1018
Early Diagnosis and Treatment Monitoring Roles of
Tumor Markers Cyfra 21-1 and TPS in Oral Squamous
Cell Carcinoma
Rafael M. Nagler, D.M.D., M.Sc., Ph.D.1,2
Mira Barak, Ph.D.3
Micha Peled, D.M.D., M.D.1
Hanna Ben-Aryeh, Ph.D.2
Margarita Filatov2
Dov Laufer, D.M.D.1
1
Department of Oral and Maxillofacial Surgery,
Rambam Medical Center and Faculty of Medicine,
Technion-Israel Institute of Technology, Haifa, Is-
rael.
2
Oral Biochemistry Laboratory, Rambam Medical
Center and Faculty of Medicine, Technion-Israel
Institute of Technology, Haifa, Israel.
3
Central Laboratories Haifa and Western Galilee,
Nesher, Israel.
The authors thank Mrs. S. Gan for statistical anal-
ysis and Mrs. M. Perlmutter for her assistance in
the preparation of this article.
Address for reprints: Rafael M. Nagler, D.M.D.,
M.Sc., Ph.D., 17 Yaara Street, 23840 Timrat, Is-
rael.
Received June 8, 1998; revision received Septem-
ber 15, 1998; accepted September 15, 1998.
© 1999 American Cancer Society
BACKGROUND. Mucosal oral squamous cell carcinoma (SCC) accounts for 3–5% of
all reported cancers, with a 5-year survival rate of approximately 50%. Unfortu-
nately, current detection means are of no value in diagnosing lesions early enough
for cure, especially when they recur after resection. Postoperative radiotherapy
and/or covering the resection site with reconstructive flaps (regional or free vas-
cularized) often makes early diagnosis an impossible task.
METHODS. The authors examined the detection and treatment monitoring capacity
of two relatively new tumor markers in the serum of SCC patients, comparing their
levels with those in patients with other oral/perioral malignancies or benign oral
tumors and with disease free, posttreatment SCC patients and healthy controls.
RESULTS. Values of sensitivity, specificity, and positive and negative prediction for
Cyfra 21-1 were 96%, 87%, 93%, and 53%, respectively, whereas those for tissue
polypeptide specific antigen (TPS) were 69%, 87%, 93%, and 54%, respectively.
Approximately 2–3 weeks after resection of the SCC lesion, Cyfra 21-1 and TPS
levels were reduced by 47% (P # 0.003) and 36% (P # 0.041), respectively. Cyfra
21-1 levels in SCC patients were significantly greater than those of healthy patients
by 73% (P # 0.0001), patients with benign tumors by 74% (P # 0.0003), and patients
in disease remission by 66% (P # 0.0002). Similarly, the TPS levels of SCC patients
were significantly greater than those of healthy patients by 59% (P # 0.0005),
patients with benign tumors by 55% (P # 0.0001), and patients in disease remission
by 59% (P # 0.0001). In two patients, a second, new SCC lesion was diagnosed
within the follow-up period, with increased tumor markers noted concomitantly
with the diagnosis.
CONCLUSIONS. The accumulated data point to the suitability of the clinical usage of
these two markers, especially Cyfra 21-1, in the early detection of oral SCC lesions
(primary, recurrent, or secondary) as well as for treatment monitoring. These
results may open new avenues for the diagnosis and follow-up of these patients
and hopefully improve their treatment outcome. Cancer 1999;85:1018 –25.
© 1999 American Cancer Society.
KEYWORDS: diagnosis, prognosis, oral, squamous cell carcinoma, tumor markers,
Cyfra 21-1, tissue polypeptide specific antigen (TPS).
O ral carcinoma (mucosal squamous cell carcinoma [SCC]) ac-
counts for .90% of malignant lesions of the mouth and is a major
and growing problem worldwide.1 The relative prevalence of oral SCC
is 3–5% of all cancer cases, similar to that of skin melanoma. Unfor-
tunately, despite enormous clinical and basic investigative efforts, the
reported 5-year survival rates of oral carcinoma remain essentially
unchanged.2,3 Primary oral carcinoma usually is superficial and easily
detected. However, deeply invading tumors and oral tumors that
 Tumor Markers in Oral Squamous Cell Carcinoma/Nagler et al.
recur after primary resection usually are detected only
when they are large and no longer curable. Further-
more, early diagnosis of recurrence is difficult if not
impossible, especially when the resected region is ir-
radiated or covered by a reconstructive flap. No fewer
than 15– 43% of T1 and T2 SCC patients4,5 have clini-
cally undetected, occult neck metastases and, accord-
ingly, do not undergo dissection until the metastases
become clinically apparent, with an attendant failure
rate of 40%.6 This unfortunate state of affairs is com-
plicated further by the fact that imaging modalities,
such as computed tomography (CT) and magnetic
resonance imaging (MRI) are not helpful in the early
stages of tumor development.7–10 Thus, an accurate
and sensitive method for detecting early SCC lesions
as well as predicting regional recurrence and/or
spreading metastases is of paramount importance.11
Tumor markers, which have been accepted as a
valuable tool for diagnosis, prognosis, and treatment
monitoring in recent years,12–18 could be an easy and
desirable means of achieving this purpose. Various
tumor markers have been examined for their value in
detecting head and neck and/or oral carcinoma.19 –31
However, the main drawback of these markers was
their low sensitivity for head and neck carcinoma,
which rendered them useless for clinical purposes.
Recently, two new markers, Cyfra 21-1 (cytokera-
tine-19 fragments) and tissue polypeptide specific an-
tigen (TPS) were shown to be highly sensitive and
specific in their detection potential and prognostic
value in head and neck malignancies.12,32,33 Unfortu-
nately, these markers were not evaluated for oral SCC.
Therefore, the purpose of this study was to examine
the serum levels of Cyfra 21-1 and TPS in oral SCC
patients before and after treatment, thus determining
the suitability of these markers for SCC diagnosis,
prognosis, and treatment monitoring.
PATIENTS AND METHODS
Both Cyfra 21-1 and TPS were analyzed in 148 patients
who were divided into the following groups: 1) 38
patients with oral SCC, which included a subgroup of
15 patients who were examined at 2–3 weeks after
tumor eradication; 2) 12 patients with other oral/peri-
oral malignancies but not SCC. The malignancies were
located in the maxilla, mandible, tongue, and the par-
toid and submandibular glands and included malig-
nant lymphoma, malignant melanoma, sarcoma, SCC,
and the following salivary malignancies: adenocarci-
noma, mucoepidermoid carcinoma, carcinoma E 1
mixed tumor, acinic cell carcinoma, and adenoid cys-
tic carcinoma; 3) 48 patients with benign oral tumors,
including fibroma, granuloma, neuroma, lipoma, os-
teoma, ameloblastoma, oncocytoma, pleomorphic
1019
and monomorphic adenoma, hemangioma, angio-
myxoma, giant cell granuloma, plasmacytoma, my-
coma, cherubism, and eosinophilic granuloma; 4) 20
patients in disease remission from oral SCC (disease
free) with no evidence of disease by history, physical
examination, biopsy specimen, and radiologic tests;
and 5) 30 healthy controls who were matched for age
and gender.
Patients in the first group also were followed pe-
riodically in an attempt at detecting recurrences. The
malignant lesions of the SCC patients were classified
as moderately or poorly differentiated according to the
histopathology and divided into Stages I and II or
Stages III and IV according to local invasiveness and
the TNM staging classification.
Venous blood samples (6 mL) were collected after
informed consent was obtained from patients admit-
ted to the Department of Oral and Maxillofacial Sur-
gery at Rambam Medical Center for treatment of their
oral SCC lesions. The samples were allowed to clot,
centrifuged at room temperature, and stored at 220 °C
until assayed. TPS was assayed using the monoclonal
immunoradiometric assay (IRMA) of Beki.34 The assay
measures the M3 epitope soluble fragments of human
cytokeratin 18. Cyfra 21-1 was evaluated using a kit
(Elsa-Cyfra 21-1 IRMA kit; CIS Bio-International, Gif-
Sur-Yvette, France). Cyfra 21-1 was developed using
two monoclonal antibodies (BM 19-21 and KS 19-1)
that react with different epitopes on cytokeratin 19
found in serum samples. The first monoclonal anti-
body was immobilized in plastic tubes, whereas the
second antibody was iodinated. When the serum sam-
ple contained cytokeratin 19 fragments, their epitopes
cross-linked both antibodies, resulting in an increase
in the radioactivity as measured by a gamma counter.
Statistical Evaluation
The results of the statistical evaluation were taken
from these 5 groups of patients: SCC (n 5 38), other
malignancies (n 5 12), benign tumors (n 5 48), dis-
ease free (n 5 20), and healthy (n 5 30). Demographic
and clinical characteristics of gender, age, histology,
and stage of disease were obtained for the SCC pa-
tients. Data from serum tests of Cyfra 21-1 and TPS
levels in each group were observed and calculated.
Means, standard deviations, and standard errors were
computed.
The site, stage, and histology levels between gen-
ders in the SCC patients were compared with the
Wilcoxon rank sum test.35 The means of Cyfra 21-1
and TPS levels between genders, disease stages, and
histology of the SCC patients were compared with the
two sample Student’s t tests for differences in means.
The mean Cyfra 21-1 and TPS levels of each of the five
1020 CANCER March 1, 1999 / Volume 85 / Number 5
groups were analyzed and compared using the analy-
sis of variance single factor test.36 A multiple compar-
ison test model was used. There were five means of the
groups for Cyfra 21-1 and five means of the group for
TPS, for all of which the Student-Newman-Keuls pro-
cedure for multiple pairwise comparison tests be-
tween all groups was performed. Ten of the two sam-
ple Student’s t tests for differences in means were
performed for the Cyfra 21-1 level and ten two sample
Student’s t tests were performed for the TPS level.
Each of the Student’s t tests tested the differences
between the means of two of the groups in the model.
The correlation between the Cyfra 21-1 level and TPS
level in each group was calculated using Pearson’s
correlation. The sensitivity, specificity, positive pre-
dictive value, and negative predictive value in the SCC
and healthy groups of patients were calculated at a
cutoff level of 0.7 ng/mL for the Cyfra 21-1 marker and
at 35 ng/mL for the TPS marker. These cutoff levels,
(i.e., the upper limits of normal) were determined by
calculating the mean value plus one standard devia-
tion. The differences in the Cyfra 21-1 level and the
TPS level for 15 patients before and after treatment
were compared with the Student’s t test for paired
differences.
RESULTS
SCC Group
The mean age of the SCC group (n 5 38) (Table 1) was
66.6 6 2.7 years and included 24 males with a mean
age of 69.2 6 9.3 years and 14 females with a mean age
of 61.4 6 17 years. Four patients had a recurrent lesion
whereas the other 33 patients had a primary lesion
when examined. The mean levels of Cyfra 21-1 and
TPS in this group were 1.71 6 0.29 ng/mL and 66.7 6
8.3 ng/mL, respectively (Tables 1 and 2). No signifi-
cant differences were found for either Cyfra 21-1 or
TPS levels between the following subgroups: males
versus females, various sites of the lesions, different T
or N scoring, Stages I and II versus Stages III and IV,
and patients with primary lesions versus patients with
recurrent lesions (Table 3). The lack of correlation
between the stage of the tumor and its concomitant
level of markers is significant, because more advanced
lesions would be expected to be characterized by
higher levels of markers.
After surgical resection of the tumors at 2–3
weeks, the serum levels of both Cyfra 21-1 and TPS
were obtained for 15 patients; significant reductions in
both markers were demonstrated in 14 patients (93%).
The Cyfra 21-1 levels were reduced from 1.61 6 0.19
ng/mL to 0.85 6 0.18 ng/mL (47%) (P # 0.003). The
TPS level was reduced from 66.2 6 10.7 ng/mL to
42.1 6 11.7 ng/mL (36%) (P # 0.041) (Fig. 1). Two
patients died soon after the surgical procedure,
whereas the remaining 13 patients were monitored at
later time points (mean follow-up period, 11.5
months). It is interesting to note that the mean Cyfra
21-1 and TPS values during this follow-up period for
the 11 disease free patients were not statistically dif-
ferent from those noted at 2–3 weeks after resection or
those of the control group. A second diagnosis of an
SCC lesion was made during the follow-up period in
two patients, with a concomitant increase in Cyfra
21-1.
Values of sensitivity, specificity, positive predic-
tive value, and negative predictive value for Cyfra 21-1
were 84%, 93%, 97%, and 70%, respectively, whereas
those for TPS were 69%, 87%, 93%, and 54%, respec-
tively. These values corresponded with cutoff levels of
0.7 ng/mL and 40 ng/mL for Cyfra 21-1 and TPS,
respectively. The mean levels of Cyfra 21-1 and TPS in
the 30 healthy controls were 0.49 6 0.8 ng/mL and
28.5 6 2.1 ng/mL, respectively. The correlation be-
tween Cyfra 21-1 and TPS in the SCC group was found
to be very high according to Pearson’s analysis (P 5
0.65 and P # 0.0001, respectively).
Comparison of SCC and Other Groups
The Cyfra 21-1 and TPS levels of four different control
groups were computed and compared with those of
the SCC group. These groups included 20 disease free
patients (in disease remission from oral SCC with no
evidence of disease on physical examination, biopsy
specimen, and radiologic tests 2–7 years after surgical
resection of the neoplasm), 48 patients with benign
oral tumors, 12 patients with other malignancies in
the oral/perioral region, and 30 healthy controls. The
mean levels of both Cyfra 21-1 and TPS in the various
groups are presented in Table 2. The Cyfra 21-1 level
in the SCC patients was significantly higher than those
in healthy controls by 73% (P # 0.0001), patients with
benign tumors by 64% (P # 0.0003), and patients in
disease remission by 66% (P # 0.0002) (Fig. 2). Simi-
larly, the TPS level in SCC patients was significantly
higher than those in healthy controls by 59% (P #
0.0005), patients with benign tumors by 55% (P #
0.0001), and patients in disease remission by 59% (P #
0.0001) (Fig. 3). There were no significant differences
in Cyfra 21-1 and TPS levels among the three control
groups (healthy controls, patients in disease remis-
sion, and patients with benign tumors). The Cyfra 21-1
levels of patients with other malignancies were signif-
icantly greater than that of the control groups (P ,
0.05), in contrast to the TPS level, which was not
statistically different from that of the three control
groups.
 Tumor Markers in Oral Squamous Cell Carcinoma/Nagler et al. 1021

TABLE 1
Profile of the SCC Group (n 5 38)

Cyfra TPS
Patient Gender Age (yrs) Site P/S/M TNM Stage (ng/mL) (ng/mL)

1 M 75 Mandible, cheek P T4N1M0 III–IV 0.9 49.4

2 F 70 Tongue P T2N0M0 I–II 1.6 —
3 M 68 Maxilla P T2N0M0 I–II 3.6 147.4
4 F 85 Tongue P T2N0M0 I–II 2.8 86
5 M 71 Mandible, cheek P T4N0M0 III–IV 1.5 58.6
6 F 49 Tongue P T4N2M0 III–IV 0.8 46.1
7 F 82 Tongue P T3N2M0 III–IV 1.3 40.3
8 F 76 Tongue P T4N1M0 III–IV 0.6 24.9
9 M 71 Maxilla S T1N0M0 I–II 1.8 92.3
10 M 45 Floor of mouth P T3N1M0 III–IV 0.6 55.2
11 M 79 Mandible P T3N1M0 III–IV 1.9 83.3
12 F 64 Cheek S T3N1M0 III–IV 0.8 3.1
13 M 80 Maxilla P T4N3M0 III–IV 0.4 32.9
14 M — Cheek P T4N0M0 III–IV 1.6 68.6
15 M — Floor of mouth P T4N2M0 III–IV 4 211.4
16 M 70 Tongue P T1N0M0 I–II 1.6 18.8
17 M — Tongue, floor of mouth P T4N3M0 III–IV 0.5 30.7
18 M 88 Lower lip P T1N0M0 I–II 1.5 82.4
19 M 82 Tongue P T2N0M0 I–II 2.6 45.4
20 F 81 Tongue P T2N0M0 I–II 0.9 36.7
21 M 43 Tongue S T4N1M0 III–IV 0.4 45.4
22 F 66 Maxilla P T1N0M0 I–II 0.6 19.6
23 M 41 Lower lip P T2N1M0 III–IV 0.8 16.8
24 M 84 Tongue P T2N0M0 I–II 0.9 27
25 M 62 Mandible, floor of mouth P T4N1M0 III–IV 1.9 81
26 F 62 Tongue P T2N0M0 I–II 2.6 120
27 M 21 Lower lip P T1N0M0 I–II 1.3 69
28 M 70 Tongue, floor of mouth S T3N1M0 III–IV 10.8 1400
29 M 51 Floor of mouth P T1N0M0 I–II 0.9 58
30 F 77 Tongue P T2N1M0 III–IV 1.3 21
31 M 31 Maxilla P T3N0M0 III–IV 1.5 48
32 M 74 Mandible, cheek P T3N0M0 III–IV 1 50
33 F 56 Mandible P T2N1M0 III–IV 1 80
34 F 79 Tongue P T2N0M0 I–II 0.8 198
35 M 72 Tongue P T2N0M0 I–II 2 88
36 F 66 Mandible M — — 121.8 1760
37 M 72 Tongue P T4N1M0 III–IV 1.9 132
38 F — Tongue P T3N0M0 III–IV 2.1 66

SCC: squamous cell carcinoma; P: primary; S: secondary; M: metastasis; TPS: tissue polypeptide specific antigen; M: male; F: female.

DISCUSSION Thus, the potentially neoplastic active region is hid-

To our knowledge, no reliable circulating tumor den, and both direct physical examination and/or har-
marker currently is available for use in the treatment vesting of a biopsy is prevented, while the only diag-
of patients with oral SCC, despite efforts to establish nostic tools remaining (CT, ultrasound, and MRI)
such a marker.26 –30,37 This is a major drawback in the often are not sensitive enough for early lesions. This
field because the importance of the early detection of led us to search for reliable circulating tumor markers
these lesions cannot be overemphasized. Early detec- that could assist in monitoring response to therapy
tion often is difficult in a previously operated and/or and in the early detection of recurrent disease or de-
irradiated region that develops fibrotic scars. The velopment of a second primary lesion.
widely and rapidly developing use of reconstruction We decided to examine two similar antigens shar-
flaps (both pedicle and free vascularized) in the last ing relatively new tumor markers, Cyfra 21-1 and TPS.
decade38 makes the situation much more difficult and These markers, evaluated using IRMAS, recently were
complex, because relatively large volumes of recon- examined in patients with head and neck malignan-
structive tissue eventually cover the resected area. cies and demonstrated their diagnostic and prognostic
1022 CANCER March 1, 1999 / Volume 85 / Number 5

TABLE 2
Cyfra 21-1 and TPS Levels of Examined Groupsa

A
Cyfra 21-1 No. Mean SD SE P value

SCC 38 1.71 1.75 0.29

Other malignancies 12 1.23 1.04 0.30 NSb
Benign 48 0.61 0.34 0.05 , 0.01
Disease free 20 0.58 0.31 0.07 , 0.01
Healthy 15 0.49 0.21 0.08 , 0.01

B
TPS No. Mean SD SE P value

SCC 38 66.7 48.4 8.3

Other malignancies 12 41.3 43.4 13.1 NSb
Benign 48 30.1 26.8 3.9 , 0.01
Disease free 20 27.4 16.6 3.9 , 0.01
Healthy 15 28.5 11.5 2.1 , 0.01

TPS: tissue polypeptide specific antigen; SD: standard deviation; SE: standard error; SCC: squamous cell carcinoma; NS: not significant.
a
The groups examined were squamous cell carcinoma (n 5 38), other malignancies (n 5 12), benign (n 5 48), disease free (n 5 20), and healthy (n 5 30).
b
Not significant. Cyfra 21-1 and tissue polypeptide specific antigen levels of the four control groups (other malignancies, benign, disease free, and healthy) were compared with those of the squamous cell carcinoma
group.

TABLE 3
Cyfra 21-1 and TPS Levels of Subgroups of SCC Patients

Patients Cyfra 21-1 TPS

Group No. Mean SE Mean SE

Total 38 1.71 0.29 66.7 8.3

Histology—well 5 1.44 0.21 73 15.9
Histology—moderate 17 2.00 0.58 59.8 10.6
Histology—poor 11 1.24 0.26 58.7 6.7
Stages I–II 15 1.67 0.23 72.7 14.2
Stages III–IV 22 1.71 0.49 62.5 10.2
Male 24 1.91 0.43 69.2 9.3
Female 14 1.32 0.21 61.4 17.1
N50 20 1.64 0.18 73 10.8
N.0 17 1.76 0.6 59.6 12.9
T51 6 1.28 0.2 56.7 12.7
T52 12 1.74 0.28 78.7 17.4
T53 8 2.56 1.38 46.6 10.6
T54 11 1.32 0.32 71 16.7

TPS: tissue polypeptide specific antigen; SCC: squamous cell carcinoma; SE: standard error.

potential.12,32,33 Both markers are soluble forms of epitope of the tissue polypeptide antigen.32,40 The
keratins in human sera that are relatively new in the high sensitivity and usefulness in disease monitoring
field of monitoring cancer patients, and previously of TPS previously was reported in patients with breast
were examined in patients with pulmonary, bronchial, and cervical carcinomas.46 – 48
endometrial, cervical, urinary bladder, and large The presented data clearly demonstrate the po-
bowel carcinomas.39 – 43 The Cyfra 21-1 marker is a tential role of both markers in the diagnosis of oral
cytokeratin fragment (an intermediate filament pro- SCC, because the sensitivity and specificity of both
tein44) recognized by the KS 19-1 and BM 19-21 anti- markers were relatively high: 84% and 93%, respec-
bodies that was obtained by the immunization of mice tively, for Cyfra 21-1 and 69% and 87%, respectively,
with MCF-7 cells.45 The TPS marker is a specific M3 for TPS. The Cyfra 21-1 and TPS levels for SCC patients
 Tumor Markers in Oral Squamous Cell Carcinoma/Nagler et al.
FIGURE 1. Cyfra 21-1 and tissue polypeptide specific antigen (TPS) levels of
15 oral squamous cell carcinoma patients before (Pre) and 2–3 weeks after
(Post) resection of the tumors. For both markers, there was a significant
reduction. * denotes P # 0.05; ** denotes P # 0.01.
FIGURE 2. Cyfra 21-1 levels of the squamous cell carcinoma (SCC) group in
comparison with those of another oral/perioral malignant group and three
nonmalignant control groups: healthy, post-SCC disease free patients, and
patients with benign tumors. The marker levels were significantly higher (P #
0.01) than those of the healthy control group by 73%, 66%, and 64%,
respectively. The numbers above the bars denote the number of patients in
each group. ** denotes P # 0.01.
were above the cutoff values in 84% and 69% of pa-
tients, respectively. The data also clearly delineate the
potential role for treatment monitoring of both mark-
ers, as demonstrated by the significant reduction of
both Cyfra 21-1 (47%; P # 0.003) and TPS (36%; P #
041) after total resection of the neoplastic tissue. Thus,
both Cyfra 21-1 and TPS were shown to have a dual
role in diagnosis and treatment monitoring. Further-
more, it was demonstrated that this detection capacity
is applicable to malignancies in the oral/perioral re-
gion in general, but not to the other control condi-
1023
FIGURE 3. Tissue polypeptide specific antigen (TPS) levels of the squamous
cell carcinoma (SCC) group in comparison with those of another oral/perioral
malignant group and three nonmalignant control groups: healthy, post-SCC
disease free patients, and patients with benign tumors. The marker levels were
significantly higher (P # 0.01) than those of the healthy control group by 59%,
59%, and 55%, respectively. The numbers above the bars denote the number
of patients in each group. ** denotes P # 0.01.
tions; healthy groups, groups with benign tumor, or
those with disease remission had similar mean levels
that were not statistically different. The mean levels
for the 30 healthy controls were 28.5 6 2.1 ng/mL for
TPS and 0.49 6 0.08 ng/mL for Cyfra 21-1. This level of
Cyfra 21-1 is close to the minimum detectable by the
assay, 0.3 ng/mL. The demonstrated slightly higher
accuracy of Cyfra 21-1 in comparison with TPS with
regard to both sensitivity and specificity accords well
with previously published studies performed in lung
and bronchial carcinoma patients.40 – 42 No statistically
significant differences were found between the levels
of Cyfra 21-1 and TPS in patients with primary or
recurrent lesions, emphasizing the value of these
markers in the early detection of oral SCC lesions
(Tables 1 and 2).
In a recent article, Bongers et al.33 failed to dem-
onstrate a correlation between the stage of head and
neck tumors and Cyfra 21-1 levels, in contrast to
Doweck et al.,12 who found such a correlation. Similar
contradictory findings regarding the correlation be-
tween the level of another tumor marker, SCC antigen,
and head and neck carcinoma also were found.25,49
The lack of correlation between the stage of the oral
SCC lesion and the level of both Cyfra 21-1 and TPS
could either characterize the nature of the lesion (rel-
atively sensitive for early detection by the markers) or
the relatively small number of patients with a great
variation between tumor stage and intraoral sites. In
this respect, it is interesting to note that Molina et al.32
did not find a significant correlation between the TPS
and SCC antigen serum levels in patients with head
1024 CANCER March 1, 1999 / Volume 85 / Number 5
and neck malignancies and their age or tumor size,
site, or histologic grade, but they did find such a cor-
relation with lymph node invasion.
The important role of Cyfra 21-1 in posttreatment
follow-up was demonstrated in the two patients in
whom a diagnosis of either a second primary tumor or
a recurrent tumor was made during the relatively
short follow-up period. In both patients, the marker
levels again were increased concomitantly with the
diagnosis of the recurrent neoplastic lesion and de-
creased again after its second resection. This promis-
ing role for Cyfra 21-1 for patients with oral SCC
should be investigated further in a larger group be-
cause the establishment of a sensitive recurrent detec-
tor is greatly desired.
The accumulated data point to the relatively high
sensitivity and specificity of both Cyfra 21-1 and TPS
in patients with oral SCC lesions. Cyfra 21-1 is slightly
superior but both markers are suitable for clinical use
in detecting early lesions (primary, recurrent, or sec-
ondary lesions) and in monitoring the efficacy of treat-
ment. These results may open new clinical avenues for
both the diagnosis and follow-up of oral SCC patients,
thereby improving their treatment outcome. Further
evaluation of both prognostic capacity and the early
detection of secondary SCC lesions in larger groups of
patients is warranted.
REFERENCES
1. Federle DJ, Adelson R, Niessen LC, Harrison K. Oral cavity
and pharyngeal cancer among Department of Veterans Af-
fairs hospital discharges. J Public Health Dent 1995;55:143–7.
2. Schantz SP. Carcinogenesis, markers, staging and prognosis
of head and neck cancer. Curr Opin Oncol 1993;5:483–90.
3. Cancers of the oral cavity and pharynx: a Statistics Review
Monograph 1973–1987. Washington, DC: U.S. Department
of Health and Human Services, 1991.
4. Spiro RH. Epidermoid carcinoma of the mobile tongue. Am J
Surg 1971;122:707–10.
5. Bradfield JS. Carcinoma of the mobile tongue: incidence of
cervical metastasis in early lesions related to method of
primary treatment. Laryngoscope 1983;93:1332– 6.
6. Spiro RH. Surgical approach to squamous carcinoma con-
fined to the tongue and the floor of the mouth. Otolaryngol
Head Neck Surg 1986;9:27–31.
7. Close LG. Computed tomography evaluation of regional
lymph node involvement in cancer of the oral cavity and
oropharynx. Otolaryngol Head Neck Surg 1989;11:309 –17.
8. Stern WBR. Computed tomography of the clinically negative
neck. Head Neck Surg 1990;12:109 –13.
9. van den Brekel MWM. Magnetic resonance imaging vs pal-
pation of cervical lymph node metastases. Arch Otolaryngol
Head Neck Surg 1991;117:666 –73.
10. Friedman M. Rationale for elective neck dissection in 1990.
Laryngoscope 1990;100:54 –9.
11. Leedy DA, Trune DR, Kronz JD, Weidner N, Cohen JI. Tumor
angiogenesis, the P53 antigen and cervical metastasis in
squamous cell carcinoma of the tongue. Arch Otolaryngol
Head Neck Surg 1994;111:417–22.
12. Doweck I, Barak M, Greenberg E, Uri N, Kellner J, Lurie M,
et al. Cyfra 21-1: a new potential tumor marker for squa-
mous cell carcinoma of the head and neck. Arch Otolaryngol
Head Neck Surg 1995;121:177– 81.
13. Nagumo K, Okada N, Takagi M, Yamamoto H, Amagasa T,
Fujibayashi T. Squamous cell carcinoma antigen in oral
squamous cell carcinomas. Tokyo Med Dent Univ 1990;37:
27–34.
14. Mitsuhashi N, Nagai T, Okazaki A, Hayakawa K, Katoh S,
Sugiyama S, et al. The clinical usefulness of serum SCC
antigen level determinations in patients with malignant tu-
mor. J Jpn Soc Cancer Ther 1987;22:2182–90.
15. Fischbach W, Meyer T, Barthi K. Squamous cell carcinoma
antigen in the diagnosis and treatment follow-up of oral and
facial squamous cell carcinoma. Cancer 1990;65:1321– 4.
16. Yoneda K, Hirota S, Yamamoto T, Ueda E, Ozaki T. Clinical
evaluation of tumor marker in head and neck squamous cell
carcinoma. Jpn J Cancer Clin 1990;36:458 – 64.
17. Kato H, Miyauchi F, Morioka H, Fujino T, Torigoe T. Tumor
antigen of human cervical squamous cell carcinoma. Cancer
1979;43:585–90.
18. Matsubara Y, Yasuda Y, Hanawa T, Miyamoto Y, Ninomiya
K, Hatakenaka R, et al. SCC-antigen in patients with lung
cancer. J Jpn Soc Cancer Ther 1986;21:1036 – 48.
19. Hirata S, Odajima T, Kohama GI, Ishigaki S, Niitsu Y. Sig-
nificance of glutathione S-transferase-II or a tumor marker
in patients with oral cancer. Cancer 1992;70:2381–7.
20. Silverman NA, Alexander JC, Chretien PB. CEA levels in head
and neck cancer. Cancer 1976;37:2204 –11.
21. Katopodis N, Hishaut Y, Geller NL, Stock CC. Lipid-associ-
ated static acid test for the detection of human cancer.
Cancer Res 1982;42:5270 –5.
22. Wolf GT, Chretien PB, Elias EG. Serum glycoproteins in head
and neck squamous carcinoma. Am J Surg 1979;138:489 –
500.
23. Shideler CE, Johns ME, Cantrell RW. Erythrocyte polyamine
determinations in patients with head and neck cancer. Arch
Otolaryngol Head Neck Surg 1981;107:752– 4.
24. Maxim PE, Veltri RW, Sprinkle PM, Pusater RJ. Soluble im-
mune complexes in sera from head and neck cancer pa-
tients: a preliminary report. Otolaryngology 1976;86:428 –32.
25. Yoshimura Y, Oka M, Harada T. SCC-antigen for detection of
squamous cell and mucoepidermoid carcinoma after pri-
mary treatment: a preliminary report. J Oral Maxillofac Surg
1990;48:1288 –92.
26. Schroder M, Meyer T. CEA studies in squamous cell carci-
nomas of the head and neck. HNO 1986;34:334 – 42.
27. Seifert G. The importance of tumor markers in oral pathol-
ogy. II. Cell membrane and cytoplasmic antigens as tumor
markers. Pathol Res Pract 1985;179:625– 8.
28. Xing RD, Wang ZS, Li CQ, Tang QY, Jiang CB, Zhan YZ. Total
sialic acid as a tumor marker for oral cancer. Int J Biol
Markers 1994;9:239 – 42.
29. Zoller J. The value of “tumor markers” in the therapy and
aftercare of carcinoma of the oral mucosa. Dtsch Zahn
Mund Kieferheilkd Zentralbl 1992;80:351–7.
30. Okitsu M, Shimada J, Hiranuma Y, Shimazaki T, Nakasato
M, Katoh K, et al. Clinical significance of serum beta 2-mi-
croglobulin as a tumor marker of oral and maxillofacial
malignant tumors. Meikai Daigaku Shigaku Zasshi 1988;17:
191–7.
31. Kato H, Morioka M, Aramaki S. Prognostic significance of
the tumor antigen TA-4 in squamous cell carcinoma of the
uterine cervix. Am J Obstet Gynecol 1983;145:350 – 4.
 Tumor Markers in Oral Squamous Cell Carcinoma/Nagler et al.
32. Molina R, Torres MD, Moragas M, Filella X, Joi J, Gimenez N,
et al. Prognostic value of TPS in patients with head and neck
malignancies: comparison with SCC. Anticancer Res 1995;
15:479 – 84.
33. Bongers V, Braakhuis BJM, Snow GB. Circulating fragments
of cytokeratin 19 in patients with head and neck squamous
cell carcinoma. Clin Otolaryngol 1995;20:479 – 82.
34. Rylander L, Ziegler E, Bergman T, Schoberl E, Steiner G,
Bergman A-C, et al. Molecular characterization of a tissue-
polypeptide-specific-antigen epitope and its relationship to
human cytokeratin 18. Eur J Biochem 1996;241:309 –14.
35. Wilcoxon F. Individual comby ranking methods. Biometrics
1945;1:80 –3.
36. Scheffe H. The analysis of variance. New York: John Wiley &
Sons, 1959.
37. Xing R, Chen R, Wang Z, Zhang Y. Serum sialic acid levels in
patients with oral and maxillofacial malignancy. J Oral Max-
illofac Surg 1991;49:843–7.
38. Nagler RM, Braun J, Daitzman M, Peled M, Har-Shai Y,
Laufer D. Spiral CT angiography—an alternative vascular
evaluation technique for head and neck microvascular re-
construction: a preliminary experience. Plast Reconstruct
Surg 1997;100:1697–702.
39. Ferdeghini M, Gadducci A, Prontera C, Castellani C, Annic-
chiarico C, Gagetti O, et al. Determination of serum levels of
different cytokeratins in patients with uterine malignancies.
Anticancer Res 1994;14:1393–7.
40. Pujol JL, Grenier J, Parrat E, Lehmann M, Lafontaine T,
Quantin X, et al. Cytokeratins as serum markers in lung
cancer: a comparison of CYFRA 21-1 and TPS. Am J Respir
Crit Care Med 1996;154:725–33.
41. Trevisani L, Putinati S, Sartori S, Abbasciano V, Bagni B.
1025
Cytokeratin tumor marker levels in bronchial washing in the
diagnosis of lung cancer. Chest 1996;109:104 – 8.
42. Giovanella L, Ceriani L, Bandera M, Rimoldi R, Beghe B,
Roncari G. Tissue polypeptide specific antigen [tps] and
cytokeratin 19 fragment [CYFRA 21-1] immunoradiometric
assay in non small cell lung cancer evaluation. Q J Nucl Med
1995;39:285–9.
43. Plebani M, Basso D, Navaglia F, De Paoli M, Tommasini A,
Cipriani A. Clinical evaluation of seven tumor markers in
lung cancer diagnosis: can any combination improve the
results? Br J Cancer 1995;72:170 –3.
44. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The
catalog of human cytokeratins: patterns of expression in nor-
mal epithelia, tumors and cultured cells. Cell 1982;31:11–24.
45. Broers JL, Ramaekers FC, Rot MK, Oostendorp T, Huysmans
A, van Muijen GN, et al. Cytokeratins in different types of
human lung cancer as monitored by chain-specific mono-
clonal antibodies. Cancer Res 1988;48:3221–9.
46. van Dalen A. Preoperative tumor marker levels in patients
with breast cancer and their prognosis. Tumour Biol 1990;
11:189 –95.
47. Gitsch G, Kainz C, Joura E, Frolich B, Bieglmayer C, Tatral G.
Squamous cell carcinoma antigen, tumor associated trypsin
inhibitor and tissue polypeptide specific antigen in fol-
low-up of stage III cervical cancer. Anticancer Res 1992;12:
1247–50.
48. van Dalen A. TPS in breast cancer: a comparative study with
carcinoembryonic antigen and CA 15.3. Tumour Biol 1992;
13:10 –7.
49. Ebert W, Johnson JT, editors. Tumor markers in the man-
agement of SCC of the head, neck and lung. Princeton, NJ:
Excerpta Medica, 1987.