By Monica Heger
This story was originally published July 26.
With commitments for the sequencing of 17,000 tumor genomes in hand and 40 projects already underway, the International Cancer Genome Consortium is ahead of schedule to sequence 25,000 cancer genomes from 50 different body sites by 2018.
In addition, new projects have been added, including two by Cancer Research UK for esophageal and prostate cancers, and a brain cancer project from Genome Canada.
"We've doubled the number of projects since last year," Tom Hudson, president of the Ontario Institute of Cancer Research and member of the ICGC's executive committee, told Clinical Sequencing News. "We're ahead of schedule."
Hudson said that the consortium is now focused on adding projects from countries that are as of yet not represented. For example, Saudi Arabia has become the latest "observer member" and is in the planning phase for a thyroid cancer project, but "it will take another year before that project is official," he said. "We're also looking to accelerate projects in Africa and South America," he added.
Currently around half the projects are being done in the United States and Canada. Of the 17,000 committed samples, 8,000 are from North America, 6,000 from Europe, and the rest from Australia and Asia. Hudson said the consortium is especially interested in ensuring that the effort is truly international because there are many rare cancer subtypes that are specific to certain countries.
So far, 25 projects have submitted data on more than 2,800 tumors including whole-genome, exome, transcriptome, and methylome data, identifying around 9,000 genes that contain at least one nonsynonymous coding mutation.
"We're just at the start of generating massive datasets," Hudson said. The project calls for the submission of genomic, transcriptomic, and epigenomic data from all tumors. The submitted datasets are in various stages of completion, but are already "bigger than anything that's been generated," Hudson said.
For example, a Spanish team, including researchers from the University of Barcelona, recently published a study in Nature on the sequencing of four chronic lymphocytic leukemia patients as part of the ICGC (CSN 6/8/2011). Additionally, a Cancer Genome Atlas team recently published the largest cancer sequencing dataset yet of the exomes of 316 ovarian cancer patients, which will be included in the ICGC database (CSN 6/29/2011).
Hudson attributed the rapid pace of the project to advances in sequencing technology, which has reduced the cost dramatically and enabled widespread adoption of the technology. In particular, the reductions in cost are enabling teams to transition from exome sequencing to whole-genome sequencing much quicker than anticipated.
He said that he expects the project to surpass the 25,000-genome goal, but there are no plans as of yet to change the ICGC's policies or goals to reflect this increased capability. The ICGC has also not changed its timeline, despite being ahead of schedule. Most of the projects launched are for "more common cancers like breast, ovarian, and prostate," said Hudson. "But there are a lot of rare cancers, like pediatric cancers, that can take years to get a tumor collection," he said.
Additionally, obtaining consent for all the samples takes time. While the consortium has secured commitments for 17,000 tumors, so far only about 5,000 patients have consented.
The consortium is also now looking into whether to return results to patients, and if so, how. "This is a very hot topic for the ICGC," Hudson said.
The consortium's policy is that the return of results should be determined by the individual project heads because there are different legal environments in different countries. As a result, "some projects are returning some actionable data and other projects are not," he said.
For example, at the OICR, there is limited return, said Hudson. If the researchers find results they think should be returned, they will submit that data to a medical committee who will decide whether the findings are clinically important, and if so, the team will then return those results to the treating clinician.
But now, he said, "there's a shift happening." When the ICGC first began, there were very few mutations that were well enough understood to be considered medically actionable, so researchers were very conservative about returning data. Increasingly though, more data is being generated, and the list of medically actionable mutations is growing. "The opinion is moving from no or limited return of results to more return," Hudson said. "This could affect the policies of the ICGC in the future."
In fact, there is now a large study within the ICGC to address this question. The study is surveying the ICGC participants about how their protocols for returning data.
Bartha Knoppers director of the Center of Genomics and Policy at McGill University, who is helping coordinate that effort, said that the results should be tabulated this fall.
"We want to know from our researchers what exactly their practices are. Do they go to the [institutional review board], do they decide case-by-case, or do they return results to the family physician and let them handle it?" she said.
Whether or not and how to return results from next-gen sequencing studies is "becoming one of the major issues within research ethics," she added. It is also becoming more nuanced — "context matters, and you can't just have a generic policy and apply it to all disease areas."
Looking ahead, Hudson predicts ICGC data will soon begin to be translated into a clinical setting to impact patient treatment. Already, data from ICGC's prostate cancer sequencing projects has identified two subtypes — those that respond well to treatment and those that have tumor recurrence. This information could be used clinically to determine which patients should receive aggressive treatments such as surgery or chemotherapy, and those that just need to be monitored.
For example, Canada's Prostate Cancer Genome Network is sequencing frozen biopsies from prostate tumors to identify the mutations that lead to tumor recurrence and biomarkers that distinguish patients with an aggressive form of the disease from those with a more indolent form, Hudson said.
Additionally, because the data is publicly available, Hudson predicts that pharmaceutical companies will comb the datasets to find targets for compounds they have already developed or are developing, as well as to identify mutations for which new compounds could be developed.
The project is just now starting to identify the "low-hanging fruit — the biomarkers and drug targets" for which drugs already exist, he said. Going forward, he predicted that there would be an increase of "clinical translation" of the data, for example, using the findings to help stratify patients in clinical trials.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.