Paediatric recurrent laryngeal nerve reinnervation at The Children’s Hospital at Westmead: a case series

Introduction

Unilateral vocal cord paralysis (UVCP) refers to immobility of a vocal cord due to disruption of its motor supply. It can be congenital or acquired due to birth trauma, iatrogenic injury, neurologic conditions, prolonged orotracheal intubation, mediastinal or neck neoplasms, or infection (1,2). At The Children’s Hospital at Westmead and worldwide, UVCP most commonly arises in paediatric patients as a complication of corrective cardiac surgery (2). The incidence of UVCP is reported to be in the range of 0% to 43% post patent ductus arteriosus (PDA) ligation, with low-birth weight and pre-term infants carrying a higher risk (3).

Symptoms of UVCP are varied and include airway compromise and dysfunction of voice and swallow, the extent of which depends on the position of the paralysed cord and the compensatory ability of other supporting laryngeal structures to achieve adequate glottic closure. Targeted treatments, both non-surgical and surgical, have been used to improve the symptoms associated with UVCP. These include voice therapy, vocal cord augmentation, medialisation laryngoplasty and recurrent laryngeal nerve (RLN) reinnervation. Vocal cord injection, though simple and immediately effective, is encumbered by the transient nature of its results, warranting repeat procedures. Medialisation laryngoplasty has limited applicability in children given the ongoing maturation of the laryngeal framework until puberty (4).

While it does not restore vocal cord movement, non-selective RLN reinnervation aims to permanently restore muscle tone, bulk, and tension to the affected hemi-larynx, thereby improving or even normalising laryngeal function. It is performed under general anaesthetic and thus does not require the patient’s cooperation. It does not alter the laryngeal framework and thus allows for thyroplasty to be performed down the track in adulthood if required (2). However, there is a significant lag of four to 8 months from time of operation to improvement in voice (5,6). RLN reinnervation in the adult population is well documented, and its use in the paediatric population is of growing interest, particularly given the significant psychosocial impact that UVCP can have.

Methods

Participants

We conducted a retrospective review of a single surgeon series of four consecutive paediatric patients that underwent RLN reinnervation between January 2014 and November 2018. The indication for surgery for all four patients was dysphonia. One patient also had swallowing dysfunction as an indication for surgery. Relevant demographics (Table 1), co-morbidities, adjunct therapies and pre- and post-operative outcomes are discussed.

Method and analysis

All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Helsinki Declaration (as revised in 2013). Written consent for publication of this case series was obtained from the parents/legal guardians of the subjects included. A copy of the written consent is available for review by the editorial office of this journal.

Formal pre- and post-operative vocal acoustic assessment (instrumental assessment of voice quality using acoustic software) and perceptual assessment (clinician and parental judgement of voice quality) was carried out (Table 2). Several parameters were evaluated in the vocal acoustic assessment, including average fundamental frequency (the average frequency of vibration of the vocal folds during speech, which correlates with the perceived pitch of the patient’s voice) and harmonic to noise ratio (a measure that quantifies the amount of additive noise in the voice signal, correlating with vocal clarity). Maximum phonation time (MPT) was measured using the longest sustained vowel sound. Intensity (a measure of vocal loudness) was evaluated at habitual (conversational) and maximal (shouting) levels. Two patients (Patients 1 and 4) did not undergo pre- or post-operative vocal acoustic analysis. Patient 1 was too young for formal vocal acoustic analysis, and Patient 4 was lost to follow up before post-operative vocal analysis could be completed. Pre- and post-operative swallowing analysis was carried out for one patient (Patient 1). Overall pre- and post-operative GRBAS scores (an auditory-perceptual scale used to rate overall grade, roughness, breathiness, asthenia, and strain on a 4-point scale, in which 0 is normal, 1 is mild, 2 moderate and 3 severe) were also collected.

Prior to surgery, all patients underwent had undergone direct laryngoscopy and bronchoscopy and vocal cord augmentation, and during these procedures laryngeal electromyography (LEMG) was also performed to confirm RLN denervation.

Surgical technique

Under general anaesthesia, all four patients underwent left RLN re-innervation with anastomosis of a branch of ansa cervicalis to the left RLN. Two patients also received injection laryngoplasty with Gelfoam^® (Pfizer Inc., New York, USA) at the same time.

The surgical approach was via a transverse left-sided neck skin-crease incision centred over the sternocleidomastoid muscle (SCM). The anterior border of SCM was skeletonised, identifying the omohyoid. The omohyoid was then retracted and the branches of the ansa cervicalis dissected along the internal jugular vein and preserved. A vein graft was harvested from the anterior jugular or middle thyroid veins and set aside. The left RLN was identified via careful dissection in the trachea-oesophageal groove, transected and tunnelled under the SCM. The donor nerve was stimulated to ensure electrical activity was present. Using an epiperineural technique, end to end anastamosis with a size matched branch of the ansa cervicalis was performed under an operating microscope using 10/0 Nylon. The vein graft was split and wrapped around the anastomosis to form a protective cuff and prevent axonal escape, and the wound was closed in layers without a drain.

Results

Patient 1

Patient 1 was born at 25 weeks’ gestation, with left UVCP following PDA ligation at 2 weeks. He was noted to have a very weak, hoarse cry post-operatively (GRBAS 3) and left UVCP was confirmed on nasoendoscopy at 8 weeks. Several speech pathology reviews identified that he was unsafe for sucking feeds. Barium swallow at 3 months showed delayed swallowing, initiated at the pyriform sinuses. He also showed deep penetration on thin fluids with a prolonged amount of time required to clear it from his airway. He was noted to be below the third percentile for weight and length at 16 months. He underwent repeated injection laryngoplasties with Gelfoam^® (Pfizer) at 4, 7 and 10 months and hydroxyapatite at 20 months, each with temporary improvement in voice quality, though his feeding remained problematic. Given his young age, pre-operative vocal acoustic analysis was not possible.

Left RLN reinnervation was performed at 22 months of age. At 5 months post-operatively, he was noted to have significant improvement in his voice and feeding. He demonstrated excellent catch-up growth to the sixth and fourth percentile for weight and length respectively. On assessment 2 years post-operatively, he had a normal voice (GRBAS 0).

Patient 2

Patient 2 was born at 23 weeks’ gestation, with left UVCP following PDA ligation at 3 weeks. He first underwent Gelfoam^® injections to the vocal cords at 1 year of age for airway protection on swallowing, with temporary improvement in the volume of his voice. At the time of surgery, his swallowing difficulties had resolved, but he had deteriorated to a whispery, essentially aphonic voice with an inability to increase the volume of his voice to attract attention. Acoustic and functional appraisal could not be completed prior to surgery as there was insufficient voicing for measurement. Despite his attempts and motivation, he was not a suitable candidate for voice therapy because he was unable to achieve sufficient vocal fold closure to generate voice (GRBAS 3).

Left RLN reinnervation was performed at 10 years of age, with left Gelfoam^® injection laryngoplasty performed at the same time. At follow-up 1 year post-operatively, Patient 2 was noted to have significant improvement in voice despite a persistent strained vocal quality with mild roughness. Laryngeal elevation during voicing was noted, indicating recruitment of additional muscles to compensate for reduced vocal cord movement.

Acoustic assessment revealed improved volume and ability to sustain phonation. He had a conversational pitch higher than other boys his age, likely due to the strained and breathy quality of his voice. Breathy voice quality and difficulties maintaining consistent closure of his vocal cords (with voice cutting in and out at times) meant that his maximal phonation time remained below the norm. He also demonstrated reduced ability to lower pitch as he was unable to maintain sufficient closure of the vocal cords at lower pitches and his voice cut out. He was better at achieving high pitches, with some vocal strain when doing so (GRBAS 2). Post-operative nasoendoscopy revealed good bulk and tone of the paralysed left vocal cord, which enabled glottic closure with adduction of the right vocal cord on phonation.

Patient 2’s parents reported gradual improvement in volume and vocal stamina post-operatively, with less ‘gasping’ breathing during speaking. His ability to be better heard was reported to have a positive impact on his ability to functionally use his voice for socialising and engaging in school.

Patient 3

Patient 3 was a 6-year-old male born at 27 weeks’ gestation, with left UVCP following PDA ligation at 3 weeks. Nasoendoscopy performed at 8 weeks confirmed left UVCP. He underwent Gelfoam^® injection to the left vocal cord at 5 years of age with temporary improvement in his voice quality. Prior to RLN reinnervation surgery, Patient 3 was found to have a moderate-to-severely breathy voice with reduced voice volume. Mild roughness and moderate vocal strain were also noted throughout, with a very short phonation time (GRBAS 3).

Left RLN reinnervation was performed at 6 years of age, with left Gelfoam^® injection laryngoplasty performed at the same time. At follow up 9 months post-operatively, Patient 3 was found to have a mildly breathy and strained voice. He spoke quietly during conversation, however readily achieved good volume when prompted to complete structured assessment tasks (e.g., counting, saying ‘ah’). He also demonstrated increased ability to speak in one breath. His inspiration was less audible and stridulous compared to prior reviews. Overall, his voice was perceptually improved compared to previous reviews (GRBAS 1).

Patient 4

Patient 4 was born at term with hypoplastic left heart syndrome, for which he underwent staged Norwood procedure. He was noted to have a hoarse voice and aspiration post stage 2 (bi-directional Glenn operation) at 4 months. Nasoendoscopy revealed a short, paralysed left vocal cord for which he underwent Gelfoam^® laryngoplasty at 16 months. Left RLN reinnervation was performed at 3 years of age.

Discussion

This study reports on the first four cases of paediatric RLN anastomosis for iatrogenic UVCP performed at The Children’s Hospital at Westmead. Though a relatively new procedure in the paediatric population, experience in RLN reinnervation in adult patients dates back nearly one hundred years. RLN reinnervation was first performed in the 1920s by Chevalier Jackson and Charles Frazier (7). This was developed by Crumley and Izdebski in 1986, who described the ansa to RLN anastomosis technique that is most commonly used today (4). Since then, several larger retrospective case series and a meta-analysis including over 1,000 patients have demonstrated good phonatory results with this technique in adults, with attainment of near normal or normal voice quality (6,8,9). Paniello et al. noted that adults under 52 years of age had more favourable outcomes post reinnervation compared to the older age group, attributed to superior neural recovery at a molecular level (10). In their large retrospective review, Li et al. also found RLN reinnervation to be more effective in patients under 60 years (11).

The first case of paediatric RLN was reported by Crumley in 1991, who performed RLN reinnervation in an 8-year-old boy with UVCP secondary to iatrogenic injury during cardiac surgery. At 7 months post-operatively, his voice had markedly improved—reported as ‘normal’ by his parents (12). Zur et al.’s 2015 case series compared outcomes from injection laryngoplasty with RLN reinnervation in 33 children with dysphonia, with the latter group demonstrating better long-term improvements in voice quality and satisfaction (1). In 2016, Smith et al. published the largest case series to date of 35 patients aged one to 21 years who underwent RLN reinnervation. Patient or parent reported percent normal voice rating improved in all patients, from an average of 40% pre-operatively to 81% post-operatively (13). Most recently, Ongkasuwan et al. published their series of 32 paediatric patients who underwent RLN reinnervation, with significant improvement in paediatric voice related quality of life questionnaires and objective voice measures post-operatively. Time from injury did not significantly impact voice outcomes (14).

In our cohort, all patients had undergone previous cardiac surgery resulting in left vocal cord paralysis. Three patients were extremely premature (less 30 weeks’ gestation), which is known to impact neurologic, respiratory, and developmental skills (1,15). We found that the procedure was well tolerated in our patients, with short operating time, no post-operative complications and early discharge from hospital at post-operative day one. Perceptual assessment by the parents, clinicians and speech therapists involved in this study indicated noticeable and progressive improvement in voice post-RLN reinnervation. Post-operative GRBAS scores showed significant improvement. All patients had a pre-operative GRBAS of 3, consistent with severe impairment, with Patient 2 being essentially aphonic. Post-operatively, Patient 1 had a normal voice (GRBAS 0), Patient 2 had mild to moderate impairment (GRBAS 2), and Patient 3 demonstrated mild impairment (GRBAS 1).

Improvements were also noted in objective measures of voice. Patient 2’s MPT improved from unrecordable pre-operatively (aphonic) to 5.5 seconds. Patient 3’s MPT improved from 3 to 5 seconds. Interestingly, Patient 3’s maximal intensity on formal acoustic analysis was recorded at 80 dB pre-operatively, and 72 dB post-operatively despite subjective reports from his family that his overall and maximal volume had improved. He also continued to speak quietly in conversation despite being able to achieve good volume when prompted. This may reflect findings by Ongkasuwan’s group, who noted that some older children continued to use a softer, breathier voice by choice, however could become louder with prompting. It was postulated that for some children, their weaker preoperative voice becomes entwined with their identity and voice therapy is needed to increase their comfort with their new post-operative voice (13). Time from injury did not adversely impact results of reinnervation. Patients 2 and 3 underwent surgery at 10 and 6 years post-injury respectively, and both showed improved acoustic results and excellent perceptual results. At 1 year follow-up, Patient 2 was able to achieve complete closure of a large pre-operative glottic gap on phonation.

Limitations of our study

This case series is limited by incomplete pre-operative and post-operative assessments, and variable time to follow-up. This was in part due to the multiple co-morbidities suffered by all four patients, with other more pressing medical issues taking precedence over their post-operative follow-up. Age at surgery also limited the feasibility of carrying out formal pre-and post-operative vocal acoustic analysis, with Patients 1 and 4 being 22 months and 3 years of age respectively at the time of surgery. One subject (Patient 2) was unable to achieve voicing pre-operatively, making acoustic analysis impossible.

The difficulties in objective analysis of voice outcomes in the paediatric population is highlighted, as objective measures may not be feasible, and do not correlate with subjective parent-reported or assessor-reported voice outcomes. In this patient population, quality of life studies is an important avenue of research, with patient and carer satisfaction with voice and swallow outcomes providing the greatest metric of success. In future, the use of validated patient reported outcome instruments such as the paediatric voice handicap index would serve as a useful measure of post-operative symptom change in this population where obtaining accurate acoustic data can be challenging.

Conclusions

Multiple treatment modalities exist to address UVCP, and the chosen intervention must be tailored to the patient at hand. In the literature, RLN reinnervation has been reported to have excellent outcomes from both a voice and swallow perspective. We have successfully introduced RLN reinnervation for paediatric UVCP at The Children’s Hospital at Westmead, and have found it to be a safe procedure that, with appropriate patient selection, can be very effective in improving patients’ symptoms of dysphonia and dysphagia.

Acknowledgments

Our abstract was accepted for presentation at the 5^th Laryngology Society of Australasia Conference, Sydney, Australia taking place between the 10^th and 12^th of November 2023.

Funding: None.

Footnote

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-23-47/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-23-47/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-23-47/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Helsinki Declaration (as revised in 2013). Written consent for publication of this case series was obtained from the parents/legal guardians of the subjects included. A copy of the written consent is available for review by the editorial office of this journal.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Zur KB, Carroll LM. Recurrent laryngeal nerve reinnervation in children: Acoustic and endoscopic characteristics pre-intervention and post-intervention. A comparison of treatment options. Laryngoscope 2015;125:S1-15. [Crossref] [PubMed]
2. Setlur J, Hartnick CJ. Management of unilateral true vocal cord paralysis in children. Curr Opin Otolaryngol Head Neck Surg 2012;20:497-501. [Crossref] [PubMed]
3. Smith ME, King JD, Elsherif A, et al. Should all newborns who undergo patent ductus arteriosus ligation be examined for vocal fold mobility? Laryngoscope 2009;119:1606-9. [Crossref] [PubMed]
4. Marcum KK, Wright SC Jr, Kemp ES, et al. A novel modification of the ansa to recurrent laryngeal nerve reinnervation procedure for young children. Int J Pediatr Otorhinolaryngol 2010;74:1335-7. [Crossref] [PubMed]
5. Grover N, Bhattacharyya A. Unilateral pediatric vocal cord paralysis: evolving trends. J Laryngol Voice 2012;2:5-9. [Crossref]
6. Marie JP. Nerve Reconstruction. In: Remacle M, Eckel HE, editors. Surgery of Larynx and Trachea. Belgium: Springer; 2010.
7. Frazier CH. The treatment of paralysis of the recurrent laryngeal nerve by nerve anastomosis. Ann Surg 1924;79:161-71. [Crossref] [PubMed]
8. Crumley RL, Izdebski K. Voice quality following laryngeal reinnervation by ansa hypoglossi transfer. Laryngoscope 1986;96:611-6. [Crossref] [PubMed]
9. Hoey AW, Hall A, Butler C, et al. Laryngeal reinnervation for paediatric vocal cord palsy: a systematic review. Eur Arch Otorhinolaryngol 2022;279:5771-81. [Crossref] [PubMed]
10. Paniello RC, Edgar JD, Kallogjeri D, et al. Medialization versus reinnervation for unilateral vocal fold paralysis: a multicenter randomized clinical trial. Laryngoscope 2011;121:2172-9. [Crossref] [PubMed]
11. Li M, Chen D, Song X, et al. The effect of patient age on the success of laryngeal reinnervation. Eur Arch Otorhinolaryngol 2014;271:3241-7. [Crossref] [PubMed]
12. Crumley RL. Update: ansa cervicalis to recurrent laryngeal nerve anastomosis for unilateral laryngeal paralysis. Laryngoscope 1991;101:384-7; discussion 388. [Crossref] [PubMed]
13. Smith ME, Houtz DR. Outcomes of Laryngeal Reinnervation for Unilateral Vocal Fold Paralysis in Children: Associations With Age and Time Since Injury. Ann Otol Rhinol Laryngol 2016;125:433-8. [Crossref] [PubMed]
14. Ongkasuwan J, Espinosa MCL, Hollas S, et al. Predictors of voice outcome in pediatric non-selective laryngeal reinnervation. Laryngoscope 2020;130:1525-31. [Crossref] [PubMed]
15. French N, Kelly R, Vijayasekaran S, et al. Voice abnormalities at school age in children born extremely preterm. Pediatrics 2013;131:e733-9. [Crossref] [PubMed]