By Monica Heger
This story was originally published Sept. 2.
The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University is launching a personalized cancer center that will use new technologies, including next-generation sequencing, to hone in on novel genetic characteristics of each patient's cancer to guide treatment and support drug development.
The new center, funded by a $30 million gift from the Commonwealth Foundation for Cancer Research, is the latest in a growing trend of organizations seeking to bring next-gen sequencing to the clinic and advance personalized medicine.
Cancer, in particular, has served as a testing ground for the use of sequencing to guide treatment, improve diagnostics and prognostics, and identify new drug targets. While sequencing will not be the only technology used by the Center for Personalized Cancer Care at Johns Hopkins, it will play a large role, particularly for understanding extreme drug responses and individual cancer antigens.
"Certain types of data loom very large in the cancer setting," said William Nelson, director of the Kimmel Center, including sequence data from the germline, somatic normal, and tumor genomes; epigenomic data; and transcriptome sequencing data.
The new center is part of a broader university-wide focus on personalized medicine that will include not only cancer, but other diseases as well.
Exact details for the center have not yet been worked out, but would include whole-genome sequencing, methylation sequencing, and RNA-seq.
"Our ability to do [the sequencing] is there," said Nelson. The key is figuring out how to best apply it to get the most information.
Initially, sequencing studies will focus on what Nelson described as "high-yield" applications such as looking at extreme phenotypes and identifying cancer antigens.
"One of the early approaches that will be reasonably informative is to look at cases … that have strikingly different phenotypes," said Nelson.
For example, among patients with disseminated prostate cancer, some show a "spectacular response" to hormonal treatment, with the cancer "effectively disappearing for decades." Meantime, other patients who present similarly may only respond to the treatment for a couple of months.
"We think that by focusing on these extreme differences in a discovery mode we'll be a little more likely … to find things that may be more likely to be useful tools in distinguishing those patients," he added. That approach has a better chance of being successful than sequencing a large cohort of patients covering a broad spectrum of responses and trying to extract meaningful data that explains only marginal differences, he added.
The strategy not only has a better chance of success in terms of finding relevant variants, but is also more likely to lead to an immediate clinical benefit.
"[I]f someone's going to respond really well to a treatment and you know who they are, obviously you want to make sure you deliver that treatment to that person," Nelson said.
[ pagebreak ]
He noted that these kinds of effects are already evident in cancer. For example, patients with specific mutations to EGFR respond especially well to EGFR inhibitors such as gefitinib (AstraZeneca's Iressa) and erlotinib (Genentech's Tarceva), while patients without those mutations do not.
Additionally, he said, there are agents on the market, like the immune checkpoint inhibitor ipilimumab (Bristol-Myers Squibb's Yervoy) recently approved to treat metastatic melanoma, where around 10 to 20 percent of patients see a dramatic response, while other patients see no benefit at all.
Nelson hypothesizes that there are likely genomic differences between the two extremes that explain the difference in drug response.
"I think that's where we'll see yields," he said — sequencing and identifying the variants between "people who respond especially well, and those that really don't respond at all."
The team will also explore the use of exome sequencing in developing personalized cancer vaccines. "There is an interesting universe of antigens that might be used to prime anti-cancer immune responses," which are typically coding sequence changes that lead to amino acid changes, he said.
Understanding the "entire universe of these changes in the cancer that are distinct from the normal tissue" may enable personalized cancer vaccines, he said.
Next-gen sequencing is not a new undertaking at Hopkins, and a number of labs within the university have made use of it in exploring research questions. The next step will be to start answering clinical questions.
"We've learned a lot in the low-lying fruit range," said Nelson. "To take it to the next level, we have to move it more toward clinical questions regarding who benefits, who responds, what treatments should be given, rather than, 'Let's sequence 50 of these and oh, we found a mutated gene we'd never thought of before.'"
According to Nelson, there may be a few more months left of those types of revelations from sequencing, but the field will soon move toward trying to answer the more difficult questions of what it all means clinically.
As part of an effort to answer those clinical questions, the center will move to develop a CLIA-certified lab, which would include next-gen sequencing, and is also trying to figure out a protocol for returning results to patients.
The Hopkins cancer center is currently equipped with Illumina and SOLiD platforms. The university also has an Ion Torrent, but has not yet used it for any cancer sequencing projects. Additionally, it is getting a Pacific Biosciences machine.
Of those, Nelson said that they would likely only take one through CLIA certification, but hadn't yet decided which one.
Nelson said the team is also figuring out when returning results to patients is appropriate, and how to do it in such a way that the patients understand the significance of the findings. He said the group has been talking with researchers at the Genome Institute at Washington University about their consent process, and are working with the Berman Institute of Bioethics at Johns Hopkins to figure out an appropriate protocol.
"It's a critical issue," he said.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.