Researchers at Toronto's Hospital for Sick Children are forging ahead with a project to identify genomic changes that separate children with medulloblastomas into distinct groups based on differing outcomes and treatment responses.
The venture has been funded with C$9.8 million by Genome British Columbia and other partners as part of Genome Canada's Large-Scale Applied Research Project Competition, which kicked off in 2010. The research group is studying the tumors of 1,000 children using RNA sequencing and whole-genome sequencing to identify markers that can segregate the disease into distinct subgroups for personalized treatment.
The researchers hope that by identifying clear genomic subgroups in the disease and creating a test that can assess which children have these gene alterations, they will be able to treat those with the best chance of survival more conservatively — saving them painful and harmful side effects — and allow those with the worst prognosis to avoid standard regimens in favor of more experimental or less debilitating treatment.
Michael Taylor, a pediatric neurosurgeon at the hospital and one of the project's co-leaders, told PGx Reporter that medulloblastoma has a relatively high cure rate, but at severe costs to pediatric patients due to the effect that toxic and aggressive treatments have on the developing brain.
"The cure rate is between 60 and 70 percent probably, which isn't bad but really only tells half the story," he said. Standard treatment starts with aggressive brain surgery and irradiation of the entire brain and spinal cord, followed by high-dose chemotherapy, often supplemented by bone marrow transplant.
"Many kids that survive have a poor quality of life," with [lower intelligence], neurological deficits, and problems with their pituitary and thyroid. "They end up short, bald, and their bone marrow doesn't work well," Taylor said. "So even though more and more kids are being cured, the price they are paying is very large."
Taylor's group has been driving toward subgroup-based treatment of medulloblastoma through several years of microarray-based transcriptional profiling studies.
"What's become apparent over the last year or two is that this disease we used to call medulloblastoma is not really one disease but a collection of diseases that look similar under the microscope, but are epidemiologically, clinically, transcriptionally, and genetically distinct," he said.
Survival, he said, seems to differ greatly between groups, with some having really good outcomes, and others succumbing to the disease even after receiving the most aggressive treatment. With the new project, Taylor and his colleagues will deepen and expand their earlier work, using RNA sequencing and whole-genome sequencing to analyze 1,000 tumors over three years.
Taylor said the group has already begun sequencing tumors and analyzing initial sequence data. The researchers are studying tumors from the Hospital for Sick Children's own patients, as well as samples from partner institutions in the Medulloblastoma Advanced Genomics International Consortium.
He said the team hopes sequencing will allow them to discover more of the transcripts present in the disease and to identify those that may have been missed using microarrays.
"The brain has a large number of transcripts that are poorly annotated, in particular for the cerebellum, which is the part of the brain medulloblastoma grows on. So probably all kinds of genes are important for the pathogenesis of the disease and might serve as good markers for the different subgroups that we would only find by doing an unbiased approach like RNA-seq," he said.
The researchers hope to be able to select markers that can separate the groups distinctly and that they can then use to stratify patients to different treatment strategies in clinical research.
"The crux comes down to understanding the differences between groups and how many groups there are and then having biomarkers that will be reliable and will work in the setting of a clinical trial," Taylor said. "The idea being that with the [children] that have really good survival we could consider backing off on their therapy, hoping to maintain the cure rate, but decrease complications."
Meanwhile, children with a low chance of survival "should probably move sooner rather than later to more experimental therapy," he said. "There's no sense just throwing the book at them when we know it won't work."
According to Taylor, the group is also interested in gleaning information about the pathogenesis of medulloblastoma among different subgroups from mutations it discovers.
"Our early results are already starting to identify recurrent [mutation] events that are highly restricted to specific subgroups," some of which are associated with "drugs available for treatment of other neoplastic or non-neoplastic disorders," he said. "So we hope we might be able to transition those drugs rapidly to phase II for children with brain tumors."
Meanwhile, as they build a group of markers to distinguish subgroups, Taylor and his team hope within the span of the project to initiate trials of lighter treatment regimens for those groups who are either most likely to do really well without aggressive treatment, or most likely to gain little from standard treatment.
"The thing I feel really excites people about this," Taylor reflected, "is that to a certain extent for the children that have a really high survival rate, if indeed they can be cured with less therapy, we don't need to invent any new drugs. We just have to know we can identify them, and then arrange a clinical trial and just walk over to the machine and turn the dial down."
One part of the grant, he said, will go to studying whether parents will actually agree to enroll their children in such a trial or not, and whether they are willing to accept the risk of reduced treatment for potential gains in quality of life.
"It is contentious," he said. "Some people are very concerned about quality of life, some about survival rates."
"We're heading into this era of personalized medicine and I think this is a very good example of that. For families where quality is paramount, they may love this, and for those [for whom] quantity is paramount, they'll probably want to stick with standard treatment," he said.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at mashford [at] genomeweb [.] com.