Investigators at St. Jude Children’s Research Hospital have developed a strategy to speed future development of more effective and less toxic treatments for medulloblastoma, a type of brain cancer. Medulloblastomas arise in the back of the brain and account for about 20 percent of childhood brain tumors.
The new technique could also be used to identify specific pathways in other types of cancer that might be vulnerable to novel therapies, and therefore speed development of so-called molecular-targeted therapies for a wide variety of cancers. Molecular-targeted therapies work by blocking individual molecules that are key triggers of disease.
This St. Jude study is important because the aggressive combination of surgery, radiation and chemotherapy used to treat medulloblastoma fails to cure many patients, according to researchers. Therefore, there is a great need to identify alternative therapies, such as novel drugs that block signaling pathways that are abnormally activated. Such treatments could not only save lives but also eliminate the severe side effects caused by current therapies, according to Richard Gilbertson, M.D., Ph.D., an associate member of the Department of Developmental Neurobiology at St. Jude and director of the Molecular Clinical Trials Core. Gilbertson, co-director of the St. Jude Neurobiology and Brain Tumor Program, is senior author of a paper on this work that appears in the April issue of Journal of Clinical Oncology.
The technique is designed to rapidly match a specific novel drug treatment to those children most likely to respond to it. This approach would avoid trial-and-error therapy that fails in patients who are not ideal candidates for a specific treatment, researchers said. It would also reduce the chance that otherwise effective drugs would be abandoned because they failed in such patients during clinical trials.
The key to the new St. Jude strategy is the ability to determine in individual children which biochemical signaling pathway triggers and sustains the cancer by identifying key genes that are linked to that pathway. The investigators proved that this strategy is valid by demonstrating that it is possible to assign children with medulloblastoma into specific groups, depending on which biochemical signaling pathway is abnormally active. Based on this classification, novel drugs designed to block a key protein in each specific pathway could be correctly administered to the children most likely to respond to them.
"Our strategy would ensure that a new drug would get a fair trial by using it to treat only the appropriate children," Gilbertson said. "That will be important as novel drugs are developed to treat medulloblastoma."
St. Jude researchers reported in the September 2004 issue of Cancer Cell that a novel drug called ShhAntag effectively targets such a pathway in mice with medulloblastoma and dramatically reduces the size of the tumor.
The St. Jude strategy overcomes two problems that often pose barriers to molecular targeted therapies. First, not all children have the same gene mutations that cause a specific cancer, Gilbertson said, so a novel therapy that targets only a gene known to cause the cancer in one group of children will not be effective in a child whose same cancer is caused by a different mutation. Secondly, scientists have identified only a few genes that are potential targets for novel drugs in medulloblastomas, and even fewer in other brain tumors. The current solution to that problem—to identify the DNA sequence of all the genes in each child’s tumor to determine which ones are mutated—is an enormous and expensive undertaking, he said.
However, the St. Jude team used an approach for identifying the signaling pathways that cause medulloblastoma in different groups of children.
"Our strategy is somewhat analogous to the link between a home’s electrical wiring system and a light bulb," Gilbertson said. "You know if a certain part of the wiring network is working if the light bulb connected to it is on. In the medulloblastoma cell, the wiring is composed of proteins making up the specific signaling pathway that is abnormally activated by a gene mutation, and the light bulb is the group of genes that the pathway causes to be over-expressed."
The St. Jude team showed in samples of medulloblastoma from different groups of children that it is possible to determine which gene mutations are present in tumors by studying which genes are expressed by those tumors.
"Since each of these gene expression signatures is regulated by a specific pathway, these genes essentially report to us which pathway might be causing the cancer," Gilbertson said. "Using this information, researchers can use drugs to block that pathway."
The St. Jude team proved that the principle of identifying genes linked to abnormal signaling pathways is valid by generating gene expression profiles from 46 medulloblastoma samples. The investigators identified five medulloblastoma subgroups that differed from each other by their patterns of gene expression. These subgroups also differed according to their histologic (microscopic) characteristics and by how they behaved in patients.
For example, medulloblastomas in subgroup B occurred in patients who were older than 3 years and whose cancer was of the standard or classic type usually recognized under the microscope. In contrast, medulloblastomas in subgroup D occurred in children younger than 3 years, and these tumors had different characteristics from those in subgroup B.
The investigators showed that activation of different signal pathways contributes to the different medulloblastoma gene expression signatures found in each subtype. For example, the team found that genes in a pathway called WNT are especially active in tumors from subgroup B, while genes in the SHH pathway are especially active in subgroup D.
The researchers then looked at the DNA sequences in the tumors to show that gene mutations that activate the WNT pathway were only found in subgroup-B tumors, and mutations that activate the SHH pathway were restricted to subgroup-D tumors. Looking for patterns of gene expression can be done much more rapidly than looking for gene mutations; therefore, in future studies of novel drugs, such gene expression signatures could be used to quickly identify which signaling pathways are responsible for medulloblastoma in specific children. Only those children would receive the drug during clinical trials.
"Our work showed that drugs that target specific pathways are likely to be effective in distinct populations of patients with medulloblastomas," said Margaret Thompson, M.D., Ph.D., clinical fellow in the Gilbertson lab. She is the first author of the paper. "We believe our technique will provide a rapid and accurate way to select appropriate patients for clinical trials of each molecular targeted therapy for medulloblastoma."
Other authors of the study include Christine Fuller, Twala Hogg, James Dalton, David Finkelstein, Tom Curran and Amar Gajjar (St. Jude); Ching Lau, Murali Chintagumpala and Adekunle Adesina (Texas Children’s Hospital); David Ashley (Melbourne, Australia); Stewart Kellie (Sydney, Australia); and Michael Taylor (Hospital for Sick Children, Toronto).
This work was supported in part by the National Cancer Institute, the Sontag Foundation, the V-Foundation for Cancer Research, a Cancer Center (CORE) Support Grant and ALSAC.