Surgical resection of cervical schwannoma and paraganglioma: Speech and swallowing outcomes

Abstract

We conducted a retrospective study (1999 to 2009) at our tertiary care institution to evaluate speech and swallowing outcomes after the resection of cervical schwannoma or paraganglioma. Of 6 patients treated for schwannoma, 5 (83.3%) had immediate dysphonia and dysphagia. All patients with deficits received primary reinnervation (n = 2) or subsequent medialization laryngoplasty (n = 3). At 6 months, 4 patients (66.6%) still had dysphonia and dysphagia. At final follow-up (median: 10 months; range: 8 to 12 months), 4 patients (66.7%) had dysphonia and 2 (33.3%) had dysphagia. Of 10 patients treated for paraganglioma, 6 (60.0%) had immediate dysphonia and dysphagia. Four patients received subsequent medialization laryngoplasty; none had primary reinnervation. At 6 months, 3 (30%) still had dysphonia and dysphagia. At final follow-up (median: 15.5 months; range: 1.25 to 48 months), 2 (20.0%) had dysphonia and dysphagia. All patients with deficits received speech and swallowing therapy. We conclude that cervical schwannoma and paraganglioma resection was associated with high rates of immediate postoperative dysphonia and dysphagia. Schwannoma had higher initial rates and poorer recovery. Primary and/or subsequent laryngeal procedures combined with therapy led to symptom resolution in some patients.

Introduction

Extracranial neurogenic tumors of the head and neck are rare and are frequently intimately involved with critical neurovascular structures, especially when located within the parapharyngeal space. Cervical schwannoma and paraganglioma are two such tumors. Their often intimate involvement with nerves that are either exiting the cranial base or traversing the deep spaces of the neck presents unique challenges to the extirpative otolaryngologist–head and neck surgeon.

Extracranial schwannomas present in the head and neck region in 25 to 45% of cases and can grow along peripheral, cranial, or autonomic nerves.¹ They arise from peripheral myelinating cells, or Schwann cells, and are classified by their site of origin.¹

Management options for cervical schwannoma include observation, radiotherapy, and surgical resection. While radiation or radiosurgery is often employed for vestibular schwannoma, it is problematic to apply vestibular schwannoma radiation data to the treatment of extracranial tumors.²

Paragangliomas are benign neuroendocrine tumors derived from paraganglia, or neuroectoderm-derived chromaffin cells located outside the adrenal gland. While most paraganglia degenerate after birth,³ they can persist along branchiomeric, intravagal, aorticosympathetic, and visceroautonomic sites.

Management of paragangliomas can include observation, radiation therapy, stereotactic radiosurgery, and surgical excision. Chemotherapy is primarily used in the setting of metastatic or malignant paragangliomas, and a complete response for some foci has been shown.⁴ Surgery or radiation results in control rates of about 90%.⁵ Radiation is often recommended for patients who have more extensive tumors or who are poor surgical candidates, including patients with significant comorbidity, preexisting neuropathies, or bilateral tumors that might recover poorly from a procedure complicated by new neuropathies.^6,7 Stereotactic radiosurgery has also been used for paraganglioma, but most reports are for intratemporal tumors.⁵

While it has been shown that there are multiple management options for cervical schwannoma or paraganglioma, surgical resection remains the treatment of choice for most patients. Dissection of associated nerves, especially cranial nerves (CN) IX, X, and XII, may result in speech and swallowing deficits. The resulting postoperative dysphonia and dysphagia often require speech and swallowing therapy and, in some patients, procedures to minimize aspiration and to improve quality of life.

Although cervical schwannoma and paraganglioma have similar potential for morbidity after treatment, data comparing outcomes after resection of both tumors are limited. We sought to review our experience with cervical schwannoma and paraganglioma by evaluating speech and swallowing outcomes and the extent of subsequent management for postoperative sequelae. Specifically, we aimed to determine the rate of immediate postoperative dysfunction, rate of recovery, interventions required for rehabilitation, and potential differences between the two tumors.

Patients and methods

Approval from the University of Minnesota Institutional Review Board was obtained.

Pathologic diagnoses of cervical schwannoma or paraganglioma between 2003 and 2009 at a tertiary care academic institution were identified through a pathology database initiated in 2003. Additional queries by ICD-9 codes were performed from 1999–2003 to include an overall study period of 10 years. Patients were included if they had the pathologic diagnosis of cervical schwannoma or paraganglioma. Patients were excluded if tumors traversed the cranial base, if the patients had incomplete assessments of speech or swallowing function, or if they had postoperative speech and/or swallowing dysfunction without at least 6 months of follow-up.

Patient charts were reviewed for demographics, presenting signs and symptoms, radiologic evaluations, and surgical approaches, as well as intraoperative findings, postoperative speech and swallowing outcomes, workup of dysphonia and/or dysphagia, and treatments used for such dysfunction. Dysphonia and dysphagia were noted immediately postoperatively, at 6 months, and at the final follow-up appointment for each patient.

Evaluation of dysphonia and dysphagia was done by history and physical examination, including flexible laryngoscopy to evaluate vocal fold and arytenoid mobility. Dysphonia and dysphagia were screened for by qualitative patient assessment and clinical report. Validated self-reporting instruments and stroboscopy were not used.

Patients with signs and/or symptoms of aspiration were further evaluated with sensation testing and a modified barium swallow study (MBSS). Sensation was tested by documenting gag reflex through oropharyngeal digital palpation, and cough reflex through contacting each side of the unanesthetized larynx with a flexible laryngoscope. Diet advancement was based on clinical assessment and MBSS results.

Ultimately, normal swallowing was defined by tolerance of all foods by mouth (regular diet) without coughing, choking, or difficulty passing solids or liquids. If MBSS was used for evaluation, recovery was defined by resolution of symptoms as above, but also by resolution of aspiration on subsequent MBSS. Dysphagia was defined as the condition in any patient not meeting these criteria.

Rehabilitation therapies for dysphonia and/or dysphagia were chosen based on final assessments from the above evaluations. All patients with dysphonia and/or dysphagia were treated with outpatient therapy by speech and language pathology specialists. Patients were offered medialization laryngoplasty by injection within weeks of the initial procedure if no primary reinnervation had been performed and if vocal fold paralysis was evident. Thyroplasty was offered after 6 months of observation if injections had improved patient symptoms. Restricted diets were advanced based on patient symptoms combined with subsequent MBSS findings.

Results

Nineteen patients met inclusion criteria: 8 with schwannoma and 11 with paraganglioma. Patients with facial nerve schwannoma (n = 1) and supraclavicular nerve schwannoma (n = 1) were excluded from the schwannoma group, and 1 patient with a carotid body tumor with insufficient documentation of speech and swallowing function was excluded from the paraganglioma group.

Schwannoma group

Of the 6 patients evaluated for schwannoma, 2 were men and 4 were women. The median age at presentation was 42 years (range: 26 to 68 years). Presenting symptoms included painful neck mass (n = 2), dysphagia and dysphonia with neck mass (n = 2), and neck mass with other functional deficits (n = 2). All patients underwent excision via transcervical approaches. Tumors originated from the sympathetic chain (n = 1), CN IX (n = 1), and CN X (n = 4). Median follow up was 10 months (range: 8 to 12 months).

Patient operative data, immediate postoperative speech and swallowing outcome, primary or subsequent treatments, and final speech and swallowing status at final follow-up are shown in table 1. CN sacrifice resulted from 5/6 (83.3%) procedures. The immediate rate of dysphonia and dysphagia, with or without aspiration, was 5/6 (83.3%) for each symptom. The patients with speech and swallowing deficits had tumors originating from CN X (n = 4) or CN IX (n = 1); 3 of these patients underwent CN sacrifice. Primary reinnervation procedures were performed in 2 of these 5 patients (patients 1 and 2).

Table 1.

Speech and swallowing outcomes of 6 patients with cervical schwannoma who underwent resection via a transcervical approach^*

Patient No.	Site	Primary Treatment	Involved nerve(s) (condition)†	Initial outcome	Subsequent treatment	Final outcome	Follow-up
1	PPS	Ansa cervicalis inter- position graft	X (sacrificed)	VF paralysis, dysphonia, aspiration	Outpatient therapy	Improving dysphonia, regular diet	8 mo
2	CB	Ansa cervicalis inter- position graft	X (sacrificed)	VF paralysis, dysphonia, dysphagia w/o aspiration	Outpatient therapy	Strong voice, regular diet	12 mo
3	PPS	None	IX, X, XI (IX sacrificed; X traumatized but intact)	Preoperative VF paralysis, hemiglossal pare- sis, dysphonia, aspiration	VF injection ×2, thyroplasty at 6 mo	Persistent dysphonia, thickened diet	11 mo
4	PPS	None	IX, X, XI, XII (X & XII sacri- ficed, others traumatized but intact)	VF & hemi- glossal paralysis, dysphonia, aspiration	VF injection ×2, thyroplasty at 8 mo	Persistent dysphonia, PEG dependence	9 mo
5	PPS	None	IX, X, XI, XII (X & XII trauma- tized but intact)	VF paralysis, hemi- glossal paresis, dysphonia, aspiration	VF injection ×1 at 7 mo	Improving dysphonia, regular diet	8 mo
6‡	CB	None	Sympathetic nerve chain (sacrificed)	First bite syn- drome, Horner syndrome	Declined botulinum toxin injection therapy	Persistent symptoms	12 mo

^*No primary medialization thyroplasty procedures were performed.

^†Underlined nerve indicates tumor origin.

^‡This patient had deficits other than speech and swallowing deficits.

Key: PPS = parapharyngeal space; CB = carotid bifurcation; VF = vocal fold; PEG = percutaneous endoscopic gastronomy tube.

Patient 1 developed aspiration pneumonia 1 month after surgery that was successfully treated medically. She returned to a regular diet and had improving dysphonia at 8 months. Of the patients with postoperative speech and swallowing deficits, patient 2 had improved by 6 months, with a strong voice and return to a regular diet.

The 6-month rates for dysphonia or dysphagia were 4/6 (66.7%) for each. At final follow-up, 2 additional patients had improved: Patients 1 and 5 had returned to a regular diet, although they still were dysphonic. The final rate of dysphonia was 4/6 (66.7%), while that for dysphagia was 2/6 (33.3%). Each of the 4 patients with persistent dysphonia (patients 1, 3, 4, and 5) had tumors located adjacent to the skull base; 3 had sacrifice of cranial nerves.

Paraganglioma group

Of the 10 patients with paraganglioma, 3 were men and 7 were women. The median age at presentation was 58.5 years (range: 26 to 76 years). Patients most commonly presented with a painless, slowly enlarging neck mass (n = 5). Other symptoms included sore throat, otalgia, coughing, seizures, and pulsatile tinnitus. All patients underwent excision via a transcervical approach. Postoperatively, patients were followed for a median of 15.5 months (range: 1.25 to 48 months). The patients followed for 1.25 and 1.5 months had no postoperative deficits; all others had at least 6 months of follow-up.

Patient operative data, immediate postoperative speech and swallowing outcome, subsequent treatments, and final speech and swallowing status at final follow-up are shown in table 2.

Table 2.

Speech and swallowing outcomes of 10 patients with cervical paraganglioma who underwent resection via a transcervical approach^*

Patient no.	Site	Involved nerve(s) (condition)	Diagnosis	Initial outcome	Subsequent treatment	Final outcome	Follow-up
1	PPS	IX, X, XI, XII (X sacrificed)	GV	VF paralysis, dys- phonia, aspiration	VF injection ×2, thyroplasty at 10 mo	Persistent dys- phonia, PEG dependence	14 mo
2	PPS	IX, X, XI, XII (X sacrificed)	GV	Preoperative VF & hemiglossal para- lysis, dysphonia, aspiration	VF injection ×2, then thyro- plasty at 16 mo	Persistent dys- phonia, PEG dependence	17 mo
3	CB	X, XII (nerves intact)	GV	VF paralysis, dys- phonia, dysphagia w/o aspiration	VF injection ×2 at 6 mo	Strong voice, regular diet	48 mo
4	CB	X (nerve intact)	CBT	VF paralysis, dys- phonia, dysphagia w/o aspiration	VF injection ×1 at 2 mo	Strong voice, regular diet	6 mo
5	CB, PPS	X (nerve intact)	GV, CBT	VF paresis, dys- phonia, dysphagia w/o aspiration	Outpatient therapy	Strong voice, regular diet	34 mo
6	CB	Pharyngeal plexus (nerves intact)	CBT	Dysphonia, dys- phagia w/o aspi- ration†	Outpatient therapy	Strong voice, regular diet	38 mo
7	CB	None	CBT	No speech or swal- lowing deficits	No further Tx	No dysfunction	1.5 mo
8	CB	None	CBT	No speech or swal- lowing deficits	No further Tx	No dysfunction	1.5 mo
9	CB	None	CBT	No speech or swal- lowing deficits	No further Tx	No dysfunction	1.25 mo
10	CB	None	CBT	No speech or swal- lowing deficits	No further Tx	No dysfunction	40 mo

^*No primary medialization thyroplasty procedures were performed.

^†No VF paralysis.

Key: PPS = parapharyngeal space; GV = glomus vagale; VF = vocal fold; CB = carotid bifurcation; CBT = carotid body tumor; Tx = treatment.

Final diagnoses were carotid body tumor alone (n = 6), glomus vagale tumor alone (n = 3), and carotid body tumor with simultaneous glomus vagale (n = 1). Cranial nerve sacrifice occurred in 2/10 (20%) procedures; both were glomus vagale cases. The immediate rate of dysphonia was 6/10 (60%). The immediate rate of dysphagia with or without aspiration was also 6/10 (60.0%). No initial reinnervation procedures were performed. Four patients required medialization laryngoplasty to improve final functional outcomes; none was performed primarily. Other sequelae included first-bite syndrome (n = 3) and shoulder weakness (n = 1).

Of the patients with immediate postoperative speech and swallowing deficits, 3 patients had improved by 6 months: Patients 4, 5, and 6 had achieved a strong voice and had returned to a regular diet. The 6-month rates of dysphonia and dysphagia were 3/10 (30%) for each. At final follow-up, 1 additional patient had improved: Patient 3 had achieved a strong voice and had returned to a regular diet.

The final rates of dysphonia and dysphagia were 2/10 (20%) for each. Both of these patients had glomus vagale tumors removed near the cranial base with CN X sacrificed.

Table 3 displays nerve injury rates, immediate dysphonia or dysphagia rates, and recovery from dysphonia or dysphagia by tumor and location.

Table 3.

Nerve sacrifice, dysphonia, and dysphagia rates by tumor type and location

Tumor type/location	Nerve sacrificed*	Immediate dysphonia	Dysphonia recovery†	Immediate dysphagia	Dysphagia recovery†
Skull base schwannoma	3/4 (75)	4/4 (100)	0/4 (0)	4/4 (100)	2/4 (50)
Carotid bifurcation   schwannoma	2/2 (100)	1/2 (50)	1/1 (100)	1/2 (50)	1/1 (100)
Skull base paraganglioma	2/4 (50)	4/4 (100)	2/4 (50)	4/4 (100)	2/4 (50)
Carotid bifurcation   paraganglioma	0/7 (0)	3/7 (42.9)	3/3 (100)	3/7 (42.9)	3/3 (100)

^*Number/total (percent).

^†Recovery refers to return of a strong voice and regular diet by final follow-up.

Discussion

Normal speech and swallowing require the incorporation of complex afferent and efferent neural processes. Speech relies on phonation from the larynx, resonance via the vocal tract, and articulation through the modifications of sound made within the oral cavity and oropharynx. As motor inputs from the recurrent laryngeal nerve and/or superior laryngeal nerve are interrupted, vocal fold paresis or paralysis can lead to incomplete glottal closure and a resultant breathy voice consistent with unilateral vocal fold paralysis. High vagal injury will additionally affect pharyngeal sensation and movement, leading to alterations in phonation and resonance, while hypoglossal injury leads to tongue dysfunction and changes in articulation.⁸

Swallowing relies heavily on both afferent and efferent neural integration and is quite complex. Sensation is provided by a combination of CN V afferents from the oral cavity, CN IX in the oropharynx, and CN X in the larynx. These nerves interact with motor inputs that allow for both the preparation of food boluses during the voluntary portions of swallowing and the series of coordinated movements necessary for normal swallowing function during the involuntary phases.

Motor function primarily relies on CN XII for voluntary movement of the tongue during the oral phase and CN X for glottal closure, uvular/palatal function, and sequential pharyngeal constriction during the pharyngeal phase. The gag and cough reflexes are mediated through CN IX/CN X and CN X/CN X, respectively. Finally, laryngeal elevation, a protective measure to prevent aspiration during swallowing, is performed by the suprahyoid musculature under control of CNs V, VII, and XII via C1.⁹

Dysfunction of speech or swallowing can result from disruption of the neural integration of any of these above mechanisms and can result from nerve sacrifice, devascularization, or stretch injury. With regard to extracranial schwannomas or paragangliomas, tumor locations in the carotid bifurcation or high in the parapharyngeal space near the exits of CNs IX, X, and XII from the cranial base put each nerve at risk during surgical removal.

In the case of schwannoma, neural fascicles are frequently splayed over tumors, making it difficult to preserve neural function, while incomplete excision leads to slow local recurrence over the course of months to years.² It has been shown that 56% of cases of complete excision required sacrifice of the involved nerve, and of the other 44%, 64% had permanent and 29% had transient deficits.¹⁰ Horner syndrome in 3 of 8 patients and vocal fold palsies in 3 of 8 patients were noted by Colreavy et al,¹¹ while facial paralysis, deafness, hoarseness, numbness, and pain associated with cranial nerves and the sympathetic chain have been reported.^1,12 Thus, a balance of complete excision and functional preservation is sought but is often difficult to achieve.

Outcomes data for cervical paraganglioma are primarily reported by studies examining resection of carotid body tumors, as approximately 60% of paragangliomas of the head and neck arise in that location.¹³ Management strategies for CN dysfunction after treatment of carotid body tumors were discussed in a review by van der Mey et al and focused on postoperative swallowing therapy and potential thyroplasty for persistent aspiration.¹⁴ A detailed discussion of management options was offered by Sniezek et al in their review of vagal paraganglioma, where swallowing therapy, reinnervation, and multiple thyroplasty procedures were suggested in certain cases, along with a palatal adhesion procedure for velopharyngeal incompetence resulting from hemipalatal paralysis.¹⁵

We found cervical schwannoma along multiple CNs in the deep spaces of the neck, most of which were close to the cranial base. Five of 6 cases involved sacrifice of at least 1 cranial nerve. Dysphonia and dysphagia complicated initial patient recovery in 5 of the 6 patients (83.3%); in 1 patient, this did not involve nerve sacrifice. Each had additional procedures to improve speech and swallowing function, 2 primarily with reinnervation procedures and 3 with subsequent medialization laryngoplasty. One patient achieved a strong voice and returned to a regular diet by 6 months. One additional patient returned to a regular diet by final follow-up with improving voice, but continued to be dysphonic.

In contrast to cervical schwannoma, cervical paraganglioma was more frequently found away from the skull base at the carotid bifurcation (n = 8). As might be expected, nerve sacrifice in 2 of 10 patients (20%) and immediate dysphonia and dysphagia rates in 6 of 10 patients (60% for both) were lower, and further functional recovery was better. At 6 months, 3 additional patients (30%) acquired a strong voice and returned to a regular diet, while another patient improved similarly by final follow-up. Only 4 of the paraganglioma patients needed additional medialization laryngoplasty postoperatively, including both cases involving nerve sacrifice for resection of glomus vagale.

Although the limited numbers in this case series restrict our ability to statistically analyze such rates qualitatively, table 3 indicates that cervical schwannoma, when compared with cervical paraganglioma, may lead to increased risk of CN sacrifice, worse initial speech and swallowing outcomes, a greater percentage of subsequent rehabilitation procedures, and poorer recovery overall. Within the paraganglioma group, resection of glomus vagale led to worse outcomes and, in general, tumors located at the cranial base have poorer outcomes, as would be expected.

Observations from both groups include a greater functional deficit in those patients who had CN X sacrifice or tumor location near the cranial base, as expected. The data suggest a more expedient functional recovery with use of primary reinnervation or subsequent medialization laryngoplasty procedures. Comparing patients 1 and 2 for each type of tumor, it may be that if the glomus vagale patients had been treated with reinnervation procedures primarily, perhaps they would have experienced greater functional benefit and avoided subsequent medialization procedures.

Several patients with intact nerves had considerable speech and swallowing deficits, which reinforces the theory that devascularization and/or stretch injuries can cause dysfunction of similar severity and duration as nerve sacrifice. This can be additionally problematic at the cranial base, where glossal and pharyngeal/laryngeal dysfunction can result from extensive dissection. Such patients need close monitoring so that medialization procedures can be used as adjuncts.

We will continue to treat patients who are appropriate surgical candidates with surgical therapy. Our data are informative with respect to speech and swallowing dysfunction and are useful for informing patients of prognostic information. Furthermore, our data suggest that immediate reinnervation procedures should perhaps be performed when nerve sacrifice is either necessary or inadvertently performed. Our decisions on appropriate diet will continue to be based on patient reports and findings on MBSS.

Limitations of this study include the retrospective nature of a case series, which imparts certain inherent biases. The small number of cases identified did now allow for formal statistical analysis, which would have strengthened results and conclusions. Validated self-reporting instruments would have provided additional outcomes data in a more objective form.

Conclusions

In this case series, cervical schwannoma and paraganglioma were often intimately involved with neurovascular structures at risk of injury during surgical excision. Postoperative speech and swallowing deficits were common, and locations at the cranial base carried the greatest risk. Resection of schwannomas had higher rates of nerve injury, dysphonia, and dysphagia compared with paraganglioma excision, while also portending poorer functional recovery. Speech and swallowing therapy and medialization laryngoplasty led to symptom improvement by 6 months and, ultimately, a strong voice and return to regular diet in some patients. Reinnervation should be considered if CN X is sacrificed.