Carcinoembryonic antigen (CEA) levels and tumor histology in colon cancer

Relation between Carcinoembryonic Antigen Levels in Colon Cancer Tissue and Serum Carcinoembryonic Antigen Levels at Initial Surgery and Recurrence

OBJECTIVES

Carcinoembryonic antigen (CEA) is widely used for postoperative surveillance of colon cancer. Even if serum CEA is negative at initial surgery, it may turn positive at recurrence. We investigated the relation between serum CEA levels and the immunohistochemical staining status of CEA in the primary and resected metastatic tissues.

METHODS

Out of 224 patients with recurrent colon cancer between 1998 and 2012, we studied 46 patients in whom serum CEA levels were measured and immunohistochemical staining for CEA was possible in the primary and metastatic tissues.

RESULTS

The positive rate of serum CEA did not differ between initial surgery and recurrence, regardless of whether the cutoff value was set at 5 or 10 ng/ml (p = 0.829, p = 0.671). There was no relation between the CEA staining status and serum CEA level at initial surgery. However, the CEA staining status of metastatic tissue was significantly related to the serum CEA level at recurrence (p = 0.0046 and p = 0.0026).

CONCLUSIONS

The immunohistochemical staining status of CEA in metastatic tissue is closely related to the serum CEA level. This finding suggests that serum CEA levels are influenced not only by the CEA production capacity of cancer cells but also by the ability of the surrounding tissue to release CEA into the blood.

Soluble E-cadherin as a serum biomarker candidate: elevated levels in patients with late-stage colorectal carcinoma and FAP

Various strategies have been tested to identify serum biomarkers in patients with cancer. Recently, the entire of proteins released by cultured tumor cells into the media, the so-called secretome, has been suggested as a promising source for biomarker discovery. Ectodomains of membrane proteins cleaved from the cell surface represent a surprisingly abundant and apparently stable subset of this subproteome. Aiming for the detection of serum biomarkers for patients with colorectal cancer (CRC), we have previously detected significant amounts of the soluble form of E-cadherin in the secretomes of CRC cells. Here, we report a comprehensive analysis of sE-cadherin levels in sera from patients with CRC, colorectal adenoma, inflammatory bowel disease and familial adenomatous polyposis (FAP). Whereas mean sE-cadherin levels in patients with inflammatory bowel disease (mean: 4.7 μg/ml, SD: 1.5 μg/ml), with adenomas (mean: 4.6 μg/ml, SD: 3.0 μg/ml) and early stage cancers (mean: 4.9 μg/ml, SD: 4.7 μg/ml) do not significantly differ from healthy controls (mean: 4.8 μg/ml, SD: 1.9 μg/ml), patients with Stage III and Stage IV carcinomas display a significant increase (mean: 6.1 μg/ml, SD: 2.6 μg/ml). In individual patients with late-stage CRC, sE-cadherin serum levels directly reflect their disease status over time. These findings suggest a potential application of sE-cadherin as an alternative diagnostic biomarker for monitoring disease particularly in patients with carcinoembryonic antigen negative tumors. In patients with FAP, on the other hand, we also detected a significant increase of serum sE-cadherin levels (mean: 5.8 μg/ml, SD: 2.8 μg/ml), but this was regardless of their tumor load and colectomy status.

Differences in serum tumor markers between colon and rectal cancer. Comparison of CA 242 and carcinoembryonic antigen

PURPOSE

We investigated whether there are differences in serum levels of CA 242 and carcinoembryonic antigen (CEA) between patients with colon and rectal cancer.

METHODS

Preoperative serum levels of CA 242 and CEA were determined in 153 patients with colon cancer and in 107 patients with rectal cancer.

RESULTS

At the recommended cut-off levels for CA 242 and CEA, the overall sensitivity of CA 242 was 39 percent for both colon and rectal cancer, whereas the sensitivity of CEA was 40 percent for colon and 47 percent for rectal cancer. A combination of CA 242 and CEA increased overall sensitivity to 57 percent in colon cancer and to 62 percent in rectal cancer, whereas specificity decreased by 10 percent, compared with CEA alone. In colon cancer either or both markers were elevated in 38, 46, 56, and 84 percent of patients with Dukes Stages A, B, C, and D, respectively. Corresponding figures for rectal cancer were 52, 46, 71, and 87 percent, respectively.

CONCLUSIONS

CA 242 showed equal sensitivity for colon and rectal cancer. In Stages A, C, and D, sensitivity of CEA and of a combination of CEA and CA 242 was higher in rectal than in colon cancer, but the difference was not significant. Concomitant use of markers increased sensitivity sharply compared with use of a single marker both in colon and rectal cancer.

Carcinoma of the ampulla of Vater

A consecutive series of 36 Japanese patients with ampullary carcinoma who underwent a pancreatoduodenectomy at the Department of Surgery I, Kyushu University Hospital during the past 20 years were reviewed clinicopathologically to study prognostic factors. A univariate generalized Wilcoxon test showed that preoperative serum carcino-embryonic antigen (CEA) concentration, venous invasion, lymphatic permeation and perineural invasion were significant parameters. A multivariate Cox regression analysis showed that venous invasion was the only significant variable. In a sequential serum CEA follow up of 10 patients, an elevation of serum CEA levels was seen to correspond to the clinical development of a recurrence in six, while a high concentration of serum CEA was not evident despite the clinical manifestation of a recurrence in two, and serum CEA levels remained within the normal limits with no evidence of a recurrence in two others. According to the death certificates of 15 patients, where an exact site of metastasis was available, 11 died from liver metastasis, three from lung metastasis and one from peritoneal dissemination. These findings support the theory that a histologic invasion of the venous space is an independent prognostic factor and close attention should be paid to any signs of haematogenous metastasis, such as to the liver and lung, as well as to a serial serum CEA follow up.

Serum carcinoembryonic antigen in relation to survival, DNA ploidy pattern, and recurrent disease in 406 colorectal carcinoma patients

Serum carcinoembryonic antigen (CEA) levels in relation to survival, flow cytometric DNA ploidy pattern, Dukes stage, and recurrent disease was prospectively evaluated in 406 patients with colorectal carcinoma. In 246 patients (61%) the carcinomas were DNA aneuploid. Increased preoperative CEA levels (> 5 micrograms/l) were found in 151 of 363 evaluable patients (42%). Dukes stage-B patients with preoperative CEA elevation showed significantly poorer prognosis than those with normal CEA values (p = 0.001). A weak but significant correlation was found between preoperative CEA level and Dukes stage (Kendall's tau = 0.25, p < 0.01). Of 50 evaluable patients with clinical recurrence and postoperative normal or normalized CEA levels, 28 (56%) had a rise in CEA before or at the time of clinical recurrence. The sensitivity of the CEA test for primary and for recurrent disease was not significantly different in the DNA aneuploid and the DNA near-diploid groups.

Carcinoembryonic antigen levels of peripheral and draining venous blood in patients with colorectal cancer. Correlation with survival

Correlations between preoperative carcinoembryonic antigen (CEA) levels of peripheral (p-CEA) and draining blood (d-CEA), the CEA gradient between d-CEA and p-CEA (d-p CEA gradient) levels, and survival after resection of cancer lesions were examined in 94 patients with colorectal cancer. Survival rates of patients with normal p-CEA and d-CEA levels and d-p CEA gradient levels (less than 5 ng/ml) were significantly better than those of patients with abnormal levels (greater than or equal to 5 ng/ml), and the 5-year survival rates were, respectively, 62%, 69%, and 72% in the former and 42%, 41%, and 35% in the latter. The differences in the 5-year survival rates between patients with normal and abnormal d-p CEA gradient, d-CEA, and p-CEA levels were 37%, 28%, and 20%, respectively. Furthermore, the positive rates of d-CEA levels (64%) and d-p CEA gradient levels (48%) were higher than that of p-CEA levels (36%). However, some significant differences in background variables also were found between the respective groups of patients with normal and abnormal p-CEA and d-CEA levels and d-p CEA gradient levels. These results suggest that patients with poor prognoses are examined more effectively by determining their d-p CEA gradient and d-CEA levels than their p-CEA levels, and that CEA may be expressed as a quantitative sum total of various pathophysiologic variables of patients with colorectal cancer but not as an independent prognostic variable.

Colorectal cancer patients with high risk of hematogenous metastasis: correlation with CEA levels in peripheral and draining venous blood during the period of operation

Correlations between carcinoembryonic antigen (CEA) levels of peripheral (p) and draining (d) venous blood during the period of operation, and pre- and post-operatively detected hematogenous metastases were examined in 78 patients with colorectal cancer. The metastases were found in 28 patients (HM group), but not found in the other 50 patients (non-HM group). The mean values (43 and 198 ng/ml) and positive rates (61 and 96%) greater than 5 ng/ml of p- and d-CEA levels in the HM group were significantly higher than those (6 and 14 ng/ml, and 22 and 48%, respectively) in the non-HM group. The differences (mean 184 ng/ml and positive rate 49%) of d-CEA levels between both groups were more significant than those (39 ng/ml and 30%) of p-CEA levels. The mean value (155 ng/ml) and positive rate (82%) greater than 5 ng/ml of the gradient between d- and p-CEA levels (d-p CEA gradient) in the HM group were significantly higher than those (8 ng/ml and 34%) in the non-HM group. These results suggest that patients with a high risk of hematogenous metastases are more effectively checked by the determination of d-CEA levels and d-p CEA gradient than of p-CEA levels, and that they are patients with positive d-CEA and d-p CEA gradient levels.

Contents of tissue CEA and CA19-9 in colonic polyp and colorectal cancer, and their clinical significance

In precancerous states or early cancer, the serum levels of tumor markers are almost not detectable. Therefore, the tissue contents of CEA and CA19-9 were measured in 48 colonic polyps, 8 colorectal cancers and 5 normal colonic mucosa. These tissue specimens were obtained by endoscopic polypectomy, surgery or autopsy, and homogenated in normal saline (10 ml/wet g of tissue). After centrifugation, the supernatant was assayed by enzyme or radioimmunoassay. There was no correlation between serum levels and tissue contents of CEA or CA19-9 in colonic adenomas and colorectal cancers. The mean contents of tissue CEA and CA19-9 in colonic polyp and colorectal cancer were significantly higher than normal colonic mucosa, and the highest contents of CEA and CA19-9 were found in colorectal cancer. The contents of tissue CEA and CA19-9 in cancerous regions were markedly increased as compared with noncancerous regions. In adenomas, there was a relationship between the degree of histological dysplasia and the tissue content of CEA. Relationships were also found between macroscopic findings and tissue tumor markers in adenomas. These results suggest the possibility that the measurement of tissue tumor markers may be useful for borderline colonic lesions.

Carcinoembryonic antigen levels of portal blood in gastric cancer patients

Correlation between carcinoembryonic antigen (CEA) levels of peripheral venous and portal blood and six histopathologic and immunohistochemical variables was examined in 53 gastric cancer patients and in 8 patients with benign diseases, in order to clarify the elevation mechanism of CEA in the peripheral blood. Immunohistochemically, CEA was localized in 48 (90.6%) of the 53 cancer lesions. CEA levels of portal blood (with a mean of 136.5 ng/ml and positive rates greater than 5 ng/ml, 58.3%) were significantly higher than those (30.3 ng/ml and 22.9%) of peripheral blood in 48 patients with CEA localized cancer. However, CEA levels of portal blood were as low as 5 ng/ml and were not different from those of peripheral blood in all of the CEA nonlocalized cancer and benign diseases. Elevation of CEA in portal blood and also peripheral blood was most highly correlated with venous invasion, although CEA levels in portal blood were significantly associated with three other variables including tumor size, lymphatic invasion, and node metastasis. These variables relating to CEA elevation in the blood were significantly related to venous invasion, whereas a relationship between venous invasion and tumor differentiation and the CEA distributed pattern was not found. These results suggest that CEA may be mainly drained by the hematogenous portal system via the draining vein from CEA localized cancer cells in the invaded veins of gastric cancer lesions, and, additionally, that histopathologic CEA elevation-relating variables may secondarily affect the CEA elevation in the blood in association with the venous invasion.

Carcinoembryonic antigen and carbohydrate antigen 19-9 levels of peripheral and draining venous blood in colorectal cancer patients. Correlation with histopathologic and immunohistochemical variables

Correlation between carcinoembryonic antigen (CEA) and carbohydrate antigen (CA) 19-9 levels of peripheral and draining venous blood, and 11 histopathologic and immunohistochemical variables was examined in 83 patients with colorectal cancer. CEA levels of draining blood (mean 34.5 ng/ml and positive rate greater than 5 ng/ml, 60.2%) were significantly higher than those (13.0 ng/ml and 28.9%) of peripheral blood. However, CA19-9 levels (mean 576.1 U/ml and positive rate greater than 37 U/ml, 29.5%) of draining blood were not different from those (568.0 U/ml and 29.5%) of peripheral blood. Immunohistochemically, CEA was observed in all of the 83 specimens and distributed in most of all cancer cells, whereas CA19-9 was found in 52 (62.5%) of the 83 specimens and sporadically distributed in some parts of cancer lesions in general. Elevation of CEA levels in draining and peripheral blood was most highly correlated with venous invasion, although the levels were related to four other histopathologic variables including liver metastasis, invasive layer of colorectal wall, lymphatic invasion, and Dukes' classification. Significant correlation between the CEA localized pattern of cancer cells was not found. Patients with CA19-9 nonlocalized cancer showed no elevation of the antigen levels in both peripheral and draining blood. The elevation of CA19-9 levels in peripheral blood of patients with CA19-9 localized cancer was most highly associated with lymphatic invasion, although the levels were correlated with five other variables consisting of liver metastasis, tumor differentiation, invasive layer of colorectal wall, venous invasion, and Dukes' classification out of 11 histopathologic and immunohistochemical variables. CEA levels of draining blood rose from 18.2 ng/ml and 40.3% to 30.1 ng/ml and 72.6%, respectively, after operative stimuli to cancer lesions, whereas the change of CA19-9 levels in draining blood of patients with CA19-9 localized cancer was not found during the time of operation. These results suggest that CEA may be drained mainly by the hematogenous portal system by the draining vein from the cancer cells in the invasive veins and that CA19-9 may be drained by the thoracic duct of the lymphatic system. It is also suggested that the CEA and CA19-9 elevation-relating variables may secondarily affect the CEA and CA19-9 elevation in the blood in association with the venous and lymphatic invasion of cancer lesions, respectively.

Comparison of carcinoembryonic antigen levels between portal and peripheral blood in patients with colorectal cancer. Correlation with histopathologic variables

Correlation between carcinoembryonic antigen (CEA) levels of peripheral and portal blood, and eight histopathologic variables, was examined in 66 patients with colorectal cancer. The change in CEA levels in the portal blood of 40 patients during operation was also examined in relation to histopathologic variables. CEA levels of portal blood (with a mean of 26.6 ng/ml and positive rate greater than 5 ng/ml, 59.1%) were significantly higher than those of peripheral blood (8.1 ng/ml, 33.3%). Elevation of CEA levels in portal and peripheral blood was most highly correlated with the venous invasion. Although the levels in the portal blood were related to six other histopathologic variables including tumor size, tumor differentiation, node metastasis, lymphatic invasion, invasive layer of the colorectal wall, and Dukes' classification except tumor location. CEA levels rose from 19.4 ng/ml and 40% to 43.6 ng/ml and 90.2% respectively following operative stimuli to cancer lesions with venous invasion. However, the levels did not rise in the lesions without the invasion. CEA levels of peripheral blood were as low as 5 ng/ml in three out of eight patients with liver metastasis. However, the levels in portal blood were much greater than 5 ng/ml in all of the patients. These results suggest that CEA may be hematogenously drained by the portal system via the draining vein from the cancer cells in the invasive veins but not by the thoracic duct of the lymphatic system.

Prognostic value of preoperative serum CEA level compared to clinical staging. IV. Histological grading and tumor type in colorectal and gastric cancer

In a clinical investigation of observed postoperative survival, 410 patients with colorectal cancer and 269 patients with stomach cancer have been registered for primary surgical treatment connected with a long-term follow up. Histologic grading and tumor typing were examined as potential prognostic factors and compared with the prognostic information covered by operability, tumor extension, and the preoperative CEA level. Statistical treatment of the data revealed no prognostic significance of the tumor types adenocarcinoma, mucinous, and anaplastic tumours in gastric cancer. Histologic grading specified ranges associated with significant differences in survival of gastric and colorectal cancer patients. However, histologic grading did not provide prognostic information in addition to operability and tumor extension. However, histologic grading gave additional prognostic information to preoperative CEA levels in the range of 0-5 micrograms CEA/1 serum but not in the range greater than 5 micrograms CEA/1. The results indicate that the prognostic information of preoperative serum CEA level is not directly linked to the histologic grade of a tumor.

Immunohistochemical study of carcinoembryonic antigen in patients with colorectal cancer. Correlation with plasma carcinoembryonic antigen levels

Using peroxidase-antiperoxidase (PAP) method, carcinoembryonic antigen (CEA) was demonstrated in conventionally processed colorectal cancer tissues. A new immunohistochemical grading for colorectal cancers based on the mode of the localization was made in an attempt to clarify the factors responsible for elevation of plasma CEA levels in colorectal cancer patients. Most of the patients with well differentiated adenocarcinoma, in which CEA was densely distributed along the apical surface but only rarely present along the basolateral surfaces of the carcinoma cells, had very low levels of plasma CEA, whereas all patients showing CEA distribution in the stroma as well as over the entire surfaces of the cancer cells and their cytoplasm showed high plasma CEA levels. In addition, there was a good correlation between the grading and presence of the blood vessel and lymphatic invasions. Thus, the appearance of CEA in the surrounding stroma, due to abnormal distribution of CEA on the basolateral plasma membrane of cancer cells, may play a significant role in the elevation of plasma CEA levels in colorectal cancer patients.

Tissue CEA in colorectal carcinoma

Immunoperoxidase method can be used to detect cellular or tissue CEA. Forty formalin-fixed paraffin-embedded specimens of colorectal carcinoma, 19 from patients who survived for five years after resection, and 21 from patients who died within five years were studied. Cellular CEA was present in 100 per cent of the specimens. Accurate quantitative evaluation of cellular CEA is currently not feasible. Therefore, the mere presence of cellular CEA has no prognostic value.

The clinical value of immunohistochemically demonstrable CEA in breast cancer: a possible method of selecting patients for adjuvant chemotherapy

The production of carcinoembryonic antigen (CEA) by human breast cancer tissue has been studied in relation to the prognosis of patients with breast cancer. All of the patients were in a controlled trial of adjuvant chemotherapy for the treatment of operable breast cancer. CEA was studied in primary tumours and axillary node metastases from these patients using an immunoperoxidase (PAP) method. Sections of 290 primary carcinomas and 217 axillary metastases were examined for CEA. The CEA status of the primary tumours was of no value as a prognostic indicator nor in the selection of patients for chemotherapy. In contrast, patients could be divided into 3 groups on the basis of the CEA results in the axillary nodes. In one group, in which cases were strongly positive for CEA (24% of the total) the prognosis, as reflected by recurrence free survival, was relatively good and chemotherapy produced no further advantage. In another group in which cases were weakly positive for CEA (18% of the total) the prognosis was poor but chemotherapy produced significant improvement. In a third group, in which cases were negative for CEA (58% of the total) the prognosis was poor and was not improved by chemotherapy, at least in the short term. Thus, the CEA status of axillary metastases may be clinically useful.

Plasma carcinoembryonic antigen concentrations and immunohistochemical patterns of epithelial marker antigens in patients with large bowel carcinoma

Carcinoembryonic antigen (CEA), secretory component (SC), and epithelial IgA were traced by paired immunofluorescence staining in 102 large bowel carcinomas from 99 patients. The immunohistochemical results were evaluated semiquantitatively in relation to histological tumour grade, clinicopathological stage, and preoperative plasma CEA concentration. CEA expression was significantly increased (p less than 0.05) in the following order: histologically normal colon mucosa, transitional mucosa adjacent to tumours, neoplastic epithelium; the reverse was true for the expression of SC and epithelial IgA (p less than 0.01). CEA was significantly more abundant in the moderately and poorly differentiated tumors than in the well differentiated ones (p less than 0.05), whereas the latter showed better expression of SC (p less than 0.05) and epithelial IgA (p approximately 0.06). In the transitional mucosa, CEA staining tended to be inversely related to histological tumour grade, whereas SC and epithelial IgA were significantly better seen in this zone when the adjacent tumour was well differentiated than when it was moderately or poorly differentiated (p less than 0.01). Furthermore, the expression of SC and epithelial IgA in the transitional mucosa decreased with increasing invasiveness of the tumours, whereas the opposite relation was indicated for CEA expression. Plasma CEA concentrations were not clearly correlated with histological levels than the localised well differentiated tumours tended to be associated with lower levels than the localised moderately differentiated ones (p approximately 0.06). Moreover, the latter variety was associated with lower plasma CEA concentrations than disseminated tumours of comparable differentiation (p less than 0.01).

Relationship of serum CEA levels to tumour size and CEA content in nude mice bearing colonic-tumour xenografts

The relationship of serum carcinoembryonic antigen (CEA) levels to tumour size and antigen content was studied in nude mice bearing well differentiated, mucinous human colonic-tumour xenografts. Blood samples were taken from normal nude mice and others bearing xenografts, whose size had been calculated from in vivo measurements; saline and KCl extracts were made of a proportion of these tumours. Sera and tissue extracts were assayed for CEA activity by double-antibody radioimmunoassay. Extracts were also made from the livers and spleens of tumour-bearing and normal nude mice. All normal sera and 78% of sera from tumour-bearing animals had CEA values less than 11.4 ng/ml. No clear correlation was found between serum CEA levels greater than 11.4 ng/ml and tumour size or weight, or between serum CEA and tumour CEA concentrations or total CEA burden. The concentration of CEA in those tumours tested varied from 1 to 22 microgram/g. Our results confirm and extend the conclusions reached by others (Stragand et al., 1980) studying the significance of serum CEA levels with xenograft model systems. The complexity of factors contributing to circulating CEA is discussed in the light of our findings.

Carcinoembryonic antigen (CEA) in diagnosis and prognosis of colorectal carcinoma

An attempt is made to define the usefulness and limitations of carcinoembryonic antigen (CEA) radioimmunoassay for evaluation of diagnosis, tumor resection, and detection of tumor relapse in 108 patients with colorectal carcinoma. Preoperative CEA levels were correlated with pathologic stage and tumor localizations. Increasing levels of CEA were found with the advanced stage of the disease (stage C and D lesions). Our results indicate that 1) an incomplete drop in circulating CEA level one month after surgery is a bad prognostic sign and 2) relapses of colonic and rectal carcinoma can be detected by increased CEA levels months before the appearance of any clinical evidence.

Carcinoembryonic antigen in breast-cancer tissue: a useful prognostic indicator

Sections of breast carcinomas removed from 69 patients six to 13 years previously were examined using an immunoperoxidase technique to determine whether carcinoembryonic antigen (CEA) was present. Patients who had CEA-negative tumours had significantly higher five- and 10-year survival rates. The difference was not related to the stage of the disease, postoperative treatment, or histological type of tumour. These results suggest that immunohistological assessment of CEA in breast-cancer tissue may provide more precise prognostic information.

Use of CEA as an indicator of early recurrence and as a guide to a selected second-look procedure in patients with colorectal cancer

The usefulness of the CEA as an indicator of recurrence and a guide to selected second-look surgery was evaluated from a retrospective analysis of 358 patients with colorectal cancer and from a prospective experience with 16 patients all of whom had been admitted for second-look surgery because of postoperative elevations of CEA only. Our previous experience had shown that after curative resection the CEA usually returned to normal levels (less than 5 ng/ml) within one month, but became elevated at time of clinically obvious recurrence being very high in patients with liver metastases, but only moderately elevated or normal in patients with local recurrence. All 16 patients had previously had curative resection of colorectal cancer; 13 in the rectum or rectosigmoid and three in the right colon. There were 13 Dukes' C and three Dukes' B cancers. All had been followed clinically and by CEA testing at three monthly intervals and were considered free of disease (NED) at time of CEA elevation. The median disease free interval was 13 months (range 4-57 months) and the median CEA prompting admission for second-look operation was 21 ng/ml (range 10-56 ng/ml). The sites of recurrence were liver in six, lung in two and localized disease in six. Two patients had negative exploration for recurrence and were found to have cholelithiasis only (one of these later died of metastases). Resection for cure was done in seven and palliative resection or biopsy only was done in nine patients. At this time, four patients are NED (12-37 months), five are living with disease (10-16 months) and seven have died of disease (2-12 months). The CEA test provides a method of early detection of recurrence and may permit surgical retrieval in selected patients and earlier initiation of palliation in other patients. The longterm effects in patient salvage remain to be defined.

Pre- and postoperative uses of CEA

CEA plasma levels obtained prior to definitive surgery in patients with colorectal cancer in our hands have predictive ability. An elevated CEA greater than 2.5 ng/ml recorded by our laboratory means an increased risk of subsequent local recurrence or of later metastatic disease. The question as to whether or not this is additive as a prognostic variable when tested against careful histopathological staging remains. As a monitor, CEA will detect recurrence. Again, the problem as to how accurate this is remains. If we use two consecutive elevations of plasma CEA greater than 2.5 ng/ml as a criteria, we encounter about 15% false positives which must be weighed against finding disease significantly earlier in about one-third of the patients followed. Our data for second-look procedures indicate clearly that when used in patients with an elevated CEA laparotomy may be useful and further studies showed the presence of disease in 11 of 14 patients with an elevation following surgery for two consecutive tests were greater than 2.5 ng/ml. Two were operable. The significance of these findings is described.

The quantitative relationship of carcinoembryonic antigen to tumor cells of different organs

Carcinoembryonic antigen (CEA) was measured in extracts of cell suspensions from tumors of the colon, stomach, breast, lung and kidney. Cell suspensions were also obtained from normal tissue surrounding the primary tumors. The number of neoplastic cells in each cell suspension was determined by cytologic criteria. Highest CEA concentrations were found in cells obtained from carcinoma of the bowel. There also appeared to be a greater difference between the concentrations of CEA in colonic tumors and either surrounding normal tissue or serum, than in tumors of other organs. There was no apparent correlation of the concentration of CEA in tumor cells and the presence or absence of metastases. The presence of CEA in normal cells from tissue adjacent to tumors was suggestive of increased synthesis of CEA by these cells in some cases.

Carcinoembryonic antigen in colorectal practice: report of three cases

CEA, in combination with the time-honored barium enema, sigmoidoscopy, rectal examination with guaiac determinations, and thorough clinical history, is a useful tool for the colorectal surgeon in following the progression of disease in both colorectal carcinoma patients and patients who have certain premalignant conditions. In this high-risk group of patients, CEA determinations as well as colonoscopy are becoming increasingly useful. That tumors on the left side are associated with higher CEA levels than those on the right side, due to vascular invasion or the grades of the tumors, and that often CEA antigenicity and severity of symptoms in ulcerative colitis correlate well, are emphasized. Finally, following total colectomy in five cases of ulcerative colitis, all CEA titers decreased to the normal range and have remained there.

Plasma CEA, tumor CEA, and tumor histology

Carcinoembryonic antigen (CEA) has been thought to be a diagnostic and prognostic indicator of colorectal cancer. Initial descriptions of CEA as a tumor specific antigen suggests a relationship between tumor CEA and circulating plasma CEA. To define the relationship between CEA and colorectal carcinoma, we have studied the CEA concentration of preoperative plasma, tumor tissue, and normal bowel distant from tumor in 35 patients who had clinically curative resections. Tumor histology was evaluated for Dukes class, histologic grade, necrosis, and vessel invasion. Regression analysis yielded no evidence of correlation between tumor CEA and plasma CEA. No correlation could be shown between tumor concentration of CEA and the histological parameters previously noted. CEA was found in all specimens of normal bowel. Furthermore, in 34% of the cases studied, the tumor CEA was not significantly higher than in normal bowel. No significant difference was shown when histopathological findings were compared to normal and abnormal plasma CEA values. These findings suggest the following conclusions: CEA is not tumor specific. Increased levels of CEA in tumor tissue are not a constant finding in colorectal carcinoma. Tumor levels of CEA do not appear to correlate with histologic degree of tumor differentiation. Elevated plasma levels of CEA do not necessarily connote elevated tumor tissue levels of CEA, and conversely, normal plasma levels of CEA do not necessarily mean low levels of tumor CEA.

Carcinoembryonic antigen: clinical and historical aspects

To further define and determine the usefulness of CEA, 1100 CEA determinations have been made over the past two years at The Ohio State University Hospitals on patients with a variety of malignant and nonmalignant conditions. Correlation of CEA titers with history and clinical course has yielded interesting results not only in cancers of entodermally derived tissues, for which CEA has become an established adjunct in management, but also in certain other neoplasms and inflammatory states. The current total of 225 preoperative CEA determinations in colorectal carcinomas shows an 81% incidence of elevation, with postoperative titers remaining elevated in patients having only palliative surgery but falling to the negative zone after curative procedures. An excellent correlation exists between CEA levels and grade of tumor (more poorly differentiated tumors showing lower titers). Left-side colon lesions show significantly higher titers than right-side lesions. CEA values have been shown to be elevated in 90% of pancreatic carcinomas studied, in 60% of metastatic breast cancers, and in 35% of other tumors (ovary, head and neck, bladder, kidney, and prostate cancers). CEA levels in 35 ulcerative colitis patients show elevation during exacerbations (51%). During remissions titers fall toward normal, although in 31% still remaining greater than 2.5 ng/ml. In the six colectomies performed, CEA levels all fell into the negative zone postoperatively. Forty percent of adenomatous polyps showed elevated CEA titers (range 2.5-10.0) that dropped following polypectomy to the negative zone. Preoperative and postoperative CEA determinations are important in assessing the effectiveness of surgery. Serial CEA determinations are important in the follow-up period and in evaluation of the other modes of therapy (e.g., chemotherapy). These determinations of tumor antigenicity give the physician added prognostic insight into the behavior of the tumor growth. Rectal examination with guaiac determinations, sigmoidoscopy, cytology, barium enema, and a good clinical evaluation remain the primary tools for detecting colorectal disease. However, in the high-risk patient suspicious of developing cancer, CEA determinations as well as colonoscopy are now being used increasingly and provide additional highly valuable tools in the physician's armamentarium.