A Prospective Evaluation of Swallowing and Speech in Patients with Neurofibromatosis Type 2

Abstract

Objective  Neurofibromatosis type 2 (NF2) patients report that swallowing and speech problems significantly affect their quality of life, but the etiology of these phenomena is poorly understood. Swallowing and speech deficits may arise due to the neuropathy of involved nerves, due to posterior fossa tumor growth, or as iatrogenic effects from neurosurgical procedures to remove these tumors. This study aims to identify the natural history of swallowing and speech deficits in an NF2 cohort and to characterize the factors that may lead to those deficits.

Methods  Subjects ( n  = 168) were enrolled in a prospective, longitudinal study of NF2 with yearly imaging and clinical exams. The patients completed a self-reported questionnaire that included responses regarding subjective swallowing and speech dysfunction. A formal speech-language pathology evaluation and modified barium swallow (MBS) study (reported as American Speech-Language Hearing Association [ASHA] swallowing independency score from 1 through 7) was obtained when a speech/swallowing deficit was reported on the questionnaire.

Results  Of the 168 enrolled subjects, 55 (33%, median age = 31 years) reported subjective speech and/or swallowing deficits. These patients underwent one ( n  = 37) or multiple ( n  = 18) MBS studies during 44.8 ± 10.4 months follow-up. During MBS, a majority demonstrated near-normal swallowing (ASHA score >6, 82%), and no evidence of aspiration (aspiration/laryngeal penetration score = 1, 96%). Prior to initial MBS consultation, 38 (69%) patients had undergone relevant neurosurgical procedures. In those with recent (<1 week) posterior fossa surgery ( n  = 12), 2 (17%) patients had severe dysphagia and high aspiration risk on postoperative MBS. Both of these patients recovered to functionally independent swallowing status. Unilateral ( n  = 10) or bilateral ( n  = 6) tongue deficits unrelated to previous history suggestive of hypoglossal nerve injury were detected on clinical examination. There was a correlation between the presence of dysarthria and tongue deficits and tumors associated with the hypoglossal canal noted on imaging.

Conclusion  A large proportion of patients with NF2 report speech and swallow deficits that are not evident on objective measurements. We also found hypoglossal neuropathy unrelated to prior surgical interventions. Our findings suggest that swallowing and speech problems in NF2 are associated with lower cranial nerve neuropathy, some due to compressive effects of posterior fossa tumors. Adaptation over the course of the disease allows for the compensation of these deficits and subsequent normal findings on objective testing.

Introduction

Neurofibromatosis type 2 (NF2) is a multitumor syndrome involving schwannomas, meningiomas, and ependymomas of both the central and peripheral nervous systems. The prevalence in the population is estimated at 1 in 60,000 and most commonly manifests as an autosomal dominant inherited condition. ^{1 2} Variations in the mutation type and location on the NF2 gene (chromosome 22) responsible for this disease account for differences in progression and mortality within the NF2 population. ^{2 3}

NF2 patients report that swallowing and speech problems significantly affect their quality of life, ⁴ but the etiology of these phenomena is poorly understood. Peripheral neuropathy has been well described in patients with NF2. ^{5 6} However, as we have discussed earlier, ⁷ the cranial nerve (CN) correlates of peripheral neuropathy remain unknown. Best et al recently documented significant impairment in CNs VII and VIII ipsilateral to tumors involving the internal auditory canal (IAC), as well as impairment to CNs X and XII in a subset of patients. ⁸ Treatment of NF2-associated tumors often involves surgery within the posterior fossa, which may result in swallowing and speech deficits. ^{9 10 11 12 13 14 15} Swallowing and speech deficits may arise due to the neuropathy of involved nerves, due to posterior fossa tumor growth, or as iatrogenic effects from neurosurgical procedures to remove these tumors. ^{16 17 18} This prospective study aims to investigate swallowing and speech deficits in NF2 patients in the context of their natural history.

Methods

Patient Enrollment and Study Design

Patients enrolled in a prospective, longitudinal study of NF2 (NIH Protocol# 08-N-0044; ClinicalTrials.gov identifier NCT00598351) were evaluated over the course of 5 years. There were a total of 168 participants, all of whom signed informed consents and completed a self-reported questionnaire that included four pertinent questions regarding subjective swallowing and speech symptoms issues ( Table 1 ). If at any time point, a patient reported “yes” to any of the four questions, suggesting swallowing or speech problems, a formal consultation to speech-language pathology (SLP) was made and modified barium swallow (MBS) study was performed. Patients were also referred for SLP consultation perioperatively per clinical need. Surgical procedures were performed as the part of a separate trial (NIH Protocol# 03-N-0164; ClinicalTrials.gov identifier NCT00060541).

Table 1. Subjective self-reported speech and swallow deficit questionnaire.

#		Question
Q1	Dysphagia	“Do you cough or choke when eating or within 20 min after eating?”
Q2	Dysphagia	“Do you feel food or liquid sticking in your neck when eating?”
Q3	Dysphagia	“When eating or drinking, do you experience throat clearing?”
Q4	Dysarthria/dysphonia	“Have you noticed a change in your speech and voice?”

Note : All patients on the protocol received a questionnaire at each visit to the neurosurgery clinic. Four of the questions were directly relevant to swallowing and speech issues. If at any time point, a patient answered “yes” to any of the four questions, suggesting swallowing or speech problems, a consultation to speech-language pathology was made and modified barium swallow study was performed.

Evaluation of Speech and Swallow Function

MBS study included multitexture (solid, liquid, and puree) assessment of swallowing function. Results were quantified according to the American Speech-Language Hearing Association (ASHA) 2003 swallowing independency score ( Supplementary Table 1 ). ¹⁹ The ASHA score is based on functional independence in eating/drinking, with a score of 1 corresponding to unsafe swallowing and a score of 7 reflecting functional independence. Aspiration and laryngeal penetration was assessed using a one-to-eight score obtained from Rosenbek et al, with a score of 1 corresponding to normal/no material entry into airway and a score of 8 corresponding to material entering the airway and passing below the vocal folds with no effort made to eject ( Supplementary Table 2 ). ²⁰ Speech and oral-motor function testing included assessment of vocal strength and quality, lip strength (CN VII), palate deviation (CN X), vocal cord movements (CN X), and tongue function (CN XII). ²⁰ Endoscopic evaluation of the larynx was performed to evaluate vocal cord movement. Neuro-otology clinic visit included complete CN examination. Deficits were characterized by laterality and severity. For example, if a tongue deficit was observed (CN XII dysfunction), the deficit was classified as atrophy, weakness, fasciculation, or deviation of affected side. In patients who did not receive neuro-otology consultation, CN examination data were obtained from evaluation by the neurosurgery service.

Delineation of Anatomic Changes on Imaging

All patients also underwent neuroimaging with high-resolution magnetic resonance imaging (MRI) pre/post-gadolinium contrast at each visit. Relevant abnormalities were classified as IAC lesion, hypoglossal nerve thickening, hypoglossal canal tumor, jugular bulb postsurgical changes, other jugular foramen abnormalities, and posterior fossa crowding. These abnormalities were deemed relevant based on correspondence to clinical manifestations (observed CN deficits). MR images were read by both the neurosurgery and neuro-otology services and discrepancies were resolved by consensus.

Results

Self-Reported Swallowing and Speech Deficits Do Not Correspond to Objective Measures

Of the 168 patients enrolled in the natural history study, 55 reported speech and swallow deficits on questionnaires ( Fig. 1 ). These deficits were noted to be stable over time with repeated administration of questionnaire ( Fig. 2A ). Patient demographics, NF2 status, and prior relevant surgical history ( Table 2 ) were characterized. Patients underwent one (37 patients, 67.27%) or more than one (18 patients, 32.73%) MBS to determine whether these self-reported deficits resulted in objective physiologic abnormalities. During MBS study, a majority of patients demonstrated functionally independent swallowing ( Fig. 2B ) as well as little evidence of aspiration ( Fig. 2C ). Based on MBS results, dietary restrictions for solids and liquids were recommended in 20 and 2% of patients, respectively. Of the 55 patients in the study cohort, 12 patients underwent 18 surgical procedures relevant to speech and language dysfunction including cervical laminectomies ( n  = 2), retrosigmoid craniotomies ( n  = 9), and supratentorial surgery ( n  = 1). Speech and swallowing testing was performed in the postsurgical period, and repeated as clinically indicated. Of the patients with a recent (≤ 1 week) history of posterior fossa surgery, only two demonstrated unsafe swallowing (ASHA score = 1) and high aspiration risk (aspiration/laryngeal penetration score = 8) on postoperative MBS. Both of these patients recovered functionally independent swallowing status without any evidence of aspiration within 2 weeks postoperatively ( Supplementary Fig. 1 ).

Fig. 1.

Study design and patient selection. Of patients were enrolled in the NF2 natural history study, a subset qualified for further evaluation of speech/swallow deficits due to self-reported symptoms on questionnaire or postoperative complaints. NF2, neurofibromatosis type 2.

Fig. 2.

Reported speech and swallow deficits occur irrespective of surgical intervention and do not correlate with quantifiable oral-motor function changes on examination. ( A ) Positive responses to questionnaire evaluating dysphagia (Q1–Q3) and dysarthria/dysphonia (Q4). ( B ) ASHA and ( C ) aspiration/laryngeal penetration score were recorded in patients reporting these deficits. ASHA, American Speech-Language Hearing Association.

Table 2. Baseline demographic and clinical characteristics.

Variable	n  = 55 (100%)
Patient overview
Age at diagnosis (y)
Mean (SD)	20.30 (13.60)
Median (range)	31 (9, 63)
Sex, n (%)
Female	38 (69.09%)
Male	17 (30.91%)
KPS at baseline, n (%)
< 90	35 (63.64%)
≥ 90	20 (36.36%)
NF2 mutational status
Deletion	1 (1.82%)
Nonsense	11 (20.00%)
Splice mutation (acceptor site)	2 (3.64%)
Splice mutation (donor site)	1 (1.82%)
Unknown	25 (45.45%)
No mutation detected	5 (9.09%)
Posterior fossa surgery prior to first MBS
Any prior posterior fossa surgery, n (%)	19 (34.55%)
Retrosigmoid surgery, n (%)
Left-sided	13 (23.64%)
Right-sided	10 (18.18%)
Translabyrinthine surgery, n (%)
Left-sided	13 (23.64%)
Right-sided	18 (32.73%)
Middle fossa surgery, n (%)
Left-sided	3 (5.45%)
Right-sided	2 (3.64%)
Facial reanimation surgery, n (%)
Left-sided	1 (1.82%)
Right-sided	0 (0.00%)
Cervical laminectomy, n (%)	1 (1.82%)
Other skull base surgery, n (%)	0 (0.00%)

Abbreviations: KPS, karnofsky performance score; MBS, modified barium swallow; NF2, neurofibromatosis type 2.

Note : Patient demographics, NF2 status, and prior relevant surgical history characterized.

NF2 Patients Demonstrate Lower Cranial Neuropathy Independent of Injury

We then reviewed MRI to determine whether anatomical correlates or abnormalities explaining these deficits could be delineated ( Table 3 ). Thirty-one patients (56.36%) had bilateral vestibular schwannomas and 13 patients (23.64%) had unilateral vestibular schwannomas on MRI ( Supplementary Fig. 2 ). Additionally, a subset of patients had tumors associated with the hypoglossal canal and jugular bulb.

Table 3. Presurgical MRI anatomical characterization of all patients.

Variable	n  = 55 (100%)
Hypoglossal nerve abnormalities, n (%)
Tumor
Unilateral	15 (27.27%)
Bilateral	3 (5.45%)
Jugular foraminal abnormalities, n (%)
Tumor
Unilateral	17 (30.91%)
Bilateral	3 (16.36%)
Bulb occlusion
Unilateral	1 (1.82%)
Bilateral	0 (0.00%)
Internal auditory canal abnormalities, n (%)
Vestibular schwannoma
Bilateral	31 (56.36%)
Unilateral	13 (23.64%)
Other, n (%)
Sinus occlusion	20 (36.36%)
Encephalomalacia	17 (30.91%)
Meningioma(s) present	14 (25.45%)

Abbreviation: MRI, magnetic resonance imaging.

Note : Characterization of anatomical abnormalities on MRI in all patients on study.

We found that several NF2 patients in the study (25/55) demonstrated unilateral (8/55) or bilateral (17/55) tongue atrophy ( Fig. 3A ). A majority of these patients (17/25) had not undergone prior cranial or skull base neurosurgical interventions that could be attributed as a cause of such deficits due to surgical hypoglossal nerve injury ( Fig. 3B ). Objective tongue deficits included atrophy, fasciculations, weakness, and/or deviation ( Fig. 3C ). We reviewed the course of hypoglossal nerve on MRI to determine whether compressive lesions of the hypoglossal nerve could explain the high frequency of hypoglossal nerve deficits, especially in patients with no relevant neurosurgical history. Of the17 patients with objective tongue deficits and no prior surgical history, MRI studies of 2 patients were of inadequate quality to evaluate the hypoglossal nerves. Of the remaining 15 patients, 10 had tumors associated with, compressing, or growing into the hypoglossal canal ipsilateral to the noted deficit ( Fig. 4A–C ). However, in the rest, objective tongue deficits were not associated with observable compressive lesions along the hypoglossal nerve. Similarly, we found hypoglossal canal lesions in patients without tongue deficits ( Table 4 ). Overall, the odds of tongue deficits were elevated (odds ratio: 10.67, 95% confidence interval: 2.8–46.5, p  < 0.001) with the presence of compressive lesions along the course of the hypoglossal nerve. However, upon multivariable analysis, we did not find significant association between objective tongue deficits and hypoglossal canal tumors. This suggests that other factors, including intrinsic NF2-related CN neuropathy may be responsible. As expected, there were significant associations between speech dysarthria scores and presence of compressive lesions along the course of hypoglossal nerve ( Supplementary Table 3 ). However, such lesions adjacent to the jugular bulb were not associated with swallowing or speech deficits.

Fig. 3.

Objective tongue abnormalities are noted in a subset of patients with speech and swallow complaints. ( A, B ) Tongue atrophy noted preoperatively in some patients with NF2. ( C ) Specific abnormalities characterized in this group of patients. NF2, neurofibromatosis type 2.

Fig. 4.

Patients with tongue abnormalities have corresponding positive findings on imaging. Example of patients with ( A ) normal hypoglossal canal and ( B ) hypoglossal canal with compressive tumor ipsilateral to tongue abnormality. ( C ) The number of patients with positive findings on MRI was quantified. Positive findings were defined as any tumor associated with, compressing, or growing into the hypoglossal canal. MRI, magnetic resonance imaging.

Table 4. Delineation of tongue deficits and correlative findings on MRI.

	Positive MRI findings	Negative MRI findings	MRI equivocal	Total
Tongue deficits present	10	5	2	17
No tongue deficits present	6	32	0	38
Total	16	37	2	55

Abbreviation: MRI, magnetic resonance imaging.

Note : Patients tongue deficits including fasciculations, weakness, deviation, or atrophy were tabulated against findings of structural lesions associated with the hypoglossal nerve on MRI (total n  = 55).

We found that several patients had jugular bulb abnormalities including tumors, nerve thickening, and postsurgical changes. However, we found that neither unilateral ( n  = 14/55, 25%) nor bilateral ( n  = 7/55, 12%) jugular bulb abnormalities were associated any clinical deficits (speech dysarthria) or swallowing dysfunction (ASHA score, aspiration/laryngeal penetration [ASP] score).

Discussion

Though self-reported swallowing and/or speech deficits were the primary reason for initial MBS study, the ASHA swallowing independency score at initial visits did not demonstrate objective functional deficits. Despite these findings on MBS, patients continued to report deficits over the course of the study. The progression of these deficits was stable and independent of neurosurgical or medical intervention. Furthermore, in cases where neurosurgical intervention caused an additional deficit, there was a subsequent return to baseline within a short time span.

We sought to understand the mechanisms of swallowing and speech deficits in our patients. Swallowing is primarily controlled by CNs V, VII, IX, X, and XII (hypoglossal nerve). ²¹ The mechanism itself is divided into four stages: oral preparatory, oral, pharyngeal, and esophageal stages. ²² The oral preparatory and oral phases are coordinated by CN VII and hypoglossal nerve, which innervate muscles of the lips and the tongue. ²³ The pharyngeal and esophageal phases, which protect the airway while pushing the bolus to through the pharynx and into the esophagus, are largely controlled by CNs IX and X, with some CN XI involvement. Speech is primarily controlled by CN X and hypoglossal nerve. Tongue deficits have been noted in patients with NF2. ^{24 25} During clinical examinations, hypoglossal deficits are readily observed as tongue atrophy and/or fasciculations. This makes the tongue examination a reliable marker of hypoglossal neuropathy specifically and lower CN neuropathy in general.

The etiology of tongue deficits remains unknown. Prior work has suggested that existing ipsilateral lesions and subsequent surgical resection are risk factors for lower CN injury, demonstrated by increased CN X deficits. ⁸ The findings of this study additionally suggested that lower CN symptoms reported by NF2 patients were a component of the natural progression of NF2. A proportion of patients reporting swallow and speech deficits did demonstrate ipsilateral compressive lesions of the hypoglossal nerve on imaging. Indeed, such lesions significantly increased the odds of finding ipsilateral tongue deficit. However, we also found the presence of tongue deficits in the absence of observable compressive lesions. This suggested an intrinsic NF2-related neuropathic component to the progression of these symptoms. We suspect that other CNs involved in swallowing and speech (CNs V, VII, IX, and X) also suffer from chronic, progressive neuropathy in patients with NF2. An accumulation of neuropathies may be perceived by NF2 patients as progressive difficulties with swallowing and speech function. That the patients were able to maintain a normal functional status on MBS suggested that there was a possible adaptive mechanism to account for these anatomical changes. Even though CN function deficits progressed, these deficits did not cause significant objective dysfunction likely due to compensatory processes.

Further investigation revealed that although speech verbal intelligibility (assessed by dysarthria score) was within functional limits for the majority of patients, there was a significant correlation between speech dysarthria score and presence of hypoglossal nerve lesion on MRI. This is likely due to the tongue's crucial involvement in normal speech articulation. Other oral-motor outcomes demonstrated multicranial neuropathy in patients with NF2. Palate deviation and vocal cord dysfunction that was observed in a subset of patients may reflect neuropathy of CN X.

Conclusion

Subjective complaints of swallowing or speech dysfunction are reported frequently even though deficits are not clinically significant in the vast majority of cases. Swallowing deficits are evident on occasion, but reflect transient postsurgical findings in patients with NF2 undergoing relevant posterior fossa surgery. Though fluctuations exist, NF2 patients overall maintain high levels of swallowing and speech functional status throughout their lives.

Tongue deficits are seen in NF2 patients both with and without any kind of relevant surgical history. Hypoglossal dysfunction should be considered a component of natural disease progression that could be exacerbated by surgical interventions. Patients with hypoglossal nerve symptoms (tongue deficits) may or may not present with notable radiographic findings, but when an anatomical abnormality on imaging is observed, the finding often corresponds to the observed deficit in laterality. Clinicians should consider screening for hypoglossal canal abnormalities during their typical radiographic examination and routinely assess nerve function on physical exam.

Funding Statement

Funding Sources This study was supported by the Intramural Research Programs of the National Institute of Neurological Diseases and Stroke, the National Institute for Deafness and Communication Disorders, and the National Institutes of Health Clinical Center, Bethesda, Maryland, United States.