Immunotherapy in Brain Cancer: Progress Continues, but Questions Remain

January 7, 2016
Andrew J. Roth

To gain insight into the research needed in glioblastoma, we interviewed Robert M. Prins, an associate professor in residence at UCLA.

In a recently published, preclinical study, a team from the UCLA Jonsson Comprehensive Cancer Center examined the combination of chemotherapy and genetically modified immune cells for the treatment of glioblastoma. This research was published in September 2015 in the journal Neuro-Oncology.

To gain insight into this work and other needed research in glioblastoma, CURE interviewed Robert M. Prins, a lead investigator on the study and associate professor in residence at UCLA.

Could you provide me with an overview of the work you're doing at UCLA?

What we're trying to do is develop immune-based approaches for the treatment of malignant gliomas. One approach that we've used is an adoptive transfer of engineered T cells — we're taking normal T cells and engineering them with a gene therapy vector to express a new T cell receptor, which changes their specificity. In this case, we changed them to be specific to a cancer testis antigen called NY-ESO-1. The reason we did that was based on some early work.

Taking a step back, one of the issues with glioblastoma is that there aren't any well-known targets that we know about for the immune system. Other cancers, like melanoma, have well-known targets but gliomas and glioblastomas do not.

What we found three or four years ago is that if you treated glioblastoma cells with a demethylating agent, decitabine, it upregulated a whole panel of these cancer testis antigens. One of them in particular that was highly upregulated was this gene called NY-ESO-1.

In this new study we published in September in Neuro-Oncology, we tested it out in preclinical treatment models. These are immuno-deficient mice that have human tumors in their brains and they were treated with a chemotherapy drug called decitabine. We showed that it upregulated NY-ESO-1 in the tumor and T cells specific for NY-ESO-1 that we engineered could treat mice bearing these tumors.

The results were really astounding, but we're going to have to confirm that this can be generalized to a whole series of human glioblastomas.

Why is this exciting — both for the medical community but also patients and survivors of glioblastoma?

It's generated a lot of enthusiasm because it's an inducible antigen that's well recognized by the immune system. By giving a simple demethylating agent, it upregulates a well-known target and then we basically adopted an engineered T cell from the surgery branch at the National Institutes of Health and could target it. We know how to do adoptive transfer immunotherapy, but in the past, we just haven't had a good antigen to target.

We have viral vectors that can be used for human patients. These are human T cells in human tumors in immuno-deficient mice. If we can prove it's generalizable, the next step would be to move it into human patients.For patients, it's a relatively straightforward approach. The decitabine that we used is already approved by the FDA for multiple myeloma and myelodysplastic syndrome, so it's a drug that's been used in patients and is available. The engineered T cell receptor that we put into the viral vector has actually already been used in humans as well. Everything we did was in human cells injected into immuno-deficient mice.

Is there anything else you'd like to mention about this study?

One of the questions in the field about adoptive transfer has been about how to deliver the T cells. In our work, we compared different routes of T cell transfer. We compared intratumoral injection of T cells with systemic adoptive transfer into the vein and found that intravenous injection was much more efficacious. Because the T cells expand in the blood and traffic better, they found their ways to isolated pockets of tumors more effectively.

Why are this and other novel strategies worth pursuing in glioblastoma?

Glioblastoma is essentially incurable with standard therapies. The standard treatment involves surgical resection to remove the main tumor mass followed by radiation and chemotherapy. However, the prognosis for these patients is still 12 to 18 months, and this hasn't changed in 30 years or so.

It's such an invasive tumor that the surgeons can only remove as much as they can see. Unfortunately, during surgery, invasive tumor cells have already spread in the brain. These are the cells that come back to cause recurrence. The concomitant treatment with radiation and chemotherapy certainly helps, but recurrence is almost universal within a couple years.

Are immunotherapies particularly difficult to develop in brain cancer? What are some of the challenges in this area?

For many years, many people did not think brain tumors would be amenable to immune-based therapies because the brain is immune-privileged. Now, we realize that the brain isn't necessarily an immune-privileged organ, but rather it just has a different type of microenvironment. Lymphocytes can get into the brain, they can get into tumors in the brain, and a lot of recent preclinical studies have changed our way of thinking. Immunotherapy is now being thought of much more positively, but we're still trying to figure out the best target and the best method for immune-based approaches.

What research still needs to be done?

Glioblastomas are not inherently immunogenic like melanoma or kidney cancer. There are not a lot of T cells normally. One of the challenges has been to induce immune responses to malignant gliomas in the brain. There are several groups around the country that have been working on this.

Even in our work, we still think there are improvements that can be made. We're not sure if anything we're doing right now is going to be the whole cure.