Neural Cell Adhesion Molecule
Expression*
Prognosis in 889 Patients With Resected Non-small
Cell Lung Cancer
Rene Hage, MD; Hans R. J. Fibers, MD, PhD;
Aart Brutel de la Riviere, MD, PhD, FCCP; and
Jules M. M. van den Bosch, MD, PhD, FCCP
Introduction: In search of factors that might predict outcome in patients with resected non-small
cell lung cancer (NSCLC), we studied the reactivity of monoclonal antibody 123C3. This marker
of neuroendocrine (NE) differentiation is directed against neural cell adhesion molecules
(NCAM). Although NCAM can often be demonstrated in small cell lung cancer and carcinoids as
a tumor antigen, not many data exist on NCAM in NSCLC.
Patients and methods: From 1983 through 1995, in 889 patients with NSCLC, who underwent
pulmonary resection, 123C3 reactivity was tested. NCAM was correlated with tumor histology,
p-TNM stage, and 5-year survival. Large cell NE carcinomas were excluded. Monoclonal
antibody-1 (MOC-1) was also tested on most specimens.
Results: Reactivity of 123C3 does not correlate with tumor histology, p-TNM stage, or 5-year
survival. In addition, MOC-1 reactivity was not significantly related to prognosis.
Conclusions: Positive reactivity with 123C3, indicating NE differentiation, does not have
predictive value in NSCLC. Also, tumor histology or stage did not correlate with 123C3 reactivity.
Reactivity of MOC-1 did not contribute to prediction of prognosis. Whether there is more
chemosensitivity in NSCLC-NE than in NSCLC without NE differentiation remains an important
(CHEST 1998; 114:1316-1320)
question that is not addressed by our present study.
Key words: lung cancer; NCAM; neuroendocrine differentiation; prognosis
Abbreviations: LCNEC large cell neuroendocrine carcinoma; MOC-1 monoclonal antibody-1; NCAM neural
cell adhesion molecule; NE neuroendocrine differentiation; NSCLC non-small cell lung carcinoma; NSCLC-NE
non-small cell lung carcinoma with neuroendocrine differentiation; SCLC small cell lung carcinoma
=
=
=
=
=
=
=
A dhesion molecules can be subdivided into four
***different classes: integrins, cadherins, selectins,
and immunoglobulin superfamily. By complex inter¬
actions, these molecules take part in intercellular
adhesion, as in embryogenesis, cell growth, and cell
differentiation. These molecules are thought to play
a critical role in carcinogenesis and the development
of metastases. Neural cell adhesion molecules
(NCAM) belong to the immunoglobulin superfamily.
By using monoclonal antibodies, such as 123C3 that
recognizes NCAM, the presence of NCAM expres¬
sion
can
be demonstrated.
*From the departments of Pulmonology (Drs. Hage and van den
Bosch), Pathology (Dr. Elbers), and Thoracic Surgery (Dr.
Brutel de la Riviere), St. Antonius Hospital, Nieuwegein, the
Netherlands.
Manuscript received December 11, 1997; revision accepted May
29, 1998.
den Bosch, MD, PhD, FCCP,
Correspondence to: J.M.M. van
St. Antonius Hospital, PO Box 2500,
of
Department
Pulmonology,
3430 EM Nieuwegein, the Netherlands.
Monoclonal antibody 123C3 stains positive when
applied to neuroendocrine (NE) cells, although
other cells can also show reactivity with 123C3.1
However, normal lung tissue does not generally react
with 123C3.
Monoclonal antibodies against tumor-associated
antigens are used to study factors that might provide
diagnostic or prognostic tools in patients with small
cell lung cancer (SCLC) and non-small cell lung
cancer
(NSCLC).2'3
Most tumor-associated
antigens
are
not
specific
for a certain type of tumor and not all tumor cells
express these antigens. However, in both the primary
tumor and its metastases, levels of tumor-associated
antigens often are elevated in comparison with nor¬
mal tissue.
Positive staining for 123C3 or monoclonal anti¬
body-1 (MOC-1) of pulmonary tumors is indicative
of NE differentiation.
1316
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21874/ on 06/15/2017
Clinical
Investigations
In most SCLC and bronchial carcinoids, both NE
NCAM expression can be demon¬
malignancies,
strated using 123C3. Also, in about 20% of NSCLC,
NCAM can be demonstrated. NCAM positivity has
reported to be associated with a poor progno¬
sis.24 Pujol et al4 demonstrated NCAM with MOC-1
and S-L 11.14 in NSCLC, while NCAM was dem¬
onstrated by 123C3 reactivity by Kibbelaar et al.2 In
the study of Kwa et al,5'6 no relationship between
NCAM and a poor prognosis was shown. However,
been
only a limited number of studies have investigated
123C3 reactivity in relation to tumor histology, TNM
tumor stage, and 5-year survival in patients with
resected NSCLC.2-4 We therefore investigated the
prognostic value of reactivity with monoclonal anti¬
body 123C3 in patients with resected NSCLC.
Materials
and
Methods
From 1983 through 1995, reactivity of monoclonal antibody
123C3 was tested in 889 patients with NSCLC, who underwent
pulmonary resection in the St Antonius Hospital Nieuwegein, the
Surgery was performed with the intention to carry
complete resection. Eighty-nine percent of the patients
(n 791) were male. Tumor histology was classified by light
microscopy, using tissue sections prepared by routine paraffin
technique. Diagnosis was based on light microscopic criteria from
the World Health Organization.78
There were 575 patients with squamous cell carcinoma, 262
with adenocarcinoma, 36 with adenosquamous cell carcinoma,
and 16 with large cell undifferentiated carcinoma. Large cell
Netherlands.
out
a
=
neuroendocrine carcinomas (LCNEC) were excluded.
After extensive routine mediastinal and subcarinal lymph node
sampling during surgery, postsurgical (p-)TNM classification was
determined, according to the International Staging Systems.9
In 889 patients, 123C3 was applied, while MOC-1 and 123C3
were tested in 799 patients. In all cases, representative tissue
blocks were snap frozen in liquid nitrogen during surgery. Frozen
section immunohistochemistry with MOC-1 (CD56, SCLC,
clone MOC-1) and 123C3 was performed in 799 patients.
Monoclonal antibody 123C3 is an IgGl murine antibody, raised
at the Netherlands Cancer Institute against a membrane fraction
of a fresh SCLC specimen. It belongs to cluster I according to the
International Workshop on Small Cell Lung Cancer Antibodies
and recognizes an epitope on NCAM. For immunohistochemis¬
try, the peroxidase labeled streptavidin (LSAB-Dako) complex
technique with diaminobenzidine chromogen as substrate was
used. The sources and dilutions used were as follows: 123C3
(Antoni van Leeuwenhoek Hospital, Amsterdam, the Nether¬
lands; 1:200); MOC-1 (Sanbio, Uden, The Netherlands; 1:20).
Incubation with unrelated antibodies or omitting the specific
antibody was used as negative control. As positive controls for
123C3 and MOC-1, slides from a typical bronchial carcinoid were
used. For both 123C3 and MOC-1, only distinct cell membranepattern staining of tumor cells was scored as positive, even when
the number of positive tumor cells was small.
We studied the relation of 123C3 reactivity with tumor histol¬
ogy, p-TNM stage, and 5-year survival. A possible relationship
between the combination of 123C3 and MOC-1 reactivity and
5-year survival was also investigated. Survival was defined as time
from surgical resection to the date of death. Postoperative
hospital deaths were excluded. For hospital survivors KaplanMeier survival analysis was performed and survival curves were
constructed according to the Kaplan-Meier method.10 Differ¬
in survival between patient groups were tested by the
log-rank test for statistical significance.11 The x~ test was used to
test differences between histologic subtypes and 123C3 reactiv¬
ity. This test was also performed to test differences between
p-TNM stages and 123C3 reactivity. Results with p values <0.05
were considered to be significant.
ences
Tumor Histology
Results
and 123C3 Reactivity
NSCLC of various histo¬
shown in Table 1. From all histologic
Reactivity of 123C3 in
logic types
subtypesforof NSCLC, only 128 tumors (14%) were
positive 123C3 staining. Between the histologic
subtypes, there was no significant inhomogeneity
with 123C3 reactivity (p 0.093, x2 test).
is
-
p-TNM stage and 123C3 Reactivity
Data on positive staining of 123C3 in relation to
p-TNM stage are shown in Table 1. There was no
between 123C3 reac¬
statistically
significant relation
and
tumor
No
tivity
stage. positive correlation could
be demonstrated between the different stages
0.18,
x2 test).
(p
=
MOC-1 and 123C3 Reactivity
We studied the association between 123C3 and
MOC-1 (Table 1). In 113 patients, 123C3 was
positive, and neuroendocrine differentiation was
confirmed by MOC-1 in 98 patients (86.7%). In 686
patients, 123C3 was negative, while MOC-1 was also
negative in 685 patients (99.9%).
Hospital Mortality
Overall hospital mortality was 3.4% (n 30). In
patients with squamous cell carcinoma, hospital mor=
Table 1.Number of Patients in Whom 123C3
Reactivity Is Tested According to Tumor Histology,
Stage, or MOC-1 Reactivity
123C3 POS, 123C3 NEG,
No. (%)
No. (%)
Total
86 (15)
Squamous cell carcinoma
Adenocarcinoma
30(11)
7 (19)
Adenosquamous cell carcinoma
(31)
Large cell carcinoma 5
Total128(14)
Stage IIIA
4 (6)
Stage IIIB
0
Stage IV
MOC-1 POS
MOC-1 NEG
Stage 5I 0(17)
Stage II40(14)
34 (14)
(0.0)
(99)
15 (2)
98
489 (85)
232(89)
29 (81)
11 (69)
761(86)
253(83)
237(86)
208 (86)
62 (94)
1 (100)
1 (1)
685 (98)
CHEST/114/5/NOVEMBER, 1998
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21874/ on 06/15/2017
575
262
36
16
889
303
277
242
66
1
99
700
1317
tality was 2.8% (n 25) of all resected NSCLC. This
was 0.45% (n
4) for patients with adenocarcinoma,
0.11% (n
1) for adenosquamous cell carcinoma,
and 0% (n 0) for large cell carcinoma.
(Table 2). We concluded that these parameters were
significantly correlated. In addition, MOC-1, as
an independent marker, was not a prognostic factor.
Cumulative Survival
Cumulative survival of patients with 123C3 posi¬
tive and negative staining of resected NSCLC is
shown in the Kaplan-Meier curve (Fig 1).
Discussion
-
=
=
not
=
Five-Year Survival and Tumor Histology:
Excluding hospital mortality,
5-year survival ac¬
is shown in Table 2. No
cording to tumor histology
significant differences in 5-year survival between
NSCLC tumors of different histology could be dem¬
onstrated.
Five-Year Survival and p-TNM
Stage:
No correlation could be demonstrated between
123C3 reactivity and 5-year survival at any p-TNM
stage (Table 2).
Five-Year Survival and Staining of Both 123C3
and MOC-1:
The possible correlation between 123C3 and
MOC-1 reactivity and 5-year survival was studied
The
important indicators of prognosis in
patients with NSCLC are TNM stage and Karnofsky
or Eastern Cooperative Oncology Group perfor¬
mance score. However, within each TNM stage,
both tumor behavior and patient characteristics may
vaiy. In search of factors that might influence treat¬
ment, or might predict prognosis of individual pa¬
tients more accurately, monoclonal antibodies have
been used to detect tumor-associated antigens.
In the present study, we investigated the prognos¬
tic significance of NCAM, in patients with resected
NSCLC. This was done using monoclonal antibody
123C3, directed against NCAM. The presence of
most
NCAM is highly indicative of NE differentiation. We
did not find a correlation between 123C3 reactivity
and NSCLC tumor histology, p-TNM stage, or
5-year survival. In addition, MOC-1 as an indepen¬
dent marker did not contribute to predicting prog¬
nosis. Therefore, the presence of NE differentiation
in NSCLC, demonstrated by immunohistochemistry,
did not correlate with prognosis. However, since all
CO
>
Z5
CO
<D
>
i?
E
o
Years after resection
Figure 1.
NSCLC.
Kaplan-Meier curve showing 123C3 positive (dotted line) and 123C3 negative (dashed line)
1318
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21874/ on 06/15/2017
Clinical
Investigations
Table 2.Five-Year Survival in Patients With 123C3
Reactivity According to Tumor Histology or p-TNM
Stage*
123C3
123C3
POS, %
NEG, %
Squamous cell carcinoma
Adenocarcinoma
Adenosquamous
cell carcinoma
Large cell carcinoma
Stage
II
Stage
Stage IIIA
Stage IIIB
Stage IV NA 0.0 NA
*NA
=
not
applicable; NS
30.5
NS40.5
37.8
NS 31.8
57.1
30.6
NS
0.0 NS
67.3
I51.4
48.5
NS
27.4
41.6
NS
14.6
NS26.5
25.0
NS19.5
Total33.0
=
not
Value
37.8
NS
significant.
patients studied were candidates for resection, pa¬
tients with clinical stage IV were excluded from
it would be
that these
study. Although
unlikely
patients would have significant impact on our results
when studied, the possibility of selection bias cannot
be excluded.
The number of studies with NCAM in NSCLC is
limited2-6 and the results are not conclusive.
In a study with 96 patients, using preheated
tissue sections, Kwa et al56 also
paraffin-embedded
concluded that NCAM was not related to survival in
surgicallyetresected NSCLC.
Pujol al,4 however, who demonstrated NCAM
expression by MOC-1 and S-L 11.14 in 94 surgically
treated patients, concluded that NCAM expression
in NSCLC does not correlate with nodal status and
indicates a poor prognosis. Furthermore, in the study
of Kibbelaar et al,2 positive 123C3 reactivity in
NSCLC showed significantly shorter survival times,
compared with 123C3 negative NSCLC. Kibbelaar
et al2 demonstrated positive 123C3 reactivity in 53 of
278 patients after surgery for NSCLC. In our study,
we investigated 889 patients, 278 of whom were
studied by Kibbelaar et al.2 Although the study of
Kibbelaar et al2 and our present study show overlap
of 278 patients, the conclusions are contradictoiy.
Differences could be explained by the smaller num¬
ber of patients of Kibbelaar et al,2 shorter survival
analysis (30 months), and lack of correction for TNM
tumor
In
separate category within the spectrum of pul¬
monary NE tumors, its place in this spectrum has
been poorly understood. At one end are the typical
carcinoids, followed by the atypical carcinoids, at the
other end, SCLC and LCNEC. Typical carcinoids
NE tumors with a low-grade malignancy, while
SCLC and LCNEC are characterised by high met¬
astatic potential and (initial) response to chemother¬
are
apy.
The fifth category in this spectrum is the NSCLCNE, comprising 10 to 15% of all NSCLC. In
NSCLC-NE, to be distinguished from LCNEC, NE
differentiation can be demonstrated only by immuor electron microscopy, as light
nohistochemistry
does
not show any evidence for NE
microscopy
differentiation.
It has been suggested that the presence of NE
markers in NSCLC is associated with an increased
chemosensitivity.14
Given the results of our
present study, we have no
indication of NSCLC-NE sharing features with
SCLC in the spectrum of NE tumors. There does
not seem to be a higher propensity to metastasize in
NSCLC-NE when compared with NSCLC without
NE differentiation. In conclusion, 123C3 has no
predictive value in NSCLC. In addition, tumor
to 123C3
histology or TNM stage is not correlated
of
MOC-1
did not
Furthermore,
reactivity.
staining
contribute to prognosis.
In other words, NE differentiation in NSCLC
(NSCLC-NE), demonstrated by NCAM expression,
does not seem to significantly alter prognosis. In
clinical practice however, this does not imply that for
patients with NSCLC-NE the same treatment as for
NSCLC is
justified.
An
important question,
not
addressed by our present study, remains whether
there is more chemosensitivity in NSCLC-NE than
in NSCLC without NE differentiation. If so, chemo¬
in patients with NCAM expression in
therapy
NSCLC might improve prognosis. Further research
is needed to see if NE markers facilitate selection of
patients with metastases for adjuvant therapy after
resection, by predicting response to chemotherapy
or
radiotherapy.
stage.
NSCLC, NE differentiation has been shown to
be associated with a higher nodal TNM stage,4 and a
worse prognosis.2412 Sundaresan et al,13 however,
could not show a correlation between NE differen¬
tiation and survival. Graziano et al14 studied four NE
markers (neuron-specific enolase, chromogranin A,
Leu-7, and synaptophysin) in patients with resected
NSCLC stage I and II. In this study, NE markers did
not
Since Travis et al16 in 1991 described NSCLC-NE
as a
predict prognosis.15
References
1 Schol
DJ,
Mooi
WJ, van der Gugten AA, et al.
Monoclonal
antibody 123C3, identifying small cell carcinoma phenotype
in lung tumours, recognizes mainly, but not exclusively,
endocrine and neuron-supporting normal tissues. Int J Can¬
cer
1988; 2:34-40
RE, Moolenaar KEC, Michalides RJAM,
2 Kibbelaar
et al.
Neural cell adhesion molecule expression, neuroendocrine
differentiation and prognosis in lung carcinoma. Eur J Cancer
1991; 27:431-435
CHEST/ 114/5 /NOVEMBER, 1998
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21874/ on 06/15/2017
1319
3 Mooi WJ, Wagenaar SS, Schol D, et al. Monoclonal antibody
123C3 in lung tumour classification: immunohistology of 358
resected lung tumours. Mol Cell Probes 1998; 2:31-37
4 Pujol JL, Simony J, Demoly P, et al. Neural cell adhesion
neuroendocrine differentiation features. J Clin Oncol 1989;
prognosis of surgically resected lung cancer.
Am Rev Respir Dis 1993; 148:1071-1075
Kwa HB, Michalides RJAM, Dijkman JH, et al. The prog¬
7:1614-1620
13 Sundaresan V, Reeve
non-small cell lung cancer.
tumours. Br J Cancer 1991; 64:333-338
14 Graziano SL, Tatum AH, Newman NB, et al. The prognostic
molecule and
5
procedures. J R Statist Soc (Series A) 1972; 135:185-198
HH, de Leij L, Poppema S, et al. Clinical
characterization of non-small cell lung cancer tumors showing
12 Berendsen
p53 and cyclin DI in resected
Lung Cancer 1996; 14:207-217
6 Kwa HB, Verheijen GMM, Litvinov SV, et al. Prognostic
factors in resected non-small cell lung cancer: an immuno¬
histochemical study of 39 cases. Lung Cancer 1996; 16:34-45
7 The World Health Organization histological typing of lung
tumours. 2nd ed. Am J Clin Pathol 1982; 77:13-136
8 Hirsch FR, Matthews MJ, Aisner S, et al. Histopathologic
classification of small cell lung cancer: changing concepts and
terminology. Cancer 1988; 62:973-977
9 Mountain CF. A new international staging system for lung
cancer. Chest 1986; 89:225S-233S
10 Kaplan EL, Meier P. Non-parametric estimation from incom¬
plete observations. J Am Stat Assoc 1958; 53:457-481
11 Peto R, Peto J. Asymptotically efficient rank invariant test
nostic value of
NCAM,
JG, Stenning S, et al. Neuroendocrine
differentiation and clinical behaviour in non-small cell lung
significance of neuroendocrine
markers and
carcinoembry-
onic antigen in patients with resected stage I and II non-small
cell lung cancer. Cancer Res 1994; 54:2908-2913
15 Graziano SL, Mazid R, Newman N, et al. The use of
neuroendocrine immunoperoxidase markers to predict che¬
motherapy response in patients with non-small-cell lung
cancer. J Clin Oncol 1989; 7:1398-1406
16 Travis WD, Linnoila RI, Tsokos MG, et al. Neuroendocrine
tumors of the lung with proposed criteria for large-cell
neuroendocrine carcinoma: an ultrastructural, immunohisto¬
chemical and flow cytometric study of 35 cases. Am J Surg
Pathol 1991; 15:529-553
1320
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21874/ on 06/15/2017
Clinical
Investigations