By Monica Heger
This article was originally published July 29.
The Ontario Institute of Cancer Research is testing the Pacific Biosciences RS machine for targeted sequencing of cancer patients with metastatic or recurrent disease, Clinical Sequencing News has learned.
Currently, OICR is sequencing 19 known cancer genes and validating the results with Sequenom's OncoCarta panel in the CLIA lab at Princess Margaret Hospital in Toronto. OncoCarta, based on Sequenom's MassArray genotyping technology, tests for 238 mutations in the same 19 genes.
The researchers are also in the process of expanding the test to 30 to 40 genes, and if successful, would eventually like to incorporate around 200 genes that provide prognostic value, are known drug targets, or are in a pathway that can be targeted.
John McPherson, director of OICR's Cancer Genomics Program and the leader of the study, said the goal is to validate the machine and develop a framework for sequencing patient tumors that can help physicians make informed decisions about treatments or clinical trial placements. OICR aims to eventually incorporate this framework into the standard of care for cancer patients, he said.
The project will enroll 80 patients with advanced cancer, and is being funded for just over C$1 million (US $1.05 million) through OICR's High Impact Clinical Trial program with a grant from Cancer Care Ontario and the Princess Margaret Hospital Foundation.
Patients will be followed for two years to see how genomic profiling of their tumor affects clinical outcomes. Not only will the researchers measure patient outcome, but they will also assess patient and clinician perception of the genomic profiling and patient management, said Janet Dancey, director of OICR's High Impact Clinical Trials program.
So far, the team has sequenced 15 patients, half of whom have contained at least one mutation in one of the 19 targeted genes. The RS has not missed mutations picked up by OncoCarta, and so far sequencing has uncovered one novel mutation.
"That's the reason we're doing the sequencing, to see things that are not included in these targeted panels," said McPherson.
Mutations detected by sequencing that are not included in the MassArray panel will be validated in the CLIA lab through Sanger sequencing.
Turnaround time is three weeks, which includes the patient signing the consent form, having a biopsy taken, DNA extraction, sequencing, analysis, validation, and a meeting with clinicians and researchers to create a report for the oncologist.
"So far we've been on time with every sample except one, which we missed by one day," said McPherson.
McPherson acknowledged that as the gene list is expanded it will become "increasingly difficult" to stay within the three-week timeframe, particularly because mutations that are not included in OncoCarta will have to be validated by Sanger sequencing.
Turnaround time is one of the main reasons McPherson's team decided to use the PacBio RS. "We've had the PacBio for about a year and we're quite comfortable with it," he said. "We also have an Ion Torrent, but we've only had that for about a month. … We're just getting used to it."
He said that the data from the PacBio has been good, with average read lengths around 1.8 kilobases. The researchers are using circular consensus sequencing to ensure accuracy and are getting several hundred-fold coverage of the amplicons.
As the team scales up, McPherson said it will be necessary to evaluate other methods for target capture. Currently, they are using PCR, but that's "not scalable." Other capture technologies, however, do not achieve complete coverage of the gene, McPherson said.
"Since we're modeling a diagnostic system, we need to make sure we have 100 percent coverage of the gene," he said. While the team will be testing various capture technologies, ultimately McPherson anticipates that they will settle on a combination of techniques, such as a capture technology with limited PCR.
Another issue the team is dealing with is whether to become CLIA certified. "To become a CLIA lab is quite an endeavor," McPherson said, and would take a couple of years.
One option is moving the PacBio RS, or another next-gen system, to the existing CLIA labs within the hospitals and helping those labs run and validate the machine.
"The CLIA labs would definitely have to validate these instruments," McPherson said. "The PacBio won't be in a CLIA environment next week, but over time I think that's the direction we're heading," he said.
He noted that benchtop machines like the Ion Torrent PGM are attractive to the CLIA labs because of their lower cost, but "it's still quite new and the data are evolving rapidly. [Clinical labs] would like to see a more stable platform before adopting it."
Both models — the OICR becoming CLIA certified, or implementing a next-gen system into an existing CLIA lab — are up for debate, said McPherson.
McPherson would also like to expand the pilot study to enroll around 1,000 patients, which would require additional funding from outside sources. Currently the project is being funded through an internal program at OICR.
An expanded study would track patients over time to determine if the tumor sequencing helped the oncologist make better decisions about what clinical trials to enroll the patient in, and whether it improved patient care.
McPherson predicts that sequencing could eventually move from a targeted model to include whole exomes or even whole genomes. Rather than a fee-for-service model, which some organizations in the US like the Fox Chase Cancer Center are adopting (CSN 6/15/2011), he said it could become implemented into Canada's public health system and offered to all cancer patients.
"I'd like to see this become the standard of care," he said.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.