Bridging Gaps Between the Recurrent Laryngeal

7/25/2019 Bridging Gaps Between the Recurrent Laryngeal

http://slidepdf.com/reader/full/bridging-gaps-between-the-recurrent-laryngeal 1/7

Bridging Gaps Between the Recurrent LaryngealNerve and Ansa Cervicalis Using AutologousNerve Grafts

Meng Li, MD, Fei Liu, MD, Song Shi, MD, Shicai Chen, MD, Donghui Chen, MD, and Hongliang Zheng, MD, Shanghai,

People’s Republic of China

Summary: Objectives/Hypothesis. We investigated the clinical efficacy of free nerve grafts in bridging gaps be-tween the recurrent laryngeal nerve (RLN) and ansa cervicalis in patients with unilateral RLN injury.

Study Design. We retrospectively reviewed the charts of 14 patients who underwent relevant free nerve grafting and

assessed the clinical outcomes of this procedure.Methods. Between January 2000 and January 2010, 14 patients with unilateral vocal fold paralysis were enrolled in

this study. In all patients, the RLN was resected and free nerve grafts were applied to bridge the gap between the distal

stump of the RLN and the anterior root of ansa cervicalis during surgery. Videostroboscopy, acoustic analysis, percep-

tual evaluation, maximum phonation time (MPT), and laryngeal electromyography (EMG) were performed both pre-

operatively and postoperatively to assess the clinical outcomes.

Results. Videostroboscopic findings showed that glottic closure, vocal fold edge, vocal fold position, phase symmetry,

and phase regularity were significantly improved postoperatively (P < 0.05), and no paradoxical movements of vocal

folds were observed. Perceptual evaluation showed that overall grade, roughness, breathiness, asthenia, and strainwere also significantly decreased postoperatively (P < 0.05). The acoustic parameters jitter (local) and shimmer (local)

and the mean noise-to-harmonics ratio were significantly lower than the corresponding preoperative values (P < 0.05).The postoperative MPT values were also significantly longer than the preoperative values. Laryngeal EMG revealed

significant improvement in voluntary motor unit recruitment during phonation postoperatively (P < 0.05).

Conclusions. Free nerve grafting is an effective procedure in bridging the gap between the RLN and ansa cervicalis in

patients with unilateral RLN injury, as well as a safe procedure without obvious morbidity. A satisfactory vocal outcome

can be obtained.

Key Words: Free nerve grafting–Vocal fold paralysis–Laryngeal reinnervation–Recurrent laryngeal nerve–Ansacervicalis.

INTRODUCTION

Injury to the recurrent laryngeal nerve (RLN) leads to perma-

nent vocal fold paralysis, which has a significant negative im-

pact on the quality of life. Functional restoration with

laryngeal reinnervation is the mainstay of treatment for patients

with unilateral vocal fold paralysis (UVFP). Numerous studies

have confirmed the excellent vocal outcome of laryngeal rein-

nervation surgery using several potential donor nerves, includ-

ing original RLN, the ansa cervicalis, the hypoglossal nerve,

and others.1–4 Direct end-to-end anastomosis of the severedRLN main trunk may restore reinnervation to laryngeal muscles

to some or to a large extent. However, because abductor and ad-

ductor fibers are distributed randomly in the RLN, original RLN

reanastomosis could result in misdirected regeneration, that is,

abductor fibers reinnervating the adductor muscles, and vice

versa, which may lead to ‘‘laryngeal synkinesis.’’ If this occurs

to a significant extent, it may result in paradoxical movements

of vocal folds or even laryngospasm.5 Anastomosis of a single

branch of ansa cervicalis-to-RLN was reintroduced by Crum-

ley2 and followed by Lee et al,3 Miyauchi et al,6 Oslon et al,7

Smith et al,8 and other investigators,9 with good voice results

reported. We modified Crumley’s procedure later by using the

main branch of the ansa cervicalis because the main branch

of ansa cervicalis in our procedure possesses more motor fibers

than a single branch.10–12 And our previous large-scale studyfound satisfactory or good voice outcomes in patients who un-

derwent main branch of ansa cervicalis-to-RLN anastomosis.11

The reinnervated vocal folds using ansa cervicalis transfer may

be fixed at midline or near-midline position, which is a typical

type I (favorable) synkinesis as described by Crumley,13 the re-

covery of vocal function may be because of the recovery of bulk, tension, and symmetry of the paralyzed vocal folds. How-

ever, in some clinical circumstances, a nerve defect is foundduring reinnervation procedure because of the injury to ansa

cervicalis so as to make tension-free ansa cervicalis-to-RLN

anastomosis impossible. Immediate bridging of RLN gaps

with autologous nerve grafts between the two severed ends of

the RLN during thyroid surgery resulted in excellent phonatory

function in previous studies.14,15 However, this procedure may

also lead to the same ‘‘misdirected reinnervation’’ that can

occur in direct end-to-end anastomosis of the RLN. For the

above reason, we modified our surgical technique on the basis

Accepted for publication January 14, 2013.M.L., F.L., S.S contributed equally to this research.This work was supported by grant no. 81100724, 81170899, 81070774 for science re-

search from National Natural Science Foundation of China and grant no. 10XD1405500from the Science and Technology Commission of Shanghai Municipality. The authorshave no other financial relationships or conflicts of interest to disclose.

From the Department of Otolaryngology-Head and Neck Surgery, Changhai Hospital,Second Military Medical University, Shanghai, People’s Republic of China.

Address correspondence and reprint requests to Hongliang Zheng, Department of Otolaryngology-Head and Neck Surgery, Changhai Hospital, Second Military MedicalUniversity, Shanghai 200433, People’s Republic of China. E-mail: [email protected]

Journal of Voice, Vol. 27, No. 3, pp. 381-3870892-1997/$36.00 2013 The Voice Foundationhttp://dx.doi.org/10.1016/j.jvoice.2013.01.009

mailto:[email protected]


http://dx.doi.org/10.1016/j.jvoice.2013.01.009

http://dx.doi.org/10.1016/j.jvoice.2013.01.009





of these previous studies, which was the first attempt to bridge

the gaps between the RLN and anterior root of ansa cervicalis

with autologous nerve grafts in 14 patients with UVFP and as-

sessed the long-term effects of this procedure on postoperative

phonatory function and evaluated whether unfavorable synki-

nesis existed.

MATERIALS AND METHODS

Patient characteristics

Our study was approved by the Institutional Review Board of

Second Military Medical University. The charts of 16 patients

with UVFP who underwent free nerve grafts bridging RLN

gaps from January 2004 to January 2010 and who were fol-

lowed up for at least 2 years were reviewed. Two patients

who were lost to follow-up were excluded from this study. A to-

tal of 14 patients were ultimately enrolled in the study (five

males, nine females; mean age, 38.0 years; range, 19–51 years).

The present study included patients with UVFP who met the

following criteria: (1) patients in whom UVFP was caused by

thyroid cancer and neck dissection (11 cases) or neck wounds(two cases) and in whom the ipsilateral ansa cervicalis was

also injured but the beginning part of the anterior root of the

ansa cervicalis was intact and (2) patients in whom UVFP

was caused by thyroid cancer and neck dissection (TC&ND)

(one case) and in whom contralateral ansa cervicalis-to-RLN

anastomosis could not guarantee a tension-free status. Informedconsent was obtained from all patients involved in this study. A

minimum of 6 months for thyroid surgery and 12 months for

neck wounds passed after the onset of RLN injury to allow

for possible spontaneous recovery or compensation. The me-

dian denervation course was 12.8 months (range, 7–20 months),

and the median follow-up period was 33.8 months (range, 24–

47 months). All patients were followed up every 2–3 monthsuntil at least 2 years postoperatively. All examinations, includ-ing laryngeal electromyography (EMG), were performed at

2 years postsurgery. Table 1 summarizes the demographic

data of the patients.

Operative procedure

Under general anesthesia, a horizontal oblique incision was

made at the original incision of the primary surgery or the orig-

inal incision was extended toward the posterior and upper part

of the neck to expose the ansa cervicalis and the beginning part

of its anterior root. The fascia between the strap muscles and

the sternomastoid muscle was opened on the side of the para-

lyzed vocal fold. The ipsilateral ansa cervicalis was explored.

The loop of the ansa cervicalis was usually seen overlying the

common carotid artery or jugular vein. Generally, several

branches arise from the loop of the ansa cervicalis and inner-

vate the strap muscles (Figure 1). In this series of cases, the ip-

silateral ansa cervicalis was injured and was not long enough

for tension-free direct anastomosis to RLN. However, the be-

ginning part of the anterior root of the ansa cervicalis remained

intact. The thyroid cartilage was engaged with a skin hook and

rotated gently anteriorly, and the RLN was dissected retro-gradely as far as possible until the injury site of the RLN

was exposed. It was then transected and transferred superiorly

to approximate the beginning part of the anterior root of the

ansa cervicalis. Because of the insufficient length of the distal

extralaryngeal RLN stump, which could not guarantee

tension-free anastomosis to the anterior root of the ansa cervi-

calis, the distal RLN stump was traced distally to the lateral

and inferior part of the cricoarytenoid joint, saving approxi-

mately 1 cm of length. The ipsilateral cervical plexus was

then dissected and exposed. The great auricular nerve or other

branches of the cervical plexus, including the transverse cuta-

neous nerve and lesser occipital nerve (which has no bifurca-

tions), were harvested for use. If the cervical plexus wasinjured in theprimary surgery, we used the suralnerve as a sub-

stitute for the graft. The distal stump of the RLN was

TABLE 1.

Demographic Data on Patients

Patient No. Age Gender Injury Course (mo) Causes of RLN Injury Types of Nerve Graft Length of Graft (cm)

1 30 M 10 TC&ND TCN 1.8

2 29 F 15 TC&ND LON 2.2

3 41 F 8 TC&ND GAN 2.2

4 51 M 11 TC&ND LON 2.5

5 36 F 10 TC&ND TCN 2.56 44 M 18 NT TCN 2.8

7 50 F 13 TC&ND and IARI TCN 2.5

8 33 F 20 TC&ND GAN 3.4

9 19 F 17 TC&ND GAN 4.0

10 49 M 7 TC&ND LON 2.5

11 40 F 10 TC&ND TCN 2.5

12 38 F 15 TC&ND SN 3.2

13 42 M 11 NT GAN 3.6

14 30 F 14 TC&ND GAN 3.5

Abbreviations: TC&ND,thyroid cancerand neck dissection; TCN,transverse cutaneousnerve;LON, lesseroccipital nerve; GAN, great auricularnerve;NT, neck

trauma; TC&ND and IARI, thyroid cancer and neck dissection with ispilateral anterior root of ansa cervicalis injury; SN, sural nerve.

Journal of Voice, Vol. 27, No. 3, 2013382



anastomosed with one end of the free nerve graft, whereas the

other end was anastomosed to the beginning part of the ante-

rior root of the ansa cervicalis (Figure 2). Finally, the wound

was closed in layers.The choice of autologous nerve grafts is dependent on the di-

ameter of the distal stump of the RLN and the length of the gap

between the RLN and ansa cervicalis ends. Generally, the prox-

imal and distal nerve ends are trimmed until the normal fascic-

ular structure is revealed. The anastomoses of the free nerve

grafts to the RLN and ansa cervicalis ends were performed us-

ing microsurgical instruments under surgical magnification. An

epineural circumferential suturing technique was adopted when

performing the anastomoses.

Videostroboscopy

All patients were observed during /eee/ phonation at a comfort-

able loudness and pitch as long as possible using a videostrobo-scope (RICHARD WOLF GmbH, model 5570, Knittlingen,

Germany), and dynamic videos were recorded. Three experi-

enced senior laryngologists who had not performed any opera-

tions reviewed the videos. The videos were randomized and

the reviewers were blinded to whether the videos were preoper-

ative or postoperative. Visual laryngeal analysis included glottal

closure (0, complete; 1, slightly incomplete; 2, moderately in-

complete; and 3, severely incomplete), vocal fold position (0,

midline; 1, paramedian; 2, intermediate; and 3, lateral), vocal

fold edge on the paralyzed side (0, straight; 1, mildly bowed;

2, moderately bowed; and 3, severely bowed), phase symmetry

(0, normal; 1, mildly asymmetrical; 2, moderately asymmetrical;

and 3, severely asymmetrical), and phase regularity (0, normal; 1,

mildly irregular; 2, moderately irregular; and 3, severely irregu-lar). Phase symmetry and phase regularity were considered to be

severely asymmetrical and irregular if the voice was too poor to

obtain a stroboscopic image. Consensus among the reviewers

was reached on the visual appearance of the larynx.

Vocal function assessment

Vocal function assessment included perceptual evaluation,acoustic analysis, and maximum phonation time (MPT). Preop-

erative and postoperative voice samples that contained the sus-

tained vowel /a/ and connected speech samples were subjected

to perceptual evaluation and acoustic analysis. The recording

equipment included a digital audiotape recorder and a dynamic

microphone (Tiger Electronics, Inc., North Reading, MA). Fivelaryngologists performed the voice perceptual evaluation using

a perceptual rating scale for voice quality and characteristics.

Theratingswereperformed in a blinded fashionwiththe samples

randomly arranged. The listeners were asked to grade connected

speech samples for overall grade, roughness, breathiness, asthe-

nia, and strain (GRBAS). This perceptual scale allowed each

listener to rate the voice quality on a scale of 0–3 (0, normal;

1, mild; 2, moderate; and 3, severe) for each of the above param-

eters. The values were averaged among the five listeners.

The acoustic parameters of the sustained vowel /a/ were eval-

uated using Praat software (Version 5.1.12; Boersma & Wee-

nink, 2011, http://www.praat.org/ ). The acoustic parameters

involved the mean noise-to-harmonics ratio (NHR) and mea-

sures of phonatory stability; namely, jitter (local) and shimmer

(local). MPT (the duration of sustained phonation of the vowel

/a/ after maximum inspiration) was measured preoperatively

and postoperatively.

Laryngeal EMG

A four-channel EMG device and concentric needle electrodes

(Dantec Counterpoint, Copenhagen, Denmark) were used for

laryngeal EMG. To test for proper needle positioning, the unaf-

fected vocal fold was examined first. The activities of the bilat-

eral thyroarytenoid (TA) muscles were recorded in two stages:

(1) when the patients breathed quietly with relaxation of thebody and (2) when the patients sustained the vowel /eee/ with

the greatest exertion. A board-certified otolaryngologist per-

formed the EMG, and a neurologist operated the EMG machine

and interpreted the results. The neurologist rated voluntary mo-

tor unit recruitment (VMUR) using the following scale: 0, full

interference; 1, mixed interference; 2, simple interference; and

3, without motor unit potential.

Statistical analysis

The perceptual evaluation data are presented as medians (lower

and upper quartiles). The acoustic analysis data and MPT

FIGURE 2. Intraoperative picture of free nerve grafting in bridging

gaps between the ansa cervicalis and RLN. Intraoperative picture

showing that the free nerve graft (great auricular nerve, b) was inter-

posed between the ends of the ansa cervicalis (c) and RLN (a). The ar-

rows show the two anastomosis sites.

FIGURE 1. Intraoperative findings of the anatomy of the cervical

plexus. The cervical plexus usually contains the transverse cervical

nerve (a), great auricular nerve (b), lesser occipital nerve (c), and

supraclavicular nerve (data not shown).

Meng Li, et al Free Nerve Grafting for RLN Injury 383

http://www.praat.org/

http://www.praat.org/



values are presented as x±s. Statistical differences between pre-

operative and postoperative videostroboscopy and EMG were

analyzed using the chi-square ðc2CMHÞ test. Statistical differ-

ences between preoperative and postoperative perceptual eval-

uations were identified using Friedman test, and statistical

differences between the preoperative and postoperative acous-

tic analyses and MPT data were analyzed by a paired t test.

All analyses were conducted using SAS software (Version9.1.3; SAS Institute Inc, USA). A P value of <0.05 was deemed

to indicate statistical significance.

RESULTS

Patients felt a sudden obvious improvement in the voice at 4–11

months after the reinnervation procedure; the voice was im-

proved to varying degrees in these patients. A stable voice

was achieved at 1–3 months after the sudden improvementwith no further improvement or deterioration.

Videostroboscopic findings

All patients in this series underwent preoperative and postoper-

ative stroboscopic examinations. Table 2 shows the preopera-

tive and postoperative laryngeal appearances of the patients

in this series. Most patients showed severe incomplete glottal

closure, a paramedian or intermediate vocal fold position,

mildly to severely bowed vocal fold edges, and severe irregular

vocal fold vibration during preoperative phonation (Figure 3A

and B). Two years after reinnervation, reinnervated vocal folds

were fixed at midline or near-midline position, no vocal folds

adduction during inspiration or vocal folds abduction during

phonation or involuntary vocal fold jerks, twitches, and jumps

was observed. Glottal closure, vocal fold position, vocal fold

edge, phase symmetry, and phase regularity were significantly

improved compared with the corresponding preoperative values

(Figure 3C and D, P < 0.05).

Vocal function assessment

The postoperative values of the five GRBAS perceptual evalu-

ation parameters were significantly improved compared with

the corresponding preoperative values (P < 0.05) (Table 3).Table 4 summarizes the acoustic analysis results and MPT

values of the patients preoperatively and postoperatively. The

postoperative jitter (local), shimmer (local), and NHR values

were significantly lower than the corresponding preoperative

values (P < 0.05), and the postoperative MPT was significantlylonger than the preoperative MPT (P < 0.05).

Electromyographic findings

All 14 patients underwent preoperative and postoperative laryn-

geal EMG. The electrical activity during laryngeal EMG was

divided into two types: spontaneous activity and VMUR. Ab-normal spontaneous activity in the TA muscles of the affected

vocal folds, including positive waves, fibrillations, or complex

repetitive discharges, were recorded in four patients preopera-

tively compared with none postoperatively. Two patients had

no motor unit potential, nine had simple interference, three

had mixed interference, and none had full interference during

phonation and during inspiration single or mixed interference

could be recorded in TA muscles in six patients preoperatively,whereas 11 patients had full interference, two had mixed inter-

ference, and one had single interference during phonation post-

operatively. These results demonstrate significant improvement

in VMUR during phonation compared with the preoperative

FIGURE 3. Videostroboscopic findings of a representative patient with UVFP with free nerve grafting. Preoperatively, the edge of the paralyzed

right vocal fold appeared mildly bowed, the vocal fold position was paramedian during inspiration (A) and a fully incomplete glottic closure pre-

sented during phonation (B). Postoperatively, the reinnervated right vocal fold was at the midline and appeared straight during inspiration (C). The

bulk of the right vocal fold appeared the same as the left side and the glottic gap completely disappeared during phonation (D).




values (P < 0.05). Though single or mixed interference could berecorded in TA muscles during inspiration in all 14 patients

postoperatively, they showed very low potentials.

DISCUSSION

In terms of peripheral nerve injury, primary end-to-end anastomo-

sis remains the most desirable approach when there is no nerve

defect or when the gap between the two ends of the nerve is

short enoughthat thetwoends canbe sutured without tension.16,17

However, some cases involve greater defects or longer gaps

between the two cut ends; in such cases, when primary

anastomosis will cause excessive tension, it should not be

performed and other alternatives should be considered. In 1927,

Bunnell reported successful nerve grafting,18 and autologous

nerve grafting has since been proven successful in the repair of pe-ripheral nerve defects after injury in many studies. It is considered

to be an effective methodfor bridging nerve gaps that are not ame-

nableto primary end-to-end anastomosis16 because nerve grafting

with no tension yields better results than primary end-to-end anas-

tomosis under tension, as proven by Millesi et al19 and Berger and

Millesi.20 It is also known that excessive tension across a nerve

repair site will impair nerve regeneration, possibly because of

the impairment of microvascular flow in the nerve tissue and

scar formation resulting from connective tissue proliferation.21

Laryngeal reinnervation using direct end-to-end anastomosis

of the RLN may restore laryngeal function.15,22 However, the

RLN has its own anatomical characteristics. The RLNcomprised abductor fibers that innervate abductor muscles and

adductor fibers that innervate adductor muscles. There is no

special segregation of the nerve fibers within the RLN.23 After

direct end-to-end anastomosis of a severed RLN, although the

nerve fibers do regenerate, this regeneration occurs in a random

fashion among adductor fibers and abductor fibers, possibly re-

sulting in ‘‘misdirected regeneration.’’24 Furthermore, because

there are more adductor than abductor fibers in the RLN and be-

cause adductors are stronger than abductorsin the larynx, there-

innervated vocal folds are usually fixed at or near the midlin e;

therefore, phonatory outcomes are significantly improved.15

However, laryngeal synkinesis including paradoxical move-

ments and laryngospasms is a possible complication of direct

RLN anastomosis.5 Crumley2 introduced a new laryngeal rein-

nervation procedure called RLN anastomosis to the ansa cervi-calis. Other researchers, including the authors of this article,

demonstrated its satisfactory phonatory results, with mild laryn-

geal synkinesis but no paradoxical movements or laryngo-

spasm.3,6–8,10 However, this kind of synkinesis does no harm

to vocal function as well as respiratory and deglutition

functions; it is a favorable synkinesis as described by

Crumley.13 On theother side, after ansa cervicalis-to-RLNanas-

tomosis, the affected vocal fold fixed at or near midline, as well

as the restoration of bulk, tension, and symmetry of vocal folds

improved the vocal function postoperatively. We have accom-

plished ansa cervicalis-to-RLN anastomosis in more than

TABLE 2.

Pre- and Postoperative Videostroboscopic Findings

Rating Rank

No. of Videostroboscopic Findings

P ValuePreoperative Postoperative

Glottal closure

Complete 0 11 <0.001

Slightly incomplete 0 3

Moderately incomplete 2 0

Severely incomplete 12 0

Vocal fold position

Midline 0 12 <0.001

Paramedian 11 2

Intermediate 3 0

Lateral 0 0

Vocal fold edge

Straight 1 12 <0.001

Mildly bowing 2 2

Moderately bowing 2 0

Severely bowing 9 0

Phase symmetry

Normal 0 12 <0.001

Mildly asymmetrical 0 2

Moderately asymmetrical 1 0

Severely asymmetrical 13 0

Regularity

Normal 0 12 <0.001

Mildly irregular 0 2

Moderately irregular 1 0

Severely irregular 13 0




300 patients with UVFP,11 14 of them were found a gap between

ansa cervicalis and the distal stump of RLN caused by trauma or

neck dissection surgery for thyroid cancer, making tension-free

direct anastomosis impossible. Previous studies have proven

that bridging severed RLN ends using autologous nerve grafts

can achieve satisfactory phonatory outcomes14,15; however,

the main disadvantage of this procedure is the possibility of

laryngeal spasm or paradoxical movements of vocal folds due

to misdirected regeneration of the abductor and adductorfibers, as may occur during direct end-to-end anastomosis of

the RLN. In the present study, we first bridged the gap by inter-

posing free nerve grafts between anterior root of ansa cervicalis

and the distal stump of RLN with satisfactory phonatory out-

comes on 2-year postoperative follow-up. The same as we re-

ported in previous ansa cervicalis-to-RLN study, no

paradoxical movements or laryngospasm was observed in this

series of patients postoperatively. Our neurophysiologic find-

ings showed effective reinnervation, motor unit recruitments

were recorded in TA muscles during phonation in most cases,

andmild actionpotentials in thesame muscle during inspiration,

in accordance with other reports.3,10

However, autologous nerve grafting has disadvantages, in-

cluding the need for a second operation to harvest the nerve

grafts and concurrent donor-site morbidity such as scarring,

pain, and permanent loss of sensation in the area supplied by

the donor nerve. The decision to perform autologous grafting

is dependent on several factors: the length of the nerve gap be-tween ansa cervicalis and RLN stump; the nature and location

of the injury; any associated donor-site morbidity; and the sur-

geon’s preference. Preoperatively, patients should be informed

of and fully understand the deficits that will result after nerve

graft harvesting. Based on our experiences, some details of

this procedure were described to help improving the surgical ef-

fects of free nerve grafting. The nerve grafts used to bridge gaps

between the distal end of RLN and anterior root of ansa cervi-

calis should meet the following criteria. First, the nerve graft

should be long enough to ensure tension-free anastomosis. Sec-

ond, it should be thin enough to match the distal stump of RLN

in diameter, which will become thinner because of atrophy after

injury, to ensure apposition of the epineural sutures, avoid axon

exposure from the suture site, and decrease scar formation.

Third, the free nerve graft trunk should not have a bifurcation

to ensure a good pathway for laryngeal nerve axons regenera-tion. Taking all the above into consideration, a sensory nerve

such as the sural nerve is commonly used as grafts in other fields

because of its relative ease of harvesting and the possibility of

obtaining a longer graft with lower donor-site morbidity. How-

ever, they are always thicker than the RLN, so they were not the

best choice in the present study. A branch of the cervical plexus

was a relatively better choice. We usually chose a branch with

no bifurcation, of a suitable length, and that matched the RLN

in diameter. In this series of cases, the free nerve graft was taken

from the great auricular nerve, lesser occipital nerve, and trans-

verse cutaneous nerve, all of which are branches of the cervical

plexus. Only in situations in which branches of the cervical

plexus were injured during the primary surgery or due to trauma

and had lost vitality because of scar formation was the sural

nerve adopted as the nerve graft.

The length of the free nerve graft could be saved by dissect-

ing the intralaryngeal segment of RLN as far as possible until to

the level below cricothyroid joint, which was then carried cra-

niomedially to be anastomosed to the anterior root of the ansa

cervicalis. To improve the surgical effect of free nerve grafts,

we also need to pay attention to the following aspects. Nerve

grafts must be placed in a suitable viable tissue bed without

scar, and the grafts should not be so long that they are redundant

or so short that they cause tension. The scar and neuroma at the

nerve ends should be carefully trimmed to expose the normal

TABLE 3.

Pre- and Postoperative Perceptual Evaluation for GRBAS

Parameter N

Preoperative Postoperative

Statistics P ValueMedian (Q L, Q U) Median (Q L, Q U)

Grade 14 2.2 (2.0, 2.2) 0.6 (0.0, 0.8) 14.000 0.0002

Roughness 14 1.7 (1.4, 2.0) 0.4 (0.0, 0.8) 14.000 0.0002

Breathiness 14 1.9 (1.6, 2.0) 0.0 (0.0, 0.4) 14.000 0.0002Asthenia 14 1.2 (0.8, 1.6) 0.0 (0.0, 0.0) 10.2857 0.0013

Strain 14 1.0 (0.0, 1.4) 0.0 (0.0, 0.0) 6.4000 0.0114

TABLE 4.

Pre- and Postoperative Vocal Function Assessment

Parameter N

Data of Vocal Function Assessment ðx±sÞ

T Value P ValuePreoperative Postoperative

Jitter 14 1.72 ± 0.46 0.30 ± 0.15 11.02 <0.001

Shimmer 14 8.56 ± 1.33 3.58 ± 2.01 7.73 <0.001

NHR 14 0.15 ± 0.06 0.02 ± 0.007 8.54 <0.001

MPT 14 5.71 ± 1.72 20.01 ± 3.50 13.73 <0.001




axons and perineurium of the ansa cervicalis and RLN stumps.

These factors are important because they not only ensure good

axon regeneration but also optimize functional recovery.16

The longest RLN defect in this series of patients was 4 cm,

and this patient regained a satisfactory voice result when

followed up 2 years after the operation. Theoretically, the lon-

ger a gap is, the more time it will take for nerve regeneration;

thus, the length of the gap is an influencing factor in the effi-ciency or degree of nerve regeneration. However, the maximal

nerve defect for which free nerve grafting can be successfully

applied remains unknown, and we are not in a position to

suggest the length limitations for the reconstruction of RLN

defects based on the present study because the number of pa-

tients was relatively limited, preventing statistical signifi-

cance. Further efforts will be made to explore the length

limits of nerve grafts in bridging gaps between the ansa cervi-

calis and RLN.

CONCLUSIONS

Free nerve grafting is an effective procedure to reinnervate la-

ryngeal muscles in bridging gaps between the RLN and ansa

cervicalis in patients with unilateral RLN injury, as well asa safe procedure without obvious morbidity, and a satisfactory

vocal outcome can be obtained with this technique.

REFERENCES1. Green DC, Ward PH. The management of the divided recurrent laryngeal

nerve. Laryngoscope. 1990;100:779–782.

2. Crumley RL. Update: ansa cervicalis to recurrent laryngeal nerve anasto-

mosis for unilateral laryngeal paralysis. Laryngoscope. 1991;101:

384–387. discussion 388.

3. Lee WT, Milstein C, Hicks D, Akst LM, Esclamado RM. Results of ansa torecurrent laryngeal nerve reinnervation. Otolaryngol Head Neck Surg.

2007;136:450–454.

4. Paniello RC. Laryngeal reinnervation with the hypoglossal nerve: II. Clin-

ical evaluation and early patient experience. Laryngoscope. 2000;110

(5 Pt 1):739–748.

5. Crumley RL. Repair of the recurrent laryngeal nerve. Otolaryngol Clin

North Am. 1990;23:553–563.

6. Miyauchi A, Yokozawa T, Kobayashi K, Hirai K, Matsuzuka F, Kuma K.

Opposite ansa cervicalis to recurrent laryngeal nerve anastomosis to restore

phonation in patients with advanced thyroid cancer. Eur J Surg. 2001;167:

540–541.

7. Olson DE, Goding GS, Michael DD. Acoustic and perceptual evaluation of

laryngeal reinnervation by ansa cervicalis transfer. Laryngoscope. 1998;

108:1767–1772.

8. Smith ME, Roy N, Stoddard K. Ansa-RLN reinnervation for unilateral

vocal fold paralysis in adolescents and young adults. Int J Pediatr Otorhi-

nolaryngol. 2008;72:1311–1316.

9. Lorenz RR, Esclamado RM, Teker AM, et al. Ansa cervicalis-to-

recurrent laryngeal nerve anastomosis for unilateral vocal fold paralysis:

experience of a single institution. Ann Otol Rhinol Laryngol. 2008;117:

40–45.

10. Zheng H, Li Z, Zhou S, Cuan Y, Wen W. Update: laryngeal reinnervation

for unilateral vocal cord paralysis with the ansa cervicalis. Laryngoscope.

1996;106:1522–1527.

11. Wang W, Chen D, Chen S, Li D, Li M, Xia S, Zheng H. Laryngeal reinner-

vation using ansa cervicalis for thyroid surgery-related unilateral vocal fold

paralysis: a long-term outcome analysis of 237 cases. PLoS One. 2011;6:

e19128.

12. WangW, Chen S, Chen D, XiaS, QiuX, LiuY, ZhengH. Contralateral ansa

cervicalis-to-recurrentlaryngealnerve anastomosis for unilateralvocal fold

paralysis: a long-term outcome analysis of 56 cases. Laryngoscope. 2011;

121:1027–1234.

13. Crumley RL. Laryngeal synkinesis revisited. Ann Otol Rhinol Laryngol.

2000;109:365–371.

14. Yumoto E, Sanuki T, Kumai Y. Immediate recurrent laryngeal nerve

reconstruction and vocal outcome. Laryngoscope. 2006;116:1657–1661.15. Miyauchi A, Inoue H, Tomoda C, et al. Improvement in phonation after re-

construction of the recurrent laryngeal nerve in patients withthyroid cancer

invading the nerve. Surgery. 2009;146:1056–1062.

16. Pabari A, Yang SY, Seifalian AM, Mosahebi A. Modern surgical manage-

ment of peripheral nerve gap. J Plast Reconstr Aesthet Surg. 2010;63:

1941–1948.

17. Sunderland IR, Brenner MJ, Singham J, Rickman SR, Hunter DA,

Mackinnon SE. Effect of tension on nerve regeneration in rat sciatic nerve

transection model. Ann Plast Surg. 2004;53:382–387.

18. Bunnell S. Surgery of the nerves of the hand. Surg Gynecol Obstet . 1927;

44:145–152.

19. Millesi H, MeisslG, BergerA. Theinterfascicular nerve-graftingof theme-

dian and ulnar nerves. J Bone Joint Surg Am. 1972;54:727–750.

20. Berger A, Millesi H. Nerve grafting. Clin Orthop Relat Res. 1978;133:

49–55.21. DriscollPJ, GlasbyMA, LawsonGM. An invivo study of peripheral nerves

in continuity: biomechanical and physiological responses to elongation.

J Orthop Res. 2002;20:370–375.

22. Chou FF, Su CY, Jeng SF, Hsu KL, Lu KY. Neurorrhaphy of the recurrent

laryngeal nerve. J Am Coll Surg. 2003;197:52–57.

23. Randolph GW. Surgical anatomy of the recurrent laryngeal nerve. In:

Randolph GW, ed. Surgery of the Thyroid and Parathyroid Glands. Phila-

delphia: Saunders; 2003:300–342.

24. Siribodhi C, Sundmaker W, Atkins JP, Bonner FJ. Electromyographic stud-

ies of laryngeal paralysisand regeneration of laryngealmotor nerve in dogs.

Laryngoscope. 1963;73:148–164.


Documents

Bridging Gaps Between the Recurrent Laryngeal