Research that directs cancer-killing immune cells against the deadliest brain tumors shows promise

Jul 15 2004

Researchers seeking to direct cancer-killing immune cells against the deadliest brain tumors have three new targets that show promise in laboratory studies and in a Phase I patient trial, according to two articles in the July 15 issue of the journal Cancer Research.

The antigens, previously associated with several other types of cancer cells, were recently found to be expressed in the most common and aggressive type of malignant brain tumor, glioblastoma multiforme (GBM). Scientists at Cedars-Sinai's Maxine Dunitz Neurosurgical Institute and the National Cancer Institute have generated cytotoxic T lymphocyte clones (cancer-killing immune cells) that recognize GBM cells expressing these antigens.

"In a Phase I clinical trial of 14 patients, we found that our dendritic cell vaccine not only generated an immune response against these antigens but it appeared to play a significant role in prolonging survival in patients with glioblastoma," said Keith L. Black, MD, director of the Institute, the Division of Neurosurgery and the Comprehensive Brain Tumor Program at Cedars-Sinai.

The median length of survival of patients with recurrent glioblastoma whose treatment included the vaccine was 133 weeks – about two and a half years. A similar group of patients receiving the same level of care but not the vaccine had a median survival of only 30 weeks – seven and a half months.

John S. Yu, MD, senior author of the articles and co-director of the Comprehensive Brain Tumor Program, said these findings represent a significant advance in the field of brain tumor immunotherapy.

"This is the first time that a specific response to brain tumor antigens has been demonstrated as the result of an immunotherapy strategy," he said. "These antigens give us specific targets to aim for and they give us potent tools with which to measure immune responses. Therefore, we have a better way of monitoring the progress of patients who undergo vaccination and we have a means of improving these therapies."

In recent years, scientists have identified several tumor-specific antigens that appear to play a role in the development of certain cancer cells. The body's natural defensive cells, T lymphocytes, have the capacity to attack "foreign" proteins, but cancer cells and the antigens they express typically evade recognition by the immune system. Therefore, cancer researchers search for new antigens that may serve as targets, devise new methods to make the targets "visible" and vulnerable to immune cells, and seek new ways to multiply the number of cancer-killing cells responding to the threat.

"These three antigens – HER2, gp100, and MAGE-1 – have been described since the 1980s but we have only recently found them to be expressed in glioblastoma cells," said Dr. Black.

HER-2 is expressed in a variety of normal tissues, but it is selectively overexpressed in a number of malignancies, including breast and ovarian tumors. Glycoprotein 100 (gp100) is an antigen linked to melanoma. MAGE-1, initially analyzed from melanomas and found to be expressed in a variety of tumor types, became the first identified tumor antigen recognized by the immune system's protective T cells.

In earlier studies at other centers, a cancer vaccine combining MAGE-1 cells with specially cultured immune system cells was able to produce a tumor-specific immune response among patients with melanoma. Clinical trials using gp100 as a target in melanoma and HER-2 as a target in several types of cancers also demonstrated that the antigens elicit a strong immune response that continues even after the vaccinations have ended.

Since then, researchers have worked to develop a number of therapies that may be used individually or in combination to target malignant brain tumors. Dendritic cell immunotherapy is intended to stimulate a patient's immune system to recognize and attack glioblastoma cells. Tumor cells that have been removed during surgery are cultured in the laboratory with dendritic cells, also called "antigen-presenting" cells, taken from the patient's blood. The resulting cells are injected back into the patient, where they are designed to identify brain tumor cells as invaders and stimulate a strong response from tumor-infiltrating T lymphocytes.

In an earlier Phase I trial, tumor cells were grown in culture, and proteins from the cell surfaces were used in preparing the vaccine. In the Phase I trial described in the Cancer Research article, this process was refined.

"Now we take the proteins directly from the surgical specimen, which ensures that we are getting the relevant proteins and not antigens or proteins that are artifacts of the culturing process. This also avoids the technical problems of trying to grow out tumor cells that have been irradiated and undergone chemotherapy," Dr. Yu said. "In addition, instead of just drawing blood to obtain a patient's dendritic cells, we're using a process that allows us to get 25-fold more dendritic cells. This may account for the dramatic prolongation of survival that we see compared to our control patients."

Dr. Black said a larger Phase II trial of the dendritic cell vaccine is now being completed and the researchers are preparing to apply for a randomized Phase III trial.

In the lab studies described in Cancer Research, seven established GBM tumor cell lines and cells from 43 GBM tumors removed from patients at Cedars-Sinai were analyzed and compared with normal brain tissue. MAGE-1 was not detected in normal tissue. Although HER-2 and gp100 were detected in normal tissue, this does not preclude their potential usefulness. In previous studies with different tumor types, HER-2 and gp100 were overexpressed in cancer cells, while in normal cells they were expressed at levels below the "threshold" for activation by the immune system.

Furthermore, because healthy neural cells are of a histocompatibility type that does not interact with cytotoxic T cells, the immune system will not launch an attack on normal tissue expressing the antigens. In contrast, however, the surfaces of GBM tumor cells were found to express Major Histocompatibility Complex (MHC) Class I molecules, the type that makes them vulnerable to interaction with cytotoxic T lymphocytes.

The researchers documented that HER-2, gp100 and MAGE-1 are frequently expressed in glioblastoma cells. They generated cytotoxic T lymphocyte clones specific to each antigen and cultured them with GBM cells. The brain tumor cells were able to naturally process the antigens, and the lymphocytes were able to process the antigen-derived peptides or proteins on the surfaces of the GBM cells.

http://www.csmc.edu/