By Jocelyn Rice
Clues to Neuroblastoma
Researchers have discovered a genetic link to a mysterious childhood cancer
By Jocelyn Rice
After exhaustively scanning the genomes of more than 1,700 neuroblastoma patients, researchers have identified a genetic tweak that doubles a child’s chances of developing the cancer.
Neuroblastoma almost exclusively affects infants and very young children, causing tumors that usually start in the abdomen or chest and creep out through nerve tissue to the rest of the body. Although it occurs in just 700 children each year in the U.S., it is one of the most common—and lethal—solid tumors among infants and children. Mild cases sometimes spontaneously disappear, but advanced, stage III or IV tumors have a survival rate of less than 40 percent. Causes and cures remain elusive.
So when it became clear that genome-wide association studies can help zero in on the genetic basis of a complex disease, John Maris, an oncologist at Children’s Hospital of Philadelphia, decided to apply this emerging tool to blood samples he had been collecting from neuroblastoma patients over the last decade. His longtime colleague, geneticist Hakon Hakonarson, hopped on a plane from Iceland to come help.
The researchers tested the genomes of roughly 1,000 neuroblastoma patients and 2,000 healthy children for the presence of nearly 500,000 genetic variations, or variants, known to occur in the general population. Then they combed through the data to identify which variants occurred more or less frequently in the cancer patients. They found three suspicious variants, but one in particular, a small change on chromosome 6, stood out from the bunch.
Once they pinpointed this abnormality, the researchers expanded their study by looking for the suspicious variant in the DNA of three other groups of neuroblastoma patients. All told, Maris and Hakonarson found that individuals born with two copies of this particular genetic change have twice the normal risk of developing aggressive, late-stage neuroblastoma.
The analysis, published in the June 12 issue of the New England Journal of Medicine, is the first genome-wide association study of neuroblastoma.
The result is very significant, according to pediatric oncologist Javed Khan, the head of oncogenomics at the National Cancer Institute’s Center for Cancer Research. The genetic abnormality lies within a hypothetical gene dubbed FLJ22536. While this gene has no known function, Hakonarson says its sequence suggests it may belong to a family of genes known to help build the blood vessels that feed a growing tumor. By characterizing FLJ22536 and probing its activities, the researchers hope to uncover valuable information about how neuroblastoma develops.
“No one really knows what that gene does,” says Khan, but “it points the way to a lot of biological understanding of neuroblastoma.”
Characterizing the mystery gene may also help identify biological targets for new drugs. If the gene does contribute to a tumor’s blood supply, says Hakonarson, blocking its action might starve the tumor and prevent further spread. “We may be after something very meaningful with respect to potential for new therapies,” he says.
Despite its promise for basic research into the disease’s causes and mechanisms, the new finding probably won’t lead to a genetic test for neuroblastoma anytime soon. Not everyone with the genetic abnormality will develop cancer. Doubled risk may seem substantial, but experts agree that the baseline risk for neuroblastoma is so low that testing all children would be neither practical nor informative.
In certain cases, when there is a family history of neuroblastoma or related cancers, Hakonarson suspects genetic testing may one day be valuable. Any individual variant—including the one identified in the new study—would not be sufficient to assess a child’s risk, but a cluster of variants could combine to paint a fuller picture. Since publishing the chromosome 6 result, the team has already identified two other mutations associated with the disease.
“It’s going to be a long process to get this into clinical medicine,” says Hakonarson, “but this certainly is a very, very strong step in that direction.”