Antibody targeting in metastatic colon cancer: a phase I study of monoclonal antibody F19 against a cell-surface protein of reactive tumor stromal fibroblasts.

PURPOSE: To define the toxicity, imaging, and biodistribution characteristics of iodine 131-labeled monoclonal antibody F19 (131I-mAbF19). MAbF19 recognizes the fibroblast activation protein (FAP), a cell-surface glycoprotein not present in most normal tissues, but abundantly expressed by reactive stromal fibroblasts of epithelial cancers, including more than 95% of primary and metastatic colorectal carcinomas. PATIENTS AND METHODS: 131I-mAbF19 was administered intravenously to 17 patients with hepatic metastases from colorectal carcinoma who were scheduled for resection of localized metastases or insertion of hepatic artery catheter for regional chemotherapy. Seven to 8 days before surgery, patients received 131I-mAbF19 at three dose levels, with at least four patients entered at each level. RESULTS: No toxicity associated with intravenous 131I-mAbF19 administration was observed. Tumor images were obtained on planar and single-photon emission tomography (SPECT) scans in 15 of 17 patients with hepatic metastases, tumor-infiltrated portal lymph nodes, and/or recurrent pelvic disease. The smallest lesion visualized was 1 cm in diameter. The optimal time for tumor imaging was 3 to 5 days after 131I-mAbF19 administration. The use of image registration techniques allowed precise anatomic localization of 131I-mAbF19 accumulation. Immunohistochemical analysis of biopsy tissues showed expression of FAP in the tumor stroma (but not in normal liver) in all patients studied and confirmed that the FAP-positive tumor stromal fibroblasts were interposed between the tumor capillaries and the malignant colon epithelial cells. At the time of surgery, tumor-to-liver ratios up to 21:1 and tumor-to-serum ratios up to 9:1 were obtained. The fraction of the injected 131I-mAbF19 dose per gram tumor (%ID/g tumor) localized to hepatic metastases at the time of surgery ranged from 0.001% to 0.016%. CONCLUSION: The FAP tumor fibroblast antigen is highly expressed in primary and metastatic colorectal carcinomas and shows limited expression in normal adult tissues. This highly selective expression pattern allows imaging of colorectal carcinoma lesions as small as 1 cm in diameter on 131I-mAbF19 scans. Because of the consistent presence of FAP in the stroma of epithelial cancers and the accessibility of FAP-positive tumor stromal fibroblasts to circulating monoclonal antibodies (mAbs), this study suggests possible diagnostic and therapeutic applications of humanized mAbF19 and mAbF19 constructs with novel immune and nonimmune effector functions.

Antibody localization in human renal cell carcinoma: a phase I study of monoclonal antibody G250.

PURPOSE: To define the imaging and biodistribution characteristics of iodine 131-labeled monoclonal antibody (mAb) G250 (131I-mAbG250), which recognizes a cell-surface antigen expressed by human renal cell carcinoma (RCC). PATIENTS AND METHODS: G250 is a cell-surface antigen recognized by mAbG250 expressed by RCC but not detected in normal kidney. Clear-cell RCC, the most frequent form of RCC, shows homogeneous expression of G250, whereas non-clear-cell RCC and cancers derived from other organs generally do not express G250. Expression in normal tissues is highly restricted and limited to large bile ducts and gastric epithelium. 131I-mAbG250 was administered intravenously (IV) to 16 patients with RCC 7 to 8 days before surgery at five dose levels, with at least three patients entered at each dose level. RESULTS: Clear tumor images were observed in 12 patients with G250-positive tumors and in one of three patients with G250-negative tumors. Imaged lesions in the peritoneal cavity were confirmed at surgery. The smallest lesion visualized was 8 mm in diameter. The specificity of 131I-mAbG250 localization to tumor tissue was established by radioactivity measurements, autoradiography, and immunohistochemistry of biopsied tissues, and technetium 99-human serum albumin blood-flow studies. The fraction of the injected 131I-mAbG250 dose per gram tumor (%ID/g tumor) localized in G250-positive tumors showed a broad range, but reached levels as high as 0.02% to 0.12%. CONCLUSION: 131I-mAbG250 localized specifically to G250 antigen-positive RCC and seems to have considerable potential as an imaging agent in RCC patients. 131I-mAbG250 uptake in the tumors, relative as well as absolute, are among the highest reported for tumor biopsies obtained 8 days after IV mAb administration. Based on the specific localization and high accumulation, mAb G250 may have therapeutic potential.

Immune and nonimmune effector functions of IgG3 mouse monoclonal antibody R24 detecting the disialoganglioside GD3 on the surface of melanoma cells.

R24, an IgG3 mouse monoclonal antibody reactive with the disialoganglioside GD3, was found to be a potent mediator of human complement cytotoxicity and human effector cell cytotoxicity. Cytotoxicity correlated with the degree of antibody binding (GD3 cell surface expression) for each of the melanoma cell lines and melanocyte cell cultures tested. Melanoma cell lines binding low amounts of R24 (low GD3 cell surface expressors) were not lysed in R24-directed immune reactions, suggesting that a threshold number of R24 molecules bound per cell is necessary to initiate these cytotoxic mechanisms. Since both complement- and cell-mediated reactions lysed the same subpopulations of cells in each cell line, both mechanisms appeared to depend on similar threshold quantities of bound R24 molecules. However, due to the heterogeneity of R24 binding in each cell line, the numerical value for this threshold could not be determined. Only in cell lines binding greater than 10(7) R24 molecules per cell were greater than 90% of the cells lysed. Normal melanocytes in culture were not lysed by R24-directed immune mechanisms, due to their low GD3 expression, indicating that monoclonal antibodies such as R24 may show tumor specificity with regard to effector functions even though normal cells express the relevant antigen. In contrast to the potent in vitro activity of R24, treatment of nu/nu mice bearing human melanoma grafts resulted in tumor inhibition only when started within 3 days of tumor cell inoculation. No effect was seen on established tumors. Thus, this in vivo mouse model failed to predict the clinical and pathological findings observed in treatment trials of R24 in human melanoma patients--urticaria involving skin metastases, cellular infiltration of tumor tissue, and tumor regression. In addition to activating immunologic effector functions, R24 had direct effects on melanoma cells, blocking their ability to attach to surfaces and causing tumor cell aggregation. These effects were again related to the number of R24 molecules bound to the cell surface; no aggregation was seen with cell lines binding less than 4 X 10(5) molecules per cell. Both immune and nonimmune effector functions may be involved in the tumor inhibitory activity of R24 in humans.

Nonhematopoietic cells selected for resistance to tumor necrosis factor produce tumor necrosis factor.

TNF-resistant lines of L cells can be derived from TNF-sensitive populations by repeated exposure to TNF, and these resistant L cells, in contrast to sensitive L cells and other types of cells, lack demonstrable cell surface receptors for TNF. We have now found that TNF-resistant L cells produce a factor that is cytotoxic for L cells and has the following distinguishing characteristics of mouse TNF: it is a protein of 43 kD, composed of 16 kD subunits, that competes with TNF for receptor binding, induces hemorrhagic necrosis of the TNF-sensitive mouse sarcoma Meth A, has synergistic cytotoxic action with interferon, and its activity is neutralized by antibody to TNF. The two conclusions of this study are that cells selected for TNF resistance spontaneously produce a molecule resembling macrophage TNF, and that cells of nonhematopoietic origin are capable of producing TNF.

Purification, characterization, and antitumor activity of nonrecombinant mouse tumor necrosis factor.

Mouse tumor necrosis factor (TNF) was purified from serum through a series of steps, and each step was monitored for L-cell cytotoxicity in vitro and tumor-necrotizing activity in vivo. The two activities copurified and could not be dissociated. Purified mouse TNF has a specific activity of 2.2 X 10(7) (L-cell assay in the absence of actinomycin D) and 1 microgram causes necrosis of the standard TNF-sensitive sarcoma Meth A. TNF has a Mr of 39,000 +/- 2000 by gel filtration and a Mr of 16,000-18,000 by NaDodSO4/PAGE. Both molecular weight forms display cytotoxic and necrotizing activities. TNF has a pI of 3.9 and is destroyed by trypsin, protease, elastase, and alpha-chymotrypsin but not by neuraminidase or papain. These characteristics of nonrecombinant mouse TNF clearly resemble those of recombinant human and mouse TNF.

Purification and characterization of a human tumor necrosis factor from the LuKII cell line.

A factor with tumor necrosis factor (TNF) activity produced by the LuKII human lymphoblastoid cell line [designated TNF(LuKII)] was purified sequentially by using controlled-pore glass, lentil lectin-Sepharose, and procion red agarose chromatography, yielding TNF with a specific activity of 1.5 X 10(7) units per mg of protein and an isoelectric point of approximately equal to 6.7. Purified TNF(LuKII) fractionated by NaDodSO4/PAGE under reducing as well as nonreducing conditions was found to contain seven protein bands of Mr 80,000, 70,000, 43,000, 25,000, 23,000, 21,000, and 19,000. The proteins of Mr 80,000 and 70,000 could not be dissociated into lower molecular weight components. Peptide mapping analysis and immunoblotting analysis revealed that the seven protein bands in the purified TNF(LuKII) preparations are related. After fractionation of TNF(LuKII) by NaDodSO4/PAGE under reducing conditions, TNF activity was recovered from the regions of Mr 70,000 and 19,000-25,000. Purified human TNF(LuKII) (i) produces hemorrhagic necrosis of Meth A mouse sarcoma in the standard in vivo mouse TNF assay; (ii) has the same pattern of reactivity as mouse TNF (cytotoxic/cytostatic/no effect) on a panel of human cancer cell lines; and (iii) has its anticellular effect potentiated by interferon, also a feature of mouse TNF.

High affinity binding of 125I-labeled human tumor necrosis factor (LuKII) to specific cell surface receptors.

125I-labeled TNF(LuKII) (tumor necrosis factor) binds specifically to human and mouse cell lines sensitive to the cytotoxic effect of TNF, but not to cells made resistant to TNF. TNF-sensitive cells have cell surface receptors with a high affinity for TNF(LuKII). Mouse TNF competes with TNF(LuKII) for receptor binding. Scatchard analysis of the binding data yielded linear plots and suggests that TNF(LuKII) binds to homogeneous receptor sites. The number of TNF(LuKII) receptors on two TNF-sensitive cell lines is 200-300 per cell and the affinity constant of the receptor for TNF(LuKII) is approximately 1 X 10(-10) M.

Cell cycle-specific effects of tumor necrosis factor.

The immediate effect of tumor necrosis factor (TNF) added to cultures of L-cells is cytostasis, manifested as cell arrest in G2. This effect prevails during the initial 4 hr when the number of G2 cells increases markedly in the absence of any significant cell death in the culture. Shortly thereafter, the cytolytic effect becomes apparent; extensive cell lysis can be detected after 7 hr of exposure to TNF. After 24 hr nearly all cells are lysed. Results of experiments in which the effects of TNF on the kinetics of cell progression through various phases of the cell cycle were studied indicate that in the presence of TNF cells do progress through G2, although with considerable delay, and reach mitosis. Most cells die (undergo lysis) specifically at late stages of mitosis (telophase) or soon after cytokinesis. Sensitivity of cells to the lytic effect of TNF is increased by cell arrest in mitosis with Colcemid or vinblastine. TNF exerts little effect on rates of cell progression through G1 or S phases of the cell cycle, although few cells die during S phase. Nonlysed cells from cultures treated with TNF do not show any changes in nuclear chromatin, suggesting that neither DNA nor nuclear proteins are the primary targets of the TNF. The present data implicate metabolic changes which occur during mitosis (cytokinesis), perhaps associated with synthesis or assembly of cell membrane components, as being responsible for increased cell sensitivity towards the cytolytic effect of TNF. The mechanism of the early cytostatic effect of the factor is unknown.

Human tumor necrosis factor produced by human B-cell lines: synergistic cytotoxic interaction with human interferon.

Human cell lines of hematopoietic origin were tested for production of tumor necrosis factor (TNF). B-cell lines transformed by Epstein-Barr virus release a factor (referred to as hTNF) that is cytotoxic for mouse L cells sensitive to mouse TNF but not for L cells resistant to mouse TNF. Exposure to 4 beta-phorbol 12 beta-myristate 13 alpha-acetate augmented production of hTNF. hTNF activity was not found in supernatants of cell lines of T-cell, monocytic, or promyelocytic origin. Partially purified hTNF has a molecular weight of approximately 70,000, has no interferon activity, is acid labile, is destroyed by heating at 70 degrees C for 1 hr, induces cross-resistance to mouse TNF in vitro, and causes hemorrhagic necrosis of Meth A mouse sarcoma in the standard in vivo mouse TNF assay. Tests with a panel of 23 human cancer cell lines showed that hTNF is cytotoxic for 7 cell lines, cytostatic for 5, and has no effect on 11. Comparative studies with human alpha, beta, and gamma interferons indicated that sensitivity to hTNF and interferon can be distinguished. Combined treatment with hTNF and alpha or gamma interferon resulted in a synergistic cytotoxic effect.

Possible importance of macrophage-derived mediators in acute malaria.

Tumor necrosis factor, lymphocyte-activating factor, and enhanced levels of type I interferon were found in serum samples taken 2 h after mice infected with Plasmodium vinckei subsp. petteri received a small intravenous injection of endotoxin. These three mediators are among those released when mice receive an endotoxin injection 2 weeks after Mycobacterium bovis BCG or Corynebacterium parvum have been administered. There is indirect evidence that this wider range of mediators is also released in P. vinckei subsp. petteri-infected mice given parenteral endotoxin. A recent report that endotoxin is detectable in the plasma of malaria-infected mice and children implies that these mediators may also be released in the acute phase of the natural infection. We propose that these macrophage-derived mediators may be important in the glucocorticoid antagonism, bone marrow depression, fever, hypergammaglobulinemia, splenomegaly, elevation of serum amyloid A, consumptive coagulopathy, and shock syndrome with associated organ damage which can accompany malaria. The intraerythrocytic parasite death seen at crisis in some malarias, as well as the subsequent development of specific protective immunity, may also depend on these mediators.

Endotoxin-induced tumor necrosis factor.

The serum of BCG-infected mice treated with endotoxin contains a substance (tumor necrosis factor, TNF) which mimics the tumor-necrotizing action of endotoxin itself. TNF is not residual endotoxin, but a factor released from host cells, probably macrophages. TNF induced in the same way in rats and rabbits also causes necrosis of transplanted murine tumors. Unlike endotoxin, TNF is toxic in vitro for neoplastic murine and human cell lines but not for mouse embryo culture. TNF has striking effects on immunologic reactions in vitro, some like those of endotoxin and others unlike those of endotoxin. TNF is a glycoprotein; its molecular weight is less than 70,000. Highly purified preparations do not contain lysosomal or nonlysosomal serum enzymes, interferon or prostaglandin E1.

Partial purification of a serum factor that causes necrosis of tumors.

Tumor necrosis can be induced in transplanted mouse methylcholanthrene-induced sarcoma by a tumor necrosis factor in the serum of mice infected with bacillus Calmette-Guérin and given bacterial endotoxin. Sera from normal mice, endotoxin-treated mice, and mice infected with bacillus Calmette-Guérin do not contain this factor. A 20- to 30-fold purification of the serum factor has been achieved by (NH4)2SO4 fractionation, Sephadex G-100 and G-200 gel filtration, and preparative polyacrylamide electrophoresis. Tumor necrosis factor is not bacterial endotoxin. It migrates with alpha-globulins, is made up of at least four subunits, and has a molecular weight of about 150,000. The active factor is a glycoprotein that contains sialic acid and galactosamine.

An endotoxin-induced serum factor that causes necrosis of tumors.

In studying "hemorrhagic necrosis" of tumors produced by endotoxin, it was found that the serum of bacillus Calmette--Guerin (BCG)-infected mice treated with endotoxin contains a substance (tumor necrosis factor; TNF) which mimics the tumor necrotic action of endotoxin itself. TNF-positive serum is as effective as endotoxin itself in causing necrosis of the sarcoma Meth A and other transplanted tumors. A variety of tests indicate that TNF is not residual endotoxin, but a factor released from host cells, probably macrophages, by endotoxin. Corynebacteria and Zymosan, which like BCG induce hyperplasia of the reticulo-endothelial system, can substitute for BCG in priming mice for release of TNF by endotoxin. TNF is toxic in vitro for two neoplastic cell lines; it is not toxic for mouse embryo cultures. We propose that TNF mediates endotoxin-induced tumor necrosis, and that it may be responsible for the suppression of transformed cells by activated macrophages.

Properties of reticulum-cell sarcomas in SJL/J mice. I. Proliferative response to tumor cells of T-derived lymphoid cells from normal mice.

Cells from reticulum-cell sarcomas (RCS), tumors with a probable B-cell origin in the SJL/J mouse strain, induced a high degree of proliferation in O-bearing syngeneic lymph-node and thymus cells obtained from young non-tumours mice. Although of considerably lower magnitude, a proliferation of SJL/J thymus cells to syngeneic normal lymph-node cells was also noted. Sera from normal syngeneic mice did not block either of these proliferative responses. Stimulation by RCS cells also occurred with normal lymph-node cells from F1 hybrids, SJL/J times C57Bl/6, and SJL/J times ASW and from the H-2s identical ASW mouse strain. It is suggested that these results are due either to the presence of viral antigens on the surface of RCS cells or to an exaggerated form of the normal syngeneic response of T to B lymphocytes. Another tumor of B-cell origin, the PU5 tumor of BALB/c mice, failed to induce proliferation in normal syngeneic lymph-node or thymus cells.

Serum-mediated leukemia cell destruction in AKR mice.

AKR mice with spontaneous leukemia were infused with normal serum from a variety of species. Leukemia cell destruction was produced by serum from strains of mice possessing the full spectrum of complement components, but not by serum from strains with a genetically determined deficiency of C5. Serum from guinea pigs, horses, and humans also causes destruction of leukemia cells. The antileukemic factor in normal serum was heat labile (56 degrees C for 35 min) and could be inactivated by cobra venom factor (CVF). Tests of individual complement factors from guinea pig serum and from human serum suggest that C5 is the antileukemic complement component in normal serum. Evidence was obtained that complement also plays a role in the antileukemic effect of interferon and endotoxin.