The immune system is a powerful tool for the selective destruction of unwanted cells, and researchers are a fair way down the road of engineering the activity of the immune system to form therapies. You might look at granulocyte transplant therapy as an example of the sort of tools that are under development. Here is an article on another line of research that has just reached the stage of early tests in humans:

In the research published Wednesday, doctors at the University of Pennsylvania say the treatment made the most common type of leukemia completely disappear in two of the patients and reduced it by 70 percent in the third. In each of the patients as much as five pounds of cancerous tissue completely melted away in a few weeks, and a year later it is still gone.

...

the researchers removed certain types of white blood cells that the body uses to fight disease from the patients. Using a modified, harmless version of HIV, the virus that causes AIDS, they inserted a series of genes into the white blood cells. These were designed to make to cells target and kill the cancer cells. After growing a large batch of the genetically engineered white blood cells, the doctors injected them back into the patients. In similar past experimental treatments for several types of cancer the re-injected white cells killed a few cancer cells and then died out. But the Penn researchers inserted a gene that made the white blood cells multiply by a thousand fold inside the body. The result, as researcher June put it, is that the white blood cells became "serial killers" relentlessly tracking down and killing the cancer cells in the blood, bone marrow and lymph tissue.

An editorial and research paper are available if you are interested in delving further into the details. Unfortunately this work suffers from much the same problem as efforts to develop granulocyte transplant therapies, which is that there are next to no sources of funding for research groups at the cutting edge of immunotherapy. The article relates what is a sadly common story in this part of the scientific community:

So why has this remarkable treatment been tried so far on only three patients? Both the National Cancer Institute and several pharmaceutical companies declined to pay for the research. Neither applicants nor funders discuss the reasons an application is turned down. But good guesses are the general shortage of funds and the concept tried in this experiment was too novel and, thus, too risky for consideration. The researchers did manage to get a grant from the Alliance for Cancer Gene Therapy, a charity founded by Barbara and Edward Netter after their daughter-in-law died of cancer. The money was enough to finance the trials on the first three patients.

This is a good example of how philanthropy modeled on venture investment - backing a range of early stage, high risk, high reward projects - can help break up the log-jams that result from institutional reluctance to fund the cutting edge in any field. The larger an institution, the more they will tend towards only backing the safe choices, but by doing that they ensure that the backing of their resources has little chance of producing radical change. Hopefully other projects, such as work on granulocyte based therapies, can find the connections needed to benefit from similar sources of funding and vision.