J. Cancer (1990),
Serum levels of tumour associated glycoprotein (TAG 72) in patients with
G. Scambia, P. Benedetti Panici, L. Perrone, C. Sonsini, S. Giannelli, A. Gallo, P.G. Natali' &
Department ofGynecology, Catholic University, Largo A. Gemelli 8, 00168 Rome; and 'Laboratory ofImmunology, National
Cancer Institute, Regina Elena, Rome, Italy.
malignant gynaecological conditions in order to evaluate the clinical usefulness of TAG 72 alone or in
combination with other tumour markers. Sixty-six per cent of patients with ovarian cancer showed abnormal
concentrations of TAG 72 antigen. A good correlation was also found between serial TAG 72 values and the
clinical course of disease during chemotherapy and follow-up. In cervical and endometrial cancer abnormal
TAG 72 values occurred in 23% and 14% of cases, while none of the patients with breast cancer had
abnormal TAG 72 levels. Among patients with benign disease only one out of 12 patients (8%) with benign
ovarian tumours and one of 15 patients with uterine fibromyomatosis (7%) showed high TAG 72 serum levels.
However, the determination of TAG 72 did not increase the sensitivity of CA 125 and squamous cell
carcinoma antigen (SCC), in ovarian and cervical
recombinant interferon alpha-2b to 15 patients with ovarian cancer and different basal levels of TAG 72 did
not increase serum levels of the antigen.
Serum levels of TAG 72 were measured in 726 serum samples from patients with benign and
cancer, respectively. The systemic administration of
The clinical management of gynaecological malignancies has,
like other areas of oncology, increasingly benefited from the
development of monoclonal antibodies to tumour associated
antigens. Two markers are of particular interest, namely the
OC-125 antigen associated with ovarian tumours and the
squamous cell carcinoma antigen (SCC) of cervical cancer
(Bast et al., 1983; Kato et al., 1977). Although serum levels
of both macromolecules are currently employed to monitor
the clinical course of these two malignancies, both markers
suffer from a number of limitations.
Determination of CA 125 is often associated with false
negative results in early stages of the disease and in patients
with low tumour burden (Bast et al., 1983; Niloff et al., 1985;
Berek et al., 1986; Schilthuis et al., 1986; Benedetti Panici et
1987). Only 50%
associated with the expression of the SCC antigen (Kato et
al., 1977, 1984; Mauro et al., 1985; Senekjian et al., 1987).
More recently, the pancarcinoma associated antigen TAG
72 identified by MoAb B72.3 (Colcher et al.,
received increasing attention because of its expression in
almost all ovarian epithelial tumours (Johnson et al., 1985;
Thor et al., 1986). An additional interesting feature of TAG-
experiments in vitro (Greiner et al., 1984; Guadagni et al.,
1987) and in vivo models (Greiner et al., 1986). Although
preliminary findings (Klug et al., 1986) have shown that
elevated serum levels of TAG 72
percentage of ovarian cancer patients, no extensive data are
yet available as to whether the measurement of this marker
in patients with gynaecological tumours offers advantages in
terms of specificity and sensitivity over available assays.
In the present study we addressed these issues by com-
paratively measuring TAG 72, CA 125 and SCC in a large
group of patients affected by gynaecological malignancies.
Moreover, in order to verify whether IFN treatment can also
facilitate the detection of TAG 72 in the circulation, we
recombinant IFN alpha-2b (rIFN alpha-2b) in 15 patients
with ovarian cancer.
isits upregulation by interferons (IFNs),
at least in
are present in a high
its levels during a short treatment with
Materials and methods
In this study a total of 726 serum samples from 285 patients
with benign and malignant gynaecological diseases
examined. They were collected in our department from June
1987 to July 1989. Sera collected from 66 healthy women
from 22 to 52 years old were included as a control group.
One hundred and seventy-three patients (median age 55;
range 29-75) had primary malignant tumours, including 44
patients with cancer of the ovary, 66 of the cervix, 43 of the
endometrium and 20 of the breast. The diagnosis of each
lesion was histologically confirmed.
criteria. Of the ovarian cancer patients, 15 had stage I-II, 17
stage III and 12 stage IV. Twenty-seven primary ovarian
tumours were histologically serous, seven were endometrial,
three were mucinous and seven were undifferentiated. Twenty
out of 43 patients (46%) with endometrial cancer were at
stage I, 14 patients (32%) at stage II and nine patients (21%)
at stage III and IV. Eighteen cervical cancer patients (27%)
had stage I, 24 (36%) stage II and 24 (36%) stage III and IV.
All endometrial cancers were adenocarcinomas, while among
cervical tumours 57 were squamous and nine adenocarcino-
mas. All patients with breast cancer had stage I or stage II
disease. Tumours were classified from GI to G3 according to
the degree of histological differentiation.
The first serum sample was always taken before therapy. A
postoperative sample was usually obtained 4-6 weeks after
surgery or just before the first chemotherapy cycle. Serial
TAG 72 measurements were performed before each chemo-
therapy cycle and during follow-up.
Stability, progression and regression of disease were de-
fined according to WHO criteria (1979).
Fifteen patients with primary ovarian cancer underwent a
short course of rIFN alpha-2b (Intron, ESSEX, Milano,
Italy). Patients received 3 x 106 U daily i.m. of rIFN alpha-
2b for 3 consecutive days and TAG 72 levels were measured
daily for 3 days after the first injection.
Venous blood samples for marker determinations were
separated by centrifugation and aliquots
-20°C until assayed. TAG 72 assay was performed using
two commercially available kits (Centocor CA 72.4 RIA,
Sorin B 72.3 MKS IRMA). The Centocor CA 72.4 RIA is a
standard solid phase radioimmunoassay which utilises two
monoclonal antibodies, B 72.3 and cc 49, a second genera-
tion antibody with greater affinity to B 72.3 and similar
Correspondence: S. Mancuso.
Received 14 December 1989; and in revised form 28 February 1990.
Br. J. Cancer (1990), 62, 147-151
'."Macmillan Press Ltd., 1990
G. SCAMBIA et al.
specificity. In the B 72.3 MKS IRMA, B 72.3 is used both on
the solid phase immunoabsorbent and as a radiolabelled
probe to subsequently detect and quantify the bound antigen.
Intra- and inter-assay variations were 8.1% and 11.9% for
CA 72.4 RIA and 8.3% and 11.2% for B 72.3 MKS IRMA.
The upper normal level was taken as 7 U ml-' and 5 U ml-'
for CA 72.4 RIA and B 72.3 MKS IRMA, respectively. This
level corresponded to the mean value + 2 s.d. of our control
group. CA 125 and SCC were measured using commercially
available kits from CIS, Compagnie ORIS Industrie, SA and
Abbott Laboratories. The upper limits of normal were taken
as 35 U ml' and 2.5 ng ml-
tively. Using these cut-off values, 93% and 94% of our
healthy patients had normal marker values respectively.
Marker titres were defined as increasing or decreasing
when changes were greater than 50% of the previous value.
The x2 test and Fisher's exact test were used to evaluate the
distribution of marker values according to different variables.
for CA 125 and SCC respec-
Table I shows the median level, range and percentage of
positivity of TAG 72 in patients with benign and malignant
gynaecological tumours. Although CA 72.4 RIA assay seems
to be more sensitive than B 72.3 MKS IRMA, no statistically
significant difference between the two assays was found in
any of the gynaecological malignancies examined. Moreover,
linear regression analysis revealed a statistically significant
correlation (P<0.01) between values obtained with the two
assays (data not shown). Therefore, in continuing the study
we only referred to data obtained with the CA 72.4 RIA
Abnormally high TAG 72 values were seen most fre-
quently in ovarian tumours. Elevated levels of TAG 72
occurred in 29 out of 44 patients (66%) with primary
tumours, abnormal TAG 72 levels were only found in 15
cervical and endometrial
patients, respectively. None of the 20 patients with breast
endometriosis and endometrial hyperplasia showed TAG 72
levels within the normal range. Only one of 12 (8%) patients
with benign ovarian tumours and one out of 15 (7%) with
uterine fibromyomatosis had
Table II compares the distribution of CA 72.4 and CA 125
in ovarian cancer as related to stage, histology and tumour
grade. TAG 72 positivity was not significantly related to
increased TAG 72
72-3 MKS IRMA in patients with malignant and benign gynaecological
diseases and control subjects
CA 72-4 RIA
> 7 U ml-'(%)
Serum levels ofTAG 72 measured with CA 72-4 RIA and B
B 72-3 MKS IRMA
(0; 0- 13)
Endometrial cancer 43
Median and range levels are indicated in parentheses.
Serum positivity of TAG 72 and CA 125 in ovarian cancer
patients according to stage, histology and tumour grade
> 7 U ml-' > 35 U ml-'
CA 125Two assays
aAt least one positive test.
stage, histotype or histological grading. Using a combination
of TAG 72 and CA 125 assays, the overall sensitivity was
only slightly increased, with 89% of the serum samples show-
ing a positive reaction in at least one test compared to 86%
for CA 125 alone. We compared the sensitivity of TAG 72
with that of SCC in patients with cervical cancer. As shown
in Table III, the sensitivity of TAG 72 was significantly lower
with respect to that of SCC. When a combination of two
assays was used the overall sensitivity rose to 64% compared
to 59% of the SCC alone.
One month after completion of primary therapy all patients
with no clinical evidence of disease had TAG 72 levels within
the normal range.
In patients with advanced ovarian cancer who underwent
cytoreductive surgery, a good correlation was found between
TAG 72 levels and residual tumours after surgery (Table IV).
Five out of 12 (42%) patients, with residual disease >0.5 cm
showed abnormal TAG 72 serum levels with respect to one
out of four patients (25%) without residual disease. How-
ever, CA 125 was more sensitive than TAG 72 in detecting
residual disease after surgery.
TAG 72 and clinical course of disease
In 18 patients TAG 72 levels were measured during chemo-
therapy, consisting of three courses of high dose cisplatin
(200 mgm2) (Benedetti Panici et al., 1987). Figure 1 shows
serum TAG 72 levels before and at completion of chemo-
Combined sensitivity of SCC with TAG 72 in 66 patients
with cervical cancer
No. patients (%)
SCC> 2.5 ng ml'
TAG 72> 7 U ml'
Two assays combineda
aAt least one positive test.
Correlation of TAG 72 serum levels with residual tumour
after surgery in ovarian cancer patients
TAG 72>7 U ml-' CA 125>35 Uml'
TAG 72 AND GYNAECOLOGICAL TUMOURS
according to clinical response. The dotted line shows the upper
limit of normal (7 U ml-').
Serum TAG 72 levels before and after chemotherapy,
therapy in relation to clinical response to treatment. Ten out
TAG 72 titers while five out of seven patients with stationary
or progressive disease showed persistently elevated TAG 72
serum levels after chemotherapy.
In pre-treated ovarian cancer patients, TAG 72 levels cor-
related with disease status at follow-up. Seventy-five per cent
of patients with recurrent disease had an abnormal TAG 72
value while only one out of 11 patients with no evidence of
disease had a positive value (Table V).
In patients with positive TAG 72 serum
antigen measurements correlated well with the clinical course
of the disease. However, in none of the 10 patients with high
pre-treatment TAG 72 levels and in whom marker levels were
serially measured during follow-up did changes in TAG 72
levels provide a better correlation with the eventual course of
disease than changes in CA 125 levels (data not shown).
11 patients who responded to treatment had decreasing
TAG 72 levels andfindings on second-look
After chemotherapy 19 patients with ovarian cancer under-
went second-look laparoscopy (Table VI). TAG 72 serum
levels were found to be within the normal range in all six
disease, as well as in three patients with residual disease of
less than 2 cm. Only five out of
or cytological evidence of
10 patients with residual
Correlation of TAG 72 serum levels with clinical status of
disease at follow-up in ovarian cancer patients
No evidence of
Correlation ofTAG 72 andCA 125 serum levels with disease
status at second-look in ovarian cancer patients
No evidence of
> 2 cm
disease greater than 2 cm had high TAG 72 serum levels.
Again, CA 125 proved to be a better indicator of persistent
disease at second-look than TAG 72. Moreover, two out of
three patients with positive CA 125 and no evidence of
disease at second-look relapsed within 6 months after second-
Effect ofrIFN alpha-2b on TAG 72 levels
To test whether TAG 72 serum levels could be up-regulated
by treatment with IFN (Greiner et al., 1984; Guadagni et al.,
1986), 15 patients with primary or recurrent ovarian cancer
received rIFN alpha-2b for 3 consecutive days (3 x 106 U
daily i.m.) and TAG 72 was measured daily for 3 days after
the first injection. As shown in Table VII, rIFN alpha-2b did
not significantly modify TAG 72 serum levels, irrespective of
the antigen basal levels. In only one case did TAG 72 levels
markedly decrease after IFN administration.
The need to improve the diagnosis and the clinical monitor-
ing of gynaecological tumours has encouraged the measure-
ment of tumour associated serum antigens. In this report we
analysed the clinical usefulness of a novel human pancarcin-
oma antigen called TAG 72 in patients with gynaecological
malignancies. Sixty-six per cent of patients with primary
epithelial ovarian cancer showed abnormal concentrations of
TAG 72 antigen, which were found to be correlated with the
clinical course of disease during chemotherapy and follow-
up. However, in our series, CA 125 proved to be significantly
more sensitive than TAG 72 in detecting not only primary
ovarian tumours but also residual disease after surgery or
before second-look. Moreover, the combined evaluation of
CA 72.4 and CA 125 does not result in increased sensitivity
and improved clinical usefulness. Interestingly, however, nor-
mal TAG 72 levels have been found in patients with endo-
metriosis, benign ovarian tumours and fibromyomatosis. This
is relevant since high CA 125 levels have been found in
patients with endometriosis and benign ovarian tumours
(Kenemans et al., 1988; Scambia et al., 1988). Our results
therefore are in agreement with previous histological findings
showing a high specificity of TAG 72 for malignant tissues
with respect to normal tissues and benign lesions. Thor et al.
(1986) reported B 72.3 immunoreactivity in more than 70%
of epithelial ovarian tumours, while normal ovarian tissues
and 26 out of 27 various benign ovarian tumours were
negative. Interestingly, the only benign tumours with TAG
72 expression showed
an unusual glandular complexity.
Variation of TAG 72 during recombinant interferon
Recombinant alpha-2b interferon was administered (5 x 106 U daily,
i.m.) for 3 consecutive days (days 1-111) and TAG 72 was measured
daily for 3 days after the first injection.
G. SCAMBIA et al.
Based on these findings it can be suggested that TAG 72
assay could be of some value in the differential diagnosis of
pelvic tumours in order to reduce the false positive results of
CA 125. Moreover, the high specificity of MAB B 72.3 for
malignant tissues suggests that this antibody could be useful
for tumour radioimmunodetection and possibly immuno-
therapy. It is also worth noting that Colcher et al. (1984)
reported a prolonged binding of radiolabelled B 72.3 when
used for the in situ radioimmunodetection of human colon
Surwit et al. (1989) indicated that labelled B 72.3 could be
ovarian cancer. Compared to ovarian cancer in uterine
tumours, the sensitivity of TAG 72 assay was very low: 14%
in endometrial cancer and 23% in cervical cancer. In these
neoplasias only patients with advanced disease were found to
have significantly positive TAG 72 serum levels. Moreover, in
cervical cancer quantification of TAG 72 does not improve
the sensitivity of SCC alone.
As previously reported by Klug et al. (1986), elevated
levels of TAG 72 are only rarely present in the sera of
patients with breast cancer. This is rather surprising since
MAB B 72.3 was generated using a membrane-enriched ex-
tract of a breast tumour metastasis as immunogen and has
high level of reactivity with at least 50% of all breast car-
cinomas (Colcher et al., 1981). Although the discrepancy
between the tissue and serological expression might reflect the
heterogeneity of antigen expression in malignant cells within
a tumour (Stramignoni et al., 1983), it can also be hypo-
thesised that TAG 72 might be shed into circulation by
different tumours at different rates. These findings also sug-
gest that the serum TAG 72 assay cannot be used for screen-
radiolabelled B 72.3.
Because the sensitivity of radioimmunometric methods
employing MoAb may be significantly influenced by the type
of combination of reagent employed, we analysed our panel
of sera using two different commercially available TAG 72
Our results did not show any significant advantage of
TAG 72 measurement with a double antibody assay with
respect to a single antibody assay, suggesting that at least in
gynaecological malignancies this antigen contains multiple
epitopes for B 72.3. Much interest has been focused on the
finding that the surface expression of TAG 72 is increased by
IFNs both in vitro on human breast and colon cancer cells
(Grenier et al., 1984; Guadagni et al., 1987) and in vivo in
human tumour xenograft in nude mice (Grenier et al., 1986).
However, the systemic administration of rIFN alpha-2b to 15
patients with ovarian cancer and different basal values of
TAG 72 did not stimulate TAG 72 levels. Our negative
findings can be explained as follows: (1) a more prolonged
exposure to rIFN alpha-2b and/or a longer lead time after
IFN administration could be necessary in order to observe an
increase of TAG 72 levels; (2) different types of IFNs could
be more effective than rIFN alpha-2b in the stimulation of
TAG 72 expression; (3) IFNs could stimulate the surface
expression of TAG 72 without a concomitant release of the
antigen in the circulation. The first two possibilities seem
rather unlikely. Several species of rIFN alpha, including
rIFN alpha B, are able to stimulate the surface binding of
MAB B 72.3 at very low concentrations with a lead time
between rIFN treatment and TAG 72 increase of 24-36h
(Greiner et al., 1986). It is also worth noting that using the
same experimental procedure as that employed in this study
we detected a marked increase of the circulating levels of a
90,000 Da antigen with a close temporal correlation between
in vitro and in vivo data (Iacobelli et al., 1988; Scambia et al.,
1990). Our data agree with the recent findings of Boyer et al.
(1989), who were unable to demonstrate any IFN-induced
increase of TAG 72 expression in six ovarian cancer cell lines
In conclusion, our findings suggest that the serum assay of
TAG 72 has only a limited value in the clinical management
of patients with gynaecological malignancies. Further studies
should be carried out to verify the possible role of B 72.3 in
the pre-operative differential diagnosis of pelvic masses as
well as in tumour radioimmunodetection.
BAST, R.C., KLUG, T.L., ST. JOHN, E. et al. (1983). A radioimmunoassay
using a monoclonal antibody to monitor the course of epithelial
ovarian cancer. N. Engl. J. Med., 9, 883.
BENEDE1TI PANICI, P., GREGGI, S., SCAMBIA, G. et al. (1987).
High-dose (200 mg/M2) Cisplatin-induced neurotoxicity in primary
advanced ovarian cancer patients. Cancer Treat. Rep., 7, 6669.
BENEDETTI PANICI, P., SCAMBIA, G., BAIOCCHI. G. et al. (1989).
Predictive value of multiple tumor marker assay at second look
procedures in ovarian cancer. Gynecol. Oncol., 35, 286.
BEREK, J.S., KNAPP, R.C., MALKASIAN, G.D. et al. (1986). CA 125
serum levels correlated with second-look operations among ovarian
cancer patients. Obstet. Gynecol., 67, 685.
BOYER, C.M., DAWSON, D.V., SHARON, E.V. et al. (1989). Differential
induction of interferons of major histocompatibility complex-
antigens in human breast and ovarian carcinoma cell lines. Cancer
Res., 49, 2928.
COLCHER, D., HORAN HAND, P., NUTI, M. et al. (1981). A spectrum of
monoclonal antibodies reactive with human mammary tumor cells.
Proc. Natl Acad. Sci. USA, 78, 3199.
COLCHER, D., KEENAN, A.M., LARSON, S.M. et al. (1984). Prolonged
binding ofa radiolabeled monoclonal antibody (B72-3) used for the
in situ radioimmunodetection of human colon carcinoma xeno-
grafts. Cancer Res., 44, 5744.
GREINER, J.W., HORAN HAND,
Enhanced expression of surface tumor-associated antigens on
human breast and colon tumor cells after recombinant human
leukocyte alpha interferon treatment. Cancer Res., 44, 3209.
GREINER, J.W., FISHER, P.B., PESTKA, S. et al. (1986). Differential
effects of recombinant human leukocyte interferons on cell surface
antigen expression. Cancer Res., 46, 4984.
GREINER, J.W., GUADAGNI, F., NOGUCHI, P. et al. (1986). Recombin-
ant interferon enhances monoclonal antibody targetting of car-
cinoma lesions in vivo. Science, 235, 895.
GUADAGNI, F., SCHLOM, J.,JOHNSTON, W.W. et al. (1987). Recombin-
ant human interferons mediate enhancement of tumor antigen
expression on human tumor cells isolated from pleural effusion and
ascites. J. Interferon Res., 7, 798.
IACOBELLI, S., SCAMBIA, G., NATOLI, C. et al. (1988). Recombinant
human leukocyte interferon alpha 2b stimulates the synthesis and
release of a 90K tumor-associated antigen in human breast cancer
cells. Int. J. Cancer, 42, 182.
JOHNSON, V.G., SCHLOM, J., PATERSON, A.J. et al. (1985). Analysis ofa
tumor-associated glycoprotein (TAG 72) identified by monoclonal
antibody B72-3. Cancer Res., 46, 850.
KATO, H. & MORIOKA, T. (1977). Radioimmunoassay for tumor
antigen of human cervical squamous cell carcinoma. Cancer, 40,
KATO, H., TAMAI, K., MORIOKA, H. et al. (1984). Tumor antigen TA-4
in the detection of recurrence in cervical squamous cell carcinoma.
Cancer, 54, 1544.
KENEMANS, P., BAST, R.C., YEDEME, C.A. et al. (1988). CA 125 and
polymorphic epithelial mucin as serum tumor markers. Cancer Rev.,
KLUG, T.L., SATTLER, M.A., COLCHER, D. et al. (1986). Monoclonal
antibody immunoradiometric assay for an antigenic determinant
(CA 72) on a novel pancarcinoma antigen (TAG-72). Int. J. Cancer,
MAURO, T., SHIBATA, K., KIMURA, A. et al. (1985). Tumor associated
antigen, TA-4, in the monitoring of the effects of therapy for
squamous cell carcinoma of the uterine cervix. Cancer, 56, 302.
NILOFF, J.M., BAST, R.C., SCHAETZL, E.M. et al. (1985) Predictive value
of CA 125 antigen levels in second-look procedures for ovarian
cancer. Am. J. Obstet. Gynecol., 151, 981.
SCAMBIA, G., BENEDETTI PANICI, P., BAIOCCHI, G. et al. (1988).
Measurement of a monoclonal-antibody-defined antigen (90K) in
the sera of patients with ovarian cancer. Anticancer Res., 8, 761.
TAG 72 AND GYNAECOLOGICAL TUMOURS
SCAMBIA, G., BENEDETTI PANICI, P., IACOBELLI, S. et al. (1990).
Recombinant interferon alpha-2benhances thecirculating levels ofa
90K tumor associated antigen in patients with gynecological and
breast malignancies. Cancer, 65, 1325.
SCHILTHUIS, M.S., ALDERS, J.G., BOUMA, J. et al. (1987). Serum CA
125 levels in epithelial ovarian cancer: relation with findings at
second-look operations and their role in the detection of tumor
recurrence. Br. J. Obstet. Gynaecol., 94, 202.
SENEKJIAN, E.K., YOUNG, J.M., WEISER, P.A. et al. (1987). An
evaluation of squamous cell carcinoma antigen in patients with
cervical squamous cell carcinoma. Am. J. Obstet. Gynecol., 157,433.
Differential reactivity of monoclonal antibodies with human colon
adenocarcinomas and adenomas. Int. J. Cancer, 31, 543.
SURWIT, E., CHILDERS, J., GRAHAM, V. et al. (1989). Phase II clinical
studies ofmonab B 72.3-GYK-DTPA. Proc. Am. Soc. Clin. Oncol.,
THOR, A., OHUCHI, N., SZPAK, C.A. et al. (1986). Distribution of
oncofetal antigen tumor-associated glycoprotein-72 defined by
monoclonal antibody B72.3. Cancer Res., 46, 3118.
THOR, A., GORSTEIN, F., OHUCHI, N. et al. (1986). Tumor-associated
glycoprotein (TAG 72) in ovarian carcinomas defined by mono-
clonal antibody B72.3. J. Natl Cancer Inst., 76, 995.
WORLD HEALTH ORGANIZATION (1979). WHO Handbook for
Reporting Results of Cancer Treatment. WHO: Geneva.