New Ca Vaccines May Be on the Horizon

PARIS, France--Cancer vaccines are now coming into their own,in advanced as well as early disease, Malcolm S. Mitchell, MD,said at the Fifth International Congress on Anti-Cancer Chemotherapy.Although vaccines have, to date, been most extensively investigatedin melanoma, the adenocarcinomas will soon be treated by thisapproach as well, said Dr. Mitchell, of the University of Californiaat San Diego.

The most exciting prospect, he believes, is the MUC-1 peptidebreast cancer vaccine, which will be started in clinical trialswithin the next few months. This vaccine is based on antigensto the mucin core protein, which breast cancer shares in commonwith gastrointestinal, ovarian, and lung cancer.

Another extremely fertile area for vaccine development, he predicted,is in such chemotherapy-resistant squamous cell carcinomas aslung cancer and head and neck cancer.

In Los Angeles and now in La Jolla, Dr. Mitchell's team has usedvaccines to treat 154 patients with metastatic melanoma in thelast 10 years. Although, at first glance, the 20% response ratemay appear discouraging, he said, some 10% of patients do survivefor at least 3 to 8 years with monthly maintenance vaccine injections.

"Tumor-induced immunosuppression is a reality," Dr.Mitchell said, noting that transforming growth factor-beta (TGF-beta),interleukin-10, and GM-CSF are all immunosuppressive moleculessecreted by tumor cells. To avoid this tumor-related immunosuppression,he urged that active immunotherapy be started at the earlieststages of disease, when the tumor burden is small.

The optimal tumor antigens, Dr. Mitchell said, are those thatmaximize the responses of both cytotoxic and helper T cells. "CD4+helper cells may actually control the major mechanisms by whichtumor cells are rejected in vivo, through the cytokines they produceand the other leukocytes they evoke," he said.

Also desirable for subunit vaccines, he pointed out, is that theantigens be defined by human T cells rather than by mouse monoclonalantibodies. "It's more important to know what a human seesas being an antigen than what a mouse sees on a human cell,"he stressed.

According to Dr. Mitchell, numerous strategies have been developedto increase tumor immunogenicity so that the patient's own immuneresponse can be harnessed to fight the tumor. As examples, hecited the use of cell lines that are by definition allogeneic,and the use of the HLA-B7 MHC antigen to make the tumor allogeneic.

Gene Insertion Boosts T Cells

A promising new approach, he noted, is to insert genes for co-stimulatormolecules such as B7, which helps macro-phages interact with Tcells. Inserting this gene into the tumor cells increases theproliferation and cytotoxicity of T cells, not only against theB7-transduced tumor cells but also against the parental tumorcells.

"The importance of this," he explained, "is that,if one simply gives these T cells to the patient, or generatesthem within the patient, the patient's own tumor will be rejectedwithout the necessity of putting B7 co-stimulator genes into everyone of the tumor cells."

While B7 must be in tumor cells to improve their effectivenessin vaccines, he said, injecting cytokines or complex adjuvantsmay well be more effective than simply inserting genes for thesecytokines into tumor cells. This is because gene insertion willgenerate cytokines only in the local environment.

It is not sufficient to immunize the patient, Dr. Mitchell said,underscoring the necessity of making the patient's tumor a bettertarget as well. Upregulation of HLA and tumor antigens, such aswith interferon-alfa or interferon-gamma, is key to ensuring thatthe antitumor T cells find a recognizable tumor target, he explained.Eight of 18 patients who failed to respond to vaccine alone didrespond when interferon-alfa was given immediately afterwards,he said.

As further evidence, he noted that melanoma patients with certainHLA class I alleles, such as HLA-A2, B44, and C3, responded betterto his melanoma vaccine than did patients without those phenotypes.

"Patients with two or more of these alleles had a 40% chanceof responding, as compared with 0 to 7% for patients who lackedall of these alleles," he said. "We think the reasonis that these alleles are very important in presenting antigenson the melanoma cell to T cells, and that patients who have thebest presentation will show the best response."