Report of the workshop on the carcinoembryonic antigen (cea): the present position and proposals for future investigation

Phase I/II Trial of Anticarcinoembryonic Antigen Radioimmunotherapy, Gemcitabine, and Hepatic Arterial Infusion of Fluorodeoxyuridine Postresection of Liver Metastasis for Colorectal Carcinoma

OBJECTIVES

Report the feasibility, toxicities, and long-term results of a Phase I/II trial of ⁹⁰Y-labeled anticarcinoembryonic antigen (anti-CEA) (cT84.66) radioimmunotherapy (RIT), gemcitabine, and hepatic arterial infusion (HAI) of fluorodeoxyuridine (FUdR) after maximal hepatic resection of metastatic colorectal cancer to the liver.

METHODS

Patients with metastatic colorectal cancer to the liver postresection or ablation to minimum disease were eligible. Each cohort received HAI of FUdR for 14 days on a dose escalation schedule. The maximum HAI FUdR dose level planned was 0.2 mg/kg/day, which is the standard dose for HAI FUdR alone. On day 9, ⁹⁰Y-cT84.66 anti-CEA at 16.6 mCi/m² as an i.v. bolus infusion and on days 9-11 i.v. gemcitabine at 105 mg/m² were given. Patients could receive up to three cycles every 6 weeks of protocol therapy. Four additional cycles of HAI FUdR were allowed after RIT.

RESULTS

Sixteen patients were treated on this study. A maximum tolerated dose of 0.20 mg/kg/day of HAI FUdR combined with RIT at 16.6 mCi/m² and gemcitabine at 105 mg/m² was achieved with only 1 patient experiencing grade 3 reversible toxicity (mucositis). After surgery, 10 patients had no evidence of visible disease and remained without evidence of disease after completion of protocol therapy. The remaining 6 patients demonstrated radiological visible disease after surgery and after protocol therapy 2 patients had a CR, 1 patient had PR, 2 had stable disease, and 1 had progression. With a median follow-up of 41.8 months (18.7-114.6), median progression free survival was 9.6 months. Two patients demonstrated long-term disease control out to 45+ and 113+ months.

CONCLUSION

This study demonstrates the safety, feasibility, and potential utility of HAI FUdR, RIT, and systemic gemcitabine. The trimodality approach does not have higher hematologic toxicities than seen in prior RIT-alone studies. Future efforts evaluating RIT in colorectal cancer should integrate RIT with systemic and regional therapies in the minimal tumor burden setting.

Association of Cea Test and Liver Scan in Detection of Hepatic Metastases of Gastrointestinal Carcinoma

The authors evaluate the combined use of liver scan and the CEA test in the diagnosis of hepatic metastases of carcinoma of the gastrointestinal tract. Association of the two tests is justified by the fact that the liver scan is very specific but not very sensitive, whereas the CEA test is more sensitive and not very specific. The sensitivity of the CEA test, on the other hand, can be increased by increasing the threshold of normality. However, the associated diagnostic use of the liver scan and the CEA test gives a loss of specificity with respect to the use of the liver scan alone. The present study, carried out on a series of 376 patients affected by gastrointestinal tumors of which 79 were of the stomach (9 with hepatic metastases), 133 of the colon and higher sigmoid (25 with hepatic metastases), and 164 of the lower sigmoid and rectum (29 with hepatic metastases), proposed to establish by use of a statistical method the optimal threshold of the CEA test that would give the best diagnostic specificity of the combined CEA test and liver scan without any relevant loss of sensitivity. A threshold of 26 ng/ml of the CEA test gave a specificity of 92 %, a sensitivity of 80 %, and an accuracy of 90 %. The authors think that in the detection of liver metastases of gastrointestinal tumors, the combined test can be more helpful the less the probability, for a given patient, for other metastatic localizations.

Preparation, characterization, and biological evaluation of a streptavidin-chimeric t84.66 conjugate for antibody pretargeting

In order to develop a new system for the antibody pretargeting of carcinoembryonic antigen (CEA)-positive cancers, a streptavidin (SA) conjugate of the monoclonal antibody (mAb) chimeric, T84.66, was synthesized and characterized. Antibody disulfide bonds were reduced with 1,4-dithiothreitol, and the resulting thiols were reacted with succinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate-derivatized streptavidin (SMCC-SA). The desired SA-cT84.66 conjugate was purified by iminobiotin affinity chromatography and size-exclusion high-performance liquid chromatography (HPLC). The molecular weight of the SA-cT84.66 conjugate (210 kDa) and immounoreactivity (100%) were confirmed by size-exclusion HPLC, and the conjugate bound three equivalents of (111)In-DOTA-biotin. SA-cT84.66-pretargeted (111)In-DOTA-biotin was evaluated in nude mice bearing LS174T human colon carcinoma xenografts. Tumor uptake of (111)In-DOTA-biotin peaked at 3.32% injected dose per gram after 15 minutes. Clearance from blood and normal organs was extremely rapid, and tumor-to-blood ratios were >20:1 after 24 hours. The specific tumor targeting and rapid whole-body clearance of SA-cT84.66-pretargeted (111)In-DOTA-biotin indicated that this system is promising for the imaging and therapy of CEA-positive cancers.

Embryonic reversions and lineage infidelities in tumour cells: genome-based models and role of genetic instability

Reversions to "embryonic precursor"-type cells and infidelities of tumour cell lineage (including metaplasias) have been recognized as aspects of various tumour types since the 19th century. Since then, evidence of these phenomena has been obtained from numerous clinical, biochemical, immunological and molecular biological studies. In particular, microarray studies have suggested that "aberrant" expressions of relevant genes are common. An unexplained aspect of the results of these studies is that, in many tumour types, the embryonic reversion or lineage infidelity only occurs in a proportion of cases. As a parallel development during the molecular biological investigation of tumours over the last several decades, genetic instability has been found much more marked, at least in some preparations of tumour cells, than that identified by means of previous karyotypic investigations of tumours. This study reviews examples of embryonic reversion and lineage infidelity phenomena, which have derived from the various lines of investigation of cancer over the last 150 or so years. Four categories of circumstances of the occurrence of embryonic reversions or lineage infidelities have been identified - (i) as part of the defining phenotype of the tumour, and hence being presumably integral to the tumour type, (ii) present ab initio in only some cases of the tumour type, and presumably being regularly associated with, but incidental to, the essential features of the tumour type, (iii) occurring later in the course of the disease and thus being possibly a manifestation of in vivo genetic instability and "tumour progression" and (iv) arising probably by genetic instability, during the processes, especially cell culture, associated with ex vivo investigations. Genomic models are described which might account for the origin of these phenomena in each of these circumstances.

Determination of tumor marker levels in cystic fluid of benign liver cysts

The levels of tumor markers in cystic fluid and serum were measured in six patients with benign biliary cyst of the liver. AFP in the cystic fluid was lower than the upper normal limit for serum in all cases, and CEA in the cystic fluid was higher than the upper normal limit for serum in one of the six cases. CA19-9, DU-PAN 2, and SPAN 1 in cystic fluid were much higher than the upper normal limit for serum in all cases (more than 100-fold for CA19-9, twofold for DU-PAN 2, and ninefold for SPAN 1). CA19-9, DU-PAN 2, and SPAN 1 in cystic fluid were significantly higher than the levels in the corresponding serum. Positive immunohistochemical staining against CA19-9, DU-PAN 2, and SPAN 1 was observed in the cytoplasm of the epithelial cells of the cyst wall. These results suggested that the high concentrations of CA19-9, DU-PAN 2, and SPAN 1 in the cystic fluid were due to secretion from the epithelial cells in the benign biliary cysts.

CEA and carbohydrate antigens in normal and neoplastic colon mucosa. An immunohistochemical study

Antibodies raised against tumour-associated antigens have been assessed for tumour selectivity using an indirect immunohistochemical peroxidase staining of formalin-fixed, paraffin-embedded tissue from colon carcinomas, colon polyps and normal mucosa. The following antibodies were used: 1) Unabsorbed polyclonal antibody to carcinoembryonic antigen (poly-CEA). 2) Monoclonal antibodies to CEA (mabs 3851 and 27). 3) Monoclonal antibodies to protein-bound carbohydrates (mabs C 216 and C 242) or to lipid- and protein-bound carbohydrates (C 50 and 19-9). These antibodies had been produced by hybridization of lymphocytes from mice, immunized with colon carcinoma cell lines or colon cancer tissue. All antibodies were used in one concentration only, preselected by initial titration experiments. No antibody was completely tumour-specific, but four antibodies, mabs 3851, 27, C 216 and C 242, showed statistically significant tumour selectivity. Using these antibodies, respectively 19, 19, 19, and 18 of 20 colon cancer were stained compared with 3, 4, 4, and 8 of 15 specimens of colon mucosa from normal controls. An increased frequency of staining was also noted in dysplastic polyps (statistically significant using mabs 3851 and C 216) and in dysplastic mucosa adjacent to a tumour (statistically significant using mabs 3851 and 27). The staining frequency of normal colon mucosa in cases of colon cancer did not differ from that in the normal controls. Poly-CEA and the anti-ganglioside mabs C 50 and 19-9 revealed no tumour selectivity. A pronounced goblet cell staining was seen using C 50, C 242 and 19-9.

Multi-centre study of a new CEA enzyme immunoassay using monoclonal antibodies

The use of a new monoclonal enzyme immunoassay (EIA) for the carcinoembryonic antigen (CEA) (Enzymun-Test CEA) was evaluated in a multi-centre study. Fifteen different laboratories [participated in the study. Data from the investigation were analysed in terms of precision, sensitivity, specificity and correlation with other test methods. The intra-assay coefficient of variation was between 1.3% at 23.0 microg/l CEA and 13.9% at 1.3 microg/l CEA. Inter-assay reproducibility ranged from 3.6% to 19.2%. The apparent sensitivity of the new EIA for CEA was approx. 0.5 microg/l CEA. The findings indicate that lipaemic and haemolytic sera and samples taken from icteric, rheumatic and dialysis patients did not have any influence on the results. There was no evidence that drugs commonly used in the treatment of carcinoma patients have any influence on the assay results. A good correlation between the new EIA for CEA and six other CEA enzyme immunoassay or radioimmunoassay methods was registered. These results seem to be of significance in particular for the monitoring of therapy for carcinoma patients. The new EIA for CEA exhibits a high degree of sensitivity, specificity and reproducibility.

Immunohistochemical demonstration of carcinoembryonic antigen (CEA) on tissue sections from squamous cell head and neck cancer and plasma CEA levels of the patients

45 squamous cell head and neck cancers including 36 with carcinoma of the oral cavity and 9 with carcinoma of the maxillary sinus were examined immunohistochemically for the presence of CEA. 12 of 30 carcinomas of the oral cavity and 7 of the 9 carcinomas of the maxillary sinus had tumors containing CEA. This difference in the occurrence of CEA was statistically significant. The mean plasma CEA level of 36 patients with carcinoma of the oral cavity and 7 patients with carcinoma of the maxillary sinus was 1.95 +/- 1.72 ng/ml and 3.70 +/- 3.53 ng/ml, respectively. Significant elevation of plasma CEA levels was found only in the stage-IV group patients with carcinoma of the oral cavity as compared with the stage-I group patients. In the 3 patients having plasma CEA values exceeding 5 ng/ml at the time of pretherapy, plasma CEA levels were decreased to below 2.5 ng/ml with the cancer treatment followed by the complete remission. These findings indicate that plasma CEA as a tumor marker in squamous cell head and neck cancer is meaningful in a small proportion of the cancer patient population.

The detection and evaluation of human tumor metastases

Many biochemical indices are purported to have clinical utility in the detection and management of neoplasia. Experience gained during the past decade tends to indicate their having a more important role in the detection and monitoring of metastases than of the primary lesion. From this present review of some of the commoner human tumours, it is concluded that such marker substances are important adjuncts in the management of germ cell and certain endocrine and endocrine-related tumours. The carcinoembryonic antigen (CEA) provides a marker for many gastrointestinal cancers, but there are no presently available substances with clinical usefulness for either breast or lung neoplasms. Alternative approaches to the detection of metastases are also presented. The particular use of antibody probes at an immunohistochemical level has been claimed to be able to detect micrometastastic disease in bone marrow or tumour-related monoclonal antibody probes may have application to other cancers in the future.

Immunohistochemical demonstration of carcinoembryonic antigen (CEA) and its correlation with grading and staging on tissue sections of urinary bladder carcinomas

In 150 transitional cell carcinomas (TCC) of the urinary bladder, 50 each of Grades 1, 2, and 3, the content of carcinoembryonic antigen (CEA) was examined immunocytochemically and correlated to grading and staging. Fifty-seven percent of the TCC contained CEA-positive tumor cells. Their distribution in tumor tissue is described. They were found in 24% of Grade 1 cases, in 72% of Grade 2 carcinomas, and in 76% of Grade 3 tumors. None of the Grade 1 cases contained more than 10% CEA-positive cells, whereas 34% and 40% of the Grade 2 and 3 tumors, respectively, revealed more than 10% CEA-positive tumor cells. According to the correlation found between grading and staging, the percentage of TCC containing CEA-positive tumor cells was 34% in pTA, 59% in pT1, and 80% in pT2/3 tumors. The results show a correlation between CEA content of tumor tissue and histopathologic malignancy in TCC of urinary bladder.

Monoclonal antibodies to carcinoembryonic antigen (CEA): characterization and use in a radioimmunoassay for CEA

Fusion of immune spleen cells of mice immunized with CEA with a mouse myeloma cell line resulted in several hybrid cell lines secreting antibodies to CEA as tested by radio-immunoassay. Four of these were cloned by limiting dilution and large amounts of anti-CEA antibodies were produced by growing the anti-CEA-producing cell clones intraperitoneally in mice and collecting the ascites fluids formed. Ascites fluids containing antibodies with both high and low affinity were obtained. The antibody having the highest affinity was shown to be specific to CEA, whereas one antibody showed cross-reaction with the normal cross-reacting antigen (NCA). The establishment of a competitive-inhibition doubleantibody radio-immunoassad with high sensitivity and specificity using mouse monoclonal antibodies is described.

Radioimmunoassay for carcinoembryonic antigen and alpha 1-fetoprotein in the qualitative evaluation of focal hepatic lesions in Japan

The combined tests of serum alpha 1-fetoprotein (AFP) and carcinoembryonic antigen (CEA) were routinely performed in 210 patients with focal hepatic lesions on a 99mTc-colloid liver scan in order to determine whether these could provide more useful information that AFP test alone in the qualitative evaluation of focal hepatic lesions. The predictive value of hepatoma with positive AFP alone remained 80%. However, when the negative CEA was combined with positive AFP, the predictive value of hepatoma (91%) with both was greatly increased. On the other hand, the predictive value of metastatic liver cancer with positive CEA showed 92%. The combined test of AFP and CEA may be useful for preserving high predictive values of hepatoma and metastatic liver disease.

Antigenic similarity between squamous cell carcinoma of horn (horn cancer) and normal bovine foetal tissues

Normal bovine foetal (liver and skin) and horn cancer tissue antigens were examined using double diffusion agar gel precipitation and immuno-electrophoretic tests to detect any cross reactivity among them. Rabbit horn cancer antisera absorbed with normal bovine liver, skin and horn core epithelium antigens, when tested with foetal skin and liver (4 to 6 months of gestation), revealed the presence of 2 foetal antigens in horn cancer. Immuno-chemically 2 of the horn cancer antigens were found to be identical to the bovine foetal antigens.

Measurement of carcinoembryonic antigen in patients with bronchogenic carcinoma

Estimation of CEA levels by the Z-gel method indicates that smokers, patients with limited lung cancer and patients with extensive lung cancer have higher values than nonsmoking controls. The CEA levels within each group are significantly different from one another. Use of CEA estimation for diagnostic purposes is limited because of the considerable overlap between normal controls and patients with cancer, the relatively low incidence of elevated values in patients with limited disease and the high incidence of false negatives (20%) even in patients with extensive disease. Elevated CEA values are associated with a poor prognosis and could be of clinical value as an addition to clinical staging to determine survival particularly for patients with extra-thoracic disease. Persistently high values in patients deemed clinically disease-free postoperatively are indicative of residual disease and a poor prognosis. If and when effective therapy for bronchogenic carcinoma becomes available, monitoring of CEA values may be useful in some patients as an early indication of release. Further studies are required to determine if the extraordinarily poor prognosis associated with marked elevations of CEA may be used as an additional criterion of inoperability in such patients.

The monitoring role of plasma CEA alone and in association with other tumor markers in colorectal and mammary carcinoma

During the past decade, evidence has accumulated to show that most, if not all, human tumors produce a variety of different factors which, if they pass into the blood and/or urine, may serve as tumor index substances (tumor markers).7 Tumor markers may either be: 1) tumor-derived--i.e., produced by the tumor itself, or 2) tumor-associated--i.e., produced by other tissues in response to the presence of the tumor and its local or distant effects on that tissue. Examples of this latter category include the changes in urinary hydroxyproline output in patients with bone metastases or the altered levels of serum acute phase proteins in neoplasia in general.7 Tumor-derived markers may be produced by either the tumor cell population itself, e.g., CEA, alpha-fetoprotein (AFP), and other oncofetal antigens, inappropriate hormones such as ACTH etc., or by their supporting framework (stroma), e.g., the osteolysins of human breast cancer.3

Proteolytic release of antigenic fragments corresponding to normal fecal antigen and non-specific cross-reacting antigen from carcinoembryonic antigen

Three immunogenic parts have so far been identified in the carcinoembryonic antigen (CEA) molecule. These are: determinants cross-reactint with the normal fecal antigen (NFA) (NFA determinant); determinants cross-reacting antigen (NCA) (NCA determinant); and determinants which appear to be more cancer-specific (cancer determinant). The chemical nature of these parts of the CEA molecule was investigated by digestion with proteolytic enzymes together with anti-CEA preparations with which these three immunogenic parts of CEA molecule could be identified. The CEA digest obtained with pepsin did not react in immunodiffusion and radioimmunoassay, indicating that pepsin completely destroyed all the antigenic parts. Digestion by pronase E destroyed only the cancer determinant and liberated two antigenic fragments corresponding to the NFA determinant and the NCA determinant, respectively. These results suggest that the cancer determinant may reside in a protein or a peptide part of the molecule. The chemical nature of the NFA and NCA determinant remains to be clarified.

Carcinoembryonic antigen (CEA) in the normal human small intestine: a light and electron microscopic study

An immunoperoxidase method for the demonstration of carcinoembryonic antigen (CEA) in tissues was applied to formalin-fixed paraffin embedded and glutaraldehyde-fixed resin sections of normal human small intestine. CEA could easily be demonstrated coating the surface of the small intestine, lining the crypts, and in goblet cells, indicating its presence there in considerable concentration. At the ultrastructural level CEA was localised in the glycocalyx and in mucin granules of goblet cells but not intracytoplasmically.

Serial carcinoembryonic antigen assays in patients with metastatic carcinoma of prostate being treated with chemotherapy

Serial carcinoembryonic antigen (CEA) assays were conducted in patients with endocrine-unresponsive prostatic adenocarcinoma who were being treated with multidrug chemotherapy. Changes in CEA correlated with the clinical status of the patient in 70 per cent of the determinations and were more accurate than acid phosphatase in monitoring the response to treatment.

Clinicopathological role of tumour index substances in paediatric neoplasia

Images

Carcinoembryonic antigen and cystic fibrosis protein in blood from cystic fibrosis homozygotes and heterozygote carriers

Carcinoembryonic antigen (CEA) activity was measured by radioimmunoassay in blood from cystic fibrosis (CF) homozygotes, heterozygote carriers of CF, normal healthy controls, and other patient controls with carcinomas involving gastrointestinal organs. All samples were also screened by electrofocusing for cystic fibrosis protein (CFP), a metabolic marker previously shown to be associated with the CF gene. Significantly increased levels of CEA activity were found in all CFP-positive groups; however, with one exception all patient controls with marked increases in CEA activity were CFP-negative. Immunodiffusion of perchloric acid extracts of CEA-like material from heterozygote carrier blood indicated that the CEA-like material, which was elevated in homozygotes and heterozygotes for CF, showed only partial identity with two separate CEA preparations obtained from colon carcinomas and was not identical to either A, B, or O(H) blood group substances. This glycoprotein material did, however, react with three different anti-CEA antisera. Our finding of an abnormally increased glycoprotein in cystic fibrosis, taken together with previous reports demonstrating abnormalities in the carbohydrate portion of glycoproteins found in various exocrine secretions in CF, further suggests that the primary defect in this disease is manifested partly as a defect in glycoprotein metabolism. This defect may result from an abnormality in one or more of the glycosyltransferases, possibly caused by a more primary defect in polyamine metabolism.

Carcinoembryonic antigen-like substances of human bile. Isolation and partial characterization

Three species of carcinoembryonic antigen (CEA)-like macromolecules, called the biliary glycoprotein I, II and III (BGP I, II and III) were identified in human bile. BGP I was found in normal gall-bladder bile and hepatic bile but not in bile subjected to inflammation ("white bile"). It was immunologically related to CEA and NGP (the "normal CEA-like glycoprotein". Synonyms: NCA, CCEA-2, etc.). BGP I differed immunologically from CEA in that it lacked the tumor-associated determinants of CEA. It was different from NGP (and CEA) in that it contained BGP I specific determinants as revealed by anti-BGP I antibodies. In bile from gallbladders with obstructed outlet and subjected to cholecystitis, a non-malignant inflammatory process, BGP I was replaced by BGP II and BGP III. Immunologically and physicochemically, BGP II and BGP III appeared to be closely similar to NGP and CEA, respectively.

Monkey antisera with increased specificity to carcino-embryonic antigen (CEA)

Three Macaca irus monkeys were immunized with purified human CEA. One received unmodified CEA and two were immunized with haptenated (4-hydroxy-5-iodo-3-nitrophenyl acetyl) CEA. All three monkeys formed precipitating antibodies to CEA. The antisera did not precipitate the CEA-related normal glycoprotein antigen (NCA). In contrast, CEA-immunized rabbits, sheep and goats invariably form antibodies against NCA. Radio-immunoassays showed the monkey antisera to contain trace amounts of antibodies to NCA but the titers were about 10-1,000 times lower than those in rabbit and sheep sera with comparable anti-CEA titers. These results show that monkeys produce antibodies against CEA, but respond weakly to a normal, CEA-related antigen. The weak response to NCA suggests that monkeys may possess an NCA-related antigen. Antisera lacking reactivity to normal CEA-related antigens may be helpful in establishing a more specific diagnostic test for CEA.

Cancer of the large intestine in man: progress and problems

This article reviews the high incidence of colon cancer in Western countries; the contribution of cell biology and biochemistry to our knowledge of the behaviour of the disease is discussed. The use of a series of markers to obtain early evidence of metastasis and to make the prognosis is described. Some suggestions are made how to obtain improvements in the survival percentages.

Role of serial plasma C.E.A. assays in detection of recurrent and metastatic colorectal carcinomas

Serial estimations of plasma carcinoembryonic antigen (C.E.A.) levels have been carried out in 220 patients with colorectal carcinomas who had potentially and apparently curative surgery. In a two-year follow-up period 53 patients developed recurrences or metastases. In 36 of these patients sustained rises in plasma C.E.A. titres occurred synchronously with or between three and 18 months before the clinical detection of recurrences or metastases. The use of serial plasma C.E.A. assays is therefore recommended as an additional diagnostic aid for the earlier detection of recurrent or metastatic colorectal carcinomas.

Detection of recurrence of large-bowel carcinoma by radioimmunoassay of circulating carcinoembryonic antigen (C.E.A.)

The usefulness and limitations of the carcinoembryonic antigen (C.E.A.) radioimmunoassay for the evaluation of tumour resection and for the detection of tumour relapse were studied in patients with large-bowel carcinoma. The level of plasma-C.E.A. was determined before any treatment in a group of 101 patients with histologically proven adenocarcinoma of the colon and rectum. 71% of all patients and 63% of cases with localised tumour (Dukes A and B) had a preoperative C.E.A. value of 5 ng. per ml. or higher. This limit was reached by only 1 of 90 apparently healthy, non-smoking blood-donors. Among 45 patients for whom a complete tumour resection was reported, all patients except 5 showed a drop of C.E.A. to normal values after surgery. The 5 patients whose C.E.A. did not fall to below 5 ng. per ml. showed a subsequent rise in C.E.A. level and were all found later to have a tumour relapse. The results indicate that an incomplete drop of circulating C.E.A. level one month after surgery has a bad prognostic significance. 22 of these patients were followed up by repeated C.E.A. radioimmunoassay for several months after surgery. 8 showed a progressive increase in C.E.A. levels preceding clinical diagnosis of tumour relapse by two to ten months. 6 other patients showed a moderate increase in C.E.A. levels, suggesting a tumour relapse not yet clinically detectable. The remaining 8 patients showed no increase in C.E.A. level above 5 ng. per ml. and no clinical symptoms of relapse. The results demonstrate that relapses of colon and rectum carcinoma can be detected by increased C.E.A. levels months before the appearance of any clinical evidence.