Functional restoration for recurrent laryngeal paralysis: an experimental study

Morphologic correlates for laryngeal reinnervation

OBJECTIVE

To describe morphologic correlates for laryngeal reinnervation.

STUDY DESIGN

Review of anatomic experiments dealing with laryngeal innervation performed over a 25-year period.

METHODS

Description of results from experimental studies on the cat and human laryngeal muscles and nerve supply.

RESULTS

Despite separation of abductor and adductor laryngeal motor neurons in the central nervous system, the mixture of abductor and adductor axons in the recurrent laryngeal nerve indicates that selective re-innervation of an individual laryngeal muscle must be accomplished at the neuromuscular junction (NMJ) of the muscle. The optimal time for a reinnervating neural source to re-occupy vacated NMJ is at the time of denervation. If the reinnervation procedure is attempted long (>1 mo) after denervation, extraneous end plates of other neural systems must be eliminated to provide vacant NMJ. The nerve muscle pedicle (NMP) concept is an effective model for reinnervation of a laryngeal muscle provided its activity pattern is similar to that of the denervated muscle and its insertion into vacated NMJ is timely.

CONCLUSION

NMP offers a logical method for selective laryngeal muscle reinnervation. Critical to the success of NMP are the physiological input to the NMP and timing of NMP implantation.

Electrical pacing for dynamic treatment of unilateral vocal cord paralysis. Experiment in long-denervated muscle

In order to explore the possibility of clinical application of laryngeal pacing as a treatment for unilateral vocal cord paralysis, we examined the reactivity of atrophic muscle to electrical stimulation in dogs whose recurrent laryngeal nerves were damaged by crushing, dissection followed by resuturing, or a 3-cm neurectomy. The threshold level to induce enough vocal cord adduction reached the maximum at 2 weeks after nerve injury, decreased with time, and never surpassed 7 V in each case. On the basis of results of these preliminary probings, laryngeal pacing was conducted on a dog 15 months after resection of the laryngeal nerve. Adduction of the paralyzed vocal cord for synchrony with the intact cord was achieved by 7 V of electrical stimulation of the thyroarytenoid muscle that was triggered by signals from the cricothyroid muscle.

Selective reinnervation of the abductor and adductor muscles of the canine larynx after recurrent nerve paralysis

Functional rehabilitation of the larynx after unilateral vocal cord paralysis was attempted in the dog by selective reinnervation of the laryngeal muscles. The intralaryngeal branches of the right recurrent nerve were dissected. The adductor branch was anastomosed with the ansa cervicalis; the abductor branch was anastomosed with the trunk of the phrenic nerve either within the larynx or through the recurrent nerve, the adductor branch of which was sectioned. Results could be analyzed in seven dogs: mobility of the vocal cord was checked, and electromyography, stimulation of the nerves, and histologic studies were performed. Functional reinnervation of both the adductor and abductor muscles was obtained in only one case, with good abduction. Adduction was recorded in five cases. False-positive results emphasize the necessity of collecting several types of data before concluding that functional reinnervation has been accomplished. The reliability of the procedure can and must be improved.

Functional reconstruction for unilateral recurrent laryngeal nerve paralysis caused by thyroid cancer

Surgery for unilateral recurrent laryngeal nerve paralysis, which is caused by thyroid cancer, appearing within the past six months, should be aimed at improvement of the asymmetrical vocal cord and preservation of the stiffness and the mass of the vocal fold. To accomplish these aims, reconstructive surgery of the neuromuscular systems is available and it should be performed at the same time as thyroidectomy. Our surgical procedures for unilateral recurrent laryngeal nerve paralysis caused by thyroid cancer are reported along with some experimental investigations in dogs. Experimentally, muscle atrophy was prevented by neurorrhaphy or by free nerve grafting of the recurrent laryngeal nerve, although movement of the vocals cord was not recovered due to misdirected reinnervation and reduction of the end plates. In our study of functional reconstruction for unilateral recurrent laryngeal nerve paralysis, end-to-end anastomosis or free nerve grafting between the recurrent laryngeal nerve after extirpating the abductor muscle branch of this nerve was the most effective and practical method for improvement of the adductor function. Clinically reconstructive surgery was performed on five females suffering from thyroid cancer with unilateral recurrent laryngeal nerve paralysis. The abductor branch was cut selectively in four cases, followed by free nerve grafting of the ansa cervicalis and pedicle nerve muscle graft of the thyrohyoid muscle implanted in one case. Vocal cord atrophies were not observed and phonations were good in all cases six months after of operation. However, obvious adduction of the operated vocal fold at phonation was obtained only in two cases and other two cases showed slight adduction.

Delayed reinnervation of unilateral vocal cord paralysis in dogs

The neuromuscular implantation technique for rehabilitation of unilateral vocal cord paralysis was performed in four dogs at the time of denervation and in six dogs after varying intervals of chronic denervation. As would be expected, the chronically denervated animals did not achieve the vocal cord excursion of the acute denervations, but some return of motion was noted even after six months. Histologic examination of the posterior cricoarytenoid muscle was correlated with the return of movement.

Horseradish peroxidase studies in animals with neuromuscular transpositions

Horseradish peroxidase (HRP) is used to trace axonal connections from the motor end-plate to the driving neuron. This technique has confirmed that the neurons activating the sternothyroid muscle are located in the cervical spinal cord, while those controlling the posterior cricoarytenoid (PCA) are found in the nucleus ambiguus ipsilaterally. Eight rabbits underwent a sternothyroid ansa pedicle implantation to the PCA at the time of sectioning the recurrent laryngeal nerve ipsilaterally. After two months, four of these animals received HRP injections into the previously implanted PCA. Brainstem staining HRP did not reveal any retrograde transport to the motor neurons that were known to control the sternothyroid. Possible for the failure of retrograde transport are discussed.

Accessory muscle activity and respiration

The relationship between the accessory muscle activity (sternohyoid and sternothyroid) and respiration was studied in canines. These animals do not have an omohyoid muscle such as found in primates. Therefore, chair-trained monkeys that have all three accessory muscles were used in a portion of the study. Findings in canines supported those previously reported. The sternothyroid muscle fired spontaneously with the onset of inspiration, but there was no similar activity in the sternohyoid. The activity of both muscles was observed after sectioning the recurrent laryngeal nerves and inducing hypoxia and hypercarbia. The severely hypoxic animals produced some negligible activity in the sternohyoid muscle. The data obtained from the chair-trained monkeys showed no consistent accessory muscle activity during normal respiration with recordings taken immediately after electrode placement, at 24 hours, and one week later. Respiratory activity was consistent in the omohyoid and sternothyroid but not in the sternohyoid muscle when partial airway obstruction and hypoxia were induced.