Childhood cancer experts are hoping that at least some of what drives pediatric cancers can be gleaned from the genomes of patients and their tumors. That’s the thinking behind the Pediatric Cancer Genome Project (PCGP), a three-year, $65 million effort to sequence major pediatric cancers.
The latest release of data from the PCGP, published in the journal Nature Genetics, includes 520 genome sequences from childhood cancer patients; half of the genetic material comes from their tumors, and half from their healthy tissues. By matching the tumor genomes to those of normal cells from the same patients, researchers hope to pinpoint the differences and get a better idea of where the cancer cells went awry. These discrepancies in the genetic code are also likely to be the richest targets yet for potential new therapies.
“We knew that if we only sequenced the tumor, and not normal non-tumor DNA from the patient, we would end up with thousands to hundreds of thousands of potential mutations, and no way to sort through which ones were important to cancer,” says Dr. James Downing, scientific director and leader of the PCGP at St. Jude Children’s Research Hospital, which partnered with Washington University for the project.
Ultimately, the researchers aim to use the genetic information to help develop new, more effective cancer treatments. Despite the fact that survival rates of many childhood cancers has improved to reach 80% to 90% in recent years, much of this benefit is due to early detection and quick intervention with conventional treatment, including surgery, chemotherapy and radiation therapies. There have been no new drugs to treat pediatric cancers in nearly two decades, and the rates of recurrent and new cancers among survivors remains high.
It’s hard to argue that the genetic information collected by PCGP won’t be useful, but can it change patients’ lives now? Will it help guide doctors to better treatments for childhood cancers? A recent study showed that despite the excitement over applying genetic information in improving health outcomes, adding genetic testing did not improve doctors’ ability to predict diseases such as breast cancer, diabetes or rheumatoid arthritis.
Downing acknowledges that the current depot of genetic information is but the first step. Doctors still need to be able to interpret and apply all the information gleaned from patients’ tumors and healthy cells to make smarter treatment decisions that improve survival — and perhaps even cure their cancer. “All the knowledge gained from direct analysis of DNA sequences will help as we start to apply them to personalized medicine and clinical genomics,” Downing says. “There is a lot of genetic testing going on, but does it really help? A lot of it doesn’t; there is a lot less specificity than we would ultimately need for a physician to use that sequencing information in caring for patients. But we are making good progress in interpreting the data and figuring out which mutation are important.”
Already, genetic information has provided some important clues about how best to treat certain childhood cancers: decoding the tumor genomes of a form of acute lymphoblastic leukemia (ALL), for example, showed that doctors were treating this cancer the wrong way; instead of being similar to other lymphoblastic leukemias, these tumors had more in common with acute myeloid leukemia (AML), another cancer of the bone marrow and blood cells, which is treated very differently. “So immediately we saw that we needed to modify the therapeutic approaches for patients with ALL,” says Downing.
Trials to treat patients who have relapsed on all available treatments for ALL with AML-based therapies are just beginning to launch, to establish whether the gene-based intelligence can make a difference in survival for these patients.
There are other examples, as well, in which a childhood eye cancer was traced to a mutation for which a targeted anti-cancer drug already exists.
The idea, says Downing, is to collate all the genetic information on childhood cancers into one place, so that researchers, clinicians, biologists and drug developers can analyze the data and start connecting dots toward more effective treatments. The database is currently available through the European Bioinformatics Institute portal.
While the effort was privately funded by St. Jude Children’s Research Hospital and Kay Jewelers, the PCGP will make access to the database free to researchers, as long as they use the information for research and agree to hold any publications for nine months after the data is uploaded to ensure time for any modifications to verify the sequences.
Because the final goal of the PCGP is to make DNA information useful in the clinic, Downing says that most of the project’s publications include some interpretation of the genetic material to help scientists apply it to effective clinical treatment. PCGP researchers attempt to provide background information on how the mutations identified might contribute to tumor formation, and whether there exist any studies linking mutations to bad or good outcomes for patients.
Downing acknowledges that scientists are still a long way from directly applying genomic information to cancer care, but the more samples and the more tumor DNA that is layered into the PCGP, the more easily researchers can start to pick out genetic patterns and read the hidden language of cancer cells.
“There is a wealth of information that we are generating and it will take years to extract all the valuable information,” he says. “But we need to start by getting it out into the hands of everybody in the scientific field — not only those in cancer research — so they can use it both as reference and as a discovery tool, and find new things that we have yet to find.”