By Monica Heger
This story was originally published Feb. 17.
In one of the first demonstrations of the use of whole-genome sequencing to guide treatment, researchers at the Translational Genomics Research Institute and the Mayo Clinic have sequenced the genome of a 44-year old pancreatic cancer patient, uncovering mutations that could potentially be used to direct the patient's next course of treatment.
The work was part of a clinical research project that will seek to address how whole-genome sequencing could and should be implemented in a clinical setting for cancer patients.
"Our take on this is that it is the way of the future, and what we have to do for our patients," said Keith Stewart, dean of research at the Mayo Clinic in Scottsdale, Ariz. "And these are the first steps toward understanding the mechanics of what [sequencing in the clinic] will look like — the problems, the turnaround time, and how you might apply it."
Stewart said that the project will sequence up to 20 cancer patients. The team has so far completed the sequencing for two patients, but has only analyzed results from one. The trial is open to cancer patients at the Mayo Clinic, and Stewart said he expects the sequencing to be completed within 18 to 20 months.
The study will help pave the way for implementing next-gen sequencing technology in the clinic, and also highlights the challenges that still remain — including how to interpret results, the appropriate course of action if druggable mutations are found, and ethical implications surrounding whether and how to disclose results to the patient.
Sequencing was done on the Illumina HiSeq, and both the tumor and matched normal DNA were sequenced to just over 30-fold coverage with 108-base paired end reads. The sequencing and analysis was completed within six weeks.
The researchers declined to give specifics on the results because they plan to submit the study for publication in a scientific journal and are not disclosing the results to the patient.
However, they said that they discovered mutations for which FDA-approved drugs currently exist — although not for pancreatic cancer — and they were also able to rule out a number of cancer drugs that would likely not be effective.
"We definitely identified key alterations that were likely to be driver alterations," John Carpten, TGen's director of integrated cancer genomics, told In Sequence. "There were insights gleaned through the sequencing that would inform therapy," he added. While the patient is currently already undergoing treatment, Carpten said that the results could be used to guide second-line treatment, if that is deemed necessary. "We have targets that can be exploited to treat this patient's individual tumor."
Interestingly, the drugs indicated by whole-genome sequencing are "not generally agents that would be given to a patient with advanced pancreatic cancer," Carpten said. And in fact, "there are no known targeted agents for pancreatic cancer," he added.
In addition to identifying drug candidates for the patient's specific tumor, "we found many mutations that clearly indicated that certain drugs may not be effective," said David Craig, associate professor and director of neurogenomics at TGen.
Additionally, Craig noted, whole-genome sequencing as opposed to exome sequencing was essential. Without providing details, he said the team detected large amplifications and deletions that would not have been identified with exome sequencing. They also identified point mutations and small insertions and deletions.
Using exome sequencing, "we would have missed an incredible amount of this person's tumor that [could] subsequently guide treatment," he said, "both in known cancer genes and also unexpected ones."
Second Line of Defense
Despite turning up potentially actionable mutations, the results of the sequencing study will not be used on the current patient, at least for the time being. Because the researchers were not sure exactly how long it would take to have results, or even if the sequencing would return actionable results, Stewart said they started the patient on conventional chemotherapy treatment before completing the sequencing.
[ pagebreak ]
The patient is responding well to the treatment, so doctors are not planning to switch based on the results. Plus, said Stewart, "if we were to use the information, we'd have to go and resequence the genes in a CLIA-certified lab." Additionally, he added, just because the mutations appear to be actionable, there is no guarantee that those mutations are in fact the driver mutations, or that the drugs indicated by the results would be effective.
Craig said that the results could potentially be useful if a second round of therapy is needed. In fact, he said he sees sequencing's role for guiding cancer treatment to be in second- or third-line treatment because of the length of time it takes to sequence and analyze the results. The idea, he said, would be to use chemotherapy initially, and then a targeted approach for any subsequent treatment.
Carpten agreed, and said the goal of the subsequent targeted therapy would be to improve progression-free survival time. Typically, as cancer progresses, the length of remission decreases with each subsequent treatment. So, while an initial round of chemotherapy may leave a patient disease-free for 12 months, after a second round of treatment the cancer may return in eight months, with each subsequent round of treatment leaving a narrower and narrower disease-free window.
"Say a patient has a 12-month response to the first line of treatment. [We hope to show] based on targeted therapy that we can at least meet that time frame if not extend it longer," Carpten said.
However, Stewart noted that he would be concerned about waiting too long between the initial sequencing and using the results to guide treatment, since cancer mutates as it progresses.
"It's likely that if too much time went by, you'd have to do [the sequencing] again," he said.
Regulatory and Ethical Implications
Aside from the technical hurdles of implementing sequencing into cancer patient treatment, Stewart said the study has also highlights the ethical and regulatory implications of the approach.
For instance, Stewart said that the team is not reporting the sequencing results to the patient unless the patient specifically requests them. But, he said, the results could potentially have family implications — for example, if they were to find a mutation in a patient's germline DNA. Additionally, if doctors make treatment decisions based on sequencing results, then the patient would likely be told the results.
There is also an issue of how to pay for the tests. Since insurers currently don't reimburse whole-genome sequencing, Carpten said that it raises the question of whether such approaches would be limited to wealthy patients.
How sequencing in a clinical cancer setting would be regulated by the US Food and Drug Administration is another unanswered question. "We're still working through the regulatory aspects," Carpten said. "Members of our team are working with the FDA."
The results of the next patients could help inform some of the regulatory decisions, including at what point to seek CLIA certification. Stewart said that as part of the project the Mayo Clinic would examine whether it made sense to invest in next-gen sequencing as part of a CLIA-certified lab, or whether it would make more sense for them to only confirm interesting findings at a CLIA lab elsewhere if they were planning to act on the results.
"There's still a lot of work to do in proof-of-principle studies," Carpten said.
Have topics you'd like to see covered by In Sequence? Contact the editor at mheger [at] genomeweb [.] com.