Childhood cancer is a terrifying concept.

No one wants their child to have cancer, although as physicians we can do a good job of not only treating but curing most children of their disease.

To do this, however, we need to use a toxic combination of chemotherapy, surgery and radiation therapy, which can leave some children with long-term health problems. And, these approaches don't work for everyone.

My research focuses on finding ways to use the immune system to help eliminate cancer in children.  One of our earliest successes has been in working against a very resistant form of leukemia, a blood cancer.

For the therapy, we collect T cells, which are a type of white blood cell essential to human immunity, from the patient's blood. We then then treat the cells in the lab to help them recognize the leukemia when returned to the child.  This process results in the insertion of a gene called a chimeric antigen receptor, or CAR, that allows the T cells to see and eliminate the leukemia.

CAR T-cell therapy has helped many children achieve remission from a leukemia that had been resistant to all other kinds of treatments, including bone marrow transplants.

Our goal is to use this approach for other types of cancer, but we have had difficulty in achieving the same results for other, similar childhood cancers.

In thinking about this problem, I looked at the nature of the T cells we got from leukemia and lymphoma patients and noticed a key difference.  Normally the body has a range of T cells that we consider very young (naïve T cells) to T cells we consider very old (terminal effector cells).

Young T cells were present in the patients with leukemia, and these cells seem to be critical to the success of the CAR therapy.

But the more chemotherapy a leukemia patient received, the less likely we were to find an abundance of young T cells.  In lymphoma patients, these young T cells seemed to be missing even before chemotherapy.  This appears to be one of the major challenges to getting CAR therapy to work for lymphoma, and we are actively investigating ways to get the most out of T cells from lymphoma patients.  For leukemia patients, one of the keys may be collecting T cells before the patient has had too much chemotherapy.

As we learn more about what kind of T cells make for effective CAR therapy, we hope to also understand more about the immune system function of children with cancer.  T cells are really good at eliminating pre-cancerous cells all the time, and a fundamental question we seek the answer to is: Why did the T cells allow that child's cancer to happen in the first place?

We have more answers now than we did just four short years ago, when we treated the first child with CAR therapy for leukemia.  My hope is that we can replace many aspects of current therapy with strategies that involve the immune system, reducing the long-term toxicity of current therapy and curing children we otherwise could not.

David Barrett, MD is a pediatric oncologist at The Children's Hospital of Philadelphia

Read more Diagnosis: Cancer here »