Endoscopic Ultrasonography, Fine-Needle Aspiration Biopsy Guided by Endoscopic Ultrasonography, and Computed Tomography in the Preoperative Staging of Non-Small-Cell Lung Cancer: A Comparison Study

Endoscopic Ultrasonography, Fine-Needle Aspiration Biopsy
Guided by Endoscopic Ultrasonography, and Computed
Tomography in the Preoperative Staging of Non-Small-Cell
Lung Cancer: A Comparison Study
Frank G. Gress, MD; Thomas J. Savides, MD; Alan Sandler, MD; Kenneth Kesler, MD;
Dewey Conces, MD; Oscar Cummings, MD; Praveen Mathur, MD; Steven Ikenberry, MD;
Sandy Bilderback, RN; and Robert Hawes, MD
Background: Current methods for detecting mediastinal
This paper is also available at http://www.acponline.org.
lymph node involvement with non-small-cell lung cancer
can be inaccurate and are often invasive and expensive.
Ann Intern Med. 1997;127:604-612.
Objective: To assess the utility of endoscopic ultrasonog-
From Indiana University School of Medicine, Indianapolis, Indi-
ana. For current author addresses, see end of text.
raphy, fine-needle aspiration biopsy guided by endoscopic
ultrasonography, and computed tomography for the de-
A ccurate staging of non-small-cell lung cancer
tection of metastases to the posterior mediastinal lymph
x V p l a y s a crucial role in determining the treat-
nodes in non-small-cell lung cancer.
ment options available to patients with this disease.
Design: Prospective preoperative evaluation of the diag-
The preoperative documentation of metastasis to
nostic operating characteristics of these procedures.
the mediastinal lymph nodes has therapeutic impli-
cations that have traditionally focused on palliation
Setting: Referral-based academic medical center.
but more recently have included neoadjuvant strat-
Patients: 130 consecutive patients with non-small-cell
egies (1, 2).
lung cancer who were otherwise good surgical candidates.
Metastasis to the mediastinal lymph nodes occurs
Interventions: All patients had initial computed tomog-
in nearly half of all patients with non-small-cell
raphy of the chest; those with enlarged nodes were re-
lung cancer. The development of mediastinal metas-
ferred for endoscopic ultrasonography. Endoscopic ultra-
tasis is probably the most frequent deterrent to
sonography-guided fine-needle aspiration biopsy was
cure, even when the presentation is localized. If
done on suspicious contralateral posterior mediastinal or
metastasis involves contralateral or large, bulky
subcarinal lymph nodes identified by ultrasonography. At
(>1.5 to 2.0 cm) subcarinal lymph nodes, surgery
surgery, lymph nodes were dissected and categorized by
alone may not be curative (3-9). As a result of
location and underwent histopathologic evaluation.
recent revisions to the staging systems for lung can-
Results: 52 patients were ultimately enrolled in the
cer, ipsilateral mediastinal and subcarinal lymph
study: Thirty-one had thoracotomy with mediastinal dis-
node involvement is now classified as potentially
section, and 21 had tumors considered unresectable on the
resectable, N2 disease; contralateral mediastinal
basis of preoperative evaluation. Ultrasonography with-
lymph node involvement (N3 disease) precludes re-
out aspiration biopsy had an overall accuracy of 84% for
section (10-12).
predicting metastasis to lymph nodes; computed tomogra-
Computed tomography of the chest is the current
phy had an accuracy of 49% (P < 0.025). Twenty-four
method by which mediastinal lymphadenopathy is
patients had ultrasonography-guided aspiration biopsy;
detected in non-small-cell lung cancer. However, its
14 of the 24 were ineligible for surgery because cytology
showed malignancy. Results of surgical pathology corre-
sensitivity for detection of metastasis to the lymph
lated with negative aspiration cytology results in 9 of 10
nodes varies; accuracy in previous studies has
patients; the one node with false-negative results con-
ranged from 52% to 88% (13-23). This inconsis-
tained a 2-mm focus of cancer. The accuracy of ultrasonog-
tency has been attributed to the variable correlation
raphy-guided aspiration biopsy in diagnosing metastasis
of lymph node size with the presence of malignancy.
to lymph nodes was 96%; the results of this test prompted
When enlarged contralateral or ipsilateral mediasti-
a change in management in 95% of the patients who had
nal lymph nodes are seen on computed tomography
the procedure.
of the chest, standard practice is to determine more
Conclusions: Endoscopic ultrasonography alone or with
accurate staging by performing aspiration biopsy of
fine-needle aspiration biopsy adds useful diagnostic infor-
the lymph node with computed tomographic guid-
mation in determining metastasis to posterior mediastinal
or subcarinal lymph nodes in patients with non-small-cell
lung cancer. These procedures are especially helpful in the
preoperative evaluation of patients with suspicious con-
See editorial comment on pp 643-645.
tralateral mediastinal or "bulky" subcarinal nodes.
604 © 1997 American College of Physicians

---

ance; bronchoscopy; or, less commonly, a transtho-
in 60% of patients, and the spiral technique was
racic approach. If these procedures are unsuccessful,
used in 40%. Computed tomography was done at
open biopsy is performed by using mediastinoscopy
the referring hospital or Indiana University Medical
or limited thoracotomy (24-26). If contralateral
Center; the scans were read at Indiana University
lymph nodes are positive for malignancy, surgical
Medical Center by a senior attending radiologist
resection of the primary tumor is contraindicated.
who has recognized expertise in this area and used
The development of endoscopic ultrasonography
currently accepted radiographic techniques to stage
has now made it possible to visualize, with high
the tumor. The radiologist's determination of the
resolution, not only the gastrointestinal tract but
benign or malignant nature of each lymph node was
also surrounding structures. Endoscopic ultrasonog-
recorded on a preoperative computed tomography
raphy has been shown to be superior to computed
lymph node map; the American Thoracic Society
tomography in evaluating lymph nodes for metasta-
mediastinal staging map (Figure 1) was used to
ses in esophageal, gastric, and pancreatic cancer
describe the location of each node (12). Any patient
(27-29). Promising results for detecting posterior
who had a questionably enlarged mediastinal lymph
mediastinal lymph nodes in patients with lung can-
node (>1 cm in diameter) and was considered a
cer suggest a possible role for endoscopic ultra-
surgical candidate was then scheduled for endo-
sonography in staging lymph nodes in patients with
scopic ultrasonography.
non-small-cell lung cancer (30-36). Fine-needle as-
piration biopsy guided by endoscopic ultrasonogra-
Endoscopic Ultrasonography
phy was recently reported to further improve the
Endoscopic ultrasonography was performed in an
accuracy of endoscopic ultrasonography in predict-
outpatient setting on all patients by one of three
ing malignancy of gastrointestinal masses, with rates
experienced endosonographers; the radial scanning
as high as 87% to 91% (37-42).
echoendoscope (GFUM-20, Olympus America, Mel-
We previously reported the results of endoscopic
ultrasonography in 17 patients with lung cancer.
This method was very accurate for detecting medi-
astinal lymphadenopathy; the overall accuracy was
71% compared with 41% for computed tomography
(P = 0.032) (43). During the initial study, however,
fine-needle aspiration biopsy guided by endoscopic
ultrasonography was not available. The goal of the
present study was to prospectively evaluate the ac-
curacy of endoscopic ultrasonography alone, endo-
scopic ultrasonography-guided fine-needle aspira-
tion biopsy, and computed tomography of the chest
in detecting mediastinal lymph node metastasis in
patients with non-small-cell lung cancer.
Methods
Patient Selection
The study sample consisted of all patients pre-
senting to the Indiana University Thoracic Oncology
Program between July 1993 and June 1995 with a
diagnosis of non-small-cell lung cancer. The study
was approved by the institutional review board, and
all enrolled patients gave informed consent. Patients
were excluded if they had documented unresectable
disease (that is, distant metastasis or locally ad-
vanced staged disease [stage III b]) as shown on
computed tomography of the chest or if they had a
serious medical illness and a life expectancy of less
Figure 1 . American Thoracic Society scheme for mapping medias-
tinal lymphadenopathy by anatomic location, as seen from behind
than 1 year. All patients underwent initial preoper-
with endoscopic ultrasonography. Ao = aorta; D = diaphragm; E =
ative intravenous contrast-enhanced computed to-
esophagus; inf VC = inferior vena cava; L = left; I.PA = left pulmonary
artery; PV = pulmonary vein; R = right; r.PA = right pulmonary artery; sup
mography of the chest; the axial technique was used
VC = superior vena cava. Numbers refer to specific mediastinal zones.
15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)
605

---

phy were directly referred for surgery because there
was no indication for aspiration biopsy.
Endoscopic Ultrasonography-Guided Fine-Needle
Aspiration Biopsy
Endoscopic ultrasonography-guided aspiration bi-
opsy became available after the first 17 patients
were enrolled in our pilot study. All posterior me-
diastinal lymph nodes that were suspicious for ma-
lignant involvement according to the endoscopic ul-
trasonographic criteria were noted; selected nodes
underwent biopsy during the same procedure. Many
of the patients had more than one suspicious lymph
node. We performed biopsy only on the most sus-
picious lymph node, which would have the greatest
Figure 2. A mediastinal lymph node as imaged with the linear-
effect on clinical staging (that is, determination of
array endoscopic ultrasonography system. The needle is exiting the
whether the metastasis was contralateral or subcari-
scope; the tip of the needle is in the center of the lymph node {arrow). Just
inferior to the lymph node is the pulmonary artery. One advantage of the
nal). This technique for ultrasonography-guided as-
linear-array instrument is its Doppler capability, which allows precise imag-
piration biopsy was initially developed for use with
ing of regional blood vessels and safe positioning of the needle relative to
the target lymph node.
the linear-array instrument (Figure 2) and is de-
scribed elsewhere (40-42). We recently reported a
similar technique that uses a radial scanning echo-
ville, New York) or the linear-array scanning echo-
endoscope (Figure 3) (37). Ultrasonography-guided
endoscope (FG32UA, Pentax, Orangeburg, New
aspiration biopsy involves the insertion of an aspi-
York) was used for all procedures. When done by
ration catheter needle device through the accessory
an experienced operator, endoscopic ultrasonogra-
channel port of the echoendoscope; the needle is
phy is similar to standard upper endoscopy both in
then deployed into the lymph node to be sampled
technique and duration of the procedure. When
under endoscopic ultrasonographic guidance. Aspi-
fine-needle aspiration biopsy is performed, the pro-
ration biopsy is done by introducing a specially de-
cedure is slightly prolonged. Patients were sedated
signed fine-needle aspiration catheter system that
with meperidine and midazolam, the doses of which
consists of a 4-cm long, 23-gauge needle attached to
were titrated to achieve adequate conscious seda-
a 180-cm long, 5-French aspiration catheter (Wil-
tion. The instrument was advanced into the stom-
son-Cook, Winston-Salem, North Carolina); "in and
out" movements of the catheter are used while the
ach, and the celiac axis was imaged. The probe was
then withdrawn to the gastroesophageal junction
and slowly withdrawn at 1-cm intervals. Images were
obtained with 7.5- and 12-MHz frequencies at each
interval. All imaged mediastinal lymph nodes were
mapped by location according to the American Tho-
racic Society classification scheme (12). From these
data, an objective assessment was made as to
whether the mediastinal lymphadenopathy detected
by endoscopic ultrasonography was benign or possi-
bly malignant according to the following previously
reported criteria for malignancy: round shape;
sharp, distinct borders; hypoechoic texture; and a
short-axis diameter greater than 5 mm (36-39). Ma-
lignancy was suspected if all of these criteria were
present. All patients who were studied before the
availability of endoscopic ultrasonography-guided
fine-needle aspiration biopsy underwent surgical re-
section, and endoscopic ultrasonographic findings
were correlated to surgical pathologic findings. Pa-
tients who were studied after the advent of fine-
Figure 3. Endoscopic ultrasonographic image obtained from the
needle aspiration biopsy and were found to have no
radial scanning instrument showing a large hypoechoic, oval sub-
carinal lymph node (LAO suspicious for metastatic involvement. The
suspicious lymph nodes by endoscopic ultrasonogra-
lymph node is adjacent to the aorta (Ao).
606 15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)

---

operator firmly grasps the catheter at the point at
biopsy were then determined by using a multistage
which it enters the accessory port.
screening process previously described by Chinchilli
Preliminary cytologic findings were obtained dur-
(44). In the comparison of sensitivities, patients with
ing fine-needle aspiration biopsy by a cytopatholo-
positive results (malignant cells) on pathology or
gist who was present during the procedure. Before
cytology were included in the definition of the gold
the sample was reviewed, Diff-Quik stain (Harleco,
standard. In contrast, only patients with negative
Gibbstown, New Jersey) was applied to the slide
results on surgical pathology were included in the
that contained the deposited specimen. Additional
definition of the gold standard for the comparison
passes were made until a positive cytologic result or
of specificities.
a negative result on an adequate tissue sample was
We also retrospectively compared the cost of en-
obtained (37).
doscopic ultrasonography (alone or with aspiration
Patients who were considered eligible for surgical
biopsy) with the cost of the more invasive staging
resection after staging by computed tomography and
procedures of mediastinoscopy and thoracotomy
endoscopic ultrasonography (that is, patients with a
that were aborted (because of the presence of me-
negative result on aspiration biopsy of contralateral
tastasis to lymph nodes) on the basis of the man-
or bulky subcarinal lymph nodes or those with nodes
agement options actually chosen for each patient.
that seemed to be benign according to endoscopic
ultrasonographic criteria) underwent thoracotomy for
pulmonary resection with ipsilateral mediastinal and
Results
subcarinal lymph node dissection. During mediasti-
nal dissection, each lymph node was placed in the
One hundred thirty patients with non-small-cell
compartment of a box that was labeled according to
lung cancer were evaluated during the study period.
the American Thoracic Society lymph node map;
In 52 of these patients, computed tomography showed
the compartment corresponded to the node's loca-
possible mediastinal lymphadenopathy, and no evi-
tion of origin. Histopathologic findings were deter-
dence indicated metastatic disease elsewhere (this
mined according to the location of the lymph nodes;
analysis includes the 17 patients in our pilot study
this allowed exact correlation between results of
[43]) (Figure 4). These 52 patients then had endo-
endoscopic ultrasonography, computed tomography,
and surgical pathology. The accuracy of computed
tomography, endoscopic ultrasonography alone, and
ultrasonography-guided aspiration biopsy of suspi-
cious mediastinal lymph nodes in predicting malig-
nant involvement was then compared with surgical
pathologic findings.
Statistical Analysis
The operating characteristics (sensitivity, specific-
ity, and positive and negative predictive values) and
false-positive and false-negative rates of endoscopic
ultrasonography and computed tomography in pre-
dicting nodal metastasis were first determined by
comparing the results of these procedures with the
results of surgical pathology and cytology of fine-
needle aspirates. Confidence intervals were com-
puted for this part of the study by using the "con-
fin" software program (Stat-Exact, Inc., Cambridge,
Massachusetts), which calculates exact CIs for pro-
portions. This study group included patients who
were evaluated by endoscopic ultrasonography alone
and then had surgery (most, including the initial 17
patients from our pilot study, were seen before the
development of ultrasonography-guided aspiration
biopsy) and patients who were studied with imaging
techniques followed by ultrasonography-guided as-
piration biopsy (Figure 4).
Figure 4. Flow diagram showing how patients were studied on
The operating characteristics and CIs for endo-
the basis of the study protocol. * = 8 subcarinal and 2 contralateral
mediastinal lymph nodes; t = 2 subcarinal and 12 contralateral mediastinal
scopic ultrasonography with and without aspiration
lymph nodes.
15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)
607

---

T a b l e 1 . Comparison of t h e Results of Endoscopic
results were false negative, and four were false pos-
Ultrasonography and Computed Tomography of
itive. The three false-negative results involved lymph
the Chest w i t h the Results of Surgical Pathology
or Cytology*
nodes that were 8 mm in diameter or less, each of
which had only a 1- to 2-mm focus of cancer on
Technique and Lymph Node Results of Pathology or Cytology
pathology. The four false-positive results involved
Appearance
Malignant Benign
lymph nodes that were more than 1.5 cm in diam-
Involvement
eter and appeared suspicious according to endo-
n
scopic ultrasonographic criteria but were shown by
Endoscopic ultrasonography
histopathology to have only reactive hyperplasia.
Malignant involvement 19t 4
Computed tomography also incorrectly determined
Benign 3 19
Computed tomography
that these same four nodes were positive. Endo-
Malignant involvement 14 15
scopic ultrasonography was best at accurately de-
Benign 8 8
tecting mediastinal lymph node metastasis in the
* For endoscopic ultrasonography, the accuracy was 84%, the positive predictive value
subcarinal (station 7), aortopulmonary window (sta-
was 83%, and the negative predictive value was 86%. For computed tomography, the
accuracy was 49%, the sensitivity was 64% (95% CI, 41 % to 83%), the specificity was
tion 5), paratracheal (station 4), and paraesophageal
35% (CI, 16% to 57%), the positive predictive value was 48%, and the negative
(station 8) regions (Figure 1).
predictive value was 50%.
t Includes 14 patients with positive results on fine-needle aspiration biopsy.
Computed tomographic scans of the mediastinal
lymph nodes in the same 45 patients were also
compared with pathologic results (Table 1). Com-
scopic ultrasonography. Twenty-four patients had en-
puted tomography correctly detected lymph node
doscopic ultrasonography-guided fine-needle aspira-
metastasis in 22 patients (overall accuracy, 49%;
tion biopsy of suspicious mediastinal and subcarinal
sensitivity, 64% [95% CI, 40.7% to 82.8%]). Fifteen
lymph nodes. A total of 24 lymph nodes (1 per
false-positive results and 8 false-negative results oc-
patient)—14 contralateral nodes and 10 subcarinal
curred; all of the false-positive results were ascribed
nodes—were aspirated. Fourteen patients had posi-
to the use of computed tomographic criteria that
tive cytologic findings on aspiration biopsy; this was
rely heavily on node size (13-23). The combination
accepted as representing definite nodal metastasis.
of computed tomography and endoscopic ultra-
On the basis of preoperative evaluations, 21 pa-
sonography did not improve overall accuracy be-
tients were considered to have unresectable tumors.
yond that seen with endoscopic ultrasonography
This group consisted of the 14 patients with malig-
alone because the three false-negative nodes on ul-
nant cytologic findings on ultrasonography-guided
trasonography were also false negative on computed
aspiration biopsy of contralateral or subcarinal
tomography; each of these nodes contained only a
lymph nodes and 7 patients in whom distant me-
small focus of micrometastasis.
tastasis or advanced disease (T4) was found by
endoscopic ultrasonography alone, computed to-
Tables 2 and 3 show the diagnostic findings for
mography, or other diagnostic methods. Thirty-one
the 24 patients who underwent ultrasonography-
patients ultimately had surgery (14 patients from
guided aspiration biopsy. Cytologic findings were
our initial pilot group, 10 patients with benign cy-
positive in 14 patients; these results confirmed the
tologic findings on ultrasonography-guided aspira-
presence of N3 disease in 12 patients and N2 dis-
tion biopsy, and 7 patients in whom aspiration bi-
ease (subcarinal involvement only) in 2 patients.
opsy was not performed because no suspicious Thus, these patients were ineligible for surgery. In 9
lymph nodes were detected according to the ultra-
of the 10 patients who had negative cytologic find-
sonographic criteria) and therefore had surgical
ings and ultimately underwent surgery, pathologic
pathologic results available.
The predictive value of endoscopic ultrasonogra-
phy for malignant involvement of lymph nodes in all
T a b l e 2 . Comparison of the Results of Endoscopic
patients who had surgical pathologic results is
Ultrasonography-Guided Fine-Needle Aspiration
Biopsy w i t h the Results of Surgical Pathology
shown in Table 1. These data included the ultra-
or Cytology*
sonographic results for the 14 patients who were
ultimately found to have positive cytologic results
Results of Endoscopic Results of Pathology or Cytology
Ultrasonography-Guided
on ultrasonography-guided aspiration biopsy and
Fine-Needle Aspiration Biopsy Malignant Benign
compare the nodal prediction provided by endo-
Involvement
scopic ultrasonography to the gold standard of ma-
n
lignancy seen on cytology or pathology. Of the 45
Malignant involvement 14 0
patients evaluated, endoscopic ultrasonography alone
Benign 1 9
accurately predicted mediastinal lymph node metas-
*The overall accuracy was 96%, the positive predictive value was 100%, and the
tasis in 38 patients (overall accuracy, 84%). Three
negative predictive value was 90%.
608 15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)

---

Table 3. Diagnostic Results in Patients Having Endoscopic
proach had a sensitivity of 95% (CI, 85% to 100%)
Ultrasonography-Guided Fine-Needle Aspiration
and a specificity of 81% (CI, 64% to 98%).
Biopsy and Comparison w i t h Predictions of
Mediastinal Lymph Node Status from
We retrospectively compared the direct costs of
Endoscopic Ultrasonography Alone and
ultrasonography-guided aspiration biopsy with the
Computed Tomography of the Chest
costs of mediastinoscopy and thoracotomy as they
are currently used for detecting mediastinal lymph
Patient Lymph Node Prediction FNAB Changed
Management?*
node involvement. The results of this comparison
EUS Alone EUS-Guided Computed
FNAB Tomography
(Table 5) suggest that ultrasonography-guided aspi-
ration biopsy may be a cost-saving method for ob-
1 + NSCLC + Yes
taining accurate final staging of non-small-cell lung
2 + NSCLC + Yes
3 + NSCLC - Yes
cancer.
4 + NSCLC + Yes
5 + NSCLC + Yes
6 + NSCLC + Yes
7 + NSCLC - Yes
Discussion
8 + NSCLC + Yes
9 + NSCLC + Yes
10 - Benign lymph node - Yes Our findings provide important information on
111" + Benign lymph node + No
12 + Benign lymph node + Yes
the accuracy of endoscopic ultrasonography-guided
13 + NSCLC - Yes
fine-needle aspiration biopsy for diagnosing metas-
14 + NSCLC - Yes
15 - Benign lymph node + Yes
tasis to mediastinal lymph nodes. They also support
16 - Benign lymph node + Yes
our earlier pilot data, which suggest that endoscopic
17 - Benign lymph node + Yes
18 - Benign lymph node - Yes
ultrasonography is more accurate than computed
19 - Benign lymph node + Yes
tomography in detecting metastasis to posterior me-
20 - Benign lymph node - Yes
21 + NSCLC + Yes
diastinal lymph nodes in patients with non-small-
22 + NSCLC + Yes
cell lung cancer (43). Current endoscopic ultrasono-
23 + NSCLC + Yes
24 + Benign lymph node + Yes
graphic criteria used to identify malignant lymph
nodes were correct in 84% of patients compared
* The result of fine-needle aspiration biopsy changed management in 95% of these
with 49% for computed tomographic criteria. The
patients (23 of 24). + = malignant appearance of lymph node; - = benign appear-
ance of lymph node. EUS = endoscopic ultrasonography; FNAB = fine-needle aspira-
combination of computed tomography and endo-
tion biopsy; NSCLC = non-small-cell lung cancer.
t Patient 11 had a false-negative result on fine-needle aspiration biopsy.
scopic ultrasonography did not improve overall ac-
curacy above that seen with endoscopic ultrasonog-
raphy alone for detecting lymph node involvement,
results confirmed the ultrasonographic prediction of
but it did help evaluate the extent of the lung
negative lymph nodes. In one patient with a nega-
cancer; detect distant metastasis not seen on endo-
tive cytologic result, a small focus of tumor invasion
scopic ultrasonography; and evaluate anterior and
was seen in one 8-mm lymph node (that is, the
pretracheal nodes, which are not imaged by endo-
cytologic result was false negative).
scopic ultrasonography.
The sensitivities and specificities of endoscopic
Computed tomography of the chest is considered
ultrasonography with and without aspiration biopsy
the gold standard in the preoperative evaluation of
and computed tomography were compared. Both
the mediastinum. However, this method relies pri-
the sensitivity and the specificity of endoscopic ultra-
marily on lymph node size for determining meta-
sonography alone and with aspiration biopsy were
static involvement. Our data suggest that size alone
significantly higher than those of computed tomog-
is not an accurate criterion for assessing lymph node
raphy. On the basis of cytologic findings, ultraso-
nography-guided aspiration biopsy influenced the
clinical management of 23 of 24 patients (95%)
Table 4
e
. Estimate
.
s o
s
f Sensitivit
o
y and S|
pecificity from a
Multistage !
e Screenin
!
g Analysis
because it avoided any further staging attempts with
of Endoscopic
Ultrasonogr aph
r
y and Guided I Fine-Needle
mediastinoscopy in all 24 patients and avoided un-
Aspiration E
n biops
E
y for Cytolog]
f
necessary surgery in the 14 patients who had posi-
tive cytologic results. The one patient with a false-
Results
Result o
s n
o Endoscopi
n
c
Endoscopi
Positive Pathologic
Negative Pathologic
Ultrasonography-Guided
Ultrasonography-Guide
Results
Results
negative result on cytology did not benefit from the
Fine-Needle
Fine-Needl Aspiratio
e
n
Aspiratio
ultrasonography-guided procedure and underwent
Biopsy
Biops
n
unnecessary surgery (Table 3).
Positive, positive
5
4
By using the Chinchilli approach, we estimated
Negative, negative
3
9
the sensitivities and specificities of the sequence of
Positive, negative
1
9
Positive, positive
14
0
endoscopic ultrasonography with and without aspi-
ration biopsy. The results of these calculations,
* The operating characteristics for the endoscopic ultrasc
>nography and guided fine-
needle aspiration biopsy multis tage screening process wer e as follows: sensitivity, 95%
shown in Table 4, revealed that the combined ap-
(95% CI, 85% to 100%) and specificity, 8 1 % (CI, 64%
to 98%).
15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)
609

---

Table 5. Cost Comparison among Endoscopic
eral anesthesia and inpatient recovery, thereby in-
Ultrasonography-Guided Fine-Needle Aspiration
creasing the time, cost, and risk of the staging pro-
Biopsy, Mediastinoscopy, and Thoracotomy for
Detection of Mediastinal and Subcarinal Lymph
cess. Another limitation is that mediastinoscopy
Node Metastasis in Patients w i t h Non-Small-Cell
cannot reliably evaluate subcarinal nodes.
Lung Cancer
With the advent of endoscopic ultrasonography,
guided fine-needle aspiration sampling of suspicious
Procedure Cost Average Length
of Stay
lymph nodes became possible. We postulated that
the use of this technology in conjunction with en-
$ d
doscopic ultrasonography staging could enhance the
Endoscopic ultrasonography-guided
overall accuracy of detecting metastasis of non-
fine-needle aspiration biopsy 0
small-cell lung cancer to mediastinal lymph nodes.
Professional fee 800.00
Hospital charges (includes all outpatient
Thus, once guided aspiration biopsy became avail-
charges)! 1175.00
able, all eligible patients presenting to our service
Total 1975.00 2.0
Mediastinoscopy
underwent this procedure to confirm the existence
Professional fee 1000.00
of metastasis in any suspicious lymph node seen on
Hospital charges (includes all inpatient
charges and 2-day hospital stay)t 6759.00
endoscopic ultrasonography. These patients were then
Total 7759.00
followed to assess the effect of the resulting cyto-
Thoracotomy 9.45
Professional fee 2320.00
logic findings on subsequent management. Our data
Hospital charges (includes inpatient
indicate that this new technique represents an im-
charges and 9-day hospital stay) t23 708.00
Total 26 028.00
provement over existing staging methods because it
is less invasive and has a sensitivity of 95% for de-
* To calculate the professional fee, we used the endoscopic ultrasonography CPT code,
which reimburses for the ultrasonography. The guided fine-needle aspiration biopsy
termining metastasis to posterior mediastinal lymph
portion is coded for the specific biopsy site,
nodes. In light of these results, it seems reasonable
t Average hospital charges for each of the procedures, based on a financial review of all
patients having these procedures from January 1995 to June 1995 at Indiana University
to consider using endoscopic ultrasonography and
Hospital.
ultrasonography-guided fine-needle aspiration biop-
sy in patients with suspicious posterior mediastinal
status. Size does not accurately differentiate inflam-
lymph nodes; at least for now, it might be used in
matory or reactive lymph nodes from malignant in-
patients in whom previous lymph node sampling
volvement and cannot detect normal-appearing done by use of computed tomography or broncho-
lymph nodes that contain small foci of micrometas-
scopic techniques was unsuccessful. Mediastinoscopy
tasis (13-24). Furthermore, geographic factors influ-
or partial thoracotomy could be reserved for pa-
ence the size of mediastinal lymph nodes. For ex-
tients with enlarged anterior lymph nodes or suspi-
ample, in Indiana (an area in which histoplasmosis
cious nodes that were not successfully sampled by
is endemic), the majority of the population has
computed tomography, bronchoscopy, or endoscopic
large mediastinal lymph nodes (36-40). We postu-
ultrasonography.
late that this regional phenomenon may be partly
The overall diagnostic accuracy of endoscopic
responsible for the relatively low overall accuracy of
ultrasonography alone or with fine-needle aspiration
computed tomography (49%) in evaluating such
biopsy was 96%; this method was significantly better
nodes in our patients.
than computed tomography for predicting medias-
Patients in whom non-small-cell lung cancer is
tinal lymph node metastasis, particularly when me-
diagnosed undergo routine staging with computed
tastases involved the subcarina and posterior me-
tomography of the chest and bronchoscopy. Bron-
diastinum. Furthermore, the results of fine-needle
choscopic fine-needle aspiration biopsy is commonly
aspiration biopsy contributed significantly to patient
used to evaluate suspicious paratracheal, hilar, and
management because cancer in patients who had
subcarinal lymph nodes seen on computed tomog-
positive cytologic results was staged as pathologic
raphy for malignant involvement. The role of this
III A disease (bulky subcarinal nodes). These pa-
technique in the diagnosis and staging of non-small-
tients were considered to be better served by proto-
cell lung cancer is well established (45-50). Malig-
cols that included chemotherapy and radiation rather
nant involvement of mediastinal lymph nodes, par-
than surgical resection. No complications of guided
ticularly bulky subcarinal or contralateral nodes, on
fine-needle aspiration biopsy occurred in our study.
bronchoscopic fine-needle aspiration biopsy of
Despite its many potential advantages, endo-
nodes found to be enlarged on computed tomogra-
scopic ultrasonography is not yet widely available,
phy usually precludes surgery. When lymph node
largely because of a lack of skilled endosonogra-
status according to both computed tomography and
phers; the procedure can therefore be done only in
bronchoscopy is unknown, mediastinoscopy and, in academic or tertiary referral medical centers. It will
some cases, thoracotomy are usually performed to
be some time before many institutions have an op-
clarify disease stage. These procedures require gen-
portunity to use this unique approach to staging
610 15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)

---

non-small-cell lung cancer. Until then, they must
puted tomography, endoscopic ultrasonography alone,
rely on bronchoscopic or transthoracic fine-needle
and endoscopic ultrasonography-guided fine-needle
aspiration biopsy and mediastinoscopy, which, in ex-
aspiration biopsy with the nodes found on careful
perienced hands, contribute substantially to the de-
surgical dissection, certain nodes from difficult sur-
tection of mediastinal lymph node metastasis and to
gical resection may have been mismatched. Finally,
the overall staging of non-small-cell lung cancer
our sample size is relatively small. We hope that
(51, 52). The use of endobronchial sonography to
future studies with larger samples, preferably drawn
evaluate bronchial carcinomas was recently reported
from multiple centers, will more clearly define the
(53). Even more interesting is the development of
role of ultrasonography and endoscopic ultrasonog-
an ultrasonic bronchoscope that can be used to
raphy-guided fine-needle aspiration biopsy in the
diagnose lung masses (54). Future use of this new
preoperative staging of non-small-cell lung cancer.
generation of endobronchial instruments will prob-
In summary, endoscopic ultrasonography by itself
ably enhance our ability to diagnose and locally
provided useful information about the status of pos-
stage lung tumors and to determine lymph node
terior mediastinal lymph nodes in patients with
status.
non-small-cell lung cancer. The addition of trans-
Our study has some limitations. First, as noted,
esophageal endoscopic ultrasonography-guided fine-
our patients had an increased regional prevalence of
needle aspiration biopsy further improved the over-
enlarged mediastinal and subcarinal lymph nodes.
all accuracy of lymph node staging to 96%. Thus,
Because histoplasmosis is endemic in Indiana, much
endoscopic ultrasonography-guided fine-needle as-
of the population is at risk for developing enlarged
piration biopsy seems to be an important addition
benign mediastinal nodes; this compounds the prob-
to the overall preoperative staging of patients with
lem of staging in non-small-cell lung cancer. Most
potentially resectable non-small-cell lung cancer
of these patients subsequently have more aggressive
through its ability to establish the pathologic diag-
and invasive staging procedures. The high preva-
nosis of N2 and N3 disease. Further investigations
lence of benign enlarged nodes may have contrib-
are now needed, including prospective comparisons
uted to the low accuracy of computed tomography
of cost-effectiveness and outcome, examination of
in our study sample; the diminished value of com-
the role of spiral computed tomography, and the
puted tomography in Indiana may therefore not
evaluation of patient populations in which enlarged
apply to other parts of the United States. Converse-
mediastinal lymphadenopathy are uncommon (for
ly, if the diagnostic accuracy of computed tomogra-
example, in areas where histoplasmosis is not en-
phy is improved in other regions, the advantages of
endoscopic ultrasonography may be less obvious
demic) before the role of endoscopic ultrasonogra-
elsewhere. No data support this possibility, however,
phy alone or with fine-needle aspiration biopsy in
and endoscopic ultrasonography has not been sys-
the preoperative staging of non-small-cell lung can-
tematically compared with computed tomography
cer can be considered firmly established.
for overall accuracy in detecting mediastinal nodes.
Acknowledgments: The authors thank Marilyn Datzman, MD, and
Studies must be performed to answer these ques-
Syed Zaidi, MD, for technical support and recruitment of pa-
tions and to verify whether endoscopic ultrasonog-
tients; Chris Lambert, Nancy Ruiz, and Stephen Melson for
raphy can detect lymph nodes not seen on com-
editorial and typing assistance; and Mark Hanna, MS, and Rene
Gonin, PhD, for statistical advice.
puted tomography.
Second, we used what some might consider an
Grant Support: In part by a research grant from the American
unusual gold standard—the combination of fine-
College of Gastroenterology.
needle aspiration cytology and surgical pathology.
Requests for Reprints: Frank G. Gress, MD, Division of Gastro-
We considered positive (malignant) results on cytol-
enterology and Hepatology, Winthrop-University Hospital, State
ogy to be equal to positive results on pathology
University of New York at Stony Brook, School of Medicine and
Health Science Center, 222 Station Plaza North, Suite 429, Long
because false-positive cytologic results are rare and
Island, NY 11501.
usually occur with inexperienced operators. Endo-
scopic ultrasonography-guided fine-needle aspira-
Current Author Addresses: Dr. Gress: Division of Gastroenterol-
tion biopsy done by using a 22-gauge needle pro-
ogy and Hepatology, Winthrop-University Hospital, State Uni-
versity of New York at Stony Brook, School of Medicine and
vides ample tissue, in contrast to the limited amount
Health Science Center, 222 Station Plaza North, Suite 429, Long
of tissue usually obtained with the 25-gauge ("skin-
Island, NY 11501.
ny") needles used for most cytologic sampling.
Dr. Savides: University of California at San Diego, 200 West
Arbor Drive, San Diego, CA 92103.
A third limitation is the inability of endoscopic
Drs. Sandler, Kesler, Conces, Cummings, and Mathur and Ms.
ultrasonography to image the anterior mediastinum;
Bilderback: 550 North University Boulevard, Indiana University
thus, the use of this procedure is restricted primarily
Medical Center, Indianapolis, IN 46202.
Dr. Ikenberry: 200 East Pennsylvania, Peoria, IL 45050.
to the evaluation of the posterior mediastinum. Dr. Hawes: Medical University of South Carolina, 171 Ashley
Fourth, although we correlated our findings on com-
Avenue, Charleston, SC 29425.
15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)
611

---

References
operative classification compared to the new (1987) TNM system. Gastroen-
terology. 1989;96:1478-86.
28. Rosen T, Lorenz R, Zenker K, von Wichert A, Dancygier H, Hofler H, et
1. DeVita VT, Hellman S, Rosenberg SA, eds. Important Advances in Oncol-
al. Local staging and assessment of resectability in carcinoma of the esoph-
ogy. Philadelphia: JB Lippincott; 1993:259-72.
agus, stomach, and duodenum by endoscopic ultrasonography. Gastrointest
2. Martini N, Baines MS, McCormick PM, Kaiser LR, Burt ME. Surgical
Endosc. 1992;38:460-7.
treatment in non-small cell carcinoma of the lung: the Memorial Sloan-Ket-
29. Palazzo L, Roseau G, Gayet B, Vilgrain V, Belghiti J, Fekete F, et al.
tering experience. In: Hoogstraten B, Addis BJ, Hansen HH, eds. Treatment of
Endoscopic ultrasonography in the diagnosis and staging of pancreatic ade-
Lung Tumors. Heidelberg, Germany: Springer-Verlag; 1987:111-32.
nocarcinoma. Results of a prospective study with comparison to ultrasonog-
3. Sandler AB, Buzaid AC. Lung cancer: a review of current therapeutic mo-
raphy and CT scan. Endoscopy. 1993;25:143-50.
dalities. Lung. 1992;170:249-65.
30. Lee N, Inoue K, Yamamoto R, Kinoshita H. Patterns of internal echoes in
4. Hatter J, Kohman U , Mosca RS, Graziano SL, Veit U , Coleman M.
lymph nodes in the diagnosis of lung cancer metastasis. World J Surg. 1992;
Preoperative evaluation of stage I and stage II non-small cell lung cancer. Ann
16:986-93.
Thorac Surg. 1994;58:1738-41.
3 1 . Sugimachi K, Ohno S, Fujishima H, Kuwano H, Mori M, Misawa T.
5. Shields TW. Surgical therapy for carcinoma of the lung. Clin Chest Med.
Endoscopic ultrasonographic detection of carcinomatous invasion of lymph
1993;14:121-47.
nodes in the thoracic esophagus. Surgery. 1990;107:366-71.
6. Medina Gallardo JF, Borderas Naranjo F, Torres Cansino M, Ro-
32. Schuder G, Isringhaus H, Kubale B, Seitz G, Sybrecht GW. Endoscopic
driquez-Panadero F. Validity of enlarged mediastinal nodes as markers of
ultrasonography of the mediastinum in the diagnosis of bronchial carcinoma.
involvement by non-small cell lung cancer. Am Rev Respir Dis. 1992; 146:
Thorac Cardiovasc Surg. 1991;39:299-303.
1210-2.
33. Kondo D, Imaizumi M, Abe T, Naruke T, Suemasu K. Endoscopic ultra-
7. Van Raemdonck DE, Schneider A, Ginsberg RJ. Surgical treatment for
sound examination for mediastinal lymph node metastases of lung cancer.
higher stage non-small cell lung cancer. Ann Thorac Surg. 1993;54:999-1013.
Chest. 1990;98:586-93.
8. Naruke T, Goya T, Tsuchiya R, Suemasu K. The importance of surgery to
34. Kobayashi H, Danbara T, Tamaki S, Kitamura S, Hata E, Fukushima K,
non-small cell carcinoma of lung with mediastinal lymph node metastasis.
et al. Detection of the mediastinal lymph node metastasis in lung cancer by
Ann Thorac Surg. 1988;46:603-10.
endoscopic ultrasonography. Jpn J Med. 1988;27:17-22.
9. Martini N, Flehinger BJ. The role of surgery in N2 lung cancer. Surg Clin
35. Aibe T, Ito T, Yoshida T, Noguchi T, Ohtani T, Fuji T, et al. Endoscopic
North Am. 1987;67:1037-49.
ultrasonography of lymph nodes surrounding the upper Gl tract. Scand J
10. Mountain CF. The biological operability of stage III non-small cell lung can-
Gastroenterol Suppl. 1986;123:164-9.
cer. Ann Thorac Surg. 1985;40:60-4.
36. Wiersema MJ, Hassig W M , Hawes RH, Wonn MJ. Mediastinal lymph
11. Glazer G M , Gross BH, Quint LE, Francis IR, Bookstein FL, Qrringer MB.
node detection with endosonography. Gastrointest Endosc. 1993;39:788-93.
Normal mediastinal lymph nodes: number and size according to American
37. Ikenberry S, Gress FG, Savides TA, Hawes RH. Fine needle aspiration of
Thoracic Society mapping. AJR Am J Roentgenol. 1985;144:261-5.
posterior mediastinal lymph nodes guided by radial scanning endosonogra-
12. Mountain CF. A new international staging system for lung cancer. Chest.
phy. Gastrointest Endosc. 1996;43:605-10.
1986;89(4Suppl):225S-33S.
38. Wiersema M, Hawes RH, Tao LC, Wiersema LM, Kopecky KK, Rex DK,
13. Martini N, Flehinger BJ, Zaman MB, Beattie EJ Jr. Prospective study of
et al. Endoscopic ultrasonography as an adjunct to fine needle aspiration
445 lung carcinomas with mediastinal lymph node metastases. J Thorac Car-
cytology of the upper and lower gastrointestinal tract. Gastrointest Endosc.
diovasc Surg. 1980;80:390-9.
1992;38:35-9.
14. Ingram CE, Belli A M , Lewars MD, Peznek RH, Husband JE. Normal
39. Rex DK, Tarver RD, Wiersema M, O'Connor KW, Lappas JC, Taba-
lymph node size in the mediastinum: a retrospective study in two patient
towski K. Endoscopic transesophageal fine needle aspiration of mediastinal
groups. Clin Radiol. 1989;40:35-9.
masses. Gastrointest Endosc. 1991;37:465-8.
15. Mann H. CT in the management of lung cancer. Semin Ultrasound CT MR.
40. Wiersema MJ, Kochman ML, Chak A, Cramer H M , Kesler KA. Real-time
1988;9:40-52.
endoscopic ultrasound-guided fine-needle aspiration of a mediastinal lymph
16. Tisi G M , Friedman PJ, Peters RM. Clinical staging of primary lung cancer.
node. Gastrointest Endosc. 1993;39:429-31.
Official Statement of the American Thoracic Society. Am Rev Respir Dis. 1983;
4 1 . Villman P, Hancke S, Hendricksen FW, Jacobsen GK. Endosonographi-
127:659-64.
cally-guided fine needle aspiration biopsy of malignant lesions in the upper
17. McKenna RJ Jr, Libshitz HI, Mountain CE, McMurtrey MJ. Roentgeno-
gastrointestinal tract. Endoscopy. 1993;25:523-7.
graphic evaluation of mediastinal nodes for preoperative assessment in lung
42. Chang KJ, Katz KD, Durbin TE, Erickson RA, Butler JA, Lin F, et al.
cancer. Chest. 1985;88:206-10.
Endoscopic ultrasound-guided fine-needle aspiration. Gastrointest Endosc.
18. Patterson GA, Ginsberg RJ, Poon PY, Cooper JD, Goldberg M, Jones D,
1994;40:694-9.
et al. A prospective evaluation of magnetic resonance imaging, computed
43. Hawes RH, Gress F, Kesler KA, Cummings OW, Conces DJ Jr. Endo-
tomography, and mediastinoscopy in the preoperative assessment of medias-
scopic ultrasound versus computed tomography in the evaluation of the me-
tinal node status in bronchogenic carcinoma. J Thorac Cardiovasc Surg. 1987;
diastinum in patients with non-small-cell lung cancer. Endoscopy. 1994;26:
94:679-84.
784-7.
19. Aronchick JM. CT of mediastinal lymph nodes in patients with non-small cell
lung carcinoma. Radiol Clin North Am. 1990;28:573-81.
44. Chinchilli V. Estimates of sensitivity and specificity in a multi-stage screen for
20. Glazer GM, Orringer MB, Gross BH, Quint LE. The mediastinum in non-
medical diagnosis. Biometrics. 1983;39:333-40.
small cell lung cancer: CT-surgical correlation. AJR Am J Roentgenol. 1984;
45. Schieppati E. Mediastinal lymph node puncture through the tracheal carina.
142:1101-5.
Surg Gynecol Obstet. 1958;107:243-6.
2 1 . Cybulsky IJ, Lanza LA, Ryan MB, Putnam JB Jr, McMurtrey M M , Roth
46. Wang KP, Terry P, Marsh B. Bronchoscopic needle aspiration biopsy of
JA. Prognostic significance of computed tomography in resected N2 lung
paratracheal tumors. Am Rev Respir Dis. 1978;118:17-21.
cancer. Ann Thorac Surg. 1992;54:533-7.
47. Harrow EM, Oldenburg FA Jr, Lingenfelter MS, Smith A M Jr. Trans-
22. McCloud TC, Bourgouin PM, Greenberg RW, Kosiuk JP, Templeton PA,
bronchial needle aspiration in clinical practice. A five-year experience. Chest.
Shepard JA, et al. Bronchogenic carcinoma: analysis of staging in the me-
1989;96:1268-72.
diastinum with CT by correlative lymph node mapping and sampling. Radiol-
48. Schenk DA, Bower JH, Bryan CL, Currie RB, Spence TH, Duncan CA, et
ogy. 1992;182:319-23.
al. Transbronchial needle aspiration staging of bronchogenic carcinoma. Am
23. Lewis JW Jr, Pearlberg JL, Beute GH, Alpern M, Kvale PA, Gross BH, et
Rev Respir Dis. 1986;134:146-7.
al. Can computed tomography of the chest stage lung cancer? Yes and no.
49. Harrow E, Halber M, Hardy S, Halteman W. Bronchogenic and roent-
Ann Thorac Surg. 1990;49:591-6.
genographic correlates of a positive transbronchial needle aspiration in the
24. Staples CA, Muller NL, Miller RR, Evans KG, Nelems B. Mediastinal
staging of lung cancer. Chest. 1991;100:1592-6.
nodes in bronchogenic carcinoma: comparison between CT and mediastinos-
50. Carlin BW, Harrell JH 2d, Fedullo PF. False-positive transcarinal needle
copy. Radiology. 1988;167:367-72.
aspirate in the evaluation of bronchogenic carcinoma. Am Rev Respir Dis.
25. Jolly PC, Hutchinson CH, Detterbeck F, Guyton SW, Hofer B, Anderson
1989;140:1800-2.
RP, et al. Routine computed tomographic scans, selective mediastinoscopy,
51. Wang KP, Terry PB. Transbronchial needle aspiration in the diagnosis and
and other factors in evaluation of lung cancer. J Thorac Cardiovasc Surg.
staging of bronchogenic carcinoma. Am Rev Respir Dis. 1983;127:344-7.
1991;102:266-71.
52. Midthun DE, Cortese DA. Bronchoscopic needle aspiration and biopsy. In:
26. Merav AD. The role of mediastinoscopy and anterior mediastinotomy in
Prakash UB, ed. Bronchoscopy. New York: Raven Pr; 1994:147-53.
determining operability of lung cancer: a review of published questions and
53. Hurter T, Hanrath P. Endobronchial sonography: feasibility and preliminary
answers. Cancer Invest. 1991;9:439-42.
results. Thorax. 1992;47:565-7.
27. Tio TL, Cohen P, Coene PP, Udding J, den Hartog Jager FC, Tytgat GN.
54. Ono R, Suemasu K, Matsunaka T. Bronchoscopic ultrasonography in the
Endosonography and computed tomography of esophageal carcinoma: pre-
diagnosis of lung cancer. Jpn J Clin Oncol. 1993;23:34-40.
612 15 October 1997 • Annals of Internal Medicine • Volume 127 • Number 8 (Part 1)

---