Lloyd Old, MD

A major objective of the Laboratory of Human Cancer Immunology, which became the New York branch of the Ludwig Institute for Cancer Research in 1991, is the identification of suitable antigenic targets for antibody-based and vaccine-based immunotherapies of human cancer.

Over the past decade, a number of cell-surface antigens, expressed by human cancer and identified by mouse monoclonal antibodies, have been selected for detailed serological, biochemical, and genetic characterization. These include A33, a 43k glycoprotein with selective expression in normal and malignant epithelium of the gastrointestinal tract; G250, a glycoprotein expressed by a high percentage of renal cancers; LewisY (LeY), an oligosaccharide epitope expressed on glycolipids and glycoproteins by a wide range of epithelial cancers; GD3, a ganglioside with high expression in melanoma and other neuroectodermal tumors; FAP-alpha, a 95 k glycoprotein strongly expressed in the stromal fibroblasts of epithelial cancers; and the truncated EGF receptor, a 140 k form of the EGF receptor (deleted in exons 2-7), which is expressed by a proportion of brain cancers and other tumor types.

Clinical trials of mouse monoclonal antibodies against A33, G250, and FAP-alpha have provided baseline information about the pharmacokinetics, biodistribution, and imaging characteristics of these antibodies, and 2 of the antibodies (A33 and G250) have been evaluated in Phase I/II trials of 131Iodine-based radioimmunotherapy. The mouse monoclonal antibodies against LeY and GD3 have strong biological effector functions (complement-dependent cytotoxicity and antibody-dependent cell-mediated cytotoxicity). These antibodies have also been evaluated in Phase I/II trials.

Because mouse antibodies are strongly immunogenic and elicit a humoral immune response in patients, the antibodies can be given for only a limited period of time, and this precludes assessment of their therapeutic value. For this reason, the antibodies have been genetically modified to provide chimeric (G250, GD3) or humanized (A33, LeY, F19) constructs. The A33, G250, and GD3 constructs are undergoing initial clinical evaluation, and the LeY and FAP-alpha constructs have been prepared for clinical trials.

The protocols developed for the clinical study of this new generation of antibody constructs are designed to provide the following information: pharmacokinetic and quantitative biodistribution properties, tumor localization and imaging characteristics, immunogenicity, and therapeutic activity of antibody alone or antibody as a carrier of cytotoxic agents.

With regard to antigenic targets for cancer vaccine development, these have been identified in several ways: antigens identified by antibodies, e.g., GM2 and other carbohydrate antigens; antigens coded for by mutated or amplified oncogenes or suppressor genes, e.g., p53; and T cell-recognized antigens, e.g., MAGE, BAGE, GAGE, discovered by the methodology pioneered by the Brussels branch of the Ludwig Institute.

A powerful new approach to identify immunogenic cancer antigens has been developed by M. Pfreundschuh and his group at the University of the Saarland, Homburg, Germany. The method is called SEREX (for SERological screening of cDNA EXpression libraries prepared from human tumors). The library is probed with autologous serum from the tumor donor, and reactive clones are isolated, sequenced, and catalogued as known or unknown gene products.

Initial characterization of interesting clones include expression pattern in normal tissues, frequency of expression in different types of human cancer, and frequency of antibody responses in cancer patients versus normal individuals. MAGE, tyrosinase, and mutated p53 products have been identified by SEREX analysis, indicating that this approach has general utility for identifying T cell-recognized tumor antigens as well as gene products related to malignant transformation.

A collaborative effort involving M. Pfreundschuh and his group and investigators associated with the New York University College, London and Melbourne branches of the Ludwig Institute, and the Aichi Cancer Center, Nagoya, has been established. A range of human cancer types are being subjected to SEREX analysis, and a growing list of human cancer antigens has been identified, including a novel member of the growing family of cancer-testis (CT) antigens designated NY-ESO-1.

Like other CT antigens -- such as MAGE, BAGE, GAGE, and SSX2 -- NY-ESO-1 is strongly expressed in the testis and a wide range of human cancers and is coded for by a gene on the X chromosome. NY-ESO-1 has been expressed in recombinant form, and the recombinant protein is being used to develop mouse monoclonal antibodies and serological methods for screening human serum for NY-ESO-1 antibodies. The gene for NY-ESO-1 maps to Xq28 and codes for an 18-kD protein. NY-ESO-1 mRNA expression is found in 20 to 30 percent of melanomas, lung, breast, ovarian, bladder cancers, and other tumor types but, like other CT antigens, rarely in colon or renal cancer.

In a survey of sera from normal individuals and cancer patients, antibodies to NY-ESO-1 were found in 40 to 50 percent of patients with advanced NY-ESO-1-expressing tumors. One patient with high-titered NY-ESO-1 antibodies was found to have HLA-A2-restricted cytotoxic T lymphocytes against autologous NY-ESO-1-expressing melanoma cells, and 3 HLA-A2-restricted NY-ESO-1 peptides were identified as the target epitopes recognized by these cytotoxic T lymphocytes.

We have analyzed CD8+ T cell responses by enzyme-linked immunospot (ELISPOT), cytotoxicity, tetramer assays, and the humoral immune responses by ELISA and Western blots. Our findings show that NY-ESO-1 elicits a strong, integrated humoral and cellular immune response in a high proportion of patients with NY-ESO-1-expressing tumors. NY-ESO-1 represents a new target for vaccine development, and ESO-1 protein, DNA, and viral vector vaccines are being constructed.

A Glycoconjugate Vaccine Group has also been established to identify and chemically synthesize glycolipid (carbohydrate) antigens for human cancer vaccines. This group is made up of laboratory and clinical investigators from several Ludwig Institute branches and affiliates, including R.R. Schmidt of the University of Konstanz, and will be responsible for carrying out trials using synthetic carbohydrate vaccines. The first of these vaccines, a synthetic GM2-conjugate vaccine, entered clinical trials in 1998.

In addition to antibody-based and vaccine-based cancer therapies, nonspecific immunotherapies with microbial products and cytokines continue to be pursued. The prototype agent for this approach is Coley's toxin, a mixture of heat-killed gram-negative and gram-positive bacteria, which was first used in the treatment of cancer over a century ago. The widely used treatment of superficial bladder cancer by local installation of BCG represents a modern version of Coley's approach.

There is a widely held presumption that the beneficial effects of these microbial products for cancer are mediated by cytokines, and we are focusing our attention on 4 cytokines -- TNF, IFN-gamma, GM-CSF, and IL-12. With regard to TNF, a TNF-deficient mouse has been developed, and its response to inflammatory, antigenic, and infectious challenges has been characterized. Studies are now underway to investigate the role of TNF in tumorigenesis.

A similar analysis of GM-CSF-deficient mice is being carried out and has revealed an unexpected role of GM-CSF in Tcell functions. In studies with Robert Schreiber and his colleagues at Washington University, IFN-gamma has been found to have a major role in tumor recognition and rejection. This conclusion comes from 3 lines of evidence in experimental systems: (1) tumor cells lacking IFN-gamma responses have reduced immunogenicity and become resistant to rejection; (2) IFN-gamma receptor-deficient mice are far more susceptible to tumor induction by the chemical carcinogen 3-methylcholanthrene; and (3) IL-12, a potent inducer of IFN-gamma, has a strong inhibitory effect on transplanted tumors and on tumor development induced by 3-methylcholanthrene.

Modulating the IFN-gamma system with IL-12 and other IFN-gamma inducers provides powerful new approaches to investigate the role of innate and adaptive immunity in cancer development and in maximizing the effects of specific immunotherapy with antibodies and vaccines.