Recurrent laryngeal nerve paralysis: Current concepts and treatment: Part I--Phylogenesis and physiology.
The aim of this review article is to summarize current concepts in unilateral recurrent laryngeal nerve paralysis (URLNP). Important aspects of laryngeal phylogenesis, physiology, and anatomy are reviewed. Recent advances in the neurophysiology of URLNP are discussed. Revised and updated principles of diagnosis and treatment are provided. Glottic configuration and prognosis vary according to the type of neural lesion (neurapraxia, axonotmesis, or neurotmesis). Therapeutic indications depend on glottic configuration and prognosis. Treatment options include medialization thyroplasty, vocal fold augmentation by injection, arytenoid adduction, and laryngeal reinnervation. Each treatment option is summarized, and the results reported in the medical literature are reviewed.
Unilateral recurrent laryngeal nerve paralysis (URLNP) results in glottic incompetence. The glottic configuration varies according to neurophysiologic laws that are as yet incompletely elucidated. The severity of symptoms such as weak and breathy voice, phonatory effort, inefficient cough, and aspiration varies as well. Aspiration in URLNP can be life-threatening in patients with compromised pulmonary function. Dysphonia diminishes oral communication and quality of life.
Therapeutic options in URLNP began to flourish toward the end of the 20th century. There are now four main approaches to vocal fold medialization: intracordal injection techniques, external medialization thyroplasty, arytenoid adduction, and laryngeal reinnervation. Each technique has specific indications and advantages. Knowledge of laryngeal anatomy, physiology, and neurophysiology is necessary to determine the optimum procedure for a given etiology and laryngeal configuration.
The aim of this review article is to relate current concepts in laryngeal physiology and neurophysiology and to summarize current therapeutic options for URLNP.
Several notions in laryngeal phylogenesis are helpful in understanding laryngeal physiology and pathology in humans. [1-4] From a phylogenetic point of view, the larynx is above all a sphincter situated at the inlet of the lower airways, protecting them from water and food penetration. The most primitive larynx, in certain fish, is simply a circular smooth muscle, contracted at rest, keeping water from entering the primitive lung sacs. These fish fill the primitive lung sacs with air using a swallowing motion accompanied by the relaxation of the circular sphincter.
Adaptations allow amphibians to move from a purely aquatic atmosphere to live on land. The branchial apparatus disappears. The first species to have a structured larynx are the polypterus (or lungfish). Their larynx has a rigid fibrous structure and a dilator muscle in addition to a constrictor muscle. This laryngeal structure allows for an open laryngeal configuration on land and a closed larynx under water. More evolved amphibians have more elaborate cartilaginous structures and more dilator muscles.
Reptiles are the first land-dwelling vertebrates who only breathe air. Their larynx sits high on the hyoid bone, just behind the tongue base. The intrinsic muscles are formed of a pair of dilator muscles and a pair of constrictor muscles. The external laryngeal muscles appear for the first time in reptiles. Epilaryngeal folds rise high up to the choanae, allowing the crocodile, for example, to breathe through its nostrils while its mouth is open under water drowning its prey.
Birds have a reptilian larynx but, contrary to reptiles, they can lower their larynx and breathe through the mouth. The larynx is by no means responsible for song in birds, as the arytenoid cartilages are immobile and there are no vocal folds. The larynx is still only a sphincter. Song is produced by the syrinx, a unique organ situated at the carina and around the main bronchi. A cartilaginous structure called the tympanum, or drum, forms the base for insertion of tympaniform membranes capable of vibrating. A complex group of muscles, innervated by the hypoglossus nerve, modulates the tension of the membranes and the dimensions of the drum and thus the resonant frequency of the syrinx--in other words, the melody of the song. Some birds even have physiologic diplophonia, a result of the two halves of the syrinx vibrating asynchronously.
In marsupials, the thyroid cartilage detaches from the hyoid bone with a slight descent of the larynx as compared to reptiles, in which these cartilages are welded together. The cricoid and thyroid cartilages and the cricoid and arytenoid cartilages remain solidly fixed together and immobile. Marsupials produce very little sound.
In mammals, the larynx takes on very different positions and configurations. In all species, the hyoid bone, the thyroid cartilage, the cricoid cartilage, and the arytenoids are independent and mobile. In most mammals, the larynx is in a high position just below the nasopharynx. This assures close contact for nasal breathing necessary for keen olfaction. In chimpanzees, the cricoid cartilage is at the level of the fourth cervical vertebra, compared with the fifth or sixth in adult humans. The most extreme example of a high larynx are the cetaceans, whose larynx is in the nasopharynx just below the cephalic blowhole. In other species, the epiglottis, with gustatory papillae, touches the velum, creating a direct connection between the nasopharynx and the larynx with little or no communication with the oropharynx. This allows some species to eat while keeping their olfaction free for detecting predators. This high-situated laryngeal configuration is also found in human newborns, who can breathe through the nose and swallow simultaneously.
For tree-climbing species, laryngeal occlusion is essential for stabilizing the thorax, providing a stable pivot point for the muscles of the upper limbs. And for all mammals, laryngeal occlusion permits the increase in intrathoracic pressure necessary for coughing and sneezing.
In humans, the larynx has four basic functions: air breathing, occlusion for protection of the inferior airways during swallowing, occlusion for the increase in intrathoracic (cough) or intra-abdominal (defecation) pressure and stabilization of the superior limbs, and finally phonation. The lower laryngeal position, the short velum, and the posteriorly situated tongue base are considered as the most important phylogenetic modifications allowing oral speech production in humans,  as opposed to our simian or Neanderthal ancestors. These modifications of the vocal tract accompany a flexion of the cranial base, which is flatter in primates. The modifications seem to have occurred to the detriment of the protective, sphincteric function of the larynx, as compared with other mammals.
In summary, the larynx initially developed as a sphincter protecting the lower airways. A cartilaginous structure permitted a dilated laryngeal configuration in air-breathing species. Anatomic differences among species developed to improve olfaction or to facilitate large movements of the upper limbs. Only in humans has the larynx evolved as primarily a speaking organ, due, it seems, to the lower laryngeal position, the short velum, and the posterior tongue base. It seems that the phonatory function of the larynx has developed to the detriment of its protective and olfactory functions.
The larynx is involved in four basic physiologic functions: swallowing, breathing, cough, and phonation. It is also involved in certain reflexes and in movements necessitating glottic occlusion.
Swallowing. The preceding section has shown that the first role of the larynx is protection of the lower airways. The low laryngeal position in man, as opposed to other species, appears to be a factor in the high incidence of aspiration in humans.
Swallowing can be described in four stages. The first two--oral preparatory and oral--are voluntary. The food bolus formed during the first phase is propelled toward the tongue base in the second phase. [5,6] The third phase--pharyngeal--starts when the bolus arrives at the tongue base. This phase is, in fact, a complex reflex combining cranial nerves IX, X, and XII, the reticular formation of the medulla, the respiratory centers of the brainstem, and probably the cerebral cortex itself. It is evoked by the sensory stimulation of pharyngeal and supraglottic structures.
Apnea marks the beginning of the pharyngeal phase. The velum rises, occluding the nasopharynx. The tongue base moves backward as the larynx is raised anteriorly. The larynx closes from inferior to superior with adduction of the true vocal folds, the false vocal folds, and then the arytenoids and aryepiglottic folds. The anterior laryngeal motion moves the glottis out of the direct line between the oropharynx and the hypopharynx and leads to a relative opening of the hypopharynx and the upper esophagus. Adduction of the aryepiglottic folds and posterior displacement of the epiglottis, which is pushed by the tongue base, creates a furrow on each side, directing the food bolus toward the piriform sinuses.
The last event in the pharyngeal phase is the relaxation of the upper esophageal sphincter. The entire pharyngeal phase lasts 1 second. The esophageal phase follows with peristaltic motion of the esophageal smooth muscle. Aspiration due to recurrent laryngeal nerve paralysis occurs generally during the pharyngeal phase due to insufficient laryngeal occlusion, but abnormalities in upper esophageal sphincter tone and relaxation have been documented in URLNP. 
Breathing. The glottis is divided functionally into two distinct but interdependent parts. The posterior glottis (or respiratory glottis or cartilaginous glottis) is located between the bodies and the vocal processes of the two arytenoids. It represents approximately 45% of the anterior-posterior length of the glottis.  The anterior glottis (or phonatory glottis or membranous glottis) is formed by the musculomembranous part of the vocal folds, which vibrate during phonation. Both the anterior and posterior pans of the glottis are involved in breathing and in phonation, however. During phonation, the posterior glottis regulates laryngeal aerodynamics, which in turn affect the vibratory qualities of the anterior glottis. 
Glottic closure is obtained by a complex movement of the cricoarytenoid joint. The joint is comprised of two elliptical surfaces, with their major axes directed perpendicularly. The cricoid joint surface is convex superiorly, with its major axis directed in an anterolateral direction. The smaller arytenoid surface is concave inferiorly, with its major axis directed in an anteromedial direction. This configuration allows for two types of motion: medial rocking of the body of the arytenoid along the short axis of the cricoid surface and forward sliding of the body of the arytenoid along the long axis of the cricoid surface. The combined effect of these two movements produces a medial and caudal displacement of the vocal process during adduction and a lateral and cranial displacement of the vocal process during abduction. "Rotation" of the arytenoid does not and cannot occur physiologically due to the configuration of the cricoarytenoid joint and its ligaments. [8-10] Other types of complex motion are possible, such as the posterior approximation of the arytenoid bodies. It appears that the posterior glottis can be modified almost completely independently of the anterior glottis due to the three-dimensional rocking and sliding of the arytenoid bodies.
During inspiration, the entire glottis is in abduction. The electromyographic activity of the posterior cricoarytenoid muscle (PCA) is linked to that of the diaphragm. The PCA is activated 40 to 100 ms before the inspiratory activation of the diaphragm. [11,12] PCA activity is also positively correlated with the activity of accessory inspiratory muscles, such as the infrahyoid and intercostal muscles. The cricothyroid muscle (CT) is also activated during inspiration, tensing the vocal folds.
Laryngeal adduction during expiration varies according to lung aerodynamics. Adduction produces laryngeal resistance during expiration and thus increases intra-alveolar pressure, which in turn improves gas exchange. Adduction seems to increase especially during physical effort. At rest, expiratory adduction may even be absent. Adduction during expiration is not obligatory, as tracheotomy bypasses the larynx with apparently no effect on alveolar gas exchange. 
Cough. Coughing is comprised of three phases: inspiration, compression, and expulsion. During the compression phase, thoracic pressure is increased by compression of expiratory muscles at constant volume due to laryngeal closure. The maximum level of pressure attained, and thus the efficiency of the cough, depends in part on the quality of glottic occlusion. The cough reflex is inhibited during deep sleep, and thus laryngeal events such as gastropharyngeal reflux or aspiration may go unnoticed at night. 
Phonation. "Vocal fold vibration is the result of a delicate balance of subglottal air pressure driving the folds apart, and the muscular, elastic, and Bernoulli forces that bring them together." Laryngeal phonatory mechanisms are today in part explained by the "myoelastic-aerodynamic" model developed by Dejonckere, Van den Berg, Faaborg-Andersen, Hirano, Titze, and others. Human phonation is thought to depend on four parameters: the state of the vocal fold, the state of the mucosa, glottic closure, and subglottic pressure. Parameters defining the state of the vocal fold are the position in the axial plane, the level in the frontal plane, length, thickness, the size and shape of the medial edge, mass, and elasticity. The parameters describing the state of the vocal mucosa are tension, its viscoelastic properties, and the viscoelastic properties of the superficial lamina propria (Reinke's space). Glottic closure, or aerodynamic resistance, depends on the adduction of the membranous vocal folds and on the degree of closure of the posterior glottis. Subglottic pressure is thought to be the aerodynamic driving force necessary for phonation.
Current theory is based on the "body-cover" theory first described by Hirano.  In this model, the "body" (the vocalis muscle) and the "cover" (the vocal mucosa) vibrate independently.  In theory, the vibrations of the body do not interfere with those of the cover because of the isolating effect of the superficial lamina propria. Voice quality is thus the combination of these two vibrating structures, passively driven by the aerodynamic force of the subglottic pressure. Sensory or muscle-spindle receptors have not been shown to play an active role in the initiation or the propagation of vocal fold vibration.
Laryngeal reflexes. The laryngeal closure reflex is elicited by stimulation of epicritical sensory receptors, ubiquitous but particularly rich in the arytenoid mucosa and on the laryngeal surface of the epiglottis. The tonic laryngeal closure usually recedes spontaneously, but it is also inhibited by the hypoxemia ensuing if glottic closure lasts a sufficient length of time. The closure reflex is easily elicited at birth, becomes even more active during the first year of life, and then gradually decreases in activity. The physiologic hyperactivity of this reflex seems to be a factor in sudden infant death syndrome. It is also implicated in paroxystic laryngospasm, in which the closure reflex is provoked by gastropharyngeal reflux from environmental irritation.  The closure reflex also intervenes during vomiting.
Sensory laryngeal stimulation, especially during endotracheal intubation, can also lead to bradycardia, arrhythmia, and a fall in blood pressure due to parasympathetic vagal-mediated reflexes. Certain laryngeal receptors are stimulated by negative pressure in the upper aerodigestive tract. Their stimulation in obstructive sleep apnea syndrome or upper airway resistance syndrome may be a cause of cardiovascular complications.
Glottic closure occurs during the sneezing reflex and during hiccough, a complex reflex that remains poorly understood. Laryngeal closure, reflex or voluntary, provides an increase in abdominal pressure for defecation and provides pressure-mediated stabilization of the thorax for lifting and other activities involving the arms and shoulders.
Laryngeal innervation and neurophysiology
Central motor innervation. Vagal motor neuron cell bodies are located in the brainstem in the ambiguous nucleus and in the retrofacial nucleus. The distribution of motor neurons was first demonstrated by Gacek in 1975 using the retrograde tracer horseradish peroxidase.  The ambiguous nucleus is located lateral to the reticular formation of the caudal brainstem, extending the entire length of the medulla oblongata. Its middle portion contains the vagal motor neurons, with the larger adductor neurons dorsally and the smaller abductor neurons (of the PCA muscle) ventrally. The PCA neurons are fewer (by a factor of 4) than the adductor neurons and are more compactly arranged.
This dorsoventral arrangement parallels that of the respiratory centers in the brainstem. The expiratory center is situated dorsally, whereas the inspiratory center is located ventrally in the reticular formation. The proximity of the laryngeal neurons to the respiratory neurons allows direct connection between them. The adductor function is phylogenetically the oldest, primitive larynges having only a constrictor function. The abductor function appears later in more advanced species. Its more recent phylogenetic appearance could be the origin of a higher sensitivity of the abductor neurons to central lesions and explain bilateral central abductor paralysis (or Gerhardt's syndrome).
The retrofacial nucleus is caudal to the facial nerve. It contains the motor neuron cell bodies for the CT arranged peripherally, surrounding cell bodies for the PCA located centrally.
In summary, CT and PCA innervation arise in the ambiguous and retrofacial nuclei, whereas adductor innervation arises only in the nucleus ambiguous. Laryngeal innervation is ipsilateral; no cross-innervation arises directly from these motor neurons. [11,16,17]
Current concepts in laryngeal innervation. Cortical regulation. There seem to be a certain number of direct or indirect connections between the ambiguous nucleus and certain central areas--especially Broca's area, the motor cortex, and the anterior cingulum--but their significance is poorly understood.
Subcortical regulation: periaqueductal gray matter. The periaqueductal gray matter (PAG) describes a particular subcortical array of neurons located adjacent to the aqueduct of the fourth ventricle. It has been studied extensively in the cat and monkey.  The neurons receive afferences from the limbic system and the cingulum. Efferences extend to the ambiguous nucleus, the nucleus of the solitary tract (sensory), and to respiratory centers in the brainstem.
This neuronal formation seems to play an important role in reflex vocalization (pain, danger) in the species studied. Stimulation studies in animals have also shown a central analgesic effect, with endorphin production following stimulation of the PAG. It has been hypothesized that this region could be responsible for endorphin liberation in certain types of human vocalization, such as lamenting and crying, explaining the cathartic effect of human expressions.
Brainstem regulation. Some studies have evoked the existence of bilateral connections between laryngeal motor neurons in the brainstem, but this has not yet been proven. The ambiguous nucleus is thought to receive afferences from the contralateral ambiguous and retrofacial nuclei as well as bilateral signals from the nucleus of the solitary tract. 
Interactive organization of laryngeal control. The cerebral cortex intervenes with a high level of plasticity in the various laryngeal activities. Subcortical structures (striatum, corpus callosum, globus pallidus, thalamus, and cingulum) seem to intervene bilaterally invoicing, for only bilateral subcortical lesions lead to lasting loss of voice. The cortical regulation, however, seems to be situated in the dominant hemisphere, as a unilateral lesion can lead to abnormal laryngeal function.
The different types of laryngeal activity--crying, whispering, singing, coughing, etc.--appear to have a specific and distinct cortical representation. The fibers from each cortical laryngeal "center" converge in the nucleus ambiguous. It seems that each type of laryngeal activity has a "code" of neuronal stimuli, varying by origin, chronology, frequency, and intensity. These different "codes" translated in the nucleus ambiguous would produce a variable effect on the laryngeal musculature, leading to different types of laryngeal activity. Spasmodic dysphonia is an empiric example of this central "coding" of laryngeal activity in that, in general, the spasms occur only for a specific type of laryngeal activity, usually speaking. The spasms are often absent during other types of laryngeal activity, such as whispering, shouting, or singing.
The variable laryngeal muscle activity may also be regulated by inhibiting interneurons (a brainstem equivalent of the Renshaw interneurons in the spinal cord). This remains to be more fully elucidated.
In conclusion, experimental evidence exists to show that voicing control is more complex than a simple voluntary mechanism. 
Peripheral motor innervation
Axons. The motor innervation of the larynx is supplied by the Xth cranial nerve, the vagus nerve. The motor axons leave the ambiguous and retrofacial nuclei, combine, and form one or two rostral roots for the vagus nerve. 
Axons from the ambiguous nucleus have a larger circumference than the retrofacial axons. It has been shown for skeletal muscle that large-diameter fibers provoke rapid and short muscular contractions corresponding to fast-twitch muscular fibers (type II). Small-diameter fibers correspond to slow-twitch muscle fibers (type I). The double origin, ambiguous and retrofacial, of the PCA and CT are reflected in their mixed muscular type. Type land type II fibers coexist in these muscles, but they are classified into distinct groups.  This duality may play a role in the different functions of these muscles, with fatigue-resistant contraction for abduction during breathing and rapid contraction for adduction and tensing of the vocal folds during phonation. 
Axons for the recurrent laryngeal nerve are all myelinated and remain grouped together along the entire length of the vagus nerve. Abductor and adductor fibers are randomly dispersed. In the cranial portion of the vagus nerve, the "recurrent" fibers are in the anterior part of the nerve. They turn toward the medial aspect of the nerve as the vagus descends caudally. The recurrent nerve separates from the vagus at its medial aspect. The abductor and adductor fibers separate several centimeters before entering the larynx (in the cat). Thus, Semon's theory of the greater anatomic vulnerability of abductor fibers as compared with adductor fibers seems to be false (see below).
The vagus nerve. The vagus nerve originates at the anterolateral surface of the medulla oblongata. The superior (jugular) ganglion of the vagus nerve is situated close to the origin of the nerve and contains the cell bodies for sensory and parasympathetic fibers. The vagus nerve then leaves the cranium at the skull base via the jugular foramen anterior to the internal jugular vein. The nerve follows the posteromedial surface of the vein to the plexiform ganglion, from which arises the superior laryngeal nerve. Along the posteromedial surface of the internal jugular vein, the vagus contributes to and receives anastomoses from the carotid sympathetic chain. On the left side, the vagus follows the posterior surface of the common carotid artery and crosses the anterior surface of the aortic arch. The recurrent nerve branches under the aortic arch, runs medially under the aorta, and then returns cranially in the tracheoesophageal groove. On the right, the vagus nerve follows the posterolateral surface of the comm on carotid artery to the bifurcation of the brachiocephalic artery. The recurrent nerve branches off under the subclavian artery, runs medially along the pleura, and then returns cranially behind the common carotid artery in front of the vertebral artery toward the tracheoesophageal groove. 
The recurrent laryngeal nerve. Thus, on the left, the recurrent laryngeal nerve (RLN) has an intrathoracic portion, vulnerable to cardiothoracic or mediastinal pathology. In its cervical portion, the nerve runs cranially behind the left lobe of the thyroid gland, under the inferior pharyngeal constrictor muscle, and penetrates into the larynx behind the cricothyroid joint between the thyroid ala and the cricoid lamina. Some reports refer to the intralaryngeal portion of the nerve as the inferior laryngeal nerve, but we will continue to employ the term RLN.
The RLN has an anterior motor branch and a posterior sensitive branch. The latter runs deep to the piriform sinus mucosa to form an anastomosis with the internal branch of the superior laryngeal nerve (Galen's anastomosis).
The right RLN runs obliquely from the subclavian artery, behind the right thyroid lobe, lateral to the tracheoesophageal groove. The laryngeal penetration, bifurcation, and subdivisions are the same as the left RLN. 
The anatomy of the cervical portion of the RLN may vary. Extralaryngeal division before entering the larynx has been observed in 35 to 80% of cases reported in the literature.  The RLN may be "nonrecurrent," branching from the cervical portion of the vagus nerve. Its relationship with the thyroid arteries is variable. On the left, the RLN is said to run posterior to the branches of the inferior thyroid artery. On the right, the nerve is said to run anterior to the inferior thyroid artery. In truth, their relative positions are highly variable.
The microvascularization of the RLN is principally supplied by a posterior branch of the inferior thyroid artery or, more rarely, by the thyroid ima artery, a branch of the aortic arch, or the brachiocephalic artery at the anterior surface of the trachea.
The RLN contains 500 to 1,000 motor axons, of which 250 are destined for the PCA muscle. In cats, the right RLN contains more axons than the left RLN.  Two centimeters caudal to the laryngeal penetration (in cats), 50% of the myelinated fibers are motor axons. At the point of entry, however, 80% are motor axons. Between these two points, the RLN gives off numerous tracheal and esophageal sensory fibers. The remaining 20% of myelinated fibers are sensitive fibers. The nonmyelinated fibers in the RLN are sympathetic carotid plexus fibers and parasympathetic postganglionic dendrites, whose cell bodies lie in the superior jugular and plexiform ganglions.  These nerve fibers innervate laryngeal blood vessels and submucosal glands.
External branch of the superior laryngeal nerve. Current theory holds that only the external branch of the superior laryngeal nerve (SLN) has motor activity, and only for the CT. Recently, however, the debate has been reopened inasmuch as anastomoses between the RLN and the internal branch of the SLN have been observed, which could explain the variability of the vocal fold position in RLN paralysis (see below). The SLN arises at the inferior extremity of the plexiform ganglion. It runs posterior and medially to join the upper pharyngeal constrictor, which it follows inferiorly to the greater cornu of the hyoid bone. Here the SLN divides into an external and an internal branch. The external branch runs along the lateral surface of the middle and inferior constrictor, provides the motor innervation for the two bellies of the CT, perforates the cricothyroid membrane, and assures the sensory innervation of the anterior subglottic region.  The internal branch of the SLN penetrates immediately through the thyr ohyoid membrane. Its intralaryngeal branching is described below.
Intralaryngeal motor innervation
Using Sibler's staining, Sanders et al elegantly demonstrated the intramuscular innervation of the laryngeal muscles.  The RLN divides into an anterior (motor) branch and a posterior (sensory) branch. The first branch of the motor division is destined for the PCA. The second innervates the interarytenoid muscle (IA) after running posterior to the cricoarytenoid joint and deep to the PCA. The IA is innervated by both RLN.  The third motor branch innervates the lateral cricoarytenoid muscle (LCA), passing through it to terminate in the thyroarytenoid muscle (TA). There are multiple terminal branches of the RLN in the TA, in the ventricular fold (superior TA muscle), and in the aryepiglottic muscle. The terminal branches have the highest density along the medial aspect of the medial TA (or vocalis muscle). The external branch of the SLN innervates the CT, continues its course through the CT, the TA, and then terminates on the inferior aspect of the vocal process of the arytenoid.
Anastomoses between terminal branches of the RLN and the SLN have been observed. Branches of the internal branch of the SLN connect with branches of the RLN in the IA muscle.  The function of the nerve endings from the SLN in the IA is still unknown. They may correspond to proprioceptive muscle receptors, although their communications with fibers from the RLN imply a motor function. [28-30] Motor function has yet to be demonstrated, however. The communicating nerve, originating from the extemal branch of the SLN, runs through the CT and deep to the piriform sinus to form an anastomosis with a terminal branch of the RLN in the TA muscle. The communicating nerve is thought to be the nerve of the fifth branchial arch.  In dogs, a motor function of the communicating nerve has been demonstrated.  In the study by Sanders et al, no left-right anastomoses were observed. 
In summary, anastomoses between the RLN and the SLN exist, although their role is currently unknown and motor activity has yet to be proven. The effect of the bilateral innervation of the IA on laryngeal configuration is also unknown. The organization of the muscular branches of the RLN reflect the anatomy of these muscles, arranged in distinct bellies, and organized by muscle fiber type. The complex terminal branching throughout the muscles reflects their multi-innervated nature (see below) and is particularly dense in the vocalis muscle.
Innervation of muscle fibers
Laryngeal muscle fibers differ microscopically from other skeletal muscle. Like other skeletal muscles, each fiber has a mononeuronal innervation; innervation for that fiber arises from only one axon, and thus from only one neuron.  The polyneuronal innervation found in fetal skeletal and laryngeal muscle disappears at birth. Laryngeal muscle is different in that there are several motor end plates per muscle cell (the fibers are multi-innervated). This particular characteristic is shared with the extraocular muscles and may be a factor allowing for more rapid or more sustained contractions. In addition, motor end plates are dispersed throughout the muscles instead of occupying only one specific site. This may be due to the wide insertions of the laryngeal muscles along the laryngeal cartilages and the lack of tendons as with other skeletal muscle,  and it may also play a role in rapid muscular contraction.
Anatomy of peripheral nerves. Peripheral nerves are composed of nerve fibers (axons and dendrites), supporting cells, and connective tissue. More than 50% of the nerve is, in fact, connective tissue containing collagen, elastic fibers, and extracellular fluids.  The nerve fibers are encompassed by four different envelopes. The most peripheral layer is the mesoneurium, a thin transparent layer of connective tissue that attaches the nerve to surrounding tissues. This loose layer allows the nerve to slide with movements of the surrounding structures and provides an easily dissectable plane. The next outermost layer is the epineurium, a dense white covering of collagen and elastic fibers. This layer gives the nerve elasticity; a certain amount of stretching can occur without damage to the nerve fibers within. This dense layer allows for precise suturing during nerve repair without damage to nerve fibers. Inside the epineurium, the nerve fibers (or neurites) are grouped into bundles. Each bundle, or fascicle, is surrounded by perineurium, a layer of cells with tight junctions resting on a peripheral basal membrane, which in turn is surrounded by several dense layers of collagen in a circular and longitudinal array. Epineurial connective tissue holds the fascicles together. The fascicle is the smallest unit of the nerve visible using the operating microscope, and perineurial sutures are possible. The impermeable nature of the perineurium is an efficient barrier against infection, neoplastic encroachment, and inflammation. The endoneurium, a loose connective tissue, is the fourth layer surrounding the neurites.
Each axon is surrounded by Schwann cells, which produce the myelin sheath. Each cell rests on a basement membrane. Ranvier's nodes are small spaces between Schwann cells. For well-myelinated, generally largediameter fibers, the depolarization of the axonal membrane "umps" from node to node, resulting in a rapid conduction rate. Less well-myelinated fibers, generally of smaller diameter (nociception or autonomous nerves), have a slower conduction rate because the depolarization runs along the entire length of the fiber and cannot jump.
Each nerve fiber changes its position in the nerve every few centimeters, moving in the fascicle and even changing fascicles. Nerve vascularization is provided by a network of small, interconnected blood vessels running longitudinally inside the epineurium and the perineurium. This provides a relative resistance to surgical devascularization. 
Nerve growth and regeneration. Trauma to a nerve results in a degenerative phase followed by regeneration.  Axonal degeneration is retrograde, the axon shrinking back toward the neuronal cell body. The myelin and endoneurium left behind are phagocytized by cells involved in the inflammatory response. The proliferation of neural envelope cells follows the degeneration. The Schwann cells proliferate along their basal membrane and direct the longitudinal regrowth of the axon. Schwann cell proliferation is necessary for linear axonal regrowth in the right direction. It has been postulated that Schwann cells produce neurotrophic factors, but the mechanism of their effect on the neural cytoskeleton is unknown. An exaggerated proliferation of fibroblasts may block the linearly organized proliferation of Schwann cells, resulting in a neuroma--a chaotic mix of fibroblasts, collagen, Schwann cells, and neurites.
Neural regrowth occurs distally toward the proximal end of the distal nerve segment, even if there is a long space separating the two ends. The distal segment seems to produce factors that attract the growing nerve, but the mechanism is currently poorly understood. It is also thought that denervated muscle produces neurotrophic factors. 
In the central nervous system, the neural cell body undergoes chromatolysis with peripheral shifting of the Nissl bodies (stocked metabolites) and a decrease in their number. The cell body grows larger and increases its metabolism, as shown by a major increase in RNA. It seems that synaptic functions decrease during this phase of regeneration and metabolic reorganization.
After injury, changes in the central neural control may occur. Peripherally, close to the damaged tissue, undamaged adjacent nerve fibers can migrate and reinnervate the damaged tissues, resulting in innervation by a different neuron than before the injury. This has been observed for sensory and motor nerves, especially in the case of amputation. 
Types of neural injury
One macroscopic type of nerve injury can in fact produce several different microscopic types of neurite damage.
Neurapraxis. Neurapraxis (nerve "shock") results simply in a conduction block along the neurite. The cell membranes of the nerve fibers and the nerve envelopes remain intact. The macroscopic mechanism is generally compression or excessive stretching. Large-diameter myelinated fibers (motor, sensory) undergo the most damage. Autonomic and nociceptive nerves are less damaged, and pain may occur after neurapraxis. Detection electromyography (EMO) is abnormal, reflecting the paresis and/or the hypoesthesia. Stimulation-detection EMG, however, is normal, for the neuron is intact. Spontaneous recovery occurs after several days to several weeks. 
Axonotmesis. In this case, the nerve fiber and its myelin sheath are cut, without interruption in the neural envelopes. Again, it is generally caused by violent compression or stretching. Nerve conduction is interrupted and the nerve fiber undergoes retrograde degeneration (Wallerian degeneration). Detection EMO shows fibrillation potentials, a sign of denervation. Stimulation-detection EMG obtains no response. The degeneration phase begins 3 to 7 days following the damage and lasts several weeks.
Nerve fiber regrowth then occurs at a rate of approximately 1 mm per day. The neurite must recuperate the distance lost during the degeneration phase, cross the site of the initial lesion, penetrate into the remaining distal envelopes, and continue all the way to the nerve endings. During this phase, stimulation-detection EMG shows potentials with increased latency and small amplitude.
The axon arrives at the motor end plate and sprouts nerve endings, but voluntary muscular contraction may still not be possible due to a "maturation" phase involving nerve, end plate, and muscle.
Neurotmesis. Neurotmesis is the interruption of neurites, myelin, and the nerve envelopes (at least the endoneurium and the perineurium, in crush injury, for example). Wallerian degeneration and then nerve regrowth occur in much the same way as after axonotmesis, with two differences. First, regrowth of the envelopes may occur in a chaotic manner, resulting in a neuroma--a disorganized mix of fibroblasts, collagen, elastic fibers, perineural cells, Schwann cells, and neurites. The second difference is that even the neurites that have crossed the site of the initial lesion tend to have a smaller than normal diameter and a poor-quality myelin sheath, and thus function poorly. 
Sunderland's classification. Much of the currentknowledge of peripheral nerve anatomy and physiology was observed by Sunderland, who developed the following classification for peripheral nerve lesions: grade I, neurapraxia; grade II, axonotmesis; grade III, mixed neurapraxia and axonotmetic lesion; grade IV, neurotmesis with continuity of the epineurium; grade V, complete nerve section. 
In general, types I and II usually begin to show spontaneous EMG activity after approximately 6 weeks. In the other cases, recuperation is unpredictable.
Depending on the microscopic lesions occurring in the RLN, several scenarios are possible. For pure neurapraxic lesions (after thyroidectomy for benign pathology, for example), laryngeal paresis will occur. Paresis may not even be detected clinically, and the nerve generally recuperates spontaneously in several days or weeks. For pure axonotmesis, nerve regeneration will have a longer delay, and the resulting muscular contractions may be of poorer quality than before. In theory, however, each axon regrows along its previous envelopes and reaches the same motor end plates as in its previous state.
Neurotmesis is the interruption of the fibers' envelopes, even with immediate termino-terminal anastomosis and even if just the endoneurium and perineurium are interrupted without epineural effraction. The risk is misdirected regrowth of nerve fibers. If a neuroma forms, the axons never reach the laryngeal muscles. Laryngeal paralysis is theoretically total. Detection EMG shows fibrillation potentials (denervation) and then electrical silence. After approximately 3 weeks of denervation, the muscles undergo denervation atrophy and disorganization of the muscle cell structure, and then fibrosis and/or adipose transformation. Muscle degeneration is thought to be complete after 2 to 3 years.
Axonal regrowth may reach the laryngeal motor end plates, with several possible outcomes. First, voluntary contraction is never attained, but reinnervation is sufficient to prevent denervation atrophy. Second, the reinnervation may be of poor quality due to small fiber dimension or to insufficient maturation of the motor end plates. This may lead to isometric muscular contraction without a global vector of motion. Third, axons may grow in a misdirected fashion and reach motor end plates that they initially did not innervate. Adductor fibers may innervate abductor muscles and vice versa. The simultaneous contraction in antagonist muscles that results is called "synkinesis." 
Synkinesis depends on the quantity and the quality of the reinnervation of the antagonistic muscles and the equilibrium or disequilibrium in reinnervation that has occurred. Equilibrated synkinesis can lead to a neutral position of the arytenoid without anteromedial rocking and a paramedian vocal fold position without vocalis muscle atrophy. Small trembling-like movements may be detected. This laryngeal configuration generally results in a good vocal result and is considered "favorable synkinesis."  Disequilibrated synkinesis can lead to spasmodic or paradoxical motion of the arytenoid and the vocal fold. Spasmodic dysphonia and/or spasmodic stridor may occur. This is considered "unfavorable synkinesis." 
A second phenomenon has been evoked to explain apparent misdirected reinnervation. Ephapse is the mutual stimulation of two regenerating neurites, with myelin sheaths of poor quality, like two electrical wires that have lost their plastic coating.  Normal myelin (and the other neural envelope layers) act as an electrical barrier, insulating each fiber. The poor quality of the myelin may result in cross-stimulation of adjacent nerve fibers.
Central reorganization has been demonstrated after nerve damage in limbs. Misdirected nerve fibers obtain a new function according to the muscle reinnervated. In some cases, the neuron cell body changes its central connections according to this new function in an adaptation mechanism. However, this adaptation may be incomplete, and voluntary solicitation of the "old" function may produce parasitic motion in the newly innervated muscle. Central reorganization has not yet been demonstrated for neurologic lesions of the larynx.
RLN lesions can also be mixed, axonotmetic, and neurotmetic. In this case, axons having retained their envelopes (perineurium) will reinnervate the same muscles as before the damage. The other axons may undergo misdirected regrowth with synkinesis. The variable association of these two mechanisms could explain some of the variability in laryngeal configuration after RLN damage. Partial residual normal innervation from neurapraxia may also coexist with more severe lesions and could also contribute to laryngeal configuration in some cases. 
Van Lith-Biji et al compared the evolution of RLN damage by compression and by section-anastomosis in a feline model.  Signs of laryngeal motion appeared earlier after compression of the RLN (axonotmesis). EMG activity in the PCA and TA muscles appeared earlier and abduction was of greater amplitude. Synkinesis was detected on EMG more frequently following nerve section, despite surgical anastomosis. The number of axons present distal to the lesion site was, however, the same for both types of neural damage.
Evidence for synkinesis. Flint et al, in their seminal article concerning laryngeal reinnervation in the rat, provided an elegant demonstration of misdirected nerve growth.  They employed horseradish peroxidase as a retrograde tracer to follow axons corresponding to laryngeal muscles after unilateral section-anastomosis of the RLN. Normally, the adductor neurons corresponding to the TA and the LCA are located lateral and dorsal to the abductor (PCA) neurons. Following section-anastomosis, this clear organization disappeared; the neurons corresponding to the TA, LCA, and PCA were arranged haphazardly, although with a predominance of neurons in a lateral and dorsal area. In addition, labeling was weaker than in normal rats. No labeling in the contralateral brainstem was observed. Finally, up to 15% of the neurons labeled for both the PCA and the TAILCA, meaning that the same axon had branches for the adductors and for the abductor. EMG in the experimental rats showed phasic inspiratory activity in the PCA, but also in the TA and the LCA, normally electrically silent during breathing. These observations imply a nonselective, misdirected reinnervation of the laryngeal muscles. The reinnervation seemed to be predominantly assured by the adductor fibers, four times more numerous than the abductor fibers in the RLN. [16,34] Other studies involving limbs and the facial nerve have shown that selective reinnervation may occur after section-anastomosis, but only in the case of very young animals. 
In summary, synkinesis--the simultaneous contraction of adductor and abductor muscles--seems to originate from misdirected regrowth of motor axons. A single motor neuron can innervate both an abductor and an adductor muscle following misdirected regrowth. Reinnervation seems to be assured for a major part by adductor fibers (at least in the rat). No contralateral reinnervation has as yet been demonstrated.
Laboratory of voice, biomaterials and cervicofacial oncology, CNRSUPRESA 7018, University of Paris V, Laennec Hospital, 42 rue de Sevres, 75007 Paris, France.
(*.) Reprinted from "Hartl DM, Brasnu D. Les paralysies recurrentielles: connaissances actuelles et traitements. Annales d'Otolaryngologie et de Chirurgie Cervico-Faciale 2000;117:60-84." (c) editions Masson.
(1.) Leiberman P, Crelin ES, Klatt DH. Phonetic ability and related anatomy of the newborn and adult human, Neanderthal man, and the chimpanzee. Am Anthropol 1972;74:287-306.
(2.) Kelemen G. Comparative anatomy and performance of the vocal organ in vertebrates. In: Acoustic Behavior of Animals. Amsterdam: Elsevier, 1963:489-521.
(3.) Negus VE. The Mechanism of the Larynx. London: W.M. Heinemann, 1929.
(4.) Devilliers C, Clairambeau P. Precis de Zoologie. Vertebres I. Anatomie Comparee. Paris: Masson, 1976.
(5.) Cummings CW, ed. Otolaryngology-Head and Neck Surgery. 3rd ed. Vol. 3. St. Louis: Mosby-Year Book, 1998.
(6.) Wilson JA, Pryde A, White A, et al. Swallowing performance in patients with vocal fold motion impairment. Dysphagia 1995;10:149-54.
(7.) Hirano M, Yoshida T, Tanaka S. Vibratory behaviour of human vocal folds viewed from below. In: Gauffin J, Hammarberg B, eds. Vocal Fold Physiology. Acoustic, Perceptual and Physiological Aspects of Voice Mechanisms. San Diego: Singular Publishing, 1991:1-6.
(8.) Kasperbauer JL. A biomechanical study of the human cricoarytenoid joint. Laryngoscope 1998;108:1704-11.
(9.) Von Leden H, Moore P. The mechanics of the cricoarytenoid joint. Arch Otolaryngol 1961;73:541-9.
(10.) Selbie WS, Zhang L, Levine WS, Ludlow CL. Using joint geometry to determine the motion of the cricoarytenoid joint. J Acoust Soc Am 1998;103:1115-27.
(11.) Gacek RR. Localization of laryngeal motor neurons in the kitten. Laryngoscope 1975;85:1841-61.
(12.) Fried MP, ed. The Larynx: A Multidisciplinary Approach. 2nd ed. St. Louis: Mosby-Year Book, 1996.
(13.) Blomgren M, Chen Y, Ng ML, Gilbert HR. Acoustic, aerodynamic, physiologic, and perceptual properties of modal and vocal fry registers. J Acoust Soc Am 1998;103:2649-58.
(14.) Hirano M. Clinical Examination of Voice. New York: Springer, 1981.
(15.) Titze IR. On the mechanics of vocal-fold vibration. J Acoust Soc Am 1976;60:1366-80.
(16.) Gacek RR, Malmgren LT, Lyon MJ. Localization of adductor and abductor motor nerve fibers to the larynx. Ann Otol 1977;86:770-6.
(17.) Larson CR, Yoshida Y, Sessle BJ. Higher level motor and sensory organization. In: Titze IR, ed. Vocal Fold Physiology. San Diego: Singular Publishing, 1993:143-88.
(18.) Mazo M, Erikson D, Harvey T. Emotion and expression: Temporal data on voice quality in Russian lament. In: Fujimura O, Hirano M, eds. Vocal Fold Physiology: Voice Quality Control, San Diego: Singular Publishing, 1995:173-87.
(19.) Gacek RR, Lyon MJ. Fiber components of the recurrent laryngeal nerve in the cat. Ann Otol 1976;85:460-71.
(20.) Lacau St. Guily J. Muscles intriseques du larynx de l'homme: Contribution a l'etude morphologique des fibres musculaires. Paris: These Medecine, Universite Paris VII, No. 153, 1982.
(21.) Ferner H, Staubesant J, eds. Sobotta Atlas of Human Anatomy. Vol. 1. Baltimore: Urban and Schwarzenberg, 1983.
(22.) Bouchet A, Cuilleret J. Anatomie topographique, descriptive et fonctionnelle. 2nd ed. Paris: SIMEP, 1991.
(23.) Moreau S, Goullet de Rugy M, Babin E, et al. The recurrent laryngeal nerve: Related vascular anatomy. Laryngoscope 1998;108:1351-3.
(24.) Woodson GE. Laryngeal neurophysiology and its clinical uses. Head Neck 1996;18:78-86.
(25.) Sanders I, Wu BL, Mu L, et al. The innervation of the human larynx. Arch Otolaryngol Head Neck Sung 1993;119:934-9.
(26.) Blitzer A, Jahn AF, Keidar A. Semon's law revisited: An electromyographic analysis of laryngeal synkinesis. Ann Otol Rhinol Laryngol 1996;105:764-9.
(27.) Mu L, Sanders I, Wu BL, Biller HF. The intramuscular innervation of the human interarytenoid muscle. Laryngoscope 1994;104:33-9.
(28.) Vogel PH. The innervation of the larynx of man and the dog. Am J Anat 1952;90:427-40.
(29.) Wustrow TP, Martin F, Hollander H. Additional motor innervation of the laryngeal sphincter via the internal branch of the superior laryngeal nerve [abstract]. Laryngol Rhinol Otol 1988;67:275-81.
(30.) Sanders I, Mu L. Anatomy of the human internal superior laryngeal nerve. Anat Rec 1998;252:646-56.
(31.) Nasri S, Beizai P, Ye M, et al. Cross-innervation of the thyroarytenoid muscle by a branch from the external division of the superior laryngeal nerve. Ann Otol Rhinol Laryngol 1997;106:594-8.
(32.) Perie S, Lacau St. Guily J, Callard P. Sebille A. Innervation of adult human laryngeal muscle fibers. J Neurol Sci 1997;149:81-6.
(33.) Kline DG, Hudson AR, New E. Nerve injuries: Operative results for major nerve injuries, entrapments and tumors. 1st ed. Philadelphia: W.B. Saunders, 1995.
(34.) Woodson GE. Configuration of the glottis in laryngeal paralysis II: Animal experiments. Laryngoscope 1993;103:1235-41.
(35.) Crumley RL. Laryngeal synkinesis: Its significance to the laryngologist. Ann Otol Rhinol Laryngol 1989;98:87-92.
(36.) Crumley RL. Laryngeal synkinesis revisited. Ann Otol Rhinol Laryngol 2000;109:365-71.
(37.) Crumley RL. Recurrent laryngeal nerve paralysis. J Voice 1994;8:79-83.
(38.) Woodson GE. Configuration of the glottis in laryngeal paralysis I: Clinical study. Laryngoscope 1993;103:1227-33.
(39.) Van Lith-Bijl JT, Mahieu HF, Stolk RJ, et al. Laryngeal abductor function after recurrent laryngeal nerve injury in cats. Arch Otolaryngol Head Neck Surg 1996;122:393-6.
(40.) Flint PW, Downs DH, Coltrera MD. Laryngeal synkinesis following reinnervation in the rat. Neuroanatomic and physiologic study using retrograde fluorescent tracers and electromyography. Ann Otol Rhinol Laryngol 1991;100:797-806.
(41.) Benninger MS, Gillen JB, Altman JS. Changing etiology of vocal fold immobility. Laryngoscope 1998;108:1346-50.
(42.) Perie S, Laccourreye O, Bou-Malhab F, Brasnu D. Aspiration in unilateral recurrent laryngeal nerve paralysis after surgery. Am J Otolaryngol 1998;19:18-23.
(43.) Kashima HK. Documentation of upper airway obstruction in unilateral vocal cord paralysis: Flow-volume loop studies in 43 subjects. Laryngoscope 1984;94:923-37.
(44.) Woodson GE, Hengesteg A, Rosen CA, et al. Changes in length and spatial orientation of the vocal fold with arytenoid adduction in cadaver larynges. Ann Otol Rhinol Laryngol 1997;106:552-5.
(45.) Hoffman HT, McCulloch TM. Anatomic considerations in the surgical treatment of unilateral laryngeal paralysis. Head Neck 1996;18:174-87.
(46.) Dedo HH. Injection and removal of Teflon for unilateral vocal cord paralysis. Ann Otol Rhinol Laryngol 1992;101:81-6.
(47.) Tanaka S, Hirano M, Umeno H. Laryngeal behavior in unilateral superior laryngeal nerve paralysis. Ann Otol Rhinol Laryngol 1994;103:93-7.
(48.) Sercarz JA, Nguyen L, Nasri S, et al. Physiologic motion after laryngeal nerve reinnervation: A new method. Otolaryngol Head Neck Surg 1997;116:466-74.
(49.) Jiang J, Lin E, Hanson DG. Glottographic phase difference in recurrent nerve paralysis. Ann Otol Rhinol Laryngol 2000;109:287-93.
(50.) Koufman JA, Blalock PD. Functional voice disorders. Otolaryngol Clin North Am 1991;24:1059-73.
(51.) Benninger MS, Crumley RL, Ford CN, et al. Evaluation and treatment of the unilateral paralyzed vocal fold. Otolaryngol Head Neck Surg 1994;111:497-508.
(52.) Zeitels SM, Hochman I, Hillman RE. Adduction arytenopexy: A new procedure for paralytic dysphonia with implications for implant medialization. Ann Otol Rhinol Laryngol Suppl 1998;107:1-24.
(53.) Koufman JA, Walker FO, Joharji GM. The cricothyroid muscle does not influence vocal fold position in laryngeal paralysis. Laryngoscope 1995;105:368-72.
(54.) Wani MK, Woodson GE. Paroxysmal laryngospasm after laryngeal nerve injury. Laryngoscope 1999;109:694-7.
(55.) Bridger GP. Unilateral laryngeal palsy, A histopathological study. J Laryngol Otol 1977;91:303-7.
(56.) Kirchner JA. Atrophy of laryngeal muscles in vagal paralysis. Laryngoscope 1966;26:1753-64.
(57.) Quiney RE, Micheals L. Histopathology of vocal cord palsy from recurrent laryngeal nerve damage. J Otolaryngol 1990;19:237-41.
(58.) Gacek M, Gacek RR. Cricoarytenoid joint mobility after chronic vocal cord paralysis. Laryngoscope 1996;106:1528-30.
(59.) Nomoto M, Yoshihara T, Kanda T, Kaneko T. Synapse formation by autonomic nerves in the previously denervated neuromuscular junctions of the feline intrinsic laryngeal muscles. Brain Research 1991;539:276-86.
(60.) Nomoto M, Yoshihara T, Kanda T, et al. Misdirected reinnervation in the feline intrinsic laryngeal muscles after long-term denervation. Acta Otolaryngol Suppl 1993;506:71-4.
(61.) Imaizumi S. Acoustic measures of roughness in pathological voice. J Phonetics 1986;14:457-62.
(62.) Wendler J, Rauhut A, Kruger H. Classification of voice qualities. J Phonetics 1986;14:483-8.
(63.) Paniello RC, Daham JD. Reversibility of medialization laryngoplasty. An experimental study. Ann Otol Rhinol Laryngol 1997;106:902-8.
(64.) Ward PH, Berci G. Observations on so-called idiopathic vocal cord paralysis. Ann Otol Rhinol Laryngol 1982;91:558-63.
(65.) Laccourreye O, Hans S, Menard M, et al. Resultats de l'injection intracordale de graisse autologue dans les paralysies laryngees postchirurgicales. Chirurgie 1999;124:283-7.
(66.) Pearson BW. Subtotal laryngectomy. Laryngoscope 1981;81:1904-12.
(67.) Koufman JA, Isaacson G. Laryngoplastic phonosurgery. Otolaryngol Clin North Am 1991;24:1151-77.
(68.) Tucker HM, Wanamaker J, Trott M, Hicks D. Complications of laryngeal framework surgery (phonosurgery). Laryngoscope 1993;103:525-8.
(69.) Netterville JL, Stone RE, Suken ES, et al. Silastic medialization and arytenoid adduction: The Vanderbilt experience. Ann Otol Rhinol Laryngol 1993;102:413-24.
(70.) Montgomery WW, Montgomery SK. Montgomery thyroplasty implant system. Ann Otol Rhinol Laryngol Suppl 1997;106:1-16.
(71.) Montgomery WW, Blaugrund SM, Varvares MA. Thyroplasty: A new approach. Ann Otol Rhinol Laryngol 1993;102:571-9.
(72.) McCulloch TM, Hoffman HT. Medialization laryngoplasty with expanded polytetrafluoroethylene. Surgical technique and preliminary results. Ann Otol Rhinol Laryngol 1998;107:427-32.
(73.) Cummings CW, Purcell LL, Flint PW. Hydroxyapatite laryngeal implants for medialization. Preliminary report. Ann Otol Rhinol Laryngol 1993;102:843-5l.
(74.) Maves MD, McCabe BF, Gray S. Phonosurgery: Indications and pitfalls. Ann Otol Rhinol Laryngol 1989;98:577-80.
(75.) Simpson CB, Seshul M, Lennington W, et al. Histologic findings of silastic medialization in the canine model. Laryngoscope 1999;109:1424-7.
(76.) Stein J, Eliachar I, Myles J, et al. Histopathologic study of alternative substances for vocal fold medialization. Ann Otol Rhinol Laryngol 2000;109:221-6.
(77.) Chan RW, Titze IR. Viscosities of implantable biomaterials in vocal fold augmentation surgery. Laryngoscope 1998;108:725-31.
(78.) McLean-Muse A, Montgomery WW, Hillman RE, et al. Montgomery thyroplasty implant for vocal fold immobility: Phonatory outcomes. Ann Otol Rhinol Laryngol 2000;109:393-400.
(79.) Gorham MM, Avidano MA, Crary MA, et al. Laryngeal recovery following type I thyroplasty. Arch Otolaryngol Head Neck Surg 1998;124:739-42.
(80.) Adams SG, Irish JC, Durkin LC, et al. Evaluation of vocal function in unilateral vocal fold paralysis following thyroplastic surgery. J Otolaryngol 1996;25:165-70.
(81.) Maragos NE. The posterior thyroplasty window: Anatomical considerations. Laryngoscope 1999;109:1228-31.
(82.) Bielamowicz S, Berke GS, Gerratt BR. A comparison of type I thyroplasty and arytenoid adduction, J Voice 1995;9:466-72.
(83.) Rontal E, Rontal M. Vocal cord injection techniques. Otolaryngol Clin North Am 1991;24:1141-9.
(84.) Crumley RL. Teflon versus thyroplasty versus nerve transfer: A comparison. Ann Otol Rhinol Laryngol 1990;99:759-63.
(85.) Saccogna PW, Werning JW, Setrakian S, Strauss M. Lipoinjection in the paralyzed feline vocal fold: Study of graft survival. Otolaryngol Head Neck Surg 1997;117:465-70.
(86.) Mikaelian DO, Lowry LD, Sataloff RT. Lipoinjection for unilateral vocal cord paralysis. Laryngoscope 1991;101:465-8.
(87.) Laccourreye O, Crevier-Buchman L, Bou Malhab F, et al. Injection intracordale de graisse autologue et paralysie recurrentielle unilaterale. Ann Otolaryngol Chir Cervicofac 1998;115:264-70.
(88.) Laccourreye O, Bou Malhab F, Paz Escovar MI, Brasnu D. Medialisation de la corde vocale paralysee par injection intracordale de graisse autologue par voie endoscopique. Les cahiers d'ORL 1998;23:76-8.
(89.) Bauer CA, Valentino J, Hoffman HT. Long-term result of vocal cord augmentation with autogenous fat. Ann Otol Rhinol Laryngol 1995;104:871-4.
(90.) Brandenburg JH, Kirkham W, Koschkee D. Vocal cord augmentation with autogenous fat. Laryngoscope 1992;102:495-500.
(91.) Shindo ML, Zaretsky LS, Rice DH. Autologous fat injection for unilateral vocal fold paralysis. Ann Otol Rhinol Laryngol 1996;105:602-6.
(92.) Laccourreye O, Crevier-Buchman L, Le Pimpec-Barthes F, et al. Recovery of unilateral recurrent laryngeal nerve paralysis after intracordal autologous fat injection. J Laryngol Otol 1998;112:1082-4.
(93.) Gardner GM, Shaari CM, Parnes SM. Long-term morbidity and mortality in patients undergoing surgery for unilateral vocal cord paralysis. Laryngoscope 1992;102:501-8.
(94.) Hsiung MW, Woo P, Minasian A, Mojica JS. Fat augmentation for glottic insufficiency. Laryngoscope 2000;110:1026-33.
(95.) Tucker HM. Human laryngeal reinnervation: Long-term experience with the nerve-muscle pedicle technique. Laryngoscope 1978;88:598-604.
(96.) Sanders I. Electrical stimulation of laryngeal muscle. Otolaryngol Clin North Am 1991;24:1253-73.
(97.) Goldfarb D, Keane WM, Lowry LD. Laryngeal pacing as a treatment for vocal fold paralysis. J Voice 1994;8:179-85.
(98.) Tucker HM. Combined laryngeal framework medialization and reinnervation for unilateral vocal fold paralysis. Ann Otol Rhinol Laryngol 1990;99:778-81.
(99.) Goding GS. Nerve-muscle pedicle reinnervation of the paralyzed vocal cord. Otolaryngol Clin North Am 1991;24:1239-52.
(100.) Olson DE, Goding GS, Michael DD. Acoustic and perceptual evaluation of laryngeal reinnervation by ansa cervicalis transfer. Laryngoscope 1998;108:1767-72.