Local but no systemic immunomodulation by intraperitoneal treatment of advanced ovarian cancer with autologous T lymphocytes re-targeted by a bi-specific monoclonal antibody

We have reported a 27% overall anti-tumor response using i.p. immunotherapy of advanced ovarian carcinoma with autologous, ex vivo expanded, T lymphocytes re-targeted with bi-specific monoclonal antibody OC/TR, combined with soluble OC/TR and low-dose recombinant interleukin-2 (IL-2). This treatment had no effect on extraperitoneal disease. Therefore we studied in 13 patients whether this immunotherapeutic protocol resulted only in local or also in systemic immunomodulation. The phenotype of the ex vivo expanded lymphocytes was mainly CD3+, 4-, 8+, 16-, 56-. Their OC/TR-re-targeted cytolytic activity against Igrov-1 ovarian-carcinoma cells was approximately as high in responders as in non-responders. Following most therapeutic cycles, the immunophenotype of lymphocytes recovered from the peritoneal fluid was similar to that of the infused T cells (i.e., mainly CD3+, 4-, 8+) and they were coated with OC/TR. However, cytolytic activity of the recovered lymphocytes against Igrov- 1 cells was low in direct assays, and only slightly increased after additional in vitro re-targeting with OC/TR. Systemically, the i.p. immunotherapy resulted in a transient lymphopenia lasting for about 7 days, low (i.e., 5 to 13 ng/ml) serum concentrations of free, functional OC/TR, and very weak coating of circulating T lymphocytes with OC/TR. These peripheral-blood T lymphocytes did not exert OC/TR-re-targeted cytolytic activity. Thus, locoregional OC/TR-re-targeted cellular immunotherapy resulted in substantial local immunomodulation and anti-tumor effects but virtually no systemic immunomodulation.

Bispecific antibody-dependent cellular cytotoxicity of HER2/neu-overexpressing tumor cells by Fc gamma receptor type I-expressing effector cells

A bispecific antibody, MDX-H210, was developed to target cytotoxic effector cells expressing Fc gamma receptor type I (Fc gammaRI, CD64) to HER2/neu-overexpressing tumor cells. HER2/neu is an appropriate target for immunotherapy due to the high level of expression of this proto-oncogene in a variety of malignancies. The expression of Fc gammaRI is limited primarily to cytotoxic immune cells, including monocytes, macrophages, and cytokine-activated polymorphonuclear (PMN) cells. Therefore, tumor cells bound with MDX-H210 can be selectively recognized by effector cells with cytotoxic potential. MDX-H210 was prepared by chemical conjugation of Fab' fragments derived from the HER2/neu-specific monoclonal antibody, 520C9, and the Fc gammaRI-specific monoclonal antibody, H22. This bispecific molecule demonstrated specific, dose-dependent, and saturable binding to both HER2/neu- and Fc gammaRI-expressing cells. A solid-phase immunoassay that demonstrated simultaneous and specific binding to both antigens was used to confirm the bispecific nature of MDX-H210. Monocytes and PMN cells mediated MDX-H210-dependent lysis of HER2/neu-overexpressing cell lines derived from breast, ovarian, and lung carcinomas. IFN-gamma treatment of monocytes enhanced antibody-dependent cellular cytotoxicity, whereas IFN-gamma and granulocyte colony-stimulating factor were required for PMN cell-mediated tumor cell lysis. In addition, MDX-H210 elicited tumor necrosis factor-alpha secretion from monocytes when cultured in the presence of HER2/neu-positive target cells. These in vitro data suggest that targeting tumor cells to Fc gammaRI with MDX-H210 may be an effective treatment for malignancies that overexpress HER2/neu. The in vivo cytotoxic potential of MDX-H210 may be enhanced by combination therapy with the cytokines granulocyte colony-stimulating factor and IFN-gamma, which up-regulate Fc gammaRI expression on cytotoxic effector cells.

Specific activation of resting T cells against CA19-9+ tumor cells by an anti-CD3/CA19-9 bispecific antibody in combination with a costimulatory anti-CD28 antibody

Specific activation of resting lymphocytes for tumor targeting can be achieved by bispecific monoclonal antibodies (bi-mAb) with specificity for tumor antigens and T-cell-activating antigens, respectively, in combination with a costimulatory anti-CD28 antibody. We describe the generation and function of a bi-mAb with specificity for CD3 and for the tumor antigen CA19-9. The bi-mAb OKT3/NSI19-9 was generated by somatic fusion of two hybridoma lines secreting antibodies against CA19-9 and CD3, respectively. A hybrid/hybridoma was established, and its bi-mAb was characterized. In combination with a costimulatory anti-CD28 mAb resting peripheral lymphocytes could be activated specifically with T-cell proliferation and secretion of high amounts of interferon-gamma. On specific T-cell activation, bi-mAb OKT3/NSI19-9 could also redirect the cytotoxic effects of these T cells toward CA19-9+ tumor cells in vitro. Our results indicate that specific activation of resting T cells with bi-mAb OKT3/NSI19-9 in combination with an anti-CD28 mAb can activate resting T cells specifically and leads to antigen-dependent bi-mAb-mediated cytotoxicity against CA19-9+ target cells. This approach may offer new perspectives for the specific immunotherapy of CA19-9+ tumors.

Induction of multiple anti-c-erbB-2 specificities accompanies a classical idiotypic cascade following 2B1 bispecific monoclonal antibody treatment

The bispecific monoclonal antibody (bsmAb) 2B1, targeting the extracellular domain of c-erbB-2, the protein product of the HER-2/neu proto-ocogene, and Fc gamma RIII (CD16), expressed by human natural killer cells, neutrophils and differentiated monocytes, mediates the specific cytotoxic activity of these effector cells to tumor cells. A group of 24 patients with c-erbB-2-overexpressing tumors were treated with intravenously administered 2B1 in a phase I clinical trial and followed after treatment to evaluate the diversity and extent of the 2B1-induced humoral immune responses. As expected, 17 of 24 patients developed human anti-(murine Ig) antibodies (HAMA) to whole 2B1 IgG in a range from 100 ng/ml to more than 50000 ng/ml; 10 of these patients (42%) had strong (at least 1000 ng/ml) HAMA responses, some of which were still detectable at day 191. These responses were usually associated with similar reactivity to the F(ab')2 fragments of the parental antibodies 520C9 (anti-c-erbB-2) and 3G8 (anti-CD16). We sought evidence of an idiotypic cascade induction, indicating a prolonged specific treatment-induced effect on at least one selected target of 2B1. Using competition-based enzyme-linked immunosorbent assays, specific anti-idiotypic antibodies (Ab2) were detectable against 520C9 in 11 patients and against 3G8 in 13 patients. Peak anti-idiotypic antibodies generally occurred 3-5 weeks from treatment initiation, with a downward trend thereafter. There was a statistically significant correlation among the induction of significant HAMA responses, anti-idiotypic antibody production and the development of antibodies to c-erbB-2. The anti-c-erbB-2 responses, which were distinct from anti-anti-idiotypic (Ab3) antibodies, were detected in the post-treatment sera of 6/16 patients examined. No obvious correlation could be made between the development of humoral immune responses, the dose received, and the clinical response. Future investigation involving 2B1 therapy will concentrate on investigating an association of these humoral responses to any c-erbB-2-specific cellular responses. Manipulations of 2B1 therapy effects that augment immunity to c-erbB-2 could provide additional avenues for immunotherapy with this and other bispecific antibodies.

Bispecific antibodies overcome the opsonin-receptor mismatch of cystic fibrosis in vitro: restoration of neutrophil-mediated phagocytosis and killing of Pseudomonas aeruginosa

Inflammation and infection associated with bacterial pathogens, primarily Pseudomonas aeruginosa (Pa), are the primary causes of morbidity and mortality for cystic fibrosis (CF) patients. CF patients may be predisposed to these bacterial infections by a defect in phagocytosis due to "opsonin-receptor mismatch," in which a complement receptor (CR1) and an important opsonin (iC3b) are destroyed by proteolytic enzymes. We show that opsonin-receptor mismatch can be mitigated in vitro using a bispecific Ab (bsAb) to cross-link neutrophils via the beta-chain of leukocyte integrins (CD18) to bacterial epitopes or C3d on opsonized Pa. Two chemically cross-linked bsAb were constructed with mAb specific for C3d (or the O-specific side chain of Fisher Devlin Immunotype 1 Pa) and CD18. Using an in vitro model of elastase-mediated opsonin-receptor mismatch, these bsAb specifically enhanced Pa phagocytosis and killing, with the anti-C3d-containing bsAb restoring the levels of phagocytosis to approximately those for the non-elastase-treated opsonic control. These results encourage the further investigation of bsAb as therapeutic agents for bacterial infection in the lungs of CF patients.

Treatment of refractory Hodgkin's disease with an anti-CD16/CD30 bispecific antibody

Fifteen patients with refractory Hodgkin's disease were treated in a phase I/II trial with the natural killer (NK)-cell-activating bispecific monoclonal antibody HRS-3/A9, which is directed against the Fc(gamma)-receptor III (CD16 antigen) and the Hodgkin's-associated CD30 antigen, respectively. Median counts of NK cells and of all lymphocyte subsets were considerably decreased in the patients before therapy. HRS-3/A9 was administered 4 times every 3 to 4 days, starting with 1 mg/m2. The treatment was well tolerated, and the maximum tolerated dose was not reached at 64 mg/m2, the highest dose administered because of the limited amounts of HRS-3/A9 available. Side effects were rare and consisted of fever, pain in involved lymph nodes, and a maculopapulous rash. A total of 9 patients developed human antimouse Ig antibodies, and 4 patients developed an allergic reaction after attempted retreatment. A total of 1 complete and 1 partial remission (lasting 6 and 3 months, respectively) [corrected], 3 minor responses (1 to 11+ months), and 1 mixed response were achieved. There was no clear-cut dose-side effect or dose-response correlation. Our results encourage further clinical trials with this novel immunotherapeutic approach and emphasize the necessity to reduce the immunogenicity of the murine bispecific antibodies.

Clinical significance of IgG Fc receptors and Fc gamma R-directed immunotherapies

Fc receptors for IgG (Fc gamma Rs) can trigger the inflammatory, cytotoxic and hypersensitivity functions of immune effector cells. Activation or deactivation of effector cells via Fc gamma Rs can be exploited to develop novel therapies for cancer, infectious diseases and autoimmune disorders. Initial results of clinical trials for several Fc gamma R-directed immunotherapies show the potential promise of this approach.

Preclinical studies with Fc(gamma)R bispecific antibodies and granulocyte colony-stimulating factor-primed neutrophils as effector cells against HER-2/neu overexpressing breast cancer

Immunotherapies directed to the proto-oncogene product HER-2/neu, which is overexpressed on a subset of breast and other carcinomas, currently receive considerable attention. We have investigated cell-mediated effector mechanisms of HER-2/neu antibodies against breast cancer cell lines. Compared to unfractionated control blood, whole blood from patients during granulocyte colony-stimulating factor (G-CSF) treatment exhibits significantly enhanced lysis (P < 0.001) of SK-BR-3 cells in the presence of HER-2/neu antibody 520C9. The extent of tumor cell killing correlated positively (r = 0.74) to polymorphonuclear neutrophil (PMN) blood counts. Fractionation of whole blood into plasma, mononuclear cells, and PMNs showed major killing capacity to reside in the granulocyte fraction. PMNs were efficiently cytolytic with a panel of HER-2/neu antibodies and against various breast cancer cell lines. Experiments with blocking antibodies to Fc(gamma)R documented Fc(gamma)RII (CD32) as the major trigger molecule for monoclonal antibody 502C9-mediated cytotoxicity. Killing via 520C9 was significantly influenced by an allotypic polymorphism of Fc(gamma)RIIa, the CD32 molecule expressed on PMNs. In reverse antibody-dependent cell-mediated cytotoxicity experiments with a panel of HER-2/neu-directed bispecific antibodies, Fc(gamma)RIII (CD16) proved to be an efficient trigger molecule in blood from healthy volunteers. During G-CSF treatment, however, Fc(gamma)RI (CD64)-expressed on monocytes and G-CSF primed, but not on healthy donor PMNs-became the predominant cytotoxic trigger molecule. Thus, G-CSF application increased effector cell numbers for HER-2/neu-directed immunotherapy, and G-CSF primed PMNs proved particularly effective with a [HER-2/neu x Fc(gamma)RI] bispecific antibody. These findings support clinical trials with HER-2/neu-directed antibodies in combination with G-CSF in breast cancer patients overexpressing HER-2/neu.

Synthesis and biodistribution of peptide based 99mTc/186Re-MAGIPG-D612 monoclonal antibody in nude mice bearing colon cancer xenografts

A new bifunctional chelating agent MAGIPG was synthesized in good yield using Boc-p-nitro-phenylalanine as the starting material. 99mTc-labeled-MAGIPG-D612 conjugates reactive with human colon cancer were prepared in 3% to 15% yield with a radiochemical purity of 63% to 97%. 186Re-labeled MAGIPG-D612 conjugates were prepared with a 7% to 30% yield and a radiochemical purity of 82% to 98%. The biodistribution results demonstrate that these radioimmuno-conjugates selectively localized in LS174T human colon cancer xenografts.

Targeting medullary thyroid carcinomas with bispecific antibodies and bivalent haptens. Results and clinical perspectives

The present article reviews the clinical trials that have been performed in recurrent medullary thyroid carcinoma patients with the Affinity Enhancement System. This technique uses bispecific antibodies to target radiolabelled bivalent haptens to tumour cells. Its sensitivity in the detection of known tumour sites is high (90%) and this technique also achieves good sensitivity (61%) in the detection of occult disease as revealed by abnormal thyrocalcitonin blood levels. Due to its high targeting capacity, this technique is now considered for use as a therapeutic agent in medullary thyroid carcinoma patients.

Bispecific antibodies target operationally tumor-specific antigens in two leukemia relapse models

Despite improved procedures in chemotherapy and bone marrow transplantation (BMT), post-BMT leukemia relapse rates have remained rather constant in the last decade. Immunotherapy with monoclonal or bispecific antibodies (bsAb) is a promising approach to improve this situation, but is hampered by the absence of tumor-specific antigens on the majority of tumors. To evade this problem, we developed a new tumor-specific approach in which bispecific antibodies exploit chimerism after allogeneic BMT by redirecting donor T cells against recipient-specific antigens on tumor cells. Two different leukemia relapse models were established using a T-cell lymphoma (ST-1) and a B-cell lymphoma (BCL1) to evaluate the efficiency of such a therapy. In these experiments, irradiated BALB/c (Thy-1.2+, I-Ad) mice were transplanted with C57BL/6 Thy-1.1 (I-Ab) BM cells under the protection of graft-versus-host disease-preventing monoclonal antibodies. Forty-five days after BMT, the chimeric mice were injected with either 2 x 10(4) recipient-type, Thy-1.2+, CD3- ST-1 cells or major histocompatability complex (MHC) class II+ (I-Ad)-BCL1 cells. Four days later, the mice were treated with 8 microg bsAb G2 (anti-CD3 x anti-Thy-1.2) or 10 microg (+10 microg, day 6) bsAb BiC (anti-CD3 x anti-I-Ad), respectively. These combinations guaranteed exclusive binding of the bsAbs target arms to tumor cells, leaving the surrounding, donor-type hematopoietic cells unbound. Compared with the parental antibodies, the bsAbs markedly reduced tumor mortality. Between 34% and 83% of mice survived in the bsAb groups compared with 0% of the control groups treated with parental antibodies, clearly documenting the benefit of the redirection principle. Furthermore, cytokine release (interleukin-6) after anti-CD3 antibody or bsAb treatment was decreased by administering a low-dose antibody preinjection. We have shown (1) that 6 weeks after BMT, when donor T-cell reconstitution is still in progress, T-cell-redirecting bsAb are clearly superior to parental antibodies in terms of tumor cell elimination; and (2) that the polymorphism of a common antigen such as Thy-1 or a clinically more relevant target antigen such as MHC class II can be used as an operational tumor-specific antigen after allogeneic BMT.

CD3-CD28 costimulation as a means to avoiding T cell preactivation in bispecific monoclonal antibody-based treatment of ovarian carcinoma

One of the major limitations to the immunotherapy of ovarian carcinoma based on the use of anti-CD3/antitumor bispecific monoclonal antibodies (bi-mAb) is the need for preactivation of effector cells ex vivo, because cross-linking of the T cell receptor-CD3 complex per se may lead to T-cell unresponsiveness or even apoptosis. The bi-mAb OC/TR, which recognizes the folate-binding protein (FBP) overexpressed in 90% of ovarian carcinomas and the CD3 molecule on T cells, has demonstrated efficacy in a clinical setting. Here we investigated the possibility of delivering accessory signals to OC/TR-retargeted peripheral blood mononuclear cells (PBMCs) via an anti-CD28 mAb or an anti-FBP/anti-CD28 bi-mAb. Coculture of resting PBMCs from healthy donors with OC/TR, anti-FBP/anti-CD28 bi-mAb, and FBP+ tumor cell lines resulted in a highly activated phenotype of effector cells and in a dramatic in vitro growth inhibition of the target cells without an increase in OC/TR-redirected lysis. Whereas both the CD4 and CD8 T cell subsets were involved in the growth inhibition, only the CD8 subpopulation accounted for the cytotoxic activity. The in vitro tumor growth inhibition was mediated mainly by soluble factors, which were active on both FBP+ and FBP- ("bystander effect") cell lines. Activation and antitumor activity were also observed, albeit to a lesser extent, using OC/TR and monospecific bivalent anti-CD28 mAb. In vitro analysis demonstrated that cross-linking between tumor and effector cells for at least 24 h was needed to achieve T-cell activation and development of antitumor activities. Thus, ex vivo CD3-CD28 costimulation on resting PBMCs might be of therapeutic utility for local treatment of minimal residual disease.

Mono- and bispecific single-chain antibody fragments for cancer therapy

Especially when dealing with solid cancers, single-chain antibody fragments (scFvs) have a lot of advantages. Due to their small size (27 kDa), these proteins clear more rapidly from the blood, and penetrate faster and deeper into tissues, than whole antibodies. Furthermore, the lack of constant regions ensures that they are not retained in tissues such as the liver and kidney. This reduces possible toxic side-effects. Single-chain construction is normally done by polymerase chain reaction (PCR). To decrease the overall cost of oligonucleotide primer synthesis, time-consuming primer design, multiple PCR reactions and individual PCR optimization, we designed a universal single-step overlap extension PCR protocol using hybridoma cDNA as a template. To overcome the lack of effector function, bispecific scFvs, consisting of an scFv produced against a tumour-associated antigen fused to a T cell marker-specific scFv, are being created, starting from already assembled scFv, by means of two additional PCR reactions. In this paper we describe both PCR methods that were successfully used to create scFvs against the human transferrin receptor, the human interleukin-2 receptor, the human CD3 molecule, a breast tumour-associated antigen and an anti-transferrin-anti-CD3 bispecific scFv.

Idiotype vaccination strategies against a murine B-cell lymphoma: dendritic cells loaded with idiotype and bispecific idiotype x anti-class II antibodies can protect against tumor growth

Three strategies were used to evaluate 38C13 B-cell lymphoma-specific idiotype immunization to protect against subsequent lymphoma challenge in C3H/He mice. It was observed that tumor-specific immunity could be induced by immunization with (i) KLH-conjugated 38C13 B-cell lymphoma idiotype in complete Freund's adjuvants (survival rate 80%), (ii) dendritic cells pulsed in vitro with native idiotype protein (survival rate 80%), and (iii) bispecific antibodies composed of B-lymphoma-related idiotype and an MHC class II binding moiety (survival rate 40%). Presentation of idiotype determinants by dendritic cells or bispecific antibody resulted in lymphoma-specific immunity and obviated the requirement for carrier protein or adjuvant. Moreover, primed dendritic cells induced predominant development of a tumor-specific T-cell response. Each of these immunization strategies resulted in long-term survival without the emergence of idiotype variants or the induction of tumor dormancy.

Bispecific antibody and iodine-131-labeled bivalent hapten dosimetry in patients with medullary thyroid or small-cell lung cancer

The purpose of this study was to estimate the dose delivered to tumor targets and normal tissues after two-step injection of an anti-CEA/anti-DTPA-In (F6-734) bispecific antibody and a 131I-labeled di-DTPA in-TL bivalent hapten in patients with medullary thyroid carcinoma (MTC) and small-cell lung cancer (SCLC).

METHODS

Five patients with persistent disease or recurrences of MTC and five patients with primary SCLC or relapse were studied. In a first step, 0.1 to 0.3 mg/kg of F6-734 bispecific antibody was injected intravenously. Four days later, 6 nmole (5.8 to 9.8 mCi) of 131I-labeled di-DTPA in-TL bivalent hapten were injected. Quantitative imaging was performed during one week after the second injection.

RESULTS

All 5 patients with MTC showed positive immunoscintigraphy (IS). In the smallest visualized and resected tumor (0.8 g), the fraction of injected activity per gram (% ID/g) was 0.1% at Day 3. IS was positive in 4 of the 5 patients with SCLC. The volume of the smallest visualized SCLC tumor was estimated at 11 +/- 2 ml, and tumor uptake was about 0.009% ID/g. Tumor dose estimates ranged from 4.2 to 174 cGy/mCi in patients with MTC and from 1.7 to 8 cGy/mCi in patients with SCLC.

CONCLUSION

High absorbed dose values were calculated for small MTC recurrences. For SCLC recurrences the values were smaller but in the same range as those obtained by other investigators with the one-step technique in lymphoma.

Cytokine targeting in tumors using a bispecific antibody directed against carcinoembryonic antigen and tumor necrosis factor alpha

The use of tumor necrosis factor alpha (TNFalpha) in cancer therapy is limited by its short circulatory half-life and its severe systemic side effects. To overcome these limitations, we evaluated the capability of a bispecific antibody (BAb) directed against carcinoembryonic antigen (CEA) and human TNFalpha to target this cytokine in tumors. A BAb was constructed by coupling the Fab' fragments from an anti-CEA monoclonal antibody (MAb) to the Fab' fragments from an anti-TNFalpha MAb via a stable thioether linkage. The double specificity of the BAb for CEA and TNFalpha was demonstrated using a BIAcoreTM two-step analysis. The affinity constants of the BAb for CEA immobilized on a sensor chip and for soluble TNFalpha added to the CEA-BAb complex were as high as those of the parental MAbs (1.7 x 10(9) M-1 and 6.6 x 10(8) M-1, respectively). The radiolabeled 125I-labeled BAb retained high immunoreactivity with both CEA and TNFalpha immobilized on a solid phase. In nude mice xenografted with the human colorectal carcinoma T380, the 125I-labeled BAb showed a tumor localization and biodistribution comparable to that of 131I-labeled anti-CEA parental F(ab')2 with 25-30% of the injected dose (ID)/g tumor at 24 h and 20% ID/g tumor at 48 h. To target TNFalpha to the tumor, a two-step i.v. injection protocol was used first, in which a variable dose of 125I-labeled BAb was injected, followed 24 or 48 h later by a constant dose of 131I-labeled TNFalpha (1 microg). Mice pretreated with 3 microg of BAb and sacrificed 2, 4, 6, or 8 h after the injection of TNFalpha showed a 1.5- to 2-fold increased concentration of 131I-labeled TNFalpha in the tumor as compared to control mice, which received TNFalpha alone. With a higher dose of BAb (25 microg), mice showed a better targeting of TNFalpha with a 3.2-fold increased concentration of 131I-labeled TNFalpha in the tumor: 9.3% versus 2.9% ID/g in control mice 6 h after TNFa injection. In a one-step injection protocol using a premixed BAb-TNFalpha preparation, similar results were obtained 6 h postinjection (3.5-fold increased TNFalpha tumor concentration). A longer retention time of TNFalpha was observed leading to an 8.1-fold increased concentration of TNFalpha in the tumor 14 h postinjection (4.4 versus 0.5% ID/g tumor for BAb-treated and control mice, respectively). These results show that our BAb is able, first, to localize in a human colon carcinoma and, there, to immunoabsorb the i.v.-injected TNFalpha, leading to its increased concentration at the tumor site.

Comparative targeting of human colon-carcinoma multicell spheroids using one- and two-step (bispecific antibody) techniques

In the perspective of radioimmunotherapy (RIT) of micrometastases, we compared, in multicell spheroids (MS), the uptake and retention kinetics of 125I-F(ab)'2 F6 anti-carcinoembryonic antigen (CEA) monoclonal antibody (MAb), and the affinity enhancement system (AES) using an anti-CEA/anti-DTPA-indium bispecific antibody (BsMAb) and a 125I-labeled di-DTPA-In-tyrosine-lysine bivalent hapten. We used MS of colorectal tumor cell lines expressing CEA strongly (LS 174T), weakly (HT-29) or not at all (HRT-18). Uptake and retention kinetics of 125I-F(ab)'2 F6 and 125I-BsMAb used alone gave similar results. The highest uptake values, obtained with LS 174T MS, were slightly lower with AES than with 125I-F(ab)'2 F6. However, effective retention half-lives were longer for AES than for 125I-F(ab)'2 F6 or for 111In-labeled monovalent hapten after pre-incubation of spheroids with BsMAb. Autoradiography showed the same slow and heterogeneous distribution of 125I-F(ab)'2 F6 and 125I-BsMAb. These results indicate that the 2-step technique is more favorable for RIT: uptake values were approximately the same but uptake kinetics were more rapid, and retention half-life was longer than with the one-step technique.

MUC1-specific targeting immunotherapy with bispecific antibodies: inhibition of xenografted human bile duct carcinoma growth

For the purpose of establishing a new adoptive immunotherapy for bile duct carcinoma (BDC), we synthesized two bispecific antibodies (BsAbs), MUC1 x CD3 BsAb constructed with MUSE11 (anti-MUC1 tumor antigen) and OKT-3 (anti-CD3), and MUC1 x CD28 BsAb constructed with MUSE11 and 15E8 (anti-CD28) antibodies. These two BsAbs reacted well with both MUC1-positive target tumor cells and effector lymphokine-activated killer (LAK) cells. Investigation of in vitro cytotoxicity [3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyltetrazolium bromide assay] revealed that the MUC1 x CD3 BsAb could antigen-specifically enhance the cytotoxicity of LAK cells. Addition of the two BsAbs (MUC1 x CD3 BsAb plus MUC1 x CD28 BsAb) in vitro resulted in a 60% cytotoxicity, similar to that obtained with BsAb (MUC1 x CD3) alone. Interleukin 12-induced LAK cells demonstrated far greater cytotoxicity (50%) than their interleukin 2-induced counterparts (LAK cells), and this was also enhanced by the BsAbs. When 2 x 10(7) LAK cells sensitized with both kinds of BsAbs were administered four times i.v. to BDC-grafted severe combined immunodeficient mice (tumor size 5 mm in diameter), inhibition of tumor growth was observed. Thus, BsAb-LAK therapy for control of BDC warrants clinical trials.

Targeted inhibition of tumour cell growth by a bispecific single-chain toxin containing an antibody domain and TGF alpha

Overexpression of the epidermal growth factor receptor (EGFR) and ErbB-2 has been observed in a variety of human tumours, making these receptors promising targets for directed tumour therapy. Since many tumour cells express both ErbB-2 and EGFR and these receptors synergise in cellular transformation, therapeutic reagents simultaneously binding to ErbB-2 and EGFR might offer advantages for tumour therapy. We have previously described the potent anti-tumoral activity of a bispecific antibody toxin that contains ErbB-2- and EGFR-specific single-chain Fv (scFv) domains. Here we report the construction and functional characterisation of a novel bispecific recombinant toxin, scFv(FRP5)-TGF alpha-ETA. The fusion protein consists of the antigen-binding domain of the ErbB-2-specific MAb, FRP5, and the natural EGFR ligand, TGF alpha, inserted at different positions in truncated Pseudomonas exotoxin A. ScFv(FRP5)-TGF alpha-ETA protein displayed binding to EGFR and ErbB-2, thereby inducing activation of the receptors, which was dependent on the cellular context and the level of EGFR and ErbB-2 expression. The bispecific molecule was cytotoxic in vitro for tumour cells expressing various levels of the target receptors. In vivo scFv(FRP5)-TGF alpha-ETA potently inhibited the growth of established A431 tumour xenografts in nude mice.

Critical evaluation of bispecific antibodies as targeting agents for boron neutron capture therapy of brain tumors

Boron neutron capture therapy (BNCT) is based on the nuclear capture reaction that occurs when 10B, a stable isotope, is irradiated with low energy neutrons to produce high linear energy transfer (LET) alpha particles and recoiling 7Li nuclei. In order for BNCT to be successful in treating cancer, approximately 10(9) boron atoms must be delivered per tumor cell to sustain a lethal 10B, (n,a) 7Li capture reaction. In the present study, we have produced and characterized a bispecific antibody (BsAbB8), which was reactive with both human glioma and melanoma cell lines, as well as with a variety of polyhedral borane anions (PBA). The affinity constants (KA) of BsAb-B8 with D-54 MG and M21 cells were 3.49 and 2.57 x 10(8) M-1, respectively, which were almost identical to those of the parental mAb 9.2.27 with these cell lines. In vivo tumor localizing properties were studied in nude mice bearing subcutaneous xenografts of the D-54 MG glioma. Following intravenous injection of 131I-labeled BsAb-B8, 3.4 +/- 0.2% of the injected dose/g was detected in the tumor at 24 hours, and then slowly declined to 2.0 +/- 0.4% at 96 hours compared to 1.34 +/- 0.07% and 0.03 +/- 0.01%, respectively, for normal mouse IgG. Based on the assumption that all the tumor cell antigenic receptor sites could be saturated, the following calculations have been carried out. The maximum concentration of BsAb-B8 that could be delivered to 1 g of D-54 MG glioma cells would be 99.6 micrograms, which could bind 71.7 ng of a PBA. However, since at least 500 x more boron would be required per gram of tumor to sustain a lethal 10B (n,a) 7Li capture reaction, a macromolecule containing -10(3)-10(4) boron atoms rather than a low molecular weight PBA would be required to deliver this amount. Such boron containing macromolecules have been synthesized by us, and future studies should provide information on the feasibility of using them in combination with BsAb-B8 to deliver the requisite amount of 10B.

Cytolysis of malignant glioma cells by lymphokine-activated killer cells combined with anti-CD3/antiglioma bifunctional antibody and tumor necrosis factor-alpha

With the aim of developing an effective immunotherapy for malignant glioma, glioma cells were incubated with tumor necrosis factor-alpha (TNF-alpha) to increase their susceptibility to lysis by lymphokine-activated killer (LAK) cells. Treatment with exogenous TNF-alpha induced the expression of intercellular adhesion molecule-1 (ICAM-1) on the surface of glioma cells. In addition, the cytolytic activity of LAK cells toward exogenous TNF-alpha treated glioma cells was significantly greater than LAK cell activity toward untreated glioma cells. This increase in cytolytic activity was blocked by anti-ICAM-1 monoclonal antibodies (MAb). Furthermore, co-treatment with a bifunctional antibody (BFA) composed of anti-CD3 (UCHT1) and antiglioma (G-22) antibodies synergistically increased the cytolytic activity of LAK cells towards TNF-alpha-treated glioma cells. These results indicate that a combination of exogenous TNF-alpha and anti-CD3/antiglioma BFA may provide an effective modified adoptive immunotherapy for patients with malignant glioma.

Response of B-cell lymphoma to a combination of bispecific antibodies and saporin

Observations are described using a combination of two bispecific F(ab')2 antibodies (BsAb) to deliver the ribosome-inactivating protein, saporin, in the treatment of low-grade, end-stage, B-cell lymphoma. Two BsAb were used, each having one arm directed at saporin and one at the CD22 on target B cells. The BsAb, however, recognized different, non-overlapping epitopes on each molecule, a strategy which permits high-avidity double attachment of saporin to the target. The BsAb and saporin were pre-mixed at a molar ratio of 3:1 24 h before treatment and infused intravenously over a period of 1 h. Five patients have been treated, mostly with weekly doses of between 2 and 4 mg of saporin for a period of up to 6 weeks. Toxicity was minimal. Three complained of weakness and myalgia for 1 to 2 days after treatment, without objective neurological deficit or rise in serum creatine kinase. One patient produced an anti-mouse Fab' and an anti-saporin response. All patients showed a rapid and beneficial response to treatment. When present, circulating tumor cells were cleared (4/4 patients), ascitic and pleural effusions were eliminated (2/2 patients) and one patient with splenomegaly showed a marked reduction in tumor bulk. Malignant lymph nodes showed significant, but partial, shrinkage in all patients and finally marrow responded well with tumor clearance in biopsy material and impressive resolution of pancytopenia in some patients. While these responses were mainly short-lived, with tumor progression once the treatment was stopped, their speed and magnitude, and the relative lack of associated toxicity warrants further study of this treatment to determine maximum tolerated doses and therapeutic utility.

Bispecific antibody-mediated immunotherapy of the BCL1 lymphoma: increased efficacy with multiple injections and CD28-induced costimulation

The BCL1 lymphoma in Balb/c mice can be successfully treated with bispecific (anti-CD3 x anti-idiotype) antibodies (BSABs). In these experiments, animals were injected intraperitoneally (IP) with 5 x 10(3) tumor cells (day 0) and treated with one single intravenous (IV) injection of 5 micrograms BSAB (day 9). Because cross-linking of the CD3 complex is not in itself sufficient to activate resting T cells, the therapeutic success was mainly based on the progressive retargeting of the relatively small cytotoxic T-lymphocyte effector cell pool already in existence in vivo. For this reason, the therapy lost its effectiveness at higher tumor loads. Two possibilities were explored to treat mice with a higher tumor load (10(5) tumor cells). First multiple injections of BSABs were used, but a dose-related monovalent anti-CD3 immunosuppression was induced. This approach was only beneficial when the immune system was able to recover from the previous injection of BSAB. As a second approach, CD28 costimulation together with BSABs were used in an attempt to activate resting T cells, ultimately enlarging the effector T-cell pool. However, we were repeatedly unsuccessful in attempts to improve our in vivo results using young, naive animals in which the majority of T cells are of the naive phenotype. Only when animals were primed to induce the memory T-cell type was a significantly better outcome observed, and then only with multiple injections of BSABs and anti-CD28 monoclonal antibodies (MoAbs), rather than with BSAB or anti-CD28 MoAb alone. These results suggest that this combination was able to activate memory and effector T cells and to focus them on the tumor, but was unable to activate naive T cells fully. The in vivo potency of the BSAB and CD28 costimulation was shown by the fact that one-tenth of the quantity of BSAB was required to cure animals with 20 times the tumor load.

Cure of disseminated xenografted human Hodgkin's tumors by bispecific monoclonal antibodies and human T cells: the role of human T-cell subsets in a preclinical model

Cure of a single established human Hodgkin's tumor growing subcutaneously in severe combined immunodeficient (SCID) mice can be achieved with a complex protocol using two bispecific monoclonal antibodies (Bi-MoAb) directed against the Hodgkin's associated CD30 antigen and the T-cell triggering molecules CD3 and CD28, respectively, together with human T cells prestimulated in vitro with Bi-MoAbs in the presence of CD30+ cells. To adapt this model to the clinical situation, disseminated tumors were established in SCID mice by intravenous injection of 2 x 10(7) cells of the Hodgkin's derived cell line L540CY. Treatment of SCID mice bearing disseminated CD30+ Hodgkin's tumors with the combination of CD3/CD30 and CD28/CD30 Bi-MoAbs and naive (ie, not in vitro prestimulated) human T cells resulted in the cure of all appropriately treated animals. T lymphocytes obtained from patients with advanced stage untreated Hodgkin's disease were as effective as lymphocytes from healthy controls. Treatment was effective even when delayed until 2 weeks after tumor inoculation, and application of Bi-MoAbs into SCID mice with circulating human T cells was as effective as injecting the Bi-MoAbs before the lymphocytes. Treatment results with isolated CD4+ and CD8+ human T cells suggest that both subsets are necessary for the Bi-MoAb mediated cure of xenografted human tumors in vivo. The efficacy and practicability of this preclinical immunotherapy protocol support and form the basis for the clinical evaluation of this approach in patients with Hodgkin's disease resistant to standard therapy.

Human cancer detection and immunotherapy with conjugated and non-conjugated monoclonal antibodies

Classical therapeutic modalities such as surgery, radiation and chemotherapy not only fail to cure the majority of neoplastic disease, but their employment also leads to severe and debilitating side effects. The severe cancer related morbidity is often associated with the use of radiation and chemotherapy, making them less than ideal forms of therapy. Entirely new approaches to cancer therapy that are tumor cell directed, and specifically lethal to malignant cells and less toxic to normal tissues are being observed and developed, adhering to the old prayer "Destroy the diseased tissues, preserve the normal." Following the initial advances of Ehrlich, immunotherapy as a fourth modality of cancer therapy has already been developed and proven to be quite effective. Unfortunately, the cancer cell population is not a static entity, but rather a continually changing one. Considerable variations have been determined between individual malignant cells. Our strong belief is that it is necessary for present-day clinical oncologists to become aware of the existence of immunotherapy and learn how to employ it in order to improve the efficacy and decrease the side effects of modern cancer therapy. The development of hybridoma technology and the advances in monoclonal antibody (MoAB) production have revitalized the concept concerning the existence of cancer cell-targeted, specific "magic bullets". In addition, a variety of different agents (e.g. toxins, radionuclides, chemotherapeutic drugs) have been conjugated to mouse and human MoABs for selective delivery to cancer cells. Preclinical observations in athymic, nude mice using xenografted human cancers and mouse, anti-human MoABs were more than impressive and have lead to several clinical trials. Strategies for the employment of MoABs for cancer immunotherapy include: a) Immune reaction directed destruction of cancer cells; b) Interference with the growth and differentiation of malignant cells; c) Antigen epitope directed transport of anti-cancer agents to malignant cells; d) Anti-idiotype vaccines. Phase I studies have established the safety of employing immunoconjugates in humans, but the therapeutic results were less impressive. The clinical use of mouse MoABs in humans is limited due to the development of an anti-globulin immune response to the non-human immunoglobulins by the human host. Genetically engineered chimeric human-mouse MoABs have been developed by replacing the mouse Fc region with the human constant region. Moreover, the framework regions of variable domains of rodent immunoglobulins were also experimentally replaced by their human equivalents. These antibodies can also be designed to have specificities and effector functions determined by researchers, which may not appear in nature. The astonishing immunophenotypic (IP) heterogeneity of cancer cells, the different cytotoxic activity associated with the moiety linked to given MoABs, and mostly the impressive genetic modulation capabilities of cancer cells still remain as yet unsolved difficulties in the present immunotherapy of human cancer. Antibodies with two binding ends (bispecific antibodies) provide a great improvement in targeting cancer cells. The existing inadequacies of MoABs in immunotherapy may also be improved by increasing their efficiency with chemical coupling to various agents such as bacterial or plant toxins, radionuclides or cytotoxic drugs. In writing this review article, one of our main goals is to encourage further clinical research with the use of genetically engineered rodent MoABs and various immunoconjugates in the treatment of human cancer, as well as the combination of such immunotherapy with the three conventional modalities of therapy. Finally, we propose that MoAB-based immuno-therapy be accepted as a conventional form of therapy and employed not only in terminal cancer patients, but also, for instance, during and following surgical resection.

A bivalent single-chain antibody-toxin specific for ErbB-2 and the EGF receptor

ErbB-2 and EGF receptors are often co-expressed in human tumors and have been shown to synergize in the transformation of cells in experimental model systems. Transactivation of ErbB-2 can occur via ligand-induced heterodimerization with EGF receptor or other members of the ErbB family of receptor tyrosine kinases. We have previously described the potent anti-tumoral activity of the monospecific single-chain antibody-toxins scFv(FRP5)-ETA and scFv(225)-ETA binding to, respectively, ErbB-2 and the EGF receptor. Here we report the construction and functional characterization of a novel bivalent, bispecific single-chain antibody-toxin, scFv2(FRP5/225)-ETA. The fusion protein consists of 2 scFv domains specific for ErbB-2 and the EGF receptor linked to a modified Pseudomonas exotoxin A. ScFv2(FRP5/225)-ETA displayed in vitro cell killing activity on tumor cells overexpressing either ErbB-2 or the EGF receptor similar to that of the monospecific toxins. It was more potent in vitro and in vivo in inhibiting the growth of tumor cells expressing both receptors. Treatment of A431 cells with scFv2(FRP5/225)-ETA led to an increase in EGF receptor and ErbB-2 phosphotyrosine content, most likely via the induction of receptor heterodimers. This may explain the enhanced toxicity of the bispecific antibody-toxin.

A new in vitro model of specific targeting therapy of cancer: retargeting of PWM-LAK cells with bispecific antibodies greatly enhances cytotoxicity to hepatocellular carcinoma

For the purpose of establishing a new in vitro model of adoptive immunotherapy, we synthesized two kinds of bispecific antibodies (BsAbs), i.e., (OK x L) BsAbs constructed with both OKT-3 (anti-CD3) and L-7-6 (anti-HCC), and (3G x L) BsAbs constructed with 3-G-8 (anti-CD16) and L-7-6 antibodies. These two BsAbs, having pairs of binding arms on their single molecule, showed similar binding to target cells as the parental monoclonal antibodies (OKT-3, 3-G-8 and L-7-6), when examined with FACS. Newly devised in vitro cytotoxicity tests revealed that LAK or PWM-stimulated LAK (PWM-LAK) cells did not show any significant cytotoxic activity to HCC cells, while both effector cells equally showed greatly enhanced cytotoxicity to HCC even at a low effector/target (0.3) in the presence of BsAbs (OK x L) for the efficient retargeting of the effector cells. Inasmuch as PWM-LAK cells proliferate in vitro 3-5 times faster than LAK cells, adoptive immunotherapy using PWM-LAK cells in combination with (OK x L) BsAbs should be very promising.

Pulmonary metastases neutralization and tumor rejection by in vivo administration of beta glucan and bispecific antibody

Bispecific antibody (BsAb) with specificity for tumor cell surface antigen and the CD3 molecule on T cells can redirect activated T cells to lyse tumor cells. Since the ex vivo expansion and activation of T cells is impractical and ineffective for treating established tumors, we tested whether the immune stimulant beta glucan could in situ-activate T cells, which could secondarily be retargeted with BsAbs to lyse tumor cells. To test for tumor neutralization, C3H/HeN mice were injected i.v. with Cl-62 melanoma cells and immediately treated with i.p. beta glucan and/or anti-CD3 (500A2) x anti-p97 (96.5) F(ab')2 BsAb i.v. Pulmonary metastases were counted 14 days later. To test for tumor rejection and survival in a solid tumor model, mice were injected s.c. and i.p. with Cl-62 cells and 7 days later administered beta glucan i.p. and/or F(ab')2 BsAb i.v. In the neutralization model, there was a significant reduction in the number of metastases in the beta glucan + BsAb group, as compared with controls, and with beta glucan alone. In the established tumor model, beta glucan + BsAb reduced the incidence of s.c. tumors as compared with control, with BsAb alone and with beta glucan alone. It also prolonged survival of tumor-bearing mice compared with control, BsAb alone and beta glucan alone. We conclude that T cells can be activated in vivo by beta glucan and retargeted with F(ab')2 BsAb.

Binding characteristics and antitumor properties of 1A10 bispecific antibody recognizing gp40 and human transferrin receptor

The bispecific murine monoclonal antibody (MAb) 1A10 has specificity for the human transferrin receptor (TfR) and the human tumor-associated antigen gp40. This antibody, therefore, functions as an "antigen fork" by binding to two distinct antigens on the same malignant cell. Highly purified 1A10 inhibits the growth of cells coexpressing high levels of human TfR and the tumor-associated antigen gp40 by binding to both target antigens. In SW948 cells, the majority of 1A10 binding is via its gp40 specificity, and half-maximal inhibition of cell growth by 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay requires 20-30-micrograms/ml concentrations of 1A10. The binding of 1A10 correlates with growth inhibition in the cell lines HT-29, SK-OV-3, OVCAR-2, and OVCAR-3. The growth of OVCAR-10 cells, which express little gp40 and TfR, is not inhibited by 1A10. However, SK-BR-3 cells, which express abundant gp40 and extremely high levels of TfR, are insensitive to the effects of 1A10. In some cell lines, combined exposure to 1A10 and the iron chelator deferoxamine mesylate has synergistic antiproliferative effects. A single i.p. dose of 600 micrograms 1A10 is sufficient to achieve an estimated tumor concentration of at least 30 micrograms/ml for 7 days in C.B17/Icr-scid mice bearing SW948 human tumor xenografts. Treatment of scid mice bearing day 2 or day 4 SW948 xenografts with single or multiple 1A10 doses inhibits tumor growth in a dose-related fashion. Antitumor effects are not seen with therapy using either parental antibody of 1A10. The antiproliferative properties of 1A10 in tumor cells overexpressing gp40 and TfR suggest avenues for the development of new bispecific antibody-promoted treatment strategies.

Treatment of Hodgkin's disease with bispecific antibodies

Bispecific monoclonal antibodies (Bi-MAbs) with dual specificity for tumor-associated antigens (TAA) and a triggering molecule of an immunologic effector cell, respectively, open the possibility to specifically target to and activate cytotoxic effector cells (macrophages, T-cells, NK cells) at the tumor site. Using appropriately designed Bi-MAbs and unstimulated human NK cells and T-cells, respectively, we were able to cure SCID mice xenografted with human Hodgkin's tumors. This approach was also effective in disseminated tumors and when treatment was delayed until three weeks after the inoculation of the tumor, thus establishing this approach as the most effective model of an immunomodulating therapy of human neoplasms. Early observations with an ongoing phase I/II study with CD16/CD30 Bi-MAb in patients with refractory Hodgkin's disease confirm the expected low toxicity. If these observations can be confirmed in larger clinical studies, effector cell activating Bi-MAbs could become an important weapon in the remaining fight for the conquest of Hodgkin's disease.

Development and characterization of anti-renal cell carcinoma x antichelate bispecific monoclonal antibodies for two-phase targeting of renal cell carcinoma

To test a two-step approach for radioimmunotargeting of renal cell cancer, quadroma cells secreting antichelate x anti-renal cell carcinoma bispecific antibodies were obtained by somatic cell fusion. Five monoclonal antibodies against the chelate 1,4,7-triazaheptane-N,N',N"-pentaacetic acid (DTPA) were produced and characterized. Competitive binding assays indicated that the anti-DTPA antibodies reacted with DTPA chelated with indium, yttrium, chromium, iron, or zinc. The affinity constants of the anti-DTPA antibodies for 111In-DTPA ranged from 0.19 to 0.23 nM-1. Using different chelates, a remarkable chelate specificity of the anti-DTPA antibodies was demonstrated. The chelates recognized by the antibodies DTIn1, DTIn2, and DTIn4 share a N(N")-diacetic acid group, whereas the chelates recognized by DTIn3 share a N'-acetic acid group, suggesting the presence of different essential structures within the DTPA molecule that determine the reactivity of the antibodies. Five anti-DTPA antibody-producing hybridomas were used for somatic cell fusion with hybridoma G250 directed against renal cell carcinoma, resulting in three bispecific antibody-producing quadroma cell lines. The bispecific monoclonal antibodies were purified from ascites fluid using protein A affinity chromatography followed by hydroxylapatite chromatography and/or cation exchange chromatography. Of the total IgG amount present in the ascites fluid, 10-15% represented the bispecific antibodies. These bispecific antibodies will allow testing and optimization of a two-step approach for radioimmunotargeting of chelated radionuclides.

Prelabeling of chimeric monoclonal antibody L6 with 90yttrium- and 111indium-1,4,7,10-tetraazacyclododecane-N,N',N",N"'-tetraacetic acid (DOTA) chelates for radioimmunodiagnosis and therapy

90Y and 111In have been attached to chimeric monoclonal antibody L6 with a bifunctional chelating agent (DOTA-peptide isothiocyanate). The bifunctional chelating agent was prelabeled with either radiometal and then conjugated to the antibody. Studies in human patients showed excellent 111In single-photon emission computed tomography images of breast cancer lesions 24 h after injection.

Monocyte-mediated lysis of acute myeloid leukemia cells in the presence of the bispecific antibody 251 x 22 (anti-CD33 x anti-CD64)

Immunotherapy using bispecific antibodies (BsAb) to direct immune effector cells toward target tumor cells has been shown to be effective in a number of studies. Several immune trigger molecules have been characterized. Among them, FcgammaRI appears to play an important role in antibody-dependent cellular cytotoxicity. It is expressed mainly on monocytes, macrophages, and neutrophils under certain clinical situations. The expression of FcgammaRI can be regulated by a variety of cytokines, primarily by IFN-gamma. Recent studies have shown that granulocyte-colony-stimulating factor (G-CSF) and granulocyte-macrophage-colony stimulating factor (GM-CSF) can increase the number of the FcgammaRI-positive monocytes, increase the expression of FcgammaRI on circulating neutrophils after in vivo infusion, and greatly enhance the cytotoxic activity of circulating neutrophils. CD33 is a glycoprotein expressed on the cell surface of mature monocytes, myeloid progenitor cells, and myeloid leukemic blasts, but not on the earliest hematopoietic progenitor cells and other normal tissues. Herein, we report the construction of a BsAb, 251 x 22, by conjugating an anti-CD33 mAb (mAb 251) to an anti-FcgammaRI mAb (mAb 22). The BsAb 251 x 22 is capable of enhancing the cytotoxicity of several leukemia cell lines by cytokine-activated monocytes. Our data also show that G-CSF- and GM-CSF-stimulated monocytes can mediate cytotoxicity of target leukemia cells comparable to that of IFN-gamma-stimulated monocytes. The expression of FcgammaRI on monocytes after 24-h in vitro incubation with G-CSF and GM-CSF was increased, although not significantly. Prolonged incubation of monocytes with G-CSF for 48 h significantly increased the FcgammaRI expression. Because humanized anti-CD33 and anti-FcgammaRI mAb are available, and because GM-CSF and G-CSF have been used widely for patients after chemotherapy to stimulate the recovery of myeloid hematopoiesis, additional clinical development of this project is feasible. A BsAb comprised of humanized anti-CD33 and anti-FcgammaRI could have clinical application in the treatment of myeloid leukemia, especially in the management of minimal residual disease.

Performance of CD3xCD19 bispecific monoclonal antibodies in B cell malignancy

Bispecific monoclonal antibodies, with a dual specificity for tumor associated antigens on target cells and for surface markers on immune effector cells, have been shown (in vitro) to be effective in directing and triggering effector cells to kill target cells resulting in target cell lysis. Bispecific monoclonal antibodies (BsAb) against the CD3 antigen on T cells and the CD19 antigen on B cell were developed. Data obtained by in vitro experiments might indicate that clinical responses in BsAb immunotherapy, will only be obtained in patients with minimal tumor load, and may need additional T cell stimulation via cytokines such as IL-2. Although these experiments have shown us their limitations, they also include the promise of BsAb-directed immunotherapy in B cell malignancy as further demonstrated during a Phase I trail, showing little toxicity. Clearly, much remains to be done before this BsAb is routinely used for therapy, but, the results presented show that the CD3xCD19 BsAb has a potential as a therapeutic agent in B cell malignancy. This report describes the experiments performed to test a new immunotherapeutic approach for the treatment of B cell malignancy. Bispecific antibodies are described that can target cytotoxic T cells to tumor cells and elicit a cytolytic action towards these cancer cells.

Phase I trial of 2B1, a bispecific monoclonal antibody targeting c-erbB-2 and Fc gamma RIII

2B1 is a bispecific murine monoclonal antibody (BsMAb) with specificity for the c-erbB-2 and Fc gamma RIII extracellular domains. This BsMAb promotes the targeted lysis of malignant cells overexpressing the c-erbB-2 gene product of the HER2/neu proto-oncogene by human natural killer cells and mononuclear phagocytes expressing the Fc gamma RIII A isoform. In a Phase I clinical trial of 2B1, 15 patients with c-erbB-2-overexpressing tumors were treated with 1 h i.v. infusions of 2B1 on days 1, 4, 5, 6, 7, and 8 of a single course of treatment. Three patients were treated with daily doses of 1.0 mg/m2, while six patients each were treated with 2.5 mg/m2 and 5.0 mg/m2, respectively. The principal non-dose-limiting transient toxicities were fevers, rigors, nausea, vomiting, and leukopenia. Thrombocytopenia was dose limiting at the 5.0 mg/m2 dose level in two patients who had received extensive prior myelosuppressive chemotherapy. Murine antibody was detectable in serum following 2B1 administration, and its bispecific binding properties were retained. The pharmacokinetics of this murine antibody were variable and best described by nonlinear kinetics with an average t 1/2 of 20 h. Murine antibody bound extensively to all neutrophils and to a proportion of monocytes and lymphocytes. The initial 2B1 treatment induced more than 100-fold increases in circulating levels of tumor necrosis factor-alpha, interleukin 6, and interleukin 8 and lesser rises in granulocyte-monocyte colony-stimulating factor and IFN-gamma. Brisk human anti-mouse antibody responses were induced in 14 of 15 patients. Several minor clinical responses were observed, with reductions in the thickness of chest wall disease in one patient with disseminated breast cancer. Resolution of pleural effusions and ascites, respectively, were noted in two patients with metastatic colon cancer, and one of two liver metastases resolved in a patient with metastatic colon cancer. Treatment with 2B1 BsMAb has potent immunological consequences. The maximum tolerated dose and Phase II daily dose for patients with extensive prior myelosuppressive chemotherapy was 2.5 mg/m2. Continued dose escalation is required to identify the maximally tolerated dose for patients who have been less heavily pretreated.

Bispecific monoclonal antibodies for intravenous treatment of carcinoma patients: immunobiologic aspects

Immunobiologic parameters measured during a phase I trial of intravenously (i.v.) administered bispecific monoclonal antibodies (BsmAb) in renal cell carcinoma (RCC) patients are described. The BsmAb used, BIS-1, is reactive with a pancarcinoma-associated 38 kDa transmembrane glycoprotein, EGP-2, as well with the CD3 complex. Patients received during a 2 h i.v. infusion F(ab')2 fragments of BIS-1 at doses of 1, 3, or 5 micrograms/kg body weight during concomitantly applied subcutaneous (s.c.) IL-2 treatment. Acute but transient BIS-1 F(ab')2-related toxicity was observed at the 3 and 5 micrograms/kg dose level, and the maximum tolerated dose (MTD) was set at 5 micrograms/kg. A dose-dependent binding of BIS-1 F(ab')2 to circulating T lymphocytes was found. The in vivo occupancy of CD3 molecules on T lymphocytes was highest at teh end of the infusion period and then rapidly decreased, as shown by flow cytometry. A much slower decrease of BIS-1 F(ab')2 binding was observed in vitro, suggesting migration of BIS-1 F(ab')2-loaded T lymphocytes from the circulation. A strong but transitory leukopenia was observed, in which LFA-1 alpha bright, CD3/CD8 double positive T cells left the circulation preferentially. This phenomenon was most likely induced by elevated TNF-alpha and IFN-gamma plasma levels, which were at a maximum shortly after the end of the infusion. Isolated peripheral blood mononuclear cells obtained from patients directly after treatment with BIS-1 F(ab')2 at the 3 and 5 micrograms/kg dose level showed increased EGP-2-directed antitumor activity.

In vitro and in vivo stability and anti-tumour efficacy of an anti-EGFR/anti-CD3 F(ab')2 bispecific monoclonal antibody

The in vitro and in vivo stability and anti-tumour efficacy of the anti-EGFR/anti-CD3 bispecific monoclonal antibody (biMAb), M26.1, were analysed. The interaction of the intact biMAb with Fc receptor I (Fc gamma RI) present on human leucocytes was not observed when the antibody was used as an F(ab')2 fragment. A CD8+ T-cell clone coated with M26.1 F(ab')2 was as effective as the intact biMAb in inducing IGROV1 target cell lysis when tested in a 51Cr-release assay. Variable levels of reduction of F(ab')2 to monovalent F(ab') were observed upon incubation with human ovarian cancer ascitic fluid (OCAF) or with human glioblastoma cavity fluid (GCF), but not with mouse or human sera. Activated lymphocytes coated with F(ab')2 and incubated in vitro with GCF or OCAF for 24 and 48 h respectively maintained their targeting. Thus, the F(ab')2, when present as a soluble molecule, but not when bound to T cells, might lose some functional activity as a consequence of partial reduction to F(ab'). In normal mice, M26.1 F(ab')2 retained full cytotoxic activity in the circulation, and clearance values were similar to those obtained with parental and other MAb F(ab')2. Treatment of IGROV1 tumour-bearing mice with activated human lymphocytes coated with the M26.1 F(ab')2 significantly prolonged survival of the animals compared with tumour-bearing untreated and control mice treated with lymphocytes or F(ab')2 alone. Together, these results suggest the clinical usefulness of bispecific M26.1 F(ab')2 as a targeting agent for local treatment of tumours such as glioma and ovarian cancers that express variable levels of epidermal growth factor receptor (EGFR).

Clinical trials of bispecific antibody MDX-210 in women with advanced breast or ovarian cancer that overexpresses HER-2/neu

MDX-210 is a bispecific antibody (BsAb) that recognizes Fc gamma R1 on monocytes and macrophages and the cell surface product of the HER-2/neu oncogene, which is overexpressed on some breast and ovarian cancers. Clinical trials have demonstrated that treatment with MDX-210 is well tolerated and that MDX-210 is both immunologically and clinically active. Optimization of the dose and schedule of MDX-210 and development of combination treatments with cytokines that modulate immune effector cells will greatly enhance the efficacy of this novel BsAb construct for treatment of tumours that overexpress HER-2/neu. We envision that MDX-210 will be effective for treating patients with tumors that overexpress HER-2/neu, especially in the minimal disease setting.

Toward the production of bispecific antibody fragments for clinical applications

The clinical potential of bispecific antibodies (BsAb) has been hindered by the difficulty of obtaining clinical grade material, together with the immunogenicity of rodent-derived BsAb in patients. The supply issue is being directly addressed by recombinant methods for BsAb fragment production reviewed here. The immunogenicity issue will likely be overcome by the use of humanized or human antibodies. Currently, three technologies appear suitable for the production of BsAb fragments for clinical applications: BsF(ab')2 assembled from Fab' fragments expressed in Escherichia coli, BsF(ab')2 assembled using leucine zippers, and diabodies.

A human Fc gamma RI/CD64 transgenic model for in vivo analysis of (bispecific) antibody therapeutics

The human high-affinity IgG receptor, hFc gamma RI (CD64), is exclusively expressed on myeloid cells, where it serves an important role as a (cytotoxic) trigger molecule. To establish an in vivo model for analysis of the role of hFc gamma RI in immunity, we developed a novel transgenic mouse model. The human Fc gamma RIA gene, with endogenous regulatory sequences, was used to generate two lines of transgenic FVB/N mice. Immunohistochemical and flow cytometric studies showed that hFc gamma RI expression was restricted to myeloid cells. Monocytes, macrophages, and polymorphonuclear neutrophils (PMN) expressed physiologic hFc gamma RI levels, whereas lymphocytes and mast cells lacked expression. Human Fc gamma RI expression was regulated in vivo by the cytokines IFN-gamma (exactly as in humans) and IL-10. The transgenic receptor proved functional and bound human tumor cells via anti-hFc gamma RI-based bispecific antibodies. hFc gamma RI could, furthermore, be efficiently targeted in vivo by CD64 antibodies. These data demonstrate that the hFc gamma RI transgenic mouse model closely parallels the situation in humans. This mouse model seems useful for in vivo evaluation of the therapeutic potential of novel bispecific reagents in tumor and infection models.

Treatment of heterotransplanted Hodgkin's tumors in SCID mice by a combination of human NK or T cells and bispecific antibodies

To test the feasibility and efficacy of a new immunotherapeutic approach in Hodgkin's disease, bispecific monoclonal antibodies (BsmAb) were established with specificity for the Hodgkin's-associated CD30 antigen and for CD16 (on NK cells) or CD3 and CD28 (on T lymphocytes), respectively. These BsmAb induced a specific and efficient NK cell or T cell-mediated cytotoxicity in vitro. The treatment of severe combined immunodeficiency (SCID) mice with the NK (anti-CD16/CD30) or T cell (anti-CD3/CD30 and anti-CD28/CD30) activating BsmAb followed by administration of resting human lymphocytes led to complete remission of established heterotransplanted human Hodgkin's tumors. Even disseminated tumors were cured. Studies on the mechanism responsible for tumor destruction revealed that treatment efficacy depended on lymphocyte activation at the tumor site. Localization of human lymphocytes in mice was BsmAb mediated and antigen specific as activated lymphocytes were only detected in CD30+ tumors but not in CD30- colorectal carcinomas co-established as a control in the same animal.

Clinical experience with CD3 x CD19 bispecific antibodies in patients with B cell malignancies

In extensive preclinical testing, a CD3 x CD19 bispecific antibody (BsAb) induced killing of malignant B cells by resting T cells even in an autologous situation. In a 14 day clonogenic assay using a CD19+ pre-B cell line (REH), BsAb required repeated administration together with IL-2 to achieve a 5 log kill by resting peripheral blood T cells. Intravenously administered BsAb in an intrapatient dose escalation study of 3 patients with B cell non-Hodgkin's lymphoma showed limited toxicity (WHO grade II fever and chills) due to tumor necrosis factor-alpha (TNF-alpha) release by T cells. Pharmacokinetics with 2.5 mg BsAb showed peak levels of 200-300 micrograms/ml and a t1/2 of 10.5 h. The next patient, with chronic lymphocytic leukemia (CLL), received 0.6 mg BsAb/m2 as an i.v. infusion preceded by 1 MU IL-2/m2 s.c. Improved T cell activation was noted, as indicated by an increase in IFN-gamma, IL-6, IL-8, and IL-10, in addition to high TNF-alpha increases. TNF-alpha increases were highest on the first day. Toxicity remained restricted to grade II fever and chills, observed every day after the infusion of BsAb. No clear clinical effects were seen in this chemotherapy-resistant CLL patient with a high tumor burden. If subsequent patients also show limited toxicity, treatment of patients with a lower tumor load seems to be warranted to evaluate the efficacy of CD3 x CD19 BsAb therapy.

Clinical development of 2B1, a bispecific murine monoclonal antibody targeting c-erbB-2 and Fc gamma RIII

Bispecific monoclonal antibodies (BsmAb) can be used to specifically target tumor cells for cytotoxicity mediated by defined effector cells. One such BsmAb, 2B1, targets the extracellular domains of both the c-erbB-2 protein product of the HER-2/neu oncogene and Fc gamma RIII (CD16), the Fc gamma receptor expressed by human natural killer cells, neutrophils, and differentiated mononuclear phagocytes. 2B1 promotes the conjugation of cells expressing these target antigens. It efficiently promotes the specific lysis of tumor cells expressing c-erbB-2 by human NK cells and macrophages over a broad concentration range. 2B1 selectively targets c-erbB-2-positive human tumor xenografts growing in immunodeficient SCID mice. Treatment of such mice with 2B1 plus interleukin 2 (IL-2) inhibits the growth of early, established human tumor xenografts overexpressing c-erbB-2. A phase I clinical trial of 2B1 has been initiated to determine the toxicity profile and maximum tolerated dose (MTD) of this BsmAb and to examine the biodistribution of the antibody and the biologic effects of treatment. Preliminary results of this trial indicate that the dose-limiting toxicity for patients with extensive prior bone marrow-toxic therapy is thrombocytopenia for as yet undetermined reasons. Toxicities of fevers, rigors, and associated constitutional symptoms are explained, in part, by treatment-induced systemic expression of cytokines, such as tumor necrosis factor-alpha. Circulating, functional BsmAb is easily detectible in treatment patients' sera and exhibits complex elimination patterns. HAMA and anti-idiotypic treatment-induced antibodies are induced by 2B1 treatment. Some preliminary indications of clinical activity have been observed. BsmAb therapy targeting tumor antigens and Fc gamma RIII has potent immunologic effects. Future studies will include the development of more relevant animal models for BsmAb therapy targeting human Fc gamma RIII. The ongoing phase I trial will be completed to identify the MTD for patients without extensive prior bone marrow-toxic chemotherapy and radiation. A phase II clinical trial of 2B1 therapy in women with metastatic breast cancer is planned, as is a phase I trial incorporating treatment with both 2B1 and IL-2.