Cl
R aly
p ev
o
Jo
aD
So
Na
of 14 showing voice changes. RLNP was detected in 9 of 291 (3.1%) patients without documented
thy
the
cau
bra
cri
inc
ran
000
doi
The American Journal of Surgery (2009) 197, 459–465
nerve injury; 8 recovered. Nine of 15 patients with RLN section had poor vocal function, which
improved in 8 patients after medialization of the unilateral vocal fold.
CONCLUSIONS: Patients with PTC may have vocal dysfunction from cancer or surgery-related
RLNP. Vocal evaluation and management may help improve their vocal function, thus enhancing their
quality of life.
© 2009 Published by Elsevier Inc.
Voice alteration is a morbidity associated with invasive
roid malignancies and thyroid surgery. Surgical injury to
recurrent laryngeal nerve (RLN) is regarded as the main
se of voice changes,1–3 but surgical injury to the external
nch of the superior laryngeal nerve (EBSLN) or the
cothyroid muscle also may cause voice changes.4,5 The
idence of laryngeal nerve paralysis has been reported to
ge from 0% to 30%.4,6 Laryngeal nerve paralysis has
important implications, especially for patients in voice-re-
lated professions, and is associated with most reports of
complications of thyroid surgery.7 Voice changes after thy-
roidectomy, however, may not be related to impaired func-
tion of the RLN or EBSLN.8–11 By using objective voice
measurements, the incidence of voice changes has been found
to be as high as 84%.10 Although most patients without doc-
umented nerve damage recover from the RLN paralysis
ng-Lyel Roh, M.D., Ph.D.a,*, Yeo-Hoon Yoon, M.D.b, Chan Il Park, M.D., Ph.D.b
epartment of Otolaryngology, Asan Medical Centre, University of Ulsan, College of Medicine, 388–1, Pungnap-dong,
ngpa-gu, Seoul 138-736, the Republic of Korea; bDepartment of Otolaryngology-Head and Neck Surgery, Chungnam
tional University, College of Medicine, Daejeon, the Republic of Korea
Abstract
BACKGROUND: Recurrent laryngeal nerve paralysis (RLNP) occurs in patients with thyroid ma-
lignancy. This study prospectively evaluated vocal function and management outcomes of patients with
papillary thyroid carcinomas (PTCs) and RLNP.
METHODS: Of 319 PTC patients, 256 underwent total thyroidectomy with or without neck dissec-
tion, 42 underwent lobectomy, and 21 underwent reoperation for recurrent cancers. All patients
underwent laryngoscopy and vocal function measurements before and after surgery. Patients with
RLNP and poor vocal function underwent voice surgery.
RESULTS: Temporary and permanent RLNP rates were 2.8% and .9% at nerve-at-risk–based
analysis, respectively. Of 28 patients with tumor invasion of RLN, 14 had preoperative RLNP, with 9
KEYWORDS:
Papillary thyroid
carcinoma;
Thyroid surgery;
Recurrent laryngeal
nerve;
Paralysis;
Voice management
inical Surgery-International
ecurrent laryngeal nerve par
apillary thyroid carcinomas:
f resulting vocal dysfunction
(R
RL
wa
me
* Corresponding author: Tel.: �82-2-3010-3965; fax: �82-2-489-2773.
E-mail address: rohjl@amc.ac.kr
Manuscript received April 18, 2008; revised manuscript April 22, 2008
2-9610/$ - see front matter © 2009 Published by Elsevier Inc.
:10.1016/j.amjsurg.2008.04.017
sis in patients with
aluation and management
LNP) that occurs after thyroid surgery,6,11 patients with
N sections may have impaired vocal, swallowing, and air-
y functions, suggesting that they require further manage-
nt.
tat
ner
hav
age
pro
com
car
Pa
Pa
fer
tia
inv
PT
ph
sur
20
an
sch
the
tio
mo
Ins
inf
Su
ind
the
sur
inc
mi
hu
art
cap
tio
av
dis
co
car
tie
wa
lou
pre
roi
wa
wi
fin
thy
the
tie
cra
lob
rot
ifie
wi
pa
inf
pe
RN
As
glo
pa
aft
we
ba
tes
rap
an
on
glo
Pa
RL
ary
lar
bil
glo
Vo
dex
sca
exa
nai
and
is
tot
vo
we
ph
lar
rec
(K
wi
Th
tim
an
wi
ob
eo
sco
460 The American Journal of Surgery, Vol 197, No 4, April 2009
Although the incidence, causes, natural history, and preven-
ive measures associated with voice changes and laryngeal
ve paralysis have been examined extensively,2–6,8–11 there
e been few reports on the systematic evaluation and man-
ment of RLNP in thyroid cancer patients. We therefore
spectively evaluated vocal function and management out-
es of patients with well-differentiated papillary thyroid
cinomas (PTCs) and cancer- or surgery-related RLNP.
tients and Methods
tients and study design
This was a prospective study of patients with well-dif-
entiated PTC scheduled for thyroid surgery and at poten-
l risk for RLNP because of the surgery or tumor RLN
asion. All patients were diagnosed preoperatively with
C after fine-needle aspiration cytology and ultrasonogra-
y and underwent thyroidectomy or reoperative thyroid
gery of the central neck compartment between March
03 and June 2006. Patients with poorly differentiated,
aplastic, or other thyroid malignancies, and patients not
eduled for thyroid surgery on the central compartment of
neck, were excluded. All patients underwent vocal func-
n and laryngoscopic examination before and for at least 6
nths after thyroidectomy. The study was approved by the
titutional Review Board of our institution, and written
ormed consent was obtained from each patient.
rgical procedure
Patients underwent lobectomy or total thyroidectomy, as
icated by their primary pathology, under general anes-
sia. Patients with recurrent tumors underwent reoperative
gery of the central neck. After a lower transverse cervical
ision, the strap muscles were retracted laterally from the
dline or sometimes divided for adequate exposure of
ge or infiltrative thyroid lesions. The superior thyroid
ery and vein were ligated individually on the thyroid
sule to avoid injury to the EBSLN. Routine identifica-
n of the EBSLN was not performed during surgery to
oid inadvertent nerve injury. The cricothyroid muscle was
sected carefully from the thyroid gland without electro-
agulation or manual retraction. The RLN was identified
efully and dissected unilaterally or bilaterally in all pa-
nts. When tumors directly invaded the RLN, the nerve
s meticulously shaved off the tumors using surgical
pes under magnification. When the RLN was paralyzed
operatively or infiltrated extensively by malignant thy-
d tumors and not easily shaved off the tumors, the nerve
s sacrificed.
Neck dissection was performed in selected patients. Patients
th macroscopically involved or ultrasonography-guided
e-needle aspiration cytology–positive nodes underwent total
roidectomy and therapeutic neck dissection. Prophylactic or ser
rapeutic central neck dissection was performed in 179 pa-
nts. Node clearance of the central neck was performed
nially by both superior thyroid arteries and the pyramidal
e, caudally by the innominate vein, laterally by the ca-
id sheaths, and dorsally by the prevertebral fascia. Mod-
d radical neck dissection was performed on 70 patients
th clinically positive lateral neck nodes. The lateral com-
rtment was delimited superiorly by the hypoglossal nerve,
eriorly by the subclavian vein, and laterally by the tra-
zius muscle. Intraoperative neurologic monitoring of the
L was not performed.
sessment and management of RLNP
Vocal quality, along with vocal fold mobility and the
ttal gap between both vocal folds, was assessed in each
tient before, 1 week after, and 1, 3, 6, and 12 months
er, thyroid surgery. If any abnormalities on laryngoscopy
re found, objective vocal function was assessed using a
ttery of acoustic, aerodynamic, and videostroboscopic
ts. Patients with RLNP also underwent computed tomog-
hy scanning to determine the association between RLNP
d thyroid tumors. The decision to perform voice surgery
each patient was based on each patient’s vocal function,
ttal gap, and the recovery potential of the injured RLN.
tients with poor voice quality and a wide glottal gap after
N section were indicated for voice surgery, consisting of
tenoid adduction, medialization thyroplasty, or injection
yngoplasty. Patients with a narrow glottal space and
ateral RLNP were indicated for tracheostomy or posterior
ttic cordotomy to widen the glottic airway.
cal function assessment
Voice quality was assessed using the vocal handicap in-
,12 the grade, rough, breathy, asthenic, strained (GRBAS)
le,13 acoustic and aerodynamic methods, and stroboscopic
mination. The vocal handicap index is a 30-item question-
re consisting of 3 subscales measuring functional, physical,
emotional handicaps in verbal communication. Each item
scored from 0 to 4, each subscale from 0 to 40, and the
al score is from 0 to 120; higher scores indicate greater
cal disability. Acoustic parameters and aerodynamic data
re analyzed by Computerized Speech Lab and Aero-
one II (Kay Elemetrics, Corp., Lincoln Park, NJ). The
ynxes of the subjects were examined, and images were
orded using rigid and rhino-larynx videostroboscopy
ay Elemetrics, Corp.). Scores were between 0 and 10,
th higher scores indicating higher levels of abnormality.
e GRBAS scale measurements, maximum phonation
e, and laryngoscopy assessments were performed before
d after thyroid surgery, as described previously. Patients
th voice or laryngeal abnormalities underwent further
jective assessments with acoustic, aerodynamic, and vid-
stroboscopic tests. The GRBAS scale and videostrobo-
pic results were assessed blindly by 2 independent ob-
vers.
St
use
use
tes
va
pa
co
Re
Pa
ex
aft
lar
rat
Pa
sho
22
dia
eso
(11
28
un
ser
wa
RL
pa
pa
tie
inv
pe
pa
tiv
inc
tha
54
dif
thy
we
sec
sha
Th
RL
inj
8 r
me
pa
Vo
tio
an
wi
inv
the
the
bil
pa
do
Fig
pap
cus
pre
pre
RL
and
sid
thy
com
iza
late
T
V
S
M
P
S
M
p
461J.-L. Roh et al. Papillary thyroid carcinoma and vocal dysfunction
atistical analysis
SPSS 12.0 for Windows (SPSS, Inc., Chicago, IL) was
d for statistical analysis. The Mann–Whitney U test was
d to compare continuous variables and the Fisher exact
t was used to investigate differences between categoric
riables. The Wilcoxon signed-rank test was used to com-
re intragroup results at different times. Differences were
nsidered significant when the P value was less than .05.
sults
tients, pathology, and treatment characteristics
A total of 327 patients were enrolled. Eight patients were
cluded because they were lost to follow-up evaluation
er surgery without proper assessment of their voices and
ynxes, leaving 319 eligible patients. The male:female
io was 1:4.4, and the median patient age was 46 years.
tient demographic, clinical, and pathologic data are
wn in Table 1. The median tumor size was 1.5 cm and
6 patients (70.8%) had tumors larger than 1.0 cm in
meter. Local tumor invasion of the RLN, larynx, trachea,
phagus, or great vessels was observed in 35 patients
.0%), with 28 (8.8%) showing RLN invasion. Of these
patients, 15 underwent unilateral RLN section and 13
derwent a RLN shaving procedure. EBSLN was not ob-
ved in 277 patients (86.8%), and their strap musculature
s preserved. The median MACIS score was 5.1.
NP and voice surgery
A total of 29 patients had postoperative RLNP, with 1
tient having bilateral RLNP (Fig. 1). Fourteen of these 29
tients had unilateral RLNP preoperatively. Fifteen pa-
able 1 Patient demographics and clinical characteristics
ariable
No. of
patients % of total
ex, male/female 59/260 18.5/81.5
edian age (range), y 46 (16–85)
rimary tumor
Median tumor size (range), cm 1.5 (.2–5.5)
Local invasion (T4 tumor) 35 11.0
RLN invasion 28 8.8
urgery
Reoperative surgery 21 6.6
Lobectomy 42 13.2
Total thyroidectomy 256 80.2
Central neck dissection 179 56.1
Lateral neck dissection 70 21.9
edian MACIS score (range) 5.1 (3.1–11.1)
The demographic, clinical, and pathologic data of the 319 study
atients are shown.
tio
nts underwent unilateral RLN section because of tumor
asion. After excluding these 15 patients, temporary and
rmanent RLN paralysis occurred in 4.6% and 1.3% of
tients, respectively, and in 2.8% and .9% of cases, respec-
ely, using nerve-at-risk–based analysis (Table 2). The
idence of RLNP was higher after reoperative surgery
n after initial thyroid surgery (4 of 31 [12.9%] vs 12 of
2 [2.2%]; P � .008). The incidence of RLNP did not
fer between patients who underwent lobectomy or total
roidectomy (P � .999) or between patients who under-
nt total thyroidectomy with or without central neck dis-
tion (P � .744). Of the 13 patients who underwent RLN
ving procedures, 5 had postoperative unilateral RLNP.
ree of these patients showed complete recovery of the
NP within 6 months (Fig. 2). Of the 291 patients without
ury to the RLN, 9 had temporary unilateral RLNP, with
ecovering within 3 months after thyroidectomy.
Eight patients with unilateral RLN injuries underwent
dialization procedures of the unilateral vocal fold, 5
tients at initial surgery and 3 patients 2 to 10 months later.
ice surgery was not randomized in patients with inten-
nal RLN section; only patients with poor vocal quality
d wide glottal gaps underwent voice surgery. One patient
th preoperative unilateral RLNP showed bilateral RLN
asion of the PTC; this patient underwent sectioning of
unilateral RLN and shaving of the contralateral RLN off
tumors. Laryngoscopy in the recovery room showed
ateral RLNP requiring urgent airway management. This
tient, who immediately underwent posterior glottic cor-
tomy, showed no further symptoms of airway obstruc-
ure 1 Flow chart showing the study cohort of patients with
illary thyroid carcinomas who underwent thyroid surgery, fo-
ing on their vocal function and RLNP. All patients underwent
operative laryngoscopic examination. Fourteen patients had
operative RLNP because of tumor invasion. Postoperative
NP was found in 15 of 305 patients without preoperative RLNP
in 1 of 14 patients with preoperative RLNP on the contralateral
e. RLNP recovered in 11 of 16 patients within 6 months after
roid surgery. AA � arytenoids adduction; B/L � bilateral; CT �
puted tomography; IL � injection laryngoplasty; MT � medial-
tion thyroplasty; PGC � posterior glottic cordotomy; U/L � uni-
ral.
n.
Re
RL
vo
mo
7 u
un
RL
rel
Th
sur
tiv
14
(38
an
da
for
sur
de
sur
som
thy
gro
Co
op
pe
rat
pa
ne
an
rel
T reoper
s
T
S
R
% and .
w .008).
f tumor
Fig
out
RL
inv
sur
RL
462 The American Journal of Surgery, Vol 197, No 4, April 2009
covery of vocal function in patients with RLNP
Of 15 patients with intentional sections of the unilateral
N, 9 had poor vocal function preoperatively, 5 had good
ice quality, and 1 had normal vocal function and vocal cord
bility (Fig. 1). Of the 9 patients with poor vocal function,
nderwent medialization of the unilateral vocal cord, 1
derwent posterior glottic cordectomy because of bilateral
NP, and 1 refused voice surgery. The 5 patients with
atively good voice quality did not undergo voice surgery.
e 8 patients with poor voice quality who underwent voice
gery showed improvement in their subjective and objec-
e vocal functions (Table 3). Despite the preserved RLN,
patients had postoperative unilateral RLNP: 5 of the 13
.5%) patients who underwent RLN shaving procedures
d 9 of the 291 (3.1%) patients without documented nerve
mage. Recovery of the RLNP was observed in 2 of the
able 2 Incidence of RLNP according to surgical procedure and
No. of patients
No. of nerve
at risk
otal 319 573
urgery
Lobectomy 42 42
Total thyroidectomy 256 500
No CND 77 154
CND 109 215
CND � LND 70 131
eoperation† 21 31
The incidence of postoperative temporary and permanent RLNP was 2.8
as higher after reoperative surgery than after initial thyroid surgery (P �
CND � central neck dissection; LND � lateral neck dissection.
*Calculated after excluding patients with preoperative RLNP because o
†All patients had recurrent papillary thyroid carcinomas.
ure 2 Recovery of postoperative RLNP in 14 patients with-
preoperative RLNP. These patients experienced postoperative
NP after a procedure in which the RLN was shaved off the
asive papillary thyroid carcinoma (n � 5) or after thyroid
gery with no injury to the RLN (n � 9). �, RLN shaved; e,
weN not injured.
mer and 8 of the latter within 6 months after thyroid
gery (Fig. 2). Subjective and objective vocal functions
creased in both groups during the first few weeks after
gery, although the patients without RLN damage showed
ewhat greater improvement (Table 4). Six months after
roidectomy, however, vocal function was similar in the 2
ups (P � .1).
mments
We have shown here that, of PTC patients without pre-
erative RLNP, 4.6% and 1.3% experience temporary and
rmanent RLNP, respectively, after thyroid surgery. These
es were similar to those reported previously.2,3,6 If the 15
tients with intentional nerve section because of tumor
rve invasion were included, the overall rates of temporary
d permanent RLNP increased to 9.1% and 5.6%. The
atively high RLNP rate in our patients may have been
cause of the relatively high proportion of patients with
asive PTC (11.0%) involving the upper aerodigestive
ct and the RLN, and the relatively high rate of reoperative
roid surgery (6.6%). Surgeries for thyroid cancer have
en associated with high RLNP rates.3,6 This study pro-
ctively evaluated vocal function and RLNP in a single
up of patients with PTC.
In agreement with earlier studies, we found that periop-
tive laryngoscopic examination was essential for the de-
tion of RLNP.14,15 RLNP was detected preoperatively in
of 28 (50%) patients and voice change was detected in
ly 9 of 14 (64.3%). In comparison, preoperative RLNP
s present in 70% of patients with invasive thyroid dis-
e; of these, two-thirds presented with normal voice and
ee-fourths had no suspicion of RLNP on preoperative
mputed tomography.14 These findings indicated that
ptomatic voice assessment and radiographic evaluation
ation
Preoperative
RLNP, n
Postoperative RLNP*
Temporary, % Permanent, %
14 16 (2.8) 5 (.9)
0 1 (2.4) 0
12 11 (2.2) 2 (.4)
0 4 (2.6) 0
3 5 (2.3) 2 (.9)
9 2 (1.5) 0
2 4 (12.9) 3 (9.7)
9% of cases, respectively, using nerve-at-risk–based analysis, which
invasion.
be
inv
tra
thy
be
spe
gro
era
tec
14
on
wa
eas
thr
co
sym
re insufficient to detect preoperative RLNP and sug-
ge
tie
co
pe
thu
era
co
lar
op
the
pa
RL
thy
wa
rec
Ou
29
co
(88
we
(38
rec
T tients
V †
V
P
A
A
V
erwent
v
fu
463J.-L. Roh et al. Papillary thyroid carcinoma and vocal dysfunction
sted the need for laryngoscopic examination of all pa-
nts undergoing thyroid surgery.14 In addition, laryngos-
py and voice examination during the early postoperative
riod provided accurate information on complications,
s guiding further management. One patient with postop-
tive bilateral RLNP underwent prompt posterior glottic
rdotomy because of acute airway obstruction detected by
yngoscopy in the recovery room. Therefore, routine peri-
erative laryngoscopy and voice examination can assist in
proper management of these patients, improving im-
ired vocal or airway function.
Postoperative laryngoscopy also can detect unrecognized
NP in patients who do not suffer nerve damage during
roid surgery. In a study by Lo et al,6 unrecognized RLNP
s detected in 28 of 500 (5.6%) patients, with complete
overy of vocal cord function documented in 26 (92.9%).
r findings were similar in that we observed RLNP in 9 of
1 (3.1%) patients without documented RLN damage, with
mplete recovery of vocal cord function documented in 8
.9%). We also observed recovery in patients who under-
nt RLN shaving procedures; of these 13 patients, 5
.5%) experienced postoperative unilateral RLNP, with
able 3 Preoperative and postoperative vocal function in 13 pa
ariable
Voice surgery (n � 7)*
Preoperative† Postoperative
oice handicap index
Functional 11.2 (5.4) 3.1 (2.4)
Physical 12.9 (5.6) 3.4 (3.4)
Emotional 14.2 (6.2) 4.6 (4.2)
Total 38.3 (16.6) 11.1 (9.8)
erceptual§
Overall grade 4.3 (3.4) 1.1 (1.1)
Roughness 4.1 (3.2) 1.3 (1.2)
Breathiness 4.7 (3.0) 1.3 (1.1)
Strain 3.6 (2.8) 1.5 (1.1)
coustic
F0, Hz 178.6 (59.4) 172.4 (54.2)
Jitter, % 3.1 (2.1) .9 (.