Bilateral Selective Laryngeal Reinnervation for Bilateral Vocal Fold Paralysis in Children

Key Points

Question  Can bilateral selective laryngeal reinnervation surgery be used to treat bilateral vocal fold paralysis in pediatric patients?

Findings  In this case series, 8 children aged 2 to 18 years with bilateral vocal fold paralysis underwent bilateral selective laryngeal reinnervation and 6 had improved voice quality based on postoperative GRBAS (grade, roughness, breathiness, asthenia, strain) scale scores. All 6 patients with preoperative tracheostomies and both with perioperative ones were decannulated.

Meaning  The findings of this study suggest that bilateral selective laryngeal reinnervation surgery may be an effective treatment option for bilateral vocal fold paralysis in children with tracheostomy dependence and dysphonia.

Importance  Bilateral vocal fold paralysis (BVFP) in pediatric patients is a challenging entity with multiple causes. Traditional approaches to managing BVFP include tracheostomy, arytenoidectomy, suture lateralization, cordotomy, and posterior cricoid enlargement. These interventions are used to create a stable airway but risk compromising voice quality.

Objectives  To assess the use of bilateral selective laryngeal reinnervation (SLR) surgery to manage BVFP and restore dynamic function to the larynx in pediatric patients.

Design, Setting, and Participants  In this case series performed at 2 tertiary care academic institutions, 8 pediatric patients underwent bilateral SLR to treat BVFP (5 patients with iatrogenic BVFP and 3 with congenital BVFP) from November 2004 to August 2018 with follow-up for at least 1.5 years.

Interventions  Bilateral selective laryngeal reinnervation surgery.

Main Outcomes and Measures  Flexible laryngoscopy findings, subjective and objective measures of voice quality, subjective swallowing function, and decannulation in patients who were previously dependent on a tracheostomy tube.

Results  Participants included 6 boys and 2 girls with a median age of 9.3 (range, 2.2 to 18.0) years at the time of surgery. All 8 patients were decannulated; 6 patients had preoperative tracheostomies and 2 had perioperative tracheostomies. Voice quality, as measured using the GRBAS (grade, roughness, breathiness, asthenia, strain) scale, improved in 6 of 8 patients after reinnervation, and swallowing was not impaired in any patients. In 2 patients, GRBAS scale scores remained the same before and after surgery. Inspiratory vocal fold abduction was observed on both sides in 5 patients and on 1 side in 2 patients, with no active abduction observed in 1 patient. The follow-up period was more than 5 years in 7 of 8 patients and at least 1.5 years in all patients.

Conclusions and Relevance  Bilateral SLR appears to be a promising treatment option for children with BVFP; it is currently the only option, to our knowledge, with the potential to restore abductor and adductor vocal fold movement. In patients with complete paralysis, this procedure may provide a strategy for airway management and restoration of the dynamic function of the larynx. It could be considered as a first-line technique before endolaryngeal or airway framework procedures, which carry a risk of compromising voice quality.

Pediatric vocal fold paralysis (VFP) is a disorder with varied presentation and congenital, iatrogenic, neurological, and traumatic causes.^1-3 Although the true incidence is unknown because some cases are not immediately recognized,² it is not an unusual entity, especially among the neonatal intensive care unit population. Unilateral VFP (UVFP) usually presents with voice problems, such as a weak cry, and may involve feeding difficulty and/or aspiration with feeding or stridor.^3,4 Bilateral VFP (BVFP) constitutes approximately 60% of pediatric VFP cases^1,3,5 and usually involves considerable airway obstruction and respiratory distress in addition to the problems associated with UVFP.

Although management of UVFP focuses on facilitating closure of the glottis, management of BVFP generally centers on improving the airway. Tracheostomy was traditionally performed, although this procedure is becoming less common with recognition of the possibility of spontaneous resolution and the increased use of static glottic opening procedures, such as arytenoidectomy, arytenoidopexy, and cricoid split with balloon expansion, in the pediatric population.^6-11 Although nonselective laryngeal reinnervation has gained traction as a durable treatment with good outcomes in children with UVFP, ^12-20 bilateral selective laryngeal reinnervation (SLR) for adults with BVFP is still in development. Bilateral SLR in animal models^21-23 has been established for more than 20 years. Outcomes of the first adult patients to undergo bilateral SLR were first reported in 2003.^24,25 To date, the use of this technique in children has not been reported.^26-31 To our knowledge, this study is the first case series of bilateral SLR used to treat children with BVFP.

The study participants included 8 children with BVFP who underwent bilateral SLR at 2 tertiary care academic institutions from November 2004 to August 2018. The preoperative diagnoses, causes of BVFP, surgical intervention techniques, and outcomes over the course of follow-up care were characterized. This retrospective case series was approved by the institutional review board of the University of Utah, Salt Lake City, and the requirement for informed consent was waived because deidentified data were used.

Laryngeal electromyography (EMG) was performed via direct laryngoscopy by an assisting neurologist in the operating room in all 8 cases. Bilateral SLR was performed via the transcervical approach. Although the exact technique used varied by patient, the general principle was to selectively provide independent reinnervation of the laryngeal abductors and adductors. For inspiratory vocal fold abduction, reinnervation was achieved via a jump graft (greater auricular nerve donor) from 1 of the cervical rootlet contributions to the phrenic nerve to the bilateral posterior cricoarytenoid muscles (in 7 of 8 cases) to preserve as much ipsilateral diaphragmatic innervation as possible. In the remaining case, the patient had some limited abduction on inspiration on the right, so the decision was made to preserve this motion by reinnervating only the left posterior cricoarytenoid muscle. The contralateral phrenic nerve was preserved. Reinnervation for vocal fold adduction was provided from the ansa cervicalis or the thyrohyoid branch of the hypoglossal nerve via neurorrhaphy with or without a cable graft to the adductor branch of the recurrent laryngeal nerve or a direct nerve implant into the thyroarytenoid muscle^23,32 (Figure). The bilateral SLR technique used in children is identical to that used in adults, as previously described.^22-24,27 Postoperative evaluation included flexible laryngoscopy and voice assessment using the GRBAS (grade, roughness, breathiness, asthenia, strain) scale, with each domain scored from 0 (normal) to 3 (severe).³³ The time to decannulation date was also documented.

The study participants included 8 pediatric patients (2 girls, 6 boys); the median age at diagnosis was 1.2 years, and a median age at the time of reinnervation surgery was 9.3 (range, 2.2 to 18.0) years. On preoperative evaluation, 5 patients were found to have BVFP as a result of iatrogenic injury and 3 had congenital BVFP (Table 1). None of the patients had an underlying genetic syndrome or disorder. Seven patients had complete BVFP, and 1 had complete immobility of 1 vocal fold and very limited movement of the other. Two patients did not have preoperative tracheostomies; 1 had complete BVFP with both vocal folds in the abducted position, whereas the other had 1 vocal fold in the medialized position and the other in an abducted position. The remaining 6 patients were tracheostomy dependent prior to presentation.

Preoperative laryngeal EMG was performed in all 8 patients. Results showed either no function (5 patients) or considerable paradoxical dysfunction of the posterior cricoarytenoid muscles (3 patients) (Table 2).

Long-term improvement in voice quality, which was assessed using the GRBAS scale (0, normal; 3, severe), was seen in 6 of 8 patients. In 2 patients, GRBAS scores remained the same before and after surgery. None of the patients experienced impaired swallowing for more than 2 weeks after surgery (Table 3). Six patients with tracheostomy dependence prior to surgery were decannulated (Table 4). The mean time from reinnervation to decannulation was 18.4 (17.5) months. In 1 patient, daytime tracheostomy tube capping was initiated 10.5 months after reinnervation. The patient was then lost to follow up for 3.5 years and then presented again at age 10 years, having been capped at all times for 6 months. It is impossible to assess how long after surgery this patient could have been decannulated. After excluding this patient, the mean time to decannulation was 11.2 (8.0) months.

On flexible laryngoscopy, inspiratory vocal fold abduction was observed on both sides in 5 patients and on 1 side in 2 patients (Video). No active inspiratory abduction was seen in 1 patient, although the patient’s respiratory function had improved clinically (Table 3).

Follow-up was more than 5 years in 7 of 8 patients and at least 1.5 years in all patients. In all patients, the functional improvements in voice quality and airway persisted through the last follow-up visit.

Voice, airway, and swallowing problems in children can be caused by VFP. Unilateral vocal fold paralysis is often a result of surgical trauma, although it can be congenital, neurologic, infectious, or related to birth trauma.^2,4 As the fields of neonatology and neonatal cardiothoracic surgery evolve, more medically complex patients and those with congenital heart defects are surviving past infancy, but some of these patients are left with UVFP or BVFP.^2,10,34 Although BVFP may be caused by iatrogenic injury, it is more commonly congenital or neurological in nature.^2,10,35,36

The management of BVFP focuses primarily on opening or widening the glottic aperture to improve breathing. Traditionally, children with BVFP have been treated with tracheotomy to secure the airway, although this procedure has been shown to be avoidable in a considerable portion of patients.^{1,2,6,8,10,36-38} Much of the literature regarding management of pediatric BVFP has focused on alternatives to tracheotomy in children with BVFP-related airway compromise. Techniques such as arytenoidectomy, cordotomy, suture lateralization, and endoscopic posterior cricoid split with or without balloon dilation and/or cartilage grafting have been shown to be successful in achieving extubation or decannulation.^{2,6,8,10,11,36,37,39-42} The avoidance of tracheotomy and its associated morbidities is an important consideration. However, in all of these alternative techniques, the improvement in breathing comes with the risk of deleterious consequences in terms of voice quality and the ability to protect the airway as a result of the static opening of the glottic aperture. Although glottic enlargement procedures may be the only alternative to tracheotomy for patients with bilateral impaired vocal fold mobility caused by arytenoid joint fixation or glottic stenosis, the same is not true for patients with neurogenic paralysis. Patients with BVFP with a neurological cause have the potential to regain dynamic laryngeal function with the restoration of neural input through bilateral SLR. These patients have an opportunity to regain considerable abduction as well as adduction, the latter of which cannot be improved by a glottic opening procedure. In the case of bilateral SLR failure, all of the previously described endolaryngeal and framework procedures are still options for further treatment. In contrast, once a glottic opening procedure is performed, voice quality may be permanently compromised and laryngeal scar tissue may have adverse consequences on the restoration of motion if bilateral SLR is pursued at a later stage.^43-45

Bilateral SLR is a technique that has been used in animal models and applied in adult humans for several years as an experimental procedure.^21-25 More than 80 adult patients have undergone this procedure with good results.⁴⁶ The present case series assesses the use of bilateral SLR as a method to reanimate the vocal folds, resulting in improvement in breathing without sacrificing voice quality or airway protection. To our knowledge, this procedure is the only treatment for BVFP that can restore the dynamic function of the pediatric larynx. Although various techniques for bilateral SLR in adults have been described,^24,27,29-31 to our knowledge, bilateral SLR in children has not been described. Given our increasing experience with bilateral SLR in children, we believe that this procedure may be considered safe and appropriate to treat children with BVFP as young as 2 years of age. We believe that early reinnervation is beneficial because it allows children to be decannulated and have improved voice and speech quality before they begin going to school.

The management of both pediatric UVFP and BVFP could be complicated by the potential for spontaneous resolution, making decisions regarding surgical management difficult. de Gaudemar et al⁴⁷ reported spontaneous resolution of VFP in 63.7% of children with congenital VFP that usually occurred before 6 months of age but could occur as late as 4 to 5 years of age. Children with congenital BVFP were less likely to have spontaneous resolution, although it was observed in 52% of patients. In their long-term retrospective study, Daya et al¹ found that patients with BVFP could take years to recover, with 18% of patients with idiopathic BVFP recovering after 5 to 11 years. Jabbour et al³⁶ found that spontaneous BVFP could occur in a delayed fashion, with 32% of patients with BVFP experiencing resolution at a mean of 11.6 months, whereas the condition resolved after more than 3 years in some cases. Berkowitz⁴⁸ observed that spontaneous improvement of vocal fold movement in patients with congenital BVFP was usually sufficient enough to allow decannulation to be performed between the ages of 5 to 10 years.

Because spontaneous recovery may be seen years after diagnosis, some authors recommend a conservative approach with close observation for several years before considering surgery in patients who are tracheostomy dependent.^{1,3,10,36-38,48} Other authors recommend an observation period of approximately 1 year prior to surgical intervention because delayed spontaneous recovery is the exception rather than the rule.⁴⁷ Laryngeal EMG may provide useful prognostic information on patients with UVFP or BVFP.^49-52 For patients with newly diagnosed BVFP, our practice is close observation for 1 year prior to operative airway evaluation and laryngeal EMG. If the vocal cords are mobile on palpation and there is no evidence of recovery on laryngeal EMG, observation continues until the patient is 2 to 3 years of age. Most patients who have spontaneous resolution of BVFP will show evidence of recovery within 2 years.^36-38,47 Performing bilateral SLR at 2 to 3 years of age allows functional reinnervation to take place in time for the child to be decannulated before starting school. In patients with BVFP who are tracheostomy dependent, the risks and consequences for quality of life associated with a long-term tracheostomy must be weighed against the potential for spontaneous recovery. In our experience, parents of children who are tracheostomy dependent are often eager to move toward decannulation via bilateral SLR, even in the setting of a relatively small chance of spontaneous recovery without intervention.

In this case series, after bilateral SLR, there was movement of the true vocal cord (TVC) bilaterally in 5 patients, unilaterally in 2 patients, and no motion visualized in 1 patient. Respiratory improvement was seen in all 8 patients, including the 1 in whom no movement was seen on postoperative laryngoscopy. The cause of airway improvement in this last case is unknown. It is possible that resolution of synkinetic adduction after division of the recurrent laryngeal nerves could result in more abducted positioning of the TVCs even in the absence of true reinnervation. Similarly, scar formation after dissection of the posterior cricoarytenoid muscle could cause rotation of the arytenoids, resulting in static TVC abduction. Of note, in this patient, although the bilateral SLR did not restore movement, there were no adverse effects on the airway.

This study has limitations. One limitation is that several techniques were used for bilateral SLR; however, they were all variations on the theme of independent “rewiring” of the adductor and abductor laryngeal muscles on both sides, with inspiratory supply provided by a cervical rootlet contributing to the phrenic nerve, as described in several studies.^22,24,27 Although this variation in the techniques used prevents the study cohort from being perfectly uniform, it does demonstrate that good outcomes can be achieved using more than one method of bilateral SLR. As the preoperative EMGs demonstrated, the laryngeal muscles in patients with BVFP are not always completely denervated but often have some remaining innervation, although it may be dysfunctional. The SLR procedure intentionally completely denervates the larynx by dividing both recurrent laryngeal nerves but preserving the external branch of the superior laryngeal nerve whenever possible. Currently, we favor this denervation to create a clean slate followed by reinnervation of the bilateral laryngeal abductor muscles from a unilateral phrenic nerve root via a branched cable graft, and separate reinnervation of the bilateral laryngeal abductors from the ipsilateral thyrohyoid branch of the hypoglossal nerve via jump graft as previously described in several studies.^21-25,53 This technique repopulates laryngeal motor units and muscles selectively with new nerve sources that independently control vocal fold adduction and abduction. Although the consequences of unilateral partial denervation of the diaphragm have not been measured in children, previous studies in animals and adult patients have reported that when 1 cervical rootlet contributor to the phrenic nerve is sacrificed for use as a donor nerve, normal function of the diaphragm is preserved.^25,53

Other limitations of this study include the small number of patients, the retrospective nature of this case series, and variations in evaluation and follow-up between the 2 academic sites. Potential future directions in research include further evaluation and quantification of the voice quality and swallowing outcomes with objective measures, including acoustic analysis of voice and modified barium swallow tests when indicated. Standardizing the approach to evaluation both preoperatively and postoperatively may be helpful in better characterizing the outcomes of the procedure.

Bilateral vocal fold paralysis is a challenging condition that has considerable consequences for the quality of life of pediatric patients and their families. Although traditional methods of static management to widen glottic aperture usually improve the airway and facilitate tracheostomy decannulation in patients with BVFP, these methods have potential adverse effects on voice quality and swallowing function. To our knowledge, bilateral SLR is the only method that can improve breathing, voice quality, and swallowing, and restore dynamic laryngeal function. This case series illustrates the procedure’s safe and appropriate use in children with BVFP. It may be proposed as the first treatment option, prior to vocally destructive endolaryngeal or airway framework procedures.

Accepted for Publication: December 30, 2019.

Corresponding Author: Janet W. Lee, MD, Department of Head and Neck Surgery & Communication Sciences, Duke University, DUMC 3805, Durham, NC 27710 (janet.w.lee@duke.edu).

Published Online: March 19, 2020. doi:10.1001/jamaoto.2019.4863

Author Contributions: Drs Smith and Marie had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Bon-Mardion, Smith, Marie.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Lee, Smith, Marie.

Critical revision of the manuscript for important intellectual content: All authors.

Administrative, technical, or material support: Lee, Marie.

Supervision: Smith, Marie.

Conflict of Interest Disclosures: None reported.