The fundamental problem of dealing with cancer is that cancer cells are so like normal cells that it is very hard to find differences that can be a basis for treatment without severe side effects. The key to successful cancer immunotherapy will be based on a very careful choice of cancer targets that are sufficiently cancer specific not to cause serious side effects. There are two fundamentally different ways to deploy the immune system for such cancer treatments. One is to increase the efficacy of the cancer patient's own immune system so that it attacks these differences. This has been achieved by "checkpoint blocking" which is very successful but only with a relatively small proportion of cancers. Secondly, one can produce antibodies, or T cells, whose specificity is directed against proteins expressed differentially in cancers. CART cell treatments have proved very effective for some blood cancers but not so far for common solid tumours. Humanised, unmodified monoclonal antibodies have been used extensively for the treatment of certain adenocarcinomas with modest success. However, using antibodies together with the body's own immune system to treat cancers by engineering monoclonal antibodies that are directed at both a target antigen on the cancer cell surface and also against T cells shows promise for the development of novel immunotherapies. Genes can be found which are expressed highly in some cancers but with a low or absent expression on normal tissues and so are good novel targets. It is so far, only immune-based killing that can kill bystander target negative cells, which is essential for successful treatment since hardly ever will all the cells in a cancer express any desired target. We conclude that, while there still may be many hurdles in the way, engineered bispecific T cell attracting monoclonal antibody-mediated killing of cancer cells may be the most promising approach for achieving novel effective cancer immunotherapies.
Keywords: CAR T cells; T cell receptor; bispecific antibody; immunotherapy.