By Dean A. Haycock

Special to BioWorld Today

Medical scientists have been thinking about cancer vaccines for more than 100 years, ever since New York surgeon William Coley noticed that some tumors waned after cancer patients developed bacterial infections. Thus inspired, Coley, in the early 1890s, injected bacteria directly into the tumors of some of his patients. Some of the tumors shrank.

Medicine has not forgotten Coley¿s initial foray into the field of immunotherapy. Research has continued at varying levels of intensity ever since. Today, while the FDA has yet to approve a cancer vaccine, several are making their way through the clinical trial process.

The basic concept behind these potential therapies is easy to understand. The immune system attacks foreign substances, antigens. Proteins or other molecules on invading bacteria, for example, serve as antigens that stimulate a counterattack by the body¿s immune system. If cancer cells can be made to look like foreign invaders, the immune system may be induced to attack them. To encourage such immune responses, scientists are injecting subjects with antigens obtained from cancer cells. Among these are carbohydrate antigens.

Researchers at Memorial Sloan-Kettering Cancer Center in New York now are testing several carbohydrate-based antigens against prostate, breast and ovarian cancers. Carbohydrate antigens were known to induce antibodies, but the exact degree of the response had never been measured. The scientists needed a way to isolate the antibodies elicited by their antigens so they could characterize them.

Antibodies Directed To Tumor Antigens

Samuel Danishefsky, of the center¿s Laboratory of Bio-organic Chemistry, and his co-authors describe how they solved this problem in an article titled ¿Polyclonal antibodies from patients immunized with a globo H-keyhole limpet hemocyanin vaccine: Isolation, quantification, and characterization of immune responses by using totally synthetic immobilized tumor antigens.¿ The article appears in today¿s Proceedings of the National Academy of Sciences, dated March 14, 2000.

The work entailed the preparation of an affinity column that can separate antibodies from other components of the blood of patients immunized with the cancer vaccine. ¿It is an unusual program,¿ Danishefsky said, ¿in that it is an unusual blending of front-line skills of medicine, chemical synthesis and immunology.¿ The approach involves passing sera over a column that contains permanently bound antigens. Antibodies in the sera that recognize the bound antibodies bind to them and are delayed in passing through the column. They can be collected later as they are eluted together.

¿Now,¿ Danishefsky told BioWorld Today, ¿I¿d say for the first time, we are able to readily isolate antibodies that come out of an immunization and are directed to carbohydrate-based antigens.¿

The column recognizes antibodies to a carbohydrate antigen called globo H which was conjugated to a protein called keyhole limpet hemocyanin (KLH). The antigen, as this description suggests, is a complex synthetic molecule. This appears to be the first time that such a fully synthetic antigen with such a degree of complexity has been adapted for use on a column capable of isolating antibodies, according to Danishefsky.

¿There are basically three sources of antigen that goes into the vaccination,¿ he said. ¿There is the antigen that goes into the vaccination. There is the important one on the surface of the cancer cell. And then there is the question, Can you reduce that antigen to an affinity column.¿ This paper shows that you can. That means we had to take the same epitope that went into the total synthesis that led to the vaccine and discontinue part of the synthesis, or divert part of that synthesis, to allow us to attach the same substance to an affinity column. That was not easy. Remember that the synthesis of the substance that was used to conjugate the carrier protein is of the order of 30 steps. So now having done that, you have to learn how to do the linking. And then you are taking a risk. That means in putting it on to the solid support, you might have lost or modified the structures. It is now in a setting very different from the setting of vaccination. But still, it works.¿

The successful preparation of such a column allowed the researchers to obtain quantitative data concerning the antibody response they elicited in their patients. With no previous data from cancer patients available for comparison, the researchers compared the antibody levels they observed with those reported in studies involving bacterial polysaccharide vaccines. They found that vaccination with globo H-KLH antigen yielded antibody levels that were often greater than those elicited by bacterial polysaccharides

The purifying column also allowed the researchers to determine the type of antibodies produced in the patients. They found significant amounts of both IgG and IgM antibodies. ¿We were really very pleasantly surprised at the high enzymatic titer at the expense of what is called class switch,¿ meaning IgE representation,¿ Danishefsky said. ¿Those are more sophisticated antibodies. No one could have predicted there would be that good of an IgG response.¿

These antibodies may offer promising clinical applications.

¿The major part of the experiment,¿ Danishefsky explained, ¿was to actually isolate the antibody from each patient. In principal, you could now modify those antibodies and use them in what is called a passive form of immunity. Active immunity means you inject antigen. Passive immunity means you inject antibody, which is either connected to a drug or something else.¿

Multivalent Vaccine Development Under Way

Danishefsky indicated he and his colleagues had already applied the approach to other vaccines the team is working on. This will be useful as they develop what is called a multivalent vaccine. Such a vaccine would include multiple antigens. This would be expected to reduce the chances of a cancer cell not expressing any particular antigen.

¿Hopefully upon immunization, a whole range of antibodies will be expressed,¿ Danishefsky told BioWorld Today. ¿The affinity columns will be used to distinguish between the various antibodies that have been generated. So, the chances that a [cancer] cell would not have any of these antigens is correspondingly reduced because you have a lot more antigen.¿