5-Iododeoxyuridine increases the efficacy of the radioimmunotherapy of human tumors growing in nude mice.

Recently, there has been much interest in the use of radionuclide conjugated monoclonal antibodies for the treatment of human malignancies. One way to potentially maximize the therapeutic effectiveness of radioimmunotherapy would be to sensitize tumor cells to the radiation dose delivered by the antibody. Since radioimmunotherapy can potentially treat disseminated disease, including micrometastasis, we chose to study a halogenated pyrimidine radiosensitizer, a class of compounds that affect nonhypoxic cells. 5-Iododeoxyuridine, administered with pyrimidine metabolism modulators, increased the therapeutic effectiveness of radioimmunotherapy, resulting in individual cures of human tumors growing in BALB/c nu/nu (nude) mice. 5-Iododeoxyuridine was administered with N-(phosphonacetyl)-L-aspartic acid and 5-fluoro-deoxycytidine plus tetrahydrouridine. This drug treatment was combined with radioimmunotherapy using 131I conjugated to a monoclonal antibody, Mc5. Mc5 binds to a mucin component of the human milk fat globule. This antigen is expressed on the surface of MX-1 cells, the transplantable human tumor used in this study. Tumor-bearing mice treated with both the drug protocol and 131I-Mc5 (540 microCi, 10 microCi/micrograms) showed a regression in average tumor volume. The average tumor volume was reduced below the initial size at treatment for 50 days; two of five cures were obtained. Neither cures nor regressions were observed with either the drug or antibody treatments alone. Our results indicate the potential for increasing the therapeutic effectiveness of radioimmunotherapy of human solid tumors with halogenated pyrimidines.

Quality control test for immunoreactivity of radiolabeled antibody.

A rapid method for the measurement of the immunoreactive fraction of a radiolabeled monoclonal antibody or antibody fragment has been developed. This may be used as a quality control test prior to patient administration of the radiolabeled antibody preparation. The test employs solid phase antigens and the assay is conducted under conditions of antigen excess. Assay parameters have been evaluated and a standardized procedure has been developed. The assay has been compared to a standard extrapolation method and found to give approximately the same result. The test has been used on four different radiolabeled antibodies currently in clinical trials in patients with colorectal cancer. Mean immunoreactive fractions for these radiolabeled antibodies ranged from 35 to 65% and the variability of the immunoreactive fraction ranged from 140 to 240% for different antibodies. We conclude that the quality, defined as the immunoreactive fraction, of radiolabeled antibodies is both low and highly variable, indicating the need for a quality control test of these radiopharmaceuticals in the clinic prior to patient administration.

Anti-colon/ovarian tumor antigen: localization of colon cancer xenografts in athymic rats.

Tumor localization studies in athymic rats bearing human colon tumors were performed using the radioiodinated monoclonal antibody SP-21 and its F(ab')2. Antibody preparations isolated from ascitic fluid and antibody from bioreactor effluent were used in these studies with similar radioimmunolocalization results. The intact antibody had an optimal localization time of 4-8 days after injection, while the F(ab')2 fragments had an optimal localization time of 3-4 days. Whole-body autoradiography, whole-body immunohistochemistry, and scintigraphy confirmed that the intact antibody and the antibody fragments localized preferentially in the tumor. The antibody distribution within the tumor was uniform, and not confined to the periphery, nor to focal areas within the tumor. Dose-response studies were performed with the intact antibody over a range of 10-100 micrograms/kg of total body weight with no clear-cut relationships observed. Comparisons of different radio-iodination methods indicated that the chloramine-T-based methods resulted in preparations with higher tumor uptake.

A comparison of monoclonal antibodies against distinct colon tumor-associated antigens in immunohistochemistry and in tumor localization.

Monoclonal antibodies to colon/ovary tumor antigen (COTA), carcinoembryonic antigen (CEA), colon-specific antigen (CSA), and colon-specific antigen "protein" (CSAp) were evaluated for specificity, reactivity with normal tissues, and tumor localizations using athymic rats bearing xenografted human colon tumors. Radioiodine labeled anti-CSA and anti-COTA retained immunoreactivity and effectively localized the tumors; anti-CSAp retained immunoreactivity, but localized less effectively; and anti-CEA lost most of its immunoreactivity and localized poorly. Of the antibodies tested, anti-COTA showed potential for human colorectal tumor radiolocalization.

COTA (colon-ovarian tumor antigen). An immunohistochemical study.

A goat anti-serum was prepared against mucinous ovarian cyst fluid and absorbed with normal colon and a variety of normal tissues until the only residual immunoreactivity was directed against colon cancer and ovarian tumor mucin. The set of antigenic determinants defined by this anti-serum has been called COTA, standing for colon-ovarian-tumor-antigen. This highly absorbed anti-serum (anti-COTA) was used for immunohistochemical staining of 42 different tissues in parallel with staining with a goat anti-CEA, which was also highly absorbed. The results suggest that COTA is a highly sensitive and specific antigen for colon carcinoma and may have potential for the early detection of malignant changes predictive of cancer of the colon.

Production of monoclonal antibody SP-21 to colon-ovarian tumor antigen, COTA.

An IgG monoclonal antibody, SP-21, directed against colon-ovarian tumor antigen, COTA, is reported. The antibody had no reactivity with CEA, normal colonic mucin, CSAp, ABO blood group antigens, or with normal human lung, liver, spleen, kidney, plasma and saliva in studies using the enzyme-linked immunoassay method (ELISA). Immunoperoxidase staining of colon, lung, kidney, and prostate cancer tissues and benign and inflammatory colon disease tissues revealed a specificity identical to that of the polyclonal (goat) anti-COTA antibodies.

Characterization of a common antigen of colorectal and mucinous ovarian tumors, COTA.

A new colon cancer antigen is reported. It is designated as COTA, Colon-Ovarian Tumor Antigen, because it is found in mucins produced by both tissues during malignancy. The new antigen was identified by making antibodies against human colon cancer tissue in goats. The antisera were exhaustively absorbed with lyophilized extracts of normal colon, lung, liver, spleen, kidney, plasma, and the well-known colon tumor antigen, carcinoembryonic antigen (CEA). The new antigen was identified by immunodiffusion. Studies of 28 malignant tissue extracts, 10 ovarian adenocarcinoma cyst fluids, 43 normal tissues, and 5 plasma samples revealed that this antigen is found only in colon tumors and mucinous ovarian adenocarcinomas. The antigen was not detected in serous adenocarcinoma of the ovaries, extracts of adenocarcinoma of lung, breast, kidney or stomach nor in the extracts of normal tissues. Other tests show that this antigen is not CEA, Ca 19-9, or CSAp. It is stable to heating at 65 degrees for 5 minutes; it elutes from an ion exchange matrix (DEAE) with 0.3-0.5M NaCl; it migrates to the alpha-2 region on immunoelectrophoresis; and its size, by exclusion chromatography on Sepharose 4B, is 3-15 million daltons. Anti-COTA stains colon cancer tissue sections indicating that COTA is present in goblet-cell mucin.

Colon-specific antigen-p (CSAp). II. Further characterization in colorectal and pancreatic cancer.

The physicochemical and immunological characteristics of colon-specific antigen-p (CSAp) in plasma and in colorectal and pancreatic tumors were investigated. CSAp in the plasma of a rectal cancer patient and in a colonic carcinoma xenografted in hamsters (GW-39 tumor) appeared to have similar chromatographic properties, being of a molecular size of 4 million or more. The activities of CSAp in both plasma and tumor were similarly destroyed by treatment with a thiol reagent. Finally, identical immunological reactions in radioimmunoassay and gel diffusion tests were obtained between the CSAp's in patient plasma and in GW-39 tumor tissue. Also CSAp in human pancreatic cancers xenografted in nude mice showed a precipitin line of complete identity with CSAp extracted from GW-39 human colonic carcinoma transplants. Thus, the CSAp's found in colorectal cancer patient plasma, in colonic carcinoma, and in pancreatic cancer tissues appear to be immunologically identical.

Colon-specific antigen-p (CSAp). I. Initial clinical evaluation as a marker for colorectal cancer.

A radiometric immunoassay for detecting colon-specific antigen-p (CSAp) in the blood of patients suspected of having colorectal cancer has been developed and evaluated in 272 subjects of various disease entities. Using 10 units/ml as the cutoff value for normalcy, the results indicate that the highest number of elevated CSAp titers occurred in patients with advanced colorectal cancer (61%). Only one of 12 colonic adenoma patients had an elevated CSAp titer, and this was slightly above the 10 units/ml cutoff. Other nonneoplastic gastrointestinal disorders showed an 18% abnormal CSAp titer frequency, of which more than half bordered the upper limit of normalcy. CSAp elevations were also found in gastric cancer (20%), pancreatic cancer (20%), breast cancer (5%), and normal individuals (3%). CSAp was compared to carcinoembryonic antigen (CEA) in 44 colorectal cancer patients, in 12 patients with colonic adenomas, and in 62 patients with diverse gastrointestinal disorders. Using a CEA cutoff of 5 ng/ml, CSAp could increase the diagnostic accuracy of the CEA plasma test in colorectal cancer patients by 14%. In patients with colonic adenomas, the CSAp titer was normal when the CEA value was elevated in 25% of the cases. However, both were simultaneously elevated in only 3% of the patients with benign gastrointestinal disorders. Since the CSAp test was less frequently elevated (0-7%) in patients with breast, ovarian, lung, or cervical cancer than was found for CEA (39-75% elevated), it seems that the CSAp blood assay detects colorectal cancer more specifically than the more generally distributed CEA. These results suggest that the combined use of blood CEA and CSAp could enhance the discrimination of colorectal cancers from other nonmalignant gastrointestinal diseases, and could also serve to enhance the colorectal cancer-specificity of the CEA assay. Furthermore, serial monitoring of both markers in four advanced colorectal cancer patients indicated that they reflect disease activity, falling after successful treatment and rising again with recurrence and disease progression.

Characterization of colon-specific antigen-p and isolation of immunologically active tryptic peptides.

CSAp was originally detected with antisera to an unfractionated extract of GW-39 human colonic carcinoma xenografts, and was determined to be restricted to normal and neoplastic gastrointestinal tissues and certain ovarian tumors. Gel filtration chromatography of GW-39 extract on Sepharose 4B columns reveals that more than 90% of the CSAp is associated with the void volume fraction. The smaller m.w. CSAp fraction is a population of fragments heterogeneous in respect to size but immunologically identical to the void CSAp. Efforts to dissociate intact CSAp from the void fraction by treatment with solutions of high ionic strength, SDS, non-ionic detergents, and 1 M lithium bromide have not been successful. CSAp is a glycoprotein that binds to Sepharose-concanavalin A and is distinct from other known tumor-associated antigens in immunologic and physicochemical properties. The antigen shows sensitivity to high concentrations of chaotropic reagents and especially to sulfhydryl reagents, even in low concentrations, which supports earlier results indicating that the CSAp antigenic determinant is associated with the polypeptide chain rather than with a carbohydrate moiety. CSAp has been reduced in size by either sonication or partial tryptic digestion. The former produces immunologically active fragments, heterogeneous in respect to size, that resemble the smaller size CSAp, whereas the tryptic digest generates 2 distinct peptides. The major tryptic peptide is found to have an approximate molecular size of 120,000, as determined by gel filtration.

Experimental studies of tumor radioimmunodetection using antibody mixtures against carcinoembryonic antigen (CEA) and colon-specific antigen-p (CSAp).

Experiments with the GW-39 human colonic carcinoma growing in hamsters showed that injection of radioactive antibody to a colorectal-specific, tumor-associated antigen, CSAp, results in better tumor radiolocalization than was seen previously with radioantibodies to carcinoembryonic antigen (CEA). However, a mixture of both radioactive antibodies resulted in potentiation of CEA-tumor radioimmunodetection without affecting CSAp-tumor radiolocalization. Hence, multi-marker antibody mixtures may be the method of choice in cancer radioimmunodetection.

Further characterization of CSAp, an antigen associated with gastrointestinal and ovarian tumors.

Colon-specific antigen-p, or CSAp, was originally extracted from GW-39 tumors, which are human colonic carcinomas serially transplanted in golden hamsters, and antibodies to CSAp have been produced in the same animal hosts. By means of immunodiffusion and a hemagglutination-inhibition assay, CSAp has been found to be restricted to adult and fetal small intestine, neoplastic gastric and colonic tissues, inflamed colon, and cystic mucinous tumors of the ovary. CSAp was shown to be distinct from blood group antigens, including Lea and Leb blood group substances, liver ferritin, AFP, CEA, CSA, CMA, ZGM, and BOFA, and to have the electrophoretic mobility of an alpha2-globulin. Gel filtration studies indicated that CSAp in GW-39 tumor, primary human colonic carcinoma, and ovarian cancer mucinous cyst fluid had a peak molecular size range of 70,000--110,000. Quantitation of CSAp in 214 tissue specimens by the hemagglutination-inhibition assay revealed a progressive increase in fetal, inflamed, and neoplastic intestine, such that CSAp in colonic tumors was increased over normal colon tissue. Thus, CSAp appears to be an organ-specific antigen showing increased levels in some gastrointestinal and ovarian neoplasms, as well as in specimens with colitis.

Identification of beta-oncofetal antigen in cervical squamous cancer and its demonstration in neoplastic and normal tissues.

An extract of a human cervical squamous carcinoma was used to produce rabbit antiserum with immunoreactivity against an antigen in several types of normal and neoplastic tissues. This antigen was abundant in cervical cancer as well as in normal adult and fetal kidney and liver. The antigen had a beta mobility in immunoelectrophoresis and a molecular weight range of 74, 000 to 90,000 as determined by gel chromatography. Since some of its properties were similar to those of the beta-oncofetal antigen described by Fritsche and Mach, a comparison was undertaken that indeed revealed identical immunoreactivity of the anti-beta-oncofetal antigen and anti-cervical cancer antisera when reacted in immunodiffusion against a cervical cancer extract. These results do not support the designation of this antigen as an oncofetal antigen.

A putatively new antigen (CSAp) associated with gastrointestinal and ovarian neoplasia.

CSAp is an antigen originally identified in the GW-39 human colonic carcinoma xenograft, and also found in gastric and colonic cancers, fetal colon, normal and inflammatory adult colon, and in some ovarian tumors. However, it appears to be increased primarily in inflammatory, benign , malignant, and fetal human intestine, gastric cancer, and ovarian tumors, as determined by an hemagglutination-inhibition assay. Gel immunodiffusion patterns show that CSAp is immunologically distinct from CEA, NCA, AFP, BOFA, and human liver ferritin. CSAp thus appears to be a putatively new fetal substance with a high degree of specificity for gastric, colonic, and ovarian tissues.

Antigens associated with normal and malignant gastrointestinal tissues.

Immunization of hamsters with phenol-alcohol extracts of GW-39 human colonic tumor tissues has resulted in the identification of three gastrointestinal tissue-associated antigens, on the basis of precipitin immunoreactivity. Sephadex G-200 and Bio-Gel A-15m chromatography of normal colonic tissue and GW-39 tumor extracts revealed antigen immunoreactivity in the 46,000 (low-molecular-weight), 170,000 to 900,000 (high-molecular-weight), and 5 to 10 million (very high-molecular-weight) ranges or low-molecular-weight colon-specific antigen (LMW/CSA), high-molecular-weight colon-specific antigen (HMW/CSA), and very-high-molecular-weight colon-specific antigen (VHMW/CSA), respectively. Immunodifussion reactions indicated that the HMW/CSA was human gastrointestinal tissue-specific, increasing in concentration from the esophagus to the colon [for which reason the term colon-specific antigen (CSA) has been retained], whereas the LMW/CSA was found in human gastrointestinal tissues, hamster and rat colon, human saliva, and normal human cervix. Colon-specific antigen (CSA) could be demonstrated in human gastrointestinal tumors, including the LS-174T colonic cancer cell line, but not in cancers of other sites tested. Likewise, CSA's were found in fetal human gut tissue. Whereas HMW/CSA and VHMW/CSA showed partial identity in immunodiffusion, HMW/CSA and VHMW/CSA, as well as LMW/CSA and VHMW/CSA, showed distinct immunoprecipitin bands, respectively. The immunoelectrophoretic mobility of VHMW/CSA was similar to an alpha-globulin, whereas HMW/CSA and LMW/CSA migrated to the prealbumin region. CSA appeared in immunofluorescence of GW-39 tumor cells and in the goblet cells of human colon predominantly as a cell-surface component. Staining with periodic acid-Schiff and solubility characteristics of the CSA's suggest that they are glycoprotein in nature. These studies thus support the view that organ-specific and organ-associated antigens of the colon can be maintained and expressed in human colonic carcinomas, including the xenografted GW-39 human colonic tumor system.