Bei-Lian Wu

Quantifying the spread of botulinum toxin through muscle fascia

Botulinum toxin was recently approved for treating several head and neck dystonias. Paralysis of neighboring muscles is the major complication of its use. Spread of toxin from the injected muscle has been suggested as an etiology. This study examines how botulinum toxin crosses muscle fascia by a novel method of quantifying muscular paralysis.

The innervation of the human posterior cricoarytenoid muscle: Evidence for at least two neuromuscular compartments

Recent work has demonstrated that the dog posterior cricoarytenoid (PCA) muscle is composed of three neuromuscular compartments: a vertical, an oblique, and a horizontal. In this study, the human PCA muscle was examined for evidence of neural compartments. Fifteen human PCA muscles were processed by Sihlers stain, which renders the muscle translucent while counterstaining the nerve supply. The results clearly show that in all specimens the nerve supply of the human PCA muscle is separated into at least two main branches: one supplies the horizontal compartment and a second further subdivides to innervate both the vertical and oblique compartments. In 10 of the specimens, these nerve branches arose as separate branches from the recurrent laryngeal nerve. In all specimens, the nerve branch to the horizontal compartment was either combined or connected with the nerve branch to the interarytenoid muscle. The results suggest that the different compartments of the PCA muscle have distinct functions. In addition, the strong connections with the interarytenoid nerve complicate reinnervation procedures to reanimate a paralyzed or transplanted larynx.

The intramuscular innervation of the human interarytenoid muscle

The nerve supply of the human interarytenoid (IA) muscle has been controversial for more than a century. In this study the contribution of the recurrent and superior laryngeal nerves to the IA was investigated in 10 adult human larynges. The larynges were obtained from autopsies and processed with the modified Sihlers technique which clears soft tissue while staining nerve. The IA muscles were dissected off the specimens and transilluminated to demonstrate their nerve supply. The results demonstrated that all 10 IA muscles were bilaterally innervated by both recurrent laryngeal nerves (RLNs) as well as branches of both superior laryngeal nerves (SLNs). These nerves combined within the IA muscles to form a dense anastomotic plexus which was highly variable between specimens. The exact nature of the internal SLN neurons, whether motor or sensory, their innervation targets, or their function, were not discernible. Additional anatomic findings were the presence of large neural communications directly between the SLN and RLN, and smaller neural connections from side to side. All of these results disagree with currently accepted descriptions of laryngeal neuroanatomy.

Botulinum Toxin Decreases Salivation from Canine Submandibular Glands

The objective of this study was to determine whether botulinum toxin types A and D reduced the production of saliva from the submandibular glands of 18 dogs. The left sub-mandibular glands of 8 dogs were injected with increasing doses of botulinum type A toxin (range 10 to 70 units), and the left glands of 10 dogs were injected with botulinum type D toxin (50 or 100 units). The right gland of each dog was injected with equivalent volumes of saline solution to serve as control. Six days after the injection, the lingual nerve was electrically stimulated for 10 minutes (3 mAmp, 20 Hz). The resulting volume of saliva was collected and weighed. Overall, the glands injected with types A or D toxin produced significantly less saliva than comparable glands injected with saline solution. Six of 8 dogs injected with type A toxin showed a significant decrease in saliva production (range 10.1% to 19.2%, one-sided p value = 0.0375) when compared with the controls. Nine of 10 dogs injected with type D toxin demonstrated a highly significant reduction in saliva production (total average decrease = 60%, two-sided p value = 0.001) when compared with the controls. We concluded that intraglandular injections of botulinum toxin types A and D significantly reduced the production of saliva from canine submandibular glands. The potential applications of intraglandular injections of botulinum toxin are discussed.

Quantitative mapping of the effect of botulinum toxin injections in the thyroarytenoid muscle

Spasmodic dysphonia has been successfully treated by thyroarytenoid muscle injections of botulinum toxin (Botox) with dosages ranging from 0.625 to 25 U. In some patients, excessive paralysis with resulting breathiness and aspiration have been noted. In order to maximize the efficiency of Botox injections, the histologic effects of various Botox dosages were examined in the dog. Nine canine thyroarytenoid muscles were injected with 0.5 to 12.5 U of Botox. After 24 hours, the recurrent laryngeal nerve to the injected muscle was electrically stimulated in order to deplete the glycogen within the muscle fibers. Frozen sections of this muscle were then stained for glycogen. Those fibers that retained their glycogen were presumed paralyzed by the Botox injection. The extent of paralysis was found to be dose-related from 1.0 to 7.5 U. At 10 U and above the muscle was completely paralyzed. Spread of the toxin to the lateral cricoarytenoid muscle was seen at doses as low as 1.0 U. Clearly, doses less than 10 U appear sufficient for clinical paralysis.

The Intramuscular Nerve Supply of the Human Lateral Cricoarytenoid Muscle

The lateral cricoarytenoid (LCA) muscle is one of the adductors of the vocal cords; however, some investigators believe that the lateral edge of the muscle may be involved in abduction. The possibility of functionally distinct compartments within the LCA was investigated by observing the pattern of the intramuscular nerve supply. This technique has previously clearly demonstrated neural compartments in the posterior cricoarytenoid, thyroarytenoid and cricothyroid muscles. Five adult human larynges were processed by the Sihlers stain which clears all soft tissue while counterstaining the nerves. The results of our study showed that the innervation pattern of the human LCA muscle is composed of a homogenous nerve plexus localized to the middle region of the muscle. This pattern correlates with the location of motor endplates described by prior investigators. The consistent neural pattern suggests that the LCA is composed of a single neuromuscular compartment.

Overcoming laryngospasm by electrical stimulation of the posterior cricoarytenoid muscle.

The intent of this study was to demonstrate that the technique of transmucosal electrical stimulation of laryngeal muscles may be of clinical use in airway management. Specifically, its ability to overcome laryngospasm was evaluated. Laryngospasm was induced in eight tracheotomized dogs by hyperventilating each dog, and then applying 0.1 M ammonia to the laryngeal mucosa while administering continuous positive airway pressure (CPAP). Laryngospasm was defined by steady apposition of the vocal cords, massive electromyographic activity in the laryngeal adductor muscles, absence of such activity in the posterior cricoarytenoid muscle (PCA), and intraglottic pressure greater than 80 mm Hg. Upon transmucosal application of 10 mAmp current to the PCA bilaterally, the vocal cords abducted for the duration of the stimulus. We theorize that overcoming laryngospasm by electrostimulation involves a reflexive inhibition of the laryngeal adductors. This study provides an objective model for laryngospasm, and demonstrates that electrical manipulation of the vocal cords may have clinical relevance.

Transmucosal Electrical Stimulation of Laryngeal Muscles

A new technique is described that enables discrete activation of individual laryngeal muscles by electrical stimulation across overlying mucosa. In 15 dogs, we defined six distinct motor points by transmucosal stimulation at 3 mA while observing the resulting characteristic position of the arytenoid and true vocal cord. Five dogs were then paralyzed with succinylcholine in order to simulate bilateral vocal cord paralysis. Application of a 3-mA stimulus at each motor point yielded no motion of the cords, but when the current was increased to 20 mA, characteristic responses were elicited. In five other dogs, botulinum toxin was injected directly into laryngeal muscles. Stimulation was used in an attempt to quantify the degree of neuromuscular blockade. In the last group of five dogs, we simulated cricoarytenoid arthritis by scarifying the joint. The extent and nature of the joints impairment could be demonstrated by stimulation. Transmucosal stimulation appears promising as a clinical technique for correlating particular vocal cord movements and thresholds of activation with specific laryngeal disorders. Additionally, such a technique may be useful in clarifying how each laryngeal muscle acts upon the cricoarytenoid joint.

Retrograde tracing of motor axons within the internal superior laryngeal nerve

The human and canine posterior cricoarytenoid muscles appear to have at least two distinct compartments, the vertical and the horizontal. These compartments have different angles of insertion onto the muscular process of the arytenoid cartilage, different proportions of fastand slow-twitch muscle fibers, and different responses to electrical stimulation. These results suggest that these compartments have distinct functions. This study was designed to test this hypothesis by tracing these compartments back to the nucleus ambiguus to see whether their motor neurons pools are separate. Nine dogs had their posterior cricoarytenoid compartments injected with horseradish peroxidase as follows: vertical/horizontal (three dogs); vertical/both (two dogs); horizontal/both (two dogs); vertical/vertical (one dog); horizontal/horizontal (one dog). Dogs were sacrificed 4 days after injection, and the brain stems were removed, serially sectioned, stained with tetramethylbenzidine, and counterstained with neutral red. Labeled cell bodies were compared from side to side by measuring their position from the midline and from a horizontal line drawn through the floor of the fourth ventricle, which was perpendicular to the midline. Results showed that the motor neurons supplying the vertical compartment were usually located in a separate, although adjacent, location just dorsal to those supplying the horizontal compartment. In addition, the vertical motor neurons were oval, whereas the shape of the horizontal motor neurons was multipolar. These results support the possibility of different functions for these two compartments. (Supported in part by grant 01764 from the National Institute on Deafness and Other Communication Disorders.)

Poster 56: Retrograde Tracing of Canine Posterior Cricoarytenoid Muscle Compartments

The human and canine posterior cricoarytenoid muscles appear to have at least two distinct compartments, the vertical and the horizontal. These compartments have different angles of insertion onto the muscular process of the arytenoid cartilage, different proportions of fastand slow-twitch muscle fibers, and different responses to electrical stimulation. These results suggest that these compartments have distinct functions. This study was designed to test this hypothesis by tracing these compartments back to the nucleus ambiguus to see whether their motor neurons pools are separate. Nine dogs had their posterior cricoarytenoid compartments injected with horseradish peroxidase as follows: vertical/horizontal (three dogs); vertical/both (two dogs); horizontal/both (two dogs); vertical/vertical (one dog); horizontal/horizontal (one dog). Dogs were sacrificed 4 days after injection, and the brain stems were removed, serially sectioned, stained with tetramethylbenzidine, and counterstained with neutral red. Labeled cell bodies were compared from side to side by measuring their position from the midline and from a horizontal line drawn through the floor of the fourth ventricle, which was perpendicular to the midline. Results showed that the motor neurons supplying the vertical compartment were usually located in a separate, although adjacent, location just dorsal to those supplying the horizontal compartment. In addition, the vertical motor neurons were oval, whereas the shape of the horizontal motor neurons was multipolar. These results support the possibility of different functions for these two compartments. (Supported in part by grant 01764 from the National Institute on Deafness and Other Communication Disorders.)