Approach to diagnosis of vocal fold immobility: a literature review

Vocal fold immobility (VFI) is a sign of disease and not a final diagnosis. VFI can have a profound impact on a patient′s quality of life. The most important objective in evaluating a patient with VFI is to exclude the existence of a treatable and potentially life-threatening primary disease as the cause of VFI. Thorough evaluation of these cases is mandatory for proper decision-making and planning of therapy. This paper addresses the established diagnostic workup for VFI and critically evaluates the contribution of different modalities to VFI diagnosis. To achieve this goal, a comprehensive review of the available literature regarding the diagnostic approaches to VFI was conducted. Eligible studies were identified by searching PubMed, Google Scholar, Springer, and ScienceDirect databases for relevant articles by combining the MESH heading term ′vocal fold immobility′ with the words ′diagnosis, evaluation, paralysis, electromyography, imaging′.


Introduction
Normal laryngeal physiology depends upon highly coordinated motor function, and disruption of this elegantly balanced system by vocal fold (VF) motion impairment may aff ect any of the vital laryngeal functions, including respiration, swallowing, phonation, and cough production [1]. VF mobility can be aff ected by disorders of the cricoarytenoid joint, the parts of the brain and nerves that supply the larynx, or the muscles of the larynx [2].
Vocal fold immobility (VFI) is the term that describes restricted movement of VFs secondary to mechanical fi xation or neurological involvement. Mobility of the VFs may be decreased or absent, and it may be unilateral or bilateral. From the standpoint of etiology, choice of treatment, and prognosis, it is important to diff erentiate between hypomobility and immobility, as well as between unilateral and bilateral involvement [3].
VFI is a sign of disease and not a fi nal diagnosis, and its exact cause should be actively investigated in all patients because its etiology determines its prognosis and management [4]. Studies [5][6][7] have suggested that nonlaryngeal neoplasms may be the most common cause of unilateral VFI.
Over 1500 studies on vocal fold paralysis (VFP) exist in the medical literature, although only a small percentage report on the diagnostic evaluation to investigate the underlying cause [8]. It is clinically important to diagnose the primary disease in cases of VFI because many of its potential causes, such as symptom-free malignant tumors, can be fatal or may cause serious morbidity if detected late [9].
Th e diagnostic approach for evaluation of the patient with VFI diff ers among clinicians, and there is a lack of consensus regarding the most appropriate initial workup of these patients. MacGregor et al. [10] surveyed ENT surgeons in the UK and found great diff erence in the diagnostic evaluation of patients with unilateral VFI.
Ideally, the evaluation should not only detect or exclude the underlying etiology but should consider cost as well [11]. Terris et al. [12] surveyed otolaryngologists and found a signifi cant diff erence in the cost of evaluation between those with less than 5 years of experience and those with more than 25 years of experience ($2300 vs. $300, respectively).
Th e purpose of the present paper was to critically address the diagnostic procedures of VFI and evaluate the contribution of diff erent modalities to VFI diagnosis to provide a rational basis for the implementation of appropriate therapeutic intervention.

Anatomy of the vagus nerve
Th e vagus nerve leaves the medulla of the brainstem as several rootlets. Th ese converge into a single root that leaves the skull through the jugular foramen. Th e vagus nerve presents two enlargements: a superior (jugular) ganglion caudal to the jugular foramen and an inferior (nodose) ganglion at the level of the hyoid bone. In the Approach to diagnosis of vocal fold immobility: a literature review Omayma E. Afsah Vocal fold immobility (VFI) is a sign of disease and not a nal diagnosis. VFI can have a profound impact on a patient's quality of life. The most important objective in evaluating a patient with VFI is to exclude the existence of a treatable and potentially life-threatening primary disease as the cause of VFI. Thorough evaluation of these cases is mandatory for proper decisionmaking and planning of therapy. This paper addresses the established diagnostic workup for VFI and critically evaluates the contribution of different modalities to VFI diagnosis. To achieve this goal, a comprehensive review of the available literature regarding the diagnostic neck, the vagus nerve has several branches that control the voice and speech mechanism. Th ese branches are the pharyngeal nerve, superior laryngeal nerve (SLN), and recurrent laryngeal nerve (RLN) [5,6].
Th e right RLN leaves the vagus nerve at the anterior surface of the right subclavian artery and runs inferiorly, looping around the subclavian artery, and ascending medially in the tracheoesophageal groove. Th e left recurrent nerve branches more caudally from the vagus nerve on the anterior surface of the aortic arch. It passes inferiorly around the arch through the aortopulmonary window to ascend in the tracheoesophageal groove [7]. Th e left RLN has a longer intrathoracic course than the right nerve, coming into contact with the mediastinal surface of the left lung, continuing along the mediastinal lymph nodes, and fi nally looping around the aortic arch [13]. As each recurrent nerve ascends, it is closely related to the corresponding inferior thyroid artery [7].

Etiology of vocal fold paralysis
VFP is usually classifi ed by either the site of the lesions (supranuclear, bulbar, peripheral nerve, myoneural junctions, or laryngeal muscles) or the nature of the disorder (infl ammatory, neoplastic, traumatic, postsurgical, systemic, or idiopathic) [14]. Lesions of the cortex or supranuclear corticobulbar pathways uncommonly aff ect the larynx. Lesions from the medulla to the muscle are more common and generally produce a fl accid paralysis [14]. VFP may be unilateral or bilateral, congenital or acquired [6,15]. On the basis of the level of the lesion, the etiology of VFP can be classifi ed into the following.

Supranuclear lesions
Bilateral upper motor neuron lesions result in pseudobulbar palsy.

Nuclear lesions
Th e causes of nuclear lesions are vascular disease, neoplastic disease, motor neuron disease, polio, and syringobulbia. Lesions of the nucleus ambiguous result in ipsilateral palatal, pharyngeal, and laryngeal paralysis that is usually associated with aff ection of other cranial nerve nuclei, roots, and long tracts. Th e most common cause is occlusion of the posterior inferior cerebellar arteries resulting in a lateral medullary infarct (Wallenberg syndrome). When only the more cephalad portion of the nucleus ambiguous is injured, laryngeal function is spared (palatopharyngeal paralysis of Avellis) [16].

High (proximal) vagal lesions
Th ese refer to lesions from the nodose ganglion up (damage to the vagus nerve before it innervates the soft palate). Cranial nerves IX, XI, and XII travel with the vagus in its uppermost segment; therefore, lesions aff ecting the vagus superiorly usually (95%) aff ect one or all of these cranial nerves as well [17]. High vagal lesions include the following: (1) Posterior cranial fossa and jugular foramen lesions.
Cranial nerves IX, X, XI, and XII are all included in the contents of the suprahyoid carotid space (carotid sheath and adjacent structures). A wide range of abnormalities, including benign and malignant tumors, infl ammatory processes, and vascular lesions, can result in vagal nerve involvement with subsequent VFP. Examples of these lesions include paragangliomas/glomus vagale, vagal schwannomas, and metastatic lymph nodes [19].

Low (distal) vagal lesions
Th ese refer to lesions in the infrahyoid neck and mediastinum. Damage is below the level of the pharyngeal branch of the vagus and therefore the soft palate is spared. Th e vagus nerve in the carotid space can be aff ected by vagal schwannoma or neurofi broma. Th e close proximity of the RLN to the esophagus, trachea, and thyroid in the tracheoesophageal groove makes it vulnerable to injury secondary to pathologies involving these structures (such as esophageal and thyroid malignancies). Th e left RLN is longer than the right RLN and has a longer intrathoracic course, which makes it more vulnerable to injury secondary to mediastinal abnormalities (such as lung carcinoma, aortic aneurysm, and mediastinal lymphadenopathy) [19,20].

Myoneural junction diseases
Myasthenia gravis causes bilateral weakness of the intrinsic laryngeal muscles. Sometimes, muscles supplied by the vagus nerve are the fi rst to be aff ected by this disease. Most often, myasthenia gravis produces the full complement of fl accid dysarthria [21].

Disease of muscle
Muscular dystrophies or myopathies can cause fl accid dysphonia [21].

Patterns of injury
Th e RLN contains 500-1000 motor axons, and all are myelinated [22]. Th e nerve trunk is a collection of fascicles. Th e epineurium covers the entire nerve, whereas the perineurium surrounds individual fascicles, and the endoneurium surrounds the nerve fi bers [23].
Nerve injuries are categorized according to morphological alterations, nerve functionality, and the ability for spontaneous recovery. Th ere are two commonly used classifi cation schemes for peripheral nerve injury: the Seddon [24] and the Sunderland [25] ( Table 1). Th e Sunderland classifi cation is more complex, but more useful.
Seddon [24] divided injuries into neurapraxia, axonotmesis, and neurotmesis. Neurapraxia results simply in a conduction block along the nerve fi ber. Th e macroscopic mechanism is generally compression or excessive stretching. Axons are nonfunctional because of segmental demyelination, without axon disruption or Wallerian degeneration. In axonotmesis, the nerve fi ber and its myelin sheath are cut, without interruption in the neural envelopes. It is generally caused by violent compression or stretching. Nerve conduction is interrupted and the nerve fi ber undergoes retrograde degeneration (Wallerian degeneration) beginning within 24 h of injury. Neurotmesis is the interruption of nerve fi bers, myelin, and nerve envelopes due to nerve section or severe crush (e.g. at least the endoneurium and the perineurium, in crush injury). Wallerian degeneration occurs [27].
Sunderland [25] divided axonotmesis into grades 2-4 depending on the extent of nerve aff ected by axonal injury. Given that axonotmesis represents a broad spectrum of injury, the functional outcome can vary but is generally with little synkinesis [28].

Recurrent laryngeal nerve regrowth
For neurapraxic lesions, spontaneous full recovery follows remyelinization by Schwann cells. Th is occurs after several days to several weeks [27]. For axonotmesis with intact endoneurium, recovery ensues because regenerating axons enter their native endoneurial tubes leading back to original target muscles. In more severe injuries, reinnervation may be inappropriate, inadequate, or nonexistent. Axons may grow in a misdirected manner. Adductor fi bers may innervate abductor muscles and vice versa. Th e simultaneous contraction in antagonist muscles that results is called 'synkinesis'. An inadequate number of regenerating axons successfully reach their endoneurial conduits. Voluntary contraction is never attained, but reinnervation is suffi cient to prevent denervation atrophy. Regrowth of the envelopes may occur in a chaotic manner, resulting in a neuroma. Muscle degeneration is thought to be complete after 2-3 years [29].
Unless a result of severe trauma like transection, each case diff ers in terms of degree of neural impairment, and features a mix of injury types among its nerve fi bers with variable degrees and patterns of reinnervation [30].
Th us, VFP is probably best considered as a continuum of neurogenic dysfunction encompassing partial denervation, complete denervation, and variable degrees and patterns of reinnervation. Th is has two important clinical implications. First, it accounts for variability in the position of the paralyzed VF. As stated by Kotby et al. [31], the immobile VF may adopt several positions not only dependent on the type of the nerve muscle aff ection but also dependent on the degree of that aff ection, and on the anatomical peculiarities of the muscles and joints. Second, it accounts for the general trend for the voice to improve over time in unilateral vocal fold paralysis (UVFP). Reinnervation likely acts to restore or preserve muscle bulk and tone, leading to a voice close to normal, although the VF itself remains immobile [32,33].

Assessment of vocal fold paralysis
Among the patients who present with VFP as the initial sign, the frequency of detection of a malignant tumor as the primary lesion is relatively high. Hence, it is clinically important to diagnose the primary disease in cases of VFP [34].
On the basis of the protocol for assessment of voice disorders [35], diagnosis of VFP proceeds through the following steps: elementary diagnostic procedures, clinical diagnostic aids, and additional instrumental measures.

Elementary diagnostic procedures Key features of the history
A thorough history of complaints has to be obtained, including voice, swallowing, and breathing symptomatology. Th e voice can vary from simple vocal fatigue in mild or well-compensated cases to almost complete aphonia in severe cases.
Swallowing diffi culties are often encountered, specifi cally aspiration of liquids in the acute phase of VFP, but complete compensation is said to develop later [36]. Some dysphagia for solids may also be present, especially in brainstem or high vagal injuries, due to the concomitant denervation of the pharyngeal constrictors. Risk of aspiration is heightened in these instances as well, because of the loss of ipsilateral laryngeal sensation from SLN involvement [37].
Kashima [38] reported that over half of the patients with UVFP suff er from at least mild dyspnea. In bilateral abductor paralysis, the most serious symptom is respiratory obstruction, which is evidenced by stridor, dyspnea, retraction, and nasal fl aring [15].
A complete medical history should be taken, including a thorough history of neurologic or rheumatologic disease, previous surgeries, prolonged intubations, trauma, pulmonary symptoms suggestive of tuberculosis or malignancy, and exposure to neurotoxic agents like solid tumor chemotherapy (e.g. vincristine, vinplastine, and cisplatin) [39].

Auditory perceptual assessment
Th e abducted VF in UVFP results in a voice quality that is characterized by breathiness, diplophonia, decreased loudness, reduced phonation duration, and a restricted pitch range [40]. Th e voice is breathy and weak because of incomplete glottic closure and subsequent air escape. Th e voice may also have a watery or 'gurgly' quality if secretions are retained in the pyriform sinus, as is typical in high vagal injuries. Diplophonia may occur when the immobile VF is at a diff erent tension from that of the contralateral side, resulting in each VF vibrating at a diff erent frequency. Supraglottic hyperfunctional compensation is common, leading to irregular, low pitched voice. In contrast, other patients, often women, may develop an unnaturally high-pitched voice that is breathy in quality, referred to as a 'paralytic falsetto'. It is a common fi nding in patients with RLN paralysis caused by compensatory contraction of the ipsilateral cricothyroid muscle, which remains innervated. Th e presence of this phenomenon suggests an intact SLN [41].
Th e phoniatrician should also listen for disturbances in articulation, prosody, and resonance, which would raise suspicion for a proximal neurologic injury or underlying neurologic disorder [42].

Physical examination
General, neurological, chest, and heart examinations should be conducted. A thorough and systematic neurologic examination of the head and neck should be performed for all patients with UVFP [8]. Th e neck should be carefully examined for thyroid enlargement, masses, lymphadenopathy, and surgical scars. Examination of cranial nerves should be done with special attention to the spinal accessory and hypoglossal nerves, which share the jugular foramen with the vagus. Involvement of these adjacent cranial nerves warrants a thorough radiographic evaluation of the base of the skull. Vocal tract examination to determine the patient's gag refl ex, palatal elevation, and uvular deviation to evaluate the vagus nerve function are carried out. Palatal paralysis (the uvula deviates to the intact side) in combination with ipsilateral VFP may indicate a 'high' vagal lesion. Examination of tongue mobility is also important for possible associated hypoglossal nerve lesions [37]. Indirect mirror laryngoscopy should be performed for identifying gross VF mobility and pooled secretions.

Clinical diagnostic aids
1 Documentation of auditory perceptual assessment by high-fi delity video recording. 2 Documentation of visual assessment using rigid laryngoscopy or fl exible transnasal laryngoscopy.
Laryngeal examination is best accomplished with transnasal fl exible laryngoscopy, because this enables observation of pharyngeal and laryngeal motor function. Although rigid laryngoscopy enables excellent visualization of the endolarynx, the pharynx and soft palate are not examined in as much detail with this technique and tongue traction required may alter the laryngeal posture and lead to inaccurate assessment of laryngeal function. Th e examiner should look for asymmetric movement, VF bowing, horizontal and vertical position of the VFs, glottic gap on phonation, presence of prolapsed arytenoids, and supraglottic hyperfunction.
Th e most consistent glottic fi ndings in UVFP are a shortened and bowed VF [32]. Th e position of the paralyzed VF carries no signifi cance with respect to the site of injury or prognosis. Th e examiner should take care not to be misled by small amounts of VF motion that may be caused by the interarytenoid muscle still partially innervated from the contralateral nerve or by an intact cricothyroid muscle [37]. In such cases, there may appear to be slight adduction on phonatory eff ort, but the VF will not abduct from its position of rest.
Th e vertical level of the paralyzed VF may be either lower or higher than the normal VF or it may show the same horizontal level as does the normal VF during phonation ( Fig. 1) [43]. Th is fi nding should be considered when planning phonosurgery to correct dysphonia due to UVFP.
In UVFP, the glottal gap may be of two principal confi gurations: spindle-shaped (involving principally the membranous portion of the VF) or V-shaped (marked by greater distance between the vocal processes of the arytenoids cartilage).
Th e presence of a prolapsed arytenoid suggests profound denervation with loss of muscular support for the cartilage. Th e reasons for this malposition include: (1) Residual activity of intrinsic laryngeal muscles unaff ected by the paralysis, (2) Passive action of ligaments within the larynx, (3) Reinnervation of the aff ected laryngeal muscles in the form of synkinesis, and (4) Restriction of joint movement possibly by intrajoint infl ammation and fi brosis [44].
Th is overhanging, anteriorly displaced arytenoid cartilage is sometimes mistaken for an arytenoid cartilage dislocation (Fig. 1). Electromyography (EMG) data from patients with an immobile VF accompanied by an 'overhanging' arytenoid cartilage, however, will almost always show complete denervation or poor reinnervation of the thyroarytenoid (TA) muscle. Th erefore, in nontraumatic cases, anterior displacement or an overhanging, sagging arytenoid cartilage should not raise the suspicion for arytenoid dislocation (Fig. 2) [45].
In longstanding UVFP, supraglottic hyperfunction may obscure visualization of the VFs. Maneuvers such as humming can serve to relax the ventricular folds (unloading the larynx) to permit a more thorough evaluation of the glottis closure [46].
SLN damage is tricky to identify. Th e controversial list of laryngoscopic signs include an obliquely shaped glottis, rotation of the posterior commissure toward the weak side, lower VF height on the involved side, and bowed, thin and shortened VF.
In addition to the larynx, the following structures are examined during fl exible laryngoscopy: the soft palate for velopharyngeal closure; and the tongue base, pharynx, and hypopharynx for asymmetry, abnormal spontaneous movements, and pooling of secretions in the pyriform sinus and vallecula on the aff ected side [47]. Loss of sensation in the hypopharynx secondary to SLN compromise can result in pooling of The contact nature between paralyzed and normal vocal folds (VFs) during phonation. Equal contact level (left) and higher normal VF overlapped on lower paralyzed VF (middle) and lower normal VF (right). Arrows show paralyzed VF [43].

Figure 1
Laryngoscopic view of a patient with left vocal fold immobility and an overhanging arytenoid. Laryngeal electromyography con rmed that the immobility was caused by a neuropathic process, and not by cricoarytenoid joint dislocation or subluxation [45]. secretions in the pyriform sinus and delayed relaxation of the cricopharyngeus (Fig. 3).

3-Stroboscopy and videostroboscopy
Videostroboscopy has been the standard tool for the evaluation of VF vibration in patients with UVFP [48]. During VF vibration observed in stroboscopy, the denervated side shows wide undulating amplitudes like the fl uttering of a fl ag in the wind (although amplitude could be radically reduced) and a loss of mucosal waves. Asymmetry in phase and amplitude is obvious between the normal and denervated side [49].
Studying the mucosal waves by means of stroboscopy may be recommended for monitoring the course of laryngeal paralysis. When the VFs were shown to be denervated in the EMG, there was a complete lack of mucosal waves [50]. A return of the mucosal waves was always correlated with signs of reinnervation in the EMG. Most frequently, however, the VFs stayed immobilized for the entire observation time of more than 1 year, despite the fact that the EMG showed good reinnervation and the vibratory pattern was completely normalized with excellent mucosal waves. Fex and Elmqvist [50] explained this fi nding by misdirected reinnervation [49].
Sercarz et al. [51] stated that stroboscopy cannot reliably distinguish RLN paralysis from vagal paralysis. Th ey also stated that, in VFP, the mucosal wave is always aff ected but not invariably absent in RLN and vagal paralysis. Th ey explained the absent mucosal waves in some patients by poor glottic closure that reduced the degree of VF contact, decreasing the ability of stroboscopy to detect the very subtle mucosal wave that generally occurs on the paralyzed VF.

4-High-speed laryngeal imaging
High-speed laryngeal imaging has evolved as a method of laryngeal visualization that overcomes many of the limitations of videostroboscopy [52][53][54]. A method for automatic diagnosis of VFP has been developed by use of image analysis technology based on highspeed imaging [55]. Although high-speed imaging off ers many exciting possibilities in the evaluation of dysphonic patients, the expense of the equipment prevents widespread use in everyday practice, and its clinical utility in VF motion impairment remains to be seen. Certainly, in most cases of UVFP, standard fl exible laryngoscopy is adequate for diagnosis and treatment planning. Techniques such as videostroboscopy and high-speed imaging may be helpful for evaluation of more subtle cases of VF paresis and for monitoring the vibratory function of the VFs before and after treatment [56].

5-Videokymography
Videokymography (VKG) is a high-speed method for examination of the VF vibrations. Th e major advantages of VKG over high-speed imaging systems are lower cost, excellent spatial resolution, excellent image rate, unlimited duration of recordings, and fewer data to be stored and processed [57]. VKG fi ndings in VFP include the following: a value of zero for closed phase and closed quotient; increase in open phase; increased opening and closing time of the paralyzed VF; decreased amplitude on the paralyzed VF; asymmetry in amplitude and in the timing of opening and closing phases [58]. Th e disadvantages of VKG, compared with high-speed imaging systems, arise mainly because only a single image line is monitored in VKG [57]. To compensate for the lack of a full image, it appears useful to use VKG as a complementary method to videolaryngostroboscopy (Fig. 4).

6-Electroglottography
Electroglottography (EGG) contributes information about the quality and duration of VF contact in patients with VFP and allows documentation and objective evaluation of consecutive examinations [59].
In patients with VFP, at least one of the following characteristics of EGG waveforms can be observed: prolonged closing phase, shortened opening phase, and prolonged open phase (Fig. 5) [60].
Zagólski [59] indicated that % irregularity (which is a frequency perturbation measure) best represents the changes in VF function in elderly individuals with VFP. Mueller [61] stated that EGG waveforms recorded in patients with VFP should be interpreted with great caution, particularly in elderly individuals, who may have considerable vocal perturbations even when VF mobility is unimpaired.
Variable laryngoscopic ndings for two individuals with left vocal fold paralysis. Laryngoscopy for patient A (a) reveals anterior displacement of the arytenoid, accidity of the true vocal fold (VF), and salivary pooling in the left piriform sinus. In contrast, patient B (b) has a fairly upright arytenoid with atrophic VF with only minimal accidity [1].

Figure 3 b a
Owing to the lack of generally accepted norms for the interpretation of EGG waveforms, Zagólski [59] mentioned that the data and waveforms should be analyzed and compared for each individual and always related to other clinical data for that person.

Additional instrumental measures A-Acoustic and aerodynamic measures
Th e acoustic and aerodynamic measurements provide an objective qualifi cation of voice in patients with UVFP [62]. Eskenazi et al. [63] compared six acoustic parameters with perceptual evaluation by experienced jury listeners using the GRBAS method on a scale from 1 to 7. Regression analysis showed that jitter was well correlated with grade G (overall dysphonia) and grade R (roughness), whereas signal-to-noise ratio was correlated with grade B (breathiness). In their study on patients with UVFP, Hartl et al. [62] stated that breathiness was best correlated with airfl ow measurements.
Th e acoustic fi ndings that were observed in patients with VFP included reduction of dynamic range both in loudness and frequency, loss of high-frequency harmonics and high-frequency formants on the spectrogram, increased aperiodicity of the vibratory period (increased jitter), and increased noise-toharmonic ratio [42].
Th e aerodynamic fi ndings observed in patients with VFP included reduction of maximum phonation time (generally<10 s) and increased mean airfl ow. Subglottic air pressure is usually diminished, the person being unable to raise the pressure below the VFs because of the poor glottic closure [42].
Th e major advantage of the acoustic and aerodynamic measurements is that they permit a longitudinal followup of each case by using these measures in pretherapy and post-therapy evaluations, whether the therapy was voice therapy or surgery. Netterville et al. [64] noted that postoperative measurements should be performed after about 3 months from surgery.
Yu et al. [65] stated that the use of a single acoustic or aerodynamic parameter for objective assessment of dysphonia is not appropriate. Acoustic measurements cover only part of the information contained in perceptual analysis. For this reason several teams have proposed a multiparametric approach comprising acoustic and aerodynamic parameters to enhance the scope of data.

B-Laboratory examinations
Laboratory testing has been described as part of the evaluation of patients with UVFP without a clear etiology. Such tests include rheumatoid factor, Lyme titer, erythrocyte sedimentation rate, antinuclear antibody, angiotensin-converting enzyme, syphilis testing, and blood chemistry analysis including  An example of a electroglottography waveform recorded in a patient with right vocal fold paralysis. Prolonged closing and opening phase. Duration of vocal fold (VF) contact differs between the cycles. This waveform shows periodicity but cycles of unequal length, amplitude, and contact irregularity re ecting abnormal VF contact patterns [59]. glucose [8]. Although results from these studies may be abnormal for medical diseases associated with UVFP, no studies support their routine use in the absence of a high clinical suspicion for a particular disease [12].

C-Laryngeal electromyography Approach
For standard clinical purposes, the percutaneous approach proved to be the most suitable in the hands of many investigators [66,67].

Muscles investigated
Th e TA and the cricothyroid muscles on each side are investigated, as a representative of the recurrent and SLN-muscle systems. Kotby and Haugen [66] suggested that the fi rst sitting of laryngeal electromyography (LEMG) should include recordings from the posticus muscle [31].

Activities recorded
EMG recordings from the laryngeal muscles are obtained at rest (during quiet breathing), during deep breathing, and during phonation at comfortable pitch and loudness levels. EMG recordings also are obtained with diff erent sphincteric functions of the larynx (straining, coughing, and swallowing). Th e latter recordings are usually done last to avoid displacement of the laryngeal electrodes [31].

Guidelines for interpretation of laryngeal electromyography
(1) At rest: Th e constant background postural activity of the laryngeal muscles makes it hard to interpret muscle activity triggered by inserting the needle. Equally, detecting pathological spontaneous activity is much more diffi cult than in most other muscles. At the same time, a lesion of the nerve also reduces the background activity and thus often makes the interpretation easier than in physiologically innervated laryngeal muscles [68]. (2) Interference pattern: In laryngeal movements, unlike peripheral skeletal muscles, there is little chance for gradation of the degree of contraction. Accordingly, the comparison between partial versus full interference patterns is diffi cult. In unilateral neurogenic lesions of the larynx, the detection of a signifi cant diff erence in the interference pattern in identical muscles on both sides of the larynx can be considered as pointing to a neuropathic lesion of the muscle. (3) Motor unit action potentials (MUAPs) of laryngeal muscles are normally biphasic or triphasic, of short mean duration (2-4 ms), and of small amplitude (150-800 v) [31].

Value of laryngeal electromyography in vocal fold immobility 1-Diagnosis Paralyzed vocal fold versus xed vocal fold
LEMG is an important diagnostic tool for this purpose. Timing of LEMG should be at least 14 days after onset of immobility, to enable Wallerian degeneration to occur and spontaneous activity to emerge, and is most useful within 6 months of injury [69]. Normal electrical activity patterns of the LEMG support a diagnosis of arytenoid fi xation [70], whereas abnormal electrical activity patterns, including patterns of denervation or reinnervation, support the diagnosis of VFP [71].
Th e following are the signs of denervation: (1) Spontaneous activity during rest. Spontaneous activity can include fi brillation potentials, increased insertional fi brillations (insertional activity), myotonic discharges, complex repetitive discharges, fasciculations, and positive sharp waves. (2) Spontaneous activity may occur only 10-14 days after the injury and such spontaneous activity implies that the muscle is degenerating or that the nerve has been injured and that the injury is ongoing.  [68].
Th e detection of spontaneous fi brillation potentials and the reduction in the number of motor units on activation (refl ected through interference patterns) are some of the most reliable features in the diagnosis of neuromuscular pathologies (especially neurogenic lesions). Th ese features cannot be used reliably in clinical LEMG. Accordingly, one is apt to rely heavily on the features (parameters) of the MUAPs. Th ese MUAPs are known to show an increase in mean duration and voltage in cases of neurogenic lesions [31].
Th e following are the signs of reinnervation: (1) Polyphasic units are large-amplitude units with fi ve or more baseline crossings of increased duration compared with normal units. Nascent units are defi ned as low-amplitude units with increased number of baseline crossings and do not meet the amplitude or duration criteria of polyphasic units. Th ey can be observed at the earliest 2 months after onset and indicate ongoing reinnervation (recent injury) [72].
(2) In later stages (5-12 months after onset of nerve degeneration), spontaneous denervation potentials progressively disappear. Th e polyphasia and duration decrease but the amplitude increases, producing a giant wave. Th ese large-amplitude motor units indicate an old but stable peripheral nerve injury [73].
LEMG interpretation is based upon recruitment, waveform morphology, and spontaneous activity. Th ese variables off er fi ve interpretations ( Table 2).

Upper motor neuron versus lower motor neuron
In the fi rst 3-6 weeks after an upper motor neuron injury, there may be abnormal spontaneous activity.
After that time, examination shows electrical silence at rest with no abnormal spontaneous activity. However, EMG is neither sensitive nor specifi c in the diagnosis of upper motor neuron lesions [74].

Vagus versus recurrent laryngeal nerve
Comparison of fi ndings between the cricothyroid and the TA indicates the site of lesion. If there are neuropathic fi ndings in both muscles, a high injury is suspected, and imaging of the skull base and central nervous system is warranted. If there are abnormalities in only the TA, the RLN is implicated, directing attention to the lower neck and mediastinum [74].

2-Prognosis
For neurapraxia, the diagnostic criterion in LEMG is the detection of a rarefi ed recruitment pattern or of single action potentials during voluntary contraction without pathological spontaneous activity (Figs. 6 and 7). Axonotmesis should be suspected if spontaneous activity, indicating neural degeneration, is detected (Figs. 8 and 9). Th is classifi cation includes a Neurapraxy. Note the extremely rare ed ring pattern on phonation but no spontaneous activity at rest [75].

Figure 8
Neurapraxy. Note the slightly reduced recruitment on phonation, with no spontaneous activity at rest [75].

Figure 7
certain level of prognosis on recovery, as neurapraxia is most likely to recover completely within 8-12 weeks, whereas axonotmesis is thought to have only a poor chance of recovery to a functional level [76]. If reinnervation occurs following axonotmesis, it is usually associated with sequelae, such as synkinesis, due to neuronal misdirection [68]. In neurotmesis, EMG shows fi brillation potentials and then electrical silence.
For evaluation of synkinetic reinnervation, the interference pattern of the TA muscle is evaluated during vowel vocalization and then during sniffi ng. If synkinesis has occurred, there will be an electrical signal with both tasks [68].
EMG studies can detect early stages of recovery of partial laryngeal nerve lesion far before clinical detection of return of movement in the immobile VF. Th e increase in the number of MUAPs with an increasing percentage of polyphasic potentials denotes early muscle reinnervation [66].
A meta-analysis including 503 patients that evaluated the utility of LEMG for accurate prediction of prognosis of motion recovery found that the LEMG was able to predict poor prognosis with 91% accuracy but was only able to predict good recovery 56% of the time [69].
Th e LEMG has its limits. Clinical experience has shown that the results of LEMG are often inconclusive. Fibrillation potentials (denervation pattern) are diffi cult to detect because of competing noise from neighboring muscles and because of the small size of the laryngeal muscles. A majority of patients with RLN paralysis do not have electrical silence in the early days after nerve trauma, because the nerve lesion is incomplete. Normal interference patterns can be recorded, whereas laryngeal motion is absent for the same reason. Patients with reduced interference pattern or with polyphasic potentials certainly have a nerve lesion, but unfortunately the likelihood of recovery cannot be predicted. If recovery has to take place, it is usually observable in the fi rst 3 months and rarely after 1 year following onset of paralysis [42].

D-Radiologic studies
When a clear-cut temporal relation exists between surgical iatrogenic trauma and VFP, no additional radiologic workup is necessary. When no cause can be found for the VFP, imaging studies are essential [12].

Chest X-ray versus computed tomography scan
Terris et al. [12] found that chest X-ray (CXR) was the most useful diagnostic tool as it identifi ed a diagnosis in more than one-third of patients with previously undiagnosed UVFP. Altman and Benninger [11] stated that CXR is a useful screening study, but it is not suffi cient for the full evaluation of VFP. In their study, there were three cases in which history, examination, and CXR failed to identify the etiology. In all of these cases, a computed tomography (CT) scan of the neck identifi ed the etiology: esophageal cancer in one case and thyroid cancer in the other two. In the study by Glazer et al. [7], CXR missed 13 of 18 mediastinal lesions causing left UVFP, whereas history, examination, and CXR identifi ed all lesions causing right UVFP. Song et al. [77] found that, of 19 CT-detected malignant neoplasms, eight were not detected on CXR. Five of them were associated with left-side paralysis and showed a metastatic node in the aortopulmonary window on CT. Th e other three were associated with right-side paralysis, and the CT scan detected one case of right apical lung carcinoma and two cases of metastatic nodes in the right upper paratracheal region. Th e proposed algorithm by Altman and Benninger [11] was to use CXR as a screening tool; if this is negative, then a contrast-enhanced, 3-mm section, CT scan of the neck should follow. Th eir description CT of the neck extends from 'the posterior fossa/skull base through the aortic triangle' on the left and to the 'thoracic inlet' on the right, making it a CT along the course of the vagus and RLN, and not purely a neck CT [8].
Furukawa et al. [34] also advocated a 'side-specifi c' evaluation that included ultrasound of the neck and CXR; if these initial tests were not revealing, a chest CT should follow for left-sided paralysis only. Th e same authors also noted a sex distinction with a higher rate of lung and esophageal cancer in male Polyphasic, prolonged action potentials with giant amplitudes, indicating reinnervation 4 months after nerve injury [75]. patients and for thyroid cancer in female patients; this sex distinction became part of their paradigm with only male patients undergoing contrast esophagrams as a follow-up to the ultrasound and CXR for rightsided paralysis.

Neck ultrasonography
High-resolution ultrasonography can evaluate the thyroid gland, supraclavicular region, the entire jugular chain, and the cervical vagus nerve. It is useful for detecting subclinical neoplasia causing UVFP. Some thyroid tumors have deep location in the posterior part of the thyroid lobe, resulting in UVFP being the initial presentation rather than a palpable anterior neck mass.
Neck ultrasonography is the most sensitive and accurate imaging study for the diff erential diagnosis of thyroid tumors. If a suspicious tumor lesion is noted, fi ne-needle aspiration with cytologic examination should be performed immediately under ultrasound guidance [78].

Computed tomography versus magnetic resonance imaging
Jacobs et al. [79] have suggested a segmented approach to clinical-radiological examination of vagus nerve dysfunction, based on proximal and distal lesions. CT is the fi rst-line choice for imaging investigation of distal vagus neuropathy. Contrastenhanced images from the hyoid bone to the mediastinum should be obtained. In contrast, MRI is the fi rst-line choice for imaging investigation of proximal vagus neuropathy. It is useful in patients with associated palatal or pharyngeal paralysis and/or other cranial neuropathies. In such cases, gadoliniumenhanced MRI of the posterior cranial fossa should extend from the medulla to the hyoid bone. If MRI is unavailable, thin-section, contrast-enhanced CT can be used, although brainstem lesions may not always be detected on CT scans [76].

Summary of diagnostic imaging
It is diffi cult to make a practice recommendation on radiographic investigation of idiopathic VFP based on scientifi c evidence. Th e importance and potential superiority of CT scanning was advocated by some authors (e.g. Glazer et al. [7] and Song et al. [77]). Several other authors advocated that patients with UVFP be triaged to diff erent imaging modalities based on the side of the lesion (e.g. Altman and Benninger [11] and Furukawa et al. [34]) or the clinical suspicion of the ordering physician (e.g. Jacobs et al. [79]). Th is clinical distinction was made on clinical grounds such as accompanying cranial polyneuropathy. Th e advantage of the 'proximal' and 'distal' distinction may have its basis in the preference for MRI in skull base and intracranial imaging [8]. Ko et al. [4] proposed an algorithm for patients with UVFP (Fig. 10).
Richardson and Bastian [76] outlined a cost-eff ective and time and labor-effi cient method for the clinical evaluation of VFP, including a focused history, vocal capability assessment to fi nd defi cits in the function of the palate, pharynx, and larynx, and, fi nally, an intense examination under topical anesthesia to demonstrate these defi cits. In essence, it is the endoscopic version of a radiographic study from the skull base through the aortic arch. Th is method is streamlined as compared with prior protocols for evaluation of VFP, because it directs the necessary further workup according to the likely site of the lesion as indicated by the extended physical examination, and can be conducted entirely in the physician's offi ce. Radiographic workup should include CT of the skull base through the upper mediastinum if solely an RLN paralysis is present; it should include MRI of the skull base if high vagal signs and symptoms are present. If MRI is negative, CT may also be needed for complete evaluation. Neurologic signs that are not all ipsilateral require MRI of the brain and consultation with a neurologist. Esophageal obstruction combined with VFP mandates evaluation with esophagoscopy or an esophagram.

Figure 10
Approach to diagnosis of vocal fold immobility Afsah 89

E-Formal evaluation of swallowing
Modifi ed barium swallow is the recommended method for evaluation of swallowing complaints, although fl exible endoscopic evaluation of swallowing (FEES) and FEES with sensory testing (FEESST) have emerged as techniques that enable safe and convenient evaluation of swallowing that can be performed in the offi ce [47]. Patients with UVFP frequently complain of dysphagia, particularly aspiration of liquids. A persistent glottic gap, associated sensory defi cits, pharyngeal weakness, and delayed relaxation of the cricopharyngeus muscle may all contribute to dysphagia in UVFP cases [1].
In the study by Jang et al. [80], patients with VFP of central etiology showed a higher incidence of penetration and aspiration as compared with patients with peripheral etiology. In addition, they also demonstrated other oral and pharyngeal phase swallow problems -for example, delayed triggering of pharyngeal swallow, reduced pharyngeal peristalsis, and cricopharyngeal dysfunction.

Differential diagnosis of vocal fold paralysis Dislocation/subluxation of the cricoarytenoid joint with ankylosis
It is infrequent but presents following medical instrumentation used on the larynx and esophagus, and because of external neck trauma. Subluxation is joint dislocation with some remaining contact of the joint surface. Laryngoscopy is the most common fi rst step in diagnosis. Clues that can raise the suspicion of an arytenoid cartilage dislocation when evaluating unilateral VF mobility include arytenoid cartilage edema, diff erence in VF level, and absence of a 'jostle sign', which is a brief lateral movement of the arytenoid cartilage on the immobile side during glottic closure caused by contact from the mobile arytenoid [81].
Disparity in height between the vocal processes is much easier to see in slow motion under stroboscopic light at various pitches. A posteriorly dislocated arytenoid results in a high vocal process and a stretched VF. An anterior dislocation results in a low vocal process and a short VF. Ascertaining the direction of dislocation is diffi cult, but critical, because it will aff ect the technique of reduction [82].
Because none of the above is pathognomonic for arytenoid cartilage dislocation, a high degree of suspicion and further diagnostic testing are needed to confi rm this entity. Alexander et al. [83] reported the usefulness of helical CT in patients with arytenoid dislocation. Helical CT not only signifi cantly reduces the time necessary to study the larynx but enables one to perform multiple high-resolution multiplanar reconstructions. Asymmetry in the joint space, specifi cally obliteration or widening, supports the diagnosis of arytenoid cartilage dislocation (Fig. 11). Th e widened joint space is usually fi lled with soft tissue density from hemarthrosis and fi brosis. Of course, joint fi brosis from another process, such as rheumatoid arthritis or severe laryngeal refl ux, might look similar on a CT scan, but an adequate history and serologic studies should help determine the true etiology.
Rubin et al. [82] reported some limitations of CT: (i) in young patients, the cartilage is frequently not ossifi ed and therefore diffi cult to assess; and (ii) one needs highquality images with fi ne cuts and reconstructed images of the larynx, which are not available everywhere. Th ey believed that, although a positive CT scan is helpful to confi rm the diagnosis, a negative scan does not rule out dislocation. If the CT is equivocal, LEMG is useful to confi rm diagnosis. LEMG is helpful in distinguishing a neurologic from a joint abnormality as the origin of a hypomobile or immobile VF. However, it is possible for both denervation and a joint dislocation to be present. Th us, the LEMG fi ndings must be evaluated in the context of fi ndings on laryngoscopy. For example, if there is marked asymmetry in vocal process height without a jostle sign, there is probably a structural problem in addition to, or despite, the LEMG fi ndings [82]. If LEMG is not available, direct laryngoscopy with palpation of the cricoarytenoid joint shows no passive mobility.

Ankylosis of the cricoarytenoid joint -for example, in end-stage rheumatoid arthritis
Th e clinical presentation may vary from being asymptomatic to a constellation of upper aerodigestive Computed tomography scan of the larynx with left arytenoid cartilage dislocation. Compare the obliterated left joint space (thick arrow) with the intact joint space on the right (thin arrow) [82].

Figure 11
symptoms. Th e array of symptoms include odynophagia, foreign body sensation, dysphagia, sore throat, lump sensation in the throat, change in voice quality, referred otalgia, and respiratory symptoms [84]. Highresolution computed tomography is the modality of choice for assessing cricoarytenoid involvement in rheumatoid arthritis (Fig. 12) [86].
Similar to arytenoid dislocation, LEMG and direct laryngoscopy with palpation of the cricoarytenoid joint confi rm the diagnosis.

Laryngeal malignancy with involvement of the joint or the thyroarytenoid muscle
Careful fi beroptic laryngoscopy combined with a CT scan will reveal neoplastic infi ltration as the cause of VFI. LEMG may show decreased amplitude and recruitment of the TA muscle.

Posterior glottis stenosis or interarytenoid scarring
It is a common complication of prolonged endotracheal intubation. On palpation of the cricoarytenoid joint in patients with posterior glottis stenosis or interarytenoid scarring, the posterior commissure and contralateral arytenoid cartilage will be moved to the midline during the vocal process palpation because of the entire posterior glottic complex being fused with scar tissue [37].