carcinoembryonic antigen (cea) in cervical …...in practical terms, van nagell et al . (1981 )(5) ,...

Carcinoembryonic Antigen (CEA) In Cervical Neoplasia (Review)Published on OBGYN.Net (http://www.obgyn.net)

Carcinoembryonic Antigen (CEA) In Cervical Neoplasia (Review)June 28, 2011 | Laparoscopy [1], Pregnancy and Birth [2]By OBGYN.net Staff [3]

One of the important problems in the clinical management of cancer is the early detection of itsoccurrence. The demonstration of immunoreactive substances, which are specific or associated withcancer and measurable in serum, provides an interesting and promising approach to cancerdetection

CONTENTS:

Introduction

The Discovery of CEA

Ideal Marker

The Biology of CEA

Hypothetical Explanation of CEA Production in Malignant Cells

Isolation and Characterization of Carcinoembryonic Antigen [CEA]

Physicochemical Properties of Carcinoembryonic Antigen

Intracellular Localization of CEA

Immunological Criteria of Carcinoembryonic Antigen

The Antigenic Structure of Carcinoembryonic Antigen

CEA Cross Reacting Substances

Non-specific Cross-Reacting [NCA]

Normal Faecal Antigens [NFA]

Normal Faecal Antigen – 1 [NFA-1]

Normal Faecal Antigen – 2 (NFA–2)

Normal Faecal Cross-Reacting Antigen (NFCA)

Non-Specific Cross-Reacting Antigen–2 (NCA–2)

Meconium Antigen (MA)

Blood Group Determinants

Carcinoembryonic antigen in clinical practice

Carcinoembryonic antigen in serum: Immunoassay for CEA in Serum

Serum CEA Levels in Normal Individuals

of 44

http://www.obgyn.net/laparoscopy

http://www.obgyn.net/pregnancy-and-birth

http://www.obgyn.net/authors/obgynnet-staff

/node/192295#introduction

/node/192295#the_discovery

/node/192295#ideal_marker

/node/192295#the_biology

/node/192295#hypotherical_explanation

/node/192295#isolation

/node/192295#physicochemical

/node/192295#intracellular

/node/192295#immunological

/node/192295#the_antigenic

/node/192295#cea_cross_reacting

/node/192295#carcinoembryonic

/node/192295#carcinoembryonic_antigen_in_serum

/node/192295#serum_cea


Cut-Off Level for Serum CEA :Serum CEA in Gynaecological Malignancies

Carcinoembryonic Antigen in Tumour Tissues: Immunocytochemical Detection of CEA

Techniques of Immunoperoxidase Staining for CEA

Three-Step Unlabeled Antibody Technique

Two Step Labelled Antibody Techniques:

One Step Technique

Threshold CEA Tissue Staining Reaction

Tumour Tissue Susceptibility for CEA Staining

Light and electron microscopic findings in immunostained tumours for CEA

Carcinoembryonic Antigen in Cervical Neoplasms: Tissue CEA in neoplasms of the cervix

Tissue CEA in Rare Cervical Neoplasm

Patterns of Immunohistochemical Staining for CEA in Different Histological Types ofCervical Malignancy

Squamous Cell Carcinoma

Adenocarcinoma

Distinction Between Endocervical and Endometrial Adenocarcinoma Using Tissue Staining forCEA

Serum CEA in Cervical Neoplasms

Cut of Level for Carcinoembryonic Antigen in Serum

Pre – Treatment Serum CEA in Cancer Cervix

Effect of Histological Type of Cervical Carcinoma on Pre- Treatment Serum CEA Levels

Effect of Clinical Stage of the Disease on Pre-Treatment Serum CEA Levels

Post – Treatment Serum CEA in Cancer Cervix

Patterns of Change in Serum CEA Levels following Treatment

Post-Treatment Follow Up with Serum CEA

Causes of the Discrepancy Between the Results of Serum CEA in Different Investigations

Carcinoembryonic Antigen in Cervical Intraepithelial Neoplasia

Carcinoembryonic Antigen in Cervical Mucous and Cervico - Vaginal Washout Fluid

Introduction

of 44

/node/192295#cut_off_level

/node/192295#carcinoembryonic_antigen_in_tumor

/node/192295#techniques_of_immunoperoxidase

/node/192295#threshold_cea

/node/192295#tumor_tissue

/node/192295#light_and_electron

/node/192295#carcinoembryonic_antigen_in_cervical

/node/192295#distinction

/node/192295#distinction

/node/192295#serum_cea_in_cervical_neoplasms

/node/192295#cut_off_level_for_carcinoembryonic_antigen_in_serum

/node/192295#pre_treatment_serum

/node/192295#effect_of_histological

/node/192295#effect_of_clinical_stage

/node/192295#pre_treatment_serum

/node/192295#causes_of_the_discrepancy

/node/192295#carcinoembryonic_antigen_in_cervical_intraepithelial

/node/192295#carcino_embryonic_antigen_in_cervical_mucous


One of the important problems in the clinical management of cancer is the early detection of itsoccurrence. The demonstration of immunoreactive substances, which are specific or associated withcancer and measurable in serum, provides an interesting and promising approach to cancerdetection (1)

The measurement of tumour-associated antigens as tumour markers in the serum is useful for earlydiagnosis, differential diagnosis, and the determination of remission after therapy in cases ofmalignancy . However, it is important to be aware that each tumour-associated antigen reveals onlyone biologic characteristic of the tumour cells, and that a tumour mass consists of various types ofcells . It must also be understood that a tumour-associated antigen is not always present in all thecells forming the tumour and may also be detected in other tumours or normal organs (2)One of the important advantages of measuring secreted tumour markers in contrast to physicalmethods of assessment of tumour masses lies in their representative of viable tumour burden , animportant consideration when apparent mass constitutes a viable tumour (3).

The application of monoclonal antibody technology to development of cancer in humans has createdan increasing number of biologic tumour markers. However, their use in clinical medicine remainslimited because the required specificity has not been achieved . To date, no tumour associatedproteins unique or clearly specific to any neoplastic disease have been found. Because theseproteins are also produced by normal cells , interpretation of assay results depends on thequantitative value reported and not merely on the presence or absence of the . Tumour markers it isunlikely that tumour associated proteins will ever be useful for population screening when present inabnormal quantities, however, tumour markers have proved valuable in diagnosis and in followingdisease activity or response to treatment (4)

In practical terms, Van Nagell et al . (1981 )(5) , Considered a tumour marker as a substance which isselectively produced by tumour cells and then released into the blood stream in sufficient quantitiesto be measured by conventional techniques. Ideally, such a marker should be useful in tumourdiagnosis and serve as a biochemical monitor of disease status following therapy .Disappearance ofthe marker should indicate eradication of the tumour, whereas an increase in marker concentrationshould indicate tumour growth.(5)

Coombes and Powles ( 1982 )(6) , have suggested the following list for the potential clinicalapplications of circulating tumour markers . They may:

- Facilitate the primary detection of cancer.- Assist in monitoring the earlier detection of recurrence.- Assist in predicting response to treatment.- Assist in selection of patients for adjuvant chemotherapy.- Help in the localization of metastatic tumour : e.g. by using a specific biochemical parameter whichoriginates from a particular organ (e.g. liver enzymes ) .

The oncofoetal antigens comprise one particular group of markers produced by human neoplasms,these antigens have been detected in the sera of patients with gynaecological cancer. The practicaluse of such markers in the diagnosis and follow-up has been limited by the sensitivity and specificityof their tests (7). Carcinoembryonic antigen (CEA) is one of the first known tumour markers. Sincethen, many more have been described, but CEA, determined alone or in combination with others, isstill one of the most used. CEA is not organ specific and abnormal values may be found in a widerange of carcinomas (8).

The Discovery of CEA

The two best-characterized oncofoetal markers are Alpha- foetoprotein ( AFP ) and carcinoembryonicantigen (CEA ) in 1963, Abelev et al.(9) using immunodiffusion , discovered for the first time theAlpha-fetoprotein ( AFP ). This discovery started the modern era of tumour markers. CEA which wasfirst described as a gastrointestinal system-specific tumour antigen is still the most widely used andstudied member of this group. It should be noted, however, that CEA, AFP and other oncofoetalantigens such as foetal sulphoglycoprotein associated with carcinoma of the stomach, pancreatic

of 44


oncofoetal antigens, and oncofoetal antigens associated with carcinomas of the colon, lung, breast,liver and Kidney do not normally elicit an immune response in the patient.They are this not truly “antigenic” in the host and the term marker is more appropriate. Anti- CEAmonoclonal antibodies have been used in attempts to define cancer-specific epitomes of CEA, inevaluating CEA by immunohistology and as agents for radiommunoimaging .Similarly, anti- AFP monoclonal antibodies have been used to develop sensitive radio –immunoassay, to explore radiommunoimaging / immunotherapy and to study the control of AFPgene expression (10).The idea for the work on CEA devolved from two lectures. The first described the attempts andfailures to detect either components or functions unique to cancer cells and not found in any normaltissues(11). The second lecture dealt with the then relatively recent description of AcquiredImmunological Tolerance (11)--until 1953 only a theoretical concept--and the new perspectives thatthis phenomenon was creating in the understanding of the immune system and in various areas ofclinical medicine. Hence, the experiments subsequently undertaken were predicated on thesomewhat naive hope that the burgeoning tools of immunology, such as immunologic tolerance,would provide both the specificity and sensitivity that had been lacking in the previous efforts to findone or more specific tumor constituents in human cancer tissues.(11)In the 1950s, after perhaps a century of study in a variety of disciplines and using numeroustechnologies, the paradigm in the field was very simply that "tumor-specific materials" did not existin cancer tissue and, moreover, that these would not be found (12). Nevertheless, by the early1960s, studies of the rejection of well-defined transplantable tumors between highly inbred, orsyngeneic, animals had clearly demonstrated the existence of tumor-specific transplantationantigens (TSTAs) in these animal models (13). Conventional wisdom of the day maintained that suchartificially-generated could hardly be applicable to outbreed, "wild-type" humans. Notwithstandingthis, a sizable field of investigation grew up around the question of how TSTA-bearing animal tumorsescape the immune surveillance system, a question that has yet to be answered satisfactorily. It is toa field of tumor immunology in this state of flux that CEA research traces its origin.(11)In 1965, Gold and Freedman(14) described for the first time common specific antigens or antigenicdeterminants in adenocarcinoma of the human colon. Components identifiable with the colonictumour-specific antigens were found in carcinomata originating in other digestive system organs.Cancers of the lower bowel ( colon and rectum ) contained higher concentrations of tumour – specificantigens than those of the upper gastrointestinal tract (oesophagus and stomach) . Thesetumour-specific antigens could not be detected in malignant tumours of the uppermost portion of thedigestive system; i.e., in cancers of the palate and anterior portion of the floor of the . As the normalepithelium of those parts of the digestive system which developed the tumour-specific antigens uponmalignant transformation was derived from embryonic entodermal tissue (the lining of the for-, mid-and hindgut , and the parenchymal tissue of the liver and pancreas) , Gold and Freed-man suggestedthat cancerous lesions arising from various entodermally derived parts of human digestive systemwould certain tumour-specific antigens (14)(15).Colonic cancer was chosen as a model because this tumor does not extend intramurally beyond 5-6cm, either proximally or distally, from the tumor tissue in the gross. Hence, normal and tumor tissuestaken from the same individual at operation could be rationally compared without concern for theproblem of alloantigenicity that had plagued previous studies comparing tumor and normal tissuestaken from different individuals. The problem of obtaining fresh surgical specimens from theoperating room without interfering with the critical process of appropriate determination of thetumor should be noted (11). In one of the earliest series of experiments, then, adult male rabbitswere immunized with colonic cancer tissue extract and the resulting antiserum was absorbed with'an excess' of the corresponding normal material, while in a second series of studies an attempt wasmade to render neonatal rabbits tolerant to the normal tissue extract. These animals weresubsequently immunized with the corresponding tumor material, as adults. The necessary tissueextracts and antisera now in hand, these two series of studies led, with the use of immunologictechniques as indicator systems, to the demonstration of a tumor component in colonic cancer tissuethat was not found in the corresponding normal tissue (14).In a subsequent series of studies (15), it was found that the same cancer antigen was also present inall endodermally-derived gastrointestinal tumors--from the lower oesophagus, above, to theanorectal junction, below. The antigen was also demonstrable in primary tumors of the pancreas andthe liver, both derivatives of the second stage of the duodenum during embryologic development.Tumors of gut origin that underwent extra-enteric metastasis to such tissues as lung, bone andbrain, retained the antigenic activity in question, whereas tumors of all other organs that spread to

of 44


the liver, for example, were devoid of such activity. Hence, the parameter that determinedexpression of the antigenic activity described was the site of tumor origin, and not the site ofgrowth.(11)(14).A good deal of interest was generated by the "embryologic Confinement" of the tumor antigenicactivity in question. The substance appeared to be related to cancers that arose from tissue lyingbetween the embryologic stomatodeum, above, to the proctodeum, below. Thus, furtherinvestigations were done with human embryo- and fetus-derived tissues. Since this was done at atime well before abortion-on-demand, it took a fair amount of time to collect tissues fromspontaneous abortuses at various points in pregnancy. Nevertheless, it was found that gut-derivedtissue in the first two trimesters of gestation expressed the antigenic moiety of interest but thatcomparable tissue in the third trimester did not. Hence, the material was named thecarcinoembryonic antigen(s) of the human digestive system, subsequently abbreviated to CEA (14).Table (1):shows the milestone in the development of tumour markers and antigens.(9)

Table (1) : Milestones in the Development Tumour Markers and AntigensYear Author(s) Discovery/Event

1846 Bence – Jones Bence – Jones

1911 Rous Sarcoma transmission by cell-freeextracts

1928 Brown Extopic hormone syndrome1929 Woglom Concept of immunotherapy1929 Witebsky, Firzfeld Intestinal cancer – specific

antigens1930 Zondek Human chorionic gonadotropin

(HGG)1932 Cushing Adrenocortiocotropic hormone1933 Gutmann & Gutmann Prostatic acid phosphatase1940s Groes, Foley, Lewis, et al Inbred (syngeneic) strains of mice

and rats1950s Baldwin, Prehn & Main, Foley

Lewis & AptekmanPolymorphism of chemicallyinduced tumours. Absence ofcross-reactions

1949 Oh – Uti Deletion of blood-group antigens1959 Markert Isozymes1960 Newell Philadelphia chromosome1963 Abelev Alpha- Fetoprotein1965 Gold & Freedman Carcinoembryonic antigen1969 Heubner & Todaro Oncogenes1960s Potter Induction of myelomas in mice1965+ Sachs, Horibata, Lennox, Cohn Growth of myelomas in culture1975 Kohler & Milestein Monoclonal antibodies of

predefined specificity1980 Cooper, Weinberg, Bishop, et al Transfection, oncogene probes1981 Yunis Fragile sites1984 Drebin, Weinberg Green. et al Activated c-onc products

identified with monoclonalantibodies

(Daar and Lennox , 1987)(10)

Ideal Marker

Coombes and Neville ( 1978 )(16), have suggested specific criteria of an ideal marker, in that itshould:

Be easy and inexpensive to measure.

of 44


Be specific to the tumour studied and commonly associated with it.

Have a relationship between plasma level of the marker and tumour cell burden.

Have an abnormal plasma and/or urine level in the presence of micrometastases, i.e. at astage when no clinical or presently available diagnostic methods reveal their presence.

Have plasma and/or urine levels that are stable i.e. not subject to wild fluctuations.

Exist normally in a much lower concentration than that found in association with all stages ofcancer.

Coombes and Powles (1982 )(6) , have suggested the following list for the potential clinicalapplications of circulating tumour markers . They may:

Facilitate the primary detection of cancer.

Assist in monitoring the earlier detection of recurrence.

Assist in predicting response to treatment.

Assist in selection of patients for adjuvant chemotherapy.

Help in the localization of metastatic tumour: e.g. by using a specific biochemical parameterwhich originates from a particular organ (e.g. liver enzymes ).

In practical terms, van Nagell et al . (1981 )(5) , Considered a tumour marker as a substance which isselectively produced by tumour cells and then released into the blood stream in sufficient quantitiesto be measured by conventional techniques. Ideally, such a marker should be useful in tumourdiagnosis and serve as a biochemical monitor of disease status following therapy .

Disappearance of the marker should indicate eradication of the tumour, whereas an increase inmarker concentration should indicate tumour growth(5).

The Biology of CEA

The extensions of the two initial series of studies can primarily be divided between those focused onthe basic biology (localization, purification, chemical definition, functional activity and geneticcontrol) of the CEA molecule, and those directed at answering clinical questions(11). As concerns thebiology of CEA, early investigations involved the purification and characterization of the molecule,utilizing anti-CEA antiserum as a guide to the recovery and concentration of CEA at each stage(17)(18). These early studies would, years later, lead to the cloning of the CEA gene and to thepresent understanding of the functional biology of the CEA molecule and other members of itsmolecular family (19)(20)(21)(22)(23).

Hypothetical Explanation of CEA Production in Malignant Cells

In the very early stages of human embryonic cellular division , each primitive cell is endowed withthe totality of inherited genetic information this genetic potential which is passed on to allsubsequent generations of somatic cells, could theoretically allow each cell to specialize into anyhuman cellular form . As anatomical orientation is initiated under the influence of certain“organizers” distinct organ systems begin to develop. Systems of “regulator gene – repressor –operator “ act first to suppress the areas of the unclear genome not required for differentiation ofthese cells . However, the genetic potentialities necessary to allow the cells to undergo theirparticular type of specialization continue to be expressed (14)(15).

Fig . (1) : proposed model for regulation of protein synthesis in somatic cells.(24)

of 44


Combination of repressor with the operator inhibits the production of mRNA on the structural genes.The introduction of a substrate – inducer, into the cytoplasm, which combines with the repressorleads to initiation of mRNA production on the structural genes and thereby to the synthesis ofspecific molecule in the cytoplasm (24).

Upon adequate differentiation of the digestive system epithelium prior to birth, certain of the cellularcomponents needed during the course of specialization appear to be no longer required. At thisstage further repression of genetic information occurs and the production of “primitive” componentsceases. The disappearance of the carcinoembryonic antigens from foetal gut, liver and pancreas inthe Third trimester of gestation and their absence from the corresponding normal tissues of the adultis compatible with this hypothetical sequence of events.(14)(24)During the process of carcinogenesis in the gastrointestinal tract and the pancreas, depression ofpreviously repressed genetic information may occur leading to the recommencement of cytoplasmicproduction of carcinoembryonic antigens . Depression may be due to alteration of regulator oroperator genes by mutagenic carcinogens, or inactivation of certain cytoplasmic repressors by nonmutagenic carcinogens.Carcinoembryonic components seem to function in maintaining the viability of the foetal and cancercells of the gastrointestinal tract in their incompletely differentiated state, this may explain thesystem specificity of the production of these antigens . When differentiation is completed in thefoetal cells to the point where the Carcinoembryonic antigens disappear; other mechanisms maybecome operative in order to maintain the and function of the specialized cellular units. While incancer cells, the carcinoembryonic constituents have to be produced continuously in order tomaintain the life of the malignant cells (14)(15).

Isolation and Characterization of Carcinoembryonic Antigen [CEA]

Krupey et al. (1967)(17), attempted to purify and characterize the chemical composition ofcarcinoembryonic antigen (CEA). Using paper chromatography for quantities carbohydrate analysis,they revealed the presence of galactose, , mannose, focus and glucoseamine. Sialic acid was alsodemonstrated in purified CEA samples. The amino acid analysis of CEA showed the presence of eightamino acids. Tyrosine and lysins were found in relatively high proportions than the other six aminoacid residues namely: aspartic acid, threonine, serine, glutamine, glycine and alanine.(17)

Krupey et al. (1968)(18), mentioned that the degree of purity of CEA could be identified by finding asingle peak in the ultracentrifuge and a single band against untreated rabbit anti-tumour antiserumon immunoelectrophoresis. However, there were evidences that the purified preparations containedother components than carcinoembryonic antigen.

of 44


Coligan et al. (1972)(25), showed that tumours even of the same tissue type, can vary considerablyin their CEA content and CEA of two molecular sizes can exist in the same tumour as proved by gelfiltration, electrophoresis and ultra centrifugation. They believed that while CEA might exist in morethan one form, these forms represented definable chemical entities susceptible to completechemical characterization.

Martin and Martin (1970)(26), demonstrated the presence of a material immunologically identical toCEA in crude percholic acid extracts of non cancerous colon mucosa.

Lo Gerfo and Herter (1972)(27), showed by immunoelectrophoresis that CEA was present in benignpolyps of the colon and in non-malignant rectal mucosa obtained after haemorrhoidectomy.

Pusztaszeri and Mach (1973)(28), Sizaret and Martin (1973)(29), and Lo Gerfo et al. (1972)(30),demonstrated substances immunologically identical to CEA in bronchial carcinoma and normal lung.

Fritsche and Mach (1977)(31), were able to purify CEA from normal colon mucosa. It wasindistinguishable by immunological and physico-chemical criteria from reference colon carcinomaCEA. They noted that the concentration of CEA-reactive material in normal colon mucosa was about10-40 times lower than in primary large bowel carcinomas and 50-300 times lower than in met staticcolon or rectum carcinomas (31)

Martin et al. (1976)(32), found an increasing data supporting the premise that CEA was elevated inmany benign conditions and in non entodermally derived tumours. Significant CEA positively wasdetected in tumours of the , lung head and neck, reproductive organs and breast.(32)

Bale et al. (1980)(33), considered CEA to be a normal organ-specific antigen expressed to a greaterextent on tumours arising in that organ.

Physicochemical Properties of Carcinoembryonic Antigen

CEA is a glycoprotein with a molecular weight of approximately 200.000 Daltons (18), it migrateswith in the beta globulin region on electrophoresis. At neutral pH, the molecules appear in theelectron microscope as twisted rod-shaped particles

At pH4, which is near their average isoelectric points, the molecules tend to be more globular. AtpH9 they become more elongated presumably because the negative charges distributed throughoutthe molecule repel each other.

The CEA molecule is soluble in water, in percholic acid and in half saturated ammonium sulfate. Theproperties of solubility in percholic acid and half saturated ammonium sulfate form the basis of manyof the radioimmunoassays for serum CEA (34)(35).While there is a consensus between several laboratories about the molecular weight of CEAcorresponding to 180+20 kd Grunert et al (1985)(36), found two distinguishable CEA molecules inthis molecular weight range (160 kd and 180 kd) which showed a peptide composition differingmainly quantitatively (36).

Pletsch and Goldenberg (1974)(37), noted that the molecular size of CEA in the serum of patientswith cancer colon, cervix, ovary and bronchus was larger than the 200.0000 - Dalton CEA describedinitially. It was 370.000 + 4000 Daltons but it was immunologically identical to the 200.000 DaltonCEA. They put the following possibilities to explain the differences in the molecular size of thecirculating antigen:

1-The CEA in the plasma of patients with certain malignant disorders might be a dimer, orperhaps a mucin - CEA complex.2-The plasma antigen might be chemically unrelated to CEA except that it shared with it theimmunoreactive site measured in radioimmunoassay.3-The 370.000 - Dalton CEA might represent a more native form of the CEA molecule which

of 44


had not undergone enzymatic degradation to a smaller size.(37)

The most striking difference in the physicochemical properties of CEA is in the presence of widelyvariable percentage of carbohydrates ranging from 40% up to 82%. However, Terry et al (1974)(38)found that the percent of carbohydrate varied over a narrower range (45%-57%) .The carbohydrate protein ratio of CEA was usually of the order of 3/1. a ratio as high as 5/1 inpurified CEA of gastric cancer origin, and as low as 1/1 material obtained from colonic cancer tissue(39).The principal carbohydrate constituent of purified CEA preparations is N-acety1 glucosamine whichusually constitutes approximately 25% of the total weight of each preparation of purified material. Incontrast N-acetyl galactosamine is usually either or present only in very low concentrations in theCEA molecule. Some degree of variation is usually either or present only in very low concentrationsin the CEA molecule. Some degree of variation is usually noted in fucose, galactose mannose andsialic acid content between different preparations of CEA.. The sialic acid moiety is usually the mostvariable of the carbohydrate constituents, up to 5 fold difference in sialic acid content has beenfound (39).The carbohydrate moieties were distributed throughout the molecule ,30-36% of the carbohydrateportion of CEA was N-acetyl glucosamine, 21-27% galactose, 13-20% mannose and 13-20% fucose.Sialic acid comprised from 1-10% of the weight and was the most variable component Traces ofN-acety1 galactosamine were found (34).

The amino acid composition of the CEA preparations from all sources was constant withinexperimental error (Table 2).(39)(40)(41)(42)

Table (2) Amino Acid Composition of CEA Moles / 100 moles total amino acids

Amino acids Banjo et al(1972)

Terry et al(1972)

Tuberville et al(1973)

Hirai H. et al(1977)

Aspartic Acid 15.9 15.1 16.8 15.12Thereonine 9.3 8.3 10.3 9.16

Serine 11.4 9.6 10.7 10.2Glutamic Acid 11 10.1 11.1 10.42

Proline 8 11.1 8.1 8.1Glycine 5.5 5.2 5.7 5.4Alanine 5.9 5.8 6.6 5.72Valine 6.8 7 ND 6.96

Half-cystine 0 ND 0.1 0Methionine 0 0.1 0 0Isoleucine 5.1 4.9 5.7 5.04Leucine 9.3 8.8 9 8.2Tyrosine 2.5 3.9 3.8 4.2

Phenylalanine 21 2.4 2.6 2.41Lysine 2.5 2.5 3.4 2.65

Histidine 1.3 1.8 2.2 1.86Arginine 3.4 3.8 3.9 3.43

ND: not done (Hirai, 1977)(43)The amino acid sequence suggested by Terry et al (1972)(40) were confirmed by Kuroki et al

(1982)(41).

Koga et al (1985)(42), presented the following amino acid sequence of CEA, NCA-2 and NFA-2.(table3)

Table (3): Amino acid sequence of CEA, NCA-2 and NFA-2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

17 18

19 20 21 22 23 24

CEA

Lys Leu Thr Ile Glu Ser Thr Pro Phe Asn Val Ala Glu Gly Lys Glu Val Leu Leu Leu Val His Asn Leu

NCLys Leu Thr Ile Glu Ser Thr Pro Phe Asn Val Ala Glu Gly Lys Glu Val Leu Leu Leu Val His Asn Leu

of 44


A-2NFA-2

Lys Leu Thr Ile Glu Ser Thr Pro Phe Asn Val

(Koga et al 1985)(42)

Intracellular Localization of CEA

Studies with tissue cultured cells of colon cancer origin showed that CEA is a constituent of thetumour cell surface This observation was confirmed by immunofluorescence microscopy andimmunoferritin technique using electron microscopy (44). Martin et al (1976)(32), found that at leasta portion of the CEA is situated in the glycocalyx of the tumour cells immediately adjacent to thesurface membrane (32). CEA is located in the glycocalyx of the plasma membranes of colonic tumour cells, and in particularin that part of the membrane bordering the lumen of the anaplastic acini.(25)(45)

The studies of the metabolism of the glycocalyx showed that the carbohydrate layer can be replacedwithin a half life of about one day. This would account or the occurrence of CEA in plasma, if it wastrue for CEA also (45)The presence of CEA on the luminal surface of the tumour cells and the demonstration of CEA inconsiderable amounts in the mucous of hypersecretory well differentiated mucoid carcinomasproved that CEA is secreted by the tumour cells (46).

Immunological Criteria of Carcinoembryonic Antigen

Since the discovery of CEA, the definition of this antigen has been based on fulfilling certainphysicochemical, immuno-chemical and even biological criteria the identification of this substancewas solely dependant upon the ability of an antiserum to recognize a specific immunodominantgrouping as that detected by the original antiserum of Gold and Freedman (1965)(14). It has beensubsequently shown that CEA exhibits extensive heterogeneity in its physicochemical andimmunological properties.

Conventional antisera raised against CEA characteristically contain antibodies that react with agroup of substances closely related to CEA. At present the major members of this family of CEArelated antigens consist of the NCA which shares a similar tissue distribution with CEA, the NCA-2and NFA-2 found in meconium and adult faeces and the biliary glycoprotein - I present in normal bile(47).

Recent studies have demonstrated the immunological heterogenecity of CEA, polyclonal antisera andmonoclonal antibodies raised against CEA also react with CEA related antigens present in normaladult tissues such as spleen, lung, colon polymorphonuclear leukocytes, lymphocytes and RBCs.Thus CEA has been further defined as a family of "isoantigens" which have multiple antigenicdeterminants, some of them may be common to all the isoantigens, and others may be antigenicallydistinct for each of the members of CEA related antigens (48)(49)

Monoclonal antibodies provide a new powerful tool for the investigation of the immunologicrelationship between the different CEA related antigens, and may provide a better distinctionbetween tumour associated and normally occurring CEA related antigenic determinants (50)

However, Rogers et al (1981)(51) and Primus et al (1983)(47) anticipated that cross reactivity withCEA related antigens will be encountered with monoclonal antibodies to CEA analogous to theirconventional counterpart.

of 44


The Antigenic Structure of Carcinoembryonic Antigen

Chism et al (1977)(52) considered that at least 4 different antigenic determinants existed on CEAmolecule. They defined them as follows:

A determinant specific for colon cancer CEAAn antigenic Determinant specific for beast and ovarian CEAA determinant present on breast and ovarian CEA and also on colon carcinoma antigen IIIA determinant shared by all these molecules

Primus et al (1983)(47) using monoclonal antibodies developed against colonic tumour CEA andfocusing on their relative bonding to CEA, NCA (non specific cross reacting antigen) and MA(meconium antigen), were able to differentiate 3 general classes of antigenic determinants.

Class I, contained epitopes (antigenic sites) shared among all the 3 antigens and recognizedby the monoclonal antibody (NP-1)Class II, included antigenic sites shared between CEA and MA and identified by monoclonalantibodies (NP-2, NP-3)Class III, involved only antigenic determinants unique to CEA Molecule and recognized bymonoclonal antibody (NP-4)

Depending on these finding they concluded that at least 4 antigenic sites (epitomes) could bedifferentiated on colonic tumour CEA.

One was shared by CEA, NCA and MA. Two others shared by CEA and MA, and a fourth that appearedspecific for CEA. Additional CEA determinants were expected to be identified within all the 3 classes.(figure 2)(47)Fig. (2):

Wagener et al. (1983)(50), using monoclonal antibodies recognized 5 different epitomes on CEA. Allof the epitomes resided on the moiety of the molecule. By the use of these monoclonal antibodies,they were able to distinguish CEA from related antigens in liver and granulocytes.

Haskell et al. (1983)(53), also identified 5 different antigenic determinants on CEA with monoclonalantibodies. But they did not define whether these determinants were part of the protein or thecarbohydrate components of the CEA molecule. However, they found that the binding of somemonoclonal antibodies to antigens, unlike polyclonal antisera, might be quite sensitive to the ionicstrength of the reaction mixture. Their study confirmed considerable Variability in the binding of CEAby monoclonal antibodies as function of the ionic strength of the medium.(53)

of 44


Imai et al. (1984)(54) established 4 distinct monoclonal antibodies, which recognized 4 differentantigenic determinants. Two of the recognized determinants were of peptide nature, while the otherswere of carbohydrate nature. Non of the monoclonal antibodies reacted with NCA or NCA-2.(54)

Yachi et al. (1984)(55), also detected 4 antigenic determinants on CEA molecules using monoclonalantibodies. The nature of only two of these determinants were recognized, one was carbohydrateand the other a peptide one.

Muraro et al (1985)(48), generated fifteen monoclonal antibodies (col 1-15), which recognized atleast 5 epitomes on the CEA molecule and discriminated primary and met static colon carcinomasfrom adult tissues. None of the col-monoclonal antibodies were reactive with the surface ofpolymorphonuclear leukocytes. A major CEA related antigen termed NCA-1 (non specific crossreacting antigen-1) was found to be expressed by normal polymorphonuclear leukocytes. Thisantigen might explain the high levels of CEA commonly detected in the sera of patients with andalso in smokers. At least of the col-monoclonal antibodies (col-4 and col-12) were shown to reactwith distal metastases.(48)

Kurokli et al (1984)(56), using polyclonal antisera in comparative studies among CEA and 5 differentcross reacting substances, suggested that the antigenic structure of CEA could be divided at leastinto the following 4 antigenic parts (fig. 3):

1. NCA (non specific cross reacting antigen) common part: Shared with NCA, NCA-2 (inmeconium), NFCA and NFA-2.

2. NFCA (normal faecal cross reacting antigen) common part: Retained on NFCA, NFA-2 andNCA-2

3. NFA-1 (normal faecal antigen-1) common part: Shared with NFA-1, NFA-2 and NCA-2.4. CEA distinctive part. (56)

When monoclonal antibodies against CEA and related antigen were used, Kurokli et al.(1984)(56),were able to separate these antibodies into 4 distinct groups according to their reactivity’s with thestudied antigens. (figure 3)Group 1:Reacted with the epitopes on NCA common part present in CEA, NFA-2, NFCA, NCA but not withNFA-1.

Group II:Showed reactivity with the epitomes on NFCA common part found in CEA, NFA, NCA-2 and NFCA butnot with NFA-1 and NCA.

Group III:The antibodies in this group were reactive with the antigenic determinants on NFA-1 common part.Located in CEA, NFA-2, NCA-2 and NFA-1 but not NCA.

Group IV:Bound by CEA.

Fig (3):

of 44


C.P.: Common PartAll. P: Alloptic partNFCA: Normal faecal cross reacting antigen.NFA-1: Normal faecal antigen -1.NFA-2: Normal faecal antigen –2.NCA: Non specific cross reacting antigen.NCA-2: Non specific cross reacting antigen –2.The results of Kuroki et al. (1984)(56), revealed that at least eight epitomes could be identified onthe CEA molecule; two in group 1, one in group II, 3 in group III and 2 in group IV. The antigenic sitesdefined by antibodies in-group IV were referred to as CEA allotypic part. The epitomes in this partwere closely related and they might be one of the factors that gave rise to the various immunologyforms of CEA. All the eight epitomes identified in their study were in protein moiety of the CEAmolecule. Although previously it had been suggested by several investigators (57)(58) that the CEArelated antigens in carcinomas outside the colon and rectum e.g. breast, ovarian and lungcarcinomas, were cross reactive antigens rather than CEA. The use of monoclonal anti-CEAantibodies proved that the antigens present in most of the vital tumour tissues of different humancarcinomas studied were very similar to CEA as they contain all the antigenic determinantsrecognized in purified CEA preparations. So it seems that in human carcinomas, the production ofcross-reacting antigen without a concomitant production of CEA is a rare event excluding necrotictissue areas where binding heterogeneity is sometimes observed.(57)(58)(59)

The CEA related antigenic determinants might be degraded to a different extent in necrotic areas. Asa result antigens extracted from tumour containing high amounts of necrotic tissue may show apattern of partial immunological identity with CEA (60).

The epitomes reactivate of 52 well – characterized monoclonal antibodies against CEA from 11different research groups were studied using competitive solid phase immunoassays. Themonoclonal antibodies against carcinoembryonic antigen appear to react with only a limited numberof different antigenic areas of CEA.

of 44


Almost all (41/52) of the studied monoclonal antibodies are directed against carbohydrate epitomesand two gave inconclusive results. About 83% of those monoclonal antibodies could be classified intoone of five non-interacting epitomes groups containing between 4 and 15 monoclonal antibodieseach (61).

CEA Cross Reacting Substances:

Clarification of the chemical, antigenic, and development inter-relationships between CEA and CEA-like antigens is needed. They may have role tumour markers themselves or contribute to themeasurement of CEA levels in view of their strong immunological cross-reactivity with this antigen(47).

The major members of the family of CEA related antigens could be categorized as follows:

NCA (non specific cross reacting antigen)BGP – 1 biliary glycoprotein – 1)NFA in faeces, NCA –2 in meconium and faeces, CEA – like substances in gastric juice, NCWand BGP II in bile. These were found in the alimentary canal and were reported to be verysimilar to CEACEA – No in normal colon mucosa, identical to CEA in tumour tissue (47)(50)

The exact size of the CEA family is unknown. To date 13 most likely different CEA – relatedmacromolecules have been identified biochemical or by molecular biology techniques, they are:

NFA II (56)NCA – 160 “NCA of molecular weight 160.000 (62)NCA – II meconium antigen (62)CEA low “CEA of molecular weight 128.000” (63)NCA – 55 “NCA of molecular Weight 55.000” (38)NCA – 95 “NCA of molecular weight 95.000” (64)NFA – 1 (58)

Normal plasma antigen of molecular weight 114.000 (63)NCA species of molecular weight 75.000 (38)CEA related meconium antigen (63)Two new forms of granulocyte NCA of molecular weight 90.000 and 160.000 (65)CEA related foetal liver antigen (66)

These antigens show variable degrees of immunological cross – reactivity with CEA (61).

Non-specific Cross-Reacting [NCA]:

Von Kleist et al (1974)(67), were the first to describe this antigen and to give it the name of nonspecific cross reacting antigen. NCA is a B- globulin that remains present in the percholic acidextracts of colonic carcinoma. It has a molecular weight of approximately 50.000 Daltons. It contains25% carbohydrates and its location in the colonic cell is quite similar to that of CEA (i.e. it is situatedat the surface of cells and in intraglandular deposits).(67)

NCA can be ground normally in many argons such as lung, spleen, stomach and colonic mucosa. It isalso present in cancerous tissues especially of the digestive tract (32)

Several CEA like substances that had been described before were probably NCA based on immuno –chemical similarities. Among these substances were: Normal glycoprotein, CEA associated , colonicCEA-2, Colonic carcinoma antigen III, tumour associated antigen and the CEX antigen (57)(68)

As regards the faecal NCA no differences were observed between it and NCA in lungs or spleens. Themolecular weight of faecal NCA is about 50.000 to 90.000 it has at least 2 parts of antigenicdeterminants, one unique for NCA itself and other common to CEA, NFA-2 and NFCA (57).

of 44


CEA and NCA contain Homologous region, 24 amino acids in length at the N-terminus of thepolypeptide chain(69). The exact molecular relationship to CEA is not clearly understood and is stillsubject to controversy. Most difficulties in the identification of the molecules arise from the fact thatCEA and NCA have different molecular weights.(69)

For CEA, 2 molecular species of 160 and 180 kd (70) or 180 and 200 kd (56) were reported. While forNCA, molecular species of 60-80 and 100-120 kd had been reported (59)(71)(72)(73).

Normal Faecal Antigens [NFA]:

Although CEA related antigens in faeces other than faecal NCA have been inclusively NFA, it wasfound that there were 3 other molecular species of CEA related antigens in faeces, designated as:Normal faecal antigen-1 (NFA-1, NFA-2), and normal faecal cross reacting antigen (NFCA)respectively. Among these antigens NFA-1, NFA-2 and faecal NCA were isolated in pure form. NFCAhas not yet been obtained in pure form but was identified as an antigen separate from the otherthree antigens (57).

The origin and production mechanism of faecal antigens are still unclear, but it seems most probablethat NFA-1 and NFCA are produced from NFA-2 through degradation by bacterial or host enzymes inthe alimentary canal. Aseptic conditions and probable lack of proteolysis enzymes in the foetaldigestive tract may result in the absence of antigens in meconium corresponding to NFA-1 and NFCA(62).

Normal Faecal Antigen – 1 [NFA-1]:

It is antigenic ally unrelated to NCA, has a molecular weight of 20.000 to 30.000 and seems to be thesmallest CEA related antigen reported thus far The antigenic determinants corresponding to NFA-1(NFA-1 determinant) appears to be the most dominant immunogenic group among several antigenicmoieties on the CEA molecule because the amount of antibodies reactive with the NFA-1 is usuallythe largest in conventional rabbit anti-CEA antisera (57)

Normal Faecal Antigen – 2 (NFA–2):

It is the largest cross-reacting antigen in faeces with a molecular weight (160,000 to 170,000) similarto that of the CEA molecule from tumour tissues. Amino acid and carbohydrate compositions ofNFA–2 are similar to those of CEA. Antigentically as far as the reactivity with rabbit or goat antisera isconcerned, NFA–2 is indistinguishable from CEA. However, both CEA and NFA–2 has its own uniquedeterminant residing in different parts of the molecule with different chemical characteristics. Theorigin of NFA – 2 in the alimentary canal is still unclear. Fritsche and Mach (1977)(31), reported the presence of small amounts of CEA (CEA–No) in normalcolon mucosa that was identical to CEA in tumour tissues. The CEA–No might be a progenitormolecule of the NFA–2 in normal faeces. The antigenic differences between NFA–2 and CEA formtumour tissues could be an expression of molecular changes of CEA–No resulting from exposure toenzymes and / or bacteria in the digestive tract. Another similar antigen described was NCW innormal colon washings. At present it is unclear whether NCW is identical to CEA- No or NFA–2 (74)CEA–like substances in gastric juice and BGP III in bile must be taken into consideration, they seemto be very similar NFA–2 (57).

Normal Faecal Cross-Reacting Antigen (NFCA):

The antigenic determinants of this antigen can be divided into at least 2 moieties, one common toNCA and the other common to CEA or NFA–2 but unrelated to NFA–1. This antigen is not yetseparated in a pure form (57)

Non-Specific Cross-Reacting Antigen–2 (NCA–2):

In 1973, Burtin at al.(75), reported a cross-reactive antigen in meconium and faeces, which seemedto be similar to NFA. They named it NCA–2 to distinguish it from NCA.NCA-2 was detected in me conium and stools but it immunological and physiological characterization

of 44


were performed only on antigen isolated from meconium. Two molecular species of CEA relatedantigen (NCA and NCA-2) exist n me conium in contrast to4 molecular species (FA-1, NFA-2, NFCAand NCA) in adult faeces. NFA-2 and NCA-2 are considered the counterparts of CEA in adult faecesand meconium respectively. The molecular weight and the content of carbohydrate and amino acidsamong these substances are indistinguishable from those of CEA. However, antigentically each ofthem has a unique determinant.

These findings lead to a suggest that NCA-2 can be a protype of CEA glycoprotein family and that itsplace will be taken by NFA-2 in adult tissues and by CEA in tumour tissues. The presence of smallamount of molecules possessing the NFA-2 distinctive determinant in NCA-2 preparation may reflecta differentiation phase at the neonatal period during which most molecules remained as protype(NCA-2) but a small portion of them are differentiated into the adult type NFA-2 (62)The molecular weight of purified NCA-2 was estimated to be 150,000 to 170,000 Daltons. AlthoughNCA-2, NFA-2 and CEA from tumour tissues were indistinguishable from each other with conventionalanti-CEA antiserum, the CEA distinctive determinant was absent in NCA-2 and the NFA-2 distinctivedeterminant was detectable in only 20% of NCA-2 molecules (76).

NCA-2 is the CEA gene family member that is structurally most similar to CEA. There are evidencesthat this antigen is increased in the serum of many cancer patients (77). Table (4) ShowsCarbohydrate composition of CEA and CEA - related antigens.(62) Table (5) Shows Amino acidcomposition of CEA - and CEA related antigens.(62)

Table (4): Carbohydrate composition of CEA and CEA - related antigens:Percent of sample weight

MONOSACCHARIDE CEA NFA-2 NCA-2Focuse 10.2 7.8 6.5

Mannose 8.9 6.7 5Galactose 9.3 11.2 11.9Sialic acid 3.4 2.4 3.1

N-acetyl glucosamine 19.3 16.6 17.6N-acetyl galactosamine 1.6 2 1.5

(Matsouka et al 1982)(62)

Table (5): Amino acid composition of CEA - and CEA related antigens: Mole amino acid / 10 g protein

Amino Acid CEA NFA-2 NCA-2Asparagine 147 142.5 140Threonine 80.6 88.1 89.2

Serine 85.9 94.5 90.2Glutamic acid 99.7 106.1 113.1

Glycine 52.8 52 53.4Alanine 55 55 55.5

Half-cystine 2.1 13 10.6Valine 65.2 68.4 73.7

Methionine 0.9 0.7 0.4Isoleucine 48.6 50.1 51.3Leucine 79.3 88.1 92.3Tyrosine 39.1 45.2 42.1

Phenyl alanine 23.7 25.1 28.7Lysine 23.8 241 24.2

Histidine 15.3 15.6 20.1Arginine 35 37.3 43.2Proline 72.6 78.1 75.6

Tryptophan was not determined(Matsouka et al 1982)(62)

of 44


Meconium Antigen (MA):

Primus et al. (1983)(47), identified 2 NCA unrelated, CEA cross-reactive substances in meconium.One of them was termed meconium antigen (MA). The relationship or identity of MA with NCA-2, andNFA family in adult faeces, particularly NFA-2 was unclear. Preliminary characterization of thisantigen showed it to be a glycoprotein migrating as an alpha globulin and having a molecular weightsimilar to that of CEA (185,000 – 200,000 Daltons). Meconium antigen shared at least 3 epitomeswith CEA. However, the observation that MA appeared at about 6 fold higher concentration than CEAin meconium could indicate that this antigen was an earlier marker of differentiation than CEA. Hedin et al. (1983)(78), considered that NCA-2 was the meconium antigen and was unable todistinguish it from CEA using a monoclonal antibody enzyme immunoassay for serum CEA.

Two meconium antigens of molecular weight 165 KD and 105 KD were isolated ,their amino acidsequence was similar to that of CEA.(63)(79).

Zoubir et al. (1990)(80), considered that the molecular weight of meconium antigen was 160 KD andtermed it (NCA-160).

Blood Group Determinants:

Blood group determinants have been found in CEA preparations. These finding are consistent withthe idea that incomplete or unexpected glycosylation patterns occur in glycoproteins produced bytumour cells. Since antibodies directed against blood group substances cross-react withcarbohydrate determinants on CEA, clinical determination of CEA or anti-CEA levels in serum may beadversely affected. The reported occurrence of natural antibodies to CEA may be an artifact causedby the binding of anti-blood group immunoglobulins form one patient to the radio labeled CEA ofanother. The blood group antigens that cross-react with CEA include A1,B,H,Lea,Leb and MN bloodgroup determinants (81)(82).

Carcinoembryonic antigen in clinical practice

CEA survived the early over enthusiasm of commercial promoters to become the most widely used ofthe so-called tumour markers. Although its failure as a screening test for colorectal cancer wasresounding, today CEA plays a major role in staging, and particularly the monitoring of a variety ofmalignant processes. The value of CEA as a tumour marker lies in the fact that when it is expressedits level in serum usually accurately mirrors the state of the malignant process. A clinical response totreatment weather surgical intervention, chemotherapy, or radiation therapy, is usually matched bya decline in the serum CEA level. Total remission is signalled by a fall in the CEA level to within thenormal range. Use of serum CEA levels to correlate with tumour burden or to stage malignancies onless form ground (35).

Carcinoembryonic antigen in serum: Immunoassay for CEA in Serum

Thomson et al. (1969)(83), developed a radioimmunoassay technique, which could detect 2.5 mgCEA/ml. Egan et al. (1972)(84), interested in the CEA concept developed other techniques.Commercially available immunoassays for CEA include tests with radio-labelled antigens and withenzyme-labelled antibodies. The radio immunoassays are competitive binding tests in which radiolabelled CEA. Competes with endogenous CEA for binding sites on antibodies to CEA. Afterseparation of the bound form the free label, the bound portion is counted with a gamma counter,and unknown specimens are measured against a standard curve to obtain results.(84)

The enzyme immunoassays employ a sandwich procedure in which two antibodies to different siteson the CEA molecule are incubated with serum or fluids suspected to contain CEA. One antibody isbound to a solid phase; the other is labelled with an enzyme, usually horseradish peroxides. After theCEA sandwich has been formed and extraneous material has been washed away from the solidphase, a containing a chromo graphic substrate for the enzyme label is added and the colorgenerated is proportional to the amount of the enzyme and thus to the level of CEA present.(84)

of 44


The antibodies used in these assays are raised in different ways. Some are guinea pig or goatantibodies generated using CEA isolated form tumour material; others are monoclonal, raised usingCEA form tumour cells grown in tissue culture. It is inevitable that a variety of immunoassays willshow differing specificities and avidities for antigenic sites on glycoprotein molecule that may notitself are homogeneous (84).

Serum CEA Levels in Normal Individuals

Martin et al. (1976)(32), found that plasma CEA levels in normal individuals were less than 2.5 mg/mlin 97% of the studied group. The remaining 3% with elevated CEA levels were heavy smokers.Among pregnant females also 3% had elevated CEA values.

Some investigators considered 2.5 mg/ml as the upper limit of normal because they noted that about90% of normal adults had CEA levels below this limit. No differences in CEA levels related to sex, butthe levels were higher over 40 years of age.(43)(85)(86)

Others considered the normal range of serum CEA from “0” to less than 5 ng/ml.(87)(88).

Tholander et al. (1990)(89), reported that the upper limit of normal for 95% of the non smokers (oftheir control group) was 4.5 ug/litre, and for the smokers it was 7.5 mg/litre.

Meier et al. (1990)(90) and Leminen (1990)(91), considered that normal serum levels for CEA werebelow 3 ng/litre. While, Lundqvist et al. (1992)(94), took the value of 5.2 ng/ml as the upper limit ofthe normal range of serum CEA values.

Cut-Off Level for Serum CEA

Elevated levels of CEA do occur in benign and malignant disease. Most of the benign conditions thatproduce slight elevation in CEA antigenicity rarely exceed 10 ng/ml and most of these are between2.5 and 5 ng/ml. Liver cirrhosis may give slightly higher results (32). The level of 2.5 ng/ml is statedto be the upper limit of the normal for plasma CEA levels.

Values greater than 2.5 ng/ml may be found in association with cancer, in particular those ofgastrointestinal tract, pancreas, ovary, lung and breast.

Similarly raised CEA levels may, however, be detected in cigarette smokers, in patients with benignneoplasm, and in 15% to 20% of subjects with inflammatory disorders such as , Chron’s disease,pancreatitis, liver disease and pulmonary infections. Thus raised plasma CEA values are not specificfor cancer, although very high levels (e.g. > 20 mg/ml) are highly suggestive of malignancy. It isimportant that serial assays of CEA are used in reaching a clinical judgment, and not any singledetermination. Each laboratory doing CEA assays should establish its own “normal range”. Therecommended upper level of normal (2.5 mg/ml) in the population requires additional evaluation(93)(94). The level of CEA defined as the upper limit of normal markedly affects the results obtained.Raising this limit to 5 mg/ml reduced the number of elevated values in patients with benign diseaseto only 4%. Also, 59% if patients with intraepithelial cancer and 43% of patients with invasive cancerhad been excluded on taking this level. Using the limit of 2.5 mg/ml permitted the inclusion of thesepatients in the positive group. On the other hand, 11% of normal healthy volunteers and 18% ofpatients with benign gynecologic disease had plasma CEA concentration above 2.5 mg/ml. This lackof specificity makes CEA unreliable as sole screening method for patients with gynecologicmalignancy (95).

Seppala et al. (1975)(96), considered the serum CEA level 5 n/ml as a discriminating one betweenmalignant and non-malignant states. By setting this higher threshold greater specifity for cancer wasachieved, but at the cost of missing some cases, the cut-off level of 5 mg/ml was also recommended(88)(97)(98).

Serum CEA in Gynaecological Malignancies:

of 44


Lo Gerfo et al. (1971)(30), were the first to report elevated serum CEA in patients withgynaecological malignancies.

Estimation of plasma CEA levels doesn’t appear to be a worthwhile screening procedure, because ofthe detection of CEA in non-cancerous conditions. However, it is valuable in the diagnosis of somecancers (45)(99).The most significant applications of CEA to gynecologic malignancy would be in the diagnosis andmanagement of patients with ovarian cancer. Carcinoembryonic antigen from the plasma and cystfluid of patients with ovarian cystadenocarcinoma had been shown to have immunologic cancer.Elevated CEA level correlated guide well with malignancy. Elevated CEA level correlated guide wellwith malignancy. In those patients with ovarian cancer who had elevated CEA levels prior to therapy,serial CEA determinations following treatment had been reported to be of value in diagnosingnon-palpable recurrent disease(95).

Disaia et al.(1977)(100), studied a group of patients with invasive squamous cell carcinoma of thecervix and found that there was a progressive increase in the percentage of patients with positiveCEA values with the advance of the stage form 26% in stage I to 88% in stage II. Eighty five percent(85%) of the recurrent cases showed a positive CEA values.

Difficulty of marking diagnosis of recurrence is encountered particularly in patients who hadreceived radiotherapy, the presence of radiation fibrosis often makes the detection of tumoursimpossible. In these patients, the CEA assay could be of considerable value. (99)(100).

Three significant patterns of change in serum CEA levels have been shown. The first is the pattern offree disease, when there is a rapid disappearance of CEA in serum after treatment. The second isresidual cancer, if CEA is persistently detected in serum often at high levels. The third is clinicalrecurrence that is detected during close follow up of patients when CEA reappears in the serumbefore clinical detection (99). It appears that the greatest value of CEA assay lies in monitoring the results of primary treatmentand in the follow of patients whose tumour or plasma contain high antigen concentrations beforetherapy. Characteristically, plasma CEA concentrations decrease to normal levels within 4-8 weeksfollowing complete excision of cervical or ovarian tumours. In contrast, the decline of plasma CEA to normal often takes as long as 12 weeks after completion ofradiation therapy. This is presumably due to the prolonged release of antigen form the cellmembrane of radiation damaged tumour cells. The progressive increase in plasma CEA levels, duringthe follow up patients with the cervical and ovarian cancers, accurately predicted tumour recurrencein over 80% of patients whose tumour strained immunohistochemically for CEA. Rising plasma CEAvalues preceded clinical detection of recurrence by up to 6 months cases (101).Parente and Greston (1981)(102), found elevated serum CEA levels in 60% of the patients withmalignant lesions involving the female reproductive tract. They considered that CEA assay was notsuitable for screening of malignant tumours because of the large number of patients with benignhaving false-positive results.

Schwartz et al.(1987)(103), reported an elevated levels of CEA in the serum of 30-65% of ovarian,43-69% of cervical, 32-57% of vulvar cancer.

The use of CEA in the diagnosis and follow up of patients has been hampered by conflicting reportsin the literature. The proportion of ovarian carcinoma patients having an abnormal serum levels (. 5mg/ml) during the course of the disease ranges form 0 to 70%, and a variable relationship ofelevated CEA to histologic cell type was observed (104). The differences in clinical correlations mightbe due to effects of antigens cross-reacting with CEA, (e.g.: NCA, NCA-2 and BGP (105). The selectionof patients for serial plasma CEA evaluation could probably be improved by selection of patients withCEA detected in tumour tissue. Unfortunately, tissue distribution studies have shown markeddivergence as to the proportion of the cased and the type of epithelial ovarian cancer associatedwith CEA (106). Although carcinoembryonic antigen initially stimulated great interest as a possiblediagnosis marker for common epithelial malignant tumours, including gynecologic malignancies inonly a small percentage of gynaecologic cancer patients and also in some patients with benigndisease. Thus, limiting its value as a clinically useful tumour marker (107).

of 44


Carcinoembryonic Antigen in Tumour Tissues: ImmunocytochemicalDetection of CEA

Immunocytology has provided significant contributions to biological and science since its inceptionby coons et al. (1942)(108), because its unique ability to correlate the location of antigen cellularconstituents with anatomical structures. Although immunofluorescence remains the most widelyused immunocytological technique, the development of immunoenzyme procedures in which the siteof immunological reactivity is detected by histochemical reaction has permitted the localization ofnumerous substance of biological and medical interest. Most studies have used immune-peroxidestechnique. Unlike immunofluorescence, specimens stained by immunoperoxidase are permanent, can becounterstained with routine histology stains allowing direct comparison of immunologic reaction withmorphologic details, and don’t require special tissue manipulations if antigen activity is preservedfollowing formalin fixation and paraffin embedding (109).After the discovery of CEA attention was focused primarily on detection of CEA in the blood ofsuspected or proven cancer patients. It seems reasonable to determine first if the antigen is indeedpresent in the original malignant lesion before extensive and repeated monitoring is undertaken. Positive immunocytochemical reacting for CEA in the primary tumour predicts well, which cases willhave a blood CEA value reflecting disease activity. A negative CEA immunocytochemical reactiondoesn’t mean tat the tumour is devoid of CEA, but only its CEA concentration is below the thresholdof the staining reaction (110).CEA producing tumours and their metastases could be localized using photoscan detection ofradiolabeled anti – CEA antibodies even when these tumours were undetectable by otherconventional diagnostic procedures.(111)

Immunocytochemical detection of CEA in metastatic tumour tissue may be helpful in suggesting thesite of the primary tumour (105).

Techniques of Immunoperoxidase Staining for CEA: Three-Step UnlabeledAntibody Technique

The initial demonstration of CEA in conventional tissue

Sections with immunoperoxidase were done (112) using the 3-step, unlabeled, antibody technique.This method starts with reacting the primary goat CEA antiserum with the antigen containing tissuesection. After incubation, excess antiserum is removed and the tissue is treated with the “bridging”secondary antiserum, rabbit anti-goat IgG, which attaches to the primary CEA antibody with one ofits two combining sites.The remaining combining site is then free to react with the last antiserum addition,goat-antiperoxidase antibody.Both of the combining sites of the later antibody are free to react with horseradish peroxides and thesites of antigen localization are detected by a brown reaction with diaminobenzidine and hydrogenperoxide. Prior to dehydration and mounting, the tissue sections can be treated with a nuclear orcytoplasmic stain although non-counter stained tissue section may be required in specimensshowing weak CEA reactivity.(112).A variation of the 3-step procedure is peroxidase-antiperoxidase method (PAP method) ofSternberger (1974)(113). This method simplifies the 3-step procedure since performed complexes ofperoxidase-antiperoxidase are added to tissue sections previously treated with the primary andsecondary antibodies. Thus, eliminating the incubation step with free peroxides (figure 4)(114). Fig (4): Simplified Schematic Representation of Immunoperoxidase Reactions:

3- step, unlabeled antibody (A).3- step unlabeled antibody PAP, (B).2- step, unlabeled antibody (c)1- step, unlabeled antibody (D)(Primus et al 1978)(114)

of 44


Two Step Labelled Antibody Techniques:

The most commonly used immunohistochemical method, in which the tracer peroxides, is chemicallyconjugated to the second antibody (rabbit anti-goat LgG).This shortens the time required to perform the staining procedure and also obviates some of thecomplications associated with antibody dilutions encountered in the 3- step methods (114).

One Step Technique:

For immunoelectron microscopy, the direct or one-step method is used, because of the difficulties inantibody penetration, tissue manipulation procedures and structural preservation met within othermethods (114).

Threshold CEA Tissue Staining Reaction

The 3 –step unlabeled antibody immunoperoxidase reaction for CEA has been found to detect CEAconcentrations above 0.7 mg/g tissues in frozen, ethanol fixed specimens, and above 3-5 mg/g informalin – fixed paraffin embedded tissues (115)(116).

of 44


Tumour Tissue Susceptibility for CEA Staining:

There is a proclivity for CEA staining in certain tumour type, and also among different populations oftumour cells within a lesion.

First, tumours of epithelial character are particularly prone to CEA staining.Secondly, of epithelial tumours adenocarcinoma and squamous cell carcinoma express CEA.In adenocarcinoma CEA is most closely associated with the apical border of the cells liningthe lumen of the malignant glands or cysts.Finally, in cystadenocarcinoma of the ovary, tumours of the mucinous variety stain for CEAmore frequently than those of the serous types (115).

Carcinoembryonic antigen staining was most commonly seen in the more keratinized differentiatedareas of the squamous carcinomas. The localization of tissue CEA in more differentiated areas, andthe tendency of elevated serum levels to occur predominantly in poorly differentiated cases iscontroversial and can not readily be explained (117)(118).In lesions with more undifferentiated cell forms, tissue CEA may be masked and therefore difficult tobe demonstrated immunohistochemically, since antibody molecules may not combine with antigenmaterial latent molecules may not combine with antigen material latent in tissue sections.Pre-treatment of tissues with trypsin unmasks latent cellular antigens and renders them readilydemonstrable (119).Ogawa et al (1992)(120), found faint expression of CEA in a few superficial and intermediate cells ofnormal cervical squamous epithelium. However, CEA was more strongly expressed in cancer tissuesof the large cell and keratinized types.

Light and electron microscopic findings in immunostained tumours for CEA

The majority of the light microscopic studies have shown a linear labelling of the apical poles ofadenocarcinoma cells lining the glandular lumen and staining of intraglandular deposits. A diffuse orgranular cytoplasmic label is specially evident in anaplastic tumour cells. Immunoelectronmicroscopy studies with ferritin or peroxidase – labelled antibodies revealed an intimate associationof CEA with the glycocalyx. However, other evidences suggest that CEA is also a structuralglycoprotein of the plasma membrane (114).

Carcinoembryonic Antigen in Cervical Neoplasms

Tissue CEA in neoplasms of the cervix:

Carcinoembryonic antigen has been demonstrated in invasive squamous cell carcinoma andadnexorcarcinoma of the cervix (121).CEA was present in all adenosqaumous carcinomas, but only in 77% ofadenocarcinomas.(122)CEA content is related to cell type in primary and metastatic cervical cancer and may varywidely between patients. In general keratinizing squamous cell carcinomas andadenocarcinomas contain the highest antigen levels and antigen distribution is quite uniformthroughout the tumours. Conversely, non keratinizing squamous cell tumours contain littleantigen and the distribution is uneven CEA content and distribution are similar in primarycervical cancers and in their lymph node metastases.

Therefore, immunohistochemical staining of the primary tumour for CEA can help identify thosepatients who benefit most from lymphoscintography (101).Different degrees of positively to CEA were found according to the histological types of cervical Adencarcinomas. All cases of endometriosis – type Aden carcinomas were strongly positive with anti-CEA,while in the end cervical type Aden carcinomas the less differentiated cases were strongly positivewith anti-CEA than the well differentiated ones. Immunohistochemical reactions with anti-CEA weredetected in the lesions of Aden carcinoma in situ. However, Aden carcinoma in situ of the cervix has2 histological types; common or endometriosis and end cervical. The endometrioses type

of 44


adenocarcinoma in situ exhibited a more positive reaction with anti – CEA than the endocervical type(123).

Tissue CEA in Rare Cervical Neoplasma:

True mesonephric adenocarcinoma of the cervix appears to be a genuine, rarehistopathologic entity.Immunohistochemical staining for CEA doesn’t appear to be helpful in distinguishing thisvariant of cervical adenocarcinoma from the more common type, as mesonephroidadenocarcinomas of the endocervix are CEA negative (124)(125).Although CEA is rarely demonstrable in benign lesions of the cervix, it was found in a veryrare case of villous adenoma of the uterine cervix. This finding is consistent with the positivestaining of colonic villous adenomas mentioned in literature. Carcinoembryonic antigen isconsequently note useful in distinguishing villous adenomas from exophitic papillary cervicaladenocarcinomas (126).Adenoma malignum of the cervix is an extremely rare variant of adenocarcinoma. It is apotentially malignant tumour histologically characterized by adenomatous proliferation withstructural abnormality in the shape and arrangement of the glands, and budding invasivepattern in the stroma. The glands filled with mucous substances often ballooned andruptured with leakage of mucous into the stroma. Its diagnosis is rendered very difficult by itsstrong similarity to benign endocervical hyperplasia.Immunohistochemical staining for CEA has proved to be of great help in the differentiateddiagnosis (127)(128).CEA antigen present in the columnar cell of adenoma malignum, and absent in the normalendocervical glands.(129).The finding of positive CEA staining in the cytoplasm of adenoma malignum cells led tofurther study to detect CEA in the cervical mucous in careening for this tumour (128).

Patterns of Immunohistochemical Staining for CEA in Different Histological Types ofCervical Malignancy:

Immunoperoxidase histochemical techniques allow the visualization by light microscopy of tumourCEA when more than 3-5 mg/g are present. Tumour concentration of CEA correlates well withimmunoperoxidase staining. Plasma CEA titre doesn’t necessarily reflect tumour CEA content butrather total tumour burden (i.e. tumour CEA concentration x tumour mass). Over 30% of cervicalepidermoid carcinomas show positive CEA immunoperoxidase staining, whereas non of the cases ofcarcinoma in situ or dysplasia is positive(133). Wahlstrom et al (1979)(130), found immunoreactivityto CEA in 80% of cases of cervical adenocarcinoma (131).

Squamous Cell Carcinoma:

Staining for CEA was detected in the atypical epithelium on or about cell membranes with somecytoplasmic localization. Normal cervical epithelium, as well as areas of reactive basal hyperplasiaand squamous metaplasia, did not show any appreciable staining. In carcinoma in situ involvingendocervical, CEA was found with maximum intensity near the lumen and was negative near thebasal layer. Grading of non invasive lesions did not have a significant impact on the degree of thepositive reaction except that mild dysplasia did not show a strong positivity (132).

CEA was demonstrated in all studied cases of cervical intraepithelial neoplasia and in 70.6% ofinvasive squamous cell carcinomas. A heterogeneous pattern of staining was observed in differentcases and also within the tumour itself.(133)CEA was not expressed in normal squamous epithelium except for a faint expression in a fewsuperficial and intermediate cells.(120)CEA is an oncogenic antigen of cervical squamous cell carcinoma due to its strong expression inthese tumours. No significant difference in immunoreactivity was observed between primary andmetastatic lesions of carcinoma or between primary lesions with and without metastasis. Eightyeight percent (88%) of invasive squamous carcinomas showed positive staining. They exhibitedgreater variability in the number and topography of CEA positive cells than invasive lesions. Morethan half (57%) of invasive squamous carcinomas showed positively stained cell in the layers

of 44


adjacent to stroma, a feature designated as “CEA marginal effect”. This emerging pattern of CEAtissue distribution may be used to differentiate between non invasive and invasive squamouslesions. The degree of positive staining for CEA depends to some extent on the cellular types ofinvasive squamous carcinomas. Small cell carcinomas frequently lack stainable CEA while large cellnon keratinizing carcinomas show a stronger reaction than their keratinizing counterparts.(120).Metastatic lesions show the same CEA pattern as the corresponding primary tumour including themarginal effect. Goldenberg et al (1978)(110), utilized this phenomenon , proposed a photo scanningtechnique for the detection of metastatic spread of cervical carcinoma using radiolabeled ofmetastatic spread of cervical carcinoma using radiolabeled anti - CEA antibodies (110)(132).

Dallenbach and Lang (1991)(134), depending on histological and immunohistochemical examinationsubdivided invasive carcinomas and preinvasive lesions of the uterine cervix into the squamous(large cell, ectocervical) type and the reserve cell (small, large of clear cell, endocervical) type.Immunohistochemical all of the invasive and preinvasive squamous cell carcinomas were negativewith anti -CEA. Whereas most of the invasive and preinvasive reserve cell carcinomas acquiredexpression of CEA.

Toki and Yajima (1991)(135), investigated the relationship of the positive pattern of CEA staining insquamous cell carcinomas of the cervix, to the prognosis of the patients with the neoplasm. Fourpatterns of localization of CEA were observed depending on the location of CEA - positive areas incancer nests: central type, surrounding type, diffuse type and focal type. The prognosis of patientswas most excellent in the disease of central type, diffuse type and focal type.(135).The associations of CEA, tissue polypeptide antigen (TPA) and squamous cell carcinoma antigen(SCC-Ag) for detection of cervical cancers with type of cancer, clinical stage and lymph nodemetastasis were studied. The sensitivities of CEA, TPA, and SCC-Ag were 15.2%,33.6%,and49.3%.The specificities were 89.3%,96.2%,and 95.1% respectively. The positive rate of CEA wasfound significantly higher among patients with adenocarcinoma than those with squamous cellcarcinoma. The positive rate of CEA and SCC-Ag increased with advance of the clinical stages. Therewas no significant in the rate of each tumour marker between patients with lymph node metastasisand those without metastasis.(136).A total of 120 patients with squamous cell carcinoma of the cervix were prospectively monitoredwith simultaneous serum level estimations of squamous cell carcinoma antigen (SCC), CEA, andtumour-associated trypsin inhibitor (TATI) in different stages and phases of their disease. Positivevalues were observed for SCC in 72 patients(60%),for CEA in 47(39%).and for TATI in12(10%).Inaddition,28 patients, who had a renewed progression after a complete remission with normal markerlevels, showed an increase in SCC in 26 cases(92%) and CEA in 7(25%).(137).IN other study SCC was raised in 65% of squamous cell carcinoma of the cervix ,CEA was raised inless than 25%.(138).Pretreatment of SCC in conjunction with CEA is a valuable tumour marker topredict the prognosis of squamous cell carcinoma of the uterine cervix and to foresee a clinicalresponse to subsequent neoadjuvant chemotherapy.(139).

Adenocarcinoma:

Carcinoembryonic antigen immunohistochemistry may be a valuable adjunctive tool for thediagnosis of endocervical adenocarcinoma, anything more than focal positively clearly signifyingmalignancy. Negative staining strongly suggests malignancy especially if the tissue sample is large.Sixty four percent (64%) of adenocarcinomas are positive for CEA. The staining is often locallyintense near the apex of the columnar cells in well - differentiated lesions, but is more diffuse in themore poorly differentiated specimen. Not all the cells are positive in any given tumour, and theextent of differentiation of the cells.Luminal staining representing sloughed cells and/or secretion products is occasionally intense. Mostof the cases (95%) of microglandular hyperplasia are negative for CEA, few only show focal positively(140).A review of the literature concerning tumour markers used in adenocarcinoma of the cervix showsthat squamous cell carcinoma antigen, tissue polypeptide antigen, CEA, CA 125,and a few otherssuch as CA15.3,TAG72,trypsin inhibitor have been studied(140)(141)(142).There is, however, noconsensus on how useful tumour markers in adenocarcinoma of the cervix, nor is there any clearagreement on which one to use. Ngan et al 1998(144), reported that CEA was raised in 26% ofpatients with adenocarcinoma of the cervix, SCC was raised in 25%,tissue polypeptide antigen was

of 44


raised in 35%,and CA 125 was raised in 27% of these patients.(142).Kudo et al. (1990)(141), reported that 50% of in situ cervical adenocarcinomas and 55% of invasiveadenocarcinomas had positive staining reaction for CEA. Whereas 67% of the adenosqaumouscarcinomas showed a similar reaction. They also mentioned that CEA was less effective in identifyingpoorly differentiated lesions.

Distinction Between Endocervical and Endometrial Adenocarcinoma UsingTissue Staining for CEA

The distribution of CEA as determined by immunohistochemistry has been suggested to be useful inthe distinction between endocervical and endometrial adenocarcinomas. Eighty percent (80%) ofcervical adenocarcinomas and 8% of endometrial adenocarcinomas are CEA positive. Thecommonest exceptions are endocervical mesonephroid adenocarcinomas (which are CEA negative)and endometrial adenosqaumous carcinoma (which are CEA positive) (130).Maes et al (1988)(143), using well absorbed anti-CEA serum or monospecific monoclonal anti –CEAantibody, found no difference in the expression of CEA between endocervical and endometrialadenocarcinoma.

The majority of endocervical adenocarcinomas were showing abundant diffusely distributedintracellular CEA and mucin, and luminal secretions of acidic mucosubstances. Only half of theendometrial adenocarcinomas contained CEA, usually focally distributed in small amountsparticularly at the apical cell surface, with focally distributed mucin in luminal secretions.(131).

96% of endocervical and 70% of endometrial adenocarcinomas revealed positive CEA immunostain.The extent of staining did not seem to correlate with the degree of differentiation in either kind ofcancer. Among the positive cases, CEA staining was located over whole cytoplasm in 80% of positiveendocervical adenocarcinomas and on apical surface in 75% of positive endometrialadenocarcinomas. The pattern of CEA distribution in endocervical adenocarcinoma of endometrioidtype was not similar to that observed for primary endometrial adenocarcinoma.(144).

Uedal et al (1983)(145), using immunoperoxidase stain for CEA, realized that all endocervical adenoand adenosqaumous carcinomata revealed the presence of CEA, while a mesonephroidadenocarcinoma did not. The antigen was not observed in the glandular structures of endometrialadenocarcinomas, although CEA was detected in all foci of squamous metaplasia within thesetumours. CEA was also located in endocervical type of glandular epithelium within a specialendometrial adenocarcinoma containing predominantly this type of epithelium. The staining patternfor CEA in the endocervical adenocarcinoma was related to the degree of histological differentiationof tumours, in the well differentiated glandular structure CEA was located on the luminal surface,while it was detected in the whole cytoplasm of tumour cells within the moderately and poorlydifferentiated areas. Finally, they concluded that the immunoperoxidase stain for CEA would beuseful for estimating malignancy of glandular structures within the uterus, distinguishingendocervical adenocarcinoma from endometrial one, and grading of histological differentiation ofendocervical adenocarcinoma.Endometrial adenocarcinomas did not show reactivity with CEA immunostain, while all sections fromendocervical adenocarcinomas were positive for CEA. CEA was considered as a useful tumour markerin differentiating endocervical and endometrial adenocarcinomas.(133).

The discrepancy of the results of different may possible be attributed to variations in the specificityof polyclonal anti-CEA sera and the influence of cross- reactivity of these antisera with normal tissueantigens such as NCA 1, NCA2, and BGP 1 (143).

Serum CEA in Cervical Neoplasms

Plasma CEA levels are significantly elevated > 2.5 mg/ml in 48% of patients with cervical cancerprior to therapy as opposed to 11% plasma CEA elevations in healthy volunteers. Carcinoembryonicantigen values greater than 5 mg/ml are obtained in 24% of patients with cervical malignancy asopposed to 1% in healthy volunteers (`146).

of 44


The incidence of elevated plasma CEA is directly related to the stage or extent of the disease. Fortytwo percent (42%) of patients with stage I cervical cancer had increased plasma CEA concentrations,whereas 60% of patients with advanced stage disease had elevated antigenaemia. Plasma CEAlevels are more consistently elevated in patients with endocervical adenocarcinoma that in patientswith squamous cell carcinomas, CEA is more often increased in patients with large cell tumours thanthose with small cell lesions. There is a trend toward increasing frequency of abnormal plasma CEAvalues with progressive tumour dedifferentiation. Only 36% of well differentiated cervical tumourshad elevated plasma CEA levels as compared to 58% of poorly differentiated lesions (146).

Cut off Level for Carcinoembryonic Antigen in Serum

When a level of 2.5 mg/ml is taken as the cut-off point, an acceptably high false positive rate isdetected in normal controls e.g. 11% (97) ,10% (102), and 18% (111). Ito et al (1978)(147), foundthat the mean CEA level for normal controls was 2.03 mg/ml with a standard deviation of 0.49.Hence, they considered 3.01 mg/ml to be the upper limit of normal for serum CEA levels.

When a CEA level of 5 mg/ml is taken as a cut – off point, the average proportion of patients withhigh levels varies from 63% in cancer of the ovary to 36% with cervical cancer compared to 0% incontrols (148)(149)(150).

A cut – off value of 5 mg/ml is more appropriate than 2.5 mg/ml although such a high cut – off levelwill limit the possible value of CEA as a monitoring parameter post-treatment. However, if a lowercut-off value is used, e.g. 2.5 ng/ml, many cases with non specific elevations will be included, whichwill make interpretation of the test difficult and further discredit the use of CEA determination forpractical purposes (97)(151).Cauchi et al (1981)(152), found that 38% of controls had high values, compared to 54% of cancercervix patient when a cut – off value of 2.5 ug/1 was used. But with a 10 ug/1 cut-off point, non ofthe controls had a high value, compared to 8% of cancer patients.

Hansen et al (1974)(153), and Franchimont et al (1976)(156), also emphasized the importance oftaking CEA > 10 ug/1 as the cut – off level.

Braun et al (1981)(155), suggested a cut – off level of 12 mg/ml as this was the highest level of CEAencountered in a group of healthy volunteers in their study.

Leminen (1990)(91), Meier et al (1990)(90) and Leminen et al (1992)(156), took the value of 3 mg/mlas the cut-off level for serum CEA in cases of cervical carcinomas. While Lundqvist et al (1992)(92),considered 5.2 mg/ml as the suitable cut-off point.

Pre – Treatment Serum CEA in Cancer Cervix

Elevated serum levels of carcinoembryonic antigen have been regarded as an unfavourableprognostic sign in malignant disease. It appears to correlate with the total tumour burden, but it isnot known whether CEA positive tumour cells are more aggressive than those that do not expressCEA (157).

Pre-treatment serum levels of CEA exceeding the cut-off limits were the highest in advanced diseasestages and in highly differentiated forms of cervical cancer (158).TeVelde et al. (1982)(159), indicated that a pre-treatment level exceeding the individual upper limitof normal was associated with a very poor prognosis as it was encountered in about 40% of allpatients who developed recurrence.

However, Braun et al (1981)(160), could not establish a relation between the pre-treatment plasmaCEA levels and the probability of recurrence after treatment of cervical cancer.

Yokoyama et al (1987)(161), found elevated pre-treatment plasma CEA level in only 19% of patientswith cervical cancer.

of 44


Effect of Histological Type of Cervical Carcinoma on Pre- Treatment SerumCEA Levels

It is important to differentiate between tumours of different histological types. Adenocarcinomas ofthe cervix are basically different from squamous cell carcinomas in their ability to produce highlevels of plasma CEA. So a high pre-treatment level in patients with adenocarcinoma doesn’tnecessarily correspond to the high recurrence rate found in patients with squamous cell carcinomawith grossly elevated values, because these 2 tumours are biologically different and may responddifferently to treatment. In approximately 50% of all patients with squamous cell cancer of thecervix, CEA will be of predictive value. Pre-treatment levels over 5 ug/1 are highly suggestive ofmetastatic disease as they are associated with metastases in pelvic or para-aortic lymph nodes in50% of patients with stage IB disease. Also in advanced disease such as stage III and IV 48% ofpatients had a pre-therapy value exceeding 5 ug/L (151)(162).Plasma CEA is elevated in 48% of patients with invasive carcinoma of the cervix uteri. Thepercentage of patients with positive CEA values rises appreciably in those with advanced squamouscell carcinoma of the cervix uteri. Positive CEA values were detected in 92% of patients withrecurrent squamous cell carcinoma of the cervix (163).

Carcinoembryonic antigen is a reliable tumour marker in patients with adenocarcinoma of the cervix.An excellent correlation has been noted between pre-treatment plasma levels and extent of thedisease, high pre-treatment values were associated with bad prognosis. Contrary to squamous cellcarcinomas, a low value is equivocal, meaning either that the tumour volume is small or that thetumour releases little no CEA. The presence of lymph node metastases has more pronounced effecton the CEA values than the size of the primary tumour. It is unlikely that metastases have differentcapacity for CEA production than the primary tumour.The reason for the lack of antigen release from some primary tumours initially and the laterappearance of CEA in serum with recurrent disease remains unknown. These tumours are moreprobably non-releasers than non-producers (151)(164).

Meier et al (1990)(90), found an elevated serum CEA level (> 3 ng/ml) in 54% of the cases ofsquamous cell carcinoma of the cervix and in all cases of adenocarcinoma and adenosqaumouscarcinomas.

Leminen (1990)(91), reported and elevated pre-treatment serum CEA level (>3 ng/ml) in 36% and33% of the cases of adenocarcinoma and adenosqaumous carcinoma respectively. They noted andincrease in the incidence of cases with positive serum CEA levels with the increase of grade ofmalignancy being 33% in grade I, and grades II & III They recommended the determination ofpre-treatment levels of serum CEA and its subsequent use for monitoring of treatment even if theinitial level was negative, because during follow up the tumours may become marker positive.

Effect of Clinical Stage of the Disease on Pre-Treatment Serum CEA Levels

Fritsche et al (1982)(97), using a cut-off value of 4.5 mg/ml found that the incident of elevatedserum CEA in patients with squamous cell carcinoma of the uterine cervix may range from zero –10% for in situ carcinoma, 13-30% for stage I disease, 25-40% for stage II, 40-60% for stage III and30-60% for stage IV disease. Elevated plasma CEA levels (over 2.5 ng/ml) are reported in 36-80% ofinvasive carcinomas of the cervix (both epidermoid and adenocarcinomas). The frequency ofelevated CEA values increases with advancing stage of the disease reflecting tumour burden. Instage l carcinoma of the cervix 21-63% of patients have elevated levels, while in stage IV disease75-100% are plasma CEA positive frequencies of plasma CEA elevations are lower in cervicaldysplasias, with 23% positivity in moderate dysplasia and 28% positivity in severe dysplasia ascompared with 33% positivity in insitu carcinomas. Recurrent cervical tumour result in elevatedplasma CEA levels in 85% of patients (97)(131).

Braun and Leyendecker (1983)(165), taking 12 mg/ml as the upper limit of normal for CEA, foundthat only 14% of patients with stage I and 29% of those with stage II carcinoma of the cervix had

of 44


elevated serum levels. Similar results were also observed in 45% and 73% of the patients with stageIII and IV respectively.

Aaran et al (1988)(166), noted in a prolonged study that CEA levels of cervical carcinoma patientswere higher than those of normal population years before the diagnosis of cervical carcinoma wasmade.

Leminen (1990)(91), found and elevated serum CEA level (>3 ng/ml) in 32% of the cases of cervicaladenocarcinoma (stage I, II) and in 50% of the patients with stages III and IV.

Serum CEA levels were significantly higher among patients with adenocarcinoma than those withsquamous cell carcinoma. These levels increased with advance of clinical stages.(144)(167).

Ito et al (1978) (147),after studying serum CEA in patients with carcinoma of the cervix reached thefollowing conclusions:

1. Determination of serum CEA levels cannot serve as a test for the early diagnosis of cervicalcancer or pre-cancerous lesions.

2. In patients with clinical stage I or II cervical cancer, an elevated serum level indicates agreater likelihood or parametrial invasion or lymph node metastases.

3. In patients with treated carcinoma of the cervix, continued normal levels of serum CEA affirmcontinued absence of the disease, whereas development of an elevated level is highlysuspicious of occult residual or recurrent cancer.

Van Nagell et al (1978)(168), presented the relation between plasma CEA and stage of cervicalmalignancy, cell type and histological differentiation in the following 2 tables (table 6&7)

Table (6): Plasma Carcinoembryonic Antigen Related to Stage CEA ng / ml

Stage < 2.5 2.5-4.9 5-10 >10Ia 41% 36% 23% 0%Ib 63% 27% 9% 1%lla 44% 32% 12% 12%Ilb 57% 18% 16% 9%llla 50% 25% 25% 0%lllb 44% 19% 31% 6%Iva 44% 32% 12% 12%Ivb 33% 33% 1% 33%

Total 52% 24% 17% 7%(Van Nagell et al 1978) (168)

Table (7): Plasma carcinoembryonic antigen related to cell type and histologicaldifferentiation CEA ng / ml Cell type < 2.5 2.5-4.9 5-10 >10Keratinizingsquamous Cell

52% 32% 12% 4%

Small cell 82% 6% 6% 6%Large cell nonKeratinizing

52% 23% 20% 5%

Adenocarcinoma 40% 20% 20% 20%Histological differentiation

Well 64% 21% 14% 1%Moderately 51% 26% 15% 8%

of 44


Poorly 42% 21% 27% 6%(Van Nagell et al 1978) (168)

Post – Treatment Serum CEA in Cancer Cervix

Patterns of Change in Serum CEA Levels following Treatment

Plasma CEA levels returned to normal in over 90% of patients within 4 weeks following completetumour excision. This was in marked contrast to the patterns of decrease in CEA values followingradiation therapy, in which plasma antigen levels often did not return to normal until 16 weeksfollowing completion of therapy. This difference has been attributed to the protracted release ofmembrane associated antigen after radiation – induced cell damage, compared to the abruptremoval of the antigen source by surgery (168)(169).

Kjorstad and Orjasaester (1982)(151), found that the lowest post radiotherapy level was not reacheduntil many months after termination after termination of treatment, and in some patients the valuescontinued to fall for as long as 18 months. However, the CEA level measured immediately afterradiation therapy in patients with initially elevated CEA is indicative of the prognosis of therespective patients.

The median survival time of the patients with persistently elevated CEA levels was significantlyshorter than that of patients in whom initially elevated CEA levels declined into the normal rangeimmediately after therapy (155)(160).

Chemotherapy is often the only available treatment for advanced cervical carcinoma or recurrentdisease, even though the results of treatment in squamous cell carcinoma remain poor. The courseof CEA level can help in early decision, whether or not patient would profit from a continuation oftherapy (170).

A group of patients with cervical carcinoma treated with cytotoxic drugs were followed up for years.In more than 90 % of the patients CEA was elevated before the treatment, so that the course of thetumour marker levels fell rapidly to normal after one or two cycles. Non of these patients showedtumour progression while the tumour marker levels were within the normal range. Clinical remissionwas not obtained in those patients with levels, which remained high or rose again following an initialdecrease.(170).In cases of primary progression of cervical carcinomas, serum CEA levels decreasefor a short time initially following chemotherapy, but never approach normal values. In all cases, thelevels increase steeply after 2-3 months. In all these cases, chemotherapy should be discounted,even if the clinical and radiological examinations have not yet demonstrated the progression. Thecourse of CEA can help to decide early whether patients will benefit from continuation of treatmentor not. If the levels are not within the normal range after, at least, three cycles of treatment, thechemotherapy should be stopped if the tumour marker levels are normal. It should be restarted onlywhen the tumour marker levels begin to increase again (170).

Leminen et al (1992)(156), studied the clinical and tumour markers response among the cases ofcervical carcinoma which received chemotherapy as an initial treatment.The overall clinical response rate to initial chemotherapy was about 78% (80% in early and 78% inadvanced disease), according to WHO recommendations (171), a complete response (disappearanceof all clinical signs of malignancy) was achieved in 13% of the patients, and partial response(decrease in tumour size of 50% or more for at least 4 weeks) in 65% of the cases. A decreasinglevel of CEA had partial response, and 25% had no response to treatment. So they concluded thatCEA often respond to initial chemotherapy and a decrease in its level was often associated withclinical response.(171).Post-Treatment Follow Up with Serum CEA

CEA measurement can be used for follow up. In patients with initially elevated CEA levels. Thepattern of serially determined CEA in serum correlated closely with the course of the disease. Adecline in CEA serum levels during treatment concurred with a clinical regression of the tumour.While increasing or persistently elevated CEA levels provided the first indication of recurrent disease

of 44


and metastases. Also in some patients with initially normal or undetectable levels of CEA, recurrenceof the disease was preceded or accompanied by a rise in CEA levels (155).Hiura et al (1990)(172), found a good correlation between CEA concentration and the outcome oftreatment. They considered that effective serial plasma determinations of CEA, in patients withcervical carcinoma following therapy might often e useful in the evaluation of therapy as well as inthe earlier detection of recurrence.

Absolute plasma CEA concentration is related to total tumour burden, and antigen release. LikewiseCEA metabolism by the host undoubtedly affects the level of circulating CEA. Several investigatorshave emphasized the importance of hepatic function in the metabolism of CEA in both animals andhuman subjects. Nevertheless 97% of patients with progressively increasing plasma CEAconcentrations following therapy developed recurrent cancer. In nearly one half of the cases, thisprogressive of recurrence by an average of 6 months. One third of the patients experiencing a rapidincrease in plasma CEA concentration did not have levels above 2.5 mg/ml prior to therapy(101)(118).

Donaldson et al 1976)(169), found that biochemical evidence of recurrence preceded clinicaldetection of disease by 1-4 months only.Duke et al (1989)(173),noted biochemical indication of recurrence 3 months following therapy. Serialplasma CEA values did not correlate with disease states when the primary tumour contained little orno antigen.

Kjorstad and Orjasaester (1984)(164), reported that all the patients with pre-treatment levels over15 ug/1 died during the follow up. In the range between 5 – 15 ug/1, two third developed recurrence.So they suggested that for patients with levels over 5 ug/1 surgery and/or radiotherapy wereinadequate. Because of the very unfavourable prognosis such patients may be candidates forchemotherapy and sequential determination of CEA can be used for monitoring such treatment. Inpatients with initial values under 5 ug/1 serial determinations as a help in the follow up seemed to bevery useful.There was no difference in the survival rates between CEA – positive and CEA – negative patients atany time during a 10 year follow up period following treatment with radical surgery or radiotherapyfor cervical cancer(157).

Causes of the Discrepancy Between the Results of Serum CEA in DifferentInvestigations

TeVelde et al (1982)(159), attributed the differences in the published to the following factors:

The lack of adequate information both in the frequency of the follow up visits and the clinicalmethods applied routinely during follow up of the patients.In view of the relatively slow growth rate of the tumour a prolonged follow up period is aprerequisite for the correct assessment of a tumour marker in invasive cervical cancer, toavoid considering occult tumour growth incorrectly as non recurrence casesThe methodology used for establishment of recurrence has not always been described. Ifhistological confirmation cannot be established, the clinical criteria applied to the diagnosisof recurrence should be stated.

The failure to drive an appropriate cut-off which cut-off level which clearly delineates the normalfrom the abnormal range.(159). Braun et al (1981)(160), also stressed on the importance of the upper limit of normal for CEA and therelative distribution of the different clinical stages of the disease within the groups of patientsstudied as important causes of variation of results.The specificity of anti-CEA antibodies varied depending on the source, most commercial anti-CEAantibodies cross - reacted with structurally similar glycoproteins, including non-specific cross-reacting antigens (NCA) (123).

The use of monoclonal antibodies for CEA seemed to increase the positive rate for CEA in invasivecervical carcinoma than the polyclonal ones. However, Rogers et al (1981)(51) and Primus et al

of 44


(1983)(47), anticipated the appearance of potential problems in the correlation between blood levelsof CEA measured with monoclonal antibodies and disease activity due to the existence of tissularCEA determinant heterogeneity.

Hurlimann and Gloor (1984)(122), considered that the differences in results were due to the dilutionsof antisera able to detect concentrations of CEA higher or lower than those expected.

Magon et al (1981)(174), gathered the factors resulting in the discrepancy between the results ofplasma CEA in different investigations in the following:

1. Assay variability.2. Patient population selected for study.3. Tumour types, i.e. not all ovarian and cervical tumours produce CEA.4. The level of CEA regarded as positive (cut-off) level.

Carcinoembryonic Antigen in Cervical Intraepithelial Neoplasia

Intrepithial neoplasia has been considered by many investigators as an intermediate step in thebiologic progression toward invasive malignancy, particularly in the uterine cervix. The incidence ofelevation of serum CEA increased with the progression of the histological epithelial.Abnormalities from moderate dysplasia to carcinoma insitu. The association of increased plasma CEAlevels with the production of keratin by the cervical epithelium CEA expression and chronic cervicalinflammation. Both lymphoplasmacytic infiltration and benign adenomatoid hyperplasia of theendocervical glands are unrelated to the level of CEA (175).

There is little difference in tissue localization of CEA between severe dysplasia and insitu carcinoma.A few moderate dysplasias show positive reaction in isolated groups of cell, they may represent atransitional phase to severe dysplasia. No positive cellular staining is detected in mild dysplasia.Although CEA has been demonstrated in the parakeratotic surface of mild dysplasia, this stainingpattern can not be differentiated from non specific absorption of anti-CEA.. The negative results forCEA in chronic cervicitis and reserve cell hyperplasia suggest that the CEA in dysplastic epithelium isnot merely a reflection of increased cell proliferation but that it is associated with definitetransformation in the epithelial cell (119).

In CIN-1 and 2 lesions the staining reaction to CEA tended to be concentrated in the superficial andintermediate layers while in the sections of CIN-3 the positive stain became diffuse (176).The aggressive lesions of cervical intraepithelial neoplasia could be found among the cases of milddysplasia that expressed CEA. CEA was found in 24% of mild dysplasias, in 37% of severe dysplasias,and in 60% of cases classified as carcinoma in situ (157)(177).

Complete excision of intraepithelial cancer is associated with a return to normal plasma CEA titrewithin 8 weeks in over 75% of the cases. The use of CEA as an aid in determining residual in situcarcinoma of the cervix following cervical conation is potentially important. A persistently elevatedplasma CEA (> 5 mg/ml) following conation is uniformly associated with insitu carcinoma of thecervix (175).

Carcinoembryonic Antigen in Cervical Mucous and Cervico - VaginalWashout Fluid

Carcinoembryonic antigen or carcinoembryonic antigen like substances are found in cervical mucousand cervico-vaginal washout fluid in benign, pre-malignant and malignant cervical epithelial lesions.The level of CEA was below 300 mg/ml in 71% of benign cases, while in severe dysplasia andcarcinoma in situ 83% had CEA levels over 300 mg/ml (178).

The study of CEA in cervical secretions and the demonstration of CEA in cytology specimens, isuseful objective analysis complementary to cytomorphological examination of cytology specimens

of 44


and biopsies. This is specially valuable in follow up of CEA positive dysplasias treated by cone biopsy(119).

The estimation of CEA in cervico-vaginal irrigation fluid considered a more diagnostic tool for CIN-1and II than serum CEA estimation(179).

The degree of positively for CEA in cervical smears increased from pre-malignant lesions to invasivecervical cancer(180).

The cervical mucous samples from the control and leiomyoma bearing - women contain low levels ofCEA. Whereas the mucous samples from squamous cell carcinoma exhibit slightly higher levels ofCEA than those of the former two groups with significant differences. In contrast, the CEA level in thecervical mucous from cervical adenocarcinomas is extremely high with significant differences fromthat of squamous cell carcinomas. Characteristically endometrioid - type adenocarcinomas of thecervix show extremely high levels of CEA. These results suggest that certain types of cervicaladenocarcinomas may synthesize and secrete CEA into cervical mucous (181). References: References1-Khoo S. K. Daunter B. and Mackay E: Carcinoembryonic Antigen and B2 Microglobin as serumtumour markers in woman with genital cancer. Int. J. Gynaecol Obstet .,16:388, 1979

2-Yabushita H; Nakagawa Y; Ssawaguchi K; Noguchi M. and Ishihara M: Differences in themechanisms of production and release of SCC – Ag and CEA from an established uterine cervicalcarcinoma cell line (AMCC-1) Ascia – Oceania J. Obstet Gynaecol ., 14 (4): 493 – 500, 1988

3-Rustin G. J .S.: Circulating tumour markers in the management of human cancer. In: Tumourmarkers in clinical practice-concepts and applications, edited by Daar, A. S. Blackwell, Scientificpublications, oxford P 204 1987.

4-Garett P.E. and Kurtz S. R: Clinical utility of oncofoetal proteins and hormones as tumour markers.Medical Clinics of North America 70 (6): 1295-1306, 1986

5-Van Nagell J. R. Donaldson E. S Hanson M. B. Gay E. C. and Pavelik E. J: Biochemical markers in theplasma and tumours of patients with gynaecologic malignancies. Cancer .,48: 495-503, 1981

6- Coombes, R. C. and Powles T. J.: Tumour Markers in the management of human cancer In: Topicalreviews in radiotherapy and oncology, Edited by Deeley, T. J. Wright P. S. G. Bristol, P. 39,1982.

7- Kraly J and Wawrzkiewicz M : Evaluation of diagnostic usefulness of CEA, HCG and SCC antigens incervical cancer patient’s. Eur. J. Gynecol. Oncol I0 (5) : 319 - 322. 1989

8-Ballesta AM, Molina R, Filella X, JO J, Gimenez N: Carcinoembryonic antigen in staging and follow upof patients with solid tumours. Tumour Biol.16(1):32-41,1995

9- Abelev G.I: Perova S.D. Khramokora, N I, Prostnikova Z.A and Irin T.S: Transplantation, 1, 1974Citing, Daar A.S. and Lennox E.S.: Tumour markers in clinical practice – concepts and applications,edited by Daar, A.S Blackwell Scientific Publications, Oxford p. 16, 1987.

10- Daar A.S. and Lennox E.S . Tumour markers and antigens, In Tumour Markers in clinical practice– concepts and applications, edited by Daar, A.S. Blackwell scientific publications, Oxford P.4 1987

11-Phil Gold C C and Goldenberg N A :The Carcinoembryonic antigen (CEA): Past, Present, andFuture. MJM ,3:46, 1997

12-Hauschka T S: Immunologic aspects of cancer :review. Cancer Research 12:615;1952.

of 44


13-Oettgen HF, Hellström KE. Tumor Immunology. In: Holland JF, Frei E, eds. Cancer Medicine.Philadelphia: Lea and Feibiger;: 951. 1973

14-Gold P, Freedman SO. Demonstration of tumor-specific antigens in human colonic carcinomata byimmunological tolerance and absorption techniques. Journal of Experimental Medicine 121: 439;1965.

15-Gold P, Freedman SO. Specific carcinoembryonic antigens of the human digestive system. Journalof Experimental Medicine 122: 467; 1965.

16- Coombes, R.C and Neville A.M: Significance of tumour index substances in management. In:secondary spread in breast cancer, edited by Stoll, B.A William Heinemann Medical Books, Chicago1978.

17- Krupey J, Gold P, Freedman SO. Purification and characterization of carcinoembryonic antigens ofthe human digestive system. Nature 215(96): 67-68; 1967.

18-. Krupey J, Gold P, Freedman SO. Physicochemical studies of the carcinoembryonic antigens of thehuman digestive system. Journal of Experimental Medicine 128(3): 387-398; 1968.

19-. Madjic O. Cluster report: CD66. In: Knapp W, et al., eds. Leukocyte Typing IV. Oxford, England:Oxford University Press; 838. 1989

20- Gold P, Krupey J, Ansari H. Position of the carcinoembryonic antigen of the human digestivesystem in ultrastructure of tumor cell surface. Journal of the National Cancer Institute 45(2):219-225; 1970.

21-. Rosenthal KL. Palmer JL. Harris JA, et al. Antibody-induced redistribution of CEA on the cellsurface: utilization in separation of CEA and isoantigen A. Journal of Immunology 115(4):1049-1053;1975.

22-. Pignatelli M, Durbin H, Bodmer WF: Carcinoembryonic antigen functions as an accessoryadhesion molecule mediating colon epithelial cell-collagen interactions. Proceedings of the NationalAcademy of Sciences (USA) 87(4): 1541-1545; 1990.

23- Thomson DM, Krupey J, Freedman SO, et al. The radioimmunoassay of circulatingcarcinoembryonic antigen of the human digestive system. Proceedings of the National Academy ofSciences (USA) 64(1): 161-167; 1969

24- Jacob F and Monod J: Genetic regulatory mechanism in the syntheses if proteins. J. mo; Biol .,318,1961 citing Gold P. and Freedman, S.O. Specific carcinoembryonic antigen of the human digestivesystem. J. Exp Med 122:467-481, 1965

25—Coligan J E, Lautenchleger J.T ,Egan M I, and Todd C W: Isolation and characterization ofcarcinoembryonic antigen. Immunochemistry, 9: 377-386, 1972.

26- Martin F. and Martin M.S: Demonstration of an antigen common to colonic cancer in the humandigestive system. Int. J. Cancer 6:352-360, 1970

27- Lo Gerfo P. and Herter F.P: J. Surg Oncol 4:1 1972, Citing Fritsche R. and mach J.P: Isolation andcharacterization of carcinoembryonic antigen (CEA) extracted from normal human colon mucosa.Immunochemistry, 14: 119-127, 1977

28- Pusztaderi G. and Mach J.P. Carcinoembryonic antigen (CEA) in non digestive cancerous andnormal tissues. Immunochemistry 10: 197-204, 1973

29- Sizaret P. and Martin F: J. Natl Cancer Inst. .,50:807, 1973 citing Fritsche R and Mach J.P.:Isolation and characterization of carcinoembryonic antigen (CEA) extracted from normal humancolon mucosa. Immunochemistry 14:119-127, 1977

of 44


30- Lo Gerfo P; Herter F.P. Braun J. and Hansen H.J: Ann Surg 175:495, 1972, Citing: Fritsche R. andmach H.P. Isolation and characterization of carcinoembryonic antigen (CEA) extracted from normalhuman colon mucosa. Immunochemistry 14:119-127, 1977

31- Fristsche R. and Mach J.P.: Isolation and characterization of carcinoembryonic antigen (CEA)extracted from normal human colon mucosa.Immunochemistry 14:119-127, 1977

32- Martin E W, Kibbey W E. di Vecchia L Anderson G; Catalano P. and Minton J.P: Carcinoembryonicantigen : clinical and histological aspects. Cancer 37:62-81, 1976

33- Bale W.F; Contreras M.A and Grady, E.D: Factors influencing localization of labelled anti-bodies intumours. Cancer Research, 40:2965-2972,1980

34-- Egan M.1 Coligan J E Pritchard D C Schunte W C and Todd C W: Physical characterization andstructural studies of the carcinoembryonic antigen.Cancer Research 36: 3482 – 3485, 1976

35-Garrett PE and Kurtz SR: Clinical utility of oncofoetal proteins and hormones as tumour markers.Medical Clinics of North America,70(6):1295-1306.1986

36- Grunert, F abu Harfeil Schwarz K. and Von Kliest S: Two CEA and three NCA species althoughdistinguishable by monoclonal antibodies have nearly identical peptide patterns. Int. J. Cancer 36:357-362, 1985

37- Pletsch Q. and Goldenberg D.M: Molecular size of carcinoembryonic antigen in the plasma ofpatients with malignant disease. J. Natl. Cancer Inst. 53: 1201-1204, 1974

38- Terry W.D. Henkart P.A. Coligan J.E. and Todd C.W: Carcinoembryonic antigen: characterizationand clinical applications. Transplant Rev. .,20: 100, 1974 Citing, De young N J and Ashman L K :Physicochemical and immunochemical properties of carcinoembryonic antigen (CEA) from differenttumour sources. Aust. J. Exp. Biol Med Sci 56 (3): 321-331, 1978.

39- De Young N.T and Ashman L.K: Physicochemical and immunochemical properties ofcarcinoembryonic antigen (CEA) from different tumour sources. Aust. J. Exp. Biol Med Sci .,56 (3):321-331 1978

40- Terry W.D.; Henkart P A. Coligan J E. and Todd, C. W: Structural studies of the major glycoproteinin preparations with carcinoembryonic antigen activity. J. Exp. Med 136:200-204, 1972

41- Kuroki M; Shinoda K; Takayasy T; Koga Y. and Matsuoka Y: Immunological characterization andstructural studies of normal faecal antigen – 1 related to carcinoembryonic antigen. MolecularImmunology 19:399-406, 1982

42- Koga Y; Kuroki M; Matsunaga A; Shinoda T; Takayasu T. and Matsouka Y.: Further comparativestudies on chemical properties of carcinoembryonic antigen in tumour tissues and closely relatedantigens in adu7lt faeces and meconium. Molecular Immunology 22(1): 67-73, 1985

43- Hirai H: A collaborative clinical study of carcinoembryonic antigen in Japan. Cancer Research 37:2267-2274, 1977.

44- Gold P; Krupey J. and Ansari H.: Position of CEA of the human digestive system in theultrastructure of tumour cell surface J. Natl Cancer Inst. 45:219, 1970

45- Laurence D.J. R and Neville A.M.: Foetal antigens and their role in the diagnosis and clinicalmanagement of human neoplasm: An overview . Br. J. Cancer 26: 335-355, 1972

46- Denk, H; Tappeiner G; Eckerstorfer, R. and Holzner J.H: Carcinoembryonic antigen (CEA) in

of 44


gastrointestinal and extra gastrointestinal tumours and its relationship to tumour – celldifferentiation. Int. J. cancer 10:262-272, 1972

47- Primus F.J. Freeman J.W. and Goldenberg D.M: Immunological heterogeneity of carcinoembryonicantigen:Immunohistochemical detection of carcinoembryonic antigen determinants in colonic tumours withmonoclonal antibodies. Cancer Research 43: 693-701, 1983

48- Muraro R; Wunderlich D; Thor A; Lundy J; Noguchi P; Cunningham R. and Shlom J; Definition bymonoclonal antibodies of a repertoire of epitopes on carcinoembryonic antigen differentiallyexpressed in human colon carcinomas versus normal adult tissues, Cancer Research 45: 5769-5780,1985

49- Thor A; Moraro R; Gorstein F; Ohuchi N; Viglione M; Szpak C.A. Jonston W.W and Schlom J:Adjunct to the diagnostic distinction between adenocarcinomas of the ovary and the colon utilizing amonoclonal antibody (col-4) with restricted carcinoembryonic antigen reactivity. Cancer Research,47:505 0 512, 1987

50- Wagener C; Joy Yang Y.H. Crawford, F.G. and Shively J.E: Monoclonal antibodies forcarcinoembryonic antigen and related antigens as a model system: A systematic approach for thedetermination of epitope specificities of monoclonal antibodies. The journal of Immunology, 130 (5):2308-2315, 1983

51- Rogers G.T. Rawlins G.A Keep P.A. Cooper E. II and Bahshawe K.D.: Application of monoclonalantibodies to purified CEA in clinical radioimmunoassay of human serum. Br. J. Cancer 44: 371-380,1981

52- Chism, S.E; Warner N.L; Wells J.V; Crewther P: Hunts S; Marchalonis J.J and Fundenberg H.H:Evidence for common and distinct determinants of colon carcinoembryonic antigen, colon carcinomaantigen –III , and molecules with carcinoembryonic antigen activity isolated from breast and ovariancancer.Caner Research, 37 (9):3100 – 3108, 1977

53- Haskell C.M. Buchegger F; Schreyer M; carrel S. and Mach J.P. Miniclonal antibodies tocarcinoembryonic antigen: Lonic strength as a factor in the selection of antibodies forimmunoscintigraphy. Cancer Research 43: 3857-3864, 1983

54- Imai K. Moriya Y; Fujita H; Tsujisaki M; Kawaharada M. and Yachi A: Immunologic characterizationand molecular profile of carcinoembryonic antigen detected by monoclonal antibodies. The Journal ofimmunology 132 (6)::2992-2997, 1984

55- Yachi A; Imai K; Fujita H; Moriya Y; Tanda M; Endo T; Tsujisaki M. and Kawaharada M:Immunohistochemical distribution of the antigenic determinants detected by monoclonal antibodiesto carcinoembryonic antigen.The journal of Immunology 132 (6): 2998-3004, 1984

56 Kuroki M; Kuroki M; Koga Y. and Matsuoka Y: Monoclonal antibodies to carcinoembryonic antigena systematic analysis of antibody specificities by using related normal antigens and evidence forallotypic determinants on carcinoembryonic antigen. The journal of Immunology, 133 (4):2090-2097, 1984.

57-- Kuroki M. Koga Y. and Matsuoka Y: Purification and characterization of carcinoembryonic antigen– related antigens in normal adult faeces. Cancer Research 41:713-720, 1981

58- Tillack, T.W. Rosai J. and Vervynk D.J: Immunological studies of glycoproteins isolated from cellmembranes of human lung carcinomas. J. Natl Cancer Inst. .,52: 1059-1067, 1974

59- Santen R.J. Collet J. and Franchimont P.: Partial purification of carcinoembryonic – reactiveantigen from breast neoplasms using lectin and antibody affinity chromatography. Cancer research

of 44


40:1181-1188, 1980

60--Wagener C.; Petezold, P. Kohler W. and Totvic V.: Binding of five monoclonal anti-CEA antibodieswith different epitope specificities to ovarian carcinoma tissues Int. J. Cancer ., 33: 469-475, 1984

61- Hammarstorm S; Shively J.E; Paxton R.J et al: Antigenic sites in carcinoembryonic antigen.Cancer Research 49: 4852-4585, 1989

62- Matsuoka Y; Kuroki M; Koga Y; Kuriyama H; Mori t. and Kosaki G: Immunochemical differencesamong carcinoembryonic antigen in tumour tissues and related antigens in meconium and adultfaeces. Cancer Research 42: 2012-2018, 1982

63- Neumaier M; Fenger U. and Wagener C: Delineation of four carcinoembryonic antigen ((CEA)related antigens in normal plasma by transplant studies using monoclonal anti-CEA antibodies withdifferent epitope specificity. Molecular Immunology 22: 1273-1277, 1985

64- Buchegger F; Schreyer M; Carrel, S. and Mach J.P;(1984): Monoclonal antibodies identify a CEAcross reacting antigen of 95 kd (NCA-95) distinct in antigenicity and tissue distribution from thepreviously described NCA of 55 kd.Int.j.cancer,33:643-649,1984

65- Audette, M. Buchuegger, F: Schreyer, M. and Mach, J.P(1987): Monoclonal antibody againstcarcinoembryonic antigen of molecular weight 90.000 and 160.000 in normal granulocytes.Molecular Immunology 24:1177-1168, 1987

66- Khan W.N. Osterman, a, and Hammaestrom S: Molecular cloning and expression of cDNA for acarcinoembryonic antigen – related foetal liver glycoprotein. Proc Natl. Acad Sci USA, 86: 3332-3336,1989 Citing Hammarstrom S; Shively J.E; Paxton R J et al: Antigenic sites in carcinoembryonicantigen.Cancer Research .,49:4852-4858, 1989

67- Von Kleist S; Chavanel G. and Burtin P: Characterization of a normal tissular antigen extractedfrom human colonic tumours. Immunochemistry, 11:249-253, 1974

68-Darcy DA, Tuberville C, and James R: Immunological study of carcinoembryonic antigen (CEA) anda related glycoprotein. B J Cancer 28:147-160,1973

69- Blaszczyk M; Pak K.Y. Herlyn, M: Lindergren, J; Pessano S; Steplewski Z. and Koprowski, H (1984):Characterization of gastrointestinal tumour associated carcinoembryonic antigen-related antigensdefined by monoclonal antibodies. Cancer research 44:245-253, 1984.

70-Grunert F: Luckenbach G.A. Haderly B. Schwarz K. and Von Kliest S: Comparison of colon – lungand breast derived carcinoembryonic antigen and cross reacting antigens by monoclonal and fingerprint analysis Ann N.Y. Acad Sci, 417:75-85, 1983

71- Von Kleist S; Chavanel G. and Burtin P.: Identification of an antigen from normal human tissuethat cross reacts with carcinoembryonic antigen. Proc. Nat. Acad Sci .,69: 2492-2494, 1972

72- Mach J.P. and Pusztaszeri G: Carcinoembryonic antigen (CEA) Demonstration of a partial identitybetween CEA and a normal glycoprotein. Immunochemistry 9: 1031-1034, 1972

73- Engavall E. Vuento M and Ruoslahti E: A monkey antigen cross reacting with carcinoembryonicantigen Br. J. Cancer 34:341-345, 1976

74- Egan M.L Pritchard D.G. Todd C.W and Go. V. L.W. Isolation and immunochemical and chemicalcharacterization of carcinoembryonic antigen – like substances in colon lavages of healthyindividuals. Cancer Research 37:2638-2643, 1977

75- Burtin P; Chavanal, G. and Hirsch – Marie, H: Characterization of a second normal antigen thatcross – reacts with CEA. The journal of immunology, 1111:1926-1928, 1973

of 44


76- Kuriyama , H: Antigenic structure of NCA –2 from meconium and comparison with normal faecalantigen from adult faeces and carcinoembryonic antigen from malignant tissues. Nippon GekaGakkai Zasshi .,84 (4): 301-309 1983 (Abstract).

77- Bormer O.P (1991): Major disagreements between immunoassays of carcinoembryonic antigenmay be caused by non specific cross – reacting antigen – 2(NCA-2). Clin chem .,37 (10):1736-1739,1991

78- Hedin A; Carlesson L; Berglund A Hammarstrom S: A monoclonal antibody – enzymeimmunoassay for serum carcinoembryonic antigen with increased specificity for carcinomas. Proc.Natl Acad Sci USA, 80 (11) 3470-3474, 1983

79- Siepen D; Paxton R.J. Neumaier M; Shively J.E. and Wegener C: Carcinoembryonic antigen (CEA)and two cross reacting antigens of 165 kd and 105 kd isolated from meconium exhibit identicalamino terminal sequences. Biochem Biophys Res. Common .,147 (1): 212-218, 1987

80- Zoubir F; Zermoski J; Szmeja J; Hedin A and Hammarstrom S.: Tumour specificity of monoclonalantibodies to carcinoembryonic antigen. Immunohistochemical analysis. Tumour Biology 11 (1-2:5-19, 1990

81- Mach J.P. and Holburn A: Immunochemistry and biochemistry of CEA. Ann Immunol (Inst. Pasteur)124c: 579-581, 1973

82- Pompecki R; Shively J.E and Todd, C.W: Sensitive detection of carbohydrate determinants oncarcinoembryonic antigen preparations by lectin and antibody binding using polyethylene glycol.Cancer Research 41: 1905-1909, 1981

83- Thomson D.M.P. Krupey J; Freedman S.O. and Gold P.: The radioimmunoassay of circulatingcarcinoembryonic antigen of the human digestive system.Proc. Natn Acad Sci .,64: 161 1969, citing, Laurence D.J.R. and Neville, A.M. Foetal antigens and theirrole in the diagnosis and clinical management of human neoplasms: An overview Br. J. Cancer 26:335-355, 1972

84- Egan M.L. Lautenschleger J.T Coligan J.E and Todd C.W. Radioimmune assay of carcinoembryonicantigen. Immunochemistry 9:289, 1972

85- Koh S.K. and Cauchi M.N: Prognostic significance of oncofoetal antigens in patients with ovariancancer. Aust. N.Z.J. Obstet Gynaec .,23:69-72, 1983

86- Bast R.C Klug T.L Schaetzl Lavin P. Niloff J.M. Gerber T.F. Zurawski, V.R. and Knapp R.C:Monitoring human ovarian carcinoma with a combination of CA 125, CA 19-9 and carcinoembryonicantigen. Am J. Obstet Gynaecol .,149:553-559, 1984.

87- Khoo S.K. and Mackay I.R: J. Clin Path .,26:470, 1973 citing Khoo S.K; Hill R; Daunter B. andMackay E.V: Carcinoembryonic antigen in ovarian cancer: correlation of concentrations in tumourtissue, cyst fluid, ascetic fluid and peripheral blood. Aust. N.Z. J. Obstet Gynaec .,22 (2): 65-70, 1982

88- Pulay T; Csomor S; Meszaros K; Szarka G; Csabay L and Szinnyai A.: Significance of CEA AFP SPIand HCG as tumour markers in ovarian carcinoma. Neoplasma 34 (5): 595-600, 1987

89- Tholander B; Taube A; Lidgren A; Sjoberg MO; Stendahl U; Kiviranta A; Hallman K; Holm L. WeinerE. and Tamsen L: Pre-treatment serum levels of CA 125, carcinoembryonic antigen, tissuepolypeptide antigen and placental alkaline phosphatase in patients with ovarian carcinoma:Influence of histological type, grade of differentiation, and clinical stage of disease. GynaecologicOncology 39: 26-33, 1990

90- Meier W. Eiermann W. Stieber P; Moghadam A.F. Schneider A. and Hepp. H. Squamous cellcarcinoma antigen and carcinoembryonic antigen levels as prognostic factors for the response of

of 44


cervical carcinoma to chemotherapy. Gynaecologic Oncology 38: 6-11, 1990

91- Leminen A.: Tumour markers CA125 carcinoembryonic antigen and tumour associated trypsininhibitor in patients with cervical adenocarcinoma.Gynaecologic Oncology 39: 358-363, 1990

92- Lundqvist A.E. Sjovall K; Hasson L.O. and Eneroth P: Peri-and post – operative changes in serumlevels of four tumour markers and three acute phase reactants in benign and malignantgynaecological disease. Arch Gynaecol Obstet 251: 69-78, 1992

93-Goldenberg DM Neville A.M. Carter AC Go. V. L.W Holyyoke E.D. Isselbacher K.J. Schein P.S. andSchwartz M: Carcinoembryonic antigen: its role as a marker in the management of cancer. A nationalinstitutes of health consensus development conference. Annals of internal Medicine 94 (3): 407-409,1981

94-- Goldenberg D.M.Sharkey R.M. and Primus F.J.: J. Natl Cancer Inst. 57:11 1976 citing Khoo, S.K.Hill R. Daunter B. and Mackay E.V. Carcinoembryonic antigen in ovarian cancer Correlation ofconcentrations in tumour tissue, cyst fluid ascetic fluid and peripheral blood. Aust. N.Z. J. obstetGynaecol 22(2): 65-70, 1982.

95- Van Nagell J.R; Meeker W.R. Parker J. C. and Harralson J.D.: Carcinoembryonic antigen in patientswith gynaecologic malignancy Cancer., 35: 1372-1376, 1975

96- Seppala M, Pihko H. and Ruoslahti E: Carcinoembryonic antigen and alpha – fetoprotein inmalignant tumours of the female genital tract. Cancer 35: 1377-1381, 1975

97-- Fritsche H.A. Freedman, R.S. Liu F; Acomb L.D. Asep. M.T. and Collonsworth W.L: A survey oftumour markers in patients with squamous cell carcinoma of the uterine cervix. GynaecologicOncology 14:230-235, 1982

98- Sonnendecker E.W. De Souza J.J.L and Herman A.A.B : Screening for liver metastases fromovarian cancer with serum carcinoembryonic antigen and radionuclide hepatic scintiphotography. Br.J. Obstet Gynaecol .,91:187-192, 1984

99- Khoo S.K. and Mackay E.V: Carcinoembryonic antigen by radioimmunoassay in the detection ofrecurrence during long term follow up of female genital cancer . Cancer ,.34:542-548, 1974

100-- Disaia P.J. Morrow C.P.; Haverback B.J. and Dyce B.J.: Carcinoembryonic antigen in cancer ofthe female reproductive system. Serial plasma values correlated with disease state. Cancer 392365-2370, 1977

101- Van Nagell J.R. Hudson S.; Gay E.C. Donaldson, E S. Hanson, M. Powel, D.F. and GoldenbergD.M: Carcinoembryonic antigen in carcinoma of the uterine cervix: Antigen distribution in primaryand metastatic tumours. Cancer ., 49:379-383, 1982

102- Parente J.T. and Greston W.M.: Carcinoembryonic antigen levels in the diagnosis of malignantlesions of the abdominopelvic region. Surgery Gynaecology and Obstetrics 153: 693-696, 1981

103- Schwartz P.E., chambers S.K. Chambers J.T; Guttmann J; Katopodis N. and Foemmel R:Circulating tumour markers in the monitoring of gynaecologic malignancies. Cancer 60: 353-361,1987

104- Stall K.E and Martin E.W.: Plasma carcinoembryonic antigen levels in ovarian cancer patients. Achart review and survey of published data. The journal of reproductive medicine .,26(2): 73-79 1981

105- Nap M; Ten Hoor K.A. and Fleuren G.J: Cross reactivity with normal antigens incarcinoembryonic antigen – CEA sera, used for immunohistology. The need for tissue controls andabsorptions. Am. J. Clin Pathol .,79: 25-31, 1983

of 44


106- Fleuren G.J. and Nap M: Carcinoembryonic antigen in primary and metastatic ovarian tumours.Gynaecologic Oncology 30:407-415, 1988

107- Inoue M; Fujita M; Nakazawa A; Ogawa H. and Tainzawa O: Sialy1-Tn Sialy – Lewis xi, CA19-9,CA 125 Carcinoembryonic antigen, and tissue polypeptide antigen in differentiating ovarian cancerfrom benign tumours. Obstet Gynaecol .,79:434-440, 1992

108- Coons, A.H. Creech, H. J. Jones, R.N. and Berliner E: The demonstration of pneumococcalantigen in tissue by the use of fluorescent antibody.J. Immunol .,45:159-170, 1942 citing primus F.J. Sharkey, R.M.; Hansen H.J. and Goldenberg D.M:Immunoperoxidase detection of carcinoembryonic antigen. An overview. Cancer 42:1540-1545, 1978

109- Primus F.J. Sharkey R.M. Hansen H.J. and Goldenberg D.M.: Immunoperoxidase detection ofcarcinoembryonic antigen an overview. Cancer 42: 1540-1545, 1978

110-- Goldenberg DM Deland F. and Kim E: Use of radiolabeled antibodies to carcinoembryonicantigen for the detection and localization of diverse cancers by external photoscanning. N. Engl. J.Med. 298-1384-1388, 1978

111- Donaldson E.S Van Nagel J.R Pusell S; Gay E. C Meeker W.R Kashmiri, r. and De Voodre J.V:Multiple biochemical markers in patients with gynaecologic malignancies. Cancer 45:948-953, 1980

112- Hsu K.G. Zimmerman E.A. Rudin L; Logerfo P. Bennett S.: Application of immunoperoxidasebridge technique in studies of tumour tissues with antibody to tumour associated antigen (TAA). In:proc. Second conf. Embryonic and Foetal Antigens in Cancer, edited by Anderson N.G. and CogginsJ.H. Springfield Virginia, U.S. Dept. of Commerce Vol 2, P. 147 1972 Citing Primus F.J. Sharkey R.M.Hansen H.J. and Goldenberg D.M: Immunoperoxidase detection of carcinoembryonic antigen. Anoverview Cancer 42: 1540-1545, 1978

113- Sternberger L.A.: Immunocytochemistry Englewood cliffs, New Jersey Prentice – Hall Inc. 1974citing Primus F.J; Sharkey R.M; Hansen H.J. and Goldenberg D.M: Immunoperoxidase detection ofcarcinoembryonic antigen an overview. Cancer .,42:1540-1545, 1978

114-Primus F J, Sharkey R M ,Hansen H J. and Goldenberg D M.: Immunoperoxidase detection ofcarcinoembryonic antigen an overview, Cancer b42:1540-1545,1978

115- Goldenberg D.M Sharkey R.M. and Primus F.J. Immunocytochemical detection ofcarcinoembryonic antigen in conventional histopathology specimensCancer 42: 1546-1553, 1978

116- Van Nagell J. R. Kim E; Casper S; Primus F.J; Bennett S; Deland and Goldenberg D.M.:Radioimmunodectection of primary and metastatic ovarian cancer using radiolabeled antibodies tocarcinoembryonic antigen. Cancer Research ., 40: 502-506, 1980

117- Rutanen E.M. Lidgren J; Sipponen P; Stenman U. H. Saksela, E. and Seppala M.:Carcinoembryonic antigen in malignant and non malignant gynaecologic tumours, circulating levelsand tissue localization. Cancer 42: 581-590, 1978

118- Van Nagell J.R. Donaldson E.S. Gay E.C. Rayburn P; Powell D.F. and Goldenberg D.M.:Carcinoembryonic antigen in carcinoma of the uterine cervix I. The prognostic value of serial plasmadetermination Cancer .,42: 2428-2434, 1978

119- McDicken I.W. and Rainey M: The Immunohistological demonstration of carcinoembryonicantigen in intraepithelial and invasive squamous carcinoma of the cervix. Histopathology 7: 475-485,1983

120- Ogawa H. Inoue M; Tanizawa O; Miyamoto M, and Sakuri M: Altered expression of sialyl 0 Tn,Lewis antigens and Carcinoembryonic antigen between primary and metastatic lesions of uterinecervical cancers. Histochemistry 97: 311-317, 1992

of 44


121- Yabushita H; Masuda T; Ogawa A; Noguchi M. and Ishihara M: Combination assay of CA 125,TPA, IAP, CEA and Ferritin in serum for ovarian cancer.Gynaecologic Oncology .,29:66-75, 1988

122- Hurlimann J and Gloor E: Adenocarcinoma in situ and invasive adenocarcinoma of the uterinecervix. An immunohistoligic study with antibodies specific for several epithelial markers Cancer54:103-109, 1984

123- Nanbu Y; Fujii S; Konishi I; Nonogaki H. and Mori T: Immunohistochemical localizations of CA125 carcinoembryonic antigen and CA19-9 in normal and neoplastic glandular cells of the uterinecervix. Cancer 62:2580-2588, 1988

124- Inoue S; Matsuo I;Shimokawa K; Tohyo T and Maeyama M: A case of clear cell adenocarcinomaof the uterine cervix in pregnancy. Gynaecologic Oncology 24:120-125, 1986

125- Valente P.T. and Susin M.: Cervical adenocarcinoma arising in florid mesonephric hyperplasia:Report of a case with Immunocytochemical studies. Gynaecologic Oncology 27: 58-68, 1987

126- Michael H; Sutton G; Hull M.T. and Roth L.M: Villous adenoma of the uterine cervix associatedwith invasive adenocarcinoma: A histological, ultrastructural and immunohistochemical study. Int. J.Gynaecol Pathol 5(2): 163-169, 1986

127-Schneider J; Gonzalez D.T.J. Martin M; Del Marcenteno M; Diez J. and Rodriguez E.F.J: Malignantadenoma of the cervix: a diagnostic and therapeutic dilemma. Geburtshifle Frauenheilkd, 51(6):466-468, 1991

128- Yamagata S; Yamamoto K. Isuchida S; Kawamura N; Nakamura, T and Sugawa T:Clinicopathological ultrastructural and immunohistochemical study on adenoma malignum of theuterine cervix. Nippon Sanaka Fujinka Gakkai Zasshi .,43(1) 57-64, 1991

129- Sasaki A: Clinicopathological and histochemical study of adenoam malignum of the uterinecervix. Nippon Sanka Fujinka Gakkai Zasshi .,44(4) 427-432, 1992

130- Wahlstrom T; Lingren J; Korhopen M. and Seppala M: Distinction between endocervical andendometrial adenocarcinoma with immunoperoxidase staining of carcinoembryonic antigen inroutine histological specimens.Lancet, 1: 1159-1160, 1979

131- Cohen C; Shulmann, G. and Budgeon L.R: Endocervical and endometrial adenocarcinoma. Animmunoperoxidase and histochemical study. Am J. Surg Pathol .,6:151, 1982

132- Bychkov v.; Rothman, M and Bardawill W.A: Immunocytochemical localization ofcarcinoembryonic antigen (CEA), alpha – fetoprotein (AFP), and human chorionic gonadotropin (HCG)in cervical neoplasia.Am. J. Clin Pathol .,79:414-420, 1983.

133- Aqarwal, S. and Sharma S(1990): Localization of carcinoembryonic antigen in uterine cervicalneoplasia. Indian J. Med. Res, .,92:452-455, 1990

134-- Dallenbach, H.G and Lang, G: Immunohistochemical studies in uterine tumours. I. Invasivesquamous cell carcinomas of the cervix and their precursors. Patthol. Res Pract 187 (1): 36-43, 1991(abstract)

135-Toki T and Yajima A :Immunohistochemical localization of carcinoembryonic antigen (CEA) insquamous cell carcinoma of the uterine cervix: Prognostic significance of localization pattern of CEA.Tohoku J Exp Med 165(1):25-32,1991

136-Lam CP, Yuan CC, Jeng FS, Tsai LC, Yeh SH, Ng HT: Evaluation of carcinoembryonic antigen,

of 44


tissue polypeptide antigen, and squamous cell carcinoma antigen in the detection of cervicalcancers. Chung Hua I Hsueh Tsa Chih (Taipei) 1992 Jul;50(1):7-13

137-Pectasides D, Economides N, Bourazanis J, Pozadzizou P, Gogou L, Koutsiouba P, Athanassiou A:Squamous cell carcinoma antigen, tumour- associated trypsin inhibitor, and carcinoembryonicantigen for monitoring cervical cancer. Am J Clin Oncol Aug;17(4):307-12, 1994

138-Ngan HY, Cheung AN, Lauder IJ Wang LC, Ma HK: Prognostic significance of serum tumourmarkers in carcinoma of the cervix. Eur J Gynaecol Oncol;17(6):512-7, 1996

139-Bae SN, Namkoong SE, Jung JK, Kim CJ, Park JS, Kim JW, Lee JM, Kim SJ: Prognostic significance ofpre-treatment squamous cell carcinoma antigen and carcinoembryonic antigen in squamous cellcarcinoma of the cervix. Gynecol Oncol Mar;64(3):418-24, 1997

140- Speers W.C; Picaso L.G. and Silverberg S.G: Immunohistochemical localization ofcarcinoembryonic antigen in microglandular hyperplasia and adenocarcinoma of the endocervix. AmJ. Clin Pathol .,79:105-107, 1983

141- Kudo R. Sasano H; Koizumi M. Orenstein J. M. and Silverberg S.G: Immunohistochemicalcomparison of new monoclonal antibody 1 c5 and Carcinomebryonic antigen in the differentialdiagnosis of adenocarcinoma of the uterine cervix. Int. J. Gynaecol Pathol .,9 (4): 325-336, 1990

142-Ngan HY, Cheung AN, Lauder IJ, Cheng DK, Wong LC, Ma HK: Tumour markers and theirprognostic value in adenocarcinoma of the cervix. Tumour Biol;19(6):439-44, 1998

143- Maes G; Fleuren G.J. Bara J. and Nap M: The distribution of mucins, carcinoembryonic antigenand mucous associated antigens in endocervical and endometrial adenocarcinomas. Int. J. GynaecolPathol 7(2): 112-122, 1988

144- Toki T; Oikawa N: Tase T; Wada Y; Yajima A. and Suzuki M.: Immunohistochemical localization ofcarcinoembryonic antigen in adenocarcinoma of the endocervix and the endometrium.Tohoku J. Exp.Med .,146 (1): 27 – 32, 1985

145- Ueda S; Tsubura A; Izumi H; Sasaki M. and Morii, S: Immunohistochemical studios oncarcinoembryonic antigen in adenocarcinomas of the uterus.Acta Pathol Jpn ., 33 (1): 59-59, 1983

146- Van Nagell J.R Donaldson E.S. Gay E.C. Sharkey: Carcinoembryonic antigen in ovarian epithelialcystadenocarcinomas. The prognostic value of tumour and serial plasma determinations. Cancer.,41: 2335-2340, 1978

147- Ito H; Kurihara S. and Nishimyra C: Serum carcinoembryonic antigen in patients with carcinomaof the cervix. Obstet Gynaecol .,51 (4): 468-471, 1978

148- Khoo S.K; Hill R; and Mackar E.V: Detection of carcinoembryonic antigen and alpha-fetoproteinin serum and ascetic fluid from patients with ovarian cancer. Aust. N.Z.J Obstet Gynaec .,17:94, 1977

149- Khoo S.K. Whitaker S.V Jones I.S. C and Mackay E.V: Carcinoembryonic antigen in patients withresidual ovarian cancer. Clinical and pathological correlations. Gynaecologic Oncology 7:288-295,1979

150- Sarjadi S. Daunter B; Mackay E; Magon H. and Khoo S.K.: A multiparametric approach of tumourmarkers detectable in serum in patients with carcinoma of the ovary or uterine cervix. GynaecologicOncology 10:113, 1980

151- Kjorstad, K.E, and Orjasaester H: The prognostic value of CEA determinations in the plasma ofpatients with squamous cell cancer of the cervix cancer .,50:283-287, 1982

152- Cauchi M.N. Koh S.H. Lim D. and Hay D.L: Oncofoetal antigens in cancer of the cervix and ovary

of 44


Br. J cancer 44:403-409, 1981

153- Hansen H.J. Snyder J.J. and Miller E: Carcinoembryonic antigen (CEA) assay Hum Pathol 5: 139,1974 citing Cauchi M.N. Koh S.H. Lim d. and hay D.1: Oncofoetal antigens in cancer of the cervix andovary. Br. J. Cancer 44:403-409, 1981

154-- Franchimont P; Zangerle, P.E and Nogardede J: Simultaneous assays of cancer associatedantigens in various neoplastic disorders. Cancer 38: 2287, 1976

155- Braun P and Leyendecker G (1983): Carcinoembryonic antigen in cervical cancer: A five yearfollow up. Cancer Detection and Prevention, 6:249-253 1983.

156- Leminen A; Alftan H; Stermann U.H and Lehtovirta P: Chemotherapy as initial treatment forcervical carcinoma: clinical and tumour marker response .Acta. Obstet Gynaecol Scand 71: 293-297,1992

157- Lindergren J; Wahlstrom T. and Seppala M: Tissue CEA in premalignant epithelial lesions andepidermoid carcinoma of uterine cervix: prognostic significance. Int. J. Cancer 23: 448 – 453, 1979

158- Kornafel J. and Wawrzkiewicz M: Evaluation of diagnostic usefulness of CEA, HCG and SCCantigens in cervical cancer patients.Eur. J. Gynaec Oncol 10(5): 319 – 322, 1989

159- TeVelde E.R. Persijn J.P. Ballieux R.E and Faber J.: Carcinoembryonic antigen serum levels inpatients with squamous cell carcinoma of the uterine cervix: Clinical significance. Cancer 49:1866-1873, 1982.

160- Braun P: Hildenbrand G; Izbicki J and Leyendecker G (1981): Clinical significance ofmeasurements of carcinoembryonic antigen in serum of patients with carcinoma of the uterinecervix. Arch Gynaecol .,230:273, 1981

161-Yokayama T; Kaneko H; Kawamura S; Tanaka M; Chiba T; Hiura M. and Moriwaki S: A case ofimmunohistochemical carcinoembryonic antigen producing cervical cancer. The evaluation of plasmaCEA levels in 201 patients with uterine cancer. Gan No. Rinsho 33(8): 975-980, 1987

162- Kjorstad K.E. Kolbenstvedt A; Drevvante T. and Orjassaester H.: Lymphography and CEA in thediagnosis of metastases in patients with stage IB cancer of the cervix. Diagnostic Gynaecology andObstetrics, 2 (1): 71-75, 1980

163- Disaia, P. Haverback B.J. Dyce B.J and Morrow M: Carcinoembryonic antigen in patients withsquamous cell carcinoma of the cervix uteri and vulva.Surgery Gynaecology and obstetrics, 138:542-544 1974

164- Kjorstad K.E and Orjasaester, H: The prognostic significance of carcinoembryonic antigendeterminations in patients with adenocarcinoma of the cervixGynaecologic Oncology .,19:284-289, 1984

165-- Braun P and Leyendecker G (1983): Carcinoembryonic antigen in cervical cancer: A five yearfollow up. Cancer Detection and Prevention, 6:249-253 1983

166- Aaran R.K. Lehitnen M.; Knekt, P; Aromaa A and Hakama, M: Carcinoembryonic antigen levels insera drawn 1 to 12 years before diagnosis of cervical malignancy. Cancer Detection andPrevention,13:9-12,1988

167- Lam C.P. Yuan C.C Jeng F.S. Tsai L.C. Yeh S.H. and NG H.T: Evaluation of carcinoembryonicantigen, tissue polypeptide antigen, and squamous cell carcinoma antigen in the detection ofcervical carcinoma. Chung Hua 1. Hsuch Tsa Chih .,50 (1):7-13, 1992 .

168- Van Nagell J.R. Donaldson E.S. Wood E.C and Goldenberg D.M: The clinical significance ofcarcinoembryonic antigen in the plasma and tumours of patients with gynaecologic malignancies.

of 44


Cancer .,42: 1527-1532, 1978

169- Donaldson E; Van Nagell J.R. Wood E.G. Pletsch Q. and Goldeberg D.M: Carcinoembryonicantigen in patients treated with radiation therapy for invasive squamous cell carcinoma of theuterine cervix. Am. J. Reontgenol .,127:829-831, 1976

170- Meier W; Eiermann W; Stieber P; Moghadam F. and Hepp M: The course of squamous cellcarcinoma antigen and CEA as prognostic criteria for response to chemotherapy in cervix cancer.Gebutshilfe Frauenheilkd, 49 (12) 1050-1055, 1989

171- Miller A.B Hoogstraten B; Staquest M. and Winkler A: Reporting results of cancer treatment.Cancer 47: 207-214, 1981

172--Hiura M; Yorishima M; Tanaka M; Yokoyama T; Myoga H; Chiba T and Nakanishi Y: A clinicalstudy on multiple tumour markers following therapy in patients with uterine cervical carcinoma.Nippon Sanka Fujinka Gakkai 42 (5): 471-478 1990

173- Duk J.M Aalders J.G Fleuren G.J. krans M and De Bruijn HW ,. Tumour markers CA 125 squamouscell carcinoma antigen, and carcinoembryonic antigen in patients with adenocarcinoma of theuterine cervix. Obstet Gynaecol 73 (4): 661-668 1989

174- Magon H. Daunter B; Khoo S.K. and Mackay E.V: A comparison of two radioimmunoassaymethods for the detection of carcinoembryonic antigen in patients with ovarian or cervical cancerGynaecologic Oncology 11:340-347, 1981

175- Van Nagell J.R. Meeker W.R. Parker J.C; Kashmiri, R. and McCollum A.B.: Carcinoembryonicantigen in intraepithelial neoplasia of the uterine cervix.Am. J. Obstet Gynaecol ., 126 (1): 105-109, 1976

176- Fray R.E Husain O.A. N.; to A.C.W Watts K. C. Lader S. Rogers G.T Papadimtious J.T. and MorrisN.F.: The value of immunohistochemical markers in the diagnosis of cervical neoplasia. Br. J. ObstetGynaecol 91:1037-1041, 1984

177- Lindergren J; Vesterinen E; Purola E. and Wahlstrom T.: Prognostic significance of tissuecarcinoembryonic antigen in mild dysplasia of the uterine cervix.Tumour Biology 6: 465-470, 1985

178- McDicken I.W. Mcmillan D.L. and Rainey M: Carcinoembryonic antigen levels in cervico – vaginalfluid from patients with intraepithelial and invasive carcinoma of the cervix. Eur. J. Cancer Clin Oncol18(10): 917-919 1982

179- Briese, V. Barten G; Strache R.; Buttner, H. and Voss, P : Detection of carcinoembryonic antigen(CEA) in cervico-vaginal irrigation fluids.Zentrable Gynecol 111 (4): 223-229, 1989 (Abstract)

180- Harlozinska A; Kula J. Stepinska B; Jelen M. Richter R and Slesak B: Cervical carcinoma antigen,carcinoembryonic antigen (NCA) in appraisal of uterine of cervix smears. Am J. Clin Pathol 83:301-307, 1985

181- Fujii S. Konishi I, Nanbu Y. Nonogaki H: Kobayashi F. Sagawa N. Mori t. and Endo K: Analysis ofthe levels of CA 125, carcinoembryonic antigen, and CA 19-9 in the cervical mucous for a detectionof cervical adenocarcinoma. Cancer 62:541:547, 1988

Source URL: http://www.obgyn.net/laparoscopy/carcinoembryonic-antigen-cea-cervical-neoplasia-review

Links:[1] http://www.obgyn.net/laparoscopy

of 44

http://www.obgyn.net/laparoscopy/carcinoembryonic-antigen-cea-cervical-neoplasia-review


[2] http://www.obgyn.net/pregnancy-and-birth[3] http://www.obgyn.net/authors/obgynnet-staff

of 44

carcinoembryonic antigen (cea) in cervical …...in practical terms, van nagell et al . (1981 )(5) ,...

Documents