Vocal Fold Paralysis Following Radiotherapy for Nasopharyngeal Carcinoma: Laryngeal Electromyography Findings☆

INTRODUCTION 

The cranial and peripheral nerves are considered radio-resistant in contrast to the brain and spinal cord. Injury following radiotherapy is rare and normally no specific precautions are taken to protect these nerves during head and neck irradiation.1 Nevertheless all the cranial nerves are potentially at risk of injury. The reported incidence following radiotherapy for nasopharyngeal cancer ranges from 0.3% to 9%.2., 3., 4., 5., 6., 7., 8. Apart from injury to the vestibulocochlear apparatus, the hypoglossal nerve is most at risk followed by the vagus nerve. Vocal fold paralysis resulting from vagal injury is usually unilateral9 although bilateral paralysis can occur and is significant because it has the potential to cause life-threatening airway obstruction.10., 11., 12.

Until now there have been no reported laryngeal electromyography (EMG) studies on patients with vocal fold paralysis following radiotherapy. Laryngeal EMG can provide a range of useful information in a patient with a fixed vocal fold after radiotherapy. Demonstration of neuropathic changes in the thyroarytenoid muscle can distinguish paralysis from fixation due to other causes such as cricoarytenoid joint ankylosis and posterior glottic stenosis, both potential complications of radiotherapy.13., 14. EMG can also provide information on the site of the lesion15., 16. and in unilateral paralysis on the electrophysiological status of the contralateral vagus nerve. It may be able to distinguish radiation neuropathy from that due to recurrent tumor or other causes17., 18., 19. and can sometimes provide prognostic information.20., 21., 22., 23., 24. Laryngeal EMG findings are presented from three patients who developed unilateral vocal fold paralysis following radiotherapy for nasopharyngeal carcinoma.

MATERIAL AND METHODS 

Three patients presented to the Department of Otolaryngology between 1997 and 1999 with vocal fold paralysis following radiotherapy for nasopharyngeal carcinoma. Recurrent carcinoma and other pathology were excluded by fiber optic nasopharyngoscopy, postnasal space biopsy, and computed tomography or magnetic resonance imaging from the skull base to the superior mediastinum. Laryngeal function was assessed by videolaryngoscopy using a 70-degree rigid trans-oral laryngoscope (Kay Elemetrics, Lincoln Park, NJ), videofluoroscopic swallowing evaluation if clinically indicated, and laryngeal EMG.

RESULTS 

The patients' ages ranged between 52 and 61 years. Two were males and one was female. Using the UICC/AJCC staging system,26 the primary tumor was T2 in two patients and T3 in one patient. All three patients had stage N1 nodal disease. None had distant metastases. The tumors were all left sided. Radiotherapy doses to the primary tumor and neck are summarized in Table 1. No patient had any cranial nerve abnormalities prior to radiotherapy. Onset of vocal symptoms following radiotherapy was at three years in both patients with T2 tumors and four years in the patient with a T3 tumor. All three patients presented with hoarseness. Videolaryngoscopy demonstrated immobility of the left vocal fold in two patients and of the right vocal fold in one patient. The contralateral vocal fold was mobile in all cases. One patient had symptoms of aspiration but did not demonstrate aspiration on videofluoroscopy. None of the patients had evidence of recurrent tumor or other pathology to account for the vocal fold paralysis.

TABLE 1. Total Radiotherapy Doses to Primary Tumor and Neck

		Patient 1	Patient 2	Patient 3
	Tumor stage	T2 N1 M0	T2 N1 M0	T3 N1 M0
	Dose to primary	66 Gy	66 Gy	80 Gy (70 Gy+10 Gy parapharyngal boost)
	Dose to neck	70 Gy	70 Gy	70 Gy
	Other treatment	None	None	Chemotherapy (cisplatinum+5FU)

Laryngeal EMG was performed between one and three months following diagnosis of vocal fold paralysis. The findings are summarized in Figure 1. Two patients demonstrated reduced recruitment in both the thyroarytenoid and cricothyroid muscles on the side of the vocal fold paralysis. In the other patient the ipsilateral cricothyroid muscle was spared. Both patients with combined thyroarytenoid and cricothyroid involvement had concurrent ipsilateral hypoglossal nerve paralysis, manifested by tongue wasting and deviation to the side of the lesion. Reduced recruitment was seen on the side contralateral to the vocal fold paralysis in two patients. This involved the cricothyroid muscle in one patient and the thyroarytenoid muscle in the other. No patient had absent recruitment. Abnormalities of waveform morphology and the presence of spontaneous activity corresponded well with sites of reduced recruitment. These were seen in various combinations in all seven sites with reduced recruitment but in only one of five sites with normal recruitment. Fibrillation potentials, polyphasic potentials, and positive sharp waves were most commonly seen. One patient demonstrated myokymia in the thyroarytenoid muscle on the side of the paralyzed vocal fold (Figure 2).

View Large Image Download to PowerPoint

FIGURE 1. Schematic representation of sites of nerve injury. a=right superior laryngeal nerve; b=left superior laryngeal nerve; c=right recurrent laryngeal nerve; d=left recurrent laryngeal nerve. Spontaneous activity and abnormalities of waveform morphology are indicated next to the corresponding nerve. Fib=fibrillation potentials; PSW=positive sharp waves; CRD=complex repetitive discharges; Poly=polyphasic potentials; Myo=myokymia.

View Large Image Download to PowerPoint

FIGURE 2. Myokymia: Repetitive grouped discharges at regular frequency as indicated by arrows.

DISCUSSION 

EMG is a well-established tool in assessing nerve and muscle function.27 It tests the integrity of the entire neuromuscular pathway including the upper and lower motor neurons, the neuromuscular junction, and the muscle concerned. The use of EMG in the larynx was initially described in 1957.28 It has been applied increasingly in recent years to study neuromuscular disorders of the larynx.15., 16.

Because EMG tests the entire neuromuscular pathway, it may be difficult to differentiate local injury from nerve injury alone. In a patient with vocal fold paralysis, the most important diagnostic feature indicating nerve injury on laryngeal EMG is reduced or absent recruitment. The presence of abnormal waveform morphology and spontaneous activity support neuropathy rather than cricoarytenoid joint fixation as the reason for vocal fold immobility,13 although it is possible for the two pathologies to coexist. Localized muscular inflammation has been associated with high frequency discharges14 but these were not observed in our series. Conversely, recruitment may exist despite a functionally immobile vocal fold because of incomplete nerve injury or reinnervation (synkinesis). Despite these problems, laryngeal EMG still yields useful information.

The unique anatomy of the dual nerve supply to the intrinsic muscles of the larynx makes it possible to use EMG to determine the site of the nerve lesion in a patient with vocal fold paralysis.15., 16. When the superior and recurrent laryngeal nerves are both involved, injury must have occurred either at the skull base above the takeoff of the superior laryngeal nerve, or diffusely in both nerves. Concurrent hypoglossal nerve paralysis in both patients with combined recurrent and superior laryngeal neuropathy suggests proximal injury at the level of the skull base was more likely in these patients. If the cricothyroid muscle is spared indicating only recurrent laryngeal nerve involvement, injury must have taken place below the level of takeoff of the superior laryngeal nerve either in the neck or the mediastinum. This pattern of injury was seen in the third patient. While it has been demonstrated that radiation can cause nerve injury both by direct damage29 and by mechanical compression from fibrosis30 the location of the lesion along the nerve in patients with vocal fold paralysis following radiotherapy has only been inferred clinically in the past.31 Our EMG findings suggest that injury can occur proximally at the skull base, or in isolation more distally in the neck.

In making a diagnosis of radiation-induced vocal fold paralysis, it is important to exclude other possible etiological factors, as there may be no specific features indicating radiation as the cause other than the history and, sometimes, palpable fibrosis in the neck. It is critical to rule out recurrent tumor although other causes should also be considered. Hypoglossal nerve paralysis in two patients was the only other abnormal neurological finding, making central neurological system pathology unlikely. In addition, investigations including postnasal space biopsy and computed tomography or magnetic resonance scanning along the course of the recurrent laryngeal nerve did not reveal any other lesions to account for vocal fold paralysis in the three patients. While it is possible that viral mononeuritis could be an etiological factor, the coexistence of hypoglossal nerve injury in two patients suggests otherwise. Laryngeal EMG may help to distinguish radiation-induced neuropathy from tumor infiltration or other causes. Studies on patients with brachial plexopathy after radiotherapy have shown that the presence of myokymic discharges supports the diagnosis of radiation-induced neuropathy although this is not a sufficiently specific diagnostic criteria to be used in isolation.17., 18., 19. One patient in our series demonstrated myokymia. Fibrillation potentials, polyphasic potentials and positive sharp waves were also observed but are nonspecific changes.

Laryngeal electromyography has been used to predict the likelihood of recovering recurrent laryngeal nerve function.20., 21., 22., 23., 24. Predictive rates for recovery of between 70 and 90% have been reported based on the presence and degree of recruitment, spontaneous activity, and abnormal waveform morphology. Mildly reduced recruitment with no spontaneous activity or abnormal waveform morphology has been associated with a good prognosis whereas absent recruitment with abundant spontaneous activity has been associated with a poor prognosis.24 In each of our cases recruitment was moderately reduced, and associated with varying degrees of spontaneous activity or abnormal waveform morphology, or both. The outcome can be harder to predict in such situations. None of the patients demonstrated clinical recovery after a minimum follow-up of 6 months although a larger series with longer follow up, perhaps with serial electromyography, would be required to provide more refined information on the prognostic value of EMG in vocal fold paralysis following radiotherapy. Existing studies have focused on the ability of laryngeal EMG to predict recovery of function, however no information is available on the likelihood of developing vocal fold paralysis in a patient with only electrophysiological evidence of neuropathy.

Most recent reports of vocal fold paralysis following head and neck radiotherapy have been of bilateral cases,10., 11., 12. however unilateral paralysis is probably more common.9 Unilateral paralysis typically presents with hoarseness and aspiration while bilateral paralysis can result in life-threatening airway obstruction. Our series demonstrates that patients with unilateral vocal fold paralysis following radiotherapy may also have subclinical neuropathy of the contralateral vagus nerve. To our knowledge this is the first report in the literature that has studied the electrophysiological status of the laryngeal nerves following radiotherapy. Although it is possible that there is a background incidence of subclinical neuropathy in our population, we feel this is likely to be low as our experience with laryngeal electromyography for vocal fold paralysis from other causes is that contralateral changes are not typically seen. The presence of contralateral changes is significant as it may indicate an increased risk of developing bilateral vocal fold paralysis. Although none of our patients has developed this problem so far, long-term follow-up seems prudent as the onset of paralysis can take many years.9., 10., 11., 12. This is especially important if medialization laryngoplasty is being considered because this may have the effect of narrowing an already potentially compromised airway.32

CONCLUSIONS 

Vocal fold paralysis is a possible late complication of neck irradiation. Injury to the recurrent laryngeal nerve may occur at the skull base or more distally along the nerve. In unilateral vocal fold paralysis following radiotherapy, subclinical contralateral neuropathic changes may be present. The implication is that these patients may be at risk of developing bilateral vocal fold paralysis. Long-term follow-up is recommended, especially if medialization thyroplasty is being considered.