Download Serum Fragment of Cytokeratin Subunit 19

(CANCER RESEARCH 53. 61-66. January 1. 19931
Serum Fragment of Cytokeratin Subunit 19 Measured by CYFRA 21-1
Immunoradiometric Assay as a Marker of Lung Cancer
Jean-Louis Pujol,1, Jean Grenier, Jean-Pierre Daurès,Alain Daver, Henri Pujol, and François-Bernard Michel
Clinique lÃ-esMaladies Respiratoires. Universitétie Mtmtpellier. Hôpital Aiguelonfìue.3405V Montpellier Cedex ¡J-L P.. F-B. M.]; Centre Régionalde Lune contre le Cancer,
Laboratoire de Raditi-Anal\ses, Ì4W4Montpellier Cedex ¡J.G.. H. P.¡:Départementde l'Information Médicale.Hôpital Gaston Doumergue. 30000 Nîmes¡J-P.D./: and Centre
Paul Papin. 2. Rue Moll. 4W36 Angers ¡A.D.I. France
and prognosis. The three main prognostic factors able to help treat
ment decision are histology (mainly SCLC versus NSCLC), stage of
disease, and performance status (8, 9). The search for new clinical or
biological variables as putative independent prognostic factors may
potentially improve the management of lung cancer patients.
Cytokeratins are intermediate filaments which are part of the cy-
ABSTRACT
Cytokeratin 19 is a subunit of cytokeratin intermediate filament ex
pressed in simple epithelia and their malignant counterparts. Therefore, it
is expressed by respiratory epithelium cells and has been detected in lung
cancer specimens. An immunoradiometric
assay was used to detect a
fragment of the cytokeratin 19, referred to as CYFRA 21-1, in the serum
of 165 patients with histologically proved lung cancer ( 128 non-small cell
toskeleton of both normal epithelia and their malignant counterparts
(10). Immunohistochemical studies using broad-spectrum cytokeratin
and 37 small cell lung cancers). This prospective study was conducted to
evaluate the reliability of this immunoradiometric assay and to identify
the relationship between serum CYFRA 21-1 and different features of lung
antibodies have demonstrated the expression of cytokeratins in both
SCLC and NSCLC suggesting a common endodermal lineage
(11, 12). In contrast to other intermediate filaments, cytokeratins are
a family of polypeptides, some of them having cell type specificity
(10). Cytokeratin 19 is an acidic (type I) subunit expressed in all
simple epithelia and in carcinomas such as lung cancer which arise
from them.
A fragment of cytokeratin subunit 19 can be measured in serum by
a new immunoradiometric assay, the CYFRA 21-1 using two mouse
MoAb KS 19-1 and BM 19-21. Thus, this cytokeratin 19 fragment is
referred to as CYFRA 21-1. The aims of the present prospective study
are (a) to evaluate the reliability of serum CYFRA 21-1, (b) to identify
cancer including prognosis. The minimal detectable concentration de
tected by this assay was 0.06 ng/ml. The reliability of the immunoradio
metric assay was demonstrated by the linear relationship between CYFRA
21-1 measurement and dilution of the serum, the reproducibility of the
dosage in intraassay and interassay, and the high sensitivity of the method
in discriminating low CYFRA 21-1 concentrations. Using a threshold of
3.6 ng/ml, sensitivity and specificity were 0.52 and 0.87, respectively. The
sensitivity of the marker was highest in squamous cell carcinoma and
lowest in small cell carcinoma. In non-small cell lung cancer patients, the
marker varied significantly according to both stage of the disease
(Kruskal-Wallis,
13.7; P < 0.005) and performance status (KruskalWallis, 9.16; P < 0.05) inasmuch as a high serum CYFRA 21-1 level was
any relationship between the marker and other clinical and biological
parameters in both SCLC and NSCLC and (c) to analyze the prog
nostic significance of a high CYFRA 21-1 level at time of diagnosis.
associated with advanced stages, mediastinal lymph nodes, and poor per
formance status. Consequently, the marker was significantly lower in
patients who were operated upon when compared with unresectable ones.
Lung cancer patients with serum CYFRA 21-1 over 3.6 ng/ml proved to
have a significantly shorter overall survival than those with a normal
serum level (log rank, P = 0.007; Wilcoxon, /' - 0.001). The negative
MATERIALS
prognostic effect of CYFRA 21-1 was highly significant in squamous cell
carcinomas whereas it was nonsignificant for the other histologies. In
Cox's model analysis, performance status, stage grouping, and CYFRA
21-1 were the only significant determinants
AND METHODS
Patients. One hundred sixty-five consecutive lung cancer patients referred
to the Montpellier University Hospital between February 1990 and February
1992 were prospectively entered in the study (Table 1). All patients had
pathologically confirmed lung cancer. Among them were 37 SCLC. 85 SQC,
28 adenocarcinomas, and 15 large cell carcinomas as defined by the WHO
classification (13). Performance status was estimated according to the Eastern
Cooperative Oncology Group and the percentage of weight loss during the
previous 4 months was recorded. Staging was carried out by exhaustive pro
cedure according to the 4th edition of the Union Internationale Contre le
Cancer tumor-nodes-metastasis classification (14) and the American Thoracic
of survival. This study sup
ports the use of the serum fragment of cytokeratin subunit 19 CYFRA
21-1 as an independent prognostic marker of squamous cell carcinoma of
the lung.
INTRODUCTION
Treatment of lung cancer is probably one of the great challenges of
medical oncology owing to an increasing incidence in both men and
women and poor prognosis (1, 2). SCLC2 have neuroendocrine bio
Society map of regional pulmonary nodes (15). Staging of NSCLC was es
tablished according to Mountain's stage grouping (16); for SCLC, limited
disease was defined as disease confined to one hemithorax including medias
tinal lymph nodes and/or ipsilateral supraclavicular lymph nodes; extensive
disease was defined as a disease more advanced than limited SCLC. For all
patients staging procedure included clinical examination, standard chest roent-
logical properties usually associated with a high sensitivity to chemo
therapy (2-^). Surgery offers the best probability of cure for NSCLC
but it cannot be proposed for locally advanced and metastatic stages;
therefore, other modality treatments such as radiotherapy-chemother
apy association have been proposed (5). Because metastatic disease is
the first cause of failure for both SCLC (6) and NSCLC (7), research
for new treatment modalities is an important goal. Such an experi
mental approach requires an accurate definition of groups of patients
genography. computed tomographic scan of chest and upper abdomen, fiberop
tic bronchoscopy, liver sonography. and bone scanning. Brain computed to
mographic scan was done systematically for SCLC and if clinically required
for NSCLC.
Controls. The serum CYFRA 21-1 level was measured in 104 patients with
nonmalignant pulmonary diseases (20 infectious diseases. 75 chronic obstruc
tive pulmonary diseases, and 9 miscellaneous).
Treatment Decision. Each patient was discussed by a medical panel com
posed of thoracic surgeons, chest physicians, radiologists, and medical oncol
ogists. NSCLC patients with stage I or II disease or with moderate locally
advanced NSCLC (Ilia with resectable node metastasis) underwent surgery in
an attempt to achieve complete resection. Other NSCLC patients with perfor
mance status <2 and distant metastasis (stage IV) or gross mediastinal involve
ment (stage Illb and stage Ilia with more than 2 ipsilateral mediastinal lymph
Received 7/2/92: accepted 10/20/92.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely (o indicate this fact.
1To whom requests for reprints should he addressed.
2 The abbreviations used are: SCLC. small cell lung carcinoma: NSCLC. non-small
cell lung carcinoma: MoAb. monoclonal antibodies: SQC. squamous cell carcinomas;
MDC. minima] detectable concentration; ROC, receiver operating characteristic; St. cy
tokeratin standard; LDH, láclatedehydrogenase; NX. nodal stage \.
61
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
SERUM
CYFRA
21-1 IN LUNG CANCER
Table 1 Patients' characteristics
165
146/19
62 (10,31-88)
No.
M/F
Mean age (SD, range)
128
NSCLC
Performance stalus
0-1
2-3
4(1
Slage grouping
l und II
Illa
Illb
IV
17
27
35
49
Wt loss (%)
Absence
Presence
73
55
SCLC
37
Disease extent
Limited disease
Extensive disease
21
In
Wt loss (%)
0
>5
21
16
Performance status
0-1
2-3
28
9
node métastases)were eligible for chemotherapy. Best supportive care, includ
ing palliative radiation therapy when needed, was given to patients with ad
vanced stage and poor performance status. SCLC patients underwent chemo
therapy [usually an alternating regimen containing cisplatin (17)]. Treatment
was decided upon according to routine clinical and biological findings and
without knowledge of the serum CYFRA 21-1 level. Hence, treatment was not
considered as a prognostic variable in this study.
CYFRA 21-1 Immunoradiometric Assay. A blood sample was taken from
each patient at presentation, and Ihe serum was separated and stored at -190°C
until tested.
CYFRA 21-1 (Centocor Diagnostics. Malvern, PA: Cis Biointernational.
Gif/Yvette, France) is a solid-phase immunoradiometric assay based on the
two-site sandwich method. In this method the cytokeratin 19 was recognized
by two mouse MoAb. KS 19-1 and BM 19-21. directed against two different
epitopes of a fragment of cytokeratin subunit 19. which is referred to as serum
CYFRA 21-1. These MoAb were obtained by immunization against the MCF7
cell line.
MoAb KS 19-1 -coated polystyrene spheres were incubated with 200 ul of
patient serum, control serum, or standard curve (composed of the following
concentrations of cytokeratin 19: 0, 3, 8, 25, and, 50 ng/ml). for 20 h at 2-8°C.
Afterwards, the solid phase was washed with distilled water and then incubated
with 0.85 uCi/ml of '"I-labeled BM 19-21 for 3 h at 2-8°C. Finally, the solid
phase was washed again with distilled water in order to cancel the nonfixed
labeled reagents.
Radioactivity was counted in a well-type gamma counter (Auto-Gamma:
Packard Instrument Company. Downers Grove. IL) and expressed in cpm. The
calculated concentration of cytokeratin 19 was expressed ¡nng/ml.
Control of the Immunoradiometric Assay. The sensitivity of the method
was assessed by determining the MDC. This method used two cytokeratin 19
standards containing, respectively. 0 (cytokeratin free. StO) and 3 ng/ml (Stl):
these 2 standards were measured in 6 different assays (Table 2): in each of the
6 assays, the same total radioactivity was introduced for both standards. The
bound/total radioactivity was calculated for StO and Stl. The MDC in ng/ml
was determined as:
X
The control of the reliability of the method of measurement was done using
the following tests: (a) test of linearity of the relationship between consecutive
dilutions of a serum of known high cytokeratin 19 level and results of the
titration; (b) assessment of intraassay reproducibility by measuring 7 different
sera twice during the same assay: (c) assessment of interassay reproducibility
by measuring 2 different sera in 10 different assays: (a1)analysis of the standard
curve in the low values in order to evaluate the difference of signal given by
the StO and the St l.
Total LDH assays were done following the Deusch Chemical Society
recommendations by measuring its activity using pyruvate as a substrate
(Bio-Merieux, France).
The upper limits of normal values were as follows: LDH, 330 units/liter:
alkaline phosphatase. 220 units/liter; leukocytes 8000/ul. The lower limits of
normal values were 32 g/liter for albumin and 135 mmol/liter for serum
sodium.
All sera samples were assayed blind of clinical information.
Statistics. ROC curves were constructed using both patient and control
subject serum CYFRA 21-1 levels ¡nan attempt to establish a sensitivityspecificity relationship; areas under the ROC curve were calculated (18). The
serum tumor marker was not distributed normally: thus, for each patient subset,
results were expressed as median, and variation was expressed as interquartile
range. Nonparametric statistical analyses were used: differences between two
independent groups were determined by means of the Mann-Whitney U test;
differences between more than two groups were determined by means of the
Kruskal-Wallis one-way analysis of variance: P < 0.05 was considered as
significant. Survival was defined as the time from the date of sampling to the
date of death. Survival data were updated in February 1992 and none of the
patients were lost from sight during follow-up. Probability of survival was
estimated by the method of Kaplan and Meier (19). Single variable survival
analyses were done by means of Wilcoxon and log rank tests and multivariate
regression was done with the Cox's model (20). There were 12 variables in the
model. Serum CYFRA 21-1 was examined as normal (<3.6 ng/ml) or elevated
(>3.6 ng/ml) as were the other biological variables (LDH, serum sodium,
alkaline phosphatase. albumin, and leukocytes) according to their respective
upper normal limits. Age groups were defined as less than or equal to and over
50 years old. Other variables were stage of the disease, sex, histology, weight
loss, and performance status. Survival was analyzed using the SAS software
package.
RESULTS
Sensitivity and Reliability of the Immunoradiometric Assay.
According to the data in Table 2 the calculated MDC was 0.06 ng/ml.
The test of linearity showed a linear relationship between the titrations of CYFRA 21-1 and the dilutions of a serum sample with a
high cytokeratin 19 concentration (Fig. 1). Seven sera were measured
twice during the same assay (intraassay). The results of the first and
second measurements were, respectively (ng/ml): 12, 11.4; 27, 30; 6,
7.5; 3, 2; 4.4, 4.8; 45, 41 ; 3.4, 3.2. The coefficients of variation of 2
sera tested in 10 different assays (interassays) were 7 and 9.8%,
respectively.
Analysis of the standard curve in the low values demonstrated that
the difference between 0 and 3 ng/ml was 700 cpm, a value sufficient
to prevent overlap of results.
Tumor Marker Distribution. The median and interquartile range
interquartile range, of serum CYFRA 21-1 was significantly higher in
Table 2 Results of 6 different inuminoradiotnetric assays performed in order
to determine the minimal detectable concentratitm
StO
Stl
Assay123456T" (cpm)226.473219.399208.757241.441207.643198.408cpm358976718364B/T0.020.040.040.030.0
x
where A is MDC in ng/ml: X is [standard Stl] - [standard StOJ = 3 ng/ml: Y is
mean Stl (cpm) - mean StO (cpm): and y = 2 SD of StO (cpm).
" B, bound radioactivity; T, total radioactivity.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
SERUM CYFRA 21-1 IN LUNG CANCER
>SO
level differed when the histológica! type was considered (KruskalWallis test KW = 7.69; P < 0.05). Serum CYFRA 21-1 level was
y •¿
»8.78 X - 0.17
•¿
0.998; p •¿
0.002
significantly higher in SQC when compared with each of the other
histologies (Mann-Whitney U test, P < 0.05).
CYFRA 21-1 and Clinical Variables at Presentation. In NSCLC,
the CYFRA 21-1 level, was significantly higher in patients with a
metastatic disease (7.4; interquartile range, 3.0-21.0 ng/ml) in com
parison with the limited disease patient subgroup (3.8; interquartile
range. 1.9-8.5 ng/ml; Mann-Whitney U test, P < 0.001). When the
marker was analyzed according to stage grouping, a significant in
crease of the serum level was observed from stage I to stage IV
(Kruskal Wallis. 13.7; P < 0.005; Fig. 4) but there was no significant
difference when stage Ilia and stage Illb were compared (MannWhitney U test, 500; nonsignificant). CYFRA 21-1 levels were lower
1:16 1:32
1:64
Serum dilutions
Fig. l. Dilution test of a serum with a high level of cytokeratin I1).
in patients who were operated upon when compared with inoperable
patients (2.4; interquartile range. 1.25-5.0 versus 5.8; interquartile
range, 2.7-13.0 ng/ml; Mann-Whitney U test, P < 0.001). In addition.
Table 3 Sensitivity and specificity of CYFRA 21-1 using a 3.ft-n\>/ml upper limit
of normal mines
1.0
0.76
Subsets
Sensitivity
Specificity
Non-small cell
SQC
Small cell
All
0.56
0.63
0.46
0.52
0.89
0.91
0.89
0.87
knuka
Willi!
p
no'.
0.6
g 0.28
3.6
ADE
SQC
LCC
SCLC
Fig. 3. CYFRA 21-1 distribution according to histology.
. median value: columns.
interquartile range. ADE. adenocarcinoma: LCC. large cell carcinoma.
0.26
False-Positive
0.6
0.76
1.0
Rate (1 - Specificity)
Fig. 2. Receiver operating eharaeteristie eurve for serum CYFRA 2I-1. O. squanmus
cell carcinoma, area under Ihe concentration curve = 0.83 ±0.03 (SE); •¿
non-small cell
lung cancer, area under the concentration curve = 0.80 ±0.03.
cancer patients (4.3; interquartile range, 2.3-9.5) when compared with
control subjects (1.05; interquartile range, 0.8-1.8; Mann-Whitney
U test, P < 0.001). Areas under ROC curves were 0.83 ±0.03, 0.80
±0.03, and 0.74 ±0.05 for the SQC, NSCLC. and SCLC groups,
respectively (Fig. 2). Using ROC curves, a threshold of 3.6 ng/ml
was chosen as the upper limit of normal values. The sensitivity and
specificity of this cutoff in the different patient subsets are shown in
Table 3.
CYFRA 21-1 and Histology. The median (interquartile range)
serum CYFRA 21-1 for SCLC, SQC. adenocarcinomas, and large cell
carcinomas were 3.4 (1.8-7.0), 5.6 (2.6-13.0), 2.9 (1.85-7.10), and
3.8 (1.8-7.0) ng/ml, respectively (Fig. 3). The serum CYFRA 21-1
24222018161412108642n-Krutk.l
.0.006-----nn
W.lli.
p
.-3.6
TTq_ji tÃ-—
I & II
lila
1Mb
IV
Fig. 4. CYFRA 21-1 distribution according to stage grouping in non-small eel lung
cancer.
, median value; columns, interquartile range.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
SERUM CYFRA 21-1 IN LUNG
CANCER
Table 5 Significant prognostic factors in the entire population
patients who presented mediastinal lymph nodes (N2 or N3) demon
strated higher serum CYFRA 21-1 levels (5.0; interquartile range,
2.6-12.5 ng/ml), when compared with patients without mediastinal
lymph nodes (NO or Nl, 3.2; interquartile range, 1.75-8.0 ng/ml;
Mann-Whitney U test, P < 0.05).
In NSCLC, the distribution of CYFRA 21-1 varied significantly
according to the performance status of NSCLC patients (KruskalWallis 9.16; P < 0.05; Table 4). In SCLC, CYFRA 21-1 did not
significantly differ between limited (3.0; interquartile range, 1.8-5.4
ng/ml) and extensive disease (4.55; interquartile range, 2.1-11.5 ng/
ml; Mann-Whitney U test, P = 0.07).
Survival. Univariate analysis showed that lung cancer patients
with serum CYFRA 21-1 over 3.6 ng/ml proved to have a significantly
shorter overall survival than those with a normal serum level (log
rank, P = 0.007; Wilcoxon. P = 0.001 ; Fig. 5). The negative prognostic
effect of CYFRA 21-1 was highly significant in SQC (log rank, P <
0.001; Wilcoxon, P < 0.001) whereas it was nonsignificant for the
other histologies. Separate survival analyses showed a significant
effect of stage group, weight loss, performance status, serum albumin,
LDH, and CYFRA 21-1 (Table 5). With Cox's model analysis, per
(dayslStageI
Factor and level
Median surviva 1Wilcoxon
rank0.0001
IIIliaHlbIVPerformance
and
Log
0.00010.0001
status9-12-4Wt
0.00010.0018
lossNo
loss0-5%>5%Albumin
wt
0.00150.0049
(g/liter)>32<32Láclate
0.0370.0041
(units/liter)<330>330CYFRA
dehydrogenase
0.01770.001
formance status, stage grouping, and CYFRA 21-1 were the only
significant determinants of survival (Table 6).
(ng/ml)<3.6>3.6420:?4155407148402159191367191398252402213f
21-1
0.007
DISCUSSION
Table 6 Regression analvsis results: estimated relative risk for significant variables
in = 165)
New strategies of chemotherapy intensification are under evalua
tion in SCLC patients in an attempt to circumvent secondary chemoresistance (3, 4). NSCLC is a group of different histologies having a
lower chemosensitivity when compared with SCLC. Surgery of stage
I and II NSCLC offers the best probability of cure (5, 21) but in the
more advanced stages combined modality treatments are required
Table 4 CYFRA 21-! distribution according to performance
in non-small cell lung cancers
78
319
Log-rink p -0.007
Wilcoxon p -0.001
£
O.B
t
"o
0.6
i
0.4
Z
5
s
0.2
400
600
800
days
CYFRA <•
3.6 nQ/mL
CYFRA > 3.6 ng/mL
median «urvlval
<•
3.6 ng/HL
> 3.6 ng/mL
% c«n«ortd
•¿0
40«
SS
86
fit
34
index
Stage of disease
CYFRA 21-1Regression
±0.22
0.90 ±0.29
0.49 ±0.21P<0.0001
risk2.81
0.0024
0.021Relative
2.47
1.64
Thus, the treatment of both SCLC and NSCLC lung cancer is still an
experimental approach. In this setting, an awareness of significant
variables able to predict metastatic spread and poor prognosis is
recommended (25).
Some important prognostic factors of lung cancer have already been
reported. For NSCLC, the two main variables which determine sur
vival are the stage of the disease and the performance status (8). In
several studies, tumor markers such as tissue polypeptide antigen
(26, 27), LDH (8), or carcinoembryonic antigen (26, 27) have been
demonstrated as significant prognostic factors in univariate analyses.
Other biological properties of NSCLC have been investigated as pu
tative indicators of prognosis. In different multivariate studies aneuploidy (28), lack of expression of blood group A antigen (29) or
expression of the neural cell adhesion molecule (30) have been inde
pendently demonstrated as prognostic factors; however, correlation
and interactions between these variables are not known. For SCLC,
several studies have demonstrated the univariate influence of disease
extent, performance status, LDH, and alkaline phosphatase on sur
vival (9, 31-33); moreover, some authors have suggested that the
poorest prognosis is seen in SCLC patients with bone marrow métas
tases (34), elevated serum neuron specific enolase (35), or carcino
embryonic antigen (26).
Intermediate filaments are part of the cytoskeleton which provide
important information on the cell origin (36). They are classified into
five tissue specific protein families. The cytokeratin family is ex
pressed by all epithelial cells (10) including endocrine cells of the
dispersed neuroendocrine system (37) and, it appears, therefore, to be
a general specific and useful marker of the epithelial differentiation.
Interestingly, the expression of cytokeratins is not lost by epithelial
cells during malignant transformation (10, 36), a phenomenon which
(ng/ml)1
.3-5.0
2.2-9.5
2.7-16.0
4.6-84.0n10
coefficient
(ß±SE)1.03
owing to the frequency of microscopic metastatic disease (7, 22-24).
status (PS)
range
PSO
PS1
4.36.0
PS2
10.0Interquartile
PS3Median2.7"
1Kruskal-Wallis = 9.16; P < 0.05.
VariablePerformance
Fig. 5. Probability of survival of all patients with normal and elevated pretreatment
serum CYFRA 21-1 level.
64
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
SI-RUM CYI-RA 21-1 IN l.UNG
contrasts with the well-known phenotypic instability of cancer cells
(38). Antibodies raised against some common antigenic determinants
of all cytokeratins are therefore useful in typing malignant tumors
with poor histologie-aïfeatures of differentiation (36). Immunohistochemical studies using these broad-spectrum cytokeratin antibodies
have clearly demonstrated that all four histológica! types of lung
cancer expressed cytokeratins (II, 12, 39); in particular, the classic
SCLC type which have been shown to coexpress both cytokeratin and
neurofilament whereas the variant type only express cytokeratin (40).
This supports the hypothesis of a common epithelial lineage of all
histológica! types of lung cancer. A cytokeratin is a heterotypic tetratner of protofiluments composed of two polypeptidcs: one acidic
type I subunit; and one basic type II subunit. Each type of epithelia
and their malignant counterpart express a specific cytokeratin
polypeptide pattern (10, 41); simple epithelia, including pseudostratified epithelia. express cytokeratins 7, 8. 18, and 19. Selective anti
bodies raised against simple epithelium type of cytokeratin have been
shown to react with all histologies of lung cancers (42, 43). Although
cytokeratins are part of the cytoskeleton some fragments might be
released in the serum owing to cell lysis or tumor necrosis, which
gives support to the present research on cytokeratin subunit 19 frag
ment as a tumor marker of lung cancer.
The immunoradiometric assay described in this study is able to
detect the serum fragment of cytokeratin subunit 19, referred to as
CYFRA 21-1, with both sensitivity and reliability. The threshold of
3.6 ng/ml was determined using the ROC curve analysis which
showed that this threshold allowed the best sensitivity-specificity
relationship. Our study has demonstrated a correlation between serum
CYFRA 21-1 and stage of the disease in NSCLC but not in SCLC.
Moreover, in NSCLC, the marker varied significantly according to
nodal status. These data suggest that CYFRA 21-1 might be consid
ered as a marker of tumor mass. Thus, patients who present a high
level of serum CYFRA 21-1 might require a careful search of métas
tases. Despite these results it is not possible to predict operability of
NSCLC using the serum level of the marker owing to the important
overlap of the distribution in stage Ilia (usually marginally resectable)
and stage Illb- (unresectable) NSCLC groups. The reason for this
result might be that the distinction between stage Ilia and stage Illb
mainly depends on anatomic extension rather than tumor mass. The
low incidence of a high serum CYFRA 21-1 level in the stage I-II
group clearly indicates that the marker is of no help in screening or
early diagnosis of the disease. Because CYFRA 21-1 is correlated
with both disease extent and performance status it is not surprising
that univariate analysis showed that patients with a high serum level
proved to have a significantly shorter survival. The negative effect of
CYFRA 21-1 on survival was particularly significant in SQC whereas
this effect was not demonstrated in the other histologies. Interestingly.
Cox's model demonstrated that the prognostic significance brought by
CYFRA 21-1 is independent and adds information to the well-known
and above-mentioned prognosis factors. This study supports the use of
the serum fragment of cytokeratin subunit 19 CYFRA 21-1 as an
independent prognostic marker of squamous cell carcinoma of the
lung.
3. Klastersky. J. Commentary: intensive chemotherapy in small cell lung cancer. Eur. J.
Cancer Clin. Oncol., 25: 933-937, 1989.
4. Hansen. H. H., and Kristjansen. P. E. G. Chemotherapy of small cell lung cancer. Eur.
J. Cancer. 27: 342-349, 1991.
5. Mulshine, J. L.. Glatstein, E.. and Ruckdeschel. J. C. Treatment of non-small cell lung
cancer. J. Clin. Oncol., 4: 1704-1715. 1986.
6. lannuzzi. M. C.. and Scoggin. C. H. State of the art: small cell lung cancer. Am. Rev.
Rcspir. Dis.. 134: 593-608. 1986.
7. Cox. J. D.. and Yesner, R. A. Causes of treatment failure and death in carcinoma of
the lung. Yale J. Biol. Med.. 54: 201-207. 1981.
8. O'Connel. J. P.. Kris. M. G.. Gralla. R. J.. Groshen, S., Trust, A., Fiore. J. J.. Kelsen.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
ACKNOWLEDGMENTS
31.
The authors wish to thank Jo Baissus for help in preparing Ihe manuscript.
REFERENCES
32.
1. Stjemswurd. J.. and Stanley, K. Etiology, epidemiology and prevention of lung cancer.
Lung Cancer, 4: 11-24. 1988.
2. Hansen. H. H. Lung cancer. In: H. M. Pinedo, D. L. Longo, and B. A. Chabner (eds.).
Cancer Chemotherapy and Biological Response Modifiers, pp. 222-240. Annual 10.
Amsterdam: Elsevier/North-Holland BiomédicalPress, 1988.
CANCI-R
33.
D. P., Heelan, R. T., and Golhey. R. B. Frequency and prognostic importance of
pretrcatment clinical characteristics in patients with advanced non-small cell lung
cancer treated with combination chemotherapy. J. Clin. Oncol.. 4: 1604-1614. 1986.
Osterlind. K.. [hde, D. C.. Ettinger. D. S.. Gralla. R. J.. Karrer. K.. Krauss. S., Maurer.
L. H.. Rorth. M-, Sorenson. S., and Vincent. R. Staging and prognostic factors in small
cell carcinoma of the lung. Cancer Treat. Rep.. 67: 3-9, 1983.
Moll, R.. Franke. W. W.. Schiller. D. L.. Geiber. B., and Krepier, R. The catalog of
human cytokeratins: patterns of expression in normal epithelia. tumors and cultured
cells. Cell, 31: 11-24, 1982.
Broers, J. L. V.. Rot, M. K., Oostendorp. T., Huysmans, A., Wagenaar. S. S., Wiersmavan Tilburg. A. J. M.. Vooijs. P, and Ramaekers, F. C. S. Immunohistochemical
detection of human lung heterogeneity using antibodies to epithelial, neuronal and
neuroendocrine antigens. Cancer Res., 47: 3225-3234, 1987.
Pujol, J. L.. Simony, J., Laurent. J. C., Richer, G., Mary. H.. Bousquet, J.. Godard. Ph..
and Michel, F. B. Phenotypic heterogeneity studied by immunohistochemistry and
aneuploidy in non-small cell lung cancers. Cancer Res., 4V: 2797-2802, 1989.
World Health Organization. The World Health Organization Histológica! Typing of
the Lung Tumors. Ed. 2. Am. J. Clin. Pathol., 77: 123-136. 1982.
Sobin. L. H.. Hermanek. P.. and Hutter. R. V. P. TNM classification of malignant
tumours. Ed. 4. Geneva: International Union Against Cancer. 1987.
Tisi. G. M.. Friedman, P. J.. Peters. R. M.. Pearson. G.. Carr. D.. Lee, R. E.. and
Selawry, O. American Thoracic Society: clinical staging of primary lung cancer. Am.
Rev. Respir. Dis., 125: 659-64. 1982.
Mountain. C. F. A new international staging system for lung cancer. Chest, 4(s):
225s-233s, 1986.
Pujol. J. L., Demoly, P., Gaulier. V.. Romieu. G.. Dan Aouta. M., Stenger. R.. Parrai.
IL, Tadlaoui, O., Marcenac, A. M., Benahmed. S. S.. Michel, F. B.. and Godard. P.
Phase II study of alternating combination chemotherapy in small cell lung cancer.
Lung Cancer, 7: 285-294, 1991.
Beck, J. R.. and Shultz. E. K. The use of relative operating characteristic (ROC)
curves in test performance evaluation. Arch. Pathol. Lab. Med.. Ill): 13-20, 1986.
Kaplan. E. L.. and Meier. P. Nonparametric estimation from incomplete observations.
J. Am. Slat. Assoc., 53: 457^81. 1958.
Cox. D. R. Regression models and life tables. J. R. Statist. Soc. B.. 34: 187-202,
1972.
Gail. M. H., Eagan, R. T, Feld, R.. Ginsberg, R.. Goodell, B., Hill, L.. Holmes, E. C,
Lukeman. J. M.. Mountain. C. F.. Oldhani, R. K., Pearson, F. G., Wright, P. W., and
Lake. W. H. Prognostic factors in patients with resected stage I non-small cell lung
cancer. Cancer (Phila.), 54: 1802-1813. 1984.
Pearson, F. G.. DeLarue, N. C.. lives, R.. Todd. T. R. J., and Oxiper. J. D. Significance
(if positive superior mediastinal nodes identified at mediastinoscopy in patients with
resectable cancer of the lung. J. Thorac. Cardiovasc. Surg., 83: 1-11, 1982.
Pujol, J. L.. Rossi, J. F.. Le Chevalier, T.. Daurès.J. P.. Rouanet. P.. Douillard, J. Y.,
Dubois. J. B.. Amagada. R.. Mary, H.. Godard, P., and Michel. F. B. Phase 11 pilot
study of neoadjvam ifosfamide, cisplatin. and etoposide in locally advanced nonsmall cell lung cancer. Eur. J. Cancer. 26: 798-801, 1990.
Tubiana. M. The growth and progression of human tumors: implications for man
agement strategy. Radiothcr. Oncol.. 6: 167-184. 1986.
World Health Organization. WHO Handbook for Reporting the Results of Cancer
Treatment, WHO Offset Publication No. 48. Geneva: WHO. 1979.
Buccheri. G. F. Violante, B., Sartoris, A. M., Ferrigno, D., Curcio, A., and Vola. F.
Clinical value of a multiple biomarker assay in patients with hronchogcnic carcinoma.
Cancer (Phila.), 57: 2389-23%. 1986.
Salvati. F.. Cruciani, A. R., Flore, F.. De Angelis. G.. Pigorini. F.. Antilli. A.. Pau, F..
Munno, R., and Cipri, A. Plasma carcinocmhryonic antigen and tissue polypeptide
antigen levels in lung cancer: correlation with cell types and stage. Cancer. Detect.
Prev., 8: 111-114, 1985.
Zimmerman, P. V. Hawson, G. A. T.. Bint. M. H.. and Parsons. P. G. Ploidy as a
prognostic determinant in surgically treated lung cancer. Lancet, 2: 530-533, 1987.
Lee. J. S., Ro. J. Y., Sahin. A. A.. Hong. W. K.. Brown. B. W.. Mountain, C. F. and
Hittelman. W. N. Expression of blood-group antigen A: a favorable prognostic factor
in non-small cell lung cancer. N. Engl. J. Med.. 324: 1084-1090. 1991.
Bcrendsen, H. H.. De Leij. L.. Poppema. S.. Postmus. P. E.. Boes. A.. Sluiter. H. J..
and The. H. Clinical characterization of non-small-ccll lung cancer and tumors show
ing neuroendocrine differentiation features. J. Clin. Oncol.. 7: 1614-1620. 1989.
Rawson, N. S., and Peto. J. An overview of prognostic factors in small cell lung
cancer. A report from the subcommittee for the management of lung cancer of the
united Kingdom co-ordinating committee on cancer research. Br. J. Cancer, 31:
597-604, 1990.
Osterlind. K.. and Andersen. P. K. Prognostic factors in small cell lung cancer:
multivariate rmxJel based on 778 patients treated with chemotherapy with or without
irradiation. Cancer Res., 46: 4189-4194. 1986.
Albain. K. S.. Crowley, J. J., LeBlanc. M., and Livingston, R. B. Determinants of
improved outcome in small cell lung cancer: an analysis of the 2580 patient South
west Oncology Group data base. J. Clin. Oncol.. K: 1563-1574. 1990.
65
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
SERUM
CYFRA
21-1 IN LUNG CANCER
34. Van Den Brande. P. Serum láclatedehydrogenase in small cell lung cancer. N. Engl.
¡.Med.. 320: 61, 1989.
35. Jorgensen. L. G.. Osterlind. K., Hansen. H. H., and Cooper, E. H. The prognostic
influence of serum neuron specific enolase in small cell lung cancer. Br. J. Cancer. 58:
805-807, 1988.
36. Osbom. M, and Weber, K. Intermediate filaments: cell-type-specific markers in
differentiation and pathology. Cell, 31: 303-306, 1982.
37. Hoefler, H., and Denk, H. Immunohistochemical demonstration of cytokeratins in
gastrointestinal carcinoids and their probable precursors cells. Virchows Arch. B Cell
Palhol., 403: 235-240, 1984.
38. Nicolson. G. L. Tumor cell instability, diversification, and progression to the metastatic phenotype: from oncogene to oncofetal expression. Cancer Res., 47: 14731487, 1987.
39. Van Muijen. G. N. P., Ruiter. D. J., Van Leeuwen, C. Prins, F A.. Reitsema, K., and
40.
41.
42.
43.
Wamaar, S. O. Cytokeralin and neurofilament in lung carcinomas. Am. J. Pathol., 116:
363-369. 1984.
Broers. J. L. V.. Carney. D. N.. De Ley, L., Vooijs, G. P., and Ramaekers, E C. S.
Differential expression of intermediate filament proteins distinguishes classic from
varianl small-cell lung cancer cell lines. Proc. Nati. Acad. Sci. USA. 82: 4409^1413.
1985.
Moll, R., Krepier, R., Franke, W. W. Complex cytokeratin polypeptide patterns
observed in certain human carcinomas. Differentiation, 23: 256-269, 1983.
Debus, E., Moll, R.. Franke, W. W., Weber, K., and Osbom, M. Immunohistochemical
distinction of human carcinomas by cytokeratin typing with monoclonal antibodies.
Am. J. Pathol., 114: 121-130, 1984.
Blobel, G. A.. Moll. R., Franke, W. W., and Vogt-Moykopf, I. Cylokeratins in normal
and lung carcinomas. I. Adenocarcinomas, squamous cell carcinomas and cultured
cell lines. Virchows Arch. B Cell Pathol., 45: 407-^»29,1984.
66
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.
Serum Fragment of Cytokeratin Subunit 19 Measured by CYFRA
21-1 Immunoradiometric Assay as a Marker of Lung Cancer
Jean-Louis Pujol, Jean Grenier, Jean-Pierre Daurès, et al.
Cancer Res 1993;53:61-66.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/53/1/61
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.