The mouse cells contain retroviruses that produce a viral enzyme. Genes have been added to and removed from the retroviruses so they cannot replicate on their own, but they can still break out of the mouse cells and invade nearby dividing cancer cells. Once inside the cancer cells, the retroviruses produce an enzyme that will turn an otherwise harmless drug into one that is toxic to the tumor cells.
"The strategy is to use a retrovirus that will only infect dividing
cells," says Dr. Mark Israel, Director of UCSF's Brain Tumor Research
Center and a Research Associate in the CRI. "Brain cells don't divide,
so a brain tumor occurs in an environment where there are no dividing cells
except the tumor cells. The retrovirus gets into these dividing tumor cells,
establishes itself, and begins to express this enzyme. When the patient
is treated with the drug gancyclovir, the enzyme converts it into something
that is toxic to the tumor cells, and those cells die."The treatment protocol is being done in conjunction with a number of other institutions around the country, says Israel. Led by UCSFneurosurgeons Michael Prados and Michael McDermott, the treatment is only being used for patients with recurrent glioblastoma for which there is no other available effective therapy. Patients with recurrent glioblastoma almost always die from the disease.
"We have treated two patients here and two patients have been treated elsewhere," says Israel. "It's too early to assess whether the therapy is going to be useful. We're really at the very beginning of all this. But it could open up a wide range of potential treatments through gene therapy."
Israel expects that this protocol will be completed within twelve to eighteen months. He and his colleagues hope to build upon it with another generation of protocols.
"While we're hopeful that there will be a significant therapeutic effect from this, the reality of cancer treatment is that advances are made in small increments," he says. "Our goal, in addition to helping these particular patients, is to learn something that will help us to enhance the therapy, make it more efficacious. We are trying to do something that impacts on brain tumors and, at the same time, asks fairly fundamental biological questions."
While this particular gene therapy is applicable only to tumors in the brain, Israel sees potential in gene therapy for other forms of cancer.
"With some other forms of cancer, we won't have the therapeutic opportunity that is offered by the ability to transmit the gene only to dividing cells," he says. "If you put this into a liver tumor, for example, there could be regenerating liver cells adjacent to the tumor. But there are other strategies in which, even though the virus would affect every cell type, it would only be expressed in certain cells, namely tumor cells. So there are ways to get specific targeting to the cell. The key areas for advancement that are needed are better mechanisms for delivering gene therapy and strategies for more specific expression of the transduced gene. Gene delivery and gene targeting are the key themes for the future."
While Israel and his fellow researchers are exploring ways to increase the efficacy of this gene therapy, they are intrigued by an aspect of the treatment that may have important consequences.
"It turns out that this therapy has a very interesting twist to it," he says. "Our current model for the treatment of cancer is that, to cure cancer, you have to get every single-or close to every single-cell. So one would think that, to be curative, a therapy like this would require that the virus infect every single cell. But we know from animal experiments that we probably don't infect more than one cell in ten with this therapy. Yet animals can be cured by this approach. So there's a conundrum. How is it that only one tenth of the cells get infected by the gene and yet the animals are cured?"
The reason, says Israel, has to do with something called the bystander effect.
"Somehow or other those cells into which the virus is transmitted and which go on to die, also seem to have an influence on surrounding cells," he says. "That influence is a therapeutic influence in that the cells around them also die. That is called the bystander effect.
"Many mechanisms for the bystander effect have been put forth in the medical literature. Some obvious ones are that the dying cells elicit an immune response from the host that then goes on to kill the residual cells. Another is that you kill a certain number of tumor vessels and, in doing that, the nutritional supply to the residual tumor is compromised."
But the bystander effect has not been observed in other treatment settings, and that, says Israel, is one of the mysteries that needs to be resolved.
"Why don't we see it in other settings if blood vessels and the immune response are the key things? There is reason to think that these kinds of mechanisms contribute to the bystander effect. But it is also true that in vitro-in a test tube or in a tissue culture dish, where there are no blood vessels and there is no immune system-you can observe that, even if only one-tenth of the cells are infected by the retrovirus, the entire tissue culture dish of cells dies."
Israel and his colleagues believe that the bystander effect is, in fact, the primary feature of this gene therapy that allows it to work.
"We have been studying the cellular mechanisms involved," he says, "and Dr. James Fick, a neurosurgeon working in our laboratory, has made important headway in understanding the exact mechanisms by which the gene-transduced cells are able to kill their neighbors. That work has identified certain cellular structures that are able to transmit signals to neighboring cells to die. The basic mechanism involves the transfer of molecules from one cell to the next. The molecular details of that are what will allow us to build a better gene therapy.
"We're just getting this report ready for publication," he says. "Our hope is to be able to use that information to engineer an enhanced bystander effect into this therapy."