The Ontario Institute of Cancer Research has published an interim report of a pilot project to evaluate targeted sequencing of cancer patients with metastatic or recurrent disease to influence clinical care.
The group enrolled more than 80 patients in the pilot phase of the study, sequencing 19 known cancer genes using the Pacific Biosciences RS and validating the results with Sequenom's OncoCarta panel, which tests for 238 mutations in the same 19 genes.
In a paper published in the International Journal of Cancer last month, the group reported on its experience with the first 50 of these patients, writing that the pilot has demonstrated the feasibility of incorporating real-time sequencing analysis into patient management.
Moving into the second phase of the study, the team has switched gears, and is now using Illumina's MiSeq and the RainDance Cancer HotSpot panel, which targets hot spot mutations across 54 genes, to sequence another 80 patients, according to John McPherson, director of OICR's Cancer Genomics Program.
Janet Dancey, director of OICR's High Impact Clinical Trials program, told Clinical Sequencing News this week that the team has already enrolled 25 patients into the second stage of the trial.
In an e-mail to CSN, McPherson explained that the group is switching platforms because it plans to work with the University Hospital Network, which comprises three hospitals in Toronto, to move the test into its CAP accredited and CLIA certified laboratory, and the MiSeq is a better fit in terms of size and cost. McPherson added that in terms of data, the OICR team has had a good experience with the PacBio.
Dancey also said the results in the initial phase using the PacBio were promising.
"I think everybody involved is very enthusiastic about what we've found and done [so far] and that includes the patients and the clinicians," she said.
In its report of the first 50 patients in the pilot, the OICR team wrote that average sequencing depth with the PacBio RS was 600x, and 100 percent of target bases were covered, with 99 percent showing greater than 10x coverage.
For the pilot, the group used PacBio sequencing alongside genotyping with the Sequenom MassArray to establish that the sequencing results did not miss any mutations identified using the genotyping approach, and also to demonstrate that sequencing could find actionable mutations beyond those covered by the genotyping test.
In total, the group found 19 actionable mutations, including five not included on the OncoCarta array, in 16, or 32 percent, of the patients.
"In patients whose tumors contained novel mutations solely detected by sequencing, two were prescribed matched treatment, leading to one partial response … and one minor response … These results suggest an advantage for NGS over genotyping in the clinical setting," the authors wrote.
Dancey said the comparison results have given the group solid footing for moving forward with the sequencing trial. "For our purposes now, we are quite happy that MiSeq, or PacBio previously, can identify anything that's on the Sequenom panel without any issue, and can also identify more. So we have moved to an approach, [where] anything the MiSeq identifies we will validate in a CLIA laboratory with whatever technology is available."
"That will save time and save DNA, which is a critical factor," she said.
The average time from consent to final report was 20 days among the first 50 pilot patients, the group reported. With larger numbers of target genes, sequencing may take longer, but the group wrote that as more CLIA labs adopt NGS, the requirement to confirm results may be eventually eliminated, saving additional time.
Of the 16 patients with actionable mutations identified in the pilot study, six were at a disease stage allowing changes in their treatment strategy based on the results, the team reported. This resulted in four positive outcomes: three cases of partial drug response, and one case of stable disease.
Dancey explained that only six were treated because the other patients didn't meet criteria to qualify for the potential treatment associated with their mutation.
"If we moved testing even earlier in the advanced stage of cancers, the strategy might have had a greater impact," she suggested.
Mainly, the pilot has demonstrated to the group that its approach to guiding cancer treatment using rapid-turnaround sequencing is feasible.
According to Dancey the team is planning to profile a similar number of patients — about 80 — using the new approach as they did with the PacBio. The researchers have already recruited and are evaluating the first 25, she said, and expect to finish the full 80 by the end of this year or early next year.
After this, the team intends to launch a much larger assessment project, evaluating at least 1,500 patients over four years, Dancey said. She did not say whether the group won any funding for this expanded trial yet.
At least initially, she said, the group plans to continue using the MiSeq and RainDance approach, although eventually, the researchers would like to move to sequencing patients' whole exomes or whole genomes.
McPherson said in an e-mail that the team already had the RainDance platform so the HotSpot Cancer panel "was a quick route to an expanded actionable gene set."
"We continue to test out other targeting methods in my lab and will integrate them into the clinical study as appropriate," he said.