Role of preoperative ultrasonography in
the surgical management of patients
with thyroid cancer
Maria A. Kouvaraki, MD, PhD, Suzanne E. Shapiro, MS, Bruno D. Fornage, MD,
Beth S. Edeiken-Monro, MD, Steven I. Sherman, MD, Rena Vassilopoulou-Sellin, MD,
Jeffrey E. Lee, MD, and Douglas B. Evans, MD, Houston, Tex
Background. Cervical recurrence occurs in up to 30% of patients with differentiated thyroid carcinoma.
We retrospectively compared preoperative transcutaneous ultrasonography and physical examination
(PE) results in the detection of local-regional metastases (lymph node and soft tissue) in patients with
thyroid cancer.
Methods. Data were collected retrospectively from the medical records of patients with thyroid carcinoma
who underwent preoperative ultrasonography. Patients were divided into 3 groups: group 1, those
undergoing primary thyroid/neck surgery; group 2, those undergoing reoperation for persistent disease;
and group 3, those undergoing reoperation for recurrent thyroid carcinoma. For each group, we recorded
the frequencies with which ultrasonography detected disease in a neck compartment (central or lateral)
that was normal on PE.
Results. Two hundred twelve patients underwent operation for primary, persistent, or recurrent papillary
(n = 130), medullary (n = 61), or follicular/Hu
¨rthle cell (n = 21) carcinoma. Ultrasonography
detected additional sites of metastatic disease not appreciated on PE in 21 (20%) of 107 group
1 patients, 9 (32%) of 28 group 2 patients, and 52 (68%) of 77 group 3 patients. The surgical
procedure performed was altered by the information obtained from preoperative ultrasonography in 82
(39%) of the 212 patients. Of the 107 group 1 patients, cervical recurrence has been detected in only 6
(6%) at a median follow-up of 36 months, in spite of 67 (63%) having tumors larger than 2 cm or
lymph node metastases.
Conclusions. Preoperative high-quality ultrasonography detected lymph node or soft-tissue metastases in
neck compartments believed to be uninvolved by PE in 39% of patients. Ultrasound findings altered the
operative procedure in these patients, facilitating complete resection of disease and potentially
minimizing local-regional recurrence. (Surgery 2003;134:946-55.)
From the Departments of Surgical Oncology, Diagnostic Radiology, and Endocrine Neoplasia and Hormonal
Disorders, The University of Texas M.D. Anderson Cancer Center, Houston, Tex
THYROID CANCER HAS A UNIQUE and largely unex-
with 22,000 new cases expected in the United States
plained tumor biology, characterized by early
in 2003; however, thyroid cancer-related deaths
spread to regional lymph nodes and occasional ex-
will number only 1400 (6%) in 2003.2 The modest
trathyroidal soft tissue extension, but a low in-
death rate and the visible nature of surgical compli-
cidence of distant metastases and infrequent
cations such as injury to the recurrent laryngeal
death.1-3 For example, the incidence of thyroid
nerve and hypoparathyroidism may result in a con-
cancer has been increasing over the past 3 decades,
servative surgical approach to local-regional disease.
To what degree this contributes to the high rates of
cervical recurrence is not known. However, cervical
Presented at the 24th Annual Meeting of the American
recurrences, mostly regional lymph node metasta-
Association of Endocrine Surgeons, San Diego, California,
ses, have been reported in up to 31% of patients with
May 11-14, 2003.
differentiated thyroid carcinoma (DTC)4-6 and up
Reprint requests: Douglas B. Evans, MD, Department of
to 65% of patients with medullary thyroid carci-
Surgical Oncology, Unit 444, The University of Texas M. D.
Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX
noma (MTC).7 Such high rates of cervical recur-
77030.
rence suggest that many patients have macroscopic
Ó 2003, Mosby, Inc. All rights reserved.
lymph node metastases at the time of initial surgery
0039-6060/2003/$30.00 + 0
and, if detected and removed, may prevent sub-
doi:10.1016/S0039-6060(03)00424-0
sequent reoperation for recurrence in the neck.
946 SURGERY

---

Surgery
Kouvaraki et al 947
Volume 134, Number 6
The realization that preoperative physical exam-
(benign), positive (malignant), indeterminate, or
ination (PE) is inadequate for the detection of
suspicious.
extrathyroidal cervical metastases has caused many
Ultrasound scanning of the soft tissues of the
clinicians to consider transcutaneous ultrasonogra-
neck was performed with a high-resolution ultra-
phy for preoperative staging and follow-up in
sound scanner (Sequoia, Acuson, Mountain View,
patients with thyroid cancer.8-12 High-quality ultra-
California; Elegra, Siemens, Issaquah, Washington;
sonography is an established diagnostic tool for
HDI 5000, Philips-ATL, Bothell, Washington;
detecting local-regional metastatic disease as small
or Powervision 7000, Toshiba, Tokyo, Japan)
as 2 to 3 mm in patients with thyroid cancer.9-11
equipped with a high-frequency linear-array trans-
Although there is little support for prophylactic
ducer of at least 7 MHz and up to 13 MHz. The
neck dissection to remove microscopic disease in
ultrasound examination included evaluation of the
patients with DTC, compartment-oriented surgery
lymph nodes in the lateral neck compartments and
(COS) for macroscopic disease detected by ultra-
of the thyroid (when present) and associated soft
sonography may minimize future neck recurrence
tissues in the central neck compartment. The
and patient morbidity.13 The purpose of this study
location of cervical lymph nodes on ultrasonogra-
was to examine the sensitivity of preoperative
phy was considered central (level VI) or lateral
ultrasonography in detecting local-regional metas-
(levels IIA, III, IV, or V).15 Small volume disease was
tases and the impact of preoperative ultrasonogra-
localized in the operating room with transcutane-
phy on surgical management in patients with DTC
ous ultrasonography (before the skin incision was
and MTC.
made) at the discretion of the operating surgeon. If
additional disease was detected at the time of
intraoperative ultrasonography, it was added to the
ultrasound findings for all analyses. Some patients
METHODS
with positive ultrasound findings underwent ultra-
We retrospectively reviewed the medical records
sound-guided fine-needle aspiration (FNA), which
of all patients with DTC (papillary thyroid carci-
was performed with a 20-gauge needle inserted
noma [PTC] or follicular/Hu¨rthle cell carcinoma)
obliquely along the scan plane. The needle tip was
or MTC who underwent preoperative cervical
under constant ultrasound observation during the
ultrasonography followed by cervical operation in
entire biopsy procedure.
the Department of Surgical Oncology at The Uni-
A lymph node was considered benign on
versity of Texas M.D. Anderson Cancer Center
ultrasound evaluation when the appearance was
from 1991 to 2003. Study patients were divided
oval and flattened, with a smooth cortex, and the
into 3 groups on the basis of whether their oper-
central fatty hilum paralleled the smooth cortex. A
ation was for primary, persistent, or recurrent dis-
lymph node was considered malignant when it was
ease: group 1 patients underwent thyroidectomy
rounded or had an absence or truncation of the
with or without neck dissection as their primary
central fatty hilum. The spectrum of indetermi-
surgical procedure; group 2 patients underwent
nate to suspicious was subjective and based on the
reoperation for persistent disease, which was
individual ultrasonographers’ assessments of the
defined as occurring within 6 months of a pre-
lymph nodes including characteristics such as
referral thyroid cancer operation; and group 3
the size and position of the central fatty hilum, the
patients underwent reoperation for recurrent
presence or absence of eccentric cortical widening,
disease, which was defined as occurring greater
rounded versus elongated shape, markedly de-
than 6 months following a prereferral thyroid
creased echogenicity, alteration in intranodal
cancer operation.
vascularity, and calcifications.16 Lymph node size
Initial pathologic TNM stage (American Joint
was not a diagnostic criterion for metastasis. Lymph
Committee on Cancer staging system, 6th edi-
nodes that were suspicious for metastasis were
tion14) was determined by the histopathologic
considered positive for purposes of analysis. Lymph
results from all operations performed within 6
nodes that were indeterminate for metastasis were
months of the patient’s initial operation.14 If lymph
considered negative except in one patient in whom
node status was not mentioned on the histopath-
FNA biopsy of an indeterminate lymph node
ologic report, N status was considered N0. The
confirmed cancer; therefore this ultrasound scan-
presence or absence of lymph node metastasis on
ning result was considered positive.
PE was recorded as negative or positive. The
Our surgical management of patients with MTC
presence or absence of lymph node metastasis
has been previously published.15 Our surgical
on ultrasonography was recorded as negative
approach to patients with DTC and suspected or

---

948 Kouvaraki et al
Surgery
December 2003
proven central or lateral compartment lymph node
(version 5.01) statistical software package (Abacus
metastases also emphasized COS; ‘‘node plucking’’
Concepts Inc.; Berkeley, CA). Differences were
was not performed except in patients with recurrent
considered statistically significant when the P value
disease in a previously dissected compartment.
was < .05.
When performed, central compartment dissection
involved removal of all lymph nodes and soft tissues
in level VI usually with the overlying sternothyroid
muscle. In patients with DTC (in contrast to MTC)
RESULTS
the level VI dissection was largely limited to the
From 1991 to 2003, 216 consecutive patients
ipsilateral tracheoesophageal groove, unless pre-
underwent preoperative ultrasonography and sur-
operative ultrasonography suggested metastatic
gery for primary, persistent, or recurrent thyroid
disease in the contralateral paratracheal region.
carcinoma. Four patients had incidentally discov-
Lateral compartment dissection (modified radical
ered papillary carcinomas (index nodule was
neck dissection) involved removal of all lymph
benign) and were excluded from further analysis,
nodes and soft tissues in levels IIA, III, IV, and V with
leaving a study population of 212 patients. There
preservation of the jugular vein, carotid artery,
were 67 male patients (32%) and 145 female
vagus nerve, phrenic nerve, and spinal accessory
patients (68%); the median age at diagnosis was
nerve when possible. In patients with papillary or
43 years (range, 5 to 86 years). Histopathologic
Hu¨rthle cell carcinoma and no evidence of regional
diagnoses included 130 PTC, 61 MTC, and 21
lymph node metastases, we routinely removed the
follicular/Hu¨rthle cell carcinomas. Three patients
ipsilateral paratracheal lymph nodes; however, the
who underwent reoperative completion thyroidec-
lateral compartment was not dissected in the
tomy for PTC in the absence of other sites of
absence of clinical or ultrasound evidence of lymph
cervical disease were included in group 2. Twenty-
node metastases. In contrast to selected patients
two (36%) of 61 MTC patients had hereditary
with MTC, we did not consider elective reoperation
forms of MTC (multiple endocrine neoplasia type 2
in patients with DTC for an elevation in thyroglob-
[MEN 2] or familial MTC). Five of these patients
ulin in the absence of clinical or ultrasound
underwent prophylactic surgery for MEN 2 or
evidence of lymph node metastases.
familial MTC and had histologic evidence of
Postoperative 5-mCi radioiodine scanning was
invasive MTC.
performed after surgery in patients with DTC in
Pathologic tumor stages by group and histopath-
preparation for adjuvant radioiodine therapy. The
ologic tumor type are shown in Table I. Patients
percent uptake of the ingested dose of radioiodine
with MTC presented with more advanced disease
was used to measure the adequacy of surgical
than did those with DTC (P < .0001). The primary
resection.
tumor was unilateral in 152 patients, bilateral in 55,
Cervical recurrence was defined as metastatic
and isthmic in 5.
involvement of any cervical lymph node or soft
Ultrasound examination detected lymph node
tissue and was considered local when the central
metastases or soft tissue recurrence in 113 (53%) of
compartment was involved and regional when
212 patients. Sixty-nine (61%) of the 113 patients
either lateral compartment was involved. Cervical
with positive findings on ultrasonography under-
recurrences were considered local-regional when
went FNA biopsy. Cytologic results confirmed the
both the central and lateral compartments were
presence of metastatic or recurrent disease in 64
involved.
(93%) of 69 patients. Ultrasound-positive disease
Disease status at last follow-up was categorized as
not detected on PE was found in 105 neck com-
no evidence of disease (NED; no evidence of
partments of 82 patients (39%): 52 (34%) of 151
disease by physical examination and radiographic
with DTC and 30 (49%) of 61 with MTC. These 82
imaging), alive with disease (AWD; biopsy confir-
patients included 21 patients (20%) in group 1, 9
mation of distant metastases was not routinely
(32%) in group 2, and 52 (68%) in group 3. The
performed), dead of disease (DOD), or dead of
difference between groups (PE missed more group
other causes.
3 cases) was significant (P < .0001). The location of
The statistical correlations between categorical
ultrasound-positive disease not appreciated on PE
variables were assessed by v2 and Fisher’s exact
for each patient group and histopathologic type
tests. Associations of continuous variables with
are shown in Table II. Ultrasound-positive disease
different groups were evaluated using the non-
not appreciated on PE was found in the central
parametric Mann-Whitney or Kruscal-Wallis tests.
neck compartment in 58 (27%) of the 212 patients,
All analyses were performed using the Stat View
in the ipsilateral neck compartment in 32 (15%),

---

Surgery
Kouvaraki et al 949
Volume 134, Number 6
and in the contralateral neck compartment in
Table I. Pathologic tumor staging in 212 patients
14 (7%).
with thyroid carcinoma*
One hundred eighty-three of 212 patients had
No. patients
histopathologic analysis of regional lymph nodes
Tumor type
or soft tissues. Twenty-nine patients did not
and stage
Group 1
Group 2
Group 3
Total (%)
undergo compartment dissection because of neg-
DTC
ative findings on both PE and ultrasonography.
All patients
85
18
48
151
Histopathologic study confirmed local-regional
Stage I
55
13
31
99 (66)
metastatic disease in 150 (82%) of 183 patients:
Stage II
9
0
2
11 (7)
in 50 (60%) of the 83 patients from group 1 and in
Stage III
13
2
2
17 (11)
all 24 (100%) and 76 (100%) patients in group 2
Stage IVA
3
2
6
11 (7)
and group 3, respectively. The relationship be-
Stage IVC
4
1
2
7 (5)
tween preoperative ultrasound findings and histo-
Stage not
1y
0
5z
6 (4)
assessed
pathologic findings by compartment is shown in
MTC
Table III. The sensitivity and specificity of ultraso-
All patients
22
10
29
61
nography within each compartment was as follows:
Stage I
10
0
4
14 (23)
central compartment 52% and 95%, respectively;
Stage II
0
0
3
3 (5)
ipsilateral compartment 77% and 93%, respec-
Stage III
1
1
6
8 (13)
tively; and the contralateral compartment 79% and
Stage IVA
7
9
15
31 (51)
96%, respectively. We interpreted ultrasound re-
Stage IVC
4
0
1
5 (8)
sults as false-positive on the basis of final histo-
*AJCC Cancer Staging Manual (6th ed).14
pathologic findings in 4 compartments of 4 patients.
DTC, Differentiated thyroid carcinoma; MTC, medullary thyroid
One patient had sporadic MTC and underwent
carcinoma.
yStage could not be assessed from the incomplete information on
reoperation for recurrent disease. Bulky disease was
tumor size from the initial operation at M.D. Anderson.
present in the central and ipsilateral neck com-
zStage could not be assessed from the incomplete reoperation at M.D.
partments. In the contralateral compartment, an 8
Anderson.
mm lymph node posterior to the internal jugular
vein was interpreted as suspicious because of the
Histopathologic study demonstrated metastatic
absence of a fatty hilum; histopathologic study
disease in 1 or more neck compartments of 74
demonstrated no evidence of metastatic disease in
patients (35%) with no evidence of metastatic
21 lymph nodes within this compartment. Another
disease by ultrasonography (false-negative ultra-
patient underwent thyroidectomy along with cen-
sonography result): 47 patients with DTC (all had
tral and bilateral neck dissection for biopsy-proven
PTC) and 27 with MTC (Table IV). The rate of false-
familial MTC (codon 804 RET mutation). Preoper-
negative results is influenced by the extent
ative ultrasonography demonstrated an abnormal-
of surgery; a false-negative result is not possible if
appearing 1.6-cm lymph node containing fine
lymph nodes are not removed and pathologically
calcifications, interpreted as metastatic lymphade-
analyzed. In group 1 patients with PTC or Hu¨rthle
nopathy; histopathologic study demonstrated no
cell carcinoma, it was standard practice to remove
evidence of metastatic disease in 60 cervical lymph
the ipsilateral paratracheal lymph nodes (central
nodes. A third patient had sporadic MTC and
compartment, level VI) at the time of thyroidec-
underwent reoperation for persistent disease within
tomy regardless of the findings on preoperative
both lateral neck compartments. There were also
ultrasonography. Among patients with DTC, false-
2 lymph nodes within the central neck compart-
negative ultrasound results in the lateral com-
ment that were 1 cm in diameter and interpreted as
partments occurred in patients with indeterminate
suspicious for metastatic disease on ultrasonogra-
ultrasound findings in the setting of advanced
phy; histopathologic study of the central neck
disease in one or more other neck compartments.
dissection specimen demonstrated no evidence of
Surgeon preference in these cases was to extend the
recurrent MTC. The fourth patient had sporadic
dissection to the compartment with indeterminate
MTC and underwent reoperation for persistent
findings. In group 1 patients with invasive MTC, we
disease. A 9-mm nodule in the central neck
performed central neck dissection in all cases,
compartment was interpreted as suspicious for
dissection of the ipsilateral lateral compartment in
metastatic disease on ultrasonography; histopatho-
cases of sporadic MTC, and frequently, bilateral
logic study demonstrated suture material and
neck dissection in cases with familial MTC.15
foreign body giant cell reaction with no evidence
The surgical procedure was altered by the
of MTC.
information obtained from preoperative ultra-

---

950 Kouvaraki et al
Surgery
December 2003
Table II. Recurrent or metastatic disease detected by ultrasonography but not physical examination
in 212 patients with thyroid carcinoma
No. Patients (%)
Group 1
Group 2
Group 3
Total
DTC
All patients
85
18
48
151
Patients with positive US and negative PE
15 (18)
2 (11)
35 (73)
52 (34)
Cervical compartments with positive US
and negative PE*
Central
6 (7)
1 (6)
26 (54)
33 (22)
Ipsilateral
4 (5)
1 (6)
16 (33)
21 (14)
Contralateral
8 (9)
0
3 (6)
11 (7)
MTC
All patients
22
10
29
61
Patients with positive US and negative PE
6 (27)
7 (70)
17 (59)
30 (49)
Cervical compartments with positive US
and negative PE*
Central
4 (18)
7 (70)
14 (48)
25 (41)
Ipsiltateral
2 (9)
2 (20)
7 (24)
11 (18)
Contralateral
0
3 (30)
1 (3)
4 (7)
DTC, Differentiated thyroid carcinoma; MTC, medullary thyroid carcinoma; PE, physical examination; US, ultrasonography.
*If the index tumor was bilateral, then lymph node and soft tissue in either lateral compartment was considered ipsilateral. Some patients had
metastatic-appearing lymph nodes in more than one neck compartment.
sonography in 82 (39%) of 212 patients. All
The median follow-up was 36 months (range, 1
ultrasound-positive disease was surgically excised
to 140 months). Cervical recurrence following
except in 1 patient, in whom neck dissection was
primary or reoperative surgery at M. D. Anderson
limited to the one compartment with large-volume
Cancer Center developed in 16 (8%) of 207 pa-
disease because of the presence of extensive
tients with follow up: 6 (6%) in group 1, 3 (12%) in
extracervical metastases. Of the 151 patients with
group 2, and 7 (9%) in group 3 (Table VI). Of note,
DTC, unsuspected disease was found by ultraso-
67 (65%) of 103 group 1 patients had primary
nography in 52 (34%) and altered the operation to
tumors larger than 2 cm in size or lymph node
include a more extensive dissection of the central
metastases. Cervical recurrences were local in 8
compartment in 32, the ipsilateral compartment in
patients (50%), regional in 5 (31%), and local-
21, and the contralateral compartment in 9
regional in 3 (19%). Cervical recurrences involved
patients. In group 1 patients with DTC, preopera-
lymph nodes in 11 patients, soft tissue in 1, and both
tive ultrasonography changed the primary thyroid
lymph nodes and soft tissue in 4. The median size of
cancer operation to include 6 central compartment
recurrent lesions was 1.0 cm (range, 0.5 to 2.4 cm).
lymphadenectomies and 4 ipsilateral and 8 con-
Thyroglobulin levels were reviewed for the 124
tralateral functional neck dissections. Of the 61
patients with DTC who were NED at last follow-up
patients with MTC, unsuspected disease was found
and were available for 114. Of these 114 patients,
by ultrasonography in 26 patients (43%) (exclud-
anti-thyroglobulin antibodies were present in 27
ing the 4 false-positive results) and alerted the
and not measured in 6. Of the 81 patients without
surgeon to the need for a detailed dissection of the
detectable antibodies, the most recently available
central neck compartment in 22 patients and for
thyroglobulin level was undetectable in 62 (76.5%).
inclusion of the ipsilateral neck compartment in 8
The median thyroglobulin level in the remaining
and the contralateral neck compartment in 2.
19 patients with detectable thyroglobulin levels was
Results of postoperative radioiodine scans in
3.0 with a median TSH level of 0.68 mcU/mL.
patients with DTC are summarized in Table V. The
number of patients with cervical uptake of 1% or
less of the total ingested dose is provided as
DISCUSSION
a surrogate marker of the adequacy of surgical
This study reports the impact of preoperative
therapy. In group 1 patients, 77% had cervical
ultrasonography on the operative management of
uptake of 1% or less of the ingested dose.
a consecutive series of 212 patients with thyroid

---

Surgery
Kouvaraki et al 951
Volume 134, Number 6
Table III. Ultrasonography versus histopathologic
Table IV. False-negative ultrasound results by
findings in all dissected cervical compartments
compartment for patients with DTC and MTC
No. patients by
No. patients
Site of false-negative
ultrasound findings
Histopathologic
ultrasound findings
Group I Groups II and III Total
findings by compartment
Negative
Positive
Total
DTC
Central cervical
All patients
85
66
151
compartment
Patients with
31
16
47
Negative
41
2y
43
false-negative USyz
Positive
57z
61
118
Central
31
12
43
Total
98
63
161
Ipsilateral*
1
4
5
Ipsilateral cervical
Contralateral
0
3
3
compartment*
MTC
Negative
13
1y
14
All patients
22
39
61
Positive
22z
75
97
Patients with
8
19
27
Total
35
76
111
false-negative USyz
Contralateral cervical
Central
7
7
14
compartment
Ipsilateral*
3
14
17
Negative
27
1y
28
Contralateral
0
2
2
Positive
5z
19
24
DTC, Differentiated thyroid carcinoma; MTC, medullary thyroid
Total
32
20
52
carcinoma.
*If the index tumor was bilateral, then lymph node or soft tissue
*If the index tumor was bilateral, then lymph node or soft tissue
disease in either lateral compartment was considered ipsilateral.
disease in either lateral compartment was considered ipsilateral.
ySome patients had metastatic-appearing lymph nodes in more than
yFalse-positive US results.
one neck compartment.
zFalse-negative US results.
zFalse-negative US: lymph nodes within a compartment appeared
negative on preoperative US but were found to be positive for
metastasis on histopathologic study.
cancer. Unsuspected soft tissue or lymph node
metastases were found by ultrasonography in 52
(34%) of 151 patients with DTC and in 26 (43%) of
logically within a specific compartment was the
61 patients with MTC and resulted in the perfor-
exact lymph node or soft tissue metastasis seen on
mance of more extensive surgery. These patients
preoperative ultrasonography. However, our expe-
were treated with a COS approach using standard
rience with ultrasound-guided FNA biopsy indicates
techniques of neck dissection in an effort to pre-
the high specificity of ultrasonography; cytologic
vent subsequent recurrence. At a median follow-
study results of ultrasound-guided FNA confirmed
up of 36 months, cervical recurrences developed
the presence of carcinoma in 64 (93%) of 69
in only 16 (8%) of 207 evaluable patients. An
patients. Our current practice is to limit the use of
additional measure of the adequacy of surgical
FNA biopsy to patients with suspected recurrent
resection, in patients with DTC, were the results
thyroid cancer, especially those with recurrent
of the postoperative radioiodine scans.17,18 In
disease in a previously dissected compartment. We
evaluable group 1 patients with DTC, 77% had
do not perform reoperative cervical surgery in
cervical uptake of 1% or less of the ingested dose,
a previously dissected neck compartment without
excluding those 7 patients with no uptake. This
a cytologic diagnosis of cancer. For patients un-
compares favorably with the 46% of patients
dergoing ultrasonography before their first thyroid
reported to have cervical uptake of 2% or less by
cancer operation (group 1), we do not believe it is
Hudgson et al.19
necessary to confirm cytologically a positive or sus-
As suggested by previous investigators, ultraso-
picious ultrasound finding in the central or lateral
nography was both sensitive and specific for the diag-
neck compartments. However, we are more liberal in
nosis of soft tissue or lymph node metastases.20-24
the application of ultrasound-guided FNA in patients
The specificities of ultrasonography in the central,
with indeterminate ultrasound findings and those
ipsilateral, and contralateral compartments were
without a definitive cytologic diagnosis of cancer.
95%, 93%, and 96%, respectively. All 4 of the
Pathologic evaluation of the surgical specimen
patients with false-positive ultrasound results had
demonstrated metastatic disease in one or more
MTC, and 3 of them underwent ultrasonography
neck compartments of 74 patients (35%) with no
before reoperation for persistent or recurrent
evidence of metastatic disease by ultrasonography.
MTC. Because we performed COS, we cannot
The sensitivities of ultrasonography in the central,
verify that the metastatic disease found histo-
ipsilateral, and contralateral compartments were

---

952 Kouvaraki et al
Surgery
December 2003
Table V. Results of postoperative radioiodine scans in patients with DTC*
No. patients
Group 1
Group 2
Group 3
Total
Patients with DTC
85
17
49
151
Patients with DTC who had WBS
77
14
28
119
No uptake
7
0
15
22
Cervical uptake only
63
13
12
88
Extracervical uptake only
1
0
0
1
Cervical & extracervical uptake
6
1
1
8
46/60
7/13
4/7
57/80
All patients with cervical uptake # 1%y
(77%)
(54%)
(57%)
(71%)
WBS, Whole body scan; DTC, differentiated thyroid carcinoma.
*Postoperative scans were performed with 5 mCi of 131I before 131I ablation treatment.
yNot all patients had specific uptake values reported.
Table VI. Recurrence and outcome in patients with thyroid carcinoma
No. patients
Outcome
Group 1
Group 2
Group 3
Total
DTC
All patients
84
16
49
149
Median follow-up (range)*
40 mo (1-132)
47 mo (7-110)
29 mo (1-105)
36 mo (1-132)
Patients with follow-up < 1 yr
20
1
12
33
Patients with cervical recurrence
5 (6%)
0
4 (8%)
9 (6%)
Median time to first cervical recurrence (range)
22 mo (11-49)
NA
21 mo (14-29)
22 mo (11-49)
Patients Alive, NED
71
14
39
124
Patients AWD
7
1
6
14
Patients DOD
3
1
2
6
Patients Dead of Other Causes
3
0
2
5
MTC
All patients
19
10
29
58
Median follow-up (range)*
35 mo (3-140)
58 mo (24-82)
32 mo (1-95)
40 mo (1-140)
Patients with Follow-up < 1 yr
4
0
6
10
Patients with cervical recurrence
1 (5%)
3 (30%)
3 (10%)
7 (12%)
Median time to first cervical recurrence (range)
125 mo
22 mo (20-26)
25 mo (9-34)
25 mo (9-125)
Patients Alive, NED
14
8
19
41
Patients AWD
3
1
9
13
Patients DOD
2
1
1
4
AWD, Alive with disease; DOD, dead of disease; NA, not applicable; DTC, differentiated thyroid carcinoma; MTC, medullary thyroid carcinoma;
NED, no existing disease.
*Time zero = Date of first MDACC operation. Five patients who were immediately lost to follow-up after surgery were excluded from analysis.
52%, 77%, and 79%, respectively. Most false-
sults in the lateral compartments occurred in
negative ultrasound results occurred in the central
patients with indeterminate ultrasound results,
neck compartment, and more than half were in
often in the setting of more advanced disease in
patients whose thyroid gland was still in situ (group
another compartment of the neck.
1). With the thyroid gland in situ, ultrasonography
The impact of the extent of thyroidectomy and
cannot assess the central compartment with the
regional lymph node dissection on the survival of
sensitivity possible after the thyroid is removed.
patients with DTC and MTC remains a subject
Therefore we have adopted the practice of re-
of controversy. Because of the indolent nature of
moving the ipsilateral paratracheal lymph nodes in
these malignancies, the extent of surgery will likely
patients with PTC or Hu
¨ rthle cell carcinoma.
never be the subject of a prospective, phase III,
Central compartment dissection is already an
randomized trial. Although more extensive cervical
accepted surgical procedure in any patient with
surgery may have only a limited impact on survival
invasive MTC. Most false-negative ultrasound re-
duration, it may significantly impact the quality of

---

Surgery
Kouvaraki et al 953
Volume 134, Number 6
survival by minimizing cervical recurrence.13 The
6. Simon D, Goretzki PE, Witte J, Roher HD. Incidence of
known high rates of cervical recurrence in patients
regional recurrence guiding radicality in differentiated
with DTC and MTC have alerted physicians,
thyroid carcinoma. World J Surg 1996;20:860-6; discussion 6.
7. Kallinowski F, Buhr HJ, Meybier H, Eberhardt M, Herfarth
especially endocrinologists, to the importance
C. Medullary carcinoma of the thyroid—therapeutic strat-
of long-term follow-up. The common use of serum
egy derived from fifteen years of experience. Surgery 1993;
marker (thyroglobulin and calcitonin) measure-
114:491-6.
ments, radioiodine scanning (with thyrogen stimu-
8. Sherman SI, Ball DW, Beenken SW, Byrd D, Clark OH,
lation), and ultrasonography in follow-up will likely
Daniels GH, et al. NCCN Oncology Practice Guidelines
result in the diagnosis of even more recurrences.
v.1.2003. National Comprehensive Cancer Network. Avail-
able at: http://www.nccn.org/physician_gls/f_guidelines.
Therefore, if ultrasonography is to be performed in
html. Accessed 4-28-2003
the postoperative follow-up period, it seems only
9. Simeone JF, Daniels GH, Hall DA, McCarthy K, Kopans DB,
logical to perform ultrasonography before the first
Butch RJ, et al. Sonography in the follow-up of 100 patients
surgical procedure in an effort to remove all gross
with thyroid carcinoma. AJR Am J Roentgenol 1987;148:
metastatic disease, which is frequently not palpable.
45-9.
This strategy may minimize patient confusion and
10. Antonelli A, Miccoli P, Ferdeghini M, Di Coscio G, Alberti B,
Iacconi P, et al. Role of neck ultrasonography in the follow-
physician indecision during the postoperative
up of patients operated on for thyroid cancer. Thyroid
follow-up period, prevent the indiscriminate and,
1995;5:25-8.
sometimes, unnecessary use of radioiodine therapy,
11. Franceschi M, Kusic Z, Franceschi D, Lukinac L, Roncevic S.
and minimize cervical recurrence with its negative
Thyroglobulin determination, neck ultrasonography and
impact on the physical and emotional state of the
iodine-131 whole-body scintigraphy in differentiated thy-
patient.13,25 This manuscript provides preliminary
roid carcinoma. J Nucl Med 1996;37:446-51.
12. Sherman SI. Thyroid carcinoma. Lancet 2003;361:501-11.
evidence to support such a strategy.
13. Esnaola NF, Cantor SB, Sherman SI, Lee JE, Evans DB.
Further follow-up will be necessary to determine
Optimal treatment strategy in patients with papillary
the true impact of high-quality preoperative ultra-
thyroid cancer: a decision analysis. Surgery 2001;130:921-30.
sonography and standardized COS on rates of
14. Greene FL, Page DL, Fleming ID, Fritz A, Balch CM, Haller
cervical recurrence in patients with DTC and MTC,
DG, et al. AJCC cancer staging manual. 6th ed. New York:
but our preliminary findings appear promising.
Springer-Verlag; 2002.
15. Fleming JB, Lee JE, Bouvet M, Schultz PN, Sherman SI,
Although one can argue that patients with invasive
Sellin RV, et al. Surgical strategy for the treatment of
MTC should undergo a routine dissection of at least
medullary thyroid carcinoma. Ann Surg 1999;230:697-707.
level VI, and therefore preoperative ultrasound
16. Ahuja AT, Chow L, Chick W, King W, Metreweli C.
findings may not alter the planned surgical pro-
Metastatic cervical nodes in papillary carcinoma of the
cedure, the extent and location of regional metas-
thyroid: ultrasound and histological correlation. Clin
tases will alert the surgeon to the need for a more
Radiol 1995;50:229-31.
17. Cholewinski SP, Yoo KS, Klieger PS, O’Mara RE. Absence of
detailed dissection, and those surgeons less experi-
thyroid stunning after diagnostic whole-body scanning with
enced with neck dissection may wish to refer such
185 MBq 131I. J Nucl Med 2000;41:1198-202.
patients to a center more experienced with the opera-
18. Fratkin MJ, Newsome HH Jr, Sharpe AR Jr, Tatum JL.
tive management of locally advanced thyroid cancer.
Cervical distribution of iodine 131 following total thyroid-
ectomy for thyroid cancer. Arch Surg 1983;118:864-7.
19. Hodgson DC, Brierley JD, Tsang RW, Panzarella T. Pre-
REFERENCES
scribing 131Iodine based on neck uptake produces effective
thyroid ablation and reduced hospital stay. Radiother Oncol
1. Tubiana M, Schlumberger M, Rougier P, Laplanche A,
1998;47:325-30.
Benhamou E, Gardet P, et al. Long-term results and
20. Baatenburg de Jong RJ, Rongen RJ, Lameris JS, Harthoorn
prognostic factors in patients with differentiated thyroid
M, Verwoerd CD, Knegt P. Metastatic neck disease. Palpa-
carcinoma. Cancer 1985;55:794-804.
tion vs ultrasound examination. Arch Otolaryngol Head
2. Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun
Neck Surg 1989;115:689-90.
MJ. Cancer statistics, 2003. CA Cancer J Clin 2003;53:5-26.
3. Samaan NA, Maheshwari YK, Nader S, Hill CS Jr, Schultz
21. Bruneton JN, Roux P, Caramella E, Demard F, Vallicioni J,
PN, Haynie TP, et al. Impact of therapy for differentiated
Chauvel P. Ear, nose, and throat cancer: ultrasound
carcinoma of the thyroid: an analysis of 706 cases. J Clin
diagnosis of metastasis to cervical lymph nodes. Radiology
Endocrinol Metab 1983;56:1131-8.
1984;152:771-3.
4. Mazzaferri EL, Kloos RT. Clinical review 128: Current
22. Gorges R, Eising EG, Fotescu D, Renzing-Kohler K, Frilling
approaches to primary therapy for papillary and follicular
A, Schmid KW, et al. Diagnostic value of high-resolution
thyroid cancer. J Clin Endocrinol Metab 2001;86:1447-63.
B-mode and power-mode sonography in the follow-up of
5. Hay ID, Thompson GB, Grant CS, Bergstralh EJ, Dvorak CE,
thyroid cancer. Eur J Ultrasound 2003;16:191-206.
Gorman CA, et al. Papillary thyroid carcinoma managed at
23. Na DG, Lim HK, Byun HS, Kim HD, Ko YH, Baek JH.
the Mayo Clinic during six decades (1940-1999): temporal
Differential diagnosis of cervical lymphadenopathy: useful-
trends in initial therapy and long-term outcome in 2444
ness of color Doppler sonography. AJR Am J Roentgenol
consecutively treated patients. World J Surg 2002;26:879-85.
1997;168:1311-6.

---

954 Kouvaraki et al
Surgery
December 2003
24. Frasoldati A, Pesenti M, Gallo M, Caroggio A, Salvo D,
Since 1985 at Mayo, we not only regularly use
Valcavi R. Diagnosis of neck recurrences in patients with
a preoperative ultrasonography map to aid in determin-
differentiated thyroid carcinoma. Cancer 2003;97:90-6.
ing the extent of thyroid cancer operation but also as the
25. Hay ID, McConahey WM, Goellner JR. Managing patients
most cost-effective imaging method for delineating the
with papillary thyroid carcinoma: insights gained from the
presence after operation of persistent or recurrent neck
Mayo Clinic’s experience of treating 2,512 consecutive
disease. It is our hope and expectation that neck
patients during 1940 through 2000. Trans Am Clin Climatol
reexploration numbers for recurrence will be mini-
Assoc 2002;113:241-60.
mized by regular use of this powerful imaging
modality. Particularly this weekend when, for the first
time in my experience, we had a concomitant ultrasound
DISCUSSION
workshop for endocrine surgeons in North America, I
Dr Keith S. Heller (New Hyde Park, NY). Very nice
think it would be important for many centers to follow the
presentation. I have trouble with your conclusions. Many
lead of M.D. Anderson, Mayo, and others in using this
years ago a former vice president of this Society, Joseph
exciting modality to improve the lot of our patients; and
Attie, performed hundreds of elective neck dissections
second, to reduce recurrence rates in the future.
for well-differentiated thyroid cancer. As you have
Dr Irving B. Rosen (Toronto, Ontario, Canada). I
demonstrated, he found the incidence of occult positive
concur with the notion of ultrasonography and ultra-
nodes to be about 40%. Yet, he and virtually everybody
sound needle biopsy as being well established, and I have
else abandoned that approach because nobody was able
done a node sampling approach for a lengthy period of
to show any benefit from operating on these people with
time. Did these modalities really influence the extent or
clinically insignificant metastatic nodes. Do you have any
the nature of the central neck dissection that you carried
data whatsoever to suggest that the patients benefit from
out? In the contralateral positive nodes in group I, did
your aggressive surgical approach toward nonpalpable,
you have multicentric or bilateral disease?
occult disease?
Dr Kouvaraki. In all patients with invasive MTC we
Dr Kouvaraki. We do not perform elective neck
routinely perform central neck dissection, and therefore
dissection in patients with well-differentiated carcinoma
preoperative ultrasound findings may not alter the
of the thyroid. In contrast, we do perform therapeutic
planned surgical procedure. However, the extent and
dissection when gross disease is seen on preoperative
location of regional metastases may alert the surgeon to
ultrasonography. When cervical lymph node metastases
the need for a more detailed dissection, especially those
are large enough to be detected by ultrasonography, they
surgeons who are less experienced in neck dissection. In
are not considered microscopic. Thus the incidence of
addition, the extent of cervical disease, if known before
positive lymph nodes we report in this study corresponds
operation, may cause some surgeons to consider the
to gross disease.
referral of such patients to a center more experienced
Dr Allan Siperstein (Cleveland, Ohio). I very much
with the operative management of locally advanced
enjoyed this study, especially as an advocate of doing
thyroid cancer. Second, cervical metastatic disease was
preoperative ultrasound examination in all of these
considered as ipsilateral in patients with bilateral or
patients where we found sizable disease that was non-
multicentric primaries in the thyroid.
palpable. In terms of comparing palpation with ultraso-
Dr Scott Wilhelm (Chicago, Ill). In terms of the crite-
nography, when was the ultrasound scanning performed,
ria used for determining your abnormal lymph nodes,
before or after the physical examination, or was the
the ultrasound scan you showed revealed an 8-mm,
surgeon aware of the ultrasound findings before the
examination? Second, who performed the ultrasound
smooth-appearing node. I know you said that one node
scan? Was it surgically performed, as is my preference, or
in particular came back as metastatic. What criteria does
is it done in the radiology department?
your group use to define a ‘‘malignant-appearing’’ node?
Dr Kouvaraki. In all cases the physical examination
Your FNA results of the nodes were impressive. You
was done before the ultrasound scan, and thus the
had 64 of 69 patients with a 93% positive cytologic study
surgeon was not aware of the ultrasound results. All neck
result. Are you doing hematoxylin & eosin staining and
ultrasound scans at M.D. Anderson are performed in the
immunohistochemistry on the FNAs?
Radiology Department by radiologists specializing in
Last, you had 4 false-positive ultrasound scan results in
cervical ultrasonography.
your MTCs. Did those patients have FNAs or was it false-
Dr Ian D. Hay (Rochester, Minn). I stand before this
positive on the basis of your diagnostic criteria for what
august audience to comment on this study as a non-
you found was an abnormal node prompting your
aggressive nonsurgical clinician who sees about 400
dissection?
patients with differentiated thyroid cancer annually.
Dr Kouvaraki. Lymph node assessment on ultraso-
Rather like the well-documented role of ultrasonography
nography was somewhat subjective. However, a lymph
in permitting the differentiation between solitary and
node was considered malignant when it was rounded or
multinodular goiters, the preoperative use of high-
had an absence or truncation of the central fatty hilum.
resolution ultrasonography in differentiated thyroid
The spectrum of indeterminate to suspicious was sub-
cancer has left a simple neck palpator a humbled and
jective and based on characteristics such as the size and
more honest individual.
position of the central fatty hilum, the presence or

---

Document Outline

• Role of preoperative ultrasonography in the surgical management of patients with thyroid cancer
  • METHODS
  • RESULTS
  • DISCUSSION
  • REFERENCES

---