Collaborating with Foundation Medicine, researchers from Weill Cornell Medical College have demonstrated that targeted sequencing can identify potentially actionable mutations in formalin-fixed paraffin-embedded samples from advanced prostate cancers.
The group published in the Journal of the European Association of Urology this month results of a small study that sequenced samples from 45 patients. Sequencing the samples to an average depth of 900x yielded a number of potentially targetable mutations, including a novel rearrangement involving BRAF.
Mark Rubin, the study's senior author and a pathologist at Weill Cornell Medical College told Clinical Sequencing News that his lab is planning to expand on the approach, moving to whole exome-sequencing, in a much larger multi-center project to identify mutations that can guide therapy for advanced prostate cancers.
The larger upcoming effort is the seven-center University of Michigan-led "Precision Therapy for Advanced Prostate Cancer" project. Rubin said the study with Foundation Medicine in some ways suggested a path forward for this next effort, for which the sequencing will be performed by the University of Michigan and the Broad Institute.
"We were able to look at selected genes, and in those known to be associated with cancer or potential drug targets, we were able to perform very, very deep sequencing • up to 1,000-fold coverage, so if there were any mutations we could find [them,]" he said.
"That we could take these tiny samples, and do this sequencing, was an amazing thing."
In the EAU study, Rubin and his Foundation Medicine collaborators assessed a method of hybridization capture and sequencing on the Illumina HiSeq 2000 to deeply sequence samples of at least 50 nanograms of tumor DNA from FFPE biopsies.
"If you look at the entire literature of studies in prostate cancer," Rubin said, "it probably would be fair to say that 99.9 percent of the studies that look at molecular alterations are coming from prostatectomies • men who have clinically localized cancer."
"One of the great limitations," however, "has been that men who have metastatic disease do not undergo biopsy and tissue is not available for study in those patients … and if you do get samples, they are often only available after they are fixed in formalin," he said.
Using Foundation Medicine's targeted approach, the group was able to sequence FFPE samples to an average depth of 900x, the authors reported.
The group identified many mutations that Rubin said are potentially actionable or "hypothesis generating," including mutations in the androgen receptor, but only in samples that had been treated with hormonal therapies.
"AR alterations have not been detected in hormone-naïve PCa, supporting acquisition of AR alteration as a mechanism of resistance to hormonal therapies. This observation has important clinical implications," the authors wrote.
The study also identified mutations in BRCA2 and other genes that suggest some advanced prostate cancer patients could be candidates for trials of PARP inhibitors.
The study authors noted that several ongoing trials are evaluating PARP inhibitors or Akt inhibitors in patients with metastatic prostate cancer. "Incorporating NGS assays into biomarker-designed trials may help stratify patients for treatment and result in improved therapeutic response rates," the group wrote.
Finally, the team identified a novel BRAF rearrangement that it wrote may also be targetable.
"In this relatively small study we found [all these] potentially actionable mutations," Rubin said. "And this is just perhaps a first glance at what you can do with these technologies."
Rubin's larger upcoming effort with the University of Michigan and five other centers will attempt to use mutation information from several hundred patients' prostate cancer genomes to predict their response to current and developing therapies.
The project was awarded a $10 million, three-year grant this spring by Stand Up To Cancer and the Prostate Cancer Foundation. Rubin said the group plans to recruit 600 patients, of which they hope to successfully sequence about 500. According to the project announcement, the study will "systematically evaluate the prostate cancer genomes of patients enrolling in four clinical trials, evaluating novel drugs for castration-resistant prostate cancer or beginning treatment with approved drugs like abiraterone."
Rubin said the group will be following the approach his team took in the Foundation Medicine study of collecting biopsy samples from already-treated patients and sequencing them, but will be using its own protocols and will capture the whole exome instead of a targeted panel of genes. The group will perform the sequencing at the University of Michigan and the Broad Institute.
Rubin and his colleagues recently published a prostate cancer exome sequencing study in Nature Genetics. He said the approach will likely be very similar going into the Stand Up To Cancer project.
"The Foundation Medicine study started over a year ago," Rubin said. "At the time, really we needed to first demonstrate that with the smallest amount of material we could do the sequencing."
"Now the technology is moving forward very rapidly and at this point we are going for the whole exome. You can see in the [Nature Genetics] paper that there are mutations we find [with exome sequencing] that you could not have predicted."
According to Rubin, the sequencing for the Stand Up To Cancer project, which began this year, will be done in a CLIA-certified environment. All subjects will go through a consent process whereby they indicate through a menu-like system whether they want to receive results from their genomic analysis, he said.