embolization and intraoperative blood loss. The same statistical analysis was used to correlate preoperative embolization with the length of hospital stay. Additionally, sparing of the vagus nerve was compared with vagal sacrifice regarding vocal fold paralysis, aspiration, and recurrence. The use of radiotherapy was also recorded.

No. of patients/total

Family history Multicentric Catecholamines Metastatic Cervical nodes Sacrum Death

0/19 2/19 1/19 3/19 2/3 1/3 0/19

Table 3. Presenting signs in order of frequency Sign

Neck mass Pharyngeal fullness Cranial nerve X paralysis Cranial nerve XII paralysis Horner’s syndrome Bruit Hypertension Catecholamines Aspiration Sensorineural hearing loss Nystagmus

No. of patients/total

14/19 10/19 7/19 4/19 2/19 2/19 2/19 1/19 1/19 0/19 0/19

RESULTS

The charts of 19 patients with vagal paraganglioma were examined. Eight patients were male, and 11 were female. Patient ages ranged from 26 to 70 years, with an average of 49 years. The average tumor size (determined radiographically) was 5 × 4 cm. Follow-up ranged from 2.2 to 15.1 years (Table 1). Of our 19 patients, none was found to have a positive family history. Multicentric tumors were present in 2 patients. One patient (normotensive) was found to have an elevated catecholamine level. Metastatic disease was present in 3 patients, including 2 with cervical node metastasis and 1 with distant hematogenous spread to the sacrum. None of our patients died of paragangliomarelated causes (Table 2). Presenting signs on initial physical examination included a neck mass in 14 patients, pharyngeal fullness in 10, vagal nerve paralysis in 7, hypoglossal nerve weakness in 4, Horner’s syndrome in 2, a bruit in 2, hypertension in 2 (with normal catecholamine levels), elevated catecholamine level in 1 (who was asymptomatic with normal blood pressure), and aspiration in 1 patient. No patient had signs of otologic involvement (Table 3). The average duration of symptoms before presentation was 40 months (range 5-72 months). The main presenting symptom was that of a neck mass (14 patients).

484

Otolaryngology– Head and Neck Surgery April 2000

MILLER et al

Table 4. Presenting symptoms in order of frequency Symptom

No. of patients/total

Table 7. Preoperative embolization and complications Method

Neck mass Hoarseness Pharyngeal fullness Dysphagia Dysphonia Pain Cough Aspiration Palpitations Otologic

14/19 7/19 5/19 4/19 4/19 4/19 3/19 1/19 0/19 0/19

No. of patients/total

No. of patients Neurologic complications Vascular injury Postembolization pain

5/16 0/5 0/5 0/5

Table 8. Comparison of factors related to preoperative embolization

The average duration of symptoms was 40 months. Embolized

Table 5. Preoperative radiography Method

No. of patients/total

CT & angiography MRI CT, MRI, & angiography CT

12/19 3/19 2/19 2/19

Table 6. Operative statistics No. of patients/total

Patients Lateral cervical approach Midline mandibulotomy Vocal fold medialization Cricopharyngeal myotomy Tracheotomy

16/19 15/16 1/16 6/16 7/16 9/16

Other symptoms included hoarseness in 7 patients, pharyngeal fullness in 5, dysphagia in 4, dysphonia in 4, cervical pain in 4, cough in 3, and aspiration in 1. None of our 19 patients had otologic symptoms or symptoms consistent with catecholamine excess (Table 4). Preoperative radiographic evaluation included CT scan and angiography in 12 patients; MRI in 3 patients; a combination of CT, MRI, and angiography in 2 patients; and only a CT scan in 2 patients (Table 5). Sixteen patients were treated with surgery. A lateral cervical approach was used in 15 patients. A midline mandibulotomy was used in 1 case. Concurrent vocal fold medialization was performed in 6 patients, cricopharyngeal myotomy in 7, and tracheotomy in 9 (Table 6). Preoperative embolization was used in 5 cases. Of these cases, no patient had neurologic complications, vascular injury, or postembolization pain (Table 7). In the 5 patients who were embolized, the average intraop-

Estimated blood loss (mL) Operating room time (min) Hospital stay (d) Cranial nerve XII paralysis

425 334 7.4 1/5

Nonembolized

990 431 18.3 5/11

erative blood loss was 425 mL, average operative time 334 minutes, and average hospital stay 7.4 days; 1 patient had a postoperative hypoglossal nerve palsy. Of the 11 patients who were not embolized, average intraoperative blood loss was 990 mL, average operative time 431 minutes, and average hospital stay 18.3 days; 5 patients had postoperative hypoglossal nerve paralysis (Table 8). Analysis of intraoperative blood loss with the Student t test revealed a P value of 0.16. The Student t test was also used to analyze the difference in duration of hospital stay (P = 0.18). Carotid artery injury did not occur with either the lateral cervical or midline mandibulotomy approach for surgical exposure (Table 9). Postopertive morbidity included vagal nerve paralysis in 14 patients, hypoglossal paralysis in 9 patients, and Horner’s syndrome in 7 patients. Postoperative hematoma did not develop in any patients. Nine of the 16 operative patients had postoperative aspiration. Of these patients, all 9 had vagus nerve paralysis (2 cases before and 7 cases after surgery), and 8 had hypoglossal nerve paralysis (2 cases before and 6 cases after surgery). Adjunctive procedures (which did not cure postoperative aspiration) included cricopharyngeal myotomy in 7 patients and vocal fold medialization in 5 patients. Only 1 of the 9 patients with postoperative aspiration had aspiration during the preoperative phase of treatment (Table 10). Seven of the 16 operative patients remained free of postoperative aspiration. Of these patients 5 had vagus nerve paralysis (2 cases before and 3 cases after surgery), and 1 had hypoglossal nerve paralysis (after

Otolaryngology– Head and Neck Surgery Volume 122 Number 4

MILLER et al

Table 9. Operative complications: Carotid artery injury related to surgical approach Injury

No. of patients/total

485

Table 11. Absence of postoperative aspiration as related to cranial nerve dysfunction and adjunct procedures No. of patients/total

Lateral cervical excision Midline mandibulotomy

0/15 0/1

Table 10. Postoperative aspiration as related to cranial nerve dysfunction and adjunct procedures No. of patients/total

Aspiration Cranial nerve X paralysis New onset Pre-existing Cranial nerve XII paralysis New onset Pre-existing Cricopharyngeal myotomy Vocal fold medialization Preoperative aspiration Oral nutrition with speech pathology

9/16 9/9 7/9 2/9 8/9 6/9 2/9 7/9 5/9 1/9 3/9

surgery). Adjunctive procedures performed in the nonaspiration group included vocal fold medialization in 1 patient. No patient in this group underwent cricopharyngeal myotomy. Additionally, no patient in this group presented with preoperative aspiration (Table 11). A 1-tailed Fisher exact test was used to analyze the association between new-onset (postoperative) hypoglossal nerve paralysis and aspiration (P = 0.19). During tumor resection, the vagus nerve was spared in 4 of the 16 operative cases. In these 4 cases postoperative vocal fold paralysis and associated aspiration developed in 2 patients, and 2 patients maintained vocal fold mobility without aspiration developing (Table 12). Local recurrence was diagnosed in 3 of our 16 operative cases. Of these cases, 1 patient was in the vagus nerve–spared group, and 2 patients were in the vagus nerve–sacrificed group (Table 13). Radiation therapy was used in 4 of the 19 patients in our series. Three patients had unresectable disease, and 1 patient underwent radiation therapy to a site of sacral metastatic disease. DISCUSSION

Despite the rarity of the vagal paraganglioma, many series have been reported.4-6 The literature reflects a moderately well-established understanding of incidence, tumor biology, and treatment options. The purpose of this discussion was to compare our findings with previously reported data. Additionally, we will make new

No aspiration Cranial nerve X paralysis New onset Pre-existing Cranial nerve XII paralysis New onset Pre-existing Cricopharyngeal myotomy Vocal fold medialization Preoperative aspiration

7/16 5/7 3/7 2/7 1/7 1/7 0/7 0/7 1/7 0/7

Table 12. Recurrence as related to vagus nerve sparing No. of patients/total

No. of patients Vagus saved Vagus sacrificed

3/16 1/4 2/12

Table 13. Function after vagus nerve sparing No. of patients/total

No. of patients Vocal fold paralysis Aspiration Vocal fold function No aspiration

4/16 2/4 2/4 2/4 2/4

treatment recommendations based on our experience in treating patients with vagal paragangliomas. Our male/female ratio of 1:1.4 varies slightly from the 1:2.7 ratio as quoted in the literature.3 The average age at presentation (49 years) and average size of tumor on presentation (5 × 4 cm) correlate well with the corresponding numbers in the literature.7 The most common presenting sign in our series (as well as in previous reports) is that of a painless neck mass behind the angle of the mandible.1 It has also been stated that at presentation at least 50% of patients will have 1 or more cranial nerve palsies, most commonly of the vagus nerve.8 Our series concurs with this statistic, with 11 of 19 patients presenting with cranial nerve deficits (7 vagal and 4 hypoglossal). Because the fibers of the vagus nerve are splayed out over this slowly

486

MILLER et al

growing neoplasm, the body has a chance to develop a poorly understood compensatory mechanism. This gradual adaptation is thought to be the reason for a low incidence of aspiration in patients with a preoperative vagus nerve paralysis.7 Our data further support this theory, with 7 patients having vagus nerve paralysis and only 1 associated case of preoperative aspiration. Interestingly, this patient also presented with a concomitant hypoglossal nerve paralysis, pointing to the observation made by Eriksen et al4 that ipsilateral hypoglossal and vagal deficits result in dramatically ineffective deglutition and a high risk for aspiration. This association will be reiterated later in this discussion. Although the quoted incidence of familial association is approximately 8%,5 none of our patients had a positive family history. Although no conclusions can be drawn from this discrepancy, it is included out of an interest in variability. Multicentricity in the nonfamilial vagal paraganglioma has been reported to be as high as 20%, whereas the percentage in familial forms may be as high as 50%.6 Our numbers correlate well with these findings, with 2 of 19 patients having nonfamilial multicentric disease. The literature points to a 2% to 5% incidence of catecholamine-secreting or functioning paragangliomas.9 Our series identified 1 patient as having elevated catecholamine levels, correlating well with the quoted percentages. Interestingly, this patient was normotensive despite his elevated norepinephrine levels. Patients must have a 4- to 5-fold elevation in serum norepinephrine levels to produce detectable signs and symptoms.10 The discovery of elevated serum catecholamine levels requires a workup to rule out a pheochromocytoma and/or multicentric functioning tumors. A preoperative α-receptor blockade is required in these situations. Vagal paragangliomas have a higher propensity to metastasize than other neoplasms of the DNES, with an incidence that approaches 20%. Both hematogenous and lymphatic spread are possible.11 In our series regional spread to cervical lymphatics (2 patients) and hematogenous metastasis to the sacrum (1 patient) were noted. In formulation of a surgical plan, Biller et al5 emphasized that the chance of operative injury to the carotid artery increases as the size of the vagal paraganglioma increases. They stated that tumors 5 cm or greater in diameter commonly result in carotid artery displacement parallel to the skull base. In these cases, they recommend excision with a midline mandibulotomy, which allows greater exposure and a decreased risk of carotid artery injury.5 The average size of tumor at presentation in our series was 5 × 4 cm. Fifteen of the 16

Otolaryngology– Head and Neck Surgery April 2000

excisions were performed through a lateral cervical approach. Despite the relatively large size of these tumors, the lateral approach (without mandibulotomy) provided adequate exposure and zero incidence of carotid artery injury. We do believe that a midline mandibulotomy improves exposure at the skull base and that the surgeon should plan on the possibility of using this approach. However, in most cases adequate exposure and safe excision can be achieved with a lateral cervical approach (even in cases with a tumor diameter of 5 cm or greater), as demonstrated in our series. Several studies have concluded that there is no surgical advantage to performing preoperative embolization.12-14 Urquhart et al6 emphasized the possibility of neurologic sequelae caused by cerebral artery occlusion and that this factor greatly outweighs the benefits derived from decreased tumor vascularity. In our study, 5 of the 16 more recently treated patients underwent preoperative embolization. This was associated with decreased intraoperative blood loss, decreased operative time, decreased hospital stay, and decreased incidence of postoperative hypoglossal nerve paralysis. None of our patients had neurologic complications. Although these numbers did not reach statistical significance in the 95% CI, they certainly did trend toward strong clinical significance. As noted previously, minimizing the possibility of a new-onset (postoperative) hypoglossal nerve paralysis greatly facilitates effective deglutition. In addition, adequate hypoglossal nerve function helps to decrease postoperative aspiration in patients with an immobilized hemilarynx caused by vagal nerve paralysis. Therefore this study strongly supports the use of preoperative embolization performed by an experienced interventional neuroradiologist, in preparation for surgical excision of these neoplasms. Of the expected morbidities associated with excision of a vagal paraganglioma, aspiration may be the most debilitating. In a recent report, Biller et al5 stated that although cricopharyngeal myotomy and vocal fold medialization may not restore normal deglutition, these procedures will allow the patient to maintain oral nutrition without aspirating. Of our 16 surgical cases, a subset of 9 patients experienced postoperative aspiration. In this group, 7 cricopharyngeal myotomies and 5 vocal fold medialization procedures were performed. Our data suggest that in many cases vocal fold medialization and cricopharyngeal myotomy will not prevent postoperative aspiration. Aside from vagus nerve paralysis, the most common morbidity was new-onset postoperative hypoglossal nerve palsy. Interestingly, of the 7 patients who did not aspirate after surgery, only 1 underwent concomitant vocal fold

Otolaryngology– Head and Neck Surgery Volume 122 Number 4

medialization, and no patient underwent cricopharyngeal myotomy. In this group only 1 of the 7 patients had postoperative hypoglossal nerve paralysis. From these data, we have concluded that a cricopharyngeal myotomy and vocal fold medialization are not essential in the prevention of aspiration. Moreover, the most common scenario for development of aspiration was the presence of a postoperative hypoglossal nerve palsy combined with an ipsilateral vagus nerve paralysis. Historically, recommendations for surgical treatment of vagal paragangliomas have universally included en bloc excision of the involved portion of the vagus nerve. To our knowledge, this is the first report describing a vagus nerve–sparing procedure. In our series, preservation of the vagus nerve was associated with postoperative function and prevention of aspiration in 50% of cases. In addition, this procedure was not associated with an increased risk of recurrence. CONCLUSIONS

1. Although the patient and surgeon should be prepared for a midline mandibulotomy, many tumors measuring 5 cm (or greater) may be safely excised with adequate exposure by using a lateral transcervical approach without mandibulotomy. 2. In the proper hands, preoperative embolization is safe and is recommended. It may result in decreased blood loss, decreased hospital stay, decreased hypoglossal nerve paralysis, and an associated decrease in postoperative aspiration. 3. Cricopharyngeal myotomy and vocal fold medialization do not prevent aspiration. This is particularly true in patients with concomitant ipsilateral cranial nerve X and XII deficits. 4. The most common scenario associated with the development of aspiration is a vagus nerve palsy

MILLER et al

487

combined with an ipsilateral postoperative (acute) hypoglossal nerve palsy. 5. Vagus nerve sacrifice is not mandatory in all cases. In fact, preservation may result in a functional vagus nerve after surgery. This treatment option (series of 4) has not been associated with an increased risk of tumor recurrence. REFERENCES 1. Batsakis JG. Tumors of the head and neck: clinical and pathological considerations. 2nd ed. Baltimore: Williams & Wilkins; 1979. p. 369-80. 2. Glenner GC, Grumley PM. Tumors of the extra-adrenal paraganglion system (including chemoreceptors). Atlas of tumor pathology. Washington (DC): Armed Forces Institute of Pathology; 1974. Series 2, Fascicle 9. 3. Wetmore RF, Tronzo RD, Lane RJ, et al. Nonfunctional paraganglioma of the larynx: clinical and pathological considerations. Cancer 1981;48:2717-23. 4. Eriksen C, Girdhar-Gopal H, Lowry LD. Vagal paragangliomas: a report of nine cases. Am J Otolaryngol 1991;12:278-87. 5. Biller HP, Lawson W, Som P, et al. Glomus vagale tumors. Ann Otol Rhinol Laryngol 1989;98:21-6. 6. Urquhart AC, Johnson JT, Myers EN, et al. Glomus vagale: paraganglioma of the vagus nerve. Laryngoscope 1994;104:440-5. 7. Moore G, Yarington CT, Mangham GA. Vagal body tumors: diagnosis and treatment. Laryngoscope 1986;96:533-6. 8. Lawson W. Glomus bodies and tumors. NY State J Med 1980; 80:1567-75. 9. Schwaber MK, Glasscock ME, Jackson CG. Diagnosis and management of catecholamine secreting glomus tumors. Laryngoscope 1984;94:1008-15. 10. Gulya JA. The glomus tumor and its biology. Laryngoscope 1993;103(Suppl 60):7-15. 11. Heinrich MC, Harris AE, Bell WR. Metastatic intravagal paraganglioma: case report and review of the literature. Am J Med 1985;78:1017-24. 12. Forbes G, Earnest FW, Jackson IT. Therapeutic embolization angiography for extra-axial lesions in the head. Mayo Clin Proc 1986;61:427-41. 13. Fisch U. Infratemporal fossa approach for glomus tumors of the temporal bone. Ann Otol Rhinol Laryngol 1982;91:474-9. 14. Kumar AJ, Kaufman SL, Patt J. Preoperative embolization of hypervascular head and neck neoplasms using microfibrillar collagen. AJNR Am J Neuroradiol 1982;3:163-6.