GNN - Genome News Network  
  Home | About | Topics
Gene Pattern Predicts Deadliest Liver Cancers
By Adam Marcus

Scientists have found more than 150 genes in liver cancer cells that indicate which cells are likely to spread throughout the organ, and thus are most likely to be lethal. Among those genes was one that might offer a target for new treatments for the deadly tumors, as well as a potential diagnostic test.

The researchers used gene chips to analyze the activity of more than 9,000 genes in liver cancer cells taken from Chinese men and women with liver tumors caused by hepatitis B infection. Gene chips are discs of glass or plastic that can detect the activity of thousands of genes in a blood or tissue sample.

CT scan image of a liver tumor.

The study was conducted by scientists in China and the United States, and reported in Nature Medicine. The new work supports a growing body of recent studies showing that, at least in some cancers, a propensity to spread is genetically pre-programmed.

“These data appear to support the idea that metastasis is really an inherited ability that comes from the beginning” of a tumor’s formation, says Xin Wei Wang, a cancer geneticist at the US National Cancer Institute, in Bethesda, Maryland, and leader of the research.

Wang’s group found almost no differences in the gene activity of liver tumors that had spread and cells taken from the primary cancer site. But they did find 153 genes in metastatic tumors whose activity, or “expression” was either greater or less than that of cancers that had never migrated. The result: a gene “signature” with the power to predict the course of a person’s cancer.

“Such a molecular signature could help clinicians in diagnosing liver cancer with poor prognosis,” Wang says.

The next step, says Wang, is to repeat their gene chip study in a much larger group of patients, perhaps as many as 500, with a wider range of liver cancers.

In a second experiment, Wang’s group looked for genes that were overactive in tumors that had spread throughout the liver compared with normal liver tissue. One, osteopontin, was three times more active in the cancerous cells.

Osteopontin encodes a protein of the same name that in high levels has been implicated in several forms of metastatic cancer, including tumors in the breasts, colon, lungs and prostate gland.

Wang’s group found that they could almost completely prevent liver cells from spreading in a dish by treating them with a molecule that blocks the osteopontin protein. When they gave the molecule to mice with an aggressive form of liver cancer, only half developed metastatic tumors, compared with all of a group of untreated animals. The tumors that did form were also much smaller than those in the untreated mice.

“The inhibition of metastatic tumors is quite significant” in mice treated with an anti-osteopontin molecule, Wang says. Whether such a compound would work in humans remains to be seen. Even so, osteopontin is an appealing protein for cancer detection because it appears in both blood and urine, Wang adds.

While the seeds of metastatic cancers may be present in the genes of cells, Wang believes other changes to DNA are needed to trigger a breakaway. These changes involve removal of chemicals that attach to DNA and regulate the activity of genes. In liver cancer, chronic inflammation, alcohol abuse and infection with viruses could knock off the governing molecules. Similar factors also may help explain metastatic tumors in other organs, like the lungs and breasts.

Ann Chambers, a cancer expert at the London Regional Cancer Center, in Ontario, says recent studies hint that a person’s genes may have a lot to say in advance about the risk of having metastatic tumors.

“It may be that some people’s tumors are more likely to metastasize depending on the genetic make-up of the individual,” she says. In other words, “some people’s tumors are more likely to metastasize than similar tumors in other people.”

However, Chambers says, the reason for the difference may not be a vast cadre of genes, as Wang’s research suggests. Instead, some new work indicates that a smaller core of ‘metastasis suppressor genes’ may control the spread of tumors.

A few defective suppressor genes hidden among a larger collection of other genes could also help explain why DNA chip tests of primary and metastatic tumors look so much alike. “If the appropriate metastasis suppressor was not on the array chip, the primary and metastatic tumors would indeed appear to be similar,” she says.

See related GNN article
»Gene signature may signal cancer's spread

. . .

Ye QH, et al. Predicting hepatitis B virus-positive metastatic hepatocellular carcinomas using gene expression profiling and supervised machine learning. Nat Med Published online March 17, 2003.

Back to GNN Home Page