Influence of Cocktails of Labeled Monoclonal Antibodies on the Localization of Antibodies in Human Tumor Xenografts

In order to evaluate the usefulness of cocktails of labeled monoclonal antibodies (MoAbs) recognizing different antigen molecules to localize human cancer xenografts, we have compared the potential of three MoAbs recognizing representative cancer-associated CA 19–9, 17–1A and CEA antigens when administered alone or in combination. Specific binding of radioiodinated F(ab')₂ fragments of these three MoAbs was observed to human colorectal cancer cell lines SW1116, LS180 and Co-3. The percentage of in vitro cell binding of a cocktail of any two MoAbs to cancer cells was equal to the average of those obtained with the two MoAbs alone. The three MoAbs were preferentially localized in tumor tissues xenografted in nude mice. When cocktails of any two MoAbs were used, the obtained tumor-to-normal tissue ratios and percent of injected dose per gram of tumor were between the levels obtained for each MoAb when administered alone, in all three tumors transplanted in nude mice. These data suggest that, although cocktails of labeled MoAbs recognizing different antigens may extend the spectrum of tumor specificities, their use does not improve the tumor localization ability of MoAb-conjugates.     

Effect of tumor mass and antigenic nature on the biodistribution of labeled monoclonal antibodies in mice

The effect of tumor mass and antigenic nature on the biodistribution of 111In- and 125I-labeled monoclonal antibodies (MoAbs) was studied using F(ab')2 fragments of three representative anti-tumor MoAbs and SW1116 human colorectal carcinoma grown in nude mice. The 19-9, F33-104 anti-CEA, and 17-1A MoAbs showed specific binding to SW1116 cells. The former two MoAbs recognize circulating CA 19-9 with molecular weights of more than 5,000,000 and CEA of Mr 170,000-180,000, respectively, whereas 17-1A reacts with a nonshedding antigen. Both percentage injected dose per gram tumor and tumor-to-blood ratios were inversely proportional to the tumor mass in nude mice administered 111In- and 125I-labeled 19-9, but liver uptake increased as tumor size increased. Analysis of serum samples and tumor homogenates demonstrated the presence of a high-molecular-weight species, probably due to the antibody binding to CA 19-9. In the case of 111In-labeled anti-CEA MoAb, tumor uptake also decreased and liver uptake increased with tumor size, but this effect was less obvious than that of 19-9. In contrast, tumor and liver uptake of 125I-labeled anti-CEA MoAb, 111In- and 125I-labeled 17-1A and control antibodies were independent of tumor mass. The absolute tumor uptake and tumor-to-blood ratios of all 125I-labeled antibodies were lower than those of the 111In-labeled ones. And the effect of tumor mass was also weaker with 125I-labeled antibodies, probably due to in vivo dehalogenation. These results indicate that the effect of tumor size on the incorporation of labeled MoAb into tumors is dependent on the antigenic nature to be targeted and/or radionuclides used for labeling and that high concentrations of circulating high molecular weight antigens may limit in vivo use of MoAb conjugates.     

Localization and imaging of radiolabeled monoclonal antibodies against colorectal carcinoma in tumor-bearing nude mice

Four monoclonal antibodies (MoAbs) (35, 115, 17-1A, and B72.3) directed towards human carcinoma surface antigens have been studied in athymic nude mice with LS174T, CO112, or SW948 colon carcinoma xenografts or negative control melanoma (MEL-1), lymphoma (Namalwa), and breast (MCF-7) carcinoma xenografts to evaluate the effects of antigenic heterogeneity and time after administration on localization and imaging. 125I-labeled 115 showed the highest uptake of any antibody in LS174T tumors. MoAbs 35 and B72.3 showed similar but lower levels of uptake in LS174T and CO112 tumors, but B72.3 concentrated less in SW948 tumors. 17-1A showed the highest degree of accumulation in SW948 tumor xenografts. No specific uptake of the four anti-carcinoma MoAbs was observed in MEL-1, Namalwa, or MCF-7 xenografts. The specificity of the in vivo tumor localization of the four anti-carcinoma MoAbs was confirmed by the low degree of accumulation of a control MoAb against influenza virus in LS174T tumors. Imaging studies with 131I-labeled colorectal cancer MoAbs showed specific uptake and retention in LS174T tumors, with progressive clearance from the whole body. The colorectal cancer MoAbs were compared for immunohistochemical binding against biopsies from patients with colorectal cancer and adjacent normal colonic tissue. Most colorectal cancer specimens showed moderate to strong staining with the four MoAbs. The percentage of positive cells varied within and between tumors demonstrating antigenic heterogeneity. Absent to slight focal staining was seen with normal colon tissue. B72.3 showed the highest degree of staining specificity. This study indicates a difference in the immunohistochemical binding of a panel of MoAbs against biopsies of colon adenocarcinoma and a dependence of in vivo localization on the human colon cancer cell line used as target. This has important implications for future clinical diagnostic and therapeutic studies.     

Experimental therapy of human breast tumors with 131I-labeled monoclonal antibodies prepared against the human milk fat globule

Breast tumors are susceptible to attack by unconjugated anti-human milk fat globule monoclonal antibodies (MoAbs) and most particularly by their mixture (cocktail) (Cancer Res., 47: 532-540, 1987). In the present study the same MoAbs (Mc1, Mc3, Mc5, and Mc8) labeled with 131I, either singly or in cocktails, were used for a similar purpose. Biodistribution studies showed that a transplantable human breast tumor line (MX-1) implanted in BALB/c nude mice (nu/nu) had the maximum incorporation of injected 131I-MoAbs at day 4 while levels in circulation and in normal tissue declined steadily from day 1. Also, these studies showed that the amount of radiolabeled Mc3 MoAb incorporated by MX-1 tumors was greater than that for cocktail of MoAbs and MoAb Mc5. Tumor destruction by injected 131I-MoAb cocktail was shown in therapy experiments to be dose dependent. A single injection (1500 [corrected] microCi/mouse) of 131I-MoAb Mc3, or of cocktail, produced large breast tumor volume diminution and inhibition of growth for up to 30 days while a similar dose of 131I-labeled control IgG had no effect. A second dose of 1500 [corrected] microCi 131I-MoAb of Mc3 or of cocktail, injected at an appropriate interval, again diminished tumor mass significantly and inhibited its growth for another 20 days. In control experiments, non-breast tumors (colon) were marginally affected by the 131I-MoAbs. These results show that the systemic injection of radioiodinated MoAbs against human milk fat globule destroy the epithelial cells of human breast tumors and control their growth for an appreciable length of time. Radioiodoconjugated MoAbs proved to be more effective than unconjugated MoAbs in reducing breast tumor mass and also in inhibiting growth for longer periods of time at immunoglobulin doses 100 to 200 times lower. Further exploration of their role in breast cancer treatment seems warranted by these results.     

Monoclonal antibodies against human gastric cancer

Spleen cells from Balb/c mice immunized with five human gastric cancer cell lines in sequence were fused with murine myeloma cell line SP2/0, and hybridomas 3F4, 3G9 and 3H11, secreting monoclonal antibodiees (MoAbs) against gastric cancer, were obtained through selective culture and screening. These MoAbs have both good selectivity and a high positive rate of reaction for gastric cancer, reaching 5/5 and 84.8% to 93.5% for gastric cancer cells and tissues respectively. The reaction of MoAbs with normal cells and tissues was neglible. The corresponding antigens of the MoAbs were sensitive to digestion by trypsin and pronase and resistant to treatment with sodium periodate, indicating their nature as proteins. The antigen 3G9 could be visualized with Western blotting as two bands with molecular weights of 100KD and 70KD, however no band was found for antigens 3F4 and 3H11. There was a high expression of antigens in the majority of gastric cancer cells and tissues independent of histopathological type of gastric cancer. A low expression of antigens was seen with other tumors and fetal gastrointestinal tissues. These could be considered as gastric cancer-associated antigens or oncofetal antigens with a quite extensive distribution.     

Cocktails of ricin a-chain immunotoxins against different antigens on hodgkin and sternberg-reed cells have superior anti-tumor effects against H-RS cells in vitro and solid Hodgkin tumors in mice

Three ricin A-chain immunotoxins (ITs) recognizing different antigens on Hodgkin-Reed/Sternberg (H-RS) cells were evaluated for their anti-tumor effects when used in combination as “cocktails”. These ITs, BB10.dgA (CD2S), HRS3.dgA (CD30), and IRac.dgA (70 kDa), strongly inhibited the growth of L540Cy H-RS cells in vitro. The protein synthesis of this cell line was reduced more efficiently by the combination of 2 of these ITs than by BB10.dgA, HRS3.dgA or IRac.dgA alone. A cocktail of all 3 ITs was most effective in vitro. This was at least in part due to the non-homogeneous distribution of CD25, CD30 or IRac on the L540Cy H-RS target cells and to the fact that sub-populations deficient in one antigen nevertheless expressed appreciable levels of the other target antigens. IT cocktails were also superior as anti-tumor agents in nude mice with solid L540Cy tumors. Ninety percent of mice treated with cocktails containing 2 or 3 ITs had continuous complete remissions (CCR), as compared with only 40% of mice treated with the same dose of a single IT. Analysis of 7 L540Cy sub-lines re-established ex vivo from mice that relapsed after having achieved complete remission (CR) after therapy with a single IT showed that the surviving tumor cells were antigen-deficient variants which were resistant to the original IT, but which could be killed by ITs directed against other target antigens. Thus, IT cocktails give superior results against human H-RS cells, both in vitro and in vivo.     

Experimental immunotherapy of human breast carcinomas implanted in nude mice with a mixture of monoclonal antibodies against human milk fat globule components

Immunological therapy of BALB/c nude mice (nu/nu) implanted with human breast tumors, estrogen receptor negative MX-1 and estrogen receptor-positive MCF-7, was carried out with four monoclonal antibodies (MoAbs) raised against human milk fat globule membrane glycoproteins also present on normal breast epithelial cells. MoAbs injected singly or as a partial mixture arrested growth of the tumors but to a lesser extent than a mixture ("cocktail") of all four MoAbs. Two model systems were developed in order to examine the capabilities of the four MoAbs to arrest human mammary tumor growth. In the first model the ability of these MoAbs to arrest tumor growth during a 6- to 8-week period was tested by injection of the MoAbs immediately before and after implantation (passive immunization) and thereafter every other day. In the second model the effect of these MoAbs on established and growing tumors was tested. Using the cocktail in the passive immunization protocol, human mammary tumor growth in nu/nu mice was arrested either completely or averaging to one-tenth the size of the controls for those mice in which the tumors had taken. Other human carcinomas, colon and lung, under the same protocol, were not affected. Injection of cocktail every 2 days into nu/nu mice with established and growing human breast tumors (both estrogen receptor positive and negative) produced arrests of tumor growth of 44.1, 45.2, 49.8% of their controls after 7 to 8 days of treatment. Previously, it has been established that human mammary tumors are heterogeneous in expression of the human milk fat globule antigens recognized by our antibodies to the extent that some cells may have large amounts and others no detectable amount of a particular antigen. Those MX-1 tumors treated for a prolonged time with the cocktail of MoAbs that survived and continued to grow could be the result of the preferential multiplication of those cells in the heterogeneous population which had low or no antigen content. The breast tumors that did grow in the nu/nu mice after 8 weeks of injection of the cocktail revealed by immunoperoxidase staining a 90% reduction in the antigen content as recognized by these MoAbs when compared with untreated tumors. These results attest to the effectiveness of unconjugated anti-human milk fat globule MoAbs to arrest human breast tumor growth in nu/nu mice, and they also suggest that to best arrest tumor growth the use of a mixture of MoAbs should be considered.     

Imaging and biodistribution of 111In-labeled anti-carcinoembryonic antigen monoclonal antibodies in nude mice bearing human colon cancer xenografts

Anti-CEA monoclonal antibody was labeled with 111In using DTPA cyclic anhydride. The radiochemical purity of the product was more than 99% and specific binding was 38%. Imaging and biodistribution of this radiopharmaceutical in nude mice bearing human colon cancer xenografts were investigated. The results showed that a clear image of tumor was obtained by gamma camera between 24 to 120 hrs, best in 72 hrs, after injection. A high uptake of 111In-labeled monoclonal antibodies in tumor tissue was also observed, although the radioactivity in liver was considerable.     

Stability, characterization, and kinetics of 111In-labeled monoclonal antitumor antibodies in normal animals and nude mouse-human tumor models

Monoclonal antibodies (MoAbs) against carcinoembryonic antigen were successfully radiolabeled with 111In, and the radiopharmaceutical was characterized in vitro and in normal and tumor-bearing mice. The 111In-MoAb proved to be stable in vitro and in vivo under normal conditions, although instability could be induced in vitro with large quantities of iron-free transferrin. Animal distribution studies with 111In-MoAb demonstrated tumor localization superior to 67Ga and pharmacokinetics that were highly similar to those of endogenously labeled 75Se-MoAb. The 111In-MoAb followed first-order kinetics and fit a two-compartmental model when studied in nude mice bearing human colon tumors known to express carcinoembryonic antigen. Significant quantities of radiolabel appeared in tissues other than tumor, with liver and skin having the highest concentrations. Sufficient tumor/background ratios were formed for scanning purposes. The data indicate that 111In-MoAb may prove to be effective as a radiopharmaceutical for tumor imaging.     

Progression markers in metastasizing human melanoma cells xenografted to nude mice (review)

The past decade transplants of human tumors in nude mice have been increasingly used as an experimental model for local tumor growth and dissemination. A few human melanoma cell lines have been described that give rise to metastases in nude mice after subcutaneous inoculation. First we give an overview of some relevant literature with respect to the pathogenesis of tumor metastasis, models to study human cancer metastasis, neoplastic progression and the detection of antigens involved in metastasis. Finally we describe our results concerning the morphological and immunohistochemical profile of six different human melanoma cell lines and their xenograft lesions in nude mice using a set of monoclonal antibodies recognizing different categories of human melanoma-associated antigens. From the data we conclude that the nude mouse mouse model appears suitable to study the role of melanoma-associated progression markers in the pathogenesis of metastasis.     

Conformational epitopes specific to carcinoembryonic antigen defined by monoclonal antibodies raised against colon cancer xenografts

Four monoclonal antibodies (MoAbs) reactive with carcinoembryonic antigen (CEA) were obtained by hybridizing mouse myeloma cells (P3-X63-Ag8-U1) with spleen cells from nude mice (BALB/c, nu/nu) that had rejected transplanted human colonic adenocarcinomas Co-3 and Co-4 following intraperitoneal injection of spleen cells from immunocompetent mice (BALB/c). By solid-phase RIA with purified CEA and its related antigens, NCC-CO-413 (IgG2a, kappa) was shown to react with NCA and BGP-I as well as with CEA, whereas the reactivities of three other MoAbs, NCC-CO-308 (IgG1, kappa), -432 (IgG1 lambda), and -411 (IgG1, kappa) were limited to CEA. Immunohistochemical reactivities of these MoAbs to colonic carcinomas, granulocytes, and liver bile canaliculi on acetone-fixed paraffin-embedded sections ("AMeX" sections) confirmed the specificities of these MoAbs shown by the solid-phase RIA. By competition solid-phase RIA, the epitopes recognized by NCC-CO-308 and -432 were shown to be shared or located close to each other, whereas the other MoAbs were shown to recognize different epitopes. Thus, two epitopes specific to CEA and one shared by NCA and BGP-I as well as CEA were identified. Furthermore, reactivities of MoAbs with the two CEA-specific epitopes were easily abolished by heat denaturation or reduction of CEA, as revealed by solid-phase RIA and SDS-PAGE-immunoblotting, indicating that these two CEA-specific epitopes are based on the conformational structure of the CEA molecule.     

Increased labeling of human melanoma cells in vitro using combinations of monoclonal antibodies recognizing separate cell surface antigenic determinants

A panel of mouse anti-melanoma monoclonal antibodies (MoAb) were analyzed for reactivity with human melanoma cells singly and in combination. Five MoAb, ZME-018, 96.5, P94, 4.2, and 5.1, reactive with individual cell surface melanoma-associated antigens were tested with seven melanoma cell lines and seven fresh tumor biopsies. Cells were incubated with the MoAb, indirectly stained with fluorescein-conjugated goat anti-mouse immunoglobulin, and analyzed by flow cytometry. Percentage of labeled cells and relative fluorescence intensity (FI) with individual MoAb varied with different cell lines and biopsy samples. The most reactive MoAb, ZME-018, 96.5, and P94, labeled 29-93% of the cells from cell lines with relative FI of 2-59 units, thereby demonstrating phenotypic diversity of these cells. Similar results were obtained with cells derived from tumor biopsies, where 1-73% of cells were labeled and relative FI ranged from 0-27. These variations were reduced by using a "cocktail" of MoAb which recognized different melanoma-associated antigens. In cell lines both the percentage of labeled cells (range, 82-95%) and relative FI (range, 36-115) increased substantially (P less than 0.025 and P less than 0.005, respectively) when a "cocktail" prepared from all five MoAb rather than individual MoAb was used. A cocktail of MoAb increased the percentage of labeled tumor biopsy cells (range, 53-78; P less than 0.01) and relative FI (range, 11-69; P less than 0.025). The mean FI obtained by incubating cells with a cocktail of suboptimal concentrations of three MoAb (ZME-018, 96.5, P94) was 48 +/- 12 (SD), which was significantly increased compared to the mean FI seen with suboptimal concentrations of MoAb alone (ZME-018, 7 +/- 10; 96.5, 8 +/- 7; P94, 2 +/- 2; P less than 0.005). These findings were confirmed by radioimmunoassay using a combination of two MoAb, ZME-018 and 96.5. The data suggest that cocktails of MoAb were more effective than single MoAb alone for melanoma tumor cell labeling in vitro and might be more effective for tumor imaging and therapy.     

Inhibition of tumor cell growth in vitro by murine monoclonal antibodies that recognize a proliferation-associated cell surface antigen system in rats and humans

The mouse monoclonal antibody (MoAb) B3 raised against a rat bladder cancer cell line and the MoAbs HBJ127 and HBJ98 raised against a human bladder cancer cell line recognize homologous antigens predominantly present on proliferating cells of the corresponding species. Examination of MoAb-defined antigen and epitopes revealed that both HBJ127 and HBJ98 MoAbs defined a human cell surface glycoprotein complex having an apparent molecular weight of 125,000-130,000 which was composed of a heavy subunit of a glycoprotein nature (Mr 90,000-95,000) and a disulfide-linked light subunit of protein nature (Mr 30,000-35,000), but the HBJ127 and HBJ98 MoAbs recognized a protein epitope and a sugar epitope on the heavy subunit, respectively. Likewise, the B3 MoAb recognized a protein epitope on the heavy subunit of a rat cellular glycoprotein complex of similar composition to the HBJ127/HBJ98-defined human antigen. Addition of the B3 MoAb to rat and the HBJ127 or HBJ98 MoAb to human tumor cells inhibited the nucleic acid synthesis or the proliferation of the tumor cells in vitro in a dose-dependent manner. The target tumor cells exposed to MoAb could regrow when they were freed from the antibody, indicating that the effect of these MoAbs on the tumor cells is cytostatic and reversible. These MoAbs did not cause down-regulation of the cell surface antigen and did not arrest the cell cycle in a certain phase. These observations indicate that the Mr 125,000 glycoprotein cell surface component detected in both rat and human systems may play a requisite role for cell proliferation and that our MoAbs could inhibit the function by binding to the functionally proximal region of the component.     

Pharmacokinetics of internally labeled monoclonal antibodies as a gold standard: comparison of biodistribution of 75Se-, 111In-, and 125I-labeled monoclonal antibodies in osteogenic sarcoma xenografts in nude mice

In order to know the true biodistribution of anti-tumor monoclonal antibodies, three monoclonal antibodies (OST6, OST7, and OST15) against human osteosarcoma and control antibody were internally labeled with 75Se by incubating [75Se]methionine and hybridoma cells. 75Se-labeled monoclonal antibodies were evaluated both in vitro and in vivo using the human osteogenic sarcoma cell line KT005, and the results were compared with those of 125I- and 111In-labeled antibodies. 75Se-, 125I- and 111In-labeled monoclonal antibodies had identical binding activities to KT005 cells, and the immunoreactivity was in the decreasing order of OST6, OST7, and OST15. On the contrary, in vivo tumor uptake (% injected dose/g) of 75Se- and 125I-labeled antibodies assessed using nude mice bearing human osteosarcoma KT005 was in the order of OST7, OST6, and OST15. In the case of 111In, the order was OST6, OST7, and OST15. High liver uptake was similarly seen with 75Se- and 111In-labeled antibodies, whereas 125I-labeled antibodies showed the lowest tumor and liver uptake. These data indicate that tumor targeting of antibody conjugates are not always predictable from cell binding studies due to the difference of blood clearance of labeled antibodies. Furthermore, biodistribution of both 111In- and 125I-labeled antibodies are not identical with internally labeled antibody. Admitting that internally labeled antibody is a "gold standard" of biodistribution of monoclonal antibody, high liver uptake of 111In-radiolabeled antibodies may be inherent to antibodies. Little, if any, increase in tumor-to-normal tissue ratios of antibody conjugates will be expected compared to those of 111In-labeled antibodies if stably coupled conjugates are administered i.v.     

Development of anti-idiotype monoclonal antibodies for Sialyl Le(a) antigen.

Anti-idiotype monoclonal antibodies (MoAbs) were developed for a carbohydrate antigen, Sialyl Le(a), by immunizing mice with an anti-Sialyl Le(a) MoAb, KM231 (mouse IgG1). The anti-idiotype MoAbs inhibited the binding of KM231 to Sialyl Le(a)-positive mucin protein, but did not affect the binding of another MoAb (mouse IgG1) which recognized a peptide epitope on the same mucin protein. The selected four anti-idiotype MoAbs bound to all of the five anti-Sialyl Le(a) MoAbs examined, but not to other MoAbs developed against Sialyl Le(a)-related carbohydrate antigens such as Le(a), Le(x), and Sialyl Le(x) blood type A, B, H antigens. Immunization of rats with one of the anti-idiotype MoAbs, of which the reactivity was remarkably inhibited by Sialyl Le(a) oligosaccharide, resulted in the induction of antibody (Ab3) to purified Sialyl Le(a) glycolipid. Taken together, one of the anti-idiotype MoAbs developed in this study was the first Ab2 beta-type anti-idiotype MoAb which carried the internal image of Sialyl Le(a) antigen. The positive reaction of tumor cells expressing Sialyl Le(a) antigen with the Ab3 gave rise to the possibility that the anti-idiotype MoAb would become an effective tool for active immunotherapy in patients with Sialyl Le(a)-positive tumor.     

Comparison between 131I and 111In as radiolabels for monoclonal antibodies in immunoscintigraphy of tumor bearing nude mice

F(ab')2 fragments of monoclonal antibody (MoAb) 19-9 with specificity for human colorectal adenocarcinomas were labeled with 111In or 131I and infused into nude mice bearing the human adenocarcinoma HT 29 in order to compare their preferential biodistribution according to the radiolabel used. Animal tissue distribution measured one day and five days after infusion showed that tumor accumulation was greater for 111In than for 131I. However, non specific binding of 111In labeled MoAb 19-9 was also greater in normal tissue than 131I labeled antibody, except in blood. Therefore, the tumor/normal tissue ratios were to the advantage of 131I MoAb 19-9 and a better contrast was obtained on imaging with 131I as compared to 111In labeled MoAb 19-9. Based on this experimental model 111In does not seem to be the optimal candidate for tumor imaging using radiolabeled MoAb.     

Enhancing effect by anti-cancer drugs on growth inhibition of colon carcinoma in nude mice by monoclonal antibody and complement

We investigated the possible sero-therapeutic application of monoclonal antibody-A7 against human colorectal cancer. In complement dependent cytotoxicity, A7 showed 59% cytotoxicity against SW1116 cells. In addition, the killing of tumor cells by A7 and C was enhanced when the tumor cells were pretreated with 2 micrograms/ml mitomycin and 40 micrograms/ml adriamycin. Next, we evaluated the in vivo antitumor effect of A7 alone and combined with MMC on human colon cancer (Colon-6) bearing nude mice. The group injected with A7 alone showed definite antitumor effect compared with the non-treated group. The A7+MMC group (MMC: 4mg/kg, A7: 1 mg/body, two times) showed enhanced antitumor effect compared with the groups administered A7 alone or MMC alone.     

Evaluation of monoclonal antibodies for the development of breast cancer immunotoxins

Eighty-five antibodies recognizing breast cancer-selective antigens were conjugated to ricin toxin A-chain using a disulfide linkage. The cytotoxicities of the resulting immunotoxins were determined on breast cancer cells and normal human fibroblasts. Twenty-four antibodies formed immunotoxins that were toxic to at least one breast cancer cell line at concentrations of 10 nM or less but were nontoxic to human fibroblast lines used as negative controls. Some of the breast tumor-selective immunotoxins were as toxic as a conjugate between monoclonal anti-transferrin receptor and ricin toxin A-chain (50% inhibition of cellular protein synthesis at approximately 0.1 nM). Another set of four immunotoxins were indiscriminately toxic to human breast tumor cell lines, two human fibroblast cell lines, and a human lymphoblastoid line. Several of the antibodies the toxin conjugates of which specifically killed breast cancer cell lines may be useful in cancer therapy, since they show a wide range of binding to individual breast tumors and cell lines and a limited range of binding to normal tissue types.     

Potentiation of platinum antitumor effects in human lung tumor xenografts by the angiogenesis inhibitor squalamine: effects on tumor neovascularization.

Squalamine is a novel anti-angiogenic aminosterol that is postulated to inhibit neovascularization by selectively inhibiting the sodium-hydrogen antiporter exchanger. To determine how to most effectively use this agent in patients with cancer, we examined the antitumor effects of squalamine with or without cytotoxic agents in human lung cancer xenografts and correlated these observations with the degree of tumor neovascularization. No direct cytotoxic effects of squalamine against tumor cells were observed in vitro with or without cisplatin. Squalamine was effective in inhibiting the establishment of H460 human tumors in BALBc nude mice but was ineffective in inhibiting the growth of H460, CALU-6, or NL20T-A human tumor xenografts when administered i.p. to mice bearing established tumors. However, when combined with cisplatin or carboplatin, squalamine increased tumor growth delay by > or =1.5-fold in the three human lung carcinoma cell lines compared with cisplatin or carboplatin alone. No enhancement of antitumor activity was observed when squalamine was combined with paclitaxel, vinorelbine, gemcitabine, or docetaxel. Repeated cycles of squalamine plus cisplatin administration delayed H460 tumor growth >8.6-fold. Squalamine plus cisplatin reduced CD31 vessel formation by 25% compared with controls, squalamine alone, or cisplatin alone; however, no inhibition in CD31 vessel formation was observed when squalamine was combined with vinorelbine. These data demonstrate that the combination of squalamine and a platinum analog has significant preclinical antitumor activity against human lung cancer that is related to the anti-angiogenic effects of squalamine.     

Human tumor xenografts treated with recombinant human tumor necrosis factor alone or in combination with interferons

We have studied the activity of recombinant human tumor necrosis factor (rHuTNF) on six different human tumor xenografts derived from primary breast and bowel tumors and maintained by passage in nude mice. When 5 micrograms rHuTNF was given daily intratumorally to mice with established (approximately, 0.5 cm) tumors, total tumor regression was observed by 3-4 weeks in three of six xenograft lines. In a further two lines tumor stasis or significant slowing of growth was seen. This antitumor action was not accompanied by any consistent macroscopic change in the tumor such as necrosis, but histological examination revealed tumor cell degeneration and a large peritumoral infiltration of host inflammatory cells after 4-7 days therapy. In contrast to these data, little effect was seen when the same dose of rHuTNF was administered i.p. to nude mice bearing these tumors. In only two of six lines was any significant slowing of tumor growth seen. A 5-fold increase in the i.p. dose resulted in improved activity on only one of two xenograft lines tested. Efficacy of the i.p. rHuTNF dose could, however, be enhanced by simultaneous administration of human interferon, alpha or gamma. No obvious signs of toxicity were observed at all rHuTNF doses administered and weights of control and treated mice at the end of the experiments were comparable.